Synthesis and biological evaluation of new citrate-based siderophores as potential probes for the mechanism of iron uptake in mycobacteria

Article abstract of DOI:10.1021/jm0104522

Several iron chelators containing α,β-unsaturated hydroxamic acid motifs appended to a citric acid platform were synthesized. Mycobacterium paratuberculosis was then challenged to grow in the presence of a panel of siderophores (mycobactin J, deferrioxami

Full text of DOI:10.1021/jm0104522

2056
J . Med. Chem. 2002, 45, 2056-2063
Syn th esis a n d Biologica l Eva lu a tion of New Citr a te-Ba sed Sid er op h or es a s
P oten tia l P r obes for th e Mech a n ism of Ir on Up ta k e in Mycoba cter ia
Hongyan Guo,† Saleh A. Naser,‡ George Ghobrial,‡ and Otto Phanstiel IV*,†
Center for Discovery of Drugs and Diagnostics, Department of Chemistry, and Department of Molecular and Microbiology,
University of Central Florida, Orlando, Florida 32816-2366
Received September 26, 2001
Several iron chelators containing R,â-unsaturated hydroxamic acid motifs appended to a citric
acid platform were synthesized. Mycobacterium paratuberculosis was then challenged to grow
in the presence of a panel of siderophores (mycobactin J , deferrioxamine B, acinetoferrin, and
nannochelin A) and the new synthetic agents. Of the structures tested, those containing the
trans 2-octenoyl motif were preferred over those with trans cinnamoyl groups. In addition,
derivatives containing longer tether lengths between the iron binding ligands (C5) were more
efficacious and led to higher growth index values. Perhaps most remarkable was the finding
that at 2.4 µM a trans 2-octenoylated, citrate-containing imide 6 was nearly 5-fold more effective
in stimulating growth than the native chelator, mycobactin J . In this regard, new structural
elements were identified (e.g., an imide motif or 2-octenoyl side chain), whose presence
stimulated mycobacterial growth.
In tr od u ction
The World Health Organization (WHO) estimates
that there were 8.4 million new cases of tuberculosis
(TB) in 1999 and that the annual global rate of increase
in TB incidence was 3%.1 Assuming this rate of increase
F igu r e 1. Mycolic acid architecture.
persists, there will be over 10 million new cases of TB
Consequently, penetration of antibiotics and chemo-
therapeutic agents through both the resulting tubercle
and bacterial cell wall is limited.5,6 The design and
synthesis of an efficient drug carrier, which facilitates
drug delivery, has become a major challenge in anti-
TB drug development.7,8 Therefore, understanding my-
cobacterial transport processes may offer new strategies
to combat this microorganism.
in 2005.1 Moreover, Mycobacterium tuberculosis, the
organism that causes the disease, has proven to be very
resilient. The recent emergence of multidrug resistant
(MDR) strains is of special concern due to the reported
high death rates associated with MDR TB.2,3 In addi-
tion, the current treatment of MDR strains is 100 times
more expensive than with drug-susceptible strains.2 In
short, tuberculosis continues to be a public health
threat, to the extent that the WHO has declared
tuberculosis as a global public health emergency.1
Infection of a host by M. tuberculosis results from the
inhalation of droplet nuclei containing the organism.
Colonization of the organism takes place in the alveolar
sacs in the lungs, where it encounters macrophages of
the immune system. Macrophages that are attracted to
the area by phagocytic cells, T-cells, and other immune
cells fuse to form giant cells in the vicinity of the
damaged tissue containing the bacteria.4 These mac-
rophages as well as other immune cells wall-off the
growing culture with a thick fibrin coat.4 This is the
tubercle or lesion that is associated with the disease.
Like most mycobacteria, cells of M. tuberculosis are
covered by a lipid-rich cell wall comprised mainly of a
large amount of mycolic acids containing long hydro-
carbon chains (Figure 1). The length and high degree
of saturation of the chains produce an exceptionally
tightly packed array with extremely low fluidity.5
Virulence and even survival of most microbes during
infections depend on their ability to compete effectively
for key nutrients such as iron. However, iron exists
mainly as insoluble ferric hydroxide polymers in the
biosphere and is sequestered in humans by high mo-
lecular weight proteins such as transferrin (Fe trans-
port) and ferritin (Fe storage). Microorganisms have
evolved ferric ion (Fe3+) specific chelating agents, i.e.,
siderophores,9 which can solubilize and transport iron
to the microbe. These low molecular weight ligands have
been shown to rapidly remove iron from human proteins
such as transferrin.10 Then, by modification, reduction,
or siderophore decomposition, iron is released for use
by the cell.10 Most mycobacteria produce two types of
siderophores: exochelins, which are secreted extracel-
lularly, and mycobactin, which is cell wall associated.11a
As shown in Figure 2 [mycobactin J 1, deferrioxamine
B (DFO, shown as its mesylate salt) 2 from Streptomyces
pilosus, nannochelin A 3 from Nannocystis exedens, a
hydrophilic exochelin found in M. tuberculosis cultures
(carboxymycobactin) 4,10 acinetoferrin 5 from Acineto-
bacter haemolyticus, and acinetoferrin imide 6], the
overall structures of siderophores vary depending upon
their microbial source. The iron-binding functional
groups incorporated into siderophores by most microbes
* To whom correspndence should be addressed. Tel.: 407-823-5410.
Fax: 407-823-2252. E-mail: ophansti@mail.ucf.edu.
† Center for Discovery of Drugs and Diagnostics, Department of
Chemistry, University of Central Florida.
‡ Department of Molecular and Microbiology, University of Central
Florida.
10.1021/jm0104522 CCC: $22.00 © 2002 American Chemical Society
Published on Web 04/04/2002

Citrate-Based Siderophores as Probes in Mycobacteria
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 10 2057
generation of conjugates.5b Second, the iron transport
mechanism involves an “iron-handoff” between two
siderophore families, the exochelins and the mycobac-
tins. In low iron environments, M. tuberculosis biosyn-
thesizes and secretes hydrophilic exochelins (e.g., 4) to
bind exogenous ferric ion. The iron-laden complex is
then transferred to intracellular siderophores, i.e., the
mycobactins, which are lipophilic chelators associated
with the cytoplasmic membrane.10 The mycobactin-
associated iron either remains in the cell wall as an iron
storage pool or is released into the cell by a mycobactin
reductase.10
This paper describes the synthesis and biological
evaluation of several new citrate-based siderophores as
potential probes to study the iron uptake mechanisms
of mycobacteria.12 A panel of siderophores with low,
moderate, and high growth stimulation effects would
be useful in comparing the iron uptake processes of
genetically modified mycobacterial strains with their
native counterparts. The focus of this investigation was
to develop a structure-activity relationship between
citrate-based iron chelators and their effect on the
growth rate of Mycobacterium paratuberculosis. M.
paratuberculosis was chosen as an alternative patho-
genic model for M. tuberculosis and is responsible for
J ohne’s disease (an inflammatory bowel disease in
cattle) and Crohn’s disease in humans.11 Because M.
paratuberculosis relies upon the addition of exogenous
mycobactin J (Figure 2) to grow in vitro, it serves as a
sensitive model for iron acquisition studies.13 Beyond
direct insight into possible new Crohn’s disease probes,
the findings gained with this organism should also apply
to its virulent cousin, M. tuberculosis. The expectation
was that iron-binding ligands, which were readily
utilized by M. paratuberculosis, could eventually serve
as mechanistic probes for both M. paratuberculosis and
M. tuberculosis.
Previously, two citrate derivatives were synthesized
in our laboratory, acinetoferrin 5 (from A. haemolyticus
ATCC 17906) and acinetoferrin imide 6 (Figure 2).14
Preliminary screens indicated that imide 6 was superior
to mycobactin J in “cross-feeding” M. paratuberculosis.
