Synthesis and studies of catechol-containing mycobactin S and T analogs

Article abstract of DOI:10.1039/b703116e

The syntheses of catechol-containing mycobactin S and T analogs are described. These analogs incorporate a catechol-glycine moiety in place of the phenol-oxazoline of the naturally occurring mycobactins S and T. Studies indicated that the new siderophore

Full text of DOI:10.1039/b703116e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Synthesis and studies of catechol-containing mycobactin S and T analogs
Andrew J. Walz,a Ute Mo¨llmannb and Marvin J. Miller*a,b
Received 1st March 2007, Accepted 29th March 2007
First published as an Advance Article on the web 19th April 2007
DOI: 10.1039/b703116e
The syntheses of catechol-containing mycobactin S and T analogs are described. These analogs
incorporate a catechol-glycine moiety in place of the phenol-oxazoline of the naturally occurring
mycobactins S and T. Studies indicated that the new siderophore analogs bind iron, and promote the
growth of a number of microbes, especially strains of mycobacteria, as expected.
development are being considered. Recent elegant independent
Introduction
studies by the groups of Aldrich9 and Quadri10 have shown that
The mycobactins are a unique class of microbial siderophores
found in the cell walls of all species of mycobacteria.1 The
designed inhibitors of early steps of siderophore (mycobactin)
biosyntheses by M. tuberculosis are potent antiTB agents. Our
structures and function of these interesting compounds were
laboratories and others have reported that drug conjugates of
described in an elegant series of studies by Snow in the 1960s that
siderophores can lead to the development of microbe-selective
was summarized in a superb review in 1970.2 Extended studies
“Trojan horse” compounds that utilize microbial iron assimila-
tion processes for active transport of antibiotics into bacteria.11
Additional studies from our laboratories and others have shown
that synthetic analogs of the mycobactins also serve as effective
antimycobacterial, and antiTB, agents, presumably by interfering
with iron assimilation.12 For example, we have demonstrated that
synthetic mycobactin S, with an (S)-methyl group configuration
in the butyrate linker region, inhibits the growth of Mycobacteria
tuberculosis H37Rv while mycobactin T, with an (R)-methyl group
configuration, promotes growth (Fig. 1)13 and substitution of the
butyrate linker of the natural mycobactins with a a-Boc-protected
a,b-diaminopropionate linker results in potent inhibition of the
growth of TB.14 Continuation of the search for compounds capable
of influencing essential iron assimilation processes in microbes led
us to consider syntheses of additional mycobactin analogs.
revealed that the mycobactins sequester and bind iron(III) with
three bidentate ligands: a phenol-oxazoline and two hydroxamic
acids. Besides revealing that mycobactins are important growth
factors for various strains of mycobacteria, Snow also found
that mycobactins from one type of mycobacteria antagonized the
growth of others. Over time, the potential for the development
of selective antimycobacterial agents based on altered structures
of mycobactins became known as “Snow’s hypothesis.” Demon-
strating this hypothesis became even more intriguing with further
elucidation of the complex mechanisms employed by mycobacteria
for assimilation of iron,3–5 the determination of details related
to biosyntheses of mycobactins6 and demonstrations of the
absolute dependence of mycobacterial growth and virulence on
iron acquisition.5,7
Tuberculosis (TB) remains the most severe bacterial disease
in the world and infects nearly one third of the world’s pop-
ulation (over two billion people) in latent form.8 Active forms
of TB are especially prevalent in immune compromised patients
and each year millions die. Treatment of infections due to
Mycobacterium tuberculosis, the causative agent of this dreaded
disease, remains complex and new chemotherapeutic approaches
are needed. Since M. tuberculosis also depends critically on
iron assimilation and metabolism for survival and virulence, it
expends significant energy on maintaining proper iron balance.
Under iron deficient conditions, upregulated gene expression
enhances biosynthesis of mycobacterial siderophores and other
iron sequestration machinery. Conditions of iron excess also result
in induction of down regulatory processes since iron induced free
radical processes can be degradative. Thus, the need for careful
control of a critical balance of iron may provide an “Achilles
heel” for the development of selective new tuberculosis and other
antimicrobial agents. At least three methods of utilizing the iron
assimilatory processes of mycobacteria for potential therapeutic
Of all the known Fe(III) binding ligands of naturally occurring
siderophores, the 2,3-hydroxybenzene (catechol) ligand binds most
tightly to Fe(III).15 Replacement of the natural phenol-oxazoline
iron binding component of the mycobactins with a catechol ligand
was anticipated to extend structure–activity relationship (SAR)
studies of mycobactins and analogs. Herein, we describe the
syntheses and biological studies of catechol mycobactin analogs 1
and 2. The butyrate linker regions of mycobactins S and T were
included in the syntheses of the new analogs to probe whether
or not the stereochemistry of the methyl group would affect the
activity of the analogs as it does with natural mycobactins S and T.
Results and discussion
As shown below for the retrosynthesis of 1 (Scheme 1), the assem-
bly of mycobactins typically proceeds through condensation of the
corresponding mycobactic acid and cobactin fragments. These,
in turn, consist of (L)-lysine-based hydroxamate iron binding
ligands, present in both the linear and cyclic forms, respectively.
The importance of e-N-hydroxylysine and d-N-hydroxyornithine
derivatives for the syntheses of various siderophores has prompted
considerable interest in the development of methods for gen-
eration of hydroxylamines and hydroxamic acids from amines
aDepartment of Chemistry & Biochemistry, 251 Nieuwland Science Hall,
University of Notre Dame, Notre Dame, IN, 46556, USA
bLeibniz Institute for Natural Product Research and Infection Biology—
Hans Knoell Institute, Beutenbergstrasse 11a, 07745, Jena, Germany
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 1621–1628 | 1621
©

View Article Online
Fig. 1 Structures of representative mycobactins and analogs.
Scheme 1
(Scheme 2).16–21 In this case we synthesized both the acyclic and
cyclic lysine derived hydroxamic acids using the stable, storable
and readily accessible nitrone intermediate 3 shown in Scheme 3.
Thus, condensation of Z-(L)-lysine with freshly distilled ben-
zaldehyde was followed by dry m-CPBA-mediated oxidation of
the intermediate imine to form the corresponding oxaziridine
3. Isomerization under acidic conditions led to nitrone 4 in
good overall yield from the starting protected amino acid. The
reported syntheses of the azepine hydroxamic acid 6 relied upon
TBDPS protection of the generated hydroxamic acid. Efforts to
eliminate the need for this protecting group were elaborated.
Formation of the pre-cyclization hydroxylamine HCl salt of 5 was
accomplished by the action of 1.1 equivalents of NH2OH(HCl)
on nitrone 4 at elevated temperatures in MeOH. Cyclization was
accomplished using 1.1 and 1.2 equivalents of EDC and HOAt,
respectively, under basic, high dilution conditions. After reverse
phase chromatography, hydroxamic acid 6 was obtained in 55%
overall yield from the nitrone 4.
Acyclic hydroxamic acid 9 was also readily prepared from the
lysine nitrone. Thus, treatment of nitrone 4 with NH2OH(HCl)
resulted in the formation of hydroxylamine 5. Reaction of 5 with
SOCl2 in MeOH gave methyl ester 7. The e-N-hydroxylysine
methyl ester was then treated with a large excess of palmitoyl
chloride and NaHCO3 in CH2Cl2 to provide bis-acylated hydrox-
ylamine 8 in excellent yield. According to precedent, solvolytic
removal of the O-palmitoyl group from the hydroxamate with
6% DIPEA in MeOH resulted in the formation of the known
Z-protected lysine hydroxamic acid 9.13
Using a literature procedure,21 2,3 dihydroxybenzoic acid
was reacted with BnBr and KOH in DMSO to give the 2,3-
benzyloxybenzoic acid 10 in good yield (Scheme 4). The formation
of an amide bond with this acid and glycine methyl ester has
also been reported but with no experimental detail.22 Treatment
of 10 with oxalyl chloride and catalytic DMF led to the acid
chloride, which, upon treatment with glycine methyl ester gave
amide 11 in excellent yield. Saponification of the methyl ester of 11
1622 | Org. Biomol. Chem., 2007, 5, 1621–1628
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Scheme 2
Scheme 3 Reagents and conditions: a. 1) PhCHO, KOH, MS, MeOH, rt, 16 h, 2) m-CPBA, MeOH, 0 ◦C to rt, 4 h. b. 1) TFA, CH2Cl2, rt, 1 h. 2) PhCHO,
EtOAc, 0 ◦C to rt (66% overall). c. 1) NH2OH(HCl), MeOH, 65 ◦C, 20 min. d. EDC, HOAt, NaHCO3, CH3CN, DMF, rt, 48 h (55%). e. SOCl2, MeOH,
0
◦C to rt, 12 h. f. Palmitoyl chloride, NaHCO3,CH2Cl2, rt, 12 h (95% overall from 5).
Scheme 4 Reagents and conditions: a. 1) Oxalyl chloride, DMF (cat.), benzene, 0 ◦C to rt, 3 h, 2) glycine methyl ester(HCl), NaHCO3, CH3CN, rt, 12 h
(91%). b. LiOH, THF, H2O, rt, 12 h (quant.). c. 1) 9a, EDC, HOAt, DMAP, DMF, rt, 16 h (53%). d. H2, Pd–C, MeOH, rt and pressure, 1 h (quant.).
produced the corresponding acid which was coupled to free amine
9a, obtained from hydrogenolysis of 9, to give the key catechol-
containing mycobactic acid methyl ester 12.
Completion of the syntheses of the target compounds required
the preparation of cobactins S and T. Cobactins S and T
consist of (S)- or (R)-hydroxy butyrate linked by an amide bond
to the cyclic, lysine-based hydroxamic acid derived from ester
hydrolysis of mycobactins S and T, similar to the retrosynthesis
scheme shown earlier. Using commercially available (S)- or (R)-3-
hydroxybutyric acid sodium salts, separate EDC/HOAt-mediated
coupling reactions were attempted with the amine 6a resulting
from hydrogenolytic removal of the Z-protecting group of 6.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 1621–1628 | 1623
©

View Article Online
Scheme 5 Reagents and conditions: a. H2, 10% Pd–C, MeOH, rt (quant.). b. EDC, HOAt, DMAP, DMF, rt, 24 h. c. LiOH, THF, H2O, rt, 12 h (quant.).
d. 7a, EDC, 24 h (41–55%). e. H2, 10% Pd–C, MeOH, rt (80–98%).
The reactions employed DMF as the solvent and, after the
starting materials were consumed, the aqueous work-ups proved
to be quite troublesome. Isolation of the cobactins from the
aqueous DMF/water solutions was extremely inefficient. Since
other solvents were not as effective as DMF in peptide bond
formation in the presence of hydroxamic acids, an alternative one-
pot, two coupling reaction protocol was found to be moderately
effective.
was severe and extended to the peaks corresponding to the methyl
and methylene groups of the butyrate linker. Upon hydrogenation
of the benzyl groups, giving rise to analogs 12a,b, the splitting
of peaks was lessened overall and completely eliminated from
the signals corresponding to the methyl and methylene groups
in the 1H NMR, thus indicating that peak multiplicity arose from
rotation isomers and not from diastereomers.
