Total synthesis of mycobactin analogues as potent antimycobacterial agents using a minimal protecting group strategy

Article abstract of DOI:10.1021/jo980063o

Mycobactins are a family of iron sequestering agents (siderophores) biosynthesized as growth promoters by mycobacteria including Mycobacterium tuberculosis. They are important siderophors with high affinity and specificity for Fe(III) due to the chemical nature of their component chelating functional groups. The parent compounds and their synthetic analogues can be used for studies of natural iron uptake mechanisms. It was hypothesized by Snow and co-workers that alternate and modified mycobactin analogues might serve as antagonists of mycobacterial growth and be of important therapeutic value. Efficient syntheses of four different analogues are presented. Dramatic improvements on formation of amide and ester bonds were achieved using water soluble carbodiimide (EDC-HCl)-mediated couplings in the presence of 1-hydroxy-7-azabenzotriazole (HOAt) as an additive. Using HOAt over other traditional coupling additives provided significant enhancement of the reaction rate of the desired coupling reactions and minimized side reactions. Further simplifications were made possible by minimizing the use of protecting groups during the syntheses. In fact, coupling components in the presence of free hydroxamic acids and a free phenolic hydroxyl group proceeded in excellent yields. Biological studies indicated that the resulting synthetic analogues effect moderate to high inhibition of the growth of M. tuberculosis H37Rv.

Full text of DOI:10.1021/jo980063o

4314
J . Org. Chem. 1998, 63, 4314-4322
Tota l Syn th eses of Mycoba ctin An a logu es a s P oten t
An tim ycoba cter ia l Agen ts Usin g a Min im a l P r otectin g Gr ou p
Str a tegy
Yanping Xu and Marvin J . Miller*
Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, Indiana 46556
Received J anuary 13, 1998
Mycobactins are a family of iron sequestering agents (siderophores) biosynthesized as growth
promoters by mycobacteria including Mycobacterium tuberculosis. They are important siderophores
with high affinity and specificity for Fe(III) due to the chemical nature of their component chelating
functional groups. The parent compounds and their synthetic analogues can be used for studies of
natural iron uptake mechanisms. It was hypothesized by Snow and co-workers that alternate and
modified mycobactin analogues might serve as antagonists of mycobacterial growth and be of
important therapeutic value. Efficient syntheses of four different analogues are presented.
Dramatic improvements on formation of amide and ester bonds were achieved using water soluble
carbodiimide (EDC‚HCl)-mediated couplings in the presence of 1-hydroxy-7-azabenzotriazole (HOAt)
as an additive. Using HOAt over other traditional coupling additives provided significant
enhancement of the reaction rate of the desired coupling reactions and minimized side reactions.
Further simplifications were made possible by minimizing the use of protecting groups during the
syntheses. In fact, coupling components in the presence of free hydroxamic acids and a free phenolic
hydroxyl group proceeded in excellent yields. Biological studies indicated that the resulting
synthetic analogues effect moderate to high inhibition of the growth of M. tuberculosis H37Rv.
In tr od u ction
agents in general. However, the need for safe, microbi-
ally selective, and effective drugs is apparent, and
certainly no ideal agents have yet been developed.
Continued effort must be made to promote the discovery
of new antimicrobial agents, and studies of their modes
of action are anticipated to facilitate the rational design
of new drugs. One important aspect of natural drug
resistance in TB strains is thought to be related to
impermeability of the mycobacterial cell wall, which
contains large amounts of lipids with an unusual struc-
ture. Like most mycobacteria, cells of Mycobacterium
tuberculosis are covered by a lipid-rich cell wall, which
consists of a large amount of mycolic acids with long
hydrocarbon chains. The length and high degree of
saturation of the chains produce an exceptionally tightly
packed array with extremely low fluidity. Consequently,
penetration of antibiotics and chemotherapeutic agents
into or through the cell wall is limited.7 Design and
synthesis of an efficient drug carrier which facilitates
drug delivery has become a major challenge in drug
development.
Among bacterial infections, tuberculosis is usually a
respiratory infection, though it can damage virtually any
organ in the human body. About 40% of the world’s
population is infected with the tuberculosis bacterium,
which grows slowly and in most cases is indefinitely
suppressed by the immune system. According to a new
estimate by the World Health Organization, the disease
killed more people in 1995, when effective treatment was
available, than in 1900, when no antituberculosis drugs
existed.1 Early this decade, the incidence of tuberculosis
began rising after a 33-year downward trend. In re-
sponse, the federal government greatly increased spend-
ing on tuberculosis control, from about $9 million in 1986
to nearly $120 million in 1996.1 Streptomycin, the first
antibiotic capable of killing the tuberculosis organism,
was discovered in 1944.2 Many other antituberculosis
agents were developed subsequently.3 However, the
recent emergence of drug resistant strains of tuberculosis
is of special concern.4 The World Health Organization
recently reported that the death rate of patients with
multidrug resistant tuberculosis in the U.S. was ap-
proximately 70%.5 Between 1989 and 1992 such drug
resistant strains had appeared in 17 states.6
Microbial iron chelators, or siderophores, are important
in the study of iron metabolism8 and in the development
of drugs for the treatment of iron-overloaded patients.9
The high affinity and specificity of siderophores for Fe(III)
is due to the chemical nature of the chelating functional
Over the past decade, significant advances have been
made in the discovery and development of antimicrobial
(1) Brown, D. U.S. Tuberculosis Cases Decline as Disease Increases
Worldwide. The Washington Post, March 23, 1996.
(2) Schatz, A.; Bugie, E.; Waksman, S. A. Proc. Soc. Exp. Biol. Med.
1944, 55, 66.
(3) Lemke, T. L. In Principles of Medicinal Chemistry, 4th ed.; Foye,
W. O., Lemke, T. L., Williams, D. A., Eds.; Williams & Wilkins:
Baltimore, 1995; pp 747-758.
(4) Lemonick, M. D. The Killer All Around. Time Magazine; Lem-
onick, M. D., Ed.; Time, Inc.: New York, Sept 12, 1994; p 62.
(5) Facts and Figures: Tuberculosis Today, World Health Organi-
zation’s 1995 report on tuberculosis.
(6) (a) DiGiacomo, M. A Deadly Return. Newsweek, March 16, 1992,
p 53. (b) Tuberculosis and HIV Public Health Policy: A Dual Challenge.
Washington: AIDS Action Foundation, March 1992.
(7) (a) Connell, N. D.; Nikaido, H. Membrane Permeability and
Transport in Mycobacterium Tuberculosis. In Tuberculosis: Patho-
genesis, Protection, and Control; Bloom, B. R., Ed.; ASM Press:
Washington, DC, 1994; p 333. (b) Liu, J .; Rosneberg, E. Y.; Nikaido,
H. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 11254. (c) Besra, G. S.;
Chatterjee, D. Lipid and Carbohydrates of Mycobacterium Tubercu-
losis. In Tuberculosis: Pathogenesis, Protection, and Control; Bloom,
B. R., Ed.; ASM Press: Washington, DC, 1994; p 285.
S0022-3263(98)00063-2 CCC: $15.00 © 1998 American Chemical Society
Published on Web 06/05/1998

Total Syntheses of Mycobactin Analogs
J . Org. Chem., Vol. 63, No. 13, 1998 4315
F igu r e 1.
groups, which usually consist of either catecholate or
hydroxamic acid residues. The mycobactins, 1, are
among the most structurally complex of the known
siderophores. They promote mycobacterial growth via
iron uptake processes. This family of compounds was
first isolated and characterized mainly by Snow and co-
workers.10 All members of the family of compounds
possess a nearly identical main carbon skeleton with
variations only in stereochemistry of chiral centers and
in the peripheral groups. All the mycobactins form
extremely stable hexadentate iron(III) complexes by
binding the iron with two hydroxamic acids and a
2-hydroxyphenyloxazoline residue. After detailed stud-
ies, Snow10 suggested that alternate or modified forms
of mycobactins might serve as antagonists of mycobac-
terial growth by competitively binding iron and/or inhib-
iting its assimilation by targeted forms of mycobacteria
and thus might be of important therapeutic value.