Armed with this insight, we synthesized a series of
citrate derivatives to identify which structural features
of 6 were responsible for its remarkable activity.
F igu r e 2. Structures of iron chelators.
have been conserved and are limited primarily to
hydroxamic acids, catechols, and R-hydroxy-carboxylic
acids.9 By design, the hexadentate architectures con-
taining these oxygen-rich, bidentate subunits have a
high affinity for binding ferric ion.9
Resu lts a n d Discu ssion
Because iron is essential for the growth and survival
of M. tuberculosis, several approaches have been sug-
gested for the development of antibiotics based on
siderophores and their analogues.7,8,10 One approach is
the development of conjugates (containing a lethal drug
covalently attached to a siderophore), which may lead
to new forms of drug delivery that utilize the pathogenic
organism’s own iron transport system. In this regard,
the siderophore represents a “tunable” vector for the
delivery of organism specific antibiotics.10 Another ap-
proach is to sequester the available iron into a form,
which cannot be processed by the bacteria. The success
of this latter approach relies on understanding the
molecular recognition events involved in mycobacterial
iron transport.
As shown in Figure 2, mycobactin J 13 and DFO (the
native chelator for Streptomyces pilosus)15 are predi-
cated upon R-amino acid and polyamide backbones,
respectively. The hydroxamic acid (RCON(OH)R′) and
the R-hydroxy carboxylic acid components are respon-
sible for iron chelation.9,16 Thus, each siderophore
represents a hexacoordinate ligand for iron(III). Both
acinetoferrin 5 and its imide 6 contain the unusual
trans-2-octenoyl hydroxamic acids appended to a citric
acid framework. Indeed, the ability of acinetoferrin
imide 6 to cross-feed mycobacterial strains may stem
from its appended long hydrophobic alkyl tail.5b,17 To
probe this insight, we used a modular synthetic ap-
proach,14,16,18 which allowed for alteration of the termi-
nal hydrophobic “tails” and the aliphatic spacer sepa-
rating the bidentate ligands.
Targeting the iron transport processes of M. tuber-
culosis is challenging for several reasons. First, the
complexity of the mycobactin architecture itself poses
a daunting synthetic challenge, which hampers the
Syn th esis. For comparison purposes, acinetoferrin,14
acinetoferrin tert-butyl ester 7,14 and nannochelin A 318

2058 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 10
Guo et al.
Sch em e 1a
Sch em e 2a
a
Reagents: (a) Concentrated H2SO4, MeOH; (b) concentrated
perchloric acid, tert-butyl acetate, room temperature, 3 days; (c) 2
M aqueous NaOH in MeOH, 3 h; (d) N-hydroxysuccinimide, DIC,
dry THF.
mono BOC-protected amines (12 and 13) in 82 and 75%
yield, respectively. Compounds 12 and 13 were each
dissolved in a biphasic mixture containing carbonate
buffer, pH 10.5, and BPO dissolved in CH2Cl2. After the
oxidation step, each mixture was then acylated with
trans-cinnamoyl chloride to give the desired compound
14 (55%) and 15 (52%), respectively. These represent
significant increases in yield as compared to earlier
efforts with nannochelin A, 3 (25% yield).18 Compounds
14 and 15 were then treated with a 10% concentrated
NH4OH in MeOH solution at 0 °C to liberate the “free”
hydroxamic acid 16 (81%) and 17 (94%), respectively.
Treatment of 16 and 17 with trifluoroacetic acid (TFA)
at 0 °C removed the BOC group and produced the
respective TFA salts 18 and 19. As shown in Scheme 1,
condensation of 18 and 19 with 20 in 1,4-dioxane gave
21 (90%) and 22 (85%), respectively. Finally, treatment
of 21 and 22 with TFA gave the respective chelators 8
(89%) and 9 (84%).
a
Reagents: (a) Di-tert-butyl dicarbonate, 10% TEA/MeOH; (b)
benzoyl peroxide (2 equiv)/CH2Cl2, pH ) 10.5 aqueous sodium
carbonate buffer, room temperature; (c) trans-cinnamoyl chloride,
CH2Cl2; (d) 10% NH4OH/MeOH; (e) TFA, 0 °C; (f) TEA, 20, dry
dioxane, 0 °C.
were also acquired by published procedures. To test the
side chain premise, new derivatives with appended
trans cinnamoyl groups were envisioned. The trans
cinnamoyl group represented another hydrophobic group
(in lieu of the trans 2-octenoyl tail) with the same carbon
count but with reduced conformational freedom. As
shown in Scheme 1, two carbon tethers (n ) 1 and 2)
were incorporated into the new derivatives 8 and 9,
respectively.
The synthesis of the new derivatives involved a series
of protection and deprotection steps similar to those
outlined in our total synthesis of acinetoferrin.14 In
1990, Milewska et al. illustrated the formation of an
O-benzoyl-protected hydroxamate using benzoyl perox-
ide (BPO) followed by acetyl chloride.19 This approach
led to an improved amine-oxidation method involving
biphasic conditions, which was used successfully in this
and other reports.14,20-22 The strategy involved the
construction of a N-(alkylamino)-N-hydroxyl-trans-cin-
namamide precursor via a tandem oxidation-acylation
of the primary amino group of a protected diamine
derivative. This elaboration was completed prior to
coupling with the terminal carboxylic acid groups of a
citric acid derivative. Previous experience revealed that
higher yields were obtained in the condensation step,
if the benzoyl group was removed prior to amide
formation.14,18
While the selective coupling of amino substrates with
the terminal carboxyls of citric acid can be achieved
utilizing either 2-substituted-1,3-bis-activated esters of
citric acid or anhydromethylene citric acid,16,23 the
reported imide formation associated with the latter
approach prompted the use of the former coupling
methodology. The selective activation of the 1,3-car-
boxyls of citric acid was accomplished via the terminally
bis-activated ester 2-tert-butyl-1,3-di-N-hydroxy-succin-
imidyl citrate (bis-NHS ester, 20).
