Chemical synthesis and biological evaluation of gallidermin-siderophore conjugates

Article abstract of DOI:10.1039/c0ob00846j

The lantibiotic gallidermin was modified at lysine residues by regioselective attachment of derivatives of pyochelin, agrobactin and desferrioxamine B with the objective of having siderophore receptors of Gram-negative bacteria transport the antibiotic-iron chelator conjugate through the outer membrane. All of the conjugates retained activity against the Gram-positive indicator strain, Lactococcus lactis subsp. cremoris HP. However, testing of the conjugates against several Gram-negative strains yielded unexpected results. Bacteria treated with 100 μM of the conjugates complexed with Fe³⁺ grew better than bacteria grown in iron-free media but worse than bacteria grown in the same media supplemented with 10 μM FeCl ₃. Although these findings indicate that the conjugates are unable to inhibit the growth of Gram-negative bacteria, they indicate penetration of the outer membrane and provide structure-activity information for design of other lantibiotic conjugates. The synthetic strategy is applicable for linking biomarkers or fluorescence probes to gallidermin for studies on its localization and mode of action. As there are many lantibiotics that operate with unknown mechanisms of action, this chemical approach provides a means to modify such peptides with biomarkers for biological investigations.

Full text of DOI:10.1039/c0ob00846j

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 2133
www.rsc.org/obc
PAPER
Chemical synthesis and biological evaluation of gallidermin-siderophore
conjugates
Sabesan Yoganathan, Clarissa S. Sit and John C. Vederas*
Received 6th October 2010, Accepted 13th December 2010
DOI: 10.1039/c0ob00846j
The lantibiotic gallidermin was modified at lysine residues by regioselective attachment of derivatives of
pyochelin, agrobactin and desferrioxamine B with the objective of having siderophore receptors of
Gram-negative bacteria transport the antibiotic-iron chelator conjugate through the outer membrane.
All of the conjugates retained activity against the Gram-positive indicator strain, Lactococcus lactis
subsp. cremoris HP. However, testing of the conjugates against several Gram-negative strains yielded
unexpected results. Bacteria treated with 100 mM of the conjugates complexed with Fe3+ grew better
than bacteria grown in iron-free media but worse than bacteria grown in the same media supplemented
with 10 mM FeCl3. Although these findings indicate that the conjugates are unable to inhibit the growth
of Gram-negative bacteria, they indicate penetration of the outer membrane and provide
structure–activity information for design of other lantibiotic conjugates. The synthetic strategy is
applicable for linking biomarkers or fluorescence probes to gallidermin for studies on its localization
and mode of action. As there are many lantibiotics that operate with unknown mechanisms of action,
this chemical approach provides a means to modify such peptides with biomarkers for biological
investigations.
Gram-negative bacteria by taking advantage of a class of outer
membrane receptors that recognize siderophores. Siderophores,
which are small molecule iron chelators produced by bacteria for
Introduction
Opportunistic pathogenic bacteria such as Pseudomonas aerug-
inosa cause numerous life threatening infections in humans ev-
acquiring iron, are actively transported into the cell via dedicated
ery year.1 Although there are many antibiotic treatments available,
some bacteria have found ways to develop resistance to almost
all clinically used drugs. This provides a great challenge to treat
infections caused by these resistant strains.1,2 In recent years,
lantibiotics have attracted much interest as a potential new class
of antibiotics for clinical applications and food preservation.3
Lantibiotics such as nisin and gallidermin (Fig. 1) are posttransla-
tionally modified antimicrobial peptides with potent antibacterial
activity against many pathogenic Gram-positive bacteria.4,5 More-
over, some lantibiotics exhibit a dual mode of action, by inhibiting
peptidoglycan biosynthesis and by forming cell membrane pores
that lead to leakage and cell death.4 Nisin has been used as a food
preservative in dairy products for over 50 years, and bacteria have
yet to develop significant resistance to nisin.3 Although lantibiotics
have great potential for clinical use, their inability to penetrate
the outer membranes of Gram-negative bacteria greatly limits
their use against such bacteria. As such, there is an increased
interest in identifying chemical and biological approaches to
overcome this limitation. Nature has already developed a strat-
egy to transport antibiotics through the outer membranes of
receptors.6
Albomycins,7 salmycins8 and microcin E492m9 are examples
of natural antibiotics that incorporate siderophore-mimicking
moieties within their strucures. The siderophore transporters
recogize these moieties and transport the entire molecule into the
cell, thus enabling the antibiotic to gain access to its intracellular
targets.9–12 Researchers have successfully employed a similar
strategy by linking small molecule antibiotics to siderophores;13–19
however, there is a considerable reduction in the antibacterial
activity of these conjugates. A possible explanation may be that
small molecule antibiotics are generally enzyme inhibitors whose
binding affinities for their targets could be adversely affected by
the covalent attachement of a large siderophore unit. Conversely,
lantibiotics are shown to bind to lipid II, an external target
found in the cell membrane.20 As well, attaching siderophore
moieties to lantibiotics may not have as serious consequences
with regards to their mechanism of action, since lantibiotics are
comparatively larger than siderophores. Moreover, lantibiotics
are many orders of magnitude more potent than small molecule
antibiotics against Gram-positive bacteria, so even if linking
siderophores to lantibiotics causes a small reduction in activity, the
conjugates could still be highly potent. Thus, we were interested
in pursuing the possibility of rendering lantibiotics active against
Gram-negative bacteria.
Department of Chemistry, Gunning/Lemieux Chemistry Centre, University
of Alberta, Edmonton, Alberta, T6G 2G2, Canada. E-mail: john.vederas@
ualberta.ca; Fax: +1 (780) 492-2134; Tel: +1 (780) 492-5475
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2133–2141 | 2133
©

View Article Online
Fig. 1 Simplified structure of (a) nisin and (b) gallidermin.
Our research group as well as others have reported that in-
creasing the permeability of the outer membrane sensitizes Gram-
negative bacteria towards lantibiotics.21–24 In particular, our recent
work has shown that when gallidermin and a number of other
bacteriocins from lactic acid bacteria are tested in combination
with ethylenediaminetetraacetic acid (EDTA), they exhibit activity
against Gram-negative bacteria. These findings indicate that if we
find a way to transport lantibiotics through the outer membrane,
they may be able to inhibit the growth of Gram-negative bacteria.
Encouraged by our results from the EDTA testing, we decided
to explore synthetic strategies to effectively link lantibiotics to
siderophore analogues, such that these antimicrobial peptides
could gain entry through the outer membrane. Our hypothesis
drew support from the fact that microcin E492m is an 84 amino
acid peptide with a siderophore moiety that is recognized by a
bacterial outer membrane receptor that is large enough to allow
the peptide’s entry into the cell.11,25,26 Therefore, we hoped that
the size of the lantibiotic-siderophore conjugate would not be a
limiting factor in its transport into the cell. More importantly,
the chemical methodology developed in our study will expand the
scope of modifications that can be made to lantibiotics for the pur-
pose of increasing their spectrum of activity and for studying their
mechanisms of action. For this study, three different siderophore
analogues, including pyochelin, agrobactin and desferrioxamine
B were selected. Pyochelin is secreted by Pseudomonas aeruginosa
and Burkholderia species.27,28 Agrobactin, a biscatechol containing
siderophore, and desferrioxamine B, a hydroxamate containing
siderophore, are utilized by either E. coli or Salmonella species.6,29,30
In undertaking the synthesis of gallidermin-siderophore conju-
gates, several difficulties must be addressed. Firstly, an effective
linking strategy must be identified. It is important to use a
suitable bifunctional linker that would allow us to regioselectively
link siderophores to gallidermin. Here we decided to pursue
1,5-difluoro-2,4-dinitrobenzene (DFDNB) and dimethyl squarate
as possible bifunctional linkers for this chemistry, since they
can be reacted in a stepwise fashion to link two different
molecules. Secondly, it is generally difficult to modify a natural
peptide regioselectively without the use of protecting groups.
Since gallidermin has no C-terminus and no side chain carboxyl
groups, the remaining options are to modify one of two available
lysine side chains or the N-terminus. Since modifications close
to the N-terminus tend to compromise the antibacterial activity
of lantibiotics, the best site for modification is the lysine at the
13th position. Finally, the reaction conditions used need to be
compatible with the presence of thioether bridges, dehydro amino
acid residues and the amino-vinylcysteine unit in gallidermin.