Additionally, design and synthesis of mycobactin ana-
logues, which contain potential “drug linkers”, may
provide a new class of therapeutic agents. Direct attach-
ment of antimicrobial agents to the analogues may allow
the pendant drug to be actively carried into cells by the
iron transport system or at least be targeted to myco-
bacteria. Herein, we focus on the design of mycobactin
analogues to further test the feasibility of using modified
siderophores to antagonize the growth of Mycobacterium
tuberculosis.
analogues revealed no growth inhibitory or stimulatory
activity of mycobacteria, thus confirming a need for iron
complexation. The first total synthesis of a mycobactin
analogue, mycobactin S2, which contains all iron-chelat-
ing components but lacks a long lipophilic side chain (1,
R1, R4 ) CH3; R2, R3, R5 ) H, Figure 1), was accomplished
in our group.12 Biological studies revealed that the short
alkyl group (R1) resulted in loss of lipid solubility and
siderophore activity for mycobacteria. This indicated
that a long lipophilic side chain is essential for mycobac-
tins to transport iron through cell membranes and,
therefore, to be potential drug delivery vehicles. A more
recent synthesis of a mycobactin that incorporated a long
acyl group (1, R1 ) CH3(CH2)14; R4 ) CH3; R2, R3, R5 )
H, Figure 1), mycobactin S, was accomplished by our
group.13 Interestingly, biological tests showed that syn-
thetic mycobactin S, a natural growth promoter for M.
segmatis, effects greater than 99% inhibition of the
growth M. tuberculosis H37Rv at the concentration of
12.5 µg/mL. Subsequently, the minimum inhibitory
concentration (MIC) was determined to be 3.13 µg/mL.
Compared to mycobactin T, the siderophore and growth
promoter for M. tuberculosis, mycobactin S differs only
by the stereochemistry of the methyl group at R.4 This
strongly indicates that the single chiral center plays a
critical role in controlling TB growth. Further biological
studies showed that several mycobactin components
display moderate growth promotion of mycobacteria.14
Therefore, design and synthesis of novel mycobactin
analogues would allow us not only to discover effective
inhibitors of M. tuberculosis but also to further test
Snow’s hypothesis that use of modified siderophore
structures may lead to the development of selective
mycobacterial growth inhibitors.10 The apparent com-
plexity of the structures of the mycobactins makes
A mycobactin analogue lacking the three hydroxyl
groups necessary for iron chelation was synthesized by
Carpenter and Moore in 1969.11 Biological tests of these
(8) (a) Neilands, J . B. In Microbial Iron Metabolism. A Comprehen-
sive Treatise; Neilands, J . B., Ed.; Academic Press: New York, 1974;
p 3. (b) Emery, T. Met. Ions Biol. Syst. 1978, Chapter 3, 7. (c) Emery,
T. Biochemistry 1971, 10, 1483.
(9) (a) Anderson, W. F. In Inorganic Chemistry in Biology and
Medicine; Martell, E., Ed.; American Chemical Society: Washington,
DC, 1980; Chapter 15. (b) Weatherall, D. J . In Development of Iron
Chelators for Clinical Use; Martell, A. E., Anderson, W. F., Badman,
D. G., Eds.; Elsevier/North-Holland; New York, 1981; p 3. (c) Waxman,
H. S.; Brown, E. B. Prog. Hematol. 1969, 6, 338.
(11) Carpenter, J . G.; Moore, J . W. J . Chem. Soc. C 1969, 12, 1610.
(12) Maurer, P. J .; Miller, M. J . J . Am. Chem. Soc. 1983, 105, 240.
(13) Hu, J .; Miller, M. J . J . Am. Chem. Soc. 1997, 119, 3462.
(14) Hu, J . Ph.D. Dissertation. Total Synthesis of Mycobactins,
Siderophore-Drug Conjugates, and Biological Studies. Department of
Chemistry and Biochemistry, University of Notre Dame, 1996.
(10) Snow, G. A. Bacteriol. Rev. 1970, 34, 99.

4316 J . Org. Chem., Vol. 63, No. 13, 1998
Xu and Miller
Sch em e 1. Retr osyn th etic An a lysis of Mycoba ctin
An a logs
F igu r e 2.
of lysine derivative 9 can be achieved by dimethyldiox-
irane (DMD) oxidation of the ꢀ-amino group of the
appropriate R-protected lysine ester. Oxazoline deriva-
tive 8 can be prepared by using our previously described
methodology.13 Cyclolysine 11 also can be derived from
lysine by direct DMD oxidation followed by intramolecu-
lar amidation.
Resu lts a n d Discu ssion
Our syntheses started by preparation of the four
different cobactin constituents (7a -d , Figure 2) of the
target mycobactin analogues 2-5. Treatment of cyclic
hydroxamate 1213 with 10 wt % Pd on carbon in methanol
afforded free amine 13 in quantitative yield. Using the
reaction conditions we reported earlier for reaction of 13
with â-hydroxybutyrate,13 cross-coupling of amine 13
with Cbz-protected threonine and serine using DCC
(dicyclohexylcarbodiimide)/DMAP‚HCl (4-(dimethylami-
no)pyridine hydrochloride salt) gave only moderate yields
of desired products, 14 and 15, respectively (Scheme 2).
As expected, self-coupling of the Cbz-protected â-hydroxy
amino acids produced dimeric and oligomeric esters as
the major byproducts. A number of different coupling
additives were studied in attempts to circumvent the
competitive esterification. 1-Hydroxy-7-azabenzotriazole
(HOAt), first reported by Carpino in 1993,15 was remark-
ably effective. HOAt was found to accelerate the desired
amide coupling processes and provide a visual indication
of the reaction endpoint. Although precise mechanistic
details need yet to be established, reactions of HOAt
esters appear to be enhanced relative to those of the
commonly used 1-hydroxybenzotriazole (HOBt) ana-
logues. The pronounced reactivity increase observed
upon substitution of HOAt for HOBt presumably is due
to assistance from internal hydrogen bonding of the
incoming nucleophiles by the nitrogen atom of pyridine
residue in the reagent.15 Coupling of amine 13 with Cbz-
protected threonine in the presence of HOAt afforded
amide 14 in 88% yield without any detectable competitive
ester formation (Scheme 2). Under similar reaction
conditions, serine-based cobactin analogue 15 was formed
in 90% yield.
preparation of appropriate analogues an apparently
daunting task. However, as described here, substantial
simplification of the syntheses will allow preparation of
a number of important mycobactins and derivatives in
quantities sufficient for extensive study. On the basis
of dramatic differences in the biological activities of our
previous synthetic mycobactin S and mycobactin T, our
focus in the design of 2-5 was to determine the impor-
tance of the linkage (X) between the constituent iron-
binding components and stereochemistry of the Y group
(Scheme 1). Additionally, introduction of protected linker
functionality (Z ) NHR/R ) Cbz or Boc) will facilitate
future syntheses of drug conjugates. Efficient syntheses
of four novel analogues related to 2-5 (Scheme 1) of
mycobactins and biological studies demonstrating their
antimycobacterial activities are presented.