As shown in Scheme 2, 20 was prepared using a
modification of Milewska’s original method utilizing
intermediates 23-25.16 N,N′-Diisopropylcarbodiimide
was used as the condensation reagent instead of N,N′-
dicyclohexylcarbodiimide in the last step to facilitate
isolation of the diamide.12,24
As shown in Scheme 1, the reaction of either diamine
(10 and 11) with di-tert-butyl dicarbonate afforded the

Citrate-Based Siderophores as Probes in Mycobacteria
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 10 2059
Ta ble 1. GI for M. Paratuberculosis in the Presence of
Iron-Binding Ligands at 2.4 µMa,b
further illustrates the influence of tether and other
structural features on the GI value. The tether trend is
likely reflected in Table 2 with nannochelin A 3, which
contains a C5 tether and a relative GI of 2.5. We
speculate that the longer tether allows for a more
conformationally flexible ligand to properly coordinate
to iron and provides an incremental increase in hydro-
phobicity. Such tether effects have been observed with
other siderophore systems such as DFO 2. In fact,
molecular modeling studies with DFO suggest that
tethers, which are too short (<C5), may compromise the
optimum binding geometry around iron by perturbing
the octahedral complex.9b
no. (compd)
reading
GIa
1 (mycobactin J )
2 (DFO)
5 (acinetoferrin)
7 (acinetoferrin tert-butyl ester)
8 (C3 cinnamoyl)
601
552
607
412
458
924
539
1
0.92
1
0.69
0.76
1.5
9 (C4 cinnamoyl)
21 (C3 cinnamoyl tert-butyl ester)
0.90
a
GI values were normalized to Mycobactin J growth activity.
The sterile broth media contained ferric ammonium citrate (0.04
b
g/L media).
Ta ble 2. GI for M. Paratuberculosis in the Presence of
Iron-Binding Ligands at 2.4 µMa,b
A priori, one may have expected that the more
hydrophobic esters would be uniformly more efficacious
than their carboxylic acid counterparts. This was not
the case in the limited number of systems examined.
Replacement of the free carboxylic acids in 5 and 8 with
tert-butyl esters (7 and 21, respectively) gave similar
values and no clear trend. In hindsight, this result is
not surprising as the ester does not preclude the
hexadenticity of 7 and 21 and other structural factors
may dominate such as tether length. The observation
that bulky tert-butyl esters did not dramatically alter
growth provides a possible site for future modifications
of the acyclic scaffold.
no. (compd)
reading
GIa
1 (mycobactin J )
2 (DFO)
3 (nannochelin A)
6 (acinetoferrin imide)
152
146
379
703
1
0.96
2.5
4.6
a
GI values were normalized to Mycobactin J growth activity.
The sterile broth media contained ferric ammonium citrate (0.04
b
g/L media).
Biologica l Eva lu a tion . All derivatives tested were
capable of forming hexadentate-binding architectures.
Note: the tert-butyl ester in 21 could form an overall
hexadentate ligand via its two hydroxamic acid motifs
and the available R-hydroxycarbonyl bidentate site. The
respective siderophores and synthetic derivatives (1-
3, 5-9, and 21) were evaluated for their ability to act
as growth factors for M. paratuberculosis at 2.4 µM (a
standard concentration used routinely with 1 in M.
paratuberculosis cultures).13,25 Several controls were run
in tandem. Mycobactin J (the native chelator, 1) pro-
vided the benchmark control. In this manner, systems
that provided significantly higher growth index (GI)
values than the native chelator 1 (GI ) 1) were
identified as superior growth stimulants and more
efficacious iron delivery agents (vide infra). In addition,
DFO 2 probed the effect of simply providing solubilized
ferric ion as the determining factor in stimulating
growth. If no molecular recognition events were in-
volved, all of the ligands tested would have provided
approximately the same growth index. As shown in
Tables 1 and 2, this was clearly not the case.
Comparison of the trans 2-octenoyl side chain present
in 5 with the trans cinnamoyl analogue present in 8
reveals a preference for the octenoyl subunit (GI for 5,
1.0, vs 8, 0.76). We speculate that the conformational
freedom of the octenoyl side chain may enable more
favorable membrane interactions with a cell-surface
receptor or via simple chain entanglements. Interest-
ingly, the 2-octenoyl subunit may play other roles in
mycobacteria. First, it resembles a precursor to the
mycolic acid constituents of the mycobacterial cell wall
(Figure 1) and the R1 side chain of mycobactin J (Figure
2). Second, in studies of isoniazid and ethionamide
resistant M. tuberculosis strains, a single missense
mutation in the inhA gene confers resistance to both
drugs. Recently, the corresponding InhA protein was
shown to be a NADH specific enoyl reductase, involved
in the synthesis of fatty acids in mycobacteria.26 Inter-
estingly, 2-trans-octenoyl ACP and 2-octenoyl-CoA were
also substrates for the InhA protein.26
One can compare the tabular data (Tables 1 and 2)
as all siderophores were dosed at the same concentra-
tion (2.4 µM) in the same media with the same cell type
and uniformly cultured under the same environmental
conditions (see Experimental Section). Although the
biological evaluations listed in Table 2 were performed
separately, they were performed in the presence of the
appropriate controls for proper comparison. The diverse
growth rates observed with these systems required that
measurements of “fast” growing cultures (e.g., with 6)
be conducted separately as they generated the maxi-
mum reading (999) from the BACTEC instrument at
earlier times (see Experimental Section).
As shown in Table 1, mycobactin J 1, DFO 2, acineto-
ferrin 5, and C3 ester 21 gave comparable results, while
acinetoferrin ester 7 and C3-cinnamoyl derivative 8 had
marginally lower activity. Interestingly, the C4 cin-
namoyl analogue 9 had a higher GI than 1. In general,
the systems containing longer tethers gave higher GI
values (e.g., GI for C3 8, 0.76, vs C4 9, 1.5). Table 2
Perhaps the most exciting finding from this study was
the nearly 5-fold growth increase of M. paratuberculosis
in the presence of acinetoferrin imide 6 (GI ) 4.6) vs
the control platform (mycobactin J , GI ) 1, Table 2).
Compound 6 represented another structural refinement
(e.g., imide formation) that influenced ligand efficacy
(i.e., GI value). Comparison of acyclic 5 (C3 tether, GI
) 1) and imide 6 (C3 tether, GI ) 4.6) revealed the
significance of the cyclic architecture. The fact that both
native chelators (1 and 4) as well as 6 each contain a
cyclic bidentate ligand within their architectures may
be more than a curious coincidence and deserves further
study.
The trends identified in this investigation point to
future alterations in the chelator architecture, which
should result in an even more efficacious ligand. In this
manner, future ligands could be optimized by including
the most preferred structural elements of the series
studied. Although 6 contained the preferred octenoyl

2060 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 10
Guo et al.
used in conjunction with the BACTEC brand 460 TB Analyzer.
The Middlebrook 7H9 broth base media consists of 4 mL of
broth mixture included in a sealed bottle. The bottles, there-
fore, contain 7H9 and other ingredients such as ammonium
sulfate, glutamic acid, sodium citrate, pyridoxine, biotin,
disodium phosphate, ferric ammonium citrate (0.04 g/L media),
magnesium sulfate, calcium chloride, zinc sulfate, and copper
sulfate. The 7H9 broth was also supplemented with 100 mL/L
OADC (oleic acid, bovine albumin, dextrose, and catalase
enrichment, from the Becton Dickinson Company) and 0.05%
Tween 80. Tween 80 is a detergent added to prevent clumping
(Sigma-Aldrich Chemical Co.). Mycobactin J (Allied Monitor)
at a final concentration of 2.4 µM was added to the M.
paratuberculosis strain ATCC 43015 culture as a positive
control. When the bacterium was challenged with other
siderophores, mycobactin J was not added to the culture media.