Aside from simply making use of the bifunctional linkers,
two of the siderophores, pyochelin and agrobactin, needed to
be structurally modified in order to react with DFDNB or
dimethyl squarate. We envisioned synthesizing analogues that
contain sidechains with either a hydroxyl or amine group for
the linking chemistry. A pyochelin analogue reported by Rivault
et al. was modified at the phenyl ring.28,31 Although the authors
provide no rationale for choosing to modify that specific position
with respect to binding of pyochelin to its receptor, examination
of the crystal structure of pyochelin bound to its receptor sheds
light on why the position was a suitable choice (Fig. 2).32 In the
crystal structure, the phenyl ring of pyochelin extends outward
from the receptor’s binding pocket, making it the most exposed
part of the molecule. This coupled with the fact that the phenyl
ring is not directly involved in coordinating Fe3+ makes it a logical
site to attach a side chain. Rivault et al.16 have also used the
pyochelin analogue to make a pyochelin-norfloxacin conjugate,
and have shown that it is taken in by a Pseudomonas species,
further supporting the hypothesis that this pyochelin analogue can
be recognized by its receptor. In a similar fashion, we designed an
agrobactin analogue with an amine side chain that extends from
Fig. 2 Co-crystal structure of pyochelin outer membrane receptor, FptA
from Pseudomonas aeruginosa (PDB ID : 1XKW)32 (a) full view and (b)
closer look at pyochelin binding site.
2134 | Org. Biomol. Chem., 2011, 9, 2133–2141
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
a site least likely to interfere with iron binding. Herein, we report
the attachment of these siderophore analogues to gallidermin,
and the biological evaluation of the resulting conjugates against
Gram-negative bacteria.
Results and discussion
The first step of our project was to synthesize both pyoche-
lin and agrobactin analogues. Pyochelin analogues were pre-
pared, using a modified literature procedure,28 as two sets
of interconvertable diastereomers (4¢R2¢¢R4¢¢R:4¢R2¢¢S4¢¢R:4¢S
1
2¢¢R4¢¢R:4¢S2¢¢S4¢¢R/1 : 1 : 3 : 2 by H NMR) (Scheme 1). These
pyochelin analogues contain either an amine or a hydroxyl
functionality for use in our linking chemistry.
Scheme 2 Reagents and conditions: (i) SOCl2, MeOH, reflux, 85% (ii)
BnBr, K2CO3, NaI, DMF, 63% (iii) NaOH(aq)/THF, 93%; (iv) CDI,
spermidine, CH2Cl2, 59% (v) PyBOP, HOBt, DIPEA, CH2Cl2, 91%.
Scheme 1 Reagents and conditions: (i) HBF4·Et2O, NIS, MeCN, -30 ◦
C
(ii) alkyne derivative, DIPEA, CuI, Pd(PPh3)4, DMF (iii) (R)-cysteine,
PO4-buffer/THF, 60 ◦C (iv) HN(OMe)Me, EDCI, HOBt, DIPEA, CH2Cl2
(v) LiAlH4, THF, -40 ◦C (vi) (R)-N-methylcysteine, KOAc, EtOH/H2O.
Fig. 3 Growth promotion of P. aeruginosa by the siderophore derivatives
In addition, an agrobactin analogue with an amine-containing
side chain was synthesized using the strategy shown in Scheme 2.
The compound 11 belongs to a class of bis-catechol containing
synthetic siderophores and is known as spermexatol, due to the
presence of a spermidine unit.30,33
(A: pyochelin analogue; B: agrobactin analogue; C: desferrioxamine B).
The third siderophore used in our investigation, the
hydroxamate-based desferrioxamine B, already has an amine tail
suitable for linking chemistry. Since it was commercially available,
it was used directly in reactions with the bifunctional linkers.
Before proceeding with the linking chemistry, we confirmed
that the selected siderophore analogues are recognized by outer
membrane receptors. Both growth curve studies (Fig. 3)34 and
fluorescence microscopy (Fig. 4)34 were used to evaluate the
ability of these siderophore analogues to enter the cell. For the
growth promotion studies and fluorescence evaluation, all of
the siderophores were precomplexed with Fe3+ to give a final
concentration of 10 mM and added to the test strains. All
bacteria were grown in M9 minimal media for both of these
studies. The results show that all three siderophores enter, and are
therefore utilized by, Gram-negative bacteria under iron deficient
conditions.
Fig. 4 Images of P. aeruginosa cells treated with fluorescein labeled
agrobactin analogue (a) fluorescence image and (b) overlay of fluorescence
and white-field images (scale bar, 10 mm).
the alcohol sidechain. Hence, we turned our focus to using
dimethyl squarate, which ultimately proved to be the best linker
for our studies. Controlling the pH of the reaction allows the
squarate to react with an amine-containing siderophore first, while
avoiding the formation of dimerized side products. Purification
of the squarate functionalized siderophore, and then subsequent
reaction with gallidermin affords the desired conjugates. As such,
dimethyl squarate can react in a stepwise fashion with two
different molecules that each contain an amine handle. Since this
linker reacts selectively with amino groups, the presence of other
reactive functional groups in the molecules is compatible with this
synthesis.
Following this, we went forward to explore an applicable
linking strategy. Several attempts were made to identify a suitable
bifunctional linker. In particular, DFDNB was thought to be
useful for attaching molecules together via their hydroxyl moieties.
In reality, DFDNB was found to be highly reactive but non-
selective, causing it to link onto the phenol in addition to
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2133–2141 | 2135
©

View Article Online
To employ the chosen linking strategy, siderophore analogue 11
was fully deprotected and then reacted with dimethyl squarate to
yield the monofunctional derivative 12 (Scheme 3).
positions. Once we successfully attached all three siderophore
conjugates to gallidermin regioselectively, the next step was to
evaluate the biological activity of these conjugates.
All of the conjugates retained at least 1/10 of gallidermin’s
potency in terms of antibacterial activity against the indicator
strain, Lactococcus lactis subsp. cremoris HP on a spot-on-lawn
assay. This suggests that modified gallidermin still behaves in a
similar fashion to the natural peptide. This result is encouraging
since our chemical strategy is the first successful example of
attachment of siderophore analogues to natural lantibiotics.
After completing the tests against the indicator strain, the same
spot-on-lawn overlay method was used to test conjugates 15–19
against several Gram-negative bacteria, including Pseudomonas
aeruginosa ATCC 14207, Burkholderia cepacia ATCC 25416,
E. coli DH5a and Salmonella enterica Typhimurium ATCC 13311.
All of the conjugates were complexed to Fe3+ prior to biological
evaluation. All of the conjugates showed no antibacterial activity
in both iron sufficient LB media and iron deficient M9 or succinic
media at 100 mM concentrations. To gain more insight into the
conjugates’ activity profiles, additional testing using a 64-well
plate assay in iron deficient M9 media was done. For this assay,
all of the compounds complexed to Fe3+ were tested at 100 mM
concentrations and the strains were grown in M9 minimal media
at 37 ◦C.
Scheme 3 Reagents and conditions: (i) TFA, CH2Cl2, 99% (ii) H2, Pd/C,
MeOH, 98% (iii) dimethyl squarate, MeOH, DIPEA, 86%.
In an analogous fashion, the other two siderophores were
reacted with dimethyl squarate to give the monofunctional
siderophore derivatives 13 and 14 (Fig. 5).
Fig. 7 shows the results of the well plate assay from the testing
of conjugate 18 against P. aeruginosa and B. cepacia (for other
growth curves, see Supporting Information). Conjugates 15–19
slightly promoted the growth of P. aeruginosa. Conjugates 15–18
showed a similar activity profile against B. cepacia, but compound
19 showed no effect. Both E. coli and Salmonella strains showed no
significant effect when treated with all five conjugates. From these
observations, we speculate that the siderophore-Fe(III) complexes
do enter the cell, however, there may be a very small local
concentration of the conjugates or the conjugates fail to reach
the inner membrane to show a complete killing effect. The early
growth promotion is possibly due to a small amount of ferric ion
being supplied by the conjugates. We believe that this indicates the
Fig. 5 Siderophore-monomethylsquarate esters.