The synthesis of the mycobactic acid component com-
menced with preparation of hydroxamate intermediate
19 (Scheme 3). Treatment of hydroxylamine 1613 with
excess palmitoyl chloride in the presence of NaHCO3
resulted in N- and O-acylation of the hydroxylamine
moiety. Subsequent removal of the O-palmitoyl hydrox-
amate group afforded hydroxamic acid 17. Reaction of
17 with SEMCl provided O-protected hydroxamate 18 in
As shown retrosynthetically in Scheme 1, the myco-
bactins can be constructed by coupling of the mycobactic
acid portion, 6, and the cobactin portion, 7. Mycobactic
acid 6 can be generated from coupling oxazoline 8 and
amino acid 9. The cobactins, 7, can be easily produced
by the coupling of amine 11 with acid 10. Construction
(15) (a) Carpino, L. A. J . Am. Chem. Soc. 1993, 115, 4397. (b)
Carpino, L. A. J . Org. Chem. 1995, 60, 3561. (c) Musiol, H.-J . J . Org.
Chem. 1994, 59, 6144. (d) Angell, Y. M. J . Am. Chem. Soc. 1995, 117,
7279. (e) Carpino, L. A. Tetrahedron Lett. 1994, 35, 2279. (f) Carpino,
L. A.; EI-Faham, A. J . Org. Chem. 1994, 59, 695. (g) Carpino, L. A. J .
Chem. Soc., Chem. Commun. 1994, 201.

Total Syntheses of Mycobactin Analogs
J . Org. Chem., Vol. 63, No. 13, 1998 4317
Sch em e 2
internal ester linkage. To overcome this problem a
different synthetic approach was executed. As our earlier
results indicated, effective acidity of the reaction solution
has a major influence on the outcome of coupling reac-
tions.14 On the basis of pKa values, coupling reactions
can be carried out under finely tuned conditions without
masking certain functional groups. Our earlier studies
showed that in the presence of 4-(dimethylamino)pyridine
(DMAP) and an equal molar amount of its hydrogen
chloride salt no desired coupling product (19) was ob-
tained by treatment of oxazoline 8 with amine 22 (eq 1).
The inefficiency of this reaction was attributed to com-
petitive chemistry associated with partial ionization of
the phenolic group of 8. Alternative use of EDC‚HCl and
no added base produced mycobactin analogue 19 in 65%
yield, emphasizing the importance of “acidic” coupling
conditions.13
Sch em e 3
Interestingly, phenol and hydroxamic acids have amaz-
ingly close pKa values of ca. 9. Careful control of reaction
acidity was anticipated to avoid ionization of the phenol
and hydroxamate, thereby minimizing complications due
to their competitive nucleophilicities. To test this hy-
pothesis, oxazoline derivative 8 was treated with EDC‚
HCl/HOAt and amine 23, which contains a free hydrox-
amic acid, and indeed, the desired coupled product 24a
was formed in 91% yield (Scheme 4). Under similar
conditions, free cobactin analogue 7a was prepared in
95% yield by coupling Cbz-protected L-serine and unpro-
tected cyclolysine 25. Gratifyingly, under such conditions
both reactions proceeded very cleanly without any detect-
able byproducts. These results provided opportunities
to further improve our syntheses.
96% overall yield from 16. Desired hydroxamate 19 was
produced by hydrogenation of 18 followed by EDC (1-
ethyl-3-(3-(dimethylamino)propyl)carbodiimide hydro-
chloride)-mediated coupling with oxazoline derivative 8.13
Saponification of 19 yielded the corresponding acid 20
which was subsequently coupled with modified cobactins
14 and 15 to give desired protected mycobactin analogues
21a and 21b in 85% and 91% yields, respectively.
Unfortunately, all attempts to remove the protecting
groups from 21a and 21b only resulted in total decom-
position. This was probably due to instability of the
Thus, Cbz-protected serine was again treated with free
hydroxamic acid 25 in the presence of EDC‚HCl and
HOAt. Upon 100% consumption of starting materials,
and without any further purification of 7a , mycobactic
acid 24b with another equivalent of EDC‚HCl was
introduced to the reaction mixture. The ensuing reaction

4318 J . Org. Chem., Vol. 63, No. 13, 1998
Xu and Miller
Sch em e 4
Sch em e 6
Sch em e 5
Sch em e 7
provided desired mycobactin analogue 26a in 81% overall
yield (Scheme 5). This modification not only significantly
simplified our synthetic approach by circumventing
unproductive protection and deprotection steps but also
improved the overall chemical yield. Using similar
methods, threonine analogue 26b was produced in 75%
overall yield by starting with Cbz-protected threonine
(Scheme 5).
We then turned our attention to analogues with amide
linkages. The synthesis of â-alanine analogue 4 com-
menced with preparation of modified analogue 28. Amine
13 was coupled with Cbz-protected â-alanine in the
presence of EDC‚HCl and HOAt to give 27 in 91% yield.
Reductive removal of the Cbz protecting group, followed
by the now usual EDC‚HCl/HOAt-mediated coupling
with 24b, afforded desired product 29. Subsequent
treatment of 29 with tetrabutylammonium fluoride in
acetic acid gave mycobactin analogue 4 in 85% yield
(Scheme 6).
Synthesis of 35b, a form of mycobactin analogue 5,
started with preparation of 2,3-diaminopropionic acid
derivative 32 (Scheme 7). Treatment of Boc-protected
serine with O-benzylhydroxylamine afforded hydrox-
amate 30 in 82% yield.16 Intramolecular cyclization in
the presence of triphenylphosphine and carbon tetra-
(16) Miller, M. J .; Mattingly, P. G.; Morrison, M. A.; Kerwin, J . F.,
J r. J . Am. Chem. Soc. 1980, 102, 7026.

Total Syntheses of Mycobactin Analogs
J . Org. Chem., Vol. 63, No. 13, 1998 4319
Ta ble 1. Biologica l Activities a ga in st M. Tu ber cu losis
(0.05 mmol), 1.0 equiv of HOAt (6.8 mg, 0.05 mmol), and 7
mg of DMAP (0.03 mmol) in 0.5 mL of dry DMF was added a
premixed solution of 0.055 mmol of EDC‚HCl and 0.055 mmol
of DMAP in 0.5 mL of dry DMF dropwise. The resulting
mixture was allowed to stand at room temperature for 3 h.
The reaction mixture was then diluted with 15 mL of EtOAc,
washed with water (1 × 5 mL) and brine (2 × 5 mL), dried
over Na2SO4, filtered, and concentrated. The residue was
purified by silica chromatography with 1:1 ethyl acetate/
hexanes as the eluent to afford 27.3 mg (88%) of 14 as a
colorless oil: Rf ) 0.27 (1:1 EtOAc/hexanes); IR (neat) 3315,
2950, 2865, 1735, 1530, 1456 cm-1; 1H NMR (300 MHz, CDCl3)
δ 7.74-7.70 (m, 5H), 7.46-7.33 (m, 10H), 7.04 (bd, 1H), 5.63
(bd, 1H), 5.11 (s, 2H), 4.28-4.06 (m, 3H), 3.58-3.41 (m, 2H),
3.30 (bs, 1H), 1.30-1.98 (m, 6H), 1.14 (s, 9H), 1.09 (d, 3H, J )
6.3 Hz); 13C NMR (125 MHz, CDCl3) δ 169.71, 169.38, 156.30,
136.15, 136.06, 132.01, 131.52, 130.27, 130.24, 128.52, 128.15,
128.04, 127.61, 127.53, 76.51, 67.51, 67.08, 58.51, 54.16, 30.53,
H37Rv
compound
% of inhibition
MIC (µg/mL) vs H37Rv
26a (2, R ) Cbz)
26b (3, R ) Cbz)
4
44
48
25
98
>12.5
>12.5
>12.5
<0.2
35b (5, R ) Boc)
chloride gave â-lactam 31 in 98% yield.16 Subsequent
saponification of 31 produced the desired 2,3-diamino-
propionic acid derivative 32 in quantitative yield. Co-
bactin analogue 33 was easily prepared in 91% yield by
EDC-mediated coupling of acid 32 and cyclolysine 13.
Conversion of the benzyl-protected hydroxylamine moiety
into the corresponding free amine was accomplished by
simple palladium-catalyzed hydrogenation to give 34 in
92% yield by reduction of the hydroxylamine N-O bond
in the presence of the less easily reduced hydroxamate
N-O bond.17 Subsequent coupling with unprotected
mycobactic acid 24b, followed by removal of the silyl
protecting group on the cobactin component, provided
mycobactin analogue 35b in 85% yield (Scheme 7).
27.26, 26.89, 25.29, 19.57, 18.14; HRFABMS calcd for C34H44
N3O6Si 618.2999, found 618.3025.
-
TBDP S-P r otected Coba ctin 15. To a solution of 17 mg
of amine 13 (0.045 mmol), 9.6 mg of Cbz-protected serine (0.04
mmol), 1.0 equiv of HOAt (5.4 mg, 0.04 mmol), and 5.5 mg of
DMAP (0.045 mmol) in 0.5 mL of dry DMF was added a
premixed solution of 0.045 mmol of EDC‚HCl and 0.045 mmol
of DMAP (5.5 mg) in 0.5 mL of dry DMF dropwise. The
resulting mixture was allowed to stand at the room temper-
ature for 3 h. The reaction mixture was then diluted with 10
mL of EtOAc, washed with water (1 × 5 mL) and brine (2 ×
5 mL), dried over Na2SO4, filtered, and concentrated. The
residue was purified by silica chromatography with 3:1 ethyl
acetate/hexanes as the eluent to afford 21.6 mg (90%) of 15 as
a colorless oil: Rf ) 0.40 (3:1 EtOAc/hexanes); IR (neat) 3315,
2940, 2862, 1725, 1535, 1455 cm-1; 1H NMR (300 MHz, CDCl3)
δ 7.72 (m, 5H), 7.47-7.30 (m, 10H), 6.83 (bd, 1H), 5.65 (bd,
1H), 5.11 (s, 2H), 4.30 (m, 2H), 4.20 (m, 1H), 3.89 (dd, 1H, J )
6.6, 2.2 Hz), 3.56-3.51 (m, 1H), 3.49-3.42 (m, 1H), 1.81-1.74
(m, 3H), 1.59-1.51 (m, 3H), 1.45-1.41 (m, 1H), 1.13 (s, 9H);
13C NMR (125 MHz, CDCl3) δ 169.82, 169.21, 155.92, 136.74,
135.40, 131.96, 131.51, 129.59, 127.90, 127.50, 126.96, 76.40,
67.09, 65.89, 64.57, 63.44, 54.96, 28.35, 27.56, 26.32, 19.56;
HRFABMS calcd for C33H42N3O6Si 604.2843, found 604.2823.
Biologica l Stu d ies
The four synthetic analogues (26a , 26b, 4, 35b) with
generalized structures 2-5 were submitted to the Tu-
berculosis Antimicrobial Acquisition and Coordinating
Facility of the Southern Research Institute for biological
studies with M. tuberculosis. All four types of analogues
were found to effect from moderate to high inhibitory
activity against M. tuberculosis H37Rv at a concentration
of 12.5 µg/mL (Table 1). Interestingly, the most active
analogue was 35b (MIC < 0.2 µg/mL!), the Boc-protected
version of 5. Thus, it appears that the enhanced activity
of 5 relative to 4 must be attributed to the additional
NHBoc group of 5, suggesting importance of either the
additional bulky functional group or additional hydro-
phobicity. The study of the importance of the absolute
stereochemistry on the additional NHBoc group of 5 is
in progress along with additional variation of X, Y, and
Z in the generalized structures of 2-5 with special
attention to their use as potential drug linkers.
SEM-P r otected Hyd r oxa m a te 18. To a solution of 78 mg
of 16 (0.25 mmol) and 169 mg of NaHCO3 (2.01 mmol) in 8
mL of dry CH2Cl2 was added 4 equiv of palmitoyl chloride (0.30
mL) dropwise. The resulting solution was then allowed to
stand at room temperature for 14 h. The mixture was diluted
with 20 mL of EtOAc, washed with 1 N HCl (3 × 5 mL) and
brine (3 × 5 mL), dried over Na2SO4, filtered, and concentrated.
The residue was then taken up in 25 mL of 6% Hu¨nig’s base
in methanol. The resulting mixture was then allowed to stand
at room temperature for 12 h. The mixture was concentrated.
The residue was transferred to a solution of 0.87 mL of Hu¨nig’s
base (5.0 mmol) and 20 mg of DMAP in 5 mL of toluene,
followed by addition of 10 equiv of SEMCl (0.44 mL, 2.5 mmol)
dropwise. The resulting solution was then allowed to stand
at 65 °C overnight under argon. The mixture was cooled to
room temperature and diluted with 20 mL of EtOAc. Then
the mixture was washed with 1 N HCl (2 × 10 mL), saturated
aqueous NaHCO3 (2 × 10 mL) and brine (2 × 10 mL), dried
over Na2SO4, filtered, and concentrated. The residue was
purified by silica chromatography with 1:15 ethyl acetate/
hexanes as the eluent to afford 162 mg (96%) of 18 as a
colorless oil: Rf ) 0.21 (1:15 EtOAc/hexanes); IR (neat) 3312,
Con clu sion
The total syntheses of four different mycobactin ana-
logues described here provides efficient methods for
preparing this family of compounds. Using a water
soluble carbodiimide (EDC‚HCl) in the presence of 1-hy-
droxy-7-azabenzotriazole (HOAt) under acidic conditions
proved to be an effective coupling process that allowed
minimal use of peripheral protecting groups. Further
simplification of the syntheses of analogues 2 and 3 was
accomplished by applying a tandem coupling strategy.
The biological studies of these synthetic analogues show
from moderate to excellent inhibitorial activities against
M. tuberculosis, further verifying Snow’s assumption that
alternate or modified mycobactin analogues might serve
as antagonists of mycobacterial growth. Synthetic stud-
ies of different drug conjugates and structural analogues
are under investigation.
2910, 2822, 1722, 1650 cm-1 1H NMR (300 MHz, CDCl3) δ
;
7.36-7.31(m, 5H), 5.36 (d, J ) 7.8 Hz, 1H), 5.10 (s, 2H), 4.90
(s, 2H), 4.37-4.30 (m, 1H), 3.76-3.61 (m, 7H), 2.38 (t, J ) 7.6
Hz, 2H), 1.86-1.03 (m, 32 H), 0.99-0.93 (m, 2H), 0.88 (t, J )
6.7 Hz, 3H), 0.03 (s, 9H); 13C NMR (125 MHz, CDCl3) δ 175.36,
172.69, 155.82, 136.18, 128.29, 127.89, 127.85, 99.08, 67.38,
66.68, 53.64, 52.09, 47.16, 32.43, 31.75, 29.52, 29.49, 29.38,
29.30, 29.27, 29.19, 26.10, 24.33, 22.52, 22.14, 18.02, 13.96,
-1.63; HRFABMS calcd for C37H67N2O7Si 679.4717, found
679.4702.
Exp er im en ta l Section
TBDP S-P r otected Coba ctin 14. To a solution of 21 mg
of amine 13 (0.055 mmol),10 12.5 mg of Cbz-protected threonine
(17) Arai H.; Hagmann, W. K.; Suguna, H.; Hecht, S. M. J . Am.
Chem. Soc. 1980, 102, 6631.