Siderophores were added to the BACTEC 7H12 B+ bottled
liquid media using Becton Dickinson B-D “1 cm3” sterile
syringes to a final concentration of 2.4 µM. To each of these
BACTEC bottles, 100 µL of BACTEC 7H12 B+ cultured MAP
(with a growth index of 100-300) was inoculated using Becton
Dickinson B-D 1 cm3 sterile syringes. Microbial growth activity
in this culture medium is indicated by the release of 14CO2
into the atmosphere of the sealed vial following the metabolism
of 14C-labeled palmitic acid by the microorganism.
subunits, it also contains the shortest tether (C3). A
logical extension of these studies is to change the carbon
tether to a five methylene bridge (C5), while maintain-
ing the imide framework. Ironically, the anhydrometh-
ylene synthetic pathway16,23 we originally sought to
avoid14 may, in hindsight, offer access to these novel
imide platforms. Although the imide blocks attachment
to the central citrate core via an ester linkage (e.g., 7)
and thereby limits other possible structural modifica-
tions, its efficacy warrants future investigation.
Con clu sion s
An efficient, modular approach to the construction of
R,â-unsaturated hydroxamic acids14 was applied to the
construction of several citrate derivatives. The biological
studies suggest that the iron-uptake events in M.
paratuberculosis involve molecular recognition events,
which (i) prefer trans 2-octenoyl groups over cinnamoyl
side chains (e.g., 5 vs 8) and (ii) favor an imide motif
over an acyclic architecture (e.g., 6 vs 5). In summary,
the structure-activity relationships developed in this
paper have identified new synthetically accessible
ligands, which are structurally less complex than my-
cobactin J and have different affinities for the iron-
uptake apparatus of mycobacteria. Such ligands, which
offer regulation of the initial iron delivery step, provide
the opportunity to compare the iron transport mecha-
nisms of both native and genetically modified mycobac-
teria.25
The atmosphere of the sealed BACTEC 12B vial is aspirated
from the vial via a sterile needle attached to a robotic arm
inside the BACTEC 460 TB Analyzer System. The BACTEC
Analyzer operates by initially drawing room air through a dust
14
filter, a flush valve, and an ion chamber transferring all
-
CO2 into a CO2 trap, where it is retained. This process cleans
the electrometer and leaves it ready to start the next cycle.
During the next cycle, a pair of 18G needles are heated. A
pump produces a partial vacuum in the ion chamber used to
lower the testing needles through the rubber septum of the
vial being tested. A vacuum draws culture gas from the vial
to the ion chamber. The electrometer measures the very small
current that the radioactive 14CO2 produces in the ion chamber.
Following removal of the radioactive culture gas, fresh 5% CO2
is introduced into the medium headspace every time a vial is
tested, enhancing the growth of mycobacteria. The current
measured by the electrometer is amplified and displayed as a
numeric reading. The reading is measured on a scale of 0-999
and is an indication of microbial growth activity in the bottle.
Usually, a reading of 10 or higher is an indication of definite
microbial growth. The growth indices (GI values) listed in
Tables 1 and 2 were calculated by dividing the instrument
reading obtained for the chelator by that obtained for the
mycobactin J control. As is often the circumstance when
looking at relative microbiological growth rates over time, the
error in making these measurements can be appreciable (up
to (85 reading units). Slow-growing cultures may bias the
results as their early growth is more time sensitive, whereas
faster-growing cultures may give disproportionately higher
values at early sample times. In this regard, one looks for
significant differences in growth rates and overall trends.
Exp er im en ta l Section
Ma ter ia ls a n d Meth od s. Lipophilic Sephadex LH-20 was
obtained from the Sigma Chemical Company. Silica gel (32-
63 µm) was purchased from Scientific Adsorbents, Inc. Trans-
2-octenoic acid (97.1% pure) was purchased from Lancaster
Synthesis, Inc. Reagents were purchased from the ACROS
Chemical Company and used without further purification. All
solutions are expressed in volume percent. Ammonia-contain-
ing solutions were prepared by measuring out concentrated
aqueous NH4OH in the listed volume percent. The pH 10.5
carbonate buffer solution was prepared by combining 222 mL
of 0.75 N aqueous NaHCO3 and 78 mL of 1.5 N aqueous NaOH.
“Iron-free” glassware was obtained by soaking the glassware
in a 6 N HCl bath overnight and then washing with deionized
water. Deionized water was obtained from a “B-pure” filtration
system by collecting the water when the “in-line” electrical
resistance was 17.0 MΩ cm. Iron-free silica gel was made by
washing with methanol:acetone:10 M HCl (45:45:10), followed
by 10% (w/w) Na2CO3 solution and then rinsed with deionized
1
water until pH 7 and air-dried. H and 13C nuclear magnetic
resonance (NMR) spectra were recorded at 200 MHz on a
Varian XL-200 spectrometer. The mass spectra were per-
formed at the University of Florida Department of Chemistry
N1,N 5(N 3′-Hyd r oxyl,N 3′-cin n a m oyl-3′a m in op r op yl)1,5-
citr ic Dia m id e 8. TFA (4 mL) was added dropwise to 21 (80
mg, 0.123 mmol) at 0 °C. After the addition was complete, the
solution was warmed to room temperature. TLC (15% MeOH/
CHCl3) showed that the starting material was consumed after
2 h. The volatiles were removed under high vacuum without
heating. The residue was subjected to a LH-20 column (pre-
swelled and eluted with 3% EtOH/toluene) to give 8 (65 mg,
89%). Rf ) 0.31 in 15% MeOH/CHCl3; Rf ) 0.27 in 15%
by Dr. David Powell. Mycobactin
J 1 and DFO 2 were
purchased from Allied Monitor Company in Fayette, MO, and
the Sigma-Aldrich Chemical Co., respectively. Nannochelin A
3,18 acinetoferrin 5,14 acinetoferrin imide 6,14 and acinetoferrin
tert-butyl ester 714 were acquired from published methods.
Beyond the “clean” NMR spectra obtained for both 8 and 9,
their purities were assessed by eluting each substance in two
different thin-layer chromatography (TLC) systems, which
confirmed the homogeneity of the sample. The procedures used
to make 20 and 23-25 are also included to illustrate the
changes made to Milewska’s procedures.16
Bioa ssa y. BACTEC 12B Mycobacteria also known as
Middlebrook 7H12 Medium was purchased from Becton Dick-
inson (Pittsburgh, PA). It contains 7H9 broth base, casein
hydrolysate, bovine serum albumin, catalase, and palmitic acid
labeled with 14C. It is specific for growing mycobacteria and is
1
2-propanol/ethyl acetate. H NMR (CD3OD): δ 7.64-7.23 (m,
14H, aromatic and olefinic), 3.77 (t, 4H, CH2NO), 3.22 (t, 4H,
CH2N), 2.71 (dd, 4H, CH2), 1.86 (m, 4H, CH2); 13C NMR (CD3-
OD) δ 177.2, 172.3, 168.1, 143.6, 130.2, 129.8, 128.8, 117.3,
75.2, 47.0, 44.8, 37.7, 27.7. High-resolution mass spectrum:
theory for C30H37N4O9 (M + 1), 597.2561; found M + 1,
597.2534.