These derivatives were then linked to gallidermin. The second
addition to the squarate is highly dependent on the pH of the
reaction mixture (pH~9). Therefore, the reaction was done in
borate buffer. Each of the siderophore derivatives (12, 13 and
14) was reacted with gallidermin in a mixture of methanol and
borate buffer (pH~9). We found that 13 successfully reacted with
gallidermin at 23 ◦C to give the desired monotagged product.
However, both 12 and 14 produced multiply tagged conjugates
at 23 ◦C. By lowering the reaction temperature to 4 ◦C, 12
could be used to produce monotagged conjugates (~1 : 1 mixture
of regioisomers) with gallidermin. On the other hand, when
the reaction with 14 was lowered to 4 ◦C, the reaction rates
were too slow to produce any conjugates. Ultimately, ditagged
conjugates were isolated from the reaction of 14 with gallidermin,
and these products were used for the antibacterial testing. After
extensive HPLC purification and MS/MS fragmentation analysis,
we identified that the two lysine residues were the most reactive,
giving rise to conjugates tagged at either (or both) of the two
Fig. 6 Gallidermin-siderophore conjugates.
Fig. 7 Biological evaluation of conjugate 18 against P. aeruginosa and
B. cepacia using a well plate assay.
2136 | Org. Biomol. Chem., 2011, 9, 2133–2141
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
conjugates are imported into the Gram-negative bacteria, but it
is possible that they provide a form of solubilized iron that can
be utilized by extracellular ferric ion exchange processes. Further
research is needed to explain the observed results. Although unex-
pected, this iron delivery and growth promotion are encouraging
in that they show evidence of the conjugates being recognized by
bacterial membrane receptors. This preliminary evaluation serves
as the ground work for investigating the ability of lantibiotics
to cross the outer membrane of Gram-negative bacteria. Further
investigations by attaching siderophores with better recognition
and iron binding affinity (e.g. enterobactin and salmochelins) to
gallidermin could provide a better conjugate. From this work, we
have developed an efficient synthetic strategy for attaching small
molecules to gallidermin. In turn, this approach can be extended
to synthesize other lantibiotic-siderophore conjugates.
HPLC purification methods
Method A. Purification was done using a C18 semi-preparative
column (Vydac 5 mm, 10 mm ¥ 250 mm). Flow rate: 3.0 mL/min.
Gradient: Starting from 5% CH3CN for 7 min and 1st ramp to
60% over 23 min, 2nd ramp to 90% CH3CN over 2 min, followed
by ramping down to 10% CH3CN over 1 min, then 10% CH3CN
for 4 min. Method B. Purification was done using a C18 semi-
preparative column (Vydac 5 mm, 10 mm ¥ 250 mm). Flow rate:
3.0 mL/min. Gradient: Starting from 10% CH3CN for 3 min and
1st ramp to 30% over 4 min, 2nd ramp to 90% CH3CN over 23 min,
followed by ramping down to 10% CH3CN over 1 min, then 10%
CH3CN for 4 min. Method C. Purification was done using a C8
preparative column (ZORBAX Rx 7 mm, 21.2 mm ¥ 250 mm).
Flow rate: 10.0 mL/min. Gradient: Starting from 10% CH3CN
for 3 min and 1st ramp to 40% over 27 min, 2nd ramp to 90%
CH3CN over 3 min and then 90% CH3CN for 5 min, followed
by ramping down to 10% CH3CN over 1 min, then 10% CH3CN
for 4 min. Method D. Purification was done using a C18 semi-
preparative column (Vydac 5 mm, 10 mm ¥ 250 mm). Flow rate:
3.0 mL/min. Gradient: Starting from 10% CH3CN for 3 min and
1st ramp to 30% over 4 min, 2nd ramp to 50% CH3CN over 26 min,
3rd ramp to 90% CH3CN over 2 mininutes, followed by ramping
down to 10% CH3CN over 1 min, then 10% CH3CN for 4 min.
Conclusions
An efficient synthetic approach was developed to attach
siderophores to gallidermin. The conjugates retained respectable
bioactivity against the indicator organism. However, these con-
jugates showed unexpected, slight growth promotion in ferric ion
deficient media. The antibacterial effect may be too low to observe
a complete inhibition of bacteria. This synthetic strategy can be
extended to modify other lantibiotics with various siderophores.
With the identification of numerous novel lantibiotics, there is
little known about the mechanism of action of these peptides in
literature. Therefore, it would be advantageous to use this chemical
methodology to modify such lantibiotics with biomarkers or
fluorescence probes to investigate their mechanism of action.
General procedure for the synthesis of functionalized
2-hydroxybenzonitriles (2) and (3)
A solution of 5-iodo-2-hydroxybenzonitrile (1.0 equiv.), alkyne
derivative (1.5 equiv.) and DIPEA (4.5 equiv.) in DMF
(6 mL mmol-1) was degassed for 20 min by bubbling argon through
the solution. The solution was cooled to 0 ◦C and was added with
Pd(PPh3)4 (0.05 equiv.) and CuI (0.1 equiv.). The reaction mixture
was again degassed for another 20 min and allowed to stir at 23 ◦C
for 10 h under argon atmosphere. Then the reaction mixture was
concentrated in vacuo and product was purified by silica column
chromatography (Et2O–hexanes, 1 : 2) to obtain the compounds 2
or 3.
Experimental
General
All reactions inlvolving anhydrous reaction conditions were done
in flame-dried glassware under an argon atmosphere. All commer-
cially available reagents were purchased and used without further
purification. All the solvents used for reactions were distilled
over appropriate drying reagents prior to use. Commercially
available ACS grade solvents (> 99.0%) were used for column
chromatography without any further purification. All reactions
and fractions from column chromatography were monitored by
thin layer chromatography (TLC) using glass plates coated with
silica gel 60 F254 (EMD Chemicals Inc.). Flash chromatography
was performed using 230–400 mesh silica gel (Silicycle). HPLC
purification and analysis were performed on a Varian ProStar
instrument, monitored using a dual wavelength detector (220 and
280 nm). For HPLC, water (Solvent A) and acetonitrile (Solvent
B), both containing 0.1% trifluoroacetic acid were used as eluent.
NMR spectra were recorded on an Inova 600, Inova 500, Unity
500, Inova 400 or Inova 300 spectrometer. For 1H (300, 400, 500
or 600 MHz) spectra, d values were referenced to CDCl3 (7.26
ppm), CD3OD (3.30 ppm) or (CD3)2CO (2.04 ppm). For 13C (100,
125 or 150 MHz) spectra, d values were referenced to CDCl3 (77.0
ppm), CD3OD (49.0 ppm) or (CD3)2CO (29.8 ppm) as the solvents.
Infrared spectra (IR) were recorded on a Nicolet Magna 750 or a
20SX FT-IR spectrometer. Mass spectra (MS) were recorded on a
Kratos AEIMS-50 mass spectrometer (high resolution, HRMS).
4-(3-Cyano-4-hydroxyphenyl)but-3-ynyl acetate (2). Isolated
as a yellow solid (2.6 g, 96%). nmax(cast)/cm-1 3400–3000 (br),
2962, 2233, 1740, 1708, 1607 and 1508; dH(300 MHz; CDCl3) 7.54
(d, 1H, J 2.1, H-6), 7.48 (dd, 1H, J 8.7 and 2.1, H-4), 6.91 (d, 1H,
J 8.7, H-3), 4.25 (t, 2H, J 6.9, H-10), 3.51 (s, 1H, Ar-OH), 2.73
(t, 2H, J 6.9, H-9) and 2.11 (s, 3H, CH3); dC(100 MHz; CDCl3)
171.8, 158.5, 137.7, 136.0, 116.7, 116.1, 115.7, 99.8, 85.6, 79.7,
62.5, 20.9 and 19.8; m/z(ES+) calcd for C13H11NO3Na 252.0631,
found 252.0629 [MNa+].
tert-Butyl-3-(3-cyano-4-hydroxyphenyl)prop-2-ynylcarbamate
(3). Isolated as a yellow oil (3.12 g, 85%). nmax(cast)/cm-1 3324–
3050 (br), 2979, 2934, 2231, 1681, 1606 and 1510; dH(400 MHz;
CD3OD) 7.53 (d, 1H, J 2.0, H-6), 7.45 (dd, 1H, J 2.1 and 8.7, H-
4), 6.88 (d, 1H, J 8.7, H-3), 4.00 (s, 2H, N-CH2) and 1.44 (s, 9H,
(CH3)3); dC(100 MHz; CD3OD) 161.5, 158.0, 138.7, 137.4, 117.4,
116.9, 115.9, 101.1, 81.1, 80.6, 79.4, 31.3 and 28.7; m/z(ES+) calcd
for C15H17N2O3 273.1239, found [MH+].