A one flask/two EDC/HOAt-mediated coupling reaction se-
quence has been employed in the synthesis of a mycobactin
analog.14 For the generation of the catechol analogs, the reac-
tions to form cobactins S and T were performed, in the same
manner, for 24 h (Scheme 5). Concurrently, the methyl ester
of 12 was saponified. After the 24 h reaction time, acid 12a
and another 1.1 equivalents of EDC were added. This protocol
generated the benzyl protected analogs 14a,b in 41% overall yield
for the (S)-methyl group configuration and 54% for the (R)-
methyl group configuration. Pd–C catalyzed hydrogenolysis of the
benzyl protecting groups provided the target compounds 1 and 2
cleanly.
As with many peptides, the presence of rotational isomers,
detected in the NMR spectra of synthetic analogs, is possible.
Distinguishing them from diastereomers can be quite time con-
suming. This proved to be the case with compounds 14a,b and 1,2.
Splitting of peaks in both the 1H and 13C NMR spectra was noted.
With the bis-benzylated intermediates 14a,b, the splitting of peaks
Assays of synthetic conjugates 1 and 2
The relative iron binding capacity of the desferri- or ferri-
siderophores is often determined by the chrome azurol S (CAS)
assay.23 Usually, the diameter of the resulting orange halo in mm
for a 5 lL sample reflects the extent of binding. However, as
indicated we have found that the test is not always positive with
very hydrophobic siderophores or analogs that tend to precipitate
in the assay. Also, of course, the CAS assay cannot work with iron
bound siderophores. As shown in Tables 1–3 below, the synthetic
mycobactin analogs, 1 and 2, did give a positive CAS test while
authentic mycobactin J, which contains iron, was negative, and the
more hydrophilic trihydroxamate, desferrioxamine B (Desferal),
was very responsive, as expected. With the iron binding affinity
of 1 and 2 confirmed, the compounds were also tested for their
ability to inhibit or promote the growth of various bacterial strains,
including mycobacteria. Potential antibacterial activity of the
compounds was studied by determination of minimal inhibitory
Table 1 Growth promotion of Gram-negative strains by mycobactin analogs 1 and 2, and desferrioxamine (Desferal)a
P. aeruginosa
E. coli
S. typhimurium
Compound
CAS
K799/WT
ATCC 9027
NCTC 10662
ATCC 25922
enb-7
1
++
+
++++
A
0
40
A
13
38
13
A11
40
14
12
H20
0
0
14
2
Desferal
a Samples (5 lg) were applied to 6 mm diameter discs and the growth zones are given in mm. Additional indicators are: H = inhibition of growth in mm,
A = small indication of growth.
1624 | Org. Biomol. Chem., 2007, 5, 1621–1628
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Table 2 Growth promotion of mycobacterial strains by mycobactin analogs 1 and 2, and mycobactin Ja
Mycobacterium smegmatis
Compound
CAS
SG 987
SG 987-M10
mc2155
mc2155-M24
mc2155-B1
mc2155-M24-B3
mc2155-M24-U3
1
++
+
—
32
24
25
16
16
15
17
18
16
17
15
17
17
15
15
14
8
12
H7/A14
H7/A
15
2
Mycobactin J
a Samples (5 lg) were applied to 6 mm diameter discs and the growth zones are given in mm. Additional indicators are: H = inhibition of growth in mm,
A = small indication of growth.
Table 3 Characteristics of the M. smegmatis strains used in these studies
Mycobacterium smegmatis
Bacterial strains/characteristic
SG 987
SG 987-M10
mc2155
mc2155-M24
mc2155-B1
mc2155-M24-B3
mc2155-M24-U3
Biosynthesis of:
Exochelin
Mycobactin
+
+
−
+
+
+
−
−
−
−
+
−
+
+
Exochelin permease
+
+
+
+
+
−
−
concentrations (MIC) according to the NCCLS guidelines using
the micro broth dilution method.24 Neither 1 nor 2 displayed
antibiotic activity against Gram-positive and Gram-negative
bacteria or against mycobacteria under normal assay conditions
(data not shown).
Conclusions
The synthesis of catechol containing mycobactin S and T analogs
has been accomplished. The novel catechol analogs were demon-
strated to bind iron as reflected by the standard CAS assay.
Consistent with their similarity to the natural mycobactins, they
preferentially promoted the growth of the mycobacterial strains
studied relative to other bacteria. The potential use of these
compounds as mycobacterial selective drug delivery agents is
under consideration.
The desferrimycobactin analogs 1 and 2 were also studied for
their ability to promote the growth of a series of Gram-negative
bacteria and a larger set of mycobacteria since the latter were
anticipated to be most responsive. Growth zones surrounding
the discs were read after 1 day for P. aeruginosa, E. coli and
S. typhimurium strains and after two days for M. smegmatis
strains. The results are summarized in Tables 1 and 2. Table 3
provides characteristics of the mycobacterial strains used. Indeed,
as expected, mycobacteria were more responsive to 1 and 2 as
growth promoters than were the Gram-negative bacteria. In fact,
with mycobacteria, analogs 1 and 2 were generally as effective
as natural mycobactin J. While in most cases normal zones of
growth promotion were observed, in a few instances additional
affects were noted as designated by the descriptor “H” (inhibition)
in Tables 1 and 2. This less common phenomenon is not yet
well understood, but might be due to the ability of the applied
desferrisiderophore or analog to scavenge the trace of iron that
remains in the specially prepared iron depleted media used for
the assays. Thus, at higher concentration of the siderophore
and depending on the individual mode of iron uptake for a
specific strain, the microbes are additionally stressed for the
essential nutrient. This type of growth inhibition is reflected by
an inhibition-like zone near the compound loaded paper disc
where the concentration of the siderophore is highest and before
the growth zone appears. The two analogs were also tested for
possible anti-tuberculosis activity at the Tuberculosis Acquisition
And Coordinating Facility (TAACF) and found to not inhibit
the growth of M. tuberculosis (2% and 0% inhibition for 1 and 2,
respectively, at 6.25 lg mL−1). Thus, syntheses and studies of 1
and 2 demonstrate that novel analogs of natural mycobactins can
be used by mycobacteria for microbial growth.