5
N1,N (N 4′-H yd r oxy,N 4′-cin n a m oyl-4′a m in ob u t yl)1,5-
citr ic Dia m id e 9. TFA (2 mL) was added dropwise to 22 (41

Citrate-Based Siderophores as Probes in Mycobacteria
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 10 2061
mg, 0.060 mmol) at 0 °C. The mixture was stirred at room
temperature. TLC showed the starting material was consumed
after 4 h (Rf ) 0.56 in 15% MeOH/CHCl3). The volatiles were
removed under high vacuum without heating. The residue was
subjected to a Sephadex LH-20 column (preswelled and eluted
with 4% EtOH/toluene) to give 9 (32 mg, 84% yield). Rf ) 0.38
in 15% MeOH/CHCl3; Rf ) 0.35 in 15% 2-propanol/ethyl
acetate. 1H NMR (CD3OD): δ 7.64-7.26 (m, 14H, aromatic
and olefinic), 3.75 (t, 4H, CH2NO), 3.20 (t, 4H, CH2N), 2.70 (q,
4H, CH2), 1.71 (m, 4H, CH2), 1.56 (m, 4H, CH2). 13C NMR (CD3-
OD): δ 177.2, 172.0, 168.2, 143.9, 136.3, 131.6, 130.2, 129.0,
117.7, 75.3, 45.2, 40.1, 27.7, 24.8. High-resolution mass
spectrum: theory for C32H41N4O9 (M + 1), 625.2874; found
M+1, 625.2893.
separated and the water layer was extracted twice with CH2-
Cl2. The organic layers were combined, dried over anhydrous
Na2SO4, filtered, and concentrated. The residue was subjected
to flash column chromatography, eluting with 20% ethyl
acetate/hexane to give 15 as a white solid (1.39 g, 52%). Rf )
0.41 in 20% ethyl acetate/hexane. 1H NMR (CDCl3): δ 8.17
(d, 2H, aromatic), 7.81-7.28 (m, 9H, aromatic and olefinic),
6.68 (d, 1H, olefinic), 4.60 (br s, 1H, NH), 3.98 (t, 2H, CH2-
NO), 3.18 (m, 2H, CH2N), 1.70 (m, 4H, CH2), 1.40 (s, 9H, CH3).
13C NMR (CDCl3): δ 166.7, 164.4, 156.1, 145.6, 134.8, 134.7,
130.2, 129.0, 128.8, 128.0, 126.8, 114.6, 79.0, 48.6, 40.2, 28.7,
27.6, 24.2.
N3-Hyd r oxyl-N3-cin n a m oyl-N1-BOC-p r op a n e Dia m in e
16. A solution of 10% NH4OH/MeOH (25 mL) was added
dropwise to 14 (0.9483 g, 2.23 mmol) in a pretreated iron-free
round-bottom flask at 0 °C. The mixture was stirred overnight
under N2 atmosphere. The disappearance of the starting
material was monitored by TLC (Rf ) 0.61, 50% ethyl acetate/
hexane). The mixture was concentrated, and the product was
recrystallized using ethyl acetate and hexane (1:5 volume
ratio) to give a white solid 16 (0.60 g, 81%). Rf ) 0.23 (50%
3-(ter t-Bu toxyca r bon yl a m in o)p r op yla m in e 12. 1,3-
Diaminopropane (10) (11.16 g, 150 mmol) was dissolved in 350
mL of a 10% triethylamine (TEA)/MeOH solution. A solution
of di-tert-butyl dicarbonate (10.9 g, 50 mmol) and MeOH (20
mL) was added to this mixture with vigorous stirring. The
solution was refluxed for 2 h and then concentrated under
reduced pressure to give an oil. The oil was subjected to flash
column chromatography, eluting with 4% NH4OH/MeOH to
give the mono-BOC amine 12 (7.16 g, 82%). Rf ) 0.24 in 4%
1
ethyl acetate/hexane); mp 92.0-93.0 °C. H NMR (CDCl3): δ
9.35 (br s, 1 H, OH), 7.10-7.70 (m, 7H, aromatic and olefinic),
5.21 (br s, 1H, NH), 3.81 (t, 2 H, CH2NO), 3.19 (q, 2 H, CH2N),
1.86 (m, 2H, CH2), 1.40 (s, 9 H, CH3). 13C NMR (CDCl3): δ
167.0, 157.1, 142.7, 135.2, 129.8, 128.8, 128.1, 116.4, 80.0, 46.7,
37.3, 28.2, 27.3.
1
NH4OH/MeOH. H NMR (CDCl3): δ 4.95 (br s, 1H, NH), 3.20
(m, 2H, CH2NC), 2.72 (t, 2H, CH2N), 1.62 (m, 2H, CH2), 1.44
(s, 9H, CH3).
4-(ter t-Bu toxyca r bon yla m in o)bu tyla m in e 13. 1,4-Di-
aminobutane 11 (13.2 g, 150 mmol) was dissolved in 350 mL
of 10% TEA/MeOH solution. A solution of di-tert-butyl dicar-
bonate (10.9 g, 50 mmol) and MeOH (20 mL) was added, and
the solution was refluxed with vigorous stirring for 2 h. The
solution was concentrated and subjected to flash column
chromatography, eluting with 2% NH4OH/MeOH to give the
mono-BOC amine 13 (6.96 g, 75%). Rf ) 0.27 (2% NH4OH/
4
4
N -Hyd r oxyl-N -cin n a m oyl-N1-BOC-bu ta n e Dia m in e
17. In a 50 mL pretreated iron-free round-bottom flask, 10%
NH4OH/CH3OH (20 mL) was added dropwise to 15 (0.85 g,
1.94 mmol) at 0 °C under an Ar atmosphere. The disappear-
ance of the starting material was monitored by TLC (Rf ) 0.79,
70% ethyl acetate/hexane). The mixture was concentrated, and
the product was recrystallized with methanol and hexane (1:4
volume ratio) to give 17 as a white solid (0.61 g, 94%). Rf )
0.56 (70% ethyl acetate/hexane); mp 140.0-142.0 °C. 1H NMR
(CD3OD): δ 7.65-7.23 (m, 7H, aromatic and olefinic), 3.72 (t,
2H, CH2NO), 3.10 (t, 2H, CH2N), 1.68 (m, 2H, CH2), 1.4 (m,
11 H, CH2 and tert-butyl). 13C NMR (CD3OD): δ 168.2, 158.3,
143.8, 136.3, 130.3, 129.8, 128.5, 117.7, 80.0, 48.5, 40.4, 28.6,
28.2, 25.1.
1
MeOH). H NMR (CDCl3): δ 4.95 (b s, 1H, NH), 3.13 (m, 2H,
CH2NCO), 2.72 (t, 2H, CH2N), 1.80 (m, 2H, CH2), 1.60-1.30
(m, 11H, CH2 and tert-butyl). 13C NMR (CDCl3): δ 157.8, 80.8,
43.3, 41.8, 32.3, 29.8, 29.1.