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2133–2141 | 2137
©

View Article Online
General procedure for the synthesis of Weinreb amides (4) and (5)
KOAc (5.0 equiv.) and water (8 mL mmol-1 of aldehyde). The
reaction mixture was stirred in the dark for 24 h. The reaction
mixture was diluted with water (150 mL) and washed with hexanes
(2 ¥ 100 mL). The aqueous layer was acidified with solid citric acid
to pH~4 and extracted with CH2Cl2 (3 ¥ 100 mL). The organic
layer was dried over Na2SO4, filtered, and concentrated in vacuo
to yield the pyochelin analogues 6 or 7 in pure form. The products
exist as a mixture of four diastereomers I : II : III : IV/1 : 1 : 3 : 2.
To a mixture of functionalized 2-hydroxybenzonitriles 2 or 3
(1.0 equiv.) and (R)-cysteine (2.0 equiv.) in THF (7 mL mmol-1)
was added PO4-buffer (7 mL mmol-1). The mixture was heated
under reflux at 60 ◦C for 2 days and then THF was removed
in vacuo. The suspention was diluted with water (300 mL),
acidified with citric acid (pH~2) and extracted with CH2Cl2 (3 ¥
100 mL). The combined organic layers were dried over Na2SO4
and concentrated in vacuo to yield the corresponding thiazolines
as a yellow solid. The thiazolines were used in the next step
without further purification. To a solution of thiazoline in CH2Cl2
(20 mL mmol-1) at 0 ◦C was added EDCI (1.2 equiv.), HOBt
(1.2 equiv.) and DIPEA (1.2 equiv.), immediately followed by a
solution of HN(OMe)Me∑HCl (1.5 equiv.) and DIPEA (1.5 equiv.)
in CH2Cl2 (20 mL mmol-1). The reaction mixture was allowed to
warm to 23 ◦C and after stirred for 2.5 h. The crude product
was purified uisng column chromatography (Et2O–hexanes, 1 : 1
to 2 : 1) to obtain the Weinreb amides 4 or 5.
Pyochelin analogue 6. Isolated as a yellow solid (0.291 g, 66%
over two steps). nmax(microscope)/cm-1 3322, 3055, 2954, 2230,
1704, 1618, 1572 and 1490; m/z(ES+) calcd for C18H21N2O4S2
393.0937, found 393.0940 [MH+]. Isomer I (4¢R,2¢¢R,4¢¢R):
dH(500 MHz; Acetone-d6) 7.40 (m, 2H, H-6, H-4), 6.91 (m, 1H,
H-3), 5.21 (td, 1H, J 5.5 and 9.0, H-4¢), 4.61 (d, 1H, J 5.5, H-2¢¢),
3.66-3.74 (m, 4H, H-4¢¢, H-10), 3.49-3.64 (m, 2H, H-5¢), 3.21-3.23
(m, 2H, H-5¢¢), 2.64 (s, 3H, N-CH3) and 2.58 (t, 2H, J 6.5, H-9);
dC(125 MHz; Acetone-d6) 172.9, 172.0, 159.5, 137.1, 134.1, 118.2,
117.0, 115.5, 87.2, 81.1, 80.1, 77.2, 73.3, 61.4, 41.6, 33.2, 33.0 and
24.3; Isomer II (4¢R,2¢¢S,4¢¢R): dH(500 MHz; Acetone-d6) 7.40 (m,
2H, H-6, H-4), 6.91 (m, 1H, H-3), 5.02 (q, 1H, J 8.4, H-4¢), 4.55
(d, 1H, J 8.2, H-2¢¢), 4.25 (app t, 1H, J 6.5, H-4¢¢), 3.66-3.74 (m,
3H, H-5¢, H-10), 3.42-3.46 (m, 2H, H-5¢), 3.21-3.23 (m, 2H, H-5¢¢),
2.58 (t, 2H, J 6.5, H-9) and 2.50 (s, 3H, N-CH3); dC(125 MHz;
Acetone-d6) 172.9, 172.0, 159.5, 137.1, 134.1, 118.2, 117.0, 115.5,
87.2, 81.1, 80.9, 77.8, 71.0, 61.4, 37.8, 35.1, 32.1 and 24.3; Isomer
III (4¢S,2¢¢R,4¢¢R): dH(500 MHz; Acetone-d6) 7.40 (m, 2H, H-6,
H-4), 6.91 (m, 1H, H-3), 4.84 (q, 1H, J 8.3, H-4¢¢), 4.34 (d, 1H, J
8.3, H-2¢), 4.02 (dd, 1H, J 6.9 and 5.3, H-4¢¢), 3.66-3.69 (m, 1H,
H-5¢), 3.53-3.64 (m, 1H, H-5¢¢), 3.42-3.46 (m, 1H, H-5¢), 3.23-3.30
(m, 1H, H-5¢), 2.63 (s, 3H, N-CH3) and 2.58 (t, 2H, J 6.9, H-9);
dC(125 MHz; Acetone-d6) 172.9, 172.0, 159.5, 137.1, 134.1, 118.2,
117.0, 115.5, 87.2, 83.3, 81.1, 79.3, 61.4, 44.8, 35.1, 33.6 and 24.3;
Isomer IV (4¢S,2¢¢S,4¢¢R): dH(500 MHz; Acetone-d6) 7.40 (m, 2H,
H-6, H-4), 6.91 (m, 1H, H-3), 5.28-5.34 (m, 1H, H-4¢), 5.05 (d,
1H, J 4.8, H-2¢¢), 4.25 (app t, 1H, J 6.5, H-4¢¢), 3.70-3.74 (m, 2H,
H-10),3.49-3.64 (m, 1H, H-5¢), 3.42-3.46 (m, 1H, H-5¢), 3.30-3.33
(m, 1H, H-5¢¢), 3.21-3.23 (m, 1H, H-5¢¢), 2.70 (s, 3H, N-CH3) and
2.58 (t, 2H, J 6.9, H-9); dC(125 MHz; Acetone-d6) 172.9, 172.0,
159.5, 137.1, 134.1, 118.2, 117.0, 115.5, 87.2, 81.1, 79.3, 73.5, 70.4,
61.4, 36.2, 32.4, 32.3 and 24.3.
4-(4-Hydroxy-3-(4-(methoxy(methyl)carbamoyl)-4,5-dihydro-
thiazol-2-yl)phenyl)but-3-ynyl acetate (4). Isolated as a off-white
solid (3.02 g, 79%). nmax(microscope)/cm-1 3300–2900 (br), 2940,
1739, 1672, and 1619; dH(400 MHz; CDCl3) 7.46 (d, 1H, J 2.1,
H-6), 7.36 (dd, 1H, J 8.8 and 2.1, H-4), 6.90 (d, 1H, J 8.8, H-3),
5.64 (t, 1H, J 8.7, H-4¢), 4.23 (t, 2H, J 6.9, H-10), 3.80 (s, 3H,
OCH3), 3.75 (t, 1H, J 9.4, H-5¢), 3.48 (dd, 1H, J 10.9 and 9.2,
H-5¢), 3.27 (s, 3H, N-CH3), 2.72 (t, 2H, J 6.9, H-9) and 2.08 (s,
3H, CH3); dC(125 MHz; CDCl3) 173.6, 170.8, 169.5, 158.8, 136.4,
134.0, 117.3, 116.0, 113.9, 84.1, 80.9, 74.6, 62.3, 61.7, 33.0, 32.5,
20.8 and 19.8; m/z(ES+) calcd for C18H21N2O5S 377.1166, found
377.1168 [MH+].