Experimental
Lysine nitrone 4
˚
To a solution of KOH (1.05 g, 19.0 mmol) and 5 g of 3 A molecular
sieves in 50 mL of MeOH at room temperature was added Z-(L)-
lysine (5.00 g, 17.8 mmol). When the stirred solution became clear,
benzaldehyde (2.08 g, 19.6 mmol) was added and the mixture was
stirred for an additional 16 h at room temperature. The mixture
was filtered and the solvents removed under reduced pressure to
give the crude imine product. The residue was taken up in 40 mL
of MeOH and cooled to 0 ◦C. Dry m-CPBA (3.70 g, 21.4 mmol),
dissolved in 50 mL of MeOH, was added dropwise over 20 min
and the resulting solution was stirred for an additional 2 h at
0 ◦C. The reaction mixture was filtered and the solvents removed
under reduced pressure. Isomerization of the crude oxaziridine
was accomplished by the addition of 25 mL of TFA followed by
25 mL of CH2Cl2 to the residue under a nitrogen atmosphere.
The solution was stirred for 1 h at room temperature and the
solvents removed under reduced pressure. The residue was taken
up in 60 mL of EtOAc and cooled to 0 ◦C under a nitrogen
atmosphere. Benzaldehyde (1 mL) was added and the solution was
stirred overnight while coming to room temperature. The solvents
were removed under reduced pressure and the residue purified
by flash chromatography (80% CH2Cl2 : 20% acetone to remove
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 1621–1628 | 1625
©

View Article Online
m-CPBA and PhCHO, then gradient elution to 15% MeOH : 65%
CH2Cl2 : 20% acetone) to give 4.50 g (66% yield) of the product
nitrone as a pale yellow oil. 1H NMR (300 MHz, CD3CN) d 8.24–
8.21 (m, 2H), 7.60 (s, 1H), 7.43–7.26 (m, 8H), 6.14 (d, 1H, J =
8.06 Hz), 5.35–4.85 (br s, 1H), 5.02 (s, 2H), 4.14–4.08 (m, 1H),
3.87 (t, 2H, J = 6.84 Hz), 1.94–1.34 (m, 6H); 13C NMR (75 MHz,
CD3CN) d 174.6, 157.3, 138.1, 137.1, 131.7, 131.5, 129.9, 129.5,
129.4, 128.9, 128.7, 67.1, 66.9, 58.9, 31.9, 27.8, 23.3; IR (neat)
3323, 3064, 2953, 1713, 1531 cm−1; HRFABMS for C21H25N2O5
(MH+) calcd 385.1763, found 385.1750.
at room temperature for 16 h. The solvent was removed under
reduced pressure and the residue was taken up in EtOAc. The
organic layer was washed with water, twice with 1 N HCl, and
twice with brine. The organic solution was dried with NaS2O4,
filtered and concentrated. The residue was purified by silica gel
chromatography (CH2Cl2 to 16 : 1 CH2Cl2 : EtOAc) to give 2.89 g
(95% yield) of the product as a white solid. 1H NMR (300 MHz,
CDCl3) d 7.34–7.25 (m, 5H), 5.66 (br d, 1H, J = 8.55 Hz), 5.10 (s,
2H), 4.36–4.31 (m, 1H), 3.37 (s, 3H), 3.69–3.58 (m, 2H), 2.42 (t,
2H, J = 7.45 Hz), 2.17 (br t, 2H, J = 7.32 Hz), 1.82–1.18 (m, 58H),
0.87 (t, 6H, J = 6.32 Hz); 13C NMR (75 MHz, CDCl3) d 178.1,
173.0, 171.6, 156.1, 136.4, 128.7, 128.4, 128.3, 67.2, 53.9, 52.6,
34.0, 32.5, 32.1, 32.0, 29.9, 29.8, 29.7, 29.6, 29.5, 29.4, 29.3, 26.7,
24.9, 24.5, 22.9, 22.4, 14.3; IR (neat) 3332, 2924, 2853, 1790, 1725,
1674 cm−1; HRFABMS for C47H83N2O7 (MH+) calcd 787.6200,
found 787.6218.
Azepinone hydroxamic acid 6
Lysine nitrone 4 (1.00 g, 2.87 mmol) and hydroxylamine hy-
drochloride (0.22 g, 3.16 mmol) were dissolved in 10 mL of
MeOH and heated in a 60 ◦C oil bath for 20 min and allowed
to cool. The solution was concentrated, taken up in distilled H2O
and extracted with diethyl ether until there was no more by-
product in the organic phase as detected by TLC. The aqueous
solvent was removed under reduced pressure. Residual water was
removed azeotropically under reduced pressure with benzene and
methanol. The crude hydroxylamine in 90 mL of 7 : 2 CH3CN :
DMF was added dropwise over 2 h to a stirred solution of EDC
(546 mg, 2.84 mmol), HOAt (355 mg, 2.61 mmol), and NaHCO3
(996 mg, 11.9 mmol) in 50 mL of 7 : 2 CH3CN : DMF. The
solution was stirred for 48 h at room temperature. The CH3CN
was removed under reduced pressure and EtOAc and H2O were
added. The aqueous layer was acidified to approximately pH 2 with
1 N HCl and extracted with EtOAc until no more product could
be detected by TLC. The combined organic extracts were washed
twice with 5% NaHCO3 and once with brine. The organic layer
was dried with Na2SO4, filtered and concentrated. The residue
was purified by C18 reverse phase silica gel chromatography (3 :
2 H2O : THF) to give 363 mg (55% yield) of the product as an
Dibenzyloxybenzoylglycine methyl ester 11
2,3-Benzyloxybenzoic acid 10 (0.500 g, 1.50 mmol) was added
to 5 mL of benzene and cooled to 0 ◦C. Oxalyl chloride (0.762 g,
6.00 mmol) was added followed by one drop of DMF. The reaction
was stirred for 12 h while coming to room temperature. The
solution was concentrated and dissolved in 10 mL of CH3CN. At
room temperature, NaHCO3 (1.26 g, 15.00 mmol) was added
followed by the HCl salt of glycine methyl ester (0.208 g,
1.65 mmol). The reaction was stirred for 24 h and the solvents
removed under reduced pressure. The residue was taken up in
EtOAc and washed with 1 N HCl, water, and brine. The organic
layer was dried with NaS2O4, filtered and concentrated. The
residue was purified by silica gel chromatography (4 : 1 hexanes :
EtOAc) to give 0.553 g (91% yield) of the product as a white solid.
1H NMR (300 MHz, CDCl3) d 8.52–8.44 (m, 1H), 7.73 (dd, 1H,
J = 3.59, 6.05 Hz), 7.49–7.15 (m, 12 H), 5.16 (s, 2H), 5.14 (s,
2H), 4.08 (d, 2H, J = 5.46 Hz), 3.73 (s, 3H); 13C NMR (75 MHz,
CDCl3) d 170.3, 165.5, 152.0, 147.2, 136.5, 129.2, 128.9, 128.7,
128.6, 128.5, 128.0, 126.8, 124.6, 123.5, 117.5, 76.6, 71.5, 52.4,
41.7; IR (neat) 3380, 3032, 2951, 1748, 1660 cm−1; HRFABMS for
C24H24NO5 (MH+) calcd 406.1654, found 406.1641.
1
amorphous white solid. H NMR (300 MHz, CD3OD) d 7.37–
7.28 (m, 5H), 5.08 (s, 2H), 4.35 (d, 1H, J = 11.48 Hz), 3.91 (dd,
1H, J = 11.23, 16.11 Hz), 3.66 (d, 1H, J = 15.87 Hz), 2.00–1.50
(m, 6H); 13C NMR (75 MHz, CD3OD) d 171.3, 157.9, 138.2, 129.4,
129.0, 128.8, 67.6, 54.3, 53.9, 32.0, 28.6, 26.9; IR (neat) 3333, 3125,
2917, 1688, 1650, 1622, 1529 cm−1; HRFABMS for C14H19N2O4
(MH+) calcd 279.1345, found 279.1352.
Catechol-containing mycobactic acid methyl ester analog 12
To methyl ester 11 (0.468 g, 1.15 mmol) in 10 mL of 1 : 1 THF :
water was added LiOH (0.138 g, 5.75 mmol) at room temperature.
The reaction was stirred for 12 h, diluted with water, and acidified
to pH = 2 with 1 N HCl. The aqueous layer was extracted
three times with EtOAc. The combined organic extracts were
washed with water and brine. The organic solution was dried with
NaS2O4, filtered and concentrated to give acid 11a which was used
immediately in the next reaction. In a separate flask, a solution of 9
(0.760 g, 1.39 mmol) in 10 mL of MeOH was purged with nitrogen.
Pd–C (10%, 10% weight) was added and a hydrogen atmosphere
was established with a balloon. The solution was stirred for 45 min
at room temperature and was filtered through Celite. Acid 11a
(0.400 g, 1.02 mmol), amine 9a (0.421 g, 1.02 mmol), HOAt
(0.139 g, 1.02 mmol), and catalytic DMAP were dissolved in 5 mL
of DMF at room temperature. EDC (0.215 g, 1.12 mmol) was
added and the solution was stirred for 16 h and diluted with
water. The solution was extracted three times with EtOAc. The
Bis-palmitoyl lysine hydroxylamine 8
Lysine nitrone 4 (1.48 g, 3.86 mmol) and NH2OH(HCl) (0.295 g,
4.25 mmol) were dissolved in 25 mL of MeOH and heated in a
65 ◦C oil bath for 17 min. The solution was allowed to cool to room
temperature and was placed in an ice water bath. SOCl2 (3.26 g,
27.4 mmol) was added and the solution was stirred for 16 h while
coming to room temperature. The solvents were removed under
reduced pressure and saturated NaHCO3 and CH2Cl2 were added.
The layers were separated and the aqueous layer was extracted
three times with CH2Cl2. The combined organic extracts were
dried with NaS2O4, filtered and concentrated. This residue was
purified by silica gel chromatography (9 : 1 CH2Cl2 : EtOAc
to remove impurities, then EtOAc to 95 : 5 EtOAc : MeOH).
The resulting hydroxylamine was then dissolved in 120 mL of
CH2Cl2 and NaHCO3 (2.59 g, 31.0 mmol) and palmitoyl chloride
(4.24 g, 15.4 mmol) were added and the mixture was stirred
1626 | Org. Biomol. Chem., 2007, 5, 1621–1628
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
combined organic extracts were washed with 1 N HCl, water, 5%
NaHCO3, and brine. The organic solution was dried with NaS2O4,
filtered and concentrated. The residue was purified by silica gel
chromatography (4 : 1 to 2 : 1 hexanes : EtOAc) to give 0.425 g
(53% yield) of the product as an amorphous white solid. 1H NMR
(600 MHz, CD3OD) d 7.49 (br d, 1H, J = 7.56 Hz), 7.43 (dd, 1H,
J = 1.50, 7.92 Hz), 7.39–7.22 (m, 10H), 7.13 (t, 1H, J = 8.01 Hz),
5.17 (s, 2H), 5.13 (s, 2H), 4.42 (dd, 1H, J = 4.95, 8.88 Hz), 4.00
(d, 1H, J = 16.80 Hz), 3.94 (d, 1H, J = 16.90 Hz), 3.70 (s, 3H),
3.63–3.54 (m, 5H), 2.44–2.42 (m, 2H), 1.88–1.82 (m, 1H), 1.73–
1.54 (m, 5H), 1.39–1.26 (m, 28H), 0.88 (t, 3h, J = 7.10 Hz); 13C
NMR (150 MHz, CD3OD) d 175.8, 173.6, 170.7, 167.8, 153.1,
147.4, 137.7, 137.4, 130.0, 129.3, 129.1, 129.0, 128.9, 128.7, 128.5,
125.1, 122.7, 118.2, 76.7, 71.9, 53.3, 52.4, 43.3, 32.9, 32.7, 31.7,
30.4, 30.3, 30.2, 30.1, 26.8, 25.6, 23.4, 14.1; IR (neat) 3283, 2924,
2854, 1746, 1643 cm−1; HRFABMS for C46H66N3O8 (MH+) calcd
788.4850, found 788.4870.
(neat) 3306, 2925, 1634 cm−1; HRFABMS for C55H80N5O11 (MH+)
calcd 986.5854, found 986.5884.
Catechol-containing mycobactin S analog 1
A solution of 14a (0.050 g, 0.051 mmol) in 3 mL of MeOH was
purged with nitrogen. Pd–C (10 mol% of 10% by weight) was
added and a hydrogen atmosphere was established with a balloon.
The solution was stirred for 3 h at room temperature. The mixture
was filtered through a KimWipe plug/pipette and the filtrate was
concentrated. The procedure was repeated until no catalyst was
visible in the filtrate. The residue was purified by C18 reverse phase
silica gel chromatography (1 : 1 to 85 : 15 MeOH : H2O) to give
0.033 g (98% yield) of the product as a clear glassy solid. 1H NMR
(rotational isomers present, major peaks reported when possible):
(600 MHz, CD3OD) d 7.25 (dd, 1H, J = 1.19, 8.06 Hz), 6.93 (dd,
1H, J = 1.46, 7.80 Hz), 6.72 (t, 1H, J = 8.00 Hz), 4.63 (q, 1H,
J = 11.24 Hz), 4.53–4.50 (m, 1H), 4.14–4.00 (m, 4H), 3.70–3.57
(m, 3H), 2.45 (br t, 2H, J = 7.63 Hz), 2.39 (dd, 1H, J = 7.69,
14.04 Hz), 2.33 (dd, 1H, J = 5.00, 14.16 Hz), 2.01–1.23 (m, 38H),
1.19 (d, 3H, J = 6.23 Hz), 0.89 (t, 3H, J = 7.08 Hz); 13C NMR
(rotational isomers present, major peaks reported when possible):
(150 MHz, CD3OD) d 176.3, 174.2, 173.5, 172.0, 171.8, 171.2,
170.7, 150.4, 147.5, 129.9, 119.8, 116.8, 66.2, 66.1, 54.2, 54.1, 53.8,
53.4, 53.3, 53.0, 52.7, 52.6, 46.3, 43.5, 33.4, 33.2, 31.9, 31.8, 30.9,
30.8, 30.7, 30.6, 28.8, 27.4, 26.1, 23.9, 23.8, 23.3, 14.6; IR (neat)
3307, 2924, 2853, 1636 cm−1; HRFABMS for C41H68N5O11 (MH+)
calcd 806.4915, found 806.4896.
Bis-benzylated catechol-containing mycobactin S analog 14a
A solution of azepine hydroxamic acid 6 (0.053 g, 0.188 mmol) in
5 mL of MeOH was purged with nitrogen. Pd–C (10 mol% of 10%
by weight) was added and a hydrogen atmosphere was established
with a balloon. The solution was stirred for 45 min at room
temperature and was filtered through Celite to give a solution of
6a that was used immediately in the next reaction. The filtrate was
concentrated and (R)-3-hydroxybutyric acid sodium salt (0.024 g,
0.188 mmol), HOAt (0.026 g, 0.188 mmol), catalytic DMAP, and
4 mL of DMF were added. EDC (0.040 g, 0.207 mmol) was
added and the solution was stirred at room temperature for 24 h.
Concurrently, 12 (0.150 g, 0.188 mmol) was dissolved in 5 mL of
1 : 1 THF : water. LiOH (0.014 g, 0.565 mmol) was added and the
mixture was stirred at room temperature for 12 h. The solution
was diluted with water and acidified to pH = 2 with 1 N HCl.
The aqueous solution was extracted three times with EtOAc. The
combined organic extracts were washed with water and brine. The
organic solution was dried with NaS2O4, filtered and concentrated.
The resulting acid, 12a, was added to the above reaction mixture
after 24 h followed by 1.1 equivalents of EDC. The solution
then was stirred for 24 h at room temperature and diluted with
EtOAc. The organic solution was washed with 1 N HCl, water, 5%
NaHCO3, and brine. The organic solution was dried with NaS2O4,
filtered and concentrated. The residue was purified by C18 reverse
phase silica gel chromatography (1 : 1 to 85 : 15 MeOH : H2O)
Catechol-containing mycobactin T analog 2
A solution of 14b was hydrogenated and purified as the catechol
containing mycobactin S analog to provide 0.040 g (98% yield) of
1
a clear glassy solid. H NMR (rotational isomers present, major
peaks reported when possible): (600 MHz, CD3OD) d 7.25 (dd,
1H, J = 0.85, 8.06 Hz), 6.93 (dd, 1H, J = 1.41, 7.88 Hz), 6.72 (t,
1H, J = 8.00 Hz), 4.65–4.60 (m, 1H), 4.53–4.51 (m, 1H), 4.16–4.00
(m, 4H), 3.69–3.56 (m, 3H), 2.45 (br t, 2H, J = 7.63 Hz), 2.36 (d,
1H, J = 7.45 Hz), 2.35 (d, 1H, J = 5.50 Hz), 2.00–1.27 (m, 38H),
1.20 (d, 3H, J = 6.23 Hz), 0.89 (t, 3H, J = 7.08 Hz); 13C NMR
(rotational isomers present, major peaks reported when possible):
(150 MHz, CD3OD) d 176.3, 173.4, 172.0, 171.9, 170.8, 170.7,
150.4, 147.5, 120.0, 119.8, 119.2, 116.7, 66.1, 54.1, 53.3, 53.2, 52.7,
52.6, 46.1, 43.5, 33.5, 33.2, 32.0, 31.9, 31.8, 31.0, 30.9, 30.8, 30.7,
30.6, 28.9, 28.8, 27.4, 27.3, 27.1, 23.9, 23.8, 23.6, 14.6; IR (neat)
3298, 2924, 2854, 1640 cm−1; HRFABMS for C41H68N5O11 (MH+)
calcd 806.4915, found 806.4952.
1
to give 0.100 g of the product as a clear solid in 41% yield. H
NMR (rotational isomers present, major peaks reported when
possible): (600 MHz, CD3OD) d 7.51–7.49 (m, 1H), 7.44–7.42 (m,
1H), 7.40–7.23 (m, 10H), 7.16–7.13 (m, 1H), 5.19 (s, 2H), 5.14 (s,
2H), 4.64–4.58 (m, 1H), 4.40 (dd, 1H, J = 4.65, 8.70 Hz), 4.13–4.10
(m, 1H), 4.03–3.90 (m, 3H), 3.67–3.55 (m, 3H), 2.43 (dd, 2H, J =
6.63, 14.64 Hz), 2.39 (dd, 1H, J = 7.85, 14.04 Hz), 2.33 (dd, 1H,
J = 4.97, 14.05 Hz), 1.97–1.87 (m, 2H), 1.81–1.54 (m, 8H), 1.40–
1.23 (m, 28H), 1.20 (d, 3H, J = 6.30 Hz), 0.88 (t, 3H, J = 7.05 Hz);
13C NMR (rotational isomers present, major peaks reported when
possible): (150 MHz, CD3OD) d 176.3, 173.6, 171.5, 171.2, 171.1,
168.4, 153.7, 148.0, 138.3, 138.0, 130.5, 129.8, 129.6, 129.5, 129.4,
129.2, 129.1, 125.6, 123.2, 118.7, 77.2, 72.4, 66.2, 54.2, 53.9, 53.4,
53.0, 46.4, 46.2, 43.9, 43.7, 33.5, 33.2, 32.5, 31.8, 31.0, 30.9, 30.8,
30.7, 30.6, 28.8, 27.4, 27.3, 27.1, 26.1, 24.0, 23.9, 23.3, 14.6; IR
Bacterial strains
Test organisms from Culture Collections (ATCC, NCTC) and
from the stock of the Hans Knoell Institute (SG) included: Bacillus
subtilis (ATCC 6633), Staphylococcus aureus (SG 511), Mycobac-
terium smegmatis (SG 987, mc215525 and mutants thereof26,27),
Escherichia coli (ATCC 25922), Klebsiella pneumoniae (ATTC
10031), Serratia marcescens (SG 621), Stenotrophomonas mal-
tophilia GN 12853 (kindly provided by the Episome Institute,
Gunma (Japan)),28 Pseudomonas aeruginosa (ATCC 27853, ATCC
9027, NCTC 10662), (SG 137), K799/WT, K799/61 (wild type
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 1621–1628 | 1627
©

View Article Online
and permeability mutant),29 E. coli DC 0, DC 2 (wild type and
4 C. L. Ratledge and G. Dover, Annu. Rev. Microbiol., 2000, 54, 881–41.
5 C. L. Ratledge, Tuberculosis, 2004, 84, 110–130.
6 L. E. N. Quadri, J. Sello, T. A. Keating, P. H. Weinreb and C. T. Walsh,
Chem. Biol., 1998, 5, 631–645.
7 G. M. Rodriguez, Trends Microbiol., 2004, 14, 3320–325.
8 (a) http://www.taacf.org/about-TB-epidemic.htm; (b) http://www.
who.int/mediacentre/factsheets/fs104/en/#global, WHO Fact sheet
no. 104, revised March 2006.
9 R. Somu, H. Boshoff, C. H. Qiao, E. M. Bennett, C. E. Barry and C. C.
Aldrich, J. Med. Chem., 2006, 49, 31–34.
10 J. A. Ferras, J. S. Ryu, F. Di Lello, D. S. Tan and L. E. N. Quadri, Nat.
Chem. Biol., 2005, 1, 29–32.
11 J. M. Roosenberg, Y.-M. Lin, Y. Lu and M. J. Miller, Curr. Med. Chem.,
2000, 7, 159–197.
12 G. R. Marshall, P. A. Reddy, O. F. Schall, A. Naik, D. D. Beusen, Y.
Ye and U. Slomczynska, in Advances in Supramolecular Chemistry, ed.
G. W. Gokel, JAI Press, Stamford, CT, USA, 2002, vol. 8, ch. 5, pp.
175–243.
permeability mutant).30
MIC determination
The bacteria described were grown overnight at 37 ◦C in Mueller–
Hinton broth (MHB) (Difco). 50 lL of a compound solution of
400 lg mL−1 were serially diluted by a factor of two with MHB
in microtiter plates. Then the wells were inoculated with 50 lL
of the bacterial solution to give a final concentration of 5 × 105
CFU mL−1. After the microtiter plates were incubated at 37 ◦C
for 24 h, the MIC values were read with a Nepheloscan Ascent
1.4 automatic plate reader (Labsystems, Vantaa, Finland) as the
lowest dilution of antibiotic allowing no visible growth.