N3-Cin n a m oyl-N3-b en zoyloxy-N1-(ter t-b u t oxyca r b on -
yl)p r op a n e Dia m in e 14. A solution of benzoyl peroxide (2.42
g, 10 mmol) and 50 mL CH2Cl2 was added dropwise at room
temperature to a vigorous stirred mixture of amine 12 (0.87
g, 5 mmol, dissolved in 25 mL CH2Cl2) and 75 mL of a
carbonate buffer solution (pH 10.5). The starting material was
consumed after stirring overnight as shown by TLC (4% NH4-
OH/MeOH). trans-Cinnamoyl chloride (0.83 g, 5 mmol) was
added dropwise at room temperature. After 30 min, the
disappearance of the intermediate (benzoyloxyamine) was
monitored by TLC (Rf ) 0.43 in 40% ethyl acetate/hexane).
After the acylation was complete, the organic layer was
separated and the water layer was extracted twice with CH2-
Cl2. The organic layers were combined, dried over anhydrous
Na2SO4, filtered, and concentrated. The residue was subjected
to flash column chromatography, eluting with 40% ethyl
acetate/hexane to give 14 as a white solid (1.16 g, 55%). Rf )
0.51 in 40% ethyl acetate/hexane. 1H NMR (CDCl3): δ 8.17
(d, 2H, aromatic), 7.82-7.27 (m, 9H, aromatic and olefinic),
6.66 (d, 1H, olefinic), 5.20 (br s, 1H, NH), 4.00 (t, 2H, CH2-
NO), 3.28 (q, 2H, CH2N), 1.84 (m, 2H, CH2), 1.40 (s, 9H, CH3).
13C NMR (CDCl3): δ 167.7, 164.6, 156.1, 145.0, 134.4, 134.3,
130.1, 129.0, 128.7, 128.2, 126.4, 114.6, 79.2, 46.2, 37.2, 28.3,
27.7.
N3-Hydr oxyl-N3-cin n am oyl-pr opan e Diam in e Tr iflu or o-
a cetic Acid Sa lt 18. Trifluoroacetic acid (TFA, 10 mL) was
added dropwise over 2 min to 16 (44 mg, 0.137 mmol) at 0 °C.
The ice bath was removed, and the solution was stirred for 30
min. TLC (40% ethyl acetate/hexane) showed that the reaction
was complete. The volatiles were removed under reduced
pressure to give the desired salt 18 as a crude oil, which was
consumed in the next step. 1H NMR (CD3OD): δ 7.89-7.24
(m, 7H, aromatic and olefinic), 3.88 (t, 2H, CH2NO), 2.98 (t,
2H, CH2N), 2.07 (m, 2H, CH2). 13C NMR (CD3OD): δ 169.0,
144.2, 136.2, 131.7, 129.8, 128.6, 116.7, 46.2, 38.2, 26.1.
4
4
N -Hyd r oxyl-N -cin n a m oyl-N1-BOC-bu ta n e Dia m in e
Tr iflu or oa cetic Acid Sa lt 19. Using a similar process, TFA
(15 mL) was combined with hydroxamic acid 17 (0.49 g, 1.46
mmol) at 0 °C and stirred for 2 h. TLC (60% ethyl acetate/
hexane) showed that the reaction was complete. The volatiles
were removed under reduced pressure to give the desired salt
19 as a crude oil, which was consumed in the next step. 1H
NMR (CD3OD): δ 7.66-7.28 (m, 7H, aromatic and olefinic),
3.81 (t, 2H, CH2NO), 2.97 (t, 2H, CH2N), 1.79 (m, 4H, CH2).
13C NMR (CD3OD): δ 168.3, 144.0, 136.2, 131.6, 130.0, 129.1,
117.2, 48.3, 40.1, 25.7, 24.4.
4
4
N -Cin n a m oyl-N -ben zoyloxy-N1-(ter t-bu toxyca r bon -
yl)bu ta n e Dia m in e 15. A solution of benzoyl peroxide (2.96
g, 12.2 mmol) and 50 mL of CH2Cl2 was added dropwise at
room temperature to a vigorous stirred mixture of amine 13
(1.15 g, 6.1 mmol, dissolved in 50 mL CH2Cl2) and 50 mL of a
carbonate buffer solution (pH 10.5). The starting material was
consumed after stirring overnight as shown by TLC (Rf ) 0.27
in 2% NH4OH/MeOH). trans-Cinnamoyl chloride (1.02 g, 6.1
mmol) was added dropwise at room temperature. After 40 min,
the disappearance of the intermediate benzoyloxyamine was
monitored by TLC (Rf ) 0.27 in 33% ethyl acetate/hexane).
After the acylation was complete, the organic layer was
2-ter t-Bu tyl-1,3-d i-N-(h yd r oxyl) Su ccin im id yl Citr a te
(NHS Ester ) 20. 3-tert-Butyl citrate 25 (4.60 g, 18.5 mmol)
and N-hydroxy succinimide (4.27 g, 37 mmol) were dissolved
in 100 mL of dry THF. A N,N′-diisopropylcarbodiimide (DIC)
solution (3.34 g, 29 mmol dissolved in 100 mL dry THF) was
added dropwise to the solution at room temperature. The
mixture was stirred for 2 days. A white precipitate formed.
The 1H NMR spectrum of the mixture showed that the reaction
was complete (i.e., the multiplet signal at 2.80 ppm disap-
peared). The mixture was filtered, and the white precipitate
was kept. 1H NMR showed the precipitate contained the

2062 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 10
Guo et al.
product and diisopropyl urea (doublet at 1.1 ppm). Hot CHCl3
was added to the crude solid, and part of the precipitate
dissolved. The mixture was filtered, and the solid was collected
and air-dried overnight to give the NHS ester 20 as a white
in a refrigerator until a white precipitate formed (possibly a
trimethyl citrate impurity). The precipitate was filtered and
discarded. The filtrate was concentrated to give 3-tert-butyl-
1,5-dimethyl citrate 24 as a light yellow oil (18.7 g, 75%). H
NMR (CDCl3): δ 3.70 (s, 6H, CH3), 2.84 (m, 4H, CH2), 1.52 (s,
9H, CH3).
1
1
solid (1.64 g, 20%). H NMR (DMSO-d6): δ 3.33 (d, 2H, CH2),
3.21 (d, 2H, CH2), 2.80 (s, 8H, CH2), 1.40 (s, 9H, CH3). 13C NMR
(DMSO-d6): δ 170.0, 165.0, 81.0, 72.1, 38.9, 27.1, 24.7.