4320 J . Org. Chem., Vol. 63, No. 13, 1998
Xu and Miller
Oxa zolin e Hyd r oxa m a te 19. A solution of 670 mg of 18
(0.99 mmol) in 25 mL of MeOH was purged with argon for 15
min. Then 70 mg of 10% Pd-C was added, and the system
was flushed with argon for another 10 min. The resulting
slurry was stirred under H2 (1 atm) for 3 h. After the catalyst
was filtered off through a pad of Celite, the solvent was
removed under vacuum to provide the corresponding free
amine. The amine was then treated with 202 mg of oxazoline
derivative 8 (0.98 mmol) in the presence of 187 mg of EDC‚HCl
(0.98 mmol), 133 mg of HOAt (0.98 mmol), and 25 mg of DMAP
in 30 mL of dry DMF. The resulting solution was allowed to
stand at room temperature for 3 h. The mixture was diluted
with 60 mL of EtOAc. The organic layer was washed with 1
N HCl (2 × 10 mL), saturated aqueous NaHCO3 (2 × 10 mL),
and brine (2 × 10 mL), dried over NaSO4, filtered, and
concentrated. The residue was purified by silica chromatog-
raphy with 1:1 ethyl acetate/hexanes as the eluent to afford
660 mg (91%) of 19 as a colorless oil: Rf ) 0.40 (1:1 EtOAc/
5 mL), dried over Na2SO4, filtered, and concentrated. The
residue was purified by silica chromatography with 3:1 ethyl
acetate/hexanes as the eluent to afford 83.3 mg (91%) of 21b
as a colorless oil: IR (neat) 3430, 2975, 2940, 1740, 1730, 1655
cm-1 1H NMR (500 MHz, CDCl3) δ 11.31 (bs, 1H), 7.68 (d,
;
1H, J ) 8.0 Hz), 7.54 (dd, 1H, J ) 12.5, 6.5 Hz), 7.42-7.29
(m, 5H), 7.24-7.22 (m, 1H), 7.16 (bd, 1H), 7.00 (d, 1H, J ) 8.5
Hz), 6.89 (dt, 1H, J ) 7.5, 2.5 Hz), 5.78 (m, 1H), 5.12, 5.09
(ABq, 2H, J ) 16.0 Hz), 4.96 (m, 1H), 4.92 (s, 1H), 4.87 (s,
1H), 4.65-4.55 (m, 4H), 4.28-4.25 (m, 1H), 4.22 (dd, 1H, J )
16.5, 6.5 Hz), 3.96-3.81 (m, 1H), 3.76-3.64 (m, 4H), 3.57-
3.54 (m, 1H), 2.41 (t, 1H, J ) 7.5 Hz), 2.37 (t, 1H, J ) 7.5 Hz),
2.03-1.23 (m, 38H), 1.13 (dd, 2H, J ) 12.0, 6.5 Hz), 0.98-
0.92 (m, 2H), 0.87 (t, 3H, J ) 6.5 Hz), 0.02 (s, 9H); 13C NMR
(125 MHz, CDCl3) δ 175.78, 170.88, 170.73, 167.88, 167.62,
166.84, 159.48, 155.55, 136.07, 134.05, 128.57, 128.52, 128.20,
128.15, 128.08, 119.09, 116.93, 110.32, 99.21, 69.26, 67.66,
67.15, 64.54, 53.44, 52.78, 52.35, 51.24, 50.75, 32.62, 31.91,
31.89, 31.23, 30.87, 29.66, 29.56, 29.48, 29.45, 29.35, 29.33,
27.60, 26.12, 25.70, 25.54, 24.61, 22.68, 22.65, 22.59, 22.36,
hexanes); IR (neat) 3290, 2910, 2820, 1740, 1650, 1630 cm-1
;
1H NMR (300 MHz, CDCl3) δ 7.67 (dd, J ) 7.9, 1.7 Hz, 1H),
7.39 (ddd, J ) 8.7, 7.5, 1.7 Hz, 1H), 7.00 (dd, J ) 8.5, 0.8 Hz,
1H), 6.93-6.85 (m, 1H), 4.94 (t, J ) 9.6 Hz, 1H), 4.84 (s, 2H),
4.65-4.61 (m, 2H), 4.57-4.50 (m, 1H), 3.73 (s, 3H), 3.72-3.66
(m, 2H), 3.60 (t, J ) 7.1 Hz, 2H), 2.33 (d, J ) 7.6 Hz, 2H),
1.91-1.52 (m, 6H), 1.28-1.16 (m, 26H), 0.95-0.89 (m, 2H),
0.83 (t, J ) 6.8 Hz, 3H), -0.03 (s, 9H); 13C NMR (75 MHz,
CDCl3) δ 175.45, 172.09, 170.27, 167.68, 159.72, 134.13, 128.47,
118.95, 116.88, 109.94, 99.13, 69.40, 67.97, 67.47, 52.37, 52.04,
47.29, 32.50, 31.82, 31.58, 29.59, 29.56, 29.44, 29.37, 29.26,
26.18, 24.42, 22.59, 22.35, 18.08, 14.02, -1.56; HRFABMS
calcd for C39H68N3O8Si 734.4775, found 734.4764.
18.18, 14.12, -1.47; HRFABMS calcd for
1067.6100, found 1067.6104.
C55H86N6O13Si
Mycoba ctic Meth yl Ester 24a . A slurry of 880 mg of
protected amine 17 (1.61 mmol) with 200 mg of 10% Pd-C in
20 mL of MeOH was purged with an N2 stream for 10 min
and subsequently with H2 for an additional 15 min. The
reaction mixture was stirred under H2 (1 atm) at room
temperature for 3 h. Then the palladium catalyst was filtered
off through a pad of Celite. The solution was concentrated to
yield the corresponding free amine 23. Without further
purification, the crude amine 23 was dissolved in 20 mL of a
freshly distilled DMF solution containing 330 mg of oxazoline
8 (1.59 mmol) and 217 mg of HOAt (1.59 mmol). To above
solution was added 336 mg of EDC‚HCl (1.75 mmol). The
resulting solution was allowed to stand at room temperature
for 3 h before it is diluted with 60 mL of EtOAc and 10 mL of
1 N HCl. The organic layer was washed with 0.5 N HCl (2 ×
10 mL), 5% ascorbic acid (1 × 10 mL), saturated aqueous
NaHCO3 (1 × 10 mL) and brine (2 × 10 mL), dried over
Na2SO4, filtered, and concentrated. The residue was purified
silica chromatography with 1:1 ethyl acetate/dichloromethane
as the eluent to afford 872 mg (91%) of 24a as a white solid:
P r otected Mycoba ctin 21a . To a solution of 66 mg of
amine 13 (0.173 mmol), 41 mg of Cbz-protected threonine
(0.173 mmol), 1.2 equiv of HOAt (21 mg, 0.210 mmol), and 21
mg of DMAP (0.173 mmol) in 1.0 mL of dry DMF was added
a premixed solution of 0.21 mmol of EDC (40 mg) and 0.173
mmol of DMAP (21 mg) in 0.5 mL of dry DMF dropwise. After
being allowed to stand at room temperature for 3 h, the
mixture was charged with 119 mg of acid 20 (0.165 mmol) and
25 mg of EDC‚HCl (0.13 mmol). The resulting solution was
heated at 55 °C for 42 h before it was cooled to room
temperature. The reaction mixture was then diluted with 10
mL of EtOAc, washed with water (3 × 5 mL) and brine (3 ×
5 mL), dried over Na2SO4, filtered, and concentrated. The
residue was purified by silica chromatography with 3:1 ethyl
acetate/hexanes as the solvent system to afford 151.5 mg (85%)
of 21a as a colorless oil: IR (neat) 3405, 2960, 2925, 1790,
mp 81-82.5 °C; IR (neat) 3230, 2920, 2850, 1740, 1650 cm-1
;
1H NMR (300 MHz, CDCl3/CD3OD) δ 7.69 (dd, J ) 8.0, 1.5
Hz, 1H), 7.41 (dt, J ) 8.0, 1.5 Hz, 1H), 6.96 (d, J ) 8.1 Hz,
1H), 6.91 (t, J ) 7.4 Hz, 1H), 5.02 (t, J ) 9.0 Hz, 1H), 4.65-
4.60 (m, 2H), 4.43 (dd, J ) 8.8, 4.9 Hz, 1H), 3.73 (s, 3H), 3.62-
3.55 (m, 2H), 2.43 (t, J ) 7.6 Hz, 2H), 1.95-1.26 (m, 32 H),
0.91 (t, J ) 6.6 Hz, 3H); 13C NMR (125 MHz, CDCl3) δ 176.18,
173.87, 172.95, 168.55, 161.02, 135.02, 129.50, 119.97, 117.70,
111.45, 70.43, 69.20, 53.85, 52.80, 33.26, 33.06, 31.82, 30.75,
30.62, 30.49, 30.46, 27.10, 25.95, 23.80, 23.72, 14.45; HR-
FABMS calcd for C33H53N3O7 604.3961, found 604.3975.