tert-Butyl-3-(4-hydroxy-3-(4-(methoxy(methyl) carbamoyl)-4,5-
dihydrothiazol-2-yl)phenyl)prop-2-ynylcarbamate (5). Isolated as
a yellow oil (1.30 g, 49%). nmax(microscope)/cm-1 3335, 2976, 2935,
2229, 1697, 1619, 1568 and 1488; dH(300 MHz; CDCl3) 7.49 (d,
1H, J 2.0, H-6), 7.38 (dd, 1H, J 2.0 and 8.6, H-4), 6.91 (d, 1H, J
8.6, H-3), 5.68 (t, 1H, J 8.6, H-4¢), 4.79 (s, 1H, NH), 4.12 (d, 2H, J
5.5, H-9), 3.81 (s, 3H, OCH3), 3.76 (t, 1H, J 8.0, H-5¢), 3.50 (m, 1H,
H-5¢), 3.28 (s, 3H, OCH3) and 1.46 (s, 9H, (CH3)3); dC(100 MHz;
CDCl3) 173.7, 159.2, 155.3, 136.6, 134.3, 117.5, 116.1, 113.4, 84.3,
82.2, 74.7, 61.8, 33.1, 32.6, 31.3 and 28.5; m/z(ES+) calcd for
C20H26N3O5S 420.1593, found 420.1588 [MH+].
Pyochelin analogue 7. Isolated as a yellow solid (0.950 g, 82%
over two steps). nmax(microscope)/cm-1 3374, 3053, 2979, 2934,
2230, 1709, 1619, 1572 and 1489; m/z(ES-) calcd for C22H26N3O5S2
476.1314, found 476.1319 [MH-]. Isomer I (4¢R,2¢¢R,4¢¢R):
dH(500 MHz; Acetone-d6) 7.42 (m, 2H, H-6, H-4) 6.93 (m, 1H,
H-3), 5.20 (m, 1H, H-4¢), 4.61 (d, 1H J 5.5, H-2¢¢), 4.08 (m, 2H,
H-9), 3.72 (m, 1H, H-4¢¢), 3.51 (m, 2H, H-5¢), 3.22 (m, 2H, H-
5¢¢), 2.64 (s, 3H, N-CH3) and 1.42 (s, 9H, Si(CH3)3); dC(125 MHz;
Acetone-d6) 172.9, 160.0, 156.2, 137.0, 134.2, 118.3, 117.0, 114.4,
86.4, 83.2, 81.5, 80.2, 77.1, 73.1, 41.6, 33.2, 32.6, 31.4 and 28.6;
Isomer II (4¢R,2¢¢S,4¢¢R): dH(500 MHz; Acetone-d6) 7.42 (m, 2H,
H-6, H-4) 6.93 (m, 1H, H-3), 5.02 (q, 1H, J 8.4, H-4¢), 4.55 (d,
1H, J 8.2, H-2¢¢), 4.08 (m, 2H, H-9), 3.70 (m, 1H, H-5¢), 3.52
(m, 1H, H-5¢), 3.22 (m, 2H, H-5¢¢), 2.50 (s, 3H, N-CH3) and 1.42
(s, 9H, Si(CH3)3); dC(125 MHz; Acetone-d6) 172.9, 160.0, 156.2,
137.0, 134.2, 118.3, 117.0, 114.4, 86.4, 83.2, 81.5, 77.8, 70.9, 37.8,
34.9, 32.1, 31.4 and 28.6; Isomer III (4¢S,2¢¢S,4¢¢R): dH(500 MHz;
Acetone-d6) 7.42 (m, 2H, H-6, H-4) 6.93 (m, 1H, H-3), 5.30 (m,
General procedure for the synthesis of pyochelin analogues (6) and
(7)
The Weinreb amides 4 or 5 (1.0 equiv.) were dissolved in dry
THF (16 mL mmol-1) and cooled to -40 ◦C. To the resulting
solution, LiAlH4 (1.0 M in THF) (2.0 equiv.) was added drop
wise. The reaction mixture was stirred at -20 ◦C for 1 h and
then quenched with MeOH (1 mL mmol-1 of LiAlH4), saturated
NH4Cl (3 mL mmol-1 of LiAlH4), and 5% H2SO4 (3 mL mmol-1
of LiAlH4). The reaction mixture was extracted with ethyl acetate
(3 ¥ 50 mL). The organic layer was dried over Na2SO4, filtered,
and concentrated in vacuo to yield the corresponding aldehyde,
which was used in the next step immediately without further
purification. To a solution of the aldehyde (1.0 equiv.) in ethanol
(20 mL mmol-1) was added (R)-N-methylcysteine (2.0 equiv.),
2138 | Org. Biomol. Chem., 2011, 9, 2133–2141
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
1H, H-4¢), 5.05 (d, 1H, J 4.7, H-2¢¢), 4.25 (m, 1H, H-4¢¢), 4.08 (m,
2H, H-9), 3.54 (m, 1H, H-5¢), 3.44 (m, 1H, H-5¢), 3.32 (m, 1H,
H-5¢¢), 3.22 (m, 1H, H-5¢¢), 2.70 (s, 3H, N-CH3) and 1.42 (s, 9H,
Si(CH3)3); dC(125 MHz; Acetone-d6) 172.9, 160.0, 156.2, 137.0,
134.2, 118.3, 117.0, 114.4, 86.4, 83.2, 81.5, 79.3, 73.3, 70.4, 36.2,
32.4, 32.3, 31.4 and 28.6; Isomer IV (4¢S,2¢¢R,4¢¢R): dH(500 MHz;
Acetone-d6) 7.42 (m, 2H, H-6, H-4) 6.93 (m, 1H, H-3), 4.84 (q,
1H, J 8.4, H-4¢), 4.34 (d, 1H, J 8.3, H-2¢¢), 4.08 (m, 2H, H-9), 4.02
(dd, 1H, J 6.9 and 5.3, H-4¢¢), 3.67 (m, 1H, H-5¢), 3.52 (m, 1H,
H-5¢¢), 3.44 (m, 1H, H-5¢¢), 3.31 (m, 1H, H-5¢), 2.63 (s, 3H, N-CH3)
and 1.42 (s, 9H, Si(CH3)3); dC(125 MHz; Acetone-d6) 172.9, 160.0,
156.2, 137.0, 134.2, 118.3, 117.0, 114.4, 86.4, 83.2, 81.5, 79.3, 73.8,
44.8, 35.1, 33.6, 31.4 and 28.6.
was then dried over Na2SO4, filtered, and concentrated in vacuo to
about 20 mL. Upon addition of hexanes the product precipitated
in its pure form (5.1 g, 63%) as a pale yellow solid. nmax(cast)/cm-1
3065, 3032, 2950, 2880, 1728, 1580, 1498 and 1474; dH(500 MHz;
CDCl3) 7.50 (m, 4H, Ar-H), 7.34-7.44 (m, 7H, Ar-H), 7.18 (dd, 1H,
J 1.7 and 8.1, Ar-H), 7.11 (t, 1H, J 8.2, Ar-H), 5.18 (s, 2H, OCH2),
5.16 (s, 2H, OCH2) and 3.89 (s, 3H, OCH3); dC(125 MHz; CDCl3)
166.7, 152.7, 148.2, 137.4, 128.5, 128.2, 127.9, 126.8, 123.9, 122.8,
117.9, 71.2 and 52.0; m/z(EI) calcd for C22H20O4 348.1362, found
348.1362 [M+].
2,3-Bis(benzyloxy)benzoic acid
A
solution of methyl-2,3-bis(benzyloxy)benzoate (2.48 g,
7.12 mmol) in THF (5 mL) was diluted with THF/H2O (1 : 1,
100 mL), to which was added 1.0 M KOH(aq) (15 mL). The
reaction was allowed to proceed at 23 C for 10 h. The reaction
4-(3-(tert-Butoxycarbonylamino)propylamino)-4-oxobutanoic acid
(8)
◦
A
solution of tert-butyl-3-aminopropylcarbamate (1.0 g,
mixture was then concentrated to remove volatile components
and acidified to pH ~ 2.0 with 1 M HCl(aq). The aqueous layer was
extracted with ethyl acetate (3 ¥ 50 mL) and the combined organic
layer was washed with water (30 mL) and brine (30 mL). The
organic layer was dried over Na2SO4, filtered and concentrated
in vacuo to yield the pure product (2.21 g, 93%) as a white solid.