3-ter t-Bu tyl Citr a te 25. 3-tert-Butyl-1,5-dimethyl citrate
24 (16 g, 58 mmol) was dissolved in MeOH (58 mL). A chilled
2 M NaOH aqueous solution (58 mL) was added to the solution
at 0 °C. The mixture was warmed to room temperature and
stirred for 3 h. Two layers formed. The organic layer was
discarded. The aqueous layer was acidified to pH 2 using 2 N
HCl. The aqueous layer was extracted with ethyl acetate three
times. The ethyl acetate layers were combined and concen-
trated down to give 3-tert-butyl citrate 25 as a white solid 7.89
5
N1,N (N 3′-Hyd r oxy,N 3′-cin n a m oyl-3′-a m in op r op yl)-3-
ter t-bu toxyca r bon yl-1,5-citr ic Dia m id e 21. In a pretreated
iron-free round-bottom flask, TEA (0.7 g, 6.94 mmol) was
added to a mixture of 2-tert-butyl-1,3-di-N-succinimidyl citrate
(NHS ester 20) (0.1317 g, 0.298 mmol) and TFA amine salt
18 (0.2106 g, 0.63 mmol) dissolved in 50 mL of dry dioxane at
0 °C. The solution was allowed to warm to room temperature
and stirred overnight. The solvent was removed under reduced
pressure, and the residue was dissolved in CHCl3 (20 mL) and
washed with 0.5 N HCl (20 mL). The aqueous layer was
extracted twice with CHCl3. The organic layers were combined
and concentrated. The residue was subjected to a Sephadex
LH-20 column (preswelled and eluted with 4% EtOH/toluene)
to give 21 as a solid (0.175 g, 90%). Rf ) 0.26 in 10% MeOH/
CHCl3. 1H NMR (CDCl3): δ 9.70 (br s, 1H, OH), 7.63-7.24
(m, 14H, aromatic and olefinic), 3.78 (t, 4H, CH2NO), 3.24 (q,
4H, CH2N), 2.64 (dd, 4H, CH2), 1.84 (m, 4H, CH2), 1.41 (s, 9H,
CH3). 13C NMR (CDCl3): δ 172.9, 170.3, 167.1, 142.9, 134.9,
129.9, 128.8, 127.9, 116.5, 83.2, 73.9, 46.5, 44.2, 36.6, 27.8, 26.6.
High-resolution mass spectrum: theory for C34H45N4O9 (M +
1), 653.3187; found M + 1, 653.3117.
1
g (55%). H NMR (CD3OD): δ 2.80 (dd, 4H, CH2), 1.50 (s, 9H,
CH3).
Ack n ow led gm en t. This research was supported by
an award from the University of Central Florida De-
partment of Chemistry. The 200 MHz NMR spectrom-
eter was purchased by a grant from the National Science
Foundation (CHE 8608881). The mass spectra were
provided by Dr. David H. Powell at the Department of
Chemistry at the University of Florida.
N1,N 5(N 4′-Hydr oxy,N 4′-cin n am oyl-4′am in obu tyl)-3-ter t-
bu toxyca r bon yl-1,5-citr ic Dia m id e 22. TEA (5.11 g, 50
mmol) was added dropwise to a mixture of NHS ester 20 (0.74
g, 1.67 mmol) and 19 (1.17 g, 3.37 mmol) in 60 mL of dry
dioxane at 0 °C. The solution was allowed to warm to room
temperature and stirred overnight. The volatiles were removed
under reduced pressure. The residue was subjected to a
Sephadex LH-20 column (preswelled and eluted with 5%
EtOH/toluene) to give 22 as a solid (0.97 g, 85%). Rf ) 0.30 in
Refer en ces
(1) World Health Organization. Global Tuberculosis Control; WHO
Report 2001; World Health Organization: Geneva, Switzerland;
WHO/CDS/TB/2001.287
(2) World Health Organization Tuberculosis Fact Sheet, No. 104,
2000; see http://www.who.int/inf-fs/en/fact104.html.
(3) (a) Miller, M. J . Synthesis and Therapeutic Potential of Hydrox-
amic Acid Based Siderophores and Analogues. Chem. Rev. 1989,
89, 1563-1579. (b) Haeym, B., Honore´, N.; Truffot-Pernot, C.;
Banerjee, A.; Schurra, C.; J acobs, W. R., J r.; van Embden, J . D.
A.; Grosset, J . H.; Cole, S. T. Implications of Multidrug Resis-
tance for the future of Short Course Chemotherapy of Tuber-
culosis: A Molecular Study. Lancet 1994, 344, 293-298.
(4) Salyers, A. A.; Whitt, D. D. Bacterial Pathogenesis: A Molecular
Approach, 1st ed.; ASM Press: Washington, DC, 1994.
(5) (a) Rouhi, A. M. Tuberculosis: A Tough Adversary. Chem. Eng.
News 1999, May 17, 52-69. (b) Xu, Y.; Miller, M. J . Total
Syntheses of Mycobactin Analogues as Potent Antimycobacterial
Agents Using a Minimal Protecting Group Strategy. J . Org.
Chem. 1998, 63, 4314-4322.
1
10% MeOH/CHCl3. H NMR (CDCl3): δ 9.80 (br s, 1H, OH),
7.61-7.21 (m, 14H, aromatic and olefinic), 3.72 (t, 4H, CH2-
NO), 3.17 (q, 4H, CH2N), 2.68 (dd, 4H, CH2), 1.67-1.35 (m,
14H, tert-butyl and 2 CH2). 13C NMR (CDCl3): δ 172.1, 169.3,
166.1, 141.9, 133.9, 128.9, 127.9, 127.6, 115.7, 82.1, 73.6, 47.2,
42.6, 38.0, 26.4, 25.7, 22.6. High-resolution mass spectrum:
theory for C36H49N4O9 (M + 1), 681.3500; found M + 1,
681.3474.
1,5-Dim eth yl Citr a te 23. Concentrated H2SO4 (1 mL) was
added dropwise to a solution of citric acid (50 g, 0.26 mole)
and MeOH (250 mL) at room temperature. The mixture was
refluxed for 1 h. Lime water (75 g CaCO3 suspended in 1 L
water) was used to neutralize the solution to pH 7.0. A white
precipitate formed, which was filtered off and discarded. The
filtrate was concentrated down under vacuum to give a white
residue. Water (100 mL) was added to the residue. The mixture
was sonicated and filtered. The filtered solid was discarded,
and the filtrate was kept. Concentrated HCl was added to
bring the filtrate to pH 0. A white precipitate formed and was
filtered and kept. The precipitate was dissolved in 8% (w/w)
NaHCO3 solution (25 mL). The solution was washed three
times using CHCl3. Then, concentrated HCl was added again
to bring the solution to pH 0. Again, a white precipitate was
formed. This material was filtered and dried to give 1,5-
dimethyl citrate 23 as a white solid (20.0 g, 35%). 1H NMR
(CD3OD): δ 3.67 (s, 6H, CH3), 2.87 (dd, 4H, CH2).
ter t-Bu tyl-1,5-d im eth yl Citr a te 24. Concentrated per-
chloric acid (2.7 mL, 0.03 mole) was added to a solution
containing 1,5-dimethyl citrate 23 (20 g, 0.09 mole) and tert-
butyl acetate (120 mL, 0.89 mole). The solution was stirred at
room temperature for 3 days under N2. A saturated aqueous
NaHCO3 solution was used to bring the solution to pH 6. The
organic layer was separated. The aqueous layer was extracted
three times using CH2Cl2. The organic layers were combined
and concentrated down to give a yellow residue. The residue
was dissolved in hot hexane (400 mL). The solution was
allowed to cool slowly to room temperature and then placed
(6) (a) Connell, N. D.; Nikaido, H. Membrane Permeability and
Transport in Mycobacterium Tuberculosis. In Tuberculosis:
Pathogenesis, Protection, and Control; Bloom, B. R., Ed.; ASM
Press: Washington, DC, 1994; p 333. (b) Besra, G. S.; Chatterjee,
D. Lipid and Carbohydrates of Mycobacterium Tuberculosis. In
Tuberculosis: Pathogenesis, Protection, and Control; Bloom, B.
