Discovery and synthesis of namalide reveals a new anabaenopeptin scaffold and peptidase inhibitor

Article abstract of DOI:10.1021/jm201238p

The discovery, structure elucidation, and solid-phase synthesis of namalide, a marine natural product, are described. Namalide is a cyclic tetrapeptide; its macrocycle is formed by only three amino acids, with an exocyclic ureido phenylalanine moiety at its C-terminus. The absolute configuration of namalide was established, and analogs were generated through Fmoc-based solid phase peptide synthesis. We found that only natural namalide and not its analogs containing l-Lys or l-allo-Ile inhibited carboxypeptidase A at submicromolar concentrations. In parallel, an inverse virtual screening approach aimed at identifying protein targets of namalide selected carboxypeptidase A as the third highest scoring hit. Namalide represents a new anabaenopeptin-type scaffold, and its protease inhibitory activity demonstrates that the 13-membered macrolactam can exhibit similar activity as the more common hexapeptides.

Full text of DOI:10.1021/jm201238p

Article
pubs.acs.org/jmc
Discovery and Synthesis of Namalide Reveals a New
Anabaenopeptin Scaffold and Peptidase Inhibitor
Pradeep Cheruku,† Alberto Plaza,† Gianluigi Lauro,‡ Jessica Keffer,† John R. Lloyd,† Giuseppe Bifulco,*,‡
and Carole A. Bewley*,†
†Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health,
Bethesda, Maryland 20892, United States
‡Dipartimento di Scienze Farmaceutiche e Biomediche, University of Salerno, Via Ponte Don Melillo, 84084 Fisciano, Salerno, Italy
S
* Supporting Information
ABSTRACT: The discovery, structure elucidation, and solid-
phase synthesis of namalide, a marine natural product, are
described. Namalide is a cyclic tetrapeptide; its macrocycle is
formed by only three amino acids, with an exocyclic ureido
phenylalanine moiety at its C-terminus. The absolute
configuration of namalide was established, and analogs were
generated through Fmoc-based solid phase peptide synthesis.
We found that only natural namalide and not its analogs
containing L-Lys or L-allo-Ile inhibited carboxypeptidase A at
submicromolar concentrations. In parallel, an inverse virtual
screening approach aimed at identifying protein targets of namalide selected carboxypeptidase A as the third highest scoring hit.
Namalide represents a new anabaenopeptin-type scaffold, and its protease inhibitory activity demonstrates that the 13-membered
macrolactam can exhibit similar activity as the more common hexapeptides.
group of Sephadex fractions containing what appeared to be a
novel compound on the basis of its molecular weight (ESI-MS).
In addition, these fractions showed strong inhibition of the
enzyme CPA. Active fractions were combined and purified by
C12 RP-HPLC to give just 0.4 mg of the active compound
(∼90% pure), a new natural product named namalide (1,
Figure 1). Shown by HR-ESI-MS, compound 1 had a molecular
INTRODUCTION
■
The diversity of natural products coming from marine
invertebrates and possessing interesting biological activities
has been well-documented, yet discovery of new compounds
and their associated modes of action continues apace.1 To
mention only a few examples, marine natural products that can
inhibit biological targets linked to cancer,2−5 inflammation,6
and bacterial cell division7 have been reported. Here we
describe the identification and Fmoc-based solid phase
synthesis of a new ureido-containing cyclic peptide, namalide
(1), a potent inhibitor of carboxypeptidase A (CPA). Namalide
was isolated from the same collection of the marine sponge
Siliquariaspongia mirabilis that provided the anti-HIV lip-
opeptides mirabamides A−D,8 the antitumor polyketide
mirabalin,9 and the known antifungals aurantosides A and
B.10 Namalide represents a new anabaenopeptin-type scaffold
possessing a 13-membered macrolactam core. Inverse virtual
screening and molecular docking of this novel scaffold are
consistent with observed protease inhibition and selectivity.
Figure 1. Structure of namalide (left) and key HMBC and ROE
correlations (right).
formula of C31H41N5O6 (m/z 602.2955 [M+Na]+, calcd for
C31H41N5NaO6, 602.2955) requiring 14 degrees of unsatura-
tion. A 1H−13C HSQC spectrum (CD3OD) showed
correlations for four methine signals resonating between δH
3.95−4.57 and δC 55.9−61.3, suggesting that 1 was a peptide
containing four α-amino acids. 2D NMR data including HSQC,
HMBC, HOHAHA, and DQF-COSY spectra revealed the
presence of isoleucine, lysine, and two phenylalanine residues
RESULTS
■
Isolation and Structure Elucidation of Namalide (1).