5.74 mmol) and succinic anhydride (0.69 g, 6.89 mmol) in THF
(25 mL) was heated to reflux for 16 h. The reaction mixture
was then concentrated and redissolved in CH2Cl2 (100 mL).
The organic layer was then washed with water (50 mL) and
brine (50 mL), followed by evaporation of CH2Cl2 in vacuo
to yield 8 (1.2 g, 76%) as a white solid as pure compound.
n
max(microscope)/cm-1 2945, 2667, 2576, 1689, 1598, 1577 and
n
max(microscope)/cm-1 3500–2900 (br), 2978, 1716, 1659, 1535 and
1497; dH(400 MHz; CDCl3) 7.66 (dd, 1H, J 1.7 and 7.9, Ar-H),
7.22-7.46 (m, 10H, Ar-H), 7.18 (dd, J 1.8 and 8.1), 7.10 (t, 1H, J
7.9, Ar-H), 5.18 (s, 2H, CH2) and 5.12 (s, 2H, CH2); dC(100 MHz;
CDCl3) 165.3, 151.4, 147.2, 135.9, 134.7, 129.3, 128.9, 128.6,
127.8, 125.0, 124.5, 123.1, 119.1 and 71.6; m/z(ES+) calcd for
C21H18O4Na 357.1097, found 357.1098 [MNa+].
1440; dH(400 MHz; CDCl3) 9.45 (s, br, 1H, CO2H), 7.05 (s, br, 1H,
NH), 5.22 (s, br, 1H, NH), 3.24 (m, 2H, CH2), 3.10 (m, 2H, CH2),
2.64 (m, 2H, CH2), 2.49 (m, 2H, CH2), 1.59 (m, 2H, CH2) and 1.40
(s, 9H, C(CH3)3); dC(125 MHz; CDCl3) 175.8, 172.9, 156.8, 79.5,
37.3, 36.4, 30.8, 29.8, 28.4 and 28.4; m/z(EI) calcd for C12H22N2O5
274.1529, found 274.1362 [M+].
2,3-Bis(benzyloxy)-N-(3-(4-(2,3-bis(benzyloxy)benzamido)
Methyl 2,3-dihydroxybenzoate
butylamino)propyl)benzamide (10)
To a stirring solution of 2,3-dihydroxybenzoic acid (5.0 g,
32.4 mmol) in CH3OH (25 mL), SOCl2 (2.6 mL, 35.6 mmol) was
added drop wise at 23 ◦C. The reaction mixture was heated to
reflux for 10 h, after which it was concentrated and redissolved
in ethyl acetate (200 mL). The organic layer was washed with
saturated NaHCO3 (2 ¥ 100 mL), water (100 mL) and brine,
dried over Na2SO4 and filtered. The solvent was removed in vacuo
to yield methyl-2,3-dihydroxybenzoate (3.9 g, 85%) as a white
solid. nmax(microscope)/cm-1 3600–3150 (br), 1675, 1597 and 1428;
dH(500 MHz; CDCl3) 10.91 (s, 1H, CO2H), 7.35 (dd, 1H, J 8.1
and 1.5, Ar-H), 7.11 (dd, 1H, J 7.9 and 1.5, Ar-H), 6.78 (t, 1H, J
7.9, Ar-H), and 3.94 (s, 3H, OCH3); dC(125 MHz; CDCl3) 170.8,
148.8, 145.0, 120.6, 119.9, 119.2, 112.4 and 52.4; m/z(EI) calcd for
C8H8O4 168.0423, found 168.0422 [M+].
To a solution of 2,3-bis(benzyloxy)benzoic acid (1.59 g,
4.76 mmol) in CH2Cl2 (100 mL) was added 1,1¢-
carbonyldiimidazole (0.81 g, 5.0 mmol) and the reaction
mixture was allowed to stir at 23 ◦C for 90 min. Then, a solution
of spermidine (392 mL, 2.5 mmol) in CH2Cl2 (20 mL) was added
to the flask. After 14 h, the reaction mixture was concentrated,
redissolved in ethyl acetate (200 mL) and washed with 0.5 M
NaOH(aq) (2 ¥ 100 mL), water (100 mL) and brine (50 mL).
The organic phase was then dried over Na2SO4, filtered and
concentrated in vacuo to give a pale brown oil. The crude
product was purified by flash chromatography (5% to 10%
CH3OH in CH2Cl2) to obtain 10 (1.1 g, 59% yield) as a clear oil.
n
max(microscope)/cm-1 3384, 3065, 2935, 1653, 1576 and 1559;
dH(500 MHz; CDCl3) 8.07 (t, 1H, J 5.4, NH), 7.93 (t, 1H, J 5.5,
NH), 7.71 (m, 2H, Ar–H), 7.46 (m, 4H, Ar–H), 7.28–7.41 (m,
16H, Ar–H), 7.14 (m, 4H, Ar–H), 5.15 (s, 2H, CH2), 5.14 (s, 2H,
CH2), 5.07 (s, 2H, CH2), 5.06 (s, 2H, CH2), 3.34 (q, 2H, J 6.3,
CH2), 3.26 (q, 2H, J 6.1, CH2), 2.48 (t, 2H, J 6.9, CH2), 2.41
(t, 2H, J 6.7, CH2), 1.52 (quint., 2H, J 6.8, CH2) and 1.34 (m,
4H, CH2); dC(125 MHz; CDCl3) 165.2, 165.0, 151.75, 151.72,
146.76, 146.70, 136.49, 136.45, 128.76, 128.73, 128.67, 128.63,
128.24, 127.70, 127.67, 127.61, 124.42, 124.40, 123.3, 123.2,
116.85, 116.82, 77.3, 76.3, 71.33, 71.27, 49.4, 47.4, 39.5, 37.8, 29.5,
27.4 and 27.0; m/z(ES+) calcd for C49H52N3O6 778.3851, found
778.3841 [MH+].
Methyl 2,3-bis(benzyloxy)benzoate
In a flask, methyl-2,3-dihydroxybenzoate (3.90 g, 23.2 mmol),
K2CO3 (12.8 g, 92.8 mmol), and NaI (0.300 g) were dissolved
in DMF (30 mL). To the resulting mixture was added a solution
of benzyl bromide (6.06 mL, 51.0 mmol) in DMF (10 mL). The
reaction mixture was stirred at 23 ◦C for 18 h and then filtered to
remove insoluble materials. The filtrate was concentrated in vacuo
and the residue was dissolved in ether (100 mL). The organic layer
was washed with 2% NaOH(aq) (30 mL) and water (30 mL). It
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2133–2141 | 2139
©

View Article Online
tert-Butyl-3-(4-((4-(2,3-bis(benzyloxy)benzamido)butyl)(3-(2,3-
bis(benzyloxy)benzamido)propyl)amino)-4-
oxobutanamido)propylcarbamate (11)
159.5, 137.0, 134.4, 118.3, 117.0, 115.6, 87.2, 81.1, 80.1, 77.2, 73.3,
61.1, 44.5, 41.6, 33.2 and 33.0; m/z(ES-) calcd for C22H20N3O6S2
486.0799, found 486.0796 [MH-]; HPLC retention time (method
B): 15.5 min.
A solution of 10 (610 mg, 0.78 mmol), 8 (258 mg, 0.94 mmol) and
DIPEA (273 mL, 1.57 mmol) in CH2Cl2 (50 mL) was cooled in
ice-water bath. HOBt (159 mg, 1.18 mmol) and PyBOP (614 mg,
1.18 mmol) was then added to the solution. After stirring at 0 ◦C
for 20 min, the reaction mixture was warmed to 23 ◦C and stirred
for another 14 h. The reaction mixture was diluted with CH2Cl2
(100 mL) and the organic phase was washed with 5% NaHCO3
(2 ¥ 50 mL) and brine (50 mL). The organic layer was dried over
Na2SO4, filtered and concentrated to yield a yellow oil. The crude
product was purified by flash chromatography (2% to 5% CH3OH
in CH2Cl2) to obtain 11 (738 mg, 91% yield) as a yellow oil.