R., Ed.; ASM Press: Washington, DC, 1994; p 285.
(7) Malouin, F.; Miller, M. J . Microbial Iron Chelators as Drug
Delivery Agents: The Rational Design and Synthesis of Sidero-
phore-Drug Conjugates. Acc. Chem. Res. 1993, 26, 241-249.
(8) Lin, Y.; Helquist, P.; Miller, M. J . Synthesis and Biology
Evaluation of a Siderophore-Virginiamycin Conjugate. Synthesis
1999, No. SI, 1510-1514.
(9) (a) Bergeron, R. J .; Brittenham, G. M. The Development of Iron
Chelators for Clinical Use; CRC Press: Boca Raton, FL, 1994;
pp 279-284. (b) Bergeron, R. J .; Brittenham, G. M. The
Development of Iron Chelators for Clinical Use; CRC Press: Boca
Raton, FL, 1994; p 258.
(10) Roosenberg, J . M., II; Lin, Y.-M.; Lu, Y.; Miller, M. J . Studies
and Syntheses of Siderophores, Microbial Iron Chelators, and
Analogues as Potential Drug Delivery Agents. Curr. Med. Chem.
2000, 7, 159-197 and references therein.
(11) (a) Barclay, R.; Ratledge, C. Iron binding compounds of M.
avium, M. intracellularae, M. scrofulaceum, and mycobactin-
dependent M. paratuberculosis and M. avium. J . Bacteriol. 1983,
153, 1138-1146. (b) Chamberlin, W.; Graham, D. Y.; Hulten,
K.; El-Zimaity, H. M. T.; Schwartz, M. R.; Naser, S.; Shafran,
I.; El-Zaatari, F. A. K. Mycobacterium avium subsp. Paratuber-
culosis as one cause of Crohn’s disease. Aliment Pharmacol.
Ther. 2001, 15, 337-346. (c) Chiodini, R. J .; Van Kruiningen,
H. J .; Merkal, R. S. Ruminant paratuberculosis (J ohne’s Dis-
ease): The current status and future prospects. Cornell Vet.
1984, 74, 218-262.

Citrate-Based Siderophores as Probes in Mycobacteria
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 10 2063
(12) Guo, H. Synthesis of New Citrate-Based Siderophores As
Potential Drug Delivery Agents for Mycobacterium Paratuber-
culosis and the Reactions of N-(Benzoyloxy)hexylamine with
Ketones and Aldehydes. Masters Thesis, Department of Chem-
istry, University of Central Florida, 2000.
(18) Bergeron, R. J .; Phanstiel, O., IV. The Total Synthesis of
Nannochelin: A Novel Cinnamoyl Hydroxamate Containing
Siderophore. J . Org. Chem. 1992, 57, 7140-7143.
(19) Milewska, M. J .; Chimiak, A. An Alternative Synthesis of N5-
Acetyl-N5-hydroxy-L-ornithine from L-Ornithine. Synthesis 1990,
233-234.
(13) (a) McCullough, W. G.; Merkal, R. S. Structure of Mycobactin
J . Curr. Microbiol. 1982, 7, 337-341. (b) Mathews, P. R.;
McDiarimid, J . A.; Collins, P.; Brown, A. The Dependence of
some strains of M. avium on mycobactin for initial and subse-
quent growth. J . Med. Microbiol. 1979, 11, 53-57. (c) Lambrecht,
R. S.; Collins, M. T. Mycobacterium paratuberculosis. Factors
that influence mycobactin dependence. Diagn. Microbiol. Infect.
Dis. 1992, 15 (3), 239-246.
(20) Wang, Q. X,; King, J .; Phanstiel, O., IV. An Improved Synthesis
of O-Benzoyl Protected Hydroxamates. J . Org. Chem. 1997, 62,
8104-8108.
(21) Phanstiel, O., IV; Wang, Q. X.; Powell, D. H.; Ospina, M. P.;
Leeson, B. A. Synthesis of Secondary Amines via N-(Benzoyloxy)-
amines and Organoboranes. J . Org. Chem. 1999, 64, 803-806.
(22) Badescu, V. Synthesis and Thermal Study of
a Series of
(14) Phanstiel, O., IV; Wang, Q. Total Synthesis of Acinetoferrin. J .
Org. Chem. 1998, 63, 1491-1495.
N-Benzoyloxyamines and the Scope and Limitations of an Amine
Oxidation Process. Masters Thesis, Department of Chemistry,
University of Central Florida, 2000.
(15) Bergeron, R. J .; McManis, J . S.; Phanstiel, O., IV; Vinson, J . R.
T. Total Synthesis of DFO. J . Org. Chem. 1995, 60, 109-114.
(16) Milewska, M. J .; Chimiak, A. I.; Glowacki, Z. Synthesis of
Schizokinen, Homoschizokinen, Its Imide and the Detection of
Imide with 13C NMR Spectroscopy. J . Prakt. Chem. 1987, 329
(3), 447-456.
(23) Lee, B. H.; Miller, M. J . Microbial Iron Chelators: Total
Synthesis of Aerobactin and Its Constituent Amino Acid, N 6
-
Acetyl-N 6-hydroxylysine. J . Am. Chem. Soc. 1982, 104, 3096-
3101.
(24) Benoiton, N. L.; Chen, F. M. F. Not the Alkoxycarbonylamino
acid O-Acylisourea. Chem. Commun. 1981, 543-545.
(25) Naser, S. A.; Gillespie, R.; Naser, N. A.; El-Zaatari, F. A. K. Effect
of IS900 gene of Mycobacterium paratuberculosis on Mycobac-
terium smegmatis. Curr. Microbiol. 1998, 37 (6), 373-379.
(26) J ohnsson, K.; King, D. S.; Schultz, P. G. Studies on the
mechanism of Action of Isoniazid and Ethionamide in the
Chemotherapy of Tuberculosis. J . Am. Chem. Soc. 1995, 117,
5009-5010.
(17) (a) Maurer, P. J .; Miller, M. J . Total Synthesis of a Mycobactin:
Mycobactin S2. J . Am. Chem. Soc. 1983, 105, 240-245. (b) Hu,
J .; Miller, M. J . Total Synthesis of a Mycobactin S, a Siderophore
and Growth Promoter of Mycobacterium Smegmatis, and De-
termination of its Growth Inhibitory Activity against Mycobac-
terium tuberculosis. J . Am. Chem. Soc. 1997, 119, 3462-3468.
(c) Hu, J . Total Synthesis of Mycobactins, Siderophore-Drug
Conjugates and Biological Studies. Ph.D. Dissertation, Depart-
ment of Chemistry and Biochemistry, University of Notre Dame,
1996.
J M0104522