13 J. Hu and M. J. Miller, J. Am. Chem. Soc., 1997, 119, 3462–3468.
14 Y. Xu and M. J. Miller, J. Org. Chem., 1998, 63, 4314–4322.
15 R. C. Hider, Siderophore Mediated Absorption of Iron, ed. M. J. Clarke,
J. B. Goodenough, J. A. Ibers, C. K. Jorgensen, D. M. P. Mingos,
J. B. Nielands, G. A. Palmer, D. Reinen, P. J. Sadler, R. Weiss and
R. J. P. Williams, Springer-Verlag, New York, 1984, vol. 58, pp. 25–
88.
16 J. D. Fields and P. J. Kropp, J. Org. Chem., 2000, 65, 5937–5940.
17 J. Hu and M. J. Miller, J. Org. Chem., 1994, 59, 4858–4861.
18 H. Tokuyama, T. Kuboyama, A. Amano, T. Yamashita and T.
Fukuyama, Synthesis, 2000, 1299–1304.
Siderophore assays
Utilization of compounds 1 and 2 as siderophores was determined
by a growth promotion assay as described.26,27 Strains were
suspended in the iron depleted agar media hindering normal
bacterial growth. The inoculated media were poured into Petri
dishes. Siderophore solutions (2 mM, 5 lL) were applied on paper
discs of 6 mm in diameter on the surface of the agar plates.
Growth zones surrounding the discs were read after 1 day for
P. aeruginosa, E. coli and S. typhimurium strains and after 2 days
for M. smegmatis strains.
19 (a) T. Polonoski and A. Chimiak, Tetrahedron Lett., 1974, 28, 2453–
2456; (b) H.-U. Naegeli and W. Kellelr-Schierlein, Helv. Chim. Acta,
1978, 61, 2088–2095; (c) Y.-M. Lin and M. J. Miller, J. Org. Chem.,
1999, 64, 7451–7458.
20 (a) M. J. Milewska and A. Chimiak, Synthesis, 1990, 233–234; (b) R. J.
Bergeron and O. Phanstiel, IV, J. Org. Chem., 1992, 57, 7140–7143;
(c) Q. X. Wang, J. King and O. Phanstiel, IV, J. Org. Chem., 1997, 62,
8104–8108; (d) O. Phanstiel, IV, Q. X. Wang, D. H. Powll, M. Ospina
and B. A. Leeson, J. Org. Chem., 1999, 64, 803–806.
21 B. Tse and Y. Kishi, J. Org. Chem., 1994, 59, 7807–7814.
22 G. M. Buckley, G. Pattenden and D. A. Whiting, Tetrahedron, 1994,
50, 11781–11792.
23 B. Schwynn and J. B. Neilands, Anal. Biochem., 1987, 160, 47–56.
24 National Committee for Clinical Laboratory Standards. Methods for
Dilution Antimicrobial Susceptibility Tests for Bacteria that grow
Aerobically Approved Standard, NCCLS Document M7-A5, 5th edn,
National Committee for Clnical Laboratory Standards, Villanova, PA,
2000.
25 S. B. Snapper, R. E. Melton, S. Mustafa, T. Kieser and W. R. Jacobs,
Jr., Isolation and characterization of efficient plasmid transformation
mutants of Mycobacterium smegmatis, Mol. Microbiol., 1990, 4, 1911–
1919.
Acknowledgements
Support from the NIH (GM25845 and AI054193) is gratefully
acknowledged. Anti-tuberculosis testing of 1 and 2 was kindly
performed by the Tuberculosis Acquisition and Coordinating
Facility (TAACF). We also appreciate the use of the NMR
facilities provided by the Lizzadro Magnetic Resonance Center
and the mass spectrometry services (Dr W. Boggess and Ms
N. Sevova) provided by the University of Notre Dame. The
excellent technical assistance of Irmgard Heinemann with the
assays at the HKI is gratefully appreciated. MJM gratefully
acknowledges the kind hospitality of the HKI and the University
of Notre Dame for the sabbatical opportunity during which time
this manuscript was prepared.
26 G. Schumann, U. Mo¨llmann and I. Heinemann, Mutants of Mycobac-
terium species and their use for screening of antibiotic vectors, Pat. Appl.
DE 19817021,9 (17.4.1998), Gru¨nenthal GmbH.
Notes and references
27 G. Schumann and U. Mo¨llmann, Antimicrob. Agents Chemother., 2001,
1 A. F. Verne, A. F. Walz and M. J. Miller, Nat. Prod. Rep., 2000, 17,
99–116.
2 G. A. Snow, Bacteriol. Rev., 1970, 34, 99–125.
3 C. L. Rateledge, Iron Metabolism, in Mycobacteria: Molecular Biology
and Virulence, ed. C. L. Ratledge and J. Dale, Blackwell Science, Oxford,
UK, 1999, pp. 206–286.
45, 1317–1322.
28 Y. Sumita, M. Inoue and S. Mitsuhashi, Eur. J. Clin. Microbiol. Infect.
Dis., 1998, 8, 908–916.
29 W. Zimmermann, Antimicrob. Agents Chemother., 1980, 18, 94–100.
30 M. H. Richmond, D. C. Clark and S. Wotton, Antimicrob. Agents
Chemother., 1976, 10, 215–218.
1628 | Org. Biomol. Chem., 2007, 5, 1621–1628
This journal is
The Royal Society of Chemistry 2007
©