Desferrioxamine-squarate (14). Isolated as an off-white solid
(15 mg, 88%). dH(600 MHz, CD3OD) 4.35 (s, 3H, OCH3), 3.59
(m, 7H, NCH2- and N-OH), 3.38 (t, 2H, J 6.7, CH2), 3.16 (t, 4H,
J 7.0, CH2), 2.74 (m, 4H, CH2), 2.43 (t, 4H, J 7.2, CH2), 2.08
(s, 3H, CH3), 1.63 (m, 8H, CH2) and 1.54–1.31 (m, 12H, CH2);
dC(100 MHz; CD3OD) 189.9, 185.0, 178.0, 176.2, 174.9, 174.5,
173.5, 61.1, 52.2, 52.0, 45.2, 40.3, 39.8, 31.6, 31.5, 31.0, 30.3, 30.0,
28.9, 28.8, 27.3, 27.1, 24.9, 24.5, 24.3, 24.1 and 20.2; m/z(ES+)
calcd for C30H51N6O11 671.3610, found 671.3607 [MH+]; HPLC
retention time (method C): 24.5 min.
n
max(microscope)/cm-1 3384, 3065, 2935, 1653, 1576 and 1559;
dH(500 MHz; CDCl3) 8.07 (t, 1H, J 5.4, NH), 7.93 (t, 1H, J
5.5, NH), 7.71 (m, 2H, Ar–H), 7.46 (m, 4H, Ar–H), 7.28–7.41
(m, 16H, Ar–H), 7.14 (m, 4H, Ar–H), 5.15 (s, 2H, CH2), 5.14
(s, 2H, CH2), 5.07 (s, 2H, CH2), 5.06 (s, 2H, CH2), 3.34 (q, 2H,
J 6.3, CH2), 3.26 (q, 2H, J 6.1, CH2), 2.48 (t, 2H, J 6.9, CH2),
2.41 (t, 2H, J 6.7, CH2), 1.52 (quint., 2H, J 6.8, CH2) and 1.34
(m, 4H, CH2), dC(125 MHz; CDCl3) 172.7, 171.6, 165.2, 165.0,
156.2, 151.61, 151.55, 146.7, 146.5, 136.32, 136.26, 128.64, 128.59,
128.54, 128.41, 128.12, 127.53, 127.51, 127.50, 124.30, 124.28,
123.0, 122.8, 116.8, 116.6, 78.8, 77.3, 76.4, 71.13, 71.06, 47.2, 45.4,
43.1, 38.8, 36.9, 36.1, 31.4, 29.9, 28.3, 27.5, 26.6, 25.9 and 24.9;
m/z(ES+) calcd for C61H71N5O10Na 1056.5093, found 1056.5085
[MNa+].
General procedure for the reaction of gallidermin with
siderophore-squarate
To a solution of gallidermin (1.0 mM) in borate buffer (pH =
9.0) was added a solution of siderophore-squarate (1.5 equiv.)
◦
in THF. The reaction mixture was allowed to stir at either 4 C
(agrobactin conjugates) or 23 ◦C (pyochelin and desferrioxamine
conjugates) for 14 h and then diluted with water. The crude product
was purified by HPLC.
Gallidermin-pyochelin conjugates (15) and (16). Isolated as an
off-white residue (2.3 mg, 40%, two regioisomers). Conjugate 15:
HPLC retention time (method D): 19.8 min. Conjugate 16: HPLC
retention time (method D): 21.1 min.; m/z(MALDI-TOF) calcd
for C119H159N28O28S6 2619.0, found 2620.0 [MH+].
General procedure for the reaction of siderophore with dimethyl
squarate
Gallidermin-agrobactin conjugates (17) and (18). Isolated as an
off-white residue (3.1 mg, 58%, two regioisomers). Conjugate 17:
HPLC retention time (method A): 26.6 min. Conjugate 18: HPLC
retention time (method A): 27.0 min. m/z(MALDI-TOF) calcd
for C130H179N30O33S4 2815.2, found 2815.6 [MH+].
Et3N (2 equiv.) and dimethyl squarate (1.5 equiv.) was added to the
siderophore (1 equiv.) dissolved in CH3OH. The reaction mixture
was allowed to stir at 23 ◦C for 6 h and was then diluted with water
to get a 1 : 1 (v/v) mixture of CH3OH/H2O. The crude product
was purified by HPLC (eluent: CH3CN/H2O (with 0.1% TFA)).
The product exists as a mixture of a pair of tautomers.
Gallidermin-desferrioxamine B conjugate (19). Isolated as an
off-white residue (3.3 mg, 45%, two regioisomers). Conjugate 19:
HPLC retention time (method D): 16.9 min.; m/z(MALDI-TOF)
calcd for C156H234N37O43S4 3440.6, found 3440.9 [MH+].
Agrobactin analogue-squarate (12). Isolated as a pale brown
residue (20 mg, 86%). dH(500 MHz; CD3OD) 7.20 (m, 2H, Ar–
H), 6.91 (m, 2H, Ar–H), 6.69 (m, 2H, Ar–H), 4.32 and 4.31 (s,
3H, OCH3), 3.60–3.31 (m, 10H, CH2), 3.20 (m, 2H, CH2), 2.68
(m, 2H, CH2), 2.48 (m, 2H, CH2) and 1.98–1.56 (m, 8H, CH2);
dC(175 MHz; CD3OD) 190.0, 189.8, 185.0, 184.7, 178.5, 177.9,
174.5, 173.8, 171.7, 171.5, 171.4, 163.2, 163.0, 162.8, 150.5, 150.3,
147.4, 147.3, 118.6, 118.0, 116.7, 116.0, 61.1, 61.0, 46.7, 44.3, 43.5,
43.0, 42,4, 40.0, 39.8, 38.0, 37.5, 37.3, 36.9, 32.0, 31.8, 31.4, 31.2,
29.6, 29.3, 29.1, 28.5, 27.6, 26.9 and 26.0; m/z(ES+) calcd for
C33H41N5O11 683.2803, found 683.2801 [MH+]; HPLC retention
time (method A): 25.1 min.
Antibacterial assay
All test samples were dissolved in H2O:CH3CN (4 : 1). For the
spot-on-lawn overlay assay, the indicator strain, Lactococcus lactis
subsp. cremoris HP was grown in all purpose tween (APT) media
◦
at 25 C. The gram negative bacteria were grown in Luria broth
at 37 ◦C with shaking at 200 rpm. For testing under iron sufficient
conditions, Gram-negative bacterial cultures grown in Luria broth
overnight were used. For testing under iron deficient conditions,
Gram-negative bacterial cultures were first grown in LB broth
overnight, subcultured (2% innoculum) into either M9 or succinic
minimal media and grown to an OD600 ~ 0.1 at 37 ◦C, shaking at
200 rpm. M9 media (500 mL) was prepared as follows: 100 mL
Pyochelin analogue-squarate (13). Isolated as a pale yellow
residue (8 mg, 24%). (4¢R,2¢¢R,4¢¢R): dH(600 MHz; CD3COCD3)
7.46 (m, 2H, Ar–H), 6.95 (m, 1H, Ar–H), 5.23 (q, 1H, J 6.9, H4¢),
4.61 (d, 1H, J 5.4, H2¢¢), 4.38 (m, 5H, NCH2 and OCH3), 4.02
(dd, 1H, J 6.9, 5.3 Hz, H4¢¢), 3.67 (m, 1H, H5¢), 3.52 (m, 1H,
H5¢), 3.44 (m, 1H, H5¢¢), 3.31 (m, 1H, H5¢¢), 2.63 (s, 3H, N–CH3);
dC(125 MHz; CD3COCD3) 190.0, 184.7, 175.1, 174.1, 172.9, 172.0,
∑
of 5 ¥ M9 salts (0.25 M Na2HPO4 7H2O, 0.11 M KH2PO4 and
0.043 M NaCl) was diluted with 400 mL of milli-Q water and
sterilized. Then, filter sterilized solutions of 20% glucose (5 mL),
20% (NH4)2SO4 (1.25 mL), 1 M MgSO4 (1 mL), 0.1 M CaCl2
(0.5 mL) and 10 mg mL-1 thiamine (50 mL) were added to the
2140 | Org. Biomol. Chem., 2011, 9, 2133–2141
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
media. Succinic minimal media (500 mL) consists of K2HPO4
(3 g), KH2PO4 (1.5 g), (NH4)2SO4 (0.5 g), MgSO4 7H2O (0.1
References
∑
1 M. A. Fischbach and C. T. Walsh, Science, 2009, 325, 1089–1093.
2 M. E. Falagas, I. A. Bliziotis, S. K. Kasiakou, G. Samonis, P.
Athanassopoulou and A. Michalopoulos, BMC Infect. Dis., 2005, 5, 7.
3 P. D. Cotter, C. Hill and R. P. Ross, Curr. Protein Pept. Sci., 2005, 6,
61–75.
4 J. M. Willey and W. A. Van Der Donk, Annu. Rev. Microbiol., 2007, 61,
477–501.
5 R. Kellner, G. Jung, T. Horner, H. Zahner, N. Schnell, K. D. Entian
and F. Gotz, Eur. J. Biochem., 1988, 177, 53–59.
6 M. Miethke and M. A. Marahiel, Microbiol. Mol. Biol. Rev., 2007, 71,
413–451.
7 H. P. Fiedler, F. Walz, A. Dohle and H. Zahner, Appl. Microbiol.
Biotechnol., 1985, 21, 341–347.
8 L. Vertesy, W. Aretz, H. W. Fehlhaber and H. Kogler, Helv. Chim. Acta,
1995, 78, 46–60.
9 V. Delorenzo and A. P. Pugsley, Antimicrob. Agents Chemother., 1985,
27, 666–669.
10 V. Braun, A. Pramanik, T. Gwinner, M. Koberle and E. Bohn,
BioMetals, 2009, 22, 3–13.
11 X. Thomas, D. Destoumieux-Garzon, J. Peduzzi, C. Afonso, A. Blond,
N. Birlirakis, C. Goulard, L. Dubost, R. Thai, J. C. Tabet and S.
Rebuffat, J. Biol. Chem., 2004, 279, 28233–28242.
12 V. Braun, Drug Resist. Updates, 1999, 2, 363–369.
13 M. S. Diarra, M. C. Lavoie, M. Jacques, I. Darwish, E. K. Dolence,
J. A. Dolence, A. Ghosh, M. Ghosh, M. J. Miller and F. Malouin,
Antimicrob. Agents Chemother., 1996, 40, 2610–2617.
14 M. J. Miller and F. Malouin, Acc. Chem. Res., 1993, 26, 241–249.
15 T. A. Wencewicz, U. Mollmann, T. E. Long and M. J. Miller, BioMetals,
2009, 22, 633–648.
16 F. Rivault, C. Liebert, A. Burger, F. Hoegy, M. A. Abdallah, I. J. Schalk
and G. L. A. Mislin, Bioorg. Med. Chem. Lett., 2007, 17, 640–644.
17 A. Ghosh, M. Ghosh, C. Niu, F. Malouin, U. Moellmann and M. J.
Miller, Chem. Biol., 1996, 3, 1011–1019.
18 C. Hennard, Q. C. Truong, J. F. Desnottes, J. M. Paris, N. J. Moreau
and M. A. Abdallah, J. Med. Chem., 2001, 44, 2139–2151.
19 U. Mollmann, L. Heinisch, A. Bauernfeind, T. Kohler and D. Ankel-
Fuchs, BioMetals, 2009, 22, 615–624.
20 E. Breukink and B. de Kruijff, Nat. Rev. Drug Discov., 2006, 5, 321–332.
21 L. A. Martin-Visscher, S. Yoganathan, C. S. Sit, C. Lohans and J. C.
Vederas, FEMS Microbiol. Lett., 2010 - in press.
g) and succinic acid (2 g) dissolved in milli-Q water. The pH
of the media was adjusted to 7.0 and sterilized before use. All
samples tested against Gram-negative organisms were mixed with
equimolar ferric iron prior to testing.
In the spot on lawn assays, all agar was mixed with media
(APT, LB, M9 or succinic minimal media), dissolved in water,
and sterilized before use. For each test organism, 100 mL of grown
culture was added to a tube containing 10 mL of melted soft (0.7%
W V-1) agar (40 ◦C) before being gently vortexed and poured onto
a hard (1.5% W V-1) agar plate. After the soft agar layer solidified,
the test samples (10 mL) were pipetted on to the soft agar, allowed
to air dry and incubated overnight at appropriate temperatures
(25 ◦C or 37 ◦C). Activity was detected by the appearance of a
circular zone of growth inhibition in and around the area where
the samples were spotted.
The 64-well plate assay was done under iron deficient conditions,
using M9 minimal media. In each well, 50 mL of bacterial culture
and 50 mL of test sample were added. The 64-well plate was then
◦
incubated in an automatic reader at 37 C and optical densities
were recorded at 600 nm.
Fluorescent microscopy imaging
All test organims were first grown in LB broth overnight,
subcultured (2% innoculum) in M9 minimal media and grown
to an OD600 ~ 0.1. The sampls were centrifuged and resuspended
in Tris-buffer. Cell culture (50 mL) and fluorescently labeled
siderophore-Fe(III) complexes (50 mL) were mixed, vortexed and
incubated for 1 h. Cells were isolated by centrifugation (5 mins,
13,000 rpm), washed with Tris-buffer three times, resuspended
in Tris-buffer (100 mL) and spotted on glass slides. The cells
were immobilized with poly-lysine treated coverslips. Bacterial
cells were imaged using a Nikon TE2000-E microscope, equiped
with an epifluorescence optic, using a 100¥ objective. Images were
captured using a Fastcam Super 10 K camera (Photron USA, Inc.,
San Diego, CA).
22 D. O. Ukuku and W. F. Fett, J. Food Prot., 2004, 67, 2143–2150.
23 K. A. Stevens, B. W. Sheldon, N. A. Klapes and T. R. Klaenhammer,
Appl. Environ. Microbiol., 1991, 57, 3613–3615.
24 I. S. Boziaris and M. R. Adams, J. Appl. Microbiol., 2001, 91, 715–724.
25 D. Destoumieux-Garzon, X. Thomas, M. Santamaria, C. Goulard, M.
Barthelemy, B. Boscher, Y. Bessin, G. Molle, A. M. Pons, L. Letellier,
J. Peduzzi and S. Rebuffat, Mol. Microbiol., 2003, 49, 1031–1041.
26 E. Strahsburger, M. Baeza, O. Monasterio and R. Lagos, Antimicrob.
Agents Chemother., 2005, 49, 3083–3086.
27 C. D. Cox, K. L. Rinehart, M. L. Moore and J. C. Cook, Proc. Natl.
Acad. Sci. U. S. A., 1981, 78, 4256–4260.
Acknowledgements
28 F. Rivault, V. Schons, C. Liebert, A. Burger, E. Salcr, M. A. Abdallah,
I. J. Schalk and G. L. A. Mislin, Tetrahedron, 2006, 62, 2247–2254.
29 J. M. Roosenberg, Y. M. Lin, Y. Lu and M. J. Miller, Curr. Med. Chem.,
2000, 7, 159–197.
We thank Professor Jonathan J. Dennis (University of Alberta)
for providing us with a sample of the B. cepacia strain and
Dr Caishun Li for assistance with fluorescence microscopy. We
thank Dr Marco van Belkum for helpful discussions, Victor Dong
for assistance with HPLC purifications and Dr Randy Whittal
and Jing Zheng for extensive MS/MS analyses. We are grateful
for financial support from the Natural Sciences and Engineering
Research Council of Canada (NSERC) and the Canada Research
Chair in Bioorganic and Medicinal Chemistry.
30 S. K. Sharma, M. J. Miller and S. M. Payne, J. Med. Chem., 1989, 32,
357–367.
31 A. Zamri, I. J. Schalk, F. Pattus and M. A. Abdallah, Bioorg. Med.
Chem. Lett., 2003, 13, 1147–1150.
32 D. Cobessi, H. Celia and F. Pattus, J. Mol. Biol., 2005, 352, 893–904.
33 M. Ghosh and M. J. Miller, Bioorg. Med. Chem., 1996, 4, 43–48.
34 Please see the supporting document for evaluation of other bacterial
strains.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2133–2141 | 2141
©