Previously, we reported the discovery of three separate classes
of natural products coming from the aqueous extract of a single
collection of the marine sponge S. mirabilis.8−10 Mirabamides,
mirabalin, and aurantosides were isolated from the n-BuOH
fraction of the aqueous extract after fractionation on a Sephadex
LH-20 column eluting with MeOH, followed by HPLC
purification. During these studies, we detected a separate
Received: September 16, 2011
Published: December 14, 2011
© 2011 American Chemical Society
735
dx.doi.org/10.1021/jm201238p | J. Med. Chem. 2012, 55, 735−742

Journal of Medicinal Chemistry
Article
(Table 1). A pair of HMBC correlations from H-2Phe2 (δH 4.42)
and H-2Lys (δH 4.00) to an unassigned carbon at δC 159.4
(δH 4.57) to the carbonyl resonance at δC 173.7 (C-1Ile)
completed the linear sequence of 1. Although the molecular
formula requires 1 to be a cyclic peptide, long-range
correlations that would connect the amino group of Ile to
the carboxy group of either Lys or Phe2 were not observed in
CD3OD or DMSO-d6 spectra. Residue connectivity could be
established, however, from a ROESY spectrum of 1 in DMSO-
d6 (Table S1, Supporting Information). In particular, a strong
correlation between the amide proton at δH 7.17 (NHIle) and
the α-proton of Lys at δH 3.78 (H-2Lys) indicated that 1
comprised a three-residue macrocycle N-linked to an exocyclic
Phe residue via a ureido bridge. Therefore, namalide bears
similarities to the anabaenopeptin family of peptides but is
unprecedented in its tripeptide macrocyle.
Independent evidence of the presence and placement of the
ureido-Phe moiety was provided by ESIMSn experiments.
Fragmentation of the major ion peak at m/z 602 [M+Na]+
displayed ion fragments at m/z 437 [M+Na-Phe]+ and m/z 409
[M+Na-Phe-CO]+ corresponding to loss of Phe and Phe-CO,
respectively. Further MS4 fragmentation of the daughter ion at
m/z 409 yielded a minor ion fragment at m/z 262 [M+Na-Phe-
CO-Phe]+. These fragmentation patterns were consistent with
our NMR results.
Table 1. NMR Spectroscopic Data for Namalide (1) in
CD3OD
1
a
b
c
δC
δH (J in Hz)
HMBC
Lys
1
2
3
4
5
6
175.6
56.3
30.8
20.1
30.2
38.2
4.00 m
1, 1Urea
2
1.76, 1.68, m
1.30, 1.10, m
1.61, 1.30, m
3.63, 2.86, m
Phe1
4, 5, 1Phe1
1
172.9
55.9
d
2
4.57
1, 3, 1Ile
3
37.0
3.18, 2.96, m
1, 2, 1′, 2′
1′
139.0
130.6
129.1
127.2
d
2′, 6′
3′, 5′
4′
7.25
d
7.22
d
7.17
Ile
The absolute configurations of the alpha carbons for both
Phe residues and Ile were readily established as L from LC-MS
analyses of the L- and D-FDLA (1-fluoro-2,4-dinitrophenyl-5-L/
D-leucinamide) derivatives11 of an acid hydrolysate of 1, in
conjunction with comparison of retention times with authentic
standards. However, because of our limited material or the
limitations of the method, this analysis did not allow us to
assign confidently the configuration of the Lys residue and C-3
of Ile. It is worth noting that although all anabaenopeptin-type
compounds derived from cyanobacteria contain D-Lys
exclusively,12,13 peptides reported to contain L-Lys rather than
D have been isolated from marine sponges.14−17 To address
these unknowns and provide additional compound for
biological screening, we developed a synthetic route to provide
namalide and its stereoisomers.
Solid-Phase Synthesis of Namalide and Analogs.
When designing the synthesis, we sought to develop a route
that would be amenable to standard Fmoc-based SPPS
conditions, including on-resin construction of the urea
moiety.18 To allow coupling to the free amine of the growing
peptide chain, we must couple the exocyclic C-terminal residue
to the resin through its free carboxylic acid. A retrosynthetic
analysis of 1 suggested that macrolactamization should occur
between the amino group of Ile and the carboxy group of Lys
(Scheme 1) with chain elongation occurring through the side-
chain ε amino group of Lys.
1
173.7
61.3
35.5
26.2
10.0
15.3
2
3.95 br s
1.70 m
3
4
1.48, 1.06 m
0.80 t (7.5)
0.54 (6.7)
Urea
5
3, 4
2, 3
Me-3
1
159.4
Phe2
1
178.6
57.5
2
4.42, m
1, 3, 1Urea
3
39.8
3.15, 3.06 m
1, 1′, 2′
1′
139.6
130.7
129.2
127.6
d
2′, 6′
3′, 5′
4′
7.23
d
7.22
d
7.16
a
b
c
Recorded at 125 MHz; referenced to residual CD3OD at δ 49.1 ppm.
Recorded at 500 MHz; referenced to residual CD3OD at δ 3.30 ppm.
Proton showing HMBC correlation to indicated carbon. Over-
lapping signals.
d
suggested the presence of a ureido group that must join Phe2 to
Lys and account for the remaining CO group indicated by the
molecular formula.
Additional long-range H−C correlations from the ε-
methylene protons of lysine (δH 3.63, 2.86, δC 38.2) to the
carbonyl at δC 172.9 (C-1Phe1) and from the α-proton of Phe1
The synthesis began with preparation of the reactive 4-
nitrophenylcarbamate derivative of lysine (2) by treatment of
Scheme 1. Retrosynthesis of Namalide (1)
736
dx.doi.org/10.1021/jm201238p | J. Med. Chem. 2012, 55, 735−742

Journal of Medicinal Chemistry
Article
Scheme 2. Solid Phase Synthesis of Namalide (1)
Nε-Fmoc-protected Lys with allyl bromide, followed by 4-
nitrophenylchloroformate to give the carbamate in good yield
(Supporting Information), an approach used in the synthesis of
the cyanobacterial metabolite oscillamide Y.18 Following
deprotection of Fmoc-L-Phe-Wang resin, the free amine was
treated with 2 in the presence of Hunig’s base to construct the
urea moiety. Standard Fmoc-based SPPS using O-benzotria-
zole-N,N,N′,N′-tetramethyl-uronium-hexafluoro-phosphate
(HBTU) and hydroxybenzotriazole (HOBt) as coupling
reagents furnished the protected linear peptide on resin
(Scheme 2). Palladium (0)-mediated removal of the allyl
group, followed by Fmoc deprotection with 20% piperidine,
gave the unprotected linear precursor for on-resin cyclization.
Macrolactamization was first carried out using benzotriazol-
1-yl-oxytripyrrol-idinophosphonium hexafluorophosphate
(PyBOP) and HOBt.19 Product formation at 20 min and 1,
2, 3, 6, and 18 h time points was monitored by LC-MS analysis
of the cleavage products of 5 mg resin aliquots. We found that
conversion of the linear precursor to the natural product
occurred in the highest yield after 3 h. Longer reaction times
using PyBOP/HOBt gave rise to other unidentified products
having incorrect masses. Surprisingly, LC-MS analysis of the
cleavage mixtures showed that the formation of a product with
m/z 1158, twice that of 1, was formed in an approximate 1:1
ratio relative to the desired product during the course of the
cyclization. Proton NMR of the product showed a single set of
peaks that would account for half the detected mass, thereby
indicating the product 3 to be a symmetrical namalide dimer.
Therefore, interstrand coupling must occur during the on-resin
cyclization to lead to the formation of 3.
Macrocyclization in the presence of other coupling reagents
including bromo-tris-pyrrolidino phosphoniumhexafluorophos-
phate (PyBrOP)/HOBt, diisopropylcarbodiimide (DIC)/
HOBt, HBTU/HOBt, and HATU/HOBt was also tested. In
the presence of the stronger activator PyBrOP/HOBt, the
dimeric product was formed to the exclusion of the monomeric
one. For the remaining three coupling reagents, cyclization
occurred more slowly than with PyBOP/HOBt, and no
improvement in the monomer/dimer ratio was observed.
Subsequent cleavage of the cyclic peptide from the solid
support using TFA/H2O (9:1) followed by reverse phase
HPLC purification of the crude peptide afforded 4.6 mg of 1 in
an overall yield of 8% on a scale of 0.1 mmol.
To establish configurations at C-3 of Ile and C-2 of Lys, we
synthesized two additional namalide analogs containing L-Ile/L-
Lys and L-allo-Ile/D-Lys (4, 5, Figure 2) using the same SPPS
strategy with appropriate amino acid precursors. Comparisons
of the NMR spectra and RP-HPLC retention times of the three
synthetic peptides 1, 4, and 5 with that of the natural product
showed that the isomer containing D-Lys and L-Ile
corresponded to the natural material.
Bioactivity of Namalide and Analogs. Members of the
anabaenopeptin20 family of cyclic peptides are known to inhibit
carboxypeptidases and other proteases.21−24 Although a high-
resolution structure of an anabaenopeptin-type peptide in
complex with any carboxypeptidase has not been determined,
selectivity profiles of natural products and synthetic analogs
737
dx.doi.org/10.1021/jm201238p | J. Med. Chem. 2012, 55, 735−742

Journal of Medicinal Chemistry
Article
aromatic amino acids.26 None of these peptides inhibited either
enzyme at concentrations as high as 60 μg/mL. Together, these
results demonstrate the importance of the lysine configuration
for CPA inhibition in this new tricyclic peptide scaffold. The
lack of inhibitory activity of the des-ureido-Phe analog 7 further
suggests that namalides may inhibit through a similar mode of
binding as the larger anabaenopeptin-type compounds.
Inverse Virtual Screening and Molecular Docking of
Namalide. Because namalide represents a new natural product
scaffold, we were interested in applying our recently described
inverse virtual screening in silico approach27 using Autodock
Vina software28 to assist in identifying other possible new
targets of namalide. In this method, a database of known
protein structures is searched to identify complementary ligand
binding partners on the basis of calculated binding energies
relative to those for a set of known ligands.27 For docking
calculations, a 3-D model of namalide was prepared by
performing combined Monte Carlo conformational searches
and molecular dynamics simulations, followed by energy and
geometry minimization of the obtained structure. The
minimized model was used against a panel of 159 receptors
involved in cancer processes and whose coordinates were
extracted from the Protein Data Bank (Tables S3 and S4,
Supporting Information). For normalization of the results, a
prebuilt matrix containing the affinity values of a library of 22
natural compounds (used as “blanks”) with structural and
molecular weight properties similar to those of 1 was employed
(Table S3, Supporting Information). In particular, we
calculated an average value of binding energy for each target
receptor on all of the compounds in the matrix and then
normalized the affinity values of 1 using the equation
Figure 2. Synthetic Namalide Analogs.
provide good evidence that the C-terminal ureido amino acid
confers specificity toward different proteases.24,25 Peptides 1, 3,
and 4−7 (Figure 2) were evaluated as inhibitors of bovine
pancreas CPA using N-(4-methoxyphenylazoformyl)-
phenylalanine as a colorimetric substrate in a 96-well plate
format. The results are summarized in Table 2.
V = Vo/VR
([1])
where V is the normalized value of binding energy, Vo is the
value of binding energy before the normalization, and VR is the
average value of binding energy for each targets.
a
Table 2. Inhibition of CPA
We were gratified to find that of the 159 proteins screened,
CPA was identified as the third best hit on the basis of
normalized binding energy (Table S5, Supporting Informa-
tion).29 As shown in Figure 3, molecular docking of 1 to CPA
using Autodock 4.230 places the C-terminal carboxylate in close
proximity to the Zn2+ atom and within hydrogen bonding
distance with the guanidine groups of Arg 127 and Arg 145, key
residues in the active site and specificity pocket of CPA26,31 as
well as the side chain of Asn 144. Additional interactions are
seen between the amide of Ile and the hydroxyl of Tyr 248.
Similarly, the ureido group is involved in electrostatic
interactions with Arg 127 and Tyr 248. Despite its reduced
size relative to the pentapeptide core of anabaenopeptins, both
electrostatic and aliphatic interactions are observed between
CPA and residues in the cyclic portion of 1 in the docked
model. In particular, the side chain of D-Lys is positioned close
to Phe 279, and the amide of Ile is hydrogen-bonded to the OH
of Tyr 248. Consistent with other models and the biological
activity, the C-terminal Phe is directed toward the hydrophobic
specificity pocket formed in part by Ile247, Tyr 248, and Ala
250.
peptide
IC50 (μM)
1
3
4
5
6
7
0.25 0.03
b
na
b
na
c
nt
4.5 0.9
b
na
a
b
c
Tested in duplicate. Not active at 30 μM. Poor reproducibility and
solubility.
The synthetic version of natural namalide (1) containing D-
Lys and L-Ile was the most potent inhibitor of CPA with an
IC50 value of 250 30 nM. Peptide 4, the corresponding L-Lys
analog, was inactive at concentrations as high as 30 μM, and
analog 5 bearing L-allo-Ile/D-Lys appeared to be inactive,
although its insolubility yielded poor assay results. The linear
version of namalide, 6, showed an 18-fold reduction in activity
relative to 1 with an IC50 value of 4.5 μM. In contrast cyclic
tripeptide 7, which lacks the C-terminal exocyclic ureido Phe,
and the namalide dimer 3 were inactive. For comparison,
compounds 1, 4, 6, and 7 were tested against carboxypeptidase
U (CPU), also known as activated thrombin-activable
fibrinolysis inhibitor (TAFIa), an enzyme known to recognize
C-terminal basic amino acids in its S1 pocket,25 and α-
chymotrypsin, a serine protease that recognizes aliphatic and
To investigate the specificity of namalide for CPA, we
performed molecular docking of 4, the L-Lys-containing analog,
to CPA and of 1 to CPU using the same protocol as above. In
the lowest energy model of 4 bound to CPA, interactions
between the exocyclic Phe and its carboxylate group with CPA
are preserved. However, the inverted configuration of Lys
738
dx.doi.org/10.1021/jm201238p | J. Med. Chem. 2012, 55, 735−742

Journal of Medicinal Chemistry
Article
Figure 3. Docked model of 1 to CPA. Protein is shown as a gray surface, Zn2+ atom is shown as a magenta sphere, and residues in contact with 1 are
shown as yellow (C atoms), blue (N atoms), red (O atoms),and gray (H atoms) sticks; namalide (1) is rendered as green (C atoms), blue (N
atoms), red (O atoms), and gray (H atoms) sticks and balls.
results in a flipping of the ring that moves the ureido group
further from the Zn2+ atom and reduced interactions between
the ring amino acids and CPA (cf. Supporting Information,
Figures S1 and S2). Docking of 1 to CPU failed to give any
reasonable models with 1 bound to or near the active site (cf.
Supporting Information, Figure S3). Both CPU and carbox-
ypeptidase B are exopeptidases that preferentially cleave basic
C-terminal residues, recognizing Arg and Lys as opposed to the
aromatic residues preferred by CPA. The lack of inhibitory
activity toward CPU lends further support that this scaffold
binds peptidases in a similar fashion to the larger
anabaenopeptins.12,22
In summary, we have discovered a new mini-anabaenopeptin
scaffold from a marine sponge. Structure elucidation and
synthesis of the natural product establish the absolute
configuration of all amino acids as L, with the exception of D-
Lys. In contrast with other anabaenopeptins, the macrocycle
comprises just three rather than the typical five amino acids,
leading to a 13-membered macrolactam ring versus the usual
19-membered one. The presence of this strained ring likely
accounts for the side product of a namalide dimer formed
during the on-resin cyclization employed in the synthesis of 1.
In keeping with specificity patterns described for a few peptides
of this class, the carboxypeptidase inhibitory activity depends
on the presence of D-Lys, and the exocyclic amino acid appears
to dictate specificity. Because namalide inhibits CPA with
potencies comparable to the more common hexapeptides, it
will be interesting to evaluate other designed namalide analogs
against CPA and related hydrolases.
EXPERIMENTAL SECTION
■
Sponge Material. The sponge was collected from open reef off
Nama Island, southeast of Chuuk lagoon, in the Federated States of
Micronesia, at a depth of 50 m, on 25 August 1994. The sponge was
previously identified8 to be most closely comparable to Siliquar-
iaspongia mirabilis (de Laubenfels, 1954) (Lithistid Demospongiae:
Family Theonellidae) but lacks the characteristic nonarticulated
tetracladine desmas that characterize the genus. A voucher specimen
has been deposited at the Natural History Museum, London, United
Kingdom (BMNH 2007.7.9.1). Samples were frozen immediately after
collection and shipped on dry ice to Frederick, MD, where they were
freeze-dried and extracted with H2O (OCDN 2547).
Experimental Details for Extraction and Isolation. A 6 g
portion of the extract was partitioned with n-BuOH-H2O (1:1) to
afford a dried n-BuOH extract (0.7 g) that was fractionated on a
Sephadex LH-20 column (50 × 2.5 cm) using MeOH as mobile phase.
Fractions containing peptides were combined and the solvent removed
in vacuo to give 16 mg of a pale orange film containing the peptide and
the known compounds aurantosides,10 orange-colored polyenes.
Purification by reverse-phase HPLC (Jupiter Proteo C12, 250 × 10
mm, 90 Å, 4 μ particle size, diode array detection at 220 and 280 nm)
eluting with a linear gradient of 50−80% MeOH in 0.05% TFA in 50
min yielded compound 1 (0.4 mg, tR = 14.8 min). Trace amounts of
aurantosides remained in this final fraction, where further separation
was limited due to limited quantities of material.
Experimental Details for Solid Phase Peptide Synthesis.
Solid-phase peptide synthesis was carried out manually on a 0.1 mmol
scale for 1; a 0.05 mmol scale for 4, 5, and 7; and a 0.25 mmol scale for
6 in solid-phase peptide synthesis vessels with fritted glass supports
under N2 (1 atm) using the following conditions: resins (100−200
mesh, 0.61 meq, 0.16 g, 0.1 mmol) were immersed in CH2Cl2 (10 mL)
and agitated for 2 h to allow for swelling, drained, and washed with
DMF (2 × 5 mL). Fmoc protecting groups were removed by treating
the resin with a solution of 20% piperidine/DMF (5 mL) for 30 min,
739
dx.doi.org/10.1021/jm201238p | J. Med. Chem. 2012, 55, 735−742

Journal of Medicinal Chemistry
Article
followed by washing with DMF (5 × 5 mL), CH2Cl2 (5 × 5 mL), and
DMF (5 × 5 mL). For coupling reactions, an appropriate Fmoc-
protected amino acid (4 equiv) was preactivated in the presence of
HBTU (3 equiv) and DIPEA (8 equiv) in 10 mL of DMF for 10 min,
and the resulting solution was added to the resin and agitated under
N2 for 2 h, followed by washing with DMF (5 × 5 mL), CH2Cl2 (5 × 5
mL), and DMF (5 × 5 mL). For ureido-containing peptides, the resin
was suspended in 5 mL of DMF, and a solution of 4-nitro-
phenyloxycarbonyl-D-Lys(Fmoc)−OH allyl ester (2, Supporting
Information) (173 mg, 0.32 mmol, 3eq) in DMF (5 mL) and
DIPEA (0.11 mL, 0.6 mmol, 6eq) was added to the resin slurry,
followed by shaking for 1.5 h. The resin was then washed with DMF
(5 × 5 mL), CH2Cl2 (5 × 5 mL), and DMF (5 × 5 mL). Following
each coupling reaction, a ∼5 mg aliquot of resin was cleaved and
analyzed by LC-MS to monitor the extent of the reaction. Following
each coupling step, resin was treated with a solution of 20% Ac2O in
DMF (10 mL) for 30 min and washed with DMF (5 × 5 mL) and
CH2Cl2 (5 × 5 mL).
Cleavage From Resin. Products were cleaved from the solid
support by treatment with TFA/H2O (90:10, 10 mL) for 3 h. The
resin was filtered and washed with TFA (2 × 5 mL) and CH2Cl2 (2 ×
5 mL). The combined filtrate and washes were removed in vacuo to
give the crude peptide, which was suspended in CH3CN/H2O (1:1)
and lyophilized. Purification of the crude material was carried out by
RP HPLC, and the purity of compounds used in the biological assays
was ≥95%, as determined by RP-HPLC (Table S2 of the Supporting
Information).
Experimental Details for Other Synthetic Procedures. De-
protection of Alloc Ester. After coupling of the final amino acid
followed by capping, the resin was treated with a solution of
tetrakis(triphenylphosphine) palladium (0) (116 mg, 0.1 mmol) and
dimedone (140 mg, 1.0 mmol) in dry DMF o/n.32 The reaction vessel
was protected from light during the course of deprotection, after which
the resin was washed with CH2Cl2 (5 × 5 mL), DMF (5 × 5 mL),
0.5% DIPEA + 0.5% diethyldithiocarbamic acid sodium salt in DMF (5
× 5 mL), DMF (5 × 5 mL), and CH2Cl2 (5 × 5 mL).
On-Resin Cyclization. A solution of PyBOP (280 mg, 0.6 mmol),
HOBt (81 mg, 0.6 mmol), and DIPEA (0.1 mL, 0.6 mmol) in
CH2Cl2/DMF (1:1, 10 mL) was added to the resin and the mixture
shaken for 3 h. The resin was washed with CH2Cl2 (5 × 5 mL), DMF
(5 × 5 mL), and CH2Cl2 (5 × 5 mL).
(t = 0−5 min, 80% A, 20% B; t = 5−50 min, 10% A, 90% B, t = 51
min, 100% B (A = water, 0.1% TFA, B = methanol). Flow rate = 0.5
mL/min, tR = 48.99 min); HRMS (ESI): Calcd for C31H42N5O6: m/z,
[M + H]+ 1159.6. Found: 1159.4.
Peptide 4. 1.74 mg (yield = 6%), colorless amorphous powder;
26.5
[α]D = −23.0 (c 0.1, DMSO/MeOH 1:1); IR (film) ν
3290,
max
1
2930, 2856, 1710, 1635, 1521, 1206, 1123, 695 cm−1; H NMR (500
MHz, DMSO-d6) δ 0.57 (d, J = 6.67 Hz, 3H), 0.72 (t, J = 7.26 Hz,
3H), 0.86−0.90 (m, 1H), 0.92−1.09 (m, 2H), 1.09−1.20 (m, 1H),
1.20−1.29 (m, 4H), 1.29−1.37 (m, 1H), 1.37−1.56 (m, 4H), 1.65
(brs, 1H), 2.70−2.87 (m, 2H), 2.94−3.12 (m, 2H), 3.91 (t, J = 9.32
Hz 1H), 4.01−4.08 (m, 1H), 4.31 (q, J = 7.45, 13.3 Hz 1H), 4.43 (q, J
= 8.69, 15.96 Hz 1H), 6.22 (d, J = 8.10 Hz 1H), 6.44 (d, J = 8.19 Hz
1H), 7.14−7.23 (m, 6H), 7.23−7.32 (m, 4H), 7.46 (brs, 1H), 7.66 (d,
J = 8.97 Hz 1H), 8.01 (d, J = 9.61 Hz 1H). HPLC analysis: Jupiter 4
μm Proteo 90 Å LC Column 100 × 2 mm, 220 nm (t = 0−5 min, 80%
A, 20% B; t = 5−30 min, 20% A, 80% B, t = 35 min, 100% B. (A =
water, 0.1% TFA, B = methanol). Flow rate = 0.5 mL/min, tR = 35.89
min); HRMS (ESI): Calcd for C31H42N5O6: m/z, [M + H]+ 580.3135.
Found: 580.3145.
Peptide 5. 2.6 mg (yield = 9%), colorless amorphous powder;
26.7
[α]D = −51.6 (c 0.06, DMSO/MeOH 1:1); IR (film) νmax 3298,
1
2938, 2853, 1710, 1638, 1540, 1211, 1203, 1133, 698 cm−1; H NMR
(500 MHz, DMSO-d6) δ 0.73−0.66 (m, 6H), 1.00−1.20 (m, 3H),
1.20−1.35 (m, 2H), 1.39−1.60 (m, 3H), 1.70 (brs, 1H), 2.68−2.85
(m, 2H), 2.91 (dd, J = 7.80, 13.3 Hz, 1H), 2.98−3.10 (m, 2H), 3.43−
3.54 (m, 1H), 3.72−3.89 (m, 2H), 4.34 (dd, J = 7.83, 13.29 Hz, 1H),
4.40 (dd, J = 8.73, 15.3 Hz, 1H), 6.23 (d, J = 7.42 Hz 1H), 6.39 (d, J =
5.30 Hz, 1H), 7.05−7.08 (m, 1H), 7.08−7.37 (m, 10H), 7.45 (d, J =
7.80 Hz 1H), 7.56 (brs, 1H), 8.12 (d, J = 8.79 Hz, 1H). HPLC
analysis: Jupiter 4 μm Proteo 90 Å LC Column 100 × 2 mm, 210 nm
(t = 0−5 min, 80% A, 20% B; t = 5−45 min, 20% A, 80% B, t = 46
min, 100% B (A = water, 0.1% TFA, B = methanol). Flow rate = 0.5
mL/min, tR = 37.75 min); HRMS (ESI): Calcd for C31H42N5O6: m/z,
[M + H]+ 580.3135. Found: 580.3144.
Peptide 6. CLEAR amide resin (100−200 mesh, 0.49 meq, 0.2 g,
0.1 mmol) was immersed in CH2Cl2 (10 mL) and agitated for 2 h.
SPPS of 6 was carried out using the general protocol outlined above.
The combined filtrate and washes were removed in vacuo to give the
crude peptide, which was suspended in CH3CN/H2O (1:1) and
lyophilized. Yield: 4.5 mg (30%), colorless amorphous powder;
26.9
Enzyme Assays. Peptides were tested for inhibition of CPA and
chymotrypsin using respective colorimetric substrates N-(4-methox-
yphenylazoformyl)-Phe-OH33 and N-benzoyl-L-tyrosine ethyl ester, all
of which were purchased from Sigma. Assays were performed in 96-
well plates (CPA) and 10 mm quartz cuvette (chymotrypsin) at 25 °C,
in the presence of 1 U enzyme and 10 μL of inhibitor (typically 1 mM
in DMSO) or solvent (DMSO) in a final volume of 200 μL.
Absorbance was measured after 5 min of incubation at 350 nm for
CPA and 256 nm for chymotrypsin. For CPA, inhibition curves were
plotted and best fit to the equation % inhibition = 100/(1 +
[inhibitor]/IC50) using the program Kaleidagraph; no change in
absorbance was observed for any of the peptides in the chymotrypsin
assays. TAFIa/CPU enzyme assays were performed in a similar
manner using the TAFIa kit from American Diagnostica according to
the manufacturer’s instructions.
[α]D = +6 (c 0.15 MeOH); IR (film) νmax 3304, 2943, 2832, 2518,
1645, 1446, 1201, 1188 cm−1; 1H NMR (500 MHz, DMSO-d6) δ 0.71
(t, J = 6.40 Hz, 3H), 0.68 (d, J = 6.39 Hz, 3H), 0.93−1.14 (m, 2H),
1.14−1.30 (m, 2H), 1.34−1.60 (m, 4H), 1.63−1.78 (m, 4H), 2.23−
2.43 (m, 1H), 2.71 (t, J = 7.40 Hz, 2H), 2.81−2.93 (m, 2H), 3.02−
3.10 (m, 2H), 4.03 (t, J = 6.91 Hz 1H), 4.16 (q, J = 6.73, 12.8 Hz, 1H),
4.32−4.44 (m, 2H), 6.41 (d, J = 7.96 Hz, 1H), 6.52 (d, J = 6.76 Hz,
1H), 7.03 (d, J = 8.01 Hz, 1H), 7.19−7.30 (m, 6H), 7.12−7.19 (m,
4H), 7.57 (brs, 2H), 8.0 (t, J = 8.26 Hz, 2H). HPLC analysis: Jupiter 4
μm Proteo 90 Å LC Column 100 × 2 mm, 220 nm (t = 0−5 min, 80%
A, 20% B; t = 5−45 min, 20% A, 80% B, t = 46 min, 100% B (A =
water, 0.1% TFA, B = methanol). flow rate = 0.5 mL/min, tR = 32.13
min); HRMS (ESI): Calcd for C31H42N5O6: m/z, [M + H]+ 597.3401.
Found: 597.3412.
Peptide 7. Fmoc-Phe-Wang resin (100−200 mesh, 0.61 meq, 0.4
g, 0.25 mmol) was immersed in CH2Cl2 (10 mL) and agitated for 2 h,
and SPPS of the linear precursor was carried out using the above
protocol. The product was cleaved from the solid support by treatment
with TFA/H2O (90:10, 15 mL) for 3 h. The resin was filtered and
washed with TFA (2 × 5 mL) and CH2Cl2 (2 × 5 mL). The combined
filtrate and washes were removed in vacuo to give the linear peptide,
which was suspended in CH3CN/H2O (1:1) and lyophilized.
Cyclization: to a solution of linear peptide (20 mg, 0.15 mmol) in
dry DMF (30 mL) was added PyBOP (140 mg, 0.3 mmol), HOBt (13
mg, 0.1 mmol), and DIPEA (0.1 mL, 0.6 mmol) in CH2Cl2/DMF
(1:1, 10 mL) dropwise. The reaction mixture was allowed to stir for 4
h, and the solvents were removed in vacuo. The crude Fmoc-protected
cyclic peptide was purified by silica gel column chromatography,
followed by the removal of Fmoc in 20% piperidine/CH2Cl2 (10 mL).
Synthetic Namalide (1). 4.6 mg (yield = 8%), colorless
24.5
amorphous powder; [α]D = −41.8 (c 0.03, DMSO/MeOH 1:3);
IR (film) νmax 3299, 2928, 2866, 1713, 1645, 1541, 1207, 1203, 1133,
698 cm−1; See Table S1 for NMR data; HPLC analysis: Jupiter 4 μm
Proteo 90 Å LC Column 100 × 2 mm, 220 nm (t = 0−5 min, 80% A,
20% B; t = 5−45 min, 20% A, 80% B, t = 46 min, 100% B. (A = water,
0.1% TFA, B = methanol). Flow rate = 0.5 mL/min, tR = 36.28 min);
HRMS (ESI): Calcd for C31H42N5O6: m/z, [M + H]+ 580.3135.
Found: 580.3139.
Peptide 3 (Dimeric 1). 1H NMR (400 MHz, MeOD) δ 0.43−0.74
(m, 12H), 0.76−1.01 (m, 4H), 1.03−1.84 (m, 16H), 2.75−3.03 (m,
6H), 3.09 (m, 4H), 3.86 (brs, 2H), 3.97 (brs, 2H), 4.35 (brs, 2H), 4.49
(brs, 2H), 7.01−7.36 (m, 20H), 7.58 (brs, 2H), 7.89 (brs, 1H); HPLC
analysis: Jupiter 4 μm Proteo 90 Å LC Column 100 × 2 mm, 220 nm
740
dx.doi.org/10.1021/jm201238p | J. Med. Chem. 2012, 55, 735−742

Journal of Medicinal Chemistry
Article
24.5
Yield: 1.9 mg (10%), colorless amorphous powder; [α]D = −82 (c
0.05, DMSO/MeOH 1:1); IR (film) νmax 3304, 2943, 2832, 2518,
1713, 1645, 1532, 1446, 1416, 1201, 1188 cm−1; 1H NMR (500 MHz,
DMSO-d6) δ 0.59 (d, J = 6.37 Hz, 3H), 0.79 (t, J = 7.26 Hz, 3H),
0.96−1.08 (m, 1H), 1.09−1.20 (m, 2H), 1.37−1.47 (m, 2H), 1.47−
1.73 (m, 3H), 2.23−2.43 (m, 2H), 2.67−2.75 (m, 1H), 2.83 (dd, J =
8.64 Hz 1H), 3.05 (dd, J = 6.73 Hz, 1H), 3.20 (brs, 1H), 3.41−3.52
(m, 1H), 3.61 (d, J = 7.93 Hz 1H), 6.99 (brs, 1H), 7.1 (brs, 1H),
7.12−7.21 (m, 3H, Ar), 7.21−7.3 (m, 2H, Ar), 7.56 (d, J = 7.17 Hz,
1H), 7.81 (d, J = 7.96 Hz, 1H), 8.10 (brs, 1H), 8.68 (brs, 1H); HPLC
analysis: Jupiter 4 μm Proteo 90 Å LC Column 100 × 2 mm, 220 nm
(t = 0−5 min, 80% A, 20% B; t = 5−45 min, 20% A, 80% B, t = 46
min, 100% B. (A = water, 0.1% TFA, B = methanol). flow rate = 0.5
mL/min, tR = 30.36 min); HRMS (ESI): Calcd for C31H42N5O6: m/z,
[M + H]+ 389.2553. Found: 389.2553.
ACKNOWLEDGMENTS
■
This work was supported by the NIH Intramural Research
Program (NIDDK) and FARB (University of Salerno).
ABBREVIATIONS USED
■
CPA, carboxypeptidase A; CPU, carboxypeptidase U; DIC,
diisopropylcarbodiimide; DIPEA, N,N-diisopropylethylamine;
FDLA, 1-fluoro-2,4-dinitrophenyl-5-leucinamide; HBTU, O-
benzotriazole-N,N,N′,N′-tetramethyl-uronium-hexafluoro-phos-
phate; HOBt, hydroxybenzotriazole; PyBOP, benzotriazol-1-yl-
oxytripyrrolidinophosphonium hexafluorophosphate; PyBrOP,
bromo-tris-pyrrolidino phosphoniumhexafluorophosphate;
TAFIa, activated thrombin-activable fibrinolysis inhibitor
Inverse Virtual Screening. The chemical structure of 1 was built
and processed with Macromodel 8.5 (Schrodinger, New York, 2003).
̈
REFERENCES
■
Molecular mechanics/dynamics calculations were performed on a
quad-core Intel Xeon 3.4 GHz using Macromodel 8.5 and the OPLS
force field. The Monte Carlo multiple minimum (MCMM) method
(5000 steps) was used first to allow a full exploration of the
conformational space. Molecular dynamics simulations were per-
formed at 600 K and with a simulation time of 10 ns. A constant
dielectric term, mimicking the presence of the solvent, was used in the
calculations to reduce artifacts. Finally, an optimization (Conjugate
Gradient, 0.05 Å convergence threshold) of the structures was applied
for the identification of a possible 3-D starting model of 1 for the
subsequent steps of docking calculations. The panel of protein targets
was prepared by a search of crystallized structures in the Protein Data
Bank (Tables S3−S4, Supporting Information). Water molecules were
removed, and polar hydrogens were added with AutodockTools 1.4.5.
Molecular docking calculations were performed using Autodock-Vina
software using an exhaustiveness of 64. The selected grids focused on
presumed sites of pharmacological interest on the basis of the binding
modes of crystallized ligands in the PDB files wherever possible (Table
S4, Supporting Information).
A more accurate analysis of the interaction between 1 and CPA was
conducted with the Autodock 4.2 software package. To have an
accurate weight of the electrostatics, we derived the partial charge of
Zn = 1.136 and of the amino acids involved in the catalytic center by
DFT calculations at the B3LYP level by the 6-31G(d) basis set and
ChelpG34 method for population analysis (Gaussian 03 Package
software35). Ten calculations consisting of 256 runs were performed,
obtaining 2560 structures (256 × 10). The Lamarckian genetic
algorithm was used for dockings. An initial population of 450
randomly placed individuals, a maximum number of 10.0 × 106 energy
evaluations, and a maximum number of 8.0 × 106 generations was
taken into account. A mutation rate of 0.02 and a crossover rate of 0.8
were used. Results differing by <3.0 Å in positional root-mean-square
deviation (rmsd) were clustered together. For all of the investigated
compounds, all open-chain bonds were treated as active torsional
bonds. Autodock Vina results were analyzed with Autodock Tools
1.4.5.
(1) Blunt, J. W.; Copp, B. R.; Munro, M. H.; Northcote, P. T.;
Prinsep, M. R. Marine natural products. Nat. Prod. Rep. 2011, 28,
196−268.
(2) Feling, R. H.; Buchanan, G. O.; Mincer, T. J.; Kauffman, C. A.;
Jensen, P. R.; Fenical, W. Salinosporamide A: a highly cytotoxic
proteasome inhibitor from a novel microbial source, a marine
bacterium of the new genus salinospora. Angew. Chem., Int. Ed.
2003, 42, 355−357.
(3) Manzanares, I.; Cuevas, C.; Garcia-Nieto, R.; Marco, E.; Gago, F.
Advances in the chemistry and pharmacology of ecteinascidins, a
promising new class of anti-cancer agents. Curr. Med. Chem.: Anti-
Cancer Agents 2001, 1, 257−276.
(4) Mayer, A. M. S.; Glaser, K. B.; Cuevas, C.; Jacobs, R. S.; Kem, W.;
Little, R. D.; McIