Combinatorial synthesis through disulfide exchange: Discovery of potent psammaplin A type antibacterial agents active against methicillin-resistant Staphylococcus aureus (MRSA)

Article abstract of DOI:10.1002/1521-3765(20011001)7:19<4280::AID-CHEM4280>3.0.CO;2-3

Psammaplin A is a symmetrical bromotyrosine-derived disulfide natural product isolated from the Psammaplysilla sponge, which exhibits in vitro antibacterial activity against methicillin-resistant Staphylococcus aureus (MRSA). Inspired by the structure of this marine natural product, a combinatorial scrambling strategy for the construction of heterodimeric disulfide analogues was developed and applied to the construction of a 3828-membered library starting from 88 homodimeric disulfides. These psammaplin A analogues were screened directly against various gram positive bacterial strains leading to the discovery of a series of potent antibacterial agents active against methicillin-resistant Staphylococcus aureus (MRSA). Among the most active leads derived from these studies are compounds 104, 105, 113, 115, 123, and 128. The present, catalytically-induced, disulfide exchange strategy may be extendable to other types of building blocks bearing thiol groups facilitating the construction of diverse discovery-oriented combinatorial libraries.

Full text of DOI:10.1002/1521-3765(20011001)7:19<4280::AID-CHEM4280>3.0.CO;2-3

FULL PAPER
Combinatorial Synthesis through Disulfide Exchange:
Discovery of Potent Psammaplin A Type Antibacterial Agents Active
against Methicillin-Resistant Staphylococcus aureus (MRSA)
K. C. Nicolaou,* Robert Hughes, Jeffrey A. Pfefferkorn, Sofia Barluenga,
and A. J. Roecker[a]
Abstract: Psammaplin A is a symmet-
rical bromotyrosine-derived disulfide
natural product isolated from the Psam-
maplysilla sponge, which exhibits in
vitro antibacterial activity against methi-
cillin-resistant Staphylococcus aureus
(MRSA). Inspired by the structure of
this marine natural product, a combina-
torial scrambling strategy for the con-
struction of heterodimeric disulfide ana-
logues was developed and applied to the
construction of a 3828-membered li-
brary starting from 88 homodimeric
disulfides. These psammaplin A ana-
logues were screened directly against
various gram positive bacterial strains
leading to the discovery of a series of
potent antibacterial agents active
against methicillin-resistant Staphylo-
coccus aureus (MRSA). Among the most
active leads derived from these studies
are compounds 104, 105, 113, 115, 123,
and 128. The present, catalytically-in-
duced, disulfide exchange strategy may
be extendable to other types of building
blocks bearing thiol groups facilitating
the construction of diverse discovery-
oriented combinatorial libraries.
Keywords: antibiotics
´
chemical
biology ´ combinatorial synthesis ´
psammaplin A ´ total synthesis
biological evaluation of a novel heterodimeric disulfide
library inspired by the antibacterial marine natural product
psammaplin A (1, Figure 1).
Introduction
Combinatorial chemistry has become an important tool for
both the drug discovery process and chemical biology
studies.[1] Particularly interesting has been the application of
combinatorial techniques to the synthesis and structure ± ac-
tivity optimization of biologically active natural products.[2]
Typically, the structural complexities of these molecules pose
unique challenges which necessitate the development of novel
solution or solid phase combinatorial strategies.[3] In many
cases, the methodologies developed in this way can later be
applied to the synthesis of combinatorial libraries with more
general applications. With such considerations in mind, we
undertook the solution phase combinatorial synthesis and
OH
OH
Br
Br
1: psammaplin A
O
O
S
S
N
N
H
H
N
N
HO
OH
subunit a
subunit b
Figure 1. Structure of psammaplin A, an antibacterial marine natural
product isolated from Psammaplysilla sp.
Psammaplin A (1, Figure 1) is a symmetrical bromotyro-
sine-derived disulfide dimer that was originally isolated in
1987 from the Psammaplysilla sponge.[4] Early studies re-
vealed that psammaplin A had general antibacterial and
antitumor properties. Later studies demonstrated that psam-
maplin A (1) exhibited significant in vitro antibacterial
activity against both Staphylococcus aureus (SA) and methi-
cillin-resistant Staphylococcus aureus (MRSA)[5] Given the
increasingly rapid emergence of multi-drug resistant bacterial
strains and the corresponding threat to public health, there is
signifigant interest in the development of structurally novel
antibacterial agents such as psammaplin A.[6, 7] Additionally,
[a] Prof. Dr. K. C. Nicolaou, R. Hughes, Dr. J. A. Pfefferkorn,
Dr. S. Barluenga, A. J. Roecker,
Department of Chemistry
and The Skaggs Institute for Chemical Biology
The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, California 92037 (USA)
Fax : (‡1) 858-784-2469
and
Department of Chemistry and Biochemistry
University of California San Diego
9500 Gilman Drive, La Jolla, California 92093 (USA)
E-mail: kcn@scripps.edu
Supporting information for this article is available on the WWW under
http://www.wiley-vch.de/home/chemistry/ or from the author.
4280
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4280 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 19

4280±4295
psammaplin A has been reported to exhibit weak inhibition of
a number of enzymes including DNA gyrase,[5] topoisomer-
ase II,[8] farnesyl protein transferase,[9] and leucine amino-
peptidase.[9] Despite these reports of biological activity,
however, no investigations into the structure ± activity rela-
tionships of psammaplin A have been disclosed, and its
chemical mechanism of action remains unknown. In light of
this, we undertook the synthesis of a combinatorial library of
disulfide containing molecules in order to find potent and
structurally simplified analogues which could be used as tools
for studying psammaplin Aꢁs mechanism of action and its
potential therapeutic value.
From the viewpoint of combinatorial chemistry, psamma-
plin A (1, Figure 1) is rather interesting owing to its two
identical domains (subunits a and b) which are linked through
a disulfide bridge. At the outset of this work it was unknown
whether one or both of these subunits were necessary for
antibacterial activity nor was it clear whether the disulfide
moiety played a mechanistic or structural role in the agentꢁs
biological action. Combinatorially, we envisioned utilizing the
disulfide motif as a readily exchangeable linkage[10] which
would allow rapid construction of a heterodimeric analogue
library suitable for defining structure ± activity relation-
ships.[11] It is well precedented that disulfide bonds will readily
undergo facile exchange reactions with other disulfides in
high yield under mild conditions.[12] For example, and as
shown in Figure 2, if two homodimeric disulfides, A-SS-A and
It was anticipated that a combinatorial strategy of this type
would provide an opportunity to simultaneously evaluate
modifications at remote sites of the psammaplin A architec-
ture for potential synergistic effects which might be diffi-
cult to uncover through a traditional SAR approach. This
program evolved through several stages as described in this
and the following article.[13] Initially, psammaplin A (1) and a
number of designed homodimeric analogues were individu-
ally synthesized. These synthetic homodimers, along with
other commercially available homodimeric disulfides, were
then systematically scrambled to create a large (> 3800-
membered) library of heterodimeric analogues which was
screened for antibacterial activity to identify structurally
novel lead compounds that were then re-synthesized in
homogenous form to confirm their antibacterial activities.
Subsequently, (see, following article)[13] the antibacterial
properties of these novel heterodimeric leads were further
refined through the synthesis and biological evaluation of
several promising series of simplified analogues. Finally, a
collection of analogues with structural modifications aimed at
elucidating the molecular mechanism of action of these
psammaplin-type compounds were constructed and employed
in order to better understand how this unique class of
antibacterial agents function.
Results and Discussion
Synthesis of psammaplin A and homodimeric analogues:
Prior to commencing the combinatorial aspects of this work,
we first required synthetic access to psammaplin A (1, Fig-
ure 1) and 44 designed homodimeric analogues (see Figure 3)
to be used as symmetrical disulfides in the projected
construction of the combinatorial library. To this end,
Hoshinoꢁs synthesis[14] of 1 was modified as shown in
Scheme 1 to accommodate our requirements for a general
approach to a variety of psammaplin-type structures. As
shown, l-tyrosine (2) was mono-brominated by treatment
with potassium bromate/potassium bromide to afford 3-bro-
motyrosine (3) in 81% yield.[15] Amino acid 3 was then
refluxed in TFAA (for abbreviations of reagents and protect-
ing groups, see legends in schemes) to form trifluoromethyl-
oxazolone 4, which was distilled from the reaction vessel and
hydrolyzed with aqueous TFA to provide a-keto acid 5 in
60% yield over two steps.[16] This a-keto acid was subse-
quently condensed with O-(tetrahydro-2H-pyran-2-yl)-hy-
droxylamine, affording oxime 6. After considerable optimi-
zation, it was found that crude carboxylic acid 6 could be
reliably coupled with cystamine [(H2NCH2CH2S)2] through
the intermediacy of the succinate ester. Hence, acid 6 was
treated with N-hydroxysuccinimide in the presence of EDC to
give succinate ester 7, which was isolated and immediately
treated with cystamine and Et3N to provide, through a bis-
coupling reaction, amide 8. Finally, the THP protecting groups
on the oximes were removed (HCl, MeOH, 608C) to afford
psammaplin A (1) in 36% yield over four steps. With the
psammaplin A synthesis completed, the sequence illustrated
in Scheme 1 was repeated for the construction of homodi-
meric analogues 13 ± 36 (see Figure 3) using nine different
(1)
S-S
A
A
B
A
S-S
+
A
B
A
B
+
catalyst
buffer
S-S
+
(2)
(1)
B
B
S-S
S-S
[homodimers]
[1:2:1 mixture]
Figure 2. General principle for construction of novel heterodimeric
disulfides through scrambling and equilibration of homodimeric disulfides.
B-SS-B, are mixed under basic conditions in the presence of a
suitable catalyst they will undergo rapid exchange reactions to
afford a statistical mixture of three disulfides: A-SS-A, A-SS-
B, and B-SS-B in a ratio of 1:2:1, respectively, resulting from
the degeneracy of the A-SS-B and B-SS-A products.
Abstract in Greek:
Chem. Eur. J. 2001, 7, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4281 $ 17.50+.50/0
4281

K. C. Nicolaou et al.
FULL PAPER
O
OH
Br
O
O
a. KBr, KBrO3
Br
OH
OH
b. EDC
Br
OH
O
O
NH2
NH2
HO
HO
2
3
NHR
OH
N
HO
N
O
O
O
3 : R = H
9 : R = Boc
b. TFAA, ∆
a. Boc2O
O
NHBoc
10
OH
O
Br
OH
OH
c. TFA (aq)
Br
O
N
OH
Br
Br
O
HO
CF3
HO
c. Et3N
4
5
O
O
S
S
S
N
N
H
H2N
d. THPO-NH2
H
2
NHR
NHR
11 : R = Boc
12 : R = H
d. TFA
OH
OH
Br
Br
Scheme 2. Synthesis of diamino-psammaplin
A (12). a) di-tert-Butyl
e. EDC
O
dicarbonate (2.0 equiv), NaHCO3 (5.0 equiv), THF/H2O 1:1, 238C, 12 h;
b) NHS (1.9 equiv), EDC (1.7 equiv), 1,4-dioxane, 238C, 2 h; c) Et3N
(2.0 equiv), cystamine ´ 2HCl (0.5 equiv), 1,4-dioxane/MeOH 1:1, 238C,
12 h, 46% over three steps; d) TFA (10.0 equiv), CH2Cl2, 238C, 2 h, 95%.
O
O
OH
N
N
O
O
O
OH
O
N
N
THPO
7
THPO
6
free amino analogue 12 in 95% yield. In addition, sulfon-
amide analogues 49 ± 53 (Figure 3) and amide analogues (54
and 55, Figure 3) were synthesized by standard techniques.
Finally, additional homodimeric disulfide building blocks
containing various substituted alkyl, aromatic, and heteroar-
omatic skeletons (see compounds 56 ± 99, Figure 3) were
obtained from commercial sources in order to enhance the
libraryꢁs structural diversity.
OH
Br
OH
OR
Br
f. Et3N
O
O
S
S
N
N
S
H
H
H2N
N
N
2
RO
8 : R = THP
1 : R = H
g. HCl, MeOH
Scheme 1. Total synthesis of psammaplin A (1) from l-tyrosine (2).
a) KBrO3 (0.3 equiv), KBr (2.0 equiv), H2O (0.25m HCl), 238C, 12 h, 81%;
b) TFAA (excess), 808C, 12 h; c) 70% TFA (aq), 238C, 24 h, 60% over two
steps; d) THP-ONH2 (1.5 equiv), EtOH, 238C, 12 h; e) NHS (1.9 equiv),
EDC (1.7 equiv), 1,4-dioxane, 238C, 2 h; f) Et3N (2.0 equiv), cystamine ´
2HCl (0.5 equiv), 1,4-dioxane/MeOH 1:1, 238C, 12 h; g) HCl (4.0 equiv),
CH2Cl2/MeOH (20:1), 608C, 2 h, 36% over five steps. EDC ˆ 1-[3-
(dimethylamino)propyl)]-3-ethylcarbodiimide; NHS ˆ N-hydroxysuccini-
mide; TFA ˆ trifluoroacetic acid; TFAA ˆ trifluoroacetic anhydride;
THP-ONH2 ˆ O-(tetrahydro-2H-pyran-2-yl)-hydroxylamine.
Preliminary disulfide exchange experiments: With the requi-
site symmetrical building blocks in hand, we next sought to
determine the optimal disulfide exchange conditions paying
particular attention to address the following two concerns:
a) the reaction conditions ought to afford approximately
statistical ratios of scrambled products in high yields and
purities; and b) the required reagents and their by-products
ought to be sufficiently innocuous toward bacteria so that the
disulfide products could be transferred directly from the
reaction mixture to the screening assays for antibacterial
activity without the need for any work-up or purification
procedures.
Based on ample literature precedent regarding disulfide
exchange chemistry, suitable reaction conditions for this
process were rapidly developed as illustrated for the repre-
sentative example shown in Scheme 3. Hence, the two
homodimeric disulfides 14 and 29 were mixed in a DMSO/
H2O (3:1) solvent system buffered to pH 8.3 with a standard
amino acids (phenylalanine, 4-fluorophenylalanine, 4-chloro-
phenylalanine, tyrosine, 3-chlorotyrosine, 3-bromotyrosine,
3,5-dibromotyrosine, phenylglycine, and 3-methoxytyrosine)
and three differently substituted hydroxylamines [O-(tetra-
hydro-2H-pyran-2-yl)-hydroxylamine, O-methyl-hydroxyl-
amine, and O-benzyl-hydroxylamine].
With these oxime-containing analogues in hand, several
additional homodimeric analogues were prepared in an effort
to determine the biological importance of other structural
features of the psammaplin A skeleton. These analogues
included amino acid derivatives 11, 12, and 38 ± 48 (Figure 3)
which were synthesized as illustrated in Scheme 2 for the
particular di-amino disulfide analogue 12. Hence, 3-bromo-
tyrosine 3 was N-protected by treatment with di-tert-butyl-
dicarbonate in the presence of NaHCO3 to provide N-Boc-3-
bromotyrosine (9). Crude compound 9 was then treated with
N-hydroxysuccinimide and EDC to afford succinate ester 10,
which was coupled with cystamine as described above to
provide di-amide 11 in 46% yield over three steps. Compound
11 was fully deprotected (TFA, CH2Cl2, 238C) to provide the
phosphate buffer. Subsequently,
a
catalytic amount
(0.15 equiv) of dithiothreitol (100) was added and the reaction
mixture was stirred for 12 h at 238C. Reaction progress was
monitored by HPLC analysis as illustrated in the side panels
of Scheme 3. Thus, prior to addition of reagent (Scheme 3,
panel a) only the two parent homodimers were present in the
HPLC trace, whereas upon addition of dithiothreitol, a third
product, heterodimer 101, was detected as indicated by the
corresponding HPLC trace (Scheme 3, insert panel b, 12 h,
equilibrium).[17] After normalization, integration of the three
disulfide peaks in this trace revealed their respective concen-
trations to be in a ratio of 1.0:2.1:1.0.[18] Repetition of this
4282
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4282 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 19

Combinatorial Synthesis
4280±4295
F
OH
OH
Cl
16: R = H
17: R = Me
18: R = Bn
19: R = H
20: R = Me
21: R = Bn
13: R = H
14: R = Me
15: R = Bn
22: R = H
23: R = Me
24: R = Bn
25: R = H
26: R = Me
27: R = Bn
Cl
O
O
O
O
O
O
O
O
)
)
S
)
)
)
S
S
S
S
N
2
N
H
2
N
2
N
2
N
H
2
H
H
H
N
N
N
N
N
RO
RO
RO
RO
RO
HO
MeO
OH
OH
OH
1: R = H
28: R = Me
29: R = Bn
30: R = H
31: R = Me
32: R = Bn
33: R = H
34: R = Me
35: R = Bn
Br
Br
O
Br
MeO
)
S
N
H
2
36
O
O
)
S
)
37
S
)
S
)
S
N
H
2
N
2
N
H
2
N
2
H
H
N
N
N
N
RO
RO
RO
MeO
NO2
OH
NHR
OH
Br
44: R = H
45: R = Boc
42: R = H
38: R = H
39: R = Boc
40: R = H
41: R = Boc
12: R = H
11: R = Boc
43: R = Boc
O
O
O
O
)
S
N
H
2
)
S
)
)
S
)
S
S
N
2
N
2
N
2
N
2
NHR
H
H
H
H
NHR
NHR
O
NHR
NHR
O
O
O
O
O
O
)
S
S
)
)
S
S
S
N
2
N
H
2
)
)
)
S
)
S
S
S
BocNH
H
N
H
2
N
2
2
N
H
H
MeO
Me
NH2
NH2
46
48
49
47
50
O
N
O
O
S
O O
S
O
O
O
)
S
)
)
S
S
S
S
N
H
2
N
2
Me
N
2
N
2
MeO
N
2
H
H
H
H
O2N
53
54
55
51
52
Ph
)
)
)
S
S
S
)
)
)
S
N
S
S
N
)
2
S
2
)
S
2
HN
2
2
2
N
2
2
N
Me
N
)
NH2
59
O2N
Br
Me2N
N
O2N
O2N
O
N
Me
56
57
58
60
61
62
)
S
)
S
S
S
)
S
N
2
O
N
2
)
2
S
2
2
HO
)
)
S
MeO
N
S
2
2
MeO
HO
O
63
70
64
65
66
67
74
68
69
Me
HO
HO
N
O
)
S
)
O
)
)
N
S
S
N
S
S
)
S
HO
)
S
2
)
S
2
S
2
2
2
2
N
N
O
HO2C
nPrO
71
72
73
75
76
)
)
S
NH2
NH2
2
)
)
2
S
S
S
2
2
2
2
)
S
)
HO
S
)
EtO
OH
NHMe
2
2
H2N
O
O
O
O
82
78
79
80
81
83
77
S
O
)
S
)
)
)
S
)
)
S
HO2C
S
S
O
2
2
2
2
2
)
S
86
HO
2
)
NH2
Me
S
S
Me
84
O2N
NC
2
85
88
96
89
90
98
91
99
87
)
O
)
)
)
S
)
S
S
N
S
2
H2N
2
2
2
2
2
)
S
94
95
2
nBuO
NH2
)
Cl
NH2
AcHN
S
N3
2
97
92
93
Figure 3. Structures of homodimeric disulfides utilized in library synthesis. Compounds 1, 11, 12, 13 ± 48 were synthesized as illustrated in Schemes 2 and 3.
Compounds 49 ± 55 were synthesized by standard techniques. Compounds 56 ± 99 were purchased from commercial sources.
experiment with other, structurally dissimilar, building blocks
from Figure 3 were conducted in a similar manner in order to
ensure that this reaction time and catalyst loading were
sufficient to allow for complete equilibration with no free
thiol species remaining. As discussed later, post-synthetic
analysis of representative library members did, in fact,
confirm that these reaction conditions lead generally and
reproducibly to statistical mixtures.
With the scrambling conditions defined, we next proceeded
to determine whether the disulfide products could be
Chem. Eur. J. 2001, 7, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4283 $ 17.50+.50/0
4283

K. C. Nicolaou et al.
FULL PAPER
Scheme 3. Example of disulfide exchange reaction: An equimolar (1.0 mm in each homodimer) solution of homodimers (14 and 29) in DMSO:H2O (3:1)
(insert panel a) was treated with 0.15 equiv dithiothreitol (100) and stirred at 238C for 12 h to yield homodimers 14 and 29 and heterodimer 101 (insert
panel b).
screened directly from the reaction mixture. Specifically, this
required evaluating whether or not the catalyst, dithiothreitol,
and the resulting by-product trans-1,2-dithiane-4,5-diol were
biologically inert toward bacteria at the concentration library
members were to be screened. This was confirmed by
screening both dithiothreitol and trans-1,2-dithiane-4,5-diol
against SA 6538 and MRSA 700698 strains. Encouragingly, at
1
50 mgmL , which is 50-fold greater than the concentration at
which either of these compounds would be present in the
screening mixture, neither of them exhibited any detectable
antibacterial activity. These results suggested that the pres-
ence of these compounds in trace amounts would not interfere
with the biological results of the intended assays, and
consequently, we anticipated screening library members
directly from the reaction mixtures. Similarly, it was con-
firmed that the DMSO solvent and phosphate buffer medium
used in the scrambling reaction would be sufficiently diluted
in the screening process so as to avoid any undesirable effects.
Solution phase combinatorial synthesis of psammaplin A
analogue library: With the requisite symmetrical building
blocks prepared and the reaction conditions for their ex-
change defined, construction of the combinatorial psamma-
plin A analogue library proceeded smoothly. The synthesis
was conducted in 96-well plates as illustrated in Figure 4. In
order to accommodate the 88 homodimeric starting materials
(Figure 3), a total of 88 96-well plates were arranged in an 8 Â 11
array. All wells were filled initially with a DMSO/buffer (3:1)
solution. The homodimers were then dispensed so that each
homodimer occupied one entire column and one entire
Figure 4. Schematic representation of library format.
row with 0.4 mmol of each homodimer deposited per well. For
example, homodimer 8' (Figure 4) was dispensed to each well
in the 8' row and the 8' column. Once complete, each well
4284
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4284 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 19

Combinatorial Synthesis
4280±4295
contained two different homodimers with the exception of
wells on the matrix diagonal (see below for further discussion
of symmetry properties of this matrix). Next, a solution of
dithiothreitol in DMSO was dispensed into each well
(0.15 equiv). Prior to scrambling, the final reaction concen-
tration was 1.0 mm in each homodimer. The plates were
individually agitated and then allowed to stand for 12 h at
ambient temperature. These conditions promoted scrambling
of the two disulfides in each well to afford the corresponding
heterodimer, a new compound as illustrated in Figure 4
wherein homodimers 6' and 8' are exchanged to afford a
heterodimer present at approximately twice the concentration
of either of the two original homodimers.
Notably, formatting the library in this manner resulted in
the formation of two symmetrical copies of the library. In
other words, this library is reflected across the diagonal (see
Figure 4) so that the products in well (6', 8') are identical to
those in well (8', 6'). While necessitating additional manipu-
lations, this duplicity simplified library construction. More
importantly, it was anticipated that screening two independ-
ently synthesized copies of the library for biological activity
would produce duplicate data whose accuracy would be
enhanced.
After the synthesis, it was necessary to determine the purity
and distribution of reaction products in the library by
analyzing 100 representative reaction wells. This task was
accomplished through direct HPLC analysis of the reaction
mixtures in these wells (see Supporting Information for
tabulated data). In summary, 95% of the reaction wells
sampled contained ꢀ50% of heterodimer suggesting that, in
general, the reactions proceeded smoothly and completely (as
a 50% yield of heterodimer would be expected if the
components reached a precise equilibrium).
homodimer displayed activity and, consequently, no useful
information regarding the biological activity of the hetero-
dimer could be obtained. To overcome this effect, the assay
concentration needed to be incrementally lowered to a point
below which the homodimeric component(s) was no longer
active (i.e., screening concentration <MIC value of homo-
dimer). If at this lower concentration the mixture in the well
retained activity, it was assumed that the heterodimer was
worthy of further investigation. Importantly, since the heter-
odimer is present at twice the concentration of each homo-
dimer (due to degeneracy), it was anticipated that this
deconvolution through a concentration gradient would be
relatively straightforward. To assist in this process, all
homodimeric starting materials were initially screened for
antibacterial activity, and the results are presented in Table 1.
Notably, from the synthetic homodimer subset, only com-
pounds 1 (psammaplin A), 13, 16, 22, 25, and 30 (see Figure 3
for structures) exhibited significant antibacterial activity. All
of the active synthetic homodimers contained a free a-oxime
and most were also halogenated (either bromo or chloro) on
the tyrosine. Several of the commercial homodimeric building
blocks also displayed antibacterial activities as illustrated in
Table 1 including compounds 56, 65, 66, 69 and 82 (see
Figure 3 for structures). The remaining homodimers (i.e.,
compounds not listed in Table 1) were found to have no
1
appreciable antibacterial properties at 50 mgmL .
Experimentally, evaluation of the library was accomplished
by screening the entire, symmetrically reflected library at
three concentrations (10.0, 5.0, and 2.5 mm in heterodimer)
against two bacterial strains (SA 6538 and MRSA 700698). To
accomplish this goal, aliquots were transferred from the
reaction plates to shallow well dilution plates containing
DMSO, retaining the original library format. These plates
were then screened for antibacterial activity. For each strain
and concentration, the wells of the plates were optically read
for turbidity and a graphical representation of the biological
activity in each plate was generated. The graphical represen-
tations for the individual plates were then assembled to form
the arrays illustrated in Figure 5 for the SA 6538 strain at the
three concentrations indicated.
Closer examination of the arrays in Figure 5 illustrates the
two types of screening scenarios described above. At the
highest concentration (10.0 mm) there was a significant
number of rows and columns in which almost every well was
active, indicating that at least one of the homodimers in that
position was active and overwhelming the activity of the other
compound(s) in those wells. As the concentration was halved
(i.e., 5.0 mm), the numbers of completely active rows dimin-
ished, and at the lowest concentration (i.e., 2.5 mm) any
residual effect of the homodimer(s) was lost. An identical
trend was observed for screening in the MRSA 700698 strain
(data not shown). In these arrays, white squares denote wells
with highly active compounds while black squares represent
inactive compounds. The various shades of blue denote
compounds of intermediate activity. In general, the most
active compounds were selected for further evaluation.
Active compounds were identified by visual assessment of
the arrays as shown in Figure 5. Priority was given to
compounds which exhibited activity at the lower concentra-
Antibacterial screening of psammaplin A library: Once the
library was complete, high throughput antibacterial screening
was undertaken to identify heterodimeric analogues of
psammaplin A exhibiting biological activity. This screening
posed an interesting challenge since each well contained a
mixture of three compounds in the approximate ratio of 1:2:1
[homodimer (A-SS-A):heterodimer (A-SS-B):homodimer
(B-SS-B)], and it was, therefore, expected that the total
activity of any given well would be a composite[18] of the
activities of the three compounds in the well. While we do not
generally favor mixture-based screening, in this case the
synthetic efficiency attained by synthesizing the heterodimer-
ic analogues as mixtures outweighed the necessary, but trivial,
deconvolution efforts. Indeed, such deconvolution of active
compounds from this library was quite straightforward and
fell into two scenarios. In the first, and simpler case, for a well
in which both homodimeric components were biologically
inactive, any detectable activity for that well could be
attributed to the heterodimeric product. The second case
was more complicated and occurred if one or both of the
homodimeric starting materials in a given well were to be
active (for example, when heterodimers were constructed
from psammaplin A itself). Thus, at assay concentrations
higher than the MIC (minimum inhibitory concentration)
value of the active homodimer, every well containing that
Chem. Eur. J. 2001, 7, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4285 $ 17.50+.50/0
4285

K. C. Nicolaou et al.
FULL PAPER
Table 1. Antibacterial activity (MIC mgmL 1) of psammaplin A (1) and other biologically active homodimeric compounds.
R1 R2
R3 R4
SA[a]
6538
SA[b]
13709
SA[c]
29213
SA[d]
25923
MRSA[e] MRSA[f] MRSA[g] MRSA[h] MRSA[i] AVG[j] AVG[k]
700698 43300 700787 700788 700789 SA MRSA
[a] Staphylococcus aureus ATCC 6538. [b] Staphylococcus aureus ATCC 13709. [c] Staphylococcus aureus ATCC 29213. [d] Staphylococcus aureus
ATCC 25923. [e] Staphylococcus aureus ATCC 700698, resistant to methicillin and heterogeneous susceptibility to vancomycin. [f] Staphylococcus aureus
ATCC 43300, resistant to methicillin. [g] Staphylococcus aureus ATCC 700787, resistant to methicillin and intermediate susceptibility to vancomycin.
[h] Staphylococcus aureus ATCC 700788, resistant to methicillin and intermediate susceptibility to vancomycin. [i] Staphylococcus aureus ATCC 700789,
resistant to methicillin and intermediate susceptibility to vancomycin. [j] Average MIC calculated from four strains of Staphylococcus aureus (6538, 13709,
29213, and 25923). [k] Average MIC calculated from five strains of methicillin-resistant Staphylococcus aureus (700698, 43300, 700787, 700788, and
700798).
Figure 5. Biological profile of combinatorial disulfide library in Staphylococcus aureus ATCC 6538 at the indicated concentrations. See Figure 4 and text for
a description of the library format.
tions and to compounds which were active in both the SA and
MRSA (data not shown) bacterial strains.
structure 103). Type C heterodimers bear no resemblance to
the original natural product (see structure 104). Compounds
of type A were expected to be active, particularly in cases
where both homodimeric components were biologically
active (see Table 1 for antibacterial activity of homodimers).
However, since the goal of this study was to identify
structurally simpler and more potent analogues of psamma-
plin A, it was decided that compounds of this type would not
be pursued further. Instead, efforts were focused on the
Using this method, a collection of heterodimeric disulfides
with presumed antibacterial activity was identified. These
compounds fell into three general categories with represen-
tative examples shown in Figure 6. Type A heterodimers were
comprised of two similar psammaplin-like components (see
structure 102). Type B heterodimers consisted of one psam-
maplin-like unit conjugated to an aryl or alkyl component (see
4286
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4286 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 19

Combinatorial Synthesis
4280±4295
methicillin-susceptible and methicillin-resistant strains of
Staphylococcus aureus and the results are shown in Table 2.
In general, heterodimers containing aromatic or heteroaro-
matic moieties were more potent than those containing alkyl
groups as illustrated by the fact that the three most active
compounds, 105, 113, and 115, included 4-aminophenyl,
2-quinyl, and 2-furyl groups, respectively. The most potent
heterodimer, compound 105, contained a 4-aminophenyl
subunit and was approximately four-fold more active than
psammaplin A (1) as an antibacterial agent.
Interestingly, a qualitative assessment of the validity of the
library screening method could be obtained by comparing the
concentration at which a compoundꢁs activity was visibly
detectable (i.e., Figure 5) with its average MIC values against
SA and MRSA which were determined after re-synthesis. In
general, there was a good correlation with the most potent
analogues identified at the lowest concentration level (2.5 mm)
and the least potent ones identified at the highest level
(10.0 mm), as illustrated in Figure 5.
Active compounds identified as type C are shown in
Table 3. The compounds contained an aryl sulfonamide motif
linked through the disulfide bridge to an aromatic or
heteroaromatic moiety. These compounds were usually less
potent than those of type B. Like those of the previous
category (type C), however, the MIC values obtained for
these compounds compared favorably with the concentrations
at which they were identified from screening of the initial
library. Given the structural dissimilarity between these
compounds and psammaplin A, it is currently not known
whether or not they act at the same site or through the same
mechanism as the natural product. However, the structural
simplicity of these compounds coupled with the fact that
several of them (for example sulfonamides 104 and 128) are
more potent that psammaplin A, suggests that their further
investigation is warranted. Significantly, the identification of
compounds such as 104 illustrates an important feature of this
combinatorial approach which allows for the identification of
novel, biologically active heterodimeric products formed from
inactive homodimeric components (i.e., compounds 49 and
92, see Figure 3). Indeed, it would have been highly unlikely
that such a heterodimer (i.e., 104) could have been discovered
through a traditional SAR study.
OH
Br
OH
OH
Cl
O
N
O
S
S
N
H
N
H
N
102: Type A compound
HO
OH
Br
O
S
S
N
H
N
HO
NH2
103: Type B compound
N
O
O
S
S S
N
H
Me
104: Type C compound
Figure 6. Representative examples from the three types (A ± C) of bio-
logically active heterodimers identified through library screening. See text
for a description of the three types.
equally potent compound types B and C which, owing to their
significantly lower molecular weights, might be better candi-
dates for further optimization as compared to the natural
product, psammaplin A (1).
Table 2 lists the compounds of type B identified through
this screening method. Notably, most of the compounds of this
type retained the halogenated tyrosine moiety of psammapli-
n A,with chloro- and bromo-aryl groups occurring inter-
changeably (for example, see 107 and 108). Besides the
halogenated tyrosine compounds, several phenylglycine de-
rivatives (123 ± 125) were also found to be active. After
identification, the compounds of Table 2 were individually re-
synthesized by scrambling of the two homodimeric constitu-
ents using a procedure identical to that employed for the
library construction with the exception that THF was used
instead of DMSO to simplify product isolation. The desired
heterodimer was separated from the two homodimeric start-
ing materials by preparative TLC (thin-layer chromatogra-
phy) to afford the product as a spectroscopically and
chromatographically homogenous material. It should be
noted, there was some concern as to the long-term stability
of these purified heterodimers owing to their propensity
towards slow re-equilibration to a homo/heterodimer mixture.
To minimize this potential complication, all new heterodimers
were screened immediately after purification, and subse-
quently stored at low temperature (ꢁ08C) as DMSO
solutions.
Conclusion
In this article we have described a novel strategy for the
efficient construction of a large combinatorial library of
compounds based on the structure of psammaplin A (1), a
marine-derived antibacterial agent. The disulfide exchange
chemistry described (termed combinatorial exchange syn-
thesis) provided a matrix in which each well contained two
symmetrical disulfides and produced a third, heterodimeric,
compound; the antibacterial profile of which was determined
through direct screening. This rapid protocol for the con-
struction and screening of a combinatorial library led to the
identification of several structurally distinct psammaplin A
analogues, including compounds 104, 105, 113, 115, 123, and
128 (Table 4) all of which demonstrated higher antibacterial
In order to avoid duplicity, in cases where both the bromo-
and chloro-containing compounds (for example, see 107 and
108) exhibited biological activity, only the chloro analogue
was re-synthesized for these initial rescreening studies. Given
that psammaplin A (1) and the corresponding chloro-ana-
logue (25) exhibited identical antibacterial activities (see data,
Table 1), this omission was not expected to result in a void of
structure ± activity information. Once identified and re-syn-
thesized, these compounds were assayed in a panel of
Chem. Eur. J. 2001, 7, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4287 $ 17.50+.50/0
4287

K. C. Nicolaou et al.
FULL PAPER
Table 2. Type B heterodimers and corresponding MIC [mgmL 1] values.
SA[a]
6538
SA[b]
13709
SA[c]
29213
SA[d]
25923
MRSA[e]
700698
MRSA[f]
43300
MRSA[g] MRSA[h] MRSA[i]
700787 700788 700789
AVG[j]
SA
AVG[k]
MRSA LEVEL
ID[l]
See Table 1, footnotes ([a] ± [k]) for bacterial strain information. [l] ID level denotes the screening concentration at which the compound was identified as
biologically active.
activity than the natural product. Interestingly, several of
these analogues (especially compound 105) have in vitro
antibacterial properties similar to several clinically used
antibiotics, including ciprofloxacin and vancomycin. More
importantly, these newly discovered agents demonstrate
consistent antibacterial activity, even in strains resistant to
other antibiotics. In the following paper, we describe our
efforts toward the optimization of several of these hetero-
dimeric lead compounds examining issues such as potency,
toxicity and non-specific protein binding as well as probing,
through molecular design and chemical synthesis, the mech-
anism of action of this class of antibacterial agents.[13] In
general, the simplicity, generality, and efficiency of the
described combinatorial chemistry suggests its application in
other situations. Several such extensions are currently under
consideration, including construction of other focused com-
4288
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4288 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 19

Combinatorial Synthesis
4280±4295
Table 3. Type C heterodimers and corresponding MIC [mgmL 1] values.
SA[a]
6538
SA[b]
13709
SA[c]
29213
SA[d]
25923
MRSA[e]
700698
MRSA[f]
43300
MRSA[g]
700787
MRSA[h]
700788
MRSA[i]
700789
AVG[j] AVG[k]
SA MRSA
ID[l]
LEVEL
See Table 1, footnotes ([a] ± [k]) for bacterial strain information. [l] ID level denotes the screening concentration at which the compound was identified as
biologically active.
binatorial libraries based on additional natural product
OMe
H
N
H2N
MeN
scaffolds. Although rare, several other disulfide-containing
natural products have been reported. Selected examples
include the endothelin-converting enzyme inhibitor
B-90063[20] (143, Figure 7), the alkaloid polycarpine[21] (144,
Figure 7) an antitumor agent active against human colon
cancer cells, the marine natural product citorellamine[22] (145,
Figure 7) and the antibacterial agent cortamidine oxide[23]
(146, Figure 7).
In a more general sense, the present combinatorial strategy
may also be applied to the synthesis of other discovery-
oriented (i.e., non-targeted) libraries useful in both drug
discovery and chemical biology efforts. In fact, given the
mildness and tolerance of the disulfide exchange reaction, one
could envision its application to sets of highly diverse and
structurally complex building blocks selected from the myriad
of commercially available disulfides and thiols. Arguably, this
dimerization strategy should enable the construction of more
structurally diverse discovery libraries as compared to a
conventional combinatorial strategy wherein a single central
scaffold is derivatized. Moreover, one could envision that
after identification of lead structures by this combinatorial
C5H12
C5H12
O
O
S
N
S
S
S
NMe
NH
CHO
HN
OHC
NH2
N
O
O
H
MeO
143: B-90063
144: polycarpine
S
S
N
N
H
H
145: citorellamine
Br
N
N
Br
CO2H
S
N
S
N
N
+
+
-
H
O
O
-
146: cortamidine oxide
Figure 7. Structures of other selected dimeric natural products linked
through a disulfide moiety.
strategy, the disulfide linkage might be excised and replaced
with other more desirable structural units during the normal
lead optimization process.
Chem. Eur. J. 2001, 7, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4289 $ 17.50+.50/0
4289

K. C. Nicolaou et al.
FULL PAPER
Table 4. Antibacterial activities (MIC mgmL 1) of lead heterodimers identified from combinatorial library.
SA[a]
6538
SA[b]
13709
SA[c]
29213
SA[d]
25923
MRSA[e]
700698
MRSA[f] MRSA[g]
43300 700787
MRSA[h] MRSA[i] AVG[j] AVG[k]
700788 700789 SA MRSA
See Table 1, footnotes ([a] ± [k]) for bacterial strain information.
(843 mg, 1.9 equiv, 7.33 mmol) and EDC (1.29 g, 1.7 equiv, 6.56 mmol) and
the reaction was stirred at 238C for 2 h after which time it was concentrated
and the resulting oil dissolved in EtOAc (50 mL) and washed with
saturated NaHCO3 (2 Â 50 mL), 1.0 n HCl (2 Â 50 mL) and brine (1 Â
50 mL). The organic layer was then dried over MgSO4 and concentrated
to affort succinate ester 7 which was used without further purification. This
crude ester 7 was dissolved in dioxane (10 mL) and to this solution was
added a second solution of cystamine (313 mg, 0.5 equiv, 1.39 mmol) and
Et3N (0.7 mL, 2.0 equiv, 5.5 mmol) in MeOH (10 mL). The resulting
reaction mixture was stirred at 238C for 12 h after which point it was
concentrated and the crude amide produce was passed over a short silica
gel column (50% EtOAc in hexanes) to remove any unacylated or
monoacylated cystamine. Without further purification, the resulting
coupling product 8 was dissolved in CH2Cl2/MeOH 20:1 (10 mL), placed
in a pressure tube, and anhydrous HCl (10 mL of a 1.0m ether solution) was
added. The solution was then heated to 608C for 2 h, and after cooling, was
concentrated under reduced pressure to give a crude yellow oil which was
purified by column chromatography (silica gel, 10 ! 100% EtOAc in
hexanes) to afford psammaplin A (1) (918 mg, 36% over four steps). Rf ˆ
0.10 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax ˆ 3379, 1651, 1513, 1427,
Experimental Section
General: All reactions were conducted under an argon atmosphere using
anhydrous solvents, unless otherwise noted. Anhydrous THF, diethyl ether
(ether), and methylene chloride were obtained by passing them through
commercially available alumina columns. All reagents were purchased at
the highest commercial quality and used without further purification.
Reactions were monitored by thin-layer chromatography carried out on
0.25 mm E. Merck silica glass plates (60F-254) using UV light or p-
anisaldehyde solution as visualizing agents. Flash chromatography was
performed on E. Merck silica gel (60, particle size 0.0400 ± 0.063 mm).
Preparative thin-layer chromatography was conducted on 0.25 E. Merck
silica gel plates (60F-254). All solution phase combinatorial reactions were
conducted in 2 mL, 96-well Whatman Uniplates (7701 ± 5200). All reported
yields refer to chromatographically and spectroscopically (1H NMR)
homogenous materials. NMR spectra were recorded on Bruker AMX-500
or AMX-400 instruments and calibrated using residual undeuterated
solvent as an internal reference. The following abbreviations were used to
explain the multiplicities: s ˆ singlet, d ˆ doublet, t ˆ triplet, q ˆ quartet,
m ˆ multiplet, brˆ broad. Mass spectra were recorded on a VG ZAB ZSE
mass spectrometer under MALDI-FTMS conditions using DHB as matrix.
Synthesis of psammaplin A (1): A solution of 3-bromotyrosine[14] (3)
(10.0 g, 1.0 equiv, 40.7 mmol) in trifluoroacetic anhydride (70 mL) was
heated to 408C for 12 h. The excess trifluoroacetic anhydride was then
removed under reduced pressure in order to provide trifluromethyloxazo-
lone 4. This unstable intermediate was dissolved in 70% aqueous TFA
(50 mL) and allowed to stand for 6 h during which time a white precipitate
formed. This solid was isolated by vacuum filtration and washed with cold
water (2 Â 15 mL) to provide a-keto acid 5 (6.34 g, 60%) as previously
reported.[13] To a solution of a-keto acid 5 (1.0 g, 1.0 equiv, 3.86 mmol) in
ethanol (15 mL) was added O-(tetrahydro-2H-pyran-2-yl)-hydroxylamine
(903 mg, 2.0 equiv, 0.72 mmol) and the reaction mixture was stirred at 238C
for 12 h. The reaction mixture was then concentrated under reduced
pressure and the resulting oil (9) was dissolved in EtOAc (50 mL) and
washed with 1.0n HCl (2 Â 50 mL) and brine (1 Â 50 mL). The organic
layer was then dried over MgSO4 and concentrated to afford acid 6 which
was utilized without further purification. This crude acid 6 was dissolved in
dioxane (10 mL) and to this solution was added N-hydroxylsuccinimide
1
1360, 1279, 1212, 1040, 1007 cm 1; H NMR (400 MHz, CD3OD): d ˆ 7.25
(d, J ˆ 1.8 Hz, 2H), 6.96 (dd, J ˆ 8.5, 2.0 Hz, 1H), 6.65 (d, J ˆ 8.2 Hz, 1H),
3.68 (s, 2H), 3.40 (t, J ˆ 6.8 Hz, 2H), 2.68 (t, J ˆ 6.8 Hz, 2H); 13C NMR
(100 MHz, CD3OD): d ˆ 166.0, 153.8, 153.3, 134.6, 130.7, 130.6, 117.2, 110.6,
39.7, 38.6, 28.8; HRMS (MALDI-FTMS): calcd for C22H24Br2N4O6S2
‡
[M‡Na] : 684.9396, found: 684.9426.
Homodimers 13 ± 36 were synthesized in an identical fashion to that
described for psammaplin A and selected physical data for representative
members is given below.
Compound 13: Rf ˆ 0.25 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3350, 2917, 1652, 1528, 1451, 1428, 1203, 1103, 1051 cm
;
1H NMR
(400 MHz, CDCl3): d ˆ 7.14 (d, J ˆ 8.1 Hz, 4H), 7.09 (t, J ˆ 7.9 Hz, 4H), 7.01
(t, J ˆ 7.9 Hz, 2H), 3.80 (s, 4H), 3.38 (t, J ˆ 6.7 Hz, 4H), 2.66 (t, J ˆ 6.7 Hz,
4H); 13C NMR (100 MHz, CDCl3): d ˆ 166.1, 153.3, 138.2, 130.2, 129.5,
127.4, 39.7, 38.6, 30.2; HRMS (MALDI-FTMS): calcd for C22H26N4O4S2
‡
[M‡Na] : 497.1288, found: 497.1284.
Compound 14: Rf ˆ 0.71 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3335, 2935, 1667, 1520, 1452, 1213, 1045 cm 1; 1H NMR (400 MHz, CDCl3):
4290
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001 0947-6539/01/0719-4290 $ 17.50+.50/0 Chem. Eur. J. 2001, 7, No. 19

Combinatorial Synthesis
4280±4295
d ˆ 7.30 ± 7.22 (m, 8H), 7.17 (t, J ˆ 6.2 Hz, 2H), 7.08 (t, J ˆ 5.6 Hz, 2H), 3.99
(s, 6H), 3.91 (s, 4H), 3.60 (q, J ˆ 6.4 Hz, 4H), 2.80 (t, J ˆ 6.4 Hz, 4H);
13C NMR (100 MHz, CDCl3): d ˆ 162.7, 151.7, 136.1, 129.2, 126.4, 126.3,
62.9, 38.2, 37.5, 29.7; HRMS (MALDI-FTMS): calcd for C24H30N4O4S2
Compound 25: Rf ˆ 0.11 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3409, 1655, 1512, 1427, 1287, 1215, 1013, 950 cm
;
1H NMR (400 MHz,
CD3OD): d ˆ 7.10 (d, J ˆ 1.8 Hz, 2H), 6.93 (dd, J ˆ 8.2, 2.0 Hz, 2H), 6.67 (d,
J ˆ 8.2 Hz, 2H), 3.69 (s, 4H), 3.42 (t, J ˆ 7.0 Hz, 4H), 2.69 (t, J ˆ 6.8 Hz,
4H); 13C NMR (100 MHz, CD3OD): d ˆ 166.0, 153.2, 152.7, 131.6, 130.4,
129.8, 121.4, 117.5, 102.6, 39.7, 38.6, 28.9; HRMS (MALDI-FTMS): calcd for
‡
[M‡Na] : 525.1601, found: 525.1585.
Compound 15: Rf ˆ 0.74 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
‡
C22H24Cl2N4O6S2 [M‡Na] : 597.0406, found: 597.0406.
1666, 1519, 1020, 737, 690 cm 1; 1H NMR (400 MHz, CDCl3): d ˆ 7.36 ± 7.18
(m, 20H), 7.08 (m, 2H), 5.24 (s, 4H), 3.99 (s, 4H), 3.60 (m, 4H), 2.76 (m,
4H); 13C NMR (100 MHz, CDCl3): d ˆ 162.9, 153.2, 137.3, 136.2, 130.1,
128.5, 128.4, 128.1, 128.0, 126.4, 39.2, 38.1, 30.0; HRMS (MALDI-FTMS):
Compound 26: Rf ˆ 0.14 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3388, 2938, 2360, 1660, 1525, 1427, 1047, 909, 732 cm 1; 1H NMR (400 MHz,
CDCl3): d ˆ 7.21 (d, J ˆ 2.0 Hz, 2H), 7.12 (t, J ˆ 5.8 Hz, 2H), 7.05 (dd, J ˆ
8.2, 1.8 Hz, 2H), 6.83 (d, J ˆ 8.5 Hz, 2H), 6.19 (brs, 2H), 3.99 (s, 6H), 3.79
(s, 4H), 3.59 (q, J ˆ 6.2 Hz, 4H), 2.78 (t, J ˆ 6.5 Hz 4H); 13C NMR
(100 MHz, CDCl3): d ˆ 162.7, 151.4, 150.2, 129.7, 129.2, 129.0, 119.8, 116.2,
63.1, 38.3, 37.5, 28.6; HRMS (MALDI-FTMS): calcd for C24H28Cl2N4O6S2
‡
calcd for C36H38N4O4S2 [M‡Na] : 677.227, found: 677.2250.
Compound 16: Rf ˆ 0.05 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3307, 2920, 1660, 1508, 1438, 1221, 1018, 670 cm
;
1H NMR (400 MHz,
CD3OD): d ˆ 7.30 (dd, J ˆ 8.8, 5.3 Hz, 4H), 7.16 (m, 2H), 6.93 (t, J ˆ 8.8 Hz,
4H), 3.93 (s, 4H), 3.62 (q, J ˆ 6.2H, 4H), 2.82 (t, J ˆ 6.2 Hz, 4H); 13C NMR
(100 MHz, CD3OD): d ˆ 164.2, 161.8, 153.2, 134.2, 132.0, 116.1, 39.7, 38.6,
‡
[M‡H] : 603.0900, found: 603.0892.
Compound 27: Rf ˆ 0.03 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3384, 1660, 1502, 1437, 1365, 1287, 1208, 1012 cm
;
1H NMR (400 MHz,
‡
29.3; HRMS (MALDI-FTMS): calcd for C22H24F2N4O4S2 [M‡Na] :
CDCl3): d ˆ 7.35 ± 7.30 (m, 10H), 7.23 (m, 2H), 7.07 ± 7.05 (m, 4H), 6.83 (dd,
J ˆ 8.2, 3.2 Hz, 2H), 5.21 (d, J ˆ 2.9 Hz, 2H), 3.82 (brs, 4H), 3.58 (m, 4H),
2.75 (m, 4H); 13C NMR (100 MHz, CDCl3): d ˆ 162.6, 151.9, 150.1, 136.5,
129.8, 129.4, 129.1, 128.6, 128.3, 128.2, 119.7, 126.1, 38.3, 37.4, 28.8; HRMS
533.1099, found: 533.1080.
Compound 17: Rf ˆ 0.25 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3409, 2939, 1670, 1514, 1431, 1222, 1046, 914, 823, 732 cm
;
1H NMR
‡
(400 MHz, CDCl3): d ˆ 7.25 (m, 4H), 7.08 (m, 2H), 6.91 (t, J ˆ 8.8 Hz, 4H),
3.98 (s, 6H), 3.86 (s, 4H), 3.61 (q, J ˆ 6.5 Hz, 4H), 2.80 (t, J ˆ 6.4 Hz, 4H);
13C NMR (100 MHz, CDCl3): d ˆ 162.7, 160.3, 151.6, 131.8, 130.8, 115.0,
63.0, 38.2, 37.6, 28.9; HRMS (MALDI-FTMS): calcd for C24H28F2N4O4S2
(MALDI-FTMS): calcd for C36H36Cl2N4O6S2 [M‡Na] : 777.1345, found:
777.1349.
Compound 28: Rf ˆ 0.05 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3381, 1659, 1527, 1424, 1215, 1045, 910, 731, 671 cm 1; 1H NMR (400 MHz,
CDCl3): d ˆ 7.37 (d, J ˆ 2.0 Hz, 2H), 7.12 (dd, J ˆ 8.2, 2.0 Hz, 2H), 7.10 (t,
J ˆ 7.64 Hz, 2H), 6.86 (d, J ˆ 8.2 Hz, 2H), 4.00 (s, 6H), 3.80 (s, 4H), 3.60 (q,
J ˆ 6.2 Hz, 4H), 2.79 (t, J ˆ 6.4 Hz, 4H); 13C NMR (100 MHz, CDCl3): d ˆ
162.6, 151.4, 150.9, 132.6, 130.1, 129.6, 115.9, 109.9, 63.1, 38.3, 37.6, 28.5;
‡
[M‡Na] : 561.1412, found: 561.1423.
Compound 18: Rf ˆ 0.73 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
1669, 1604, 1508, 1427, 1219, 1014, 821 cm 1; H NMR (400 MHz, CDCl3):
d ˆ 7.37 ± 7.21 (m, 14H), 7.05 (t, J ˆ 5.8 Hz, 2H), 6.90 (t, J ˆ 8.5 Hz, 4H),
5.21 (s, 4H), 3.90 (s, 4H), 3.59 (q, J ˆ 6.2 Hz, 4H), 2.78 (t, J ˆ 6.2 Hz, 4H);
13C NMR (100 MHz, CDCl3): d ˆ 162.5, 152.1, 136.6, 131.7, 130.9, 130.8,
128.5, 128.3, 128.1, 115.2, 114.9, 38.2, 37.5, 29.2; HRMS (MALDI-FTMS):
‡
HRMS (MALDI-FTMS): calcd for C24H28Br2N4O6S2 [M‡Na] : 712.9709,
found: 712.9687.
Compound 29: Rf ˆ 0.13 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
‡
1
3375, 2360, 1660, 1526, 1423, 1016 cm 1; H NMR (400 MHz, CDCl3): d ˆ
7.39 ± 7.30 (m, 8H), 7.14 (t, J ˆ 5.8 Hz, 2H), 7.09 (dd, J ˆ 8.2, 1.8 Hz, 2H),
6.82 (d, J ˆ 8.0 Hz, 2H), 5.22 (s, 4H), 3.83 (s, 4H), 3.58 (q, J ˆ 6.5 Hz, 4H),
2.74 (t, J ˆ 6.2 Hz, 4H); 13C NMR (100 MHz, CDCl3): d ˆ 162.7, 152.0,
151.5, 136.5, 132.9, 130.1, 129.7, 128.6, 128.3, 128.1, 116.8, 110.6, 38.6, 37.6,
calcd for C36H36F2N4O4S2 [M‡H] : 691.2219, found: 691.2237.
Compound 20: Rf ˆ 0.75 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3404, 2937, 1669, 1520, 1212, 1091, 1046, 915, 804, 730 cm
;
1H NMR
(400 MHz, CDCl3): d ˆ 7.23 ± 7.18 (brs, 8H), 7.07 (m, 2H), 3.98 (s, 6H), 3.86
(s, 4H), 3.61 (q, J ˆ 6.4 Hz, 4H), 2.80 (t, J ˆ 6.4 Hz, 4H); 13C NMR
(100 MHz, CDCl3): d ˆ 162.5, 151.3, 134.6, 132.2, 130.6, 128.5, 63.1, 38.2,
37.6, 29.1.
‡
28.5; HRMS (MALDI-FTMS): calcd for C36H36Br2N4O6S2 [M‡Na] :
865.0335, found: 865.0327.
Compound 30: Rf ˆ 0.20 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
Compound 21: Rf ˆ 0.74 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3397, 1734, 1639, 1458, 1381, 1229 cm 1; 1H NMR (400 MHz, CD3OD): d ˆ
7.28 (s, 4H), 3.68 (s, 4H), 3.42 (t, J ˆ 6.8 Hz, 2H), 2.70 (q, J ˆ 6.7 Hz, 2H);
13C NMR (100 MHz, CD3OD): d ˆ 165.8, 152.7, 150.8, 134.1, 132.4, 112.1,
39.8, 38.7, 28.6; HRMS (MALDI-FTMS): calcd for C22H22Br4N4O6S2
1669, 1518, 1486, 1436, 1360, 1208, 1015 cm 1; 1H NMR (400 MHz, CDCl3):
d ˆ 7.38 ± 7.33 (m, 6H), 7.28 (d, J ˆ 7.3 Hz, 4H), 7.21 (m, 8H), 7.10 (t, J ˆ
5.9 Hz, 2H), 5.22 (s, 4H), 3.92 (s, 4H), 3.60 (q, J ˆ 6.2 Hz, 4H), 2.79 (t, J ˆ
6.2 Hz, 4H); 13C NMR (100 MHz, CDCl3): d ˆ 162.9, 152.0, 136.5, 134.5,
132.2, 130.5, 128.7, 128.6, 128.4, 128.2, 77.6, 34.1, 33.3, 29.2; HRMS
‡
[M‡Na] : 840.7606, found: 840.7606.
Compound 31: Rf ˆ 0.21 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
‡
(MALDI-FTMS): calcd for C36H36Cl2N4O4S2 [M‡Na] : 745.1447, found:
1
3394, 1665, 1526, 1474, 1046, 910, 734 cm 1; H NMR (400 MHz, CDCl3):
d ˆ 7.35 (s, 4H), 7.09 (t, J ˆ 5.8 Hz, 2H), 3.99 (s, 6H), 3.76 (s, 4H), 3.61 (q,
J ˆ 6.2 Hz, 4H), 2.81 (t, J ˆ 6.4 Hz, 4H); 13C NMR (100 MHz, CDCl3): d ˆ
162.4, 150.8, 148.1, 132.7, 130.6, 109.7, 63.2, 38.3, 37.5, 28.2; HRMS
745.1478.
Compound 22: Rf ˆ 0.20 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3369, 2920, 1739, 1677, 1648, 1510, 1458, 1372, 1205, 1128, 1090 cm
;
1H NMR (400 MHz, CD3OD): d ˆ 6.95 (d, J ˆ 8.5 Hz, 4H), 6.52 (d, J ˆ
8.8 Hz, 4H), 3.68 (s, 4H), 3.38 (t, J ˆ 6.8 Hz, 4H), 2.66 (t, J ˆ 6.8 Hz, 4H);
13C NMR (100 MHz, CD3OD): d ˆ 163.6, 156.9, 132.1, 131.3, 128.9, 116.2,
39.7, 38.6, 29.2; HRMS (MALDI-FTMS): calcd for C22H26N4O6S2
‡
(MALDI-FTMS): calcd for C24H26Br4N4O6S2 [M‡Na] : 868.7919, found:
868.7882.
Compound 32: Rf ˆ 0.24 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3387, 2358, 1665, 1524, 1474, 1016 cm 1; H NMR (400 MHz, CDCl3): d ˆ
7.38 ± 7.28 (m, 7H), 7.07 (t, J ˆ 5.8 Hz, 2H), 5.90 (brs, 2H), 5.21 (s, 4H), 3.80
(s, 4H), 3.61 (q, J ˆ 6.2 Hz, 4H), 2.79 (t, J ˆ 6.2 Hz, 4H); 13C NMR
(100 MHz, CDCl3): d ˆ 162.3, 151.3, 148.0, 136.3, 132.8, 130.6, 128.7, 128.5,
128.4, 128.3, 128.2, 109.6, 77.6, 38.3, 37.5, 28.4; HRMS (MALDI-FTMS):
‡
[M‡Na] : 529.1186, found: 529.1189.
Compound 23: Rf ˆ 0.30 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3413, 1651, 1513, 1437, 1355, 1222, 1040, 1012, 950 cm
;
1H NMR
(400 MHz, CDCl3): d ˆ 7.17 (t, J ˆ 5.9 Hz, 2H), 7.07 (brd, J ˆ 8.2 Hz,
4H), 6.67 (brd, J ˆ 8.5 Hz, 4H), 3.97 (s, 6H), 3.80 (s, 4H), 3.53 (q, J ˆ
6.2 Hz, 4H), 2.71 (t, J ˆ 6.2 Hz, 4H); 13C NMR (100 MHz, CDCl3): d ˆ
163.2, 154.7, 152.1, 130.3, 127.4, 115.5, 63.0, 38.4, 37.5, 28.8; HRMS
‡
calcd for C36H34Br4N4O6S2 [M‡Na] : 1020.8545, found: 1020.8589.
Compound 34: Rf ˆ 0.34 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3277, 2936, 1652, 1530, 1443, 1257, 1051, 911, 731, 691, 648 cm 1; H NMR
(400 MHz, CDCl3): d ˆ 7.61 ± 7.58 (m, 4H), 7.39 ± 7.33 (m, 6H), 3.97 (s, 6H),
6.75 (m, 2H), 3.70 (q, J ˆ 6.2 Hz, 4H), 2.88 (t, J ˆ 6.4 Hz, 4H); 13C NMR
(100 MHz, CDCl3): d ˆ 163.4, 152.5, 131.2, 130.1, 128.6, 126.7, 62.8, 38.2,
‡
(MALDI-FTMS): calcd for C24H30N4O6S2 [M‡Na] : 557.1499, found:
577.1502.
Compound 24: Rf ˆ 0.05 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
‡
1
3347, 1659, 1513, 1225, 1015, 732 cm
;
1H NMR (400 MHz, CDCl3): d ˆ
37.3; HRMS (MALDI-FTMS): calcd for C22H26N4O4S2 [M‡Na] : 497.1288,
found: 497.1283.
7.35 ± 7.28 (m, 10H), 7.17 (m, 2H), 7.08 (d, J ˆ 8.5 Hz, 4H), 6.68 (d, J ˆ
8.6 Hz, 4H), 5.21 (s, 4H), 3.87 (s, 4H), 3.52 (m, 4H), 2.65 (m, 4H); 13C NMR
(100 MHz, CDCl3): d ˆ 163.2, 154.2, 153.1, 135.9, 131.2, 128.7, 128.5, 128.3,
115.3, 38.7, 37.5, 29.1; HRMS (MALDI-FTMS): calcd for C36H38N4O6S2
Compound 35: Rf ˆ 0.40 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3283, 1653, 1525, 1443, 1360, 1231, 1016 cm 1; 1H NMR (400 MHz, CDCl3):
d ˆ 7.61 (d, J ˆ 6.4 Hz, 4H), 7.38 ± 7.28 (m, 20H), 6.64 (t, J ˆ 5.9 Hz, 2H),
5.21 (s, 4H), 3.66 (m, 4H), 2.78 (t, J ˆ 6.4 Hz, 4H); 13C NMR (100 MHz,
‡
[M‡H] : 687.2305, found: 687.2293.
Chem. Eur. J. 2001, 7, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4291 $ 17.50+.50/0
4291

K. C. Nicolaou et al.
FULL PAPER
CDCl3): d ˆ 163.4, 152.9, 137.1, 130.1, 129.6, 128.6, 128.5, 128.3, 128.1, 77.6,
d ˆ 6.89 (q, J ˆ 10.2 Hz, 8H), 3.61 (t, J ˆ 7.3 Hz, 2H), 2.97 ± 2.88 (m, 4H),
2.69 ± 2.64 (m, 4H), 2.18 ± 2.14 (m, 4H); 13C NMR (100 MHz, CD3OD): d ˆ
169.7, 137.1, 132.5, 131.9, 124.9, 55.6, 39.6, 38.1, 37.9; HRMS (MALDI-
FTMS): calcd for C22H32N6O2S2 [M‡H] : 477.2101, found: 477.2101.
‡
38.1, 37.2; HRMS (MALDI-FTMS): calcd for C34H34N4O4S2 [M‡Na] :
649.1914, found: 649.1907.
‡
Compound 36: Rf ˆ 0.45 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3338, 1659, 1514, 1444, 1227, 1046, 731 cm 1; H NMR (400 MHz, CDCl3):
d ˆ 7.10 (m, 2H), 6.85 (brs, 2H), 6.78 (brs, 4H), 5.53 (s, 2H), 3.99 (s, 6H),
3.84 (s, 6H), 3.82 (s, 4H), 3.69 (q, J ˆ 6.2 Hz, 4H), 2.80 (t, J ˆ 6.5 Hz, 4H);
13C NMR (100 MHz, CDCl3): d ˆ 162.8, 151.2, 145.1, 144.2, 127.8, 122.0,
114.2, 112.0, 62.9, 55.8, 38.2, 37.6, 29.3; HRMS (MALDI-FTMS): calcd for
Compound 44: Rf ˆ 0.15 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
2930, 2354, 1667, 1518, 1349, 1197, 1136 cm
;
1H NMR (600 MHz,
CD3OD): d ˆ 6.64 (d, J ˆ 8.8 Hz, 4H), 5.95 (d, J ˆ 8.8 Hz, 4H), 2.58 (t,
J ˆ 7.0 Hz, 2H), 1.95 (m, 2H), 1.82 (m, 2H), 1.74 ± 1.65 (m, 4H), 1.11 (m,
4H); 13C NMR (100 MHz, CD3OD): d ˆ 169.2, 148.9, 143.4, 131.8, 124.9,
55.1, 39.3, 38.1, 37.7.
‡
C26H34N4O8S2 [M‡H] : 595.1891, found: 595.1905.
Synthesis of di-amino psammaplin A analogues 11 and 12: Di-tert-butyl-
dicarbonate (974 mg, 1.1 equiv, 4.47 mmol) was added to a solution of
3-bromotyrosine (1.0 g, 1.0 equiv, 4.06 mmol) and NaHCO3 (1.71 g,
5.0 equiv, 20.3 mmol) in THF/H2O 1:1 (10 mL). The reaction mixture was
stirred at 238C for 12 h, and the THF was then removed under reduced
pressure. To the residual reaction contents was added EtOAc (20 mL) and
the solution was then acidified to pH 4.5 using 1.0n HCl. The organic layer
was washed with brine (1 Â 20 mL), dried over MgSO4, and concentrated to
afford N-Boc-3-bromotyrosine (9) which was utilized without further
purification. This product was added dissolved in dioxane (15 mL) and N-
hydroxysuccinimide (887, 1.9 equiv, 7.71 mmol) and EDC (1.49 g, 1.7 equiv,
7.60 mmol) were added. The reaction mixture was then stirred at 238C for
2 h after which time it was concentrated under reduced pressure and the
resulting oil was taken up into EtOAc (50 mL). This solution was washed
with 0.5n HCl (1 Â 50 mL), saturated NaHCO3 (2 Â 50 mL), and brine (1 Â
50 mL) before being dried over MgSO4 and concentrated to provide
succinamate ester 10 which was utilized without further purification. This
activated ester was immediately dissolved in dioxane (10 mL) and to this
solution was added a second solution of cystamine (502 mg, 0.5 equiv,
2.23 mmol) and Et3N (1.25 mL, 2.0 equiv, 8.92 mmol) in MeOH (10 mL).
The resulting reaction mixture was stirred at 238C for 12 h and then
concentrated and purified by flash chromatography (silica gel, 25% EtOAc
in hexanes) to afford di-N-Boc-amino-psammaplin A (11) (1.55 g, 46%
Compound 45: Rf ˆ 0.18 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3312, 1679, 1652, 1519, 1347, 1166, 910, 733 cm
;
1H NMR (400 MHz,
CDCl3): d ˆ 8.10 (d, J ˆ 8.5 Hz, 4H), 7.40 (d, J ˆ 8.2 Hz, 4H), 5.79 (d, J ˆ
8.2 Hz, 2H), 3.54 ± 3.38 (m, 4H), 3.17 (dd, J ˆ 13.5, 6.5 Hz, 2H), 3.08 (dd,
J ˆ 13.5, 8.2 Hz, 2H), 2.72 ± 2.69 (m, 1H), 2.59 ± 2.56 (m, 1H), 1.35 (s, 9H);
13C NMR (100 MHz, CDCl3): d ˆ 171.8, 155.8, 146.9, 144.8, 130.3, 123.5,
80.3, 55.2, 38.7, 38.2, 37.7, 28.3; HRMS (MALDI-FTMS): calcd for
‡
C32H44N6O10S2 [M‡Na] : 759.2452, found: 759.2451.
Compound 47: Rf ˆ 0.11 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
2970, 2408, 1667, 1475, 1198, 837, 800, 722 cm
;
1H NMR (600 MHz,
CD3OD): d ˆ 3.63 ± 3.61 (m, 4H), 3.47 (m, 2H), 2.84 (m, 4H), 2.15 (m, 2H),
1.02 (t, J ˆ 6.1 Hz, 12H); 13C NMR (150 MHz, CD3OD): d ˆ 169.7, 59.8,
39.7, 38.2, 31.7, 19.0, 18.0.
Synthesis of the combinatorial library (Figure 5): Prior to synthesis, eighty-
eight 2 mL, 96-well plates were positioned in an 8 Â 11 array. To each well
was added 334 mL of a DMSO/buffer 3:1 (pH 8.3) solution. Subsequently,
8 mL of a 50 mm stock solution of each homodimer was added to every well
in the row and column designated for that compound (see Figure 4 for
further clarification). Once all homodimers were dispensed, 50 mL of a
2.4 mm solution of dithiothrietol was added to every well. The plates were
individually agitated and then allowed to stand at 238C for 12 h. Once the
synthesis was complete, the products from 100 representative wells were
sampled and analyzed by HPLC for overall purity and product distribution.
See Supporting Information for further details and tabulated data.
over three steps). Rf ˆ 0.44 (silica gel, EtOAc/hexanes 1:1); IR (film):
1
nÄmax ˆ 3309, 1655, 1366, 1249, 1165, 1045, 909, 733 cm
;
1H NMR
(400 MHz, CD3OD): d ˆ 7.32 (m, 1H), 7.01 (m, 1H), 6.85 (d, J ˆ 7.6 Hz,
1H), 5.77 (m, 1H), 4.45 (m, 1H), 3.46 ± 3.39 (m, 2H), 2.91 (brs, 2H), 2.70 ±
2.45 (m, 2H); 13C NMR (100 MHz, CD3OD): d ˆ 172.2, 155.9, 151.7, 133.2,
129.8, 116.2, 110.0, 80.3, 60.4, 55.9; HRMS (MALDI-FTMS): calcd for
Antibacterial screening of combinatorial library: Initially, three series of
dilution plates were prepared from the original synthesis plates at
concentrations of 1000mm, 500mm, and 250mm in DMSO. The library was
then screened from these three sets of dilution plates in SA 6538 (grown in
Tryhcase Soy Broth, BBL 11043) and MRSA 700698 (grown in Brain
Heart Infusion broth, DIFCO 237400). In each strain and for each
concentration, the compounds were transferred to 96-well plates contain-
ing inoculated broth so as to give final compound concentrations of 10.0mm,
5.0mm, and 2.5mm. After addition of the compounds, the plates were
incubated at 378C for 24 h and then read for turbidity at 600 nm using a
Molecular Devices Spectra MAX instrument. Graphical representations of
this data were then generated and assembled as shown in Figure 5.
‡
C32H44Br2N4O8S2 [M‡Na] : 857.0859, found: 857.0881.
A portion of compound 11 (700 mg, 1.0 equiv, 0.84 mmol) was dissolved in
CH2Cl2 (25 mL). TFA (5.0 mL) was then added and the reaction mixture
was stirred at 238C for 2 h before being concentrated under reduced
pressure to afford
a crude yellow oil which was purified by flash
chromatography (silica gel, 10 ! 100% EtOAc in hexanes) to provide
di-amino-psammaplin A 12 (505 mg, 95%). Rf ˆ 0.12 (silica gel, EtOAc/
1
hexanes 1:1); IR (film): nÄmax ˆ 3393, 1642, 1236, 1045 cm
;
1H NMR
(400 MHz, CD3OD): d ˆ 7.11 (d, J ˆ 2.0 Hz, 1H), 6.80 (dd, J ˆ 8.2 Hz, 1H),
6.61 (d, J ˆ 8.2 Hz, 1H), 3.73 (t, J ˆ 7.0 Hz, 1H), 3.30 (m, 1H), 3.12 (m, 1H),
2.75 (m, 2H); 13C NMR (100 MHz, CD3OD): d ˆ 172.2, 157.6, 137.6, 133.3,
130.4, 120.1, 113.7, 58.3, 42.0, 40.3, 39.9, 30.2; HRMS (MALDI-FTMS):
Re-synthesis of biologically active heterodimers: Heterodimers identified
as active through screening of the combinatorial were re-synthesized and
purified. These compounds were synthesized in an identical fashion to the
combinatorial reactions with the exception that THF was substituted for
DMSO to facilitate work up procedures. All compounds were purified by
preparative thin-layer chromatography and spectroscopically character-
ized. Selected physical data for representative members is given below.
‡
calcd for C22H28Br2N4O4S2 [M‡Na] : 656.9811, found: 656.9804. Homo-
dimers 38 ± 48 were synthesized in an identical fashion to that described for
compounds 11 and 12 and selected physical data for representative
members is given below.
Compound 103: Rf ˆ 0.54 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
Compound 38: Rf ˆ 0.43 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3401, 1661, 1514, 1430, 1290, 1223, 1019, 826, 707 cm 1; 1H NMR (400 MHz,
CD3OD): d ˆ 7.44 (brs, 1H), 7.31 (d, J ˆ 5.4 Hz, 2H), 7.14 (dd, J ˆ 6.5,
1.3 Hz, 1H), 6.83 (d, J ˆ 6.7 Hz, 1H), 6.70 (d, J ˆ 5.2 Hz, 2H), 3.86 (s, 2H),
3.63 (t, J ˆ 5.3 Hz, 2H), 2.88 (t, J ˆ 5.2 Hz, 2H); 13C NMR (125 MHz,
CD3OD): d ˆ 165.8, 153.7, 143.9, 142.2, 134.2, 126.6, 119.3, 117.0, 110.1, 40.9,
2945, 1670, 1480, 1199, 1137 cm 1; 1H NMR (600 MHz, CD3OD): d ˆ 5.76 ±
5.69 (m, 10H), 2.48 ± 2.39 (m, 4H), 1.93 (m, 2H), 1.80 (m, 2H), 1.57 (m,
2H), 1.51 (m, 2H), 1.10 ± 1.06 (m, 4H); 13C NMR (100 MHz, CD3OD): d ˆ
169.7, 135.6, 130.5, 130.1, 128.8, 55.8, 39.5, 38.7, 37.8; HRMS (MALDI-
‡
FTMS): calcd for C22H30N4O2S2 [M‡H] : 447.1883, found: 447.1899.
‡
38.6, 28.5; HRMS (MALDI-FTMS): calcd for C17H18BrN3O3S2 [M] :
454.9967, found: 454.9967.
Compound 39: Rf ˆ 0.45 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3299, 2975, 1657, 1528, 1366, 1247, 1169, 733 cm
;
1H NMR (600 MHz,
Compound 104: Rf ˆ 0.52 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
CDCl3): d ˆ 7.24 ± 7.20 (m, 10H), 7.00 (brs, 2H), 5.60 (brs, 2H), 4.49 (m,
2H), 3.68 (m, 2H), 3.02 (m, 4H), 2.63 (m, 2H), 2.53 (m, 2H), 1.37 (s, 18H);
13C NMR (150 MHz, CDCl3): d ˆ 172.1, 155.6, 136.8, 129.4, 128.4, 126.7,
79.9, 55.8, 38.9, 38.1, 37.6, 28.3; HRMS (MALDI-FTMS): calcd for
1
3318, 1443, 1325, 1155, 1090, 820, 667 cm 1; H NMR (500 MHz, CDCl3):
d ˆ 8.93 (m, 1H), 8.18 (d, J ˆ 8.2 Hz, 1H), 8.07 (d, J ˆ 7.4 Hz, 1H), 7.68 (m,
3H), 7.57 (t, J ˆ 7.3 Hz, 1H), 7.47 (m, 1H), 7.22 (d, J ˆ 7.36 Hz, 2H), 5.10 (t,
J ˆ 5.6 Hz, 1H), 3.32 (q, J ˆ 6.2 Hz, 2H), 2.80 (t, J ˆ 5.8 Hz, 2H), 2.39 (s,
3H); 13C NMR (100 MHz, CDCl3): d ˆ 149.6, 136.8, 136.4, 129.7, 126.9,
126.8, 125.8, 125.3, 121.9, 41.9, 37.0, 21.5; HRMS (MALDI-FTMS): calcd
‡
C32H46N4O6S2 [M‡Na] : 669.2751, found: 669.2769.
Compound 42: Rf ˆ 0.05 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3412, 2360, 1675, 1201, 1143, 1018, 952 cm 1; 1H NMR (400 MHz, CD3OD):
for C18H18N2O2S3 [M‡H] : 391.0603, found: 391.0596.
‡
4292
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4292 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 19

Combinatorial Synthesis
4280±4295
Compound 105: Rf ˆ 0.18 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
6.2 Hz, 1H), 2.84 (t, J ˆ 5.8 Hz, 1H), 2.36 (s, 3H); 13C NMR (100 MHz,
CDCl3): d ˆ 164.2, 153.1, 148.1, 141.0, 129.7, 129.5, 129.2, 116.0, 114.3, 37.7,
37.4, 28.1, 11.9; HRMS (MALDI-FTMS): calcd for C16H17ClN2O4S2
3410, 1671, 1524, 1433, 1290, 1225, 1020, 831, 707 cm 1; 1H NMR (400 MHz,
CD3OD): d ˆ 7.21 (d, J ˆ 8.5 Hz, 2H), 7.11 (d, J ˆ 2.0 Hz, 1H), 6.94 (dd, J ˆ
8.2, 2.0 Hz, 1H), 6.68 (d, J ˆ 8.2 Hz, 1H), 6.54 (d, J ˆ 8.5 Hz, 2H), 3.69 (s,
2H), 3.46 (t, J ˆ 6.8 Hz, 2H), 2.71 (d, J ˆ 7.3 Hz, 2H); 13C NMR (125 MHz,
CD3OD): d ˆ 163.5, 152.1, 151.6, 150.4, 134.7, 130.2, 128.8, 116.7, 114.6,
109.0, 108.1, 38.4, 36.5, 28.1; HRMS (MALDI-FTMS): calcd for
‡
[M‡H] : 401.0391, found: 401.0376.
Compound 117: Rf ˆ 0.18 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3383, 1664, 1513, 1429, 1290, 1222, 1020, 951, 706 cm 1; 1H NMR (400 MHz,
CD3OD): d ˆ 7.11 (d, J ˆ 2.0 Hz, 1H), 6.95 (dd, J ˆ 7.0, 2.1 Hz, 1H), 6.69 (d,
J ˆ 8.2 Hz, 1H), 3.71 (s, 2H), 3.46 (t, J ˆ 6.8 Hz, 2H), 2.83 ± 2.71 (m, 6H);
13C NMR (100 MHz, CD3OD): d ˆ 166.1, 153.3, 153.2, 131.5, 130.4, 129.9,
121.4, 119.8, 117.5, 39.6, 38.8, 34.4, 28.9, 18.2; HRMS (MALDI-FTMS):
‡
C17H18ClN3O3S2 [M] : 411.0473, found: 411.0469.
Compound 106: Rf ˆ 0.15 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3381, 1654, 1591, 1503, 1431, 1360, 1196, 670 cm
;
1H NMR (500 MHz,
‡
CDCl3): d ˆ 7.43 (d, J ˆ 8.5 Hz, 2H), 7.29 (d, J ˆ 2.0 Hz, 1H), 7.13 (dd, J ˆ
8.2, 2.0 Hz, 1H), 6.88 (d, J ˆ 8.2 Hz, 1H), 6.79 (m, 1H), 6.68 (d, J ˆ 8.8 Hz,
2H), 3.82 (s, 2H), 3.64 (d, J ˆ 6.2 Hz, 2H), 2.97 (s, 6H), 2.85 (t, J ˆ 6.2 Hz,
2H); 13C NMR (125 MHz, CDCl3): d ˆ 133.7, 129.6, 129.5, 129.3, 116.0,
113.5, 40.58, 37.47, 37.35, 28.0; HRMS (MALDI-FTMS): calcd for
calcd for C14H16ClN3O3S2 [M‡H] : 396.0214, found: 396.0205.
Compound 118: Rf ˆ 0.30 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3415, 2916, 1661, 1513, 1433, 1319, 1226, 1152, 1023, 951, 706 cm 1; 1H NMR
(400 MHz, CD3OD): d ˆ 7.65 (dd, J ˆ 4.4, 0.6 Hz, 1H), 7.51 (dd, J ˆ 3.5,
1.2 Hz, 1H), 7.11 (d, J ˆ 2.3 Hz, 1H), 7.03 (dd, J ˆ 4.4, 3.8 Hz, 1H), 6.95 (dd,
J ˆ 8.2, 2.0 Hz, 1H), 6.68 (d, J ˆ 8.5 Hz, 1H), 3.71 (s, 2H), 3.42 (t, J ˆ
6.7 Hz, 2H), 3.15 (t, J ˆ 6.4 Hz, 2H), 2.67 (m, 4H); 13C NMR (100 MHz,
CD3OD): d ˆ 166.0, 154.8, 153.1, 146.1, 133.3, 133.2, 131.6, 129.9, 128.7,
120.3, 117.5, 105.5, 43.4, 39.7, 38.9, 30.1; HRMS (MALDI-FTMS): calcd for
‡
C19H22ClN3O3S2 [M‡H] : 440.0864, found: 440.0854.
Compound 107: Rf ˆ 0.16 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3381, 1656, 1531, 1426, 1209, 1015, 907, 804, 730 cm 1; 1H NMR (400 MHz,
CDCl3): d ˆ 8.30 (s, 1H), 7.41 (d, J ˆ 8.2 Hz, 2H), 7.30 (d, J ˆ 1.7 Hz, 1H),
7.13 ± 7.10 (m, 4H), 7.03 (m, 1H), 6.88 (d, J ˆ 8.5 Hz, 1H), 5.57 (s, 1H), 3.86
(s, 2H), 3.61 (q, J ˆ 6.2 Hz, 2H), 2.82 (t, J ˆ 6.2 Hz, 2H), 2.32 (s, 3H);
13C NMR (100 MHz, CDCl3): d ˆ 162.9, 152.9, 149.9, 137.6, 133.2, 129.9,
129.7, 129.5, 129.2, 128.9, 116.1, 37.7, 37.4, 28.1, 21.0; HRMS (MALDI-
‡
C17H20ClN3O5S4 [M‡Na] : 531.9866, found: 531.9871.
Compound 119: Rf ˆ 0.22 (silica gel, EtOAc/hexanes 1:1); 1H NMR
(400 MHz, CD3OD): d ˆ 7.11 (d, J ˆ 2.0 Hz, 1H), 6.94 (dd, J ˆ 7.0, 2.3 Hz,
1H), 6.68 (d, J ˆ 8.2 Hz, 1H), 4.10 (q, J ˆ 7.0 Hz, 2H), 3.70 (m, 3H), 3.46
(td, J ˆ 7.1, 2.1 Hz, 2H), 3.00 (dd, J ˆ 7.2, 5.3 Hz, 1H), 2.86 (dd, J ˆ 7.0,
7.0 Hz, 1H), 2.76 (t, J ˆ 6.7 Hz, 2H); HRMS (MALDI-FTMS): calcd for
‡
FTMS): calcd for C18H19ClN2O3S2 [M‡H] : 411.0598, found: 411.0592.
Compound 109: Rf ˆ 0.17 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
;
1H NMR (400 MHz,
‡
3381, 1654, 1532, 1429, 1208, 1017, 737, 685 cm
C16H22ClN3O5S2 [M‡H] : 436.0762, found: 436.0774.
CDCl3): d ˆ 8.34 (s, 1H), 7.51 (dm, J ˆ 8.5 Hz, 2H), 7.32 ± 7.28 (m, 3H), 7.24
(tm, J ˆ 9.2 Hz, 1H), 7.14 (dm, J ˆ 7.3 Hz, 1H), 7.20 (m, 1H), 6.87 (dd, J ˆ
8.2, 0.88 Hz, 1H), 3.86 (s, 2H), 3.61 (q, J ˆ 6.2 Hz, 2H), 2.84 (t, J ˆ 5.6 Hz,
2H); 13C NMR (100 MHz, CDCl3): d ˆ 162.9, 152.9, 149.9, 136.7, 129.7,
129.5, 129.2, 129.1, 127.9, 127.2, 119.6, 116.1, 37.7, 37.5, 28.1; HRMS
Compound 121: 1H NMR (400 MHz, CD3OD): d ˆ 7.64 (d, J ˆ 8.2 Hz, 1H),
7.26 (d, J ˆ 7.9 Hz, 1H), 7.11 (d, J ˆ 1.7 Hz, 1H), 6.94 (dd, J ˆ 8.5, 2.1 Hz,
1H), 6.68 (d, J ˆ 8.5 Hz, 1H), 3.70 (s, 3H), 3.39 (t, J ˆ 7.0 Hz, 1H), 3.30 (t,
J ˆ 7.0 Hz, 2H), 3.05 (t, J ˆ 7.0 Hz, 1H), 2.66 ± 2.58 (m, 2H), 2.50 (t, J ˆ
7.0 Hz, 2H).
‡
(MALDI-FTMS): calcd for C17H17ClN2O3S2 [M‡H] : 397.0442, found:
Compound 122: Rf ˆ 0.23 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
397.0436.
1
3384, 2919, 1660, 1502, 1431, 1321, 1152, 1020 cm
;
1H NMR (400 MHz,
Compound 110: Rf ˆ 0.13 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
CD3OD): d ˆ 7.69 (d, J ˆ 7.0 Hz, 2H), 7.11 (d, J ˆ 2.0 Hz, 1H), 6.96 ± 6.93
(m, 3H), 6.68 (d, J ˆ 8.2 Hz, 1H), 3.75 (s, 3H), 3.70 (s, 2H), 3.40 (t, J ˆ
6.5 Hz, 2H), 3.04 (t, J ˆ 6.5 Hz, 2H), 2.66 ± 2.59 (m, 4H); 13C NMR
(100 MHz, CD3OD): d ˆ 166.0, 164.6, 153.2, 152.8, 133.4, 131.5, 130.4, 130.3,
129.8, 121.4, 117.5, 115.5, 56.3, 43.1, 39.8, 39.0, 38.4, 28.9; MS (EIS): calcd for
1
3384, 1657, 1512, 1338, 1208, 907, 847, 733 cm
;
1H NMR (400 MHz,
CDCl3): d ˆ 8.15 (d, J ˆ 4.7 Hz, 2H), 7.62 (d, J ˆ 6.8 Hz, 2H), 7.29 (d, J ˆ
2.1 Hz, 1H), 7.14 (dd, J ˆ 8.2, 2.1 Hz, 2H), 6.96 (m, 1H), 6.87 (d, J ˆ 8.2 Hz,
1H), 5.57 (brs, 1H), 3.85 (s, 2H), 3.59 (q, J ˆ 6.2 Hz, 2H), 2.90 (t, J ˆ
6.4 Hz, 2H); 13C NMR (100 MHz, CDCl3): d ˆ 163.5, 152.9, 150.0, 129.7,
129.5, 129.1, 128.0, 126.1, 124.1, 119.6, 116.1, 38.0, 37.7, 28.0, 14.2; MS (ESI):
‡
C20H24ClN3O6S3 [M‡Na] : 556, found: 556.
‡
calcd for C17H16ClN3O5S2 [M‡H] : 442, found: 442.
Compound 123: Rf ˆ 0.26 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3283, 1660, 1514, 1336, 1015, 848, 739, 692 cm
;
1H NMR (400 MHz,
Compound 112: Rf ˆ 0.45 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
;
1H NMR (400 MHz,
CDCl3): d ˆ 8.15 (d, J ˆ 9.1 Hz, 2H), 7.61 (d, J ˆ 9.1 Hz, 4H), 7.42 ± 7.32 (m,
8H), 5.25 (s, 2H), 3.71 (q, J ˆ 6.2 Hz, 2H), 2.92 (t, J ˆ 6.4 Hz, 2H);
13C NMR (100 MHz, [D6]DMSO): d ˆ 163.6, 161.0, 159.0, 152.1, 151.5,
132.2, 131.4, 130.2, 128.7, 125.6, 119.4, 116.7, 116.4, 116.2, 40.4, 38.8, 37.3,
3349, 1595, 1489, 1429, 1294, 1218, 1021, 825 cm
CD3OD): d ˆ 9.12 (s, 1H), 8.32 (dd, J ˆ 8.8, 2.6 Hz, 1H), 7.81 (d, J ˆ 9.1 Hz,
1H), 7.09 (d, J ˆ 2.0 Hz, 1H), 6.93 (dd, J ˆ 8.5, 2.0 Hz, 1H), 6.65 (d, J ˆ
8.2 Hz, 1H), 3.69 (s, 2H), 3.45 (t, J ˆ 6.4 Hz, 2H), 2.92 (t, J ˆ 6.2 Hz, 2H);
13C NMR (100 MHz, CD3OD): d ˆ 169.2, 164.1, 153.3, 152.8, 146.2, 144.0,
133.3, 131.5, 130.3, 129.8, 121.1, 117.5, 39.5, 39.2, 28.9; HRMS (MALDI-
‡
28.1; HRMS (MALDI-FTMS): calcd for C23H21N3O4S2 [M‡Na] : 490.0866,
found: 490.0866.
‡
FTMS): calcd for C16H15ClN4O5S2 [M‡H] : 443.0245, found: 443.0246.
Compound 124: Rf ˆ 0.23 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3400, 1657, 1534, 1465, 1441, 1254, 1013, 811, 693 cm 1; 1H NMR (500 MHz,
CDCl3): d ˆ 7.61 ± 7.59 (m, 2H), 7.41 ± 7.33 (m, 12H), 5.23 (s, 2H), 3.70 (q,
J ˆ 7.7 Hz, 2H), 2.86 (t, J ˆ 7.7 Hz, 2H); 13C NMR (100 MHz, CDCl3): d ˆ
163.1, 152.8, 132.2, 131.1, 130.2, 129.4, 128.6, 128.5, 128.4, 128.1, 126.8, 37.6,
Compound 113: Rf ˆ 0.10 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3356, 1656, 1495, 1040, 729 cm 1; 1H NMR (400 MHz, CDCl3): d ˆ 8.89 (d,
J ˆ 4.4 Hz, 1H), 8.17 (d, J ˆ 8.2 Hz, 1H), 8.10 (d, J ˆ 7.3 Hz, 1H), 7.64 (d,
J ˆ 8.2 Hz, 1H), 7.54 (dd, J ˆ 7.6 Hz, 1H), 7.46 (dd, J ˆ 8.5, 4.4 Hz, 1H), 7.29
(d, J ˆ 1.7 Hz, 1H), 7.16 (m, 1H), 7.12 (dd, J ˆ 8.5, 1.7 Hz, 1H), 6.86 (q, J ˆ
8.2 Hz, 1H), 3.83 (s, 2H), 3.61 (q, J ˆ 6.2 Hz, 2H); 13C NMR (100 MHz,
CDCl3): d ˆ 164.1, 153.2, 150.0, 149.5, 136.7, 136.2, 129.7, 129.3, 129.4, 128.7,
128.5, 126.8, 125.6, 125.3, 121.8, 119.7, 116.1, 38.1, 36.7, 28.1; HRMS
‡
37.5; HRMS (MALDI-FTMS): calcd for C23H21BrN2O2S2 [M‡H] :
501.0301, found: 501.0287.
Compound 125: Rf ˆ 0.63 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
2932, 1665, 1541, 1409, 1276, 1005, 911, 736, 693 cm 1; 1H NMR (400 MHz,
CDCl3): d ˆ 8.42 (m, 1H), 7.90 (s, 2H), 7.63 (dd, J ˆ 7.6, 1.2 Hz, 2H), 7.39 ±
7.32 (m, 6H), 7.29 ± 7.22 (m, 2H), 5.23 (s, 2H), 3.76 (q, J ˆ 5.5 Hz, 2H), 3.66
(t, J ˆ 6.4 Hz, 2H), 3.09 (dd, J ˆ 5.8, 4.1 Hz, 2H), 1.74 (q, J ˆ 6.7 Hz, 2H),
1.01 (t, J ˆ 7.3 Hz, 3H); 13C NMR (100 MHz, CDCl3): d ˆ 163.1, 153.7,
151.2, 144.6, 137.4, 131.1, 130.1, 129.6, 128.6, 128.3, 128.2, 128.1, 127.8, 126.5,
76.8, 70.4, 40.6, 36.9, 22.3, 10.3.
‡
(MALDI-FTMS): calcd for C16H18ClN3O3S2 [M‡H] : 448.0551, found:
448.0542.
Compound 114: Rf ˆ 0.50 (silica gel, EtOAc/hexanes 1:1); 1H NMR
(400 MHz, CDCl3): d ˆ 9.03 (m, 1H), 8.28 (dd, J ˆ 6.4, 1.2 Hz, 1H), 8.21
(dd, J ˆ 5.8, 0.8 Hz, 1H), 8.02 (brs, 1H), 7.75 (d, J ˆ 5.8 Hz, 1H), 7.65 (d,
J ˆ 6.2 Hz, 1H), 7.57 ± 7.54 (m, 1H), 7.51 (s, 2H), 3.89 (s, 2H), 3.74 (q, J ˆ
5.0 Hz, 2H), 3.00 (t, J ˆ 4.9 Hz, 2H); 13C NMR (100 MHz, CDCl3): d ˆ
163.2, 151.3, 149.6, 149.0, 141.6, 136.6, 136.5, 132.3, 131.1, 128.1, 128.0, 127.0,
126.9, 125.8, 125.3, 123.9, 123.5, 122.3, 111.7, 38.1, 36.4, 27.5.
Compound 127: Rf ˆ 0.76 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3281, 2920, 1708, 1647, 1596, 1438, 1325, 1225, 1156, 1022, 952, 744 cm
;
1H NMR (400 MHz, CD3OD): d ˆ 7.56 (d, J ˆ 8.2 Hz, 2H), 7.38 (d, J ˆ
7.0 Hz, 2H), 7.24 ± 7.15 (m, 5H), 3.0 (dd, J ˆ 7.0, 7.0 Hz, 2H), 2.62 (dd, J ˆ
7.0, 7.0 Hz, 2H), 2.30 (s, 3H); 13C NMR (100 MHz, CD3OD): d ˆ 144.8,
139.0, 138.4, 130.9, 130.4, 129.0, 128.4, 128.1, 43.3, 38.9, 21.6; HRMS
Compound 115: Rf ˆ 0.43 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3383, 1658, 1531, 1428, 1219, 1017, 732 cm 1; H NMR (400 MHz, CDCl3):
d ˆ 7.31 (brs, 1H), 7.26 (brs, 1H), 7.16 (dd, J ˆ 8.5, 1.5 Hz, 1H), 7.01 (m,
1H), 6.89 (d, J ˆ 8.5 Hz, 1H), 6.40 (brs, 1H), 3.87 (s, 2H), 3.68 (q, Jˆ
‡
(MALDI-FTMS): calcd for C15H17NO2S3 [M] : 339.0415, found: 339.0416.
Chem. Eur. J. 2001, 7, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4293 $ 17.50+.50/0
4293

K. C. Nicolaou et al.
FULL PAPER
Compound 128: Rf ˆ 0.52 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
121.9, 120.0, 114.3, 55.6, 41.8, 37.0; HRMS (MALDI-FTMS): calcd for
1
‡
3282, 1525, 1348, 1156, 1090, 810, 733, 665, 551 cm 1; H NMR (400 MHz,
CDCl3): d ˆ 8.32 (brs, 1H), 8.05 (d, J ˆ 8.2 Hz, 1H), 7.78 (d, J ˆ 7.9 Hz, 1H),
7.70 (d, J ˆ 7.9 Hz, 2H), 7.49 (t, J ˆ 8.2 Hz, 1H), 7.29 ± 7.26 (m, 2H), 4.93 (t,
J ˆ 6.2 Hz, 1H), 3.26 (q, J ˆ 6.2 Hz, 2H), 2.82 (t, J ˆ 6.4 Hz, 2H), 2.43 (s,
3H); 13C NMR (100 MHz, CDCl3): d ˆ 148.7, 143.8, 139.6, 136.7, 132.7,
129.9, 129.8, 127.0, 121.8, 121.6, 41.5, 38.0, 21.5; HRMS (MALDI-FTMS):
C18H18N2O3S3 [M‡H] : 407.0552, found: 407.0558.
Compound 137: Rf ˆ 0.63 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1443, 1319, 1225, 1149, 1048 cm 1; 1H NMR (400 MHz, CDCl3): d ˆ 7.73 (d,
J ˆ 8.2 Hz, 2H), 7.32 (d, J ˆ 8.2 Hz, 2H), 7.26 (s, 1H), 6.35 (d, J ˆ 1.2 Hz,
1H), 3.31 (q, J ˆ 6.5 Hz, 2H), 2.77 (t, J ˆ 6.2 Hz, 2H), 2.44 (s, 3H), 2.30 (s,
3H); 13C NMR (100 MHz, CDCl3): d ˆ 155.8, 143.6, 141.2, 136.9, 129.8,
127.1, 114.1, 112.5, 41.1, 37.8, 21.5, 11.9.
‡
calcd for C15H16N2O4S3 [M‡Na] : 407.0164, found: 407.0158.
Compound 129: Rf ˆ 0.22 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
Compound 138: Rf ˆ 0.74 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
1
3368, 1606, 1441, 1320, 1164, 1085 cm 1; H NMR (400 MHz, CDCl3): d ˆ
7.71 (d, J ˆ 8.2 Hz, 2H), 7.31 ± 7.29 (m, 3H), 7.16 (td, J ˆ 8.2, 1.8 Hz, 1H),
6.72 (dd, J ˆ 7.9, 1.2 Hz, 1H), 6.65 (td, J ˆ 7.3, 1.2 Hz, 1H), 4.85 (t, J ˆ
5.8 Hz, 1H), 4.36 (brs, 2H), 3.33 (q, J ˆ 6.2 Hz, 2H), 2.77 (t, J ˆ 6.4 Hz,
2H), 2.43 (s, 3H); 13C NMR (100 MHz, CDCl3): d ˆ 148.0, 143.6, 136.9,
135.5, 131.5, 129.7, 127.0, 118.6, 117.7, 115.8, 41.2, 37.4, 21.5; HRMS
3096, 2864, 1487, 1406, 1331, 1226, 1154, 1020, 951, 853, 806, 722, 586 cm
;
1H NMR (400 MHz, CDCl3): d ˆ 7.59 ± 7.55 (m, 2H), 7.38 (d, J ˆ 8.2 Hz,
2H), 7.13 (d, J ˆ 8.2 Hz, 2H), 7.07 (dd, J ˆ 5.0, 3.8 Hz, 1H), 4.93 (brt, J ˆ
5.9 Hz, 1H), 3.36 (q, J ˆ 6.2 Hz, 2H), 2.79 (t, J ˆ 6.2 Hz, 2H), 2.34 (s, 3H);
13C NMR (100 MHz, CDCl3): d ˆ 137.9, 132.9, 132.2, 132.0, 130.0, 129.2,
127.4, 41.4, 37.5, 21.1; HRMS (MALDI-FTMS): calcd for C13H15NO2S4
‡
‡
(MALDI-FTMS): calcd for C15H18N2O2S3 [M‡H] : 355.0592, found:
[M] : 344.9980, found: 344.9977.
355.0592.
Compound 139: Rf ˆ 0.87 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
2926, 2858, 1528, 1461, 1347, 1269, 1121, 1065, 876, 801, 734 cm 1; 1H NMR
(400 MHz, CDCl3): d ˆ 8.42 (t, J ˆ 2.0 Hz, 1H), 8.04 (ddd, J ˆ 8.2, 2.1,
0.8 Hz, 1H), 7.81 (d, J ˆ 7.9 Hz, 1H), 7.49 (t, J ˆ 7.9 Hz, 1H), 2.77 (t, J ˆ
7.4 Hz, 2H), 1.66 (quint, J ˆ 7.6 Hz, 2H), 1.39 ± 1.23 (m, 7H), 0.86 (t, J ˆ
6.8 Hz, 3H); 13C NMR (100 MHz, CDCl3): d ˆ 148.7, 140.9, 132.1, 129.6,
121.2, 121.1, 39.1, 31.3, 28.8, 28.1, 22.5, 13.9; MS (ESI): calcd for
Compound 130: Rf ˆ 0.72 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
3418, 1647, 1489, 1440, 1325, 1156, 1022, 809, 662 cm 1; 1H NMR (400 MHz,
CDCl3): d ˆ 7.76 (d, J ˆ 6.4 Hz, 2H), 7.44 (d, J ˆ 6.4 Hz, 2H), 7.36 (d, J ˆ
6.4 Hz, 2H), 7.20 (d, J ˆ 6.5 Hz, 2H), 4.81 (m, 1H), 3.35 (q, J ˆ 5.0 Hz, 2H),
2.82 (t, J ˆ 5.0 Hz, 2H), 2.51 (s, 3H), 2.42 (s, 3H); 13C NMR (100 MHz,
CDCl3): d ˆ 143.9, 138.3, 137.3, 133.5, 130.4, 130.2, 129.6, 127.5, 41.6, 38.1,
‡
‡
C12H17NO2S2 [M H] : 270, found: 270.
21.9, 21.5; HRMS (MALDI-FTMS): calcd for C16H19NO2S3 [M‡H] :
353.0572, found: 353.0577.
Determination of MIC values for individual compounds: After re-syn-
thesis, the minimum inhibitory concentration (MIC) values of chromato-
graphically and spectroscopically homogenous compounds were deter-
mined in a panel of bacterial strains using the two-fold serial dilution
technique in 96-well plates. The concentration of the initial dilution was
Compound 132: Rf ˆ 0.25 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1595, 1525, 1347, 1259, 1152, 1096, 1025 cm 1; 1H NMR (500 MHz, CDCl3):
d ˆ 8.31 (s, 1H), 8.06 (d, J ˆ 5.3 Hz, 1H), 7.76 (m, 3H), 7.50 (t, J ˆ 5.3 Hz,
1H), 7.25 (s, 1H), 6.94 (d, J ˆ 5.6 Hz, 2H); 13C NMR (100 MHz, CDCl3):
d ˆ 163.1, 140.3, 132.7, 131.2, 129.9, 129.2, 121.8, 121.6, 114.3, 55.6, 41.4, 38.0;
1
50 mgmL and a total of 11 two-fold dilutions were performed. The MIC
value was taken to be the lowest concentration that visible inhibited growth
of the microorganism as determined by visual inspection. SA 6538,
SA 29313, MRSA 700787, MRSA 700788 and MRSA 700789 were grown
in Tryhcase Soy Broth (BBL 11043). SA 700698 was grown in Brain Heart
Infusion broth (Difco 237400). SA 13709 and MRSA 33591 were grown in
Nutrient broth (Difco 233000). SA 25923 and MRSA 43300 were grown in
Mueller Hinton broth (Difco 0757).
‡
HRMS (MALDI-FTMS): calcd for C15H16N2O5S3 [M‡Na] : 423.0114,
found: 423.0130.
Compound 133: Rf ˆ 0.25 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3279, 1524, 1386, 1163, 1071, 835, 791, 731 cm
;
1H NMR (400 MHz,
CDCl3): d ˆ 9.04 (dd, J ˆ 4.4, 1.7 Hz, 1H), 8.41 (dd, J ˆ 7.3, 1.5 Hz, 1H),
8.31 (dd, J ˆ 8.2, 1.5 Hz, 1H), 8.21 (t, J ˆ 2.0 Hz, 1H), 8.09 (dd, J ˆ 8.2,
1.2 Hz, 1H), 8.01 (dm, J ˆ 7.9 Hz, 1H), 7.71 (d, J ˆ 7.9 Hz, 1H), 7.66 (t, J ˆ
7.9 Hz, 1H), 7.59 (dd, J ˆ 8.5, 4.4 Hz, 1H), 7.45 (t, J ˆ 7.9 Hz, 1H), 3.21 (q,
J ˆ 6.5 Hz, 2H), 2.80 (t, J ˆ 6.2 Hz, 2H); 13C NMR (100 MHz, CDCl3): d ˆ
151.4, 143.1, 139.5, 137.1, 135.6, 133.5, 132.6, 131.1, 129.9, 128.8, 125.7, 122.4,
121.7, 121.5, 41.7, 37.8; HRMS (MALDI-FTMS): calcd for C17H15N3O4S3
Acknowledgements
‡
We would like to thank Professor G. Benz of the Bayer AG for a generous
gift of ciprofloxacin. Financial support for this work was provided by The
Skaggs Institute for Chemical Biology and the National Institutes of Health
(USA), fellowships from the American Chemical Society, Division of
Organic Chemistry (sponsored by Hoechst Marion Roussel to R.H.), the
American Chemical Society, Division of Medicinal Chemistry (to R.H.),
the Department of Defense (to J.P.), the American Chemical Society,
Division of Medicinal Chemistry (to J.P.), the Ministerio de Educacion y
Cultura, Spain (to S.B.) and the Skaggs Institute for Research (to R.H.),
and grants from Amgen, Bayer AG, Boehringer-Ingelheim, Glaxo,
Hoffmann-La Roche, DuPont, Merck, Novartis, Pfizer, Schering Plough,
and Array Biopharma.
[M‡H] : 422.0303, found: 422.0303.
Compound 134: Rf ˆ 0.38 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
2961, 1542, 1406, 1327, 1277, 1163, 1010, 910, 834 cm 1; 1H NMR (400 MHz,
CDCl3): d ˆ 8.99 (dd, J ˆ 4.4, 1.8 Hz, 1H), 8.44 (dd, J ˆ 7.4, 1.5 Hz, 1H),
8.32 (s, 2H), 8.28 (dd, J ˆ 8.5, 1.8 Hz, 1H), 8.06 (dd, J ˆ 8.2, 1.5 Hz, 1H),
7.65 (dd, J ˆ 8.2, 7.4 Hz, 1H), 7.55 (dd, J ˆ 8.2, 4.1 Hz, 1H), 7.49 (m, 1H),
3.78 (d, J ˆ 6.4 Hz, 2H), 3.30 (q, J ˆ 5.6 Hz, 2H), 2.84 (t, J ˆ 5.8 Hz, 2H),
2.12 ± 2.04 (m, 1H), 1.02 (d, J ˆ 6.8 Hz, 6H); 13C NMR (100 MHz, CDCl3):
d ˆ 160.9, 151.6, 151.3, 144.9, 143.3, 136.8, 136.6, 133.3, 130.9, 125.6, 122.3,
120.0, 75.4, 41.5, 38.6, 28.2, 18.9; HRMS (MALDI-FTMS): calcd for
‡
C19H22N4O3S3 [M‡H] : 451.0927, found: 451.0938.
Compound 135: Rf ˆ 0.28 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3280, 1490, 1423, 1328, 1163, 1073, 835, 791, 685, 584 cm
;
1H NMR
(400 MHz, CDCl3): d ˆ 9.05 (dd, J ˆ 4.4, 1.5 Hz, 1H), 8.43 (dd, J ˆ 7.3,
1.5 Hz, 1H), 8.30 (dd, J ˆ 8.5, 1.8 Hz, 1H), 8.08 (dd, J ˆ 8.2, 1.2 Hz, 1H),
7.66 (t, J ˆ 7.9 Hz, 1H), 7.59 (dd, J ˆ 8.2, 4.1 Hz, 1H), 7.25 (d, J ˆ 8.5 Hz,
2H), 7.05 (d, J ˆ 8.2 Hz, 2H), 6.78 (t, J ˆ 5.8 Hz, 1H), 3.23 (q, J ˆ 6.2 Hz,
2H), 2.72 (t, J ˆ 6.2 Hz, 2H), 2.30 (s, 3H); 13C NMR (100 MHz, CDCl3):
d ˆ 151.4, 143.2, 137.7, 136.9, 135.8, 133.4, 132.7, 131.1, 129.8, 129.1, 128.7,
125.7, 122.3, 41.5, 37.4, 21.0; HRMS (MALDI-FTMS): calcd for
[1] For perspectives, see: a) R. E. Dolle, J. Comb. Chem. 2000, 2, 383 ±
433; b) P. Arya, D. T. H. Chou, M.-G. Baek, Angew. Chem. 2001, 113,
2025 ± 2031; Angew. Chem. Int. Ed. 2001, 40, 339 ± 346.
[2] For reviews, see: a) D. G. Hall, S. Manku,. F. Wang, J. Comb. Chem.
2001, 3, 125 ± 150; b) C. Watson, Angew. Chem. 1999, 111, 2025 ± 2031;
Angew. Chem. Int. Ed. 1999, 38, 1903 ± 1908.
[3] For selected recent examples, see: a) K. C. Nicolaou, J. A. Pfeffer-
korn, A. J. Roecker, G.-Q. Cao, S. Barluenga, H. J. Mitchell, J. Am.
Chem. Soc. 2000, 122, 9939 ± 9953; b) K. C. Nicolaou, R. Hughes, S. K.
Cho, N. Winssinger, C. Smethurst, H. Labischinski, R. Endermann,
Angew. Chem. 2000, 112, 3981 ± 3986; Angew. Chem. Int. Ed. 2000, 39,
3823 ± 3828; c) P. Wipf, J. T. Reeves, R. Balachandran, K. A. Guiliano,
E. Hamel, B. W. Day, J. Am. Chem. Soc. 2000, 122, 9391 ± 9395;
d) D. L. Boger, B. E. Fink, M. P. Hedrick, J. Am. Chem. Soc. 2000, 122,
6382 ± 6394; e) O. de Frutos, D. P. Curran, J. Comb. Chem. 2000, 2,
‡
C18H18N2O3S3 [M‡H] : 391.0612, found: 391.0612.
Compound 136: Rf ˆ 0.23 (silica gel, EtOAc/hexanes 1:1); IR (film): nÄmax
ˆ
1
3424, 1590, 1484, 1443, 1319, 1255, 1149, 1084 cm
;
1H NMR (400 MHz,
CDCl3): d ˆ 8.93 (d, J ˆ 2.6 Hz, 1H), 8.18 (d, J ˆ 8.2 Hz, 1H), 8.07 (d, J ˆ
7.3 Hz, 1H), 7.73 (d, J ˆ 8.8 Hz, 1H), 7.68 (d, J ˆ 7.9 Hz, 1H), 7.57 (t, J ˆ
7.9 Hz, 1H), 7.47 (dd, J ˆ 8.2, 4.4 Hz, 1H), 6.89 (d, J ˆ 8.8 Hz, 2H), 3.85 (s,
3H), 3.32 (q, J ˆ 6.2 Hz, 2H), 2.82 (t, J ˆ 6.4 Hz, 2H); 13C NMR (100 MHz,
CDCl3): d ˆ 149.6, 136.4, 136.0, 131.5, 129.2, 129.1, 128.5, 126.8, 125.8, 125.3,
4294
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4294 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 19

Combinatorial Synthesis
4280±4295
639 ± 649; f) F. Wang, S. Manku, D. G. Hall, Org. Lett. 2000, 2, 1581 ±
1583; g) D. R. Dragoli, L. A. Thompson, J. OꢁBrien, J. A. Ellman, J.
Comb. Chem. 1999, 1, 534 ± 539; h) K. C. Nicolaou, N. Winssinger, D.
Vourloumis, T. Ohshima, S. Kim, J. Pfefferkorn, J.-Y. Xu, T. Li, J. Am.
Chem. Soc. 1998, 120, 10814 ± 10826; i) K. C. Nicolaou, N. Winssinger,
J. Pastor, S. Ninkovic, F. Sarabia, Y. He, D. Vourloumis, Z. Yang, T. Li,
P. Giannakaou, E. Hamel, Nature 1997, 387, 268 ± 272.
[11] For selected other combinatorial libraries based on disulfide chem-
istry, see: a) O. Ramström, J.-M. Lehn, ChemBioChem 2000, 1, 41 ± 48;
b) S. Otto, R. L. E. Furlan, J. K. M. Sanders, J. Am. Soc. Chem. 2001,
122, 12063 ± 12064; c) D. L. Kirkpatrick, S. Watson, M. Kunkel, S.
Fletcher, S. Ulhaq, G. Powis, Anticancer Drug Dev. 1999, 14, 421 ± 432.
[12] W. J. Lees, G. M. Whitesides, J. Org. Chem. 1993, 58, 642 ± 647 and
references therein.
Á
Ä
[4] a) E. Quinoa, P. Crews, Tetrahedron Lett. 1987, 28, 3229 ± 3232; b) L.
Arabshahi, F. J. Schmitz, J. Org. Chem. 1987, 52, 3584 ± 3586.
[5] D. Kim, I. S. Lee, J. H. Jung, S.-I. Yang, Arch. Pharm. Res. 1999, 22,
25 ± 29.
[13] K. C. Nicolaou, R. Hughes, J. A. Pfefferkorn, S. Barluenga, Chem.
Eur. J. 2001, 7, 4296 ± 4310, following paper.
[14] O. Hoshino, M. Murakata, K. Yamada, Bioorg. Med. Chem. Lett. 1992,
2, 1561 ± 1562.
[6] For selected references regarding the emergence of resistance to
antibiotics, see: a) D. van der Waaij, C. E. Nord, Int. J. Antimicrob.
Agents 2000, 16, 191 ± 197; b) A. Virk, J. M. Steckelberg, Mayo Clin.
Proc. 2000, 75, 200 ± 214; c) S. Stefani, A. Agodi, Int. J. Antimicrob.
Agents 2000, 13, 143 ± 153; d) S. Mendez-Alvarez, X. Perez-Hernadez,
F. Claverie-Martin, Int. Microbiol. 2000, 3, 71 ± 80; e) K. C. Nicolaou,
C. N. C. Boddy, Sci. Am. 2001, May, 54 ± 61.
[7] For other approaches to developing novel antibiotics, see: a) M. Fines,
R. Leclercq, Curr. Opin. Anti-Infect. Invest. Drugs 1999, 1, 443 ± 452;
b) J. Zhu, Expert Opin. Ther. Pat. 1999, 9, 1005 ± 1019; c) T. J.
Dougherty, J. F. Barrett, Curr. Opin. Anti-Infect. Invest. Drugs 1999,
1, 18 ± 25; d) J. Travis, J. Potempa, Biochim. Biophys. Acta 2000, 1477,
35 ± 50.
[15] S. Sano, K. Ikai, K. Katayama, K. Takesako, T. Nakamura, A.
Obayashi, Y. Ezure, H. Enomoto, J. Antibiot. 1986, 39, 1685 ± 1696.
[16] M. L. Wilson, C. J. Coscia, J. Org. Chem. 1979, 44, 301 ± 302.
[17] The structure of heterodimer 18 was confirmed by NMR spectroscopic
analysis.
[18] Normalization was based on extinction coefficients of homodimers.
[19] The three compounds in each well were expected, in this instance, to
exhibit additive, rather than synergistic effects. For an insightful
discussion on the topic of mixture screening, see R. A. Houghten, C.
Pinilla, J. R. Appel, S. E. Blondelle, C. T. Dooley, J. Eichler, A. Nefzi,
J. M. Ostresh, J. Med. Chem. 1999, 42, 3743 ± 3778.
[20] S. Takaishi, N. Tuchiya, A. Sato, T. Negishi, Y. Takamatsu, Y.
Mataushita, T. Watanabe, Y. Iijima, H. Haruyama, T. Kinoshita, M.
Tanaka, K. Kodama, J. Antibiot. 1998, 51, 805-815.
[21] H. Kang, W. Fenical, Tetrahedron Lett. 1996, 37, 2369 ± 2372.
[22] R. M. Moriarty, D. M. Roll, Y. Y. Ku, C. Nelson, C. M. Ireland,
Tetrahedron Lett. 1987, 28, 749 ± 752.
[8] D. Kim, I. S. Lee, J. H. Jung, C. O. Lee, S.-U. Choi, Anticancer Res.
1999, 19, 4085 ± 4090.
[9] J. Shin, H.-S. Lee, Y. Seo, J. R. Rho, K. W. Cho, V. J. Paul, Tetrahedron
2000, 56, 9071 ± 9077.
[10] For an excellent review discussing disulfides and other reversible
linkages and their applications in combinatorial chemistry, see: J.-M.
Lehn, Chem. Eur. J. 1999, 5, 2455 ± 2463.
[23] G. M. Nicholas, J. W. Blunt, M. H. G. Munro, J. Nat. Prod. 2001, 64,
341 ± 344.
Received: May 17, 2001 [F3264]
Chem. Eur. J. 2001, 7, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0719-4295 $ 17.50+.50/0
4295