Coba ctin 7a . To a mixture of 50 mg of unprotected
cyclolysine 25 (0.35 mmol), 83 mg of Cbz-protected serine (0.35
mmol), and 47.3 mg of HOAt (0.35 mmol) in 6 mL of freshly
distilled DMF was added 73.2 mg of EDC‚HCl (0.38 mmol).
The reaction mixture was allowed to stand at room tempera-
ture for 2 h. The reaction mixture was then diluted with 20
mL of EtOAc. The organic layer was washed with 5% aqueous
ascorbic acid (2 × 5 mL), saturated aqueous NaHCO3 (1 × 5
mL), and brine (2 × 5 mL), dried over Na2SO4, filtered, and
concentrated. The residue was purified by silica chromatog-
raphy with 10% of methanol in dichloromethane as the eluent
to afford 121 mg (95%) of 7a as a white solid: Rf ) 0.40 (10%
methanol in dichloromethane); mp 140-142 °C; IR (neat) 3306,
2930, 2846, 1710, 1644, 1060 cm-1; 1H NMR (300 MHz, CDCl3/
CD3OD) δ 7.46 (s, 5H), 6.41 (bd, 1H), 5.24 (s, 2H), 4.59 (m,
1H), 4.51-3.27 (m, 7H), 2.19-1.41 (m, 6H); 13C NMR (75
MHz, CDCl3/CD3OD) δ 170.30, 168.39, 156.29, 128.92, 128.35,
128.00, 127.82, 66.94, 62.74, 55.77, 52.00, 51.59, 30.01, 27.45,
1725, 1705, 1660 cm-1 1H NMR (500 MHz, CDCl3) δ 11.23
;
(bs, 1H), 7.68 (td, 1H, J ) 8.1, 1.5 Hz), 7.52 (m, 1H), 7.44-
7.13 (m, 6H), 7.01 (d, 1H, J ) 8.1 Hz), 6.91-6.86 (m, 1H),
5.77-5.75 (m, 1H), 5.12, 5.09 (ABq, 2H, J ) 9.9 Hz), 5.01-
4.82 (m, 1H), 4.92 (s, 1H), 4.87 (s, 1H), 4.68-4.54 (m, 4H),
4.25-4.19 (m, 2H), 3.92-3.46 (m, 6H), 2.43-2.34 (m, 2H),
2.05-1.24 (m, 41H), 1.13 (t, 2H, J ) 6.9 Hz), 0.99-0.91 (m,
2H), 0.87 (t, 3H, J ) 6.3 Hz), 0.02 (s, 9H); 13C NMR (150 MHz,
CDCl3) δ 175.81, 170.72, 169.82, 169.13, 168.96, 167.55, 159.62,
156.53, 136.17, 134.11, 128.47, 128.40, 128.01, 127.98, 127.91,
118.92, 116.87, 109.95, 99.10, 69.06, 67.83, 67.54, 67.29, 67.01,
58.66, 53.00, 52.00, 50.97, 32.51, 31.83, 30.89, 30.77, 30.65,
29.61, 29.60, 29.58, 29.56, 29.47, 29.43, 29.38, 27.67, 27.48,
26.09, 25.82, 24.45, 22.60, 22.37, 22.16, 18.33, 18.08, 14.04,
-1.47; HRFABMS calcd for C56H89N6O13Si 1081.6257, found
1081.6227.
P r otected Mycoba ctin 21b. To a solution of 33 mg of
amine 13 (0.086 mmol), 21 mg of Cbz-protected serine (0.086
mmol), 1.0 equiv of HOAt (12 mg, 0.086 mmol), and 13 mg of
DMAP (0.045 mmol) in 1.0 mL of dry DMF was added a
premixed solution of 0.10 mmol of EDC‚HCl and 0.10 mmol
of DMAP (12.2 mg) in 0.5 mL of dry DMF dropwise. After
being allowed to stand at room temperature for 3 h, the
mixture was charged with 62 mg of acid 20 (0.086 mmol) and
19 mg of EDC‚HCl (0.10 mmol). The resulting solution was
heated at 55 °C for 30 h before it was cooled to room
temperature. The reaction mixture was then diluted with 10
mL of EtOAc, washed with water (3 × 5 mL) and brine (3 ×
25.55; HRFABMS calcd for
366.1689.
C17H24N3O6 366.1665, found
N-Cbz-P r otected Mycoba ctin An a logu e 26a . To a solu-
tion of 6.7 mg of Cbz-protected serine (0.028 mmol) and 4.0

Total Syntheses of Mycobactin Analogs
J . Org. Chem., Vol. 63, No. 13, 1998 4321
mg of cyclolysine (0.028 mmol) 25 with 1 equiv of HOAt (4.0
mg) in 1 mL of freshly distilled DMF was added 5.4 mg of
EDC‚HCl (0.028 mmol). The resulting solution was allowed
to stand at room temperature for 2.5 h. Upon disappearance
of starting material based on TLC analysis, 16.5 mg of
mycobactic acid 24b (0.028 mmol) with 0.056 mmol of
EDC‚HCl18 was introduced. After being allowed to stand at
room temperature for 12 h, the reaction mixture was diluted
with 10 mL of EtOAc. The organic portion was washed with
water (2 × 5 mL), 0.5 N HCl (1 × 5 mL) and brine (2 × 5 mL),
dried over Na2SO4, filtered, and concentrated. The residue was
purified by reverse phase C18 silica gel chromatography with
1:1 methanol/acetonitrile as the eluent to afford 21.2 mg (81%)
of 26a as a white solid: Rf ) 0.40 on reverse phase C18 silica
gel TLC (methanol/acetonitrile ) 1/1); mp 73.5-74.5 °C; IR
(neat) 3300, 2905, 2856, 1775, 1660, 1635 cm-1; 1H NMR (300
MHz, CDCl3/CD3OD) δ 7.64 (d, 1H, J ) 6.6 Hz), 7.37 (d, 1H,
J ) 7.2 Hz), 7.28-7.25 (m, 5H), 7.07 (bd, 1H), 6.98 (m, 1H),
6.89-6.84 (m, 1H), 6.03 (bs, 1H), 5.06 (s, 1H), 4.92-4.89 (m,
2H), 4.65-4.48 (m, 4H), 4.33-4.25 (m, 2H), 4.85 (bs, 1H),
3.72-3.53 (m, 3H), 3.18 (bd, 1H), 2.60-2.49 (m, 1H), 2.40-
3080, 2950, 2865, 1720, 1660 cm-1; 1H NMR (300 MHz, CDCl3)
δ 7.75-7.70 (m, 5H), 7.45-7.34 (m, 10H), 6.72 (d, 1H, J ) 6.0
Hz), 5.46 (bs, 1H), 5.08 (s, 2H), 4.18 (dd, 1H, J ) 10.5, 6.0
Hz), 3.55-3.42 (m, 4H), 2.36 (t, 2H, J ) 5.7 Hz), 1.95-1.12
(m, 6H), 1.14 (s, 9H); 13C NMR (75 MHz, CDCl3) δ 170.32,
169.31, 156.29, 136.61, 136.13, 135.99, 132.01, 131.55, 130.27,
130.23, 128.44, 127.98, 127.56, 127.49, 77.2, 66.55, 54.21,
51.62, 33.92, 31.03, 27.30, 26.87, 25.31, 19.54; HRFABMS calcd
for C33H41N3O5Si 588.2894, found 588.2874.
Mycoba ctin An a logu e 4. A slurry of 50 mg of cobactin
27 (0.082 mmol) with 15 mg of 10% Pd-C in 8 mL of MeOH
was purged with a stream of N2 over 10 min and subsequently
with H2 for an additional 15 min. The reaction mixture was
stirred under H2 (1 atm) at room temperature for 3 h. Then
palladium catalyst was filtered, off through a pad of Celite.
The solution was concentrated to yield the corresponding free
amine 28. Without further purification, amine 28 was dis-
solved in a 2 mL of a DMF solution containing 42 mg of
mycobactic acid 24b (0.073 mmol) and 10 mg of HOAt (0.073
mmol). To the above solution was added 14 mg of EDC‚HCl
(0.073 mmol). The resulting mixture was allowed to stand at
room temperature for 3.5 h. The reaction was quenched by
addition of 15 mL of EtOAc and 5 mL of H2O. The organic
mixture was then washed with water (2 × 5 mL) and brine (2
× 5 mL), dried over Na2SO4, filtered, and concentrated to give
60 mg of desired coupling product 29. To a solution of 60 mg
of 29 (0.06 mmol) with 6.9 µL of HOAc (0.12 mmol) in 2 mL of
dry THF was added 0.12 mL of 1 M TBAF (0.12 mmol) in THF
dropwise at 0 °C. Then the reaction mixture was allowed to
warm to room temperature and stand for 1 h. After removal
of solvents, the residue was purified by reverse phase C18 silica
gel chromatography with 3:2 methanol/acetonitrile as the
eluent to afford 48.6 mg (85%) of 4 as a white solid: Rf ) 0.55
on reverse phase C18 silica gel TLC (3:1 methanol/ acetonitrile);
mp 173.5-175.0 °C; IR (neat) 3300, 2935, 2850, 1680, 1500
2.38 (m, 1H), 1.97-1.41 (m, 40H), 0.83 (t, 3H, J ) 7.2 Hz); 13
C
NMR (75 MHz, CDCl3/CD3OD) δ 179.12, 176.73, 176.47,
170.47, 168.27, 167.71, 159.36, 156.18, 136.01, 134.25, 128.58,
128.47, 128.38, 128.03, 127.86, 119.12, 116.75, 109.86, 77.19,
69.36, 69.20, 69.06, 67.65, 66.99, 62.70, 55.60, 53.17, 50.21,
49.87, 49.59, 49.31, 49.11, 49.02, 48.73, 47.14, 33.91, 33.07,
31.79, 29.57, 29.36, 29.31, 29.23, 29.00, 24.65, 23.62, 23.44,
22.55, 13.96; HRFABMS calcd for C49H73N6O12 937.5286, found
937.5309.
N-Cbz-P r otected Mycoba ctin An a logu e 26b. To a solu-
tion of 17.5 mg of Cbz-protected threonine (0.069 mmol) and
10 mg of cyclolysine (0.069 mmol) 25 with 1 equiv of HOAt
(9.4 mg) in 6 mL of freshly distilled DMF was added 13.3 mg
of EDC‚HCl (0.069 mmol). The resulting solution was allowed
to stand at room temperature for 2.5 h. Upon disappearance
of starting material based on TLC analysis, 40 mg of myco-
bactic acid 24b (0.069 mmol) with 33.3 mg of EDC‚HCl (0.173
mmol)18 was introduced. After being allowed to stand at room
temperature for 12 h, the reaction mixture was diluted with
25 mL of EtOAc. The organic portion was washed with water
(2 × 5 mL), 0.5 N HCl (1 × 5 mL), 5% ascorbic acid (2 × 5
mL), and brine (2 × 5 mL), dried over Na2SO4, filtered, and
concentrated. The residue was purified by reverse phase C18
silica gel chromatography with 1:3 methanol/acetonitrile as
the eluent to afford 49 mg (75%) of 26b as a white solid: Rf )
0.50 (10% methanol in EtOAc); mp 77.0-79.0 °C; IR (neat)
3300, 2925, 2835, 1780,1655 cm-1; 1H NMR (600 MHz, CDCl3)
δ 8.58 (bs, 1H), 7.66-7.62 (m, 2H), 7.41-7.35 (m, 1H), 7.30
(s, 5H), 7.00-6.93 (m, 1H), 6.89-6.84 (m, 1H), 6.08 (bs, 1H),
6.03 (bs, 1H), 5.07 (s, 2H), 4.95-4.91 (m, 1H), 4.67-4.52 (m,
4H), 4.31-4.25 (m, 2H), 3.91-3.89 (m, 1H), 3.63-3.52 (m, 3H),
2.44-2.40 (m, 2H), 2.18 (bs, 1H), 1.95-1.23 (m, 39H), 1.12 (bs,
3H), 0.86 (t, 3H, J ) 6.6 Hz); 13C NMR (150 MHz, CDCl3) δ
170.98, 170.32, 169.82, 169.21, 168.34, 167.53, 159.48, 156.52,
136.03, 134.18, 128.41, 128.32, 127.94, 127.77, 119.01, 116.77,
116.71, 109.79, 69.10, 67.60, 67.36, 66.94, 58.83, 53.09, 52.07,
50.82, 50.11, 49.05, 47.08, 33.01, 32.36, 31.76, 30.78, 29.56,
29.34, 29.29, 29.20, 28.96, 27.60, 25.57, 25.19, 24.58, 23.58,
1
cm-1; H NMR (300 MHz, CDCl3/CD3OD) δ 7.65 (dd, 1H, J )
7.8, 1.0 Hz), 7.41-7.35 (m, 1H), 6.99-6.96 (m, 1H), 6.88 (t,
1H, J ) 7.8 Hz), 5.00-4.95 (m, 1H), 4.66-4.49 (m, 3H), 4.33-
4.30 (m, 1H), 3.89-3.51 (m, 5H), 3.22-3.15 (m, 1H), 2.48-
0.95 (m, 42H), 0.85 (t, 3H, J ) 6.5 Hz); 13C NMR (125 MHz,
CDCl3/CD3OD) δ 175.16, 172.12, 170.76, 168.52, 167.33,
159.21, 155.32, 134.11, 128.51, 119.16, 116.72, 110.34, 77.20,
68.95, 67.79, 58.68, 53.58, 51.78, 51.51, 47.24, 36.40, 36.14,
32.34, 31.86, 31.21, 29.63, 29.59, 29.51, 29.46, 29.30, 29.11,
27.73, 25.74, 24.63, 23.84, 22.62, 21.82, 19.63, 14.05, 13.54;
HRFABMS calcd for C41H67N6O9 787.4969, found 787.4969.
P r otected Dia m in op r op ion ic Acid 32. To the solution
of 1.02 g of â-lactam 3113 (3.49 mmol) in 18 mL of 50% H2O in
THF was added 125 mg of LiOH (5.22 mmol). The resulting
mixture was allowed to stand at room temperature for 3 h.
Then the solution was acidified by addition of 8 mL of 2 N
HCl. The solution was then extracted with EtOAc (3 × 10
mL). The combined organic mixture was washed with H2O (2
× 10 mL) and brine (2 × 15 mL), dried over Na2SO4, filtered,
and then concentrated to afford 1.08 g of acid 32 (100%) as a
white solid: mp 132.0-134.0 °C; IR (neat) 3400, 3250, 2960,
1
2860, 1770, 1690 cm-1; H NMR (300 MHz, CDCl3/CD3OD) δ
7.29-7.17 (m, 5H), 4.58 (s, 2H), 4.08 (bt, 1H, J ) 5.1 Hz), 3.11
(d, 2H, J ) 5.1 Hz), 1.32 (s, 9H); 13C NMR (125 MHz, CDCl3/
CD3OD) δ 176.80, 156.15, 137.20, 128.13, 128.09, 127.60, 79.45,
75.41, 53.53, 53.00, 28.03; HRFABMS calcd for C15H23N2O5
311.1607, found 311.1617.
23.38, 22.53, 21.79, 18.09, 13.98; HRFABMS calcd for C50H75
N6O12 951.5443, found 951.5453.
-
P r otected Coba ctin 27. To a mixture of 93 mg of N-Cbz-
â-alanine (0.42 mmol), 160 mg of amine 13 (0.42 mmol), and
1 equiv of HOAt (57 mg) in 3 mL of DMF was added 1.1 equiv
of EDC‚HCl (88.3 mg). After being allowed to stand at room
temperature for 2 h, the mixture was diluted with 20 mL of
EtOAc and 5 mL of saturated aqueous NaHCO3. The organic
layer was then washed with water (2 × 5 mL) and brine (2 ×
5 mL), dried over Na2SO4, filtered, and concentrated. The
residue was purified by silica chromatography with 1.5:1 ethyl
acetate/hexanes as the eluent to afford 224 mg (91%) of 27 as
a colorless oil: Rf ) 0.40 (3:1 EtOAc/hexanes); IR (neat) 3320,
P r otected Coba ctin 33. To a solution of 23 mg of amine
13 (0.061 mmol), 17 mg of Boc-protected diaminopropionic acid
(0.055 mmol) 32, 1.0 equiv of HOAt (7.5 mg, 0.055 mmol), and
7.4 mg of DMAP (0.055 mmol) in 0.5 mL of dry DMF was
added a premixed solution of 0.055 mmol of EDC and 0.055
mmol of DMAP in 0.5 mL of dry DMF dropwise. The resulting
mixture was allowed to stand at the room temperature for 6
h. The reaction mixture was then diluted with 15 mL of
EtOAc, washed with water (2 × 5 mL) and brine (2 × 5 mL),
dried over Na2SO4, filtered, and concentrated. The residue was
purified by silica chromatography with 1:1 ethyl acetate/
hexanes as the eluent to afford 33.8 mg (91%) of 33 as a
colorless oil: IR (neat) 3350, 2990, 2895, 1720, 1695 cm-1; 1H
(18) Additional EDC‚HCl was used to accelerate the reaction.

4322 J . Org. Chem., Vol. 63, No. 13, 1998
Xu and Miller
NMR (300 MHz, CDCl3) δ 7.75-7.70 (m, 4H), 7.54 (bd, 1H),
7.42-7.27 (m, 11H), 5.95 (s, 1H), 5.48 (bd, 1H), 4.68 (s, 2H),
4.31 (bs, 1H), 4.24-4.16 (m, 1H), 3.51-3.40 (m, 1H), 3.35 (bs,
1H), 3.30 (d, 1H, J ) 5.1 Hz), 3.51 (dd, 1H, J ) 13.8, 6.3 Hz),
1.82-1.18 (m, 6H), 1.44 (s, 9H), 1.14 (s, 9H); 13C NMR (125
MHz, CDCl3) δ 169.86, 169.53, 155.71, 137.42, 136.14, 136.06,
132.10, 130.22, 130.18, 129.61, 129.46, 128.54, 128.39, 127.89,
127.56, 127.48, 76.63, 76.50, 76.04, 54.15, 53.36, 51.81, 30.85,
33) of 35b as a white solid: Rf ) 0.30 on reverse phase C18
silica gel TLC(1:8 methanol/acetonitrile); mp 164.5-165.5 °C;
1
IR (neat) 3390, 3090, 2990, 2850, 1690, 1670 cm-1; H NMR
(500 MHz, CDCl3/CD3OD) δ 7.62 (d, 1H, J ) 7.5 Hz), 7.35 (t,
1H, J ) 8.0 Hz), 6.94 (d, 1H, J ) 8.0 Hz), 6.84 (t, 1H, J ) 7.5
Hz), 4.94 (t, 1H, J ) 9.0 Hz), 4.91-4.58 (m, 1H), 4.55-4.51
(m, 1H), 4.50-4.47(m, 1H), 4.12 (bs, 2H), 3.85-3.81 (m, 1H),
3.78-3.76 (m, 1H), 3.69-3.66 (m, 1H), 3.57-3.51 (m, 3H), 3.37
(t, 2H, J ) 7.5 Hz), 1.91-1.83 (m, 40 H), 1.38 (s, 9H), 0.82 (t,
3H, J ) 6.5 Hz); 13C NMR (125 MHz, CDCl3/CD3OD) δ 175.15,
172.60, 171.02, 170.68, 168.81, 167.17, 159.20, 155.21, 133.95,
128.39, 119.02, 116.60, 109.99, 80.00, 77.20, 68.97, 67.65,
54.15, 53.35, 51.96, 51.88, 47.21, 41.06, 32.25, 31.78, 30.62,
29.56, 29.52, 29.44, 29.39, 29.33, 29.22, 28.78, 28.12, 27.59,
28.29, 27.29, 26.90, 25.36, 19.58; HRFABMS calcd for C37H51
N4O6Si 675.3578, found 675.3551.
-
N-Boc-P r otected Mycoba ctin An a logu e 35b. A slurry
of 56 mg of cobactin 33 (0.083 mmol) with 15 mg of 10% Pd-C
was purged with an N2 stream for 10 min and subsequently
with H2 for an additional 15 min. The reaction mixture was
stirred under H2 (1 atm) at 56 °C for 2 h. Then the palladium
catalyst was filtered, off through a pad of Celite. The solution
was concentrated to yield 45 mg of the corresponding free
amine 34. Without further purification, amine 34 was dis-
solved in 2 mL of a DMF solution containing 49 mg of
mycobactic acid 24b (0.084 mmol) and 11.4 mg of HOAt (0.084
mmol). To the above solution was added 16 mg of EDC‚HCl
(0.083 mmol). The resulting mixture was allowed to stand at
room temperature for 3 h. The reaction was quenched by
addition of 20 mL of EtOAc and 5 mL of H2O. The organic
mixture then washed with water (2 × 5 mL) and brine (2 × 5
mL), dried over Na2SO4, filtered, and concentrated to give 84
mg of desired coupled product 35a in 89% yield. To a solution
of 37 mg of 35a (0.032 mmol) with 5.6 µL of HOAc (0.097
mmol) in 2 mL of dry THF was added 0.1 mL of 1 M TBAF
(0.1 mmol) in THF dropwise at 0 °C. Then the reaction
mixture was allowed to warm to room temperature and stand
for 1 h. After removal of solvents, the residue was purified
by reverse phase C18 silica gel chromatography with 1:8
methanol/acetonitrile as the eluent to afford 27.6 mg (85% from
25.68, 24.61, 22.55, 22.06, 13.95; HRFABMS calcd for C46H76
N7O11 902.5603, found 902.5595.
-
Ack n ow led gm en t. We gratefully acknowledge fi-
nancial support from the National Institutes of Health
(GM 25845). Biological tests of mycobactin analogues
were performed by the Tuberculosis Antimicrobial
Acquisition and Coordinating Facility under Contract
N01-AI-45246 sponsored by the National Institute of
Allergy and Infectious Diseases.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectra of new compounds (30 pages). This material is
contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
J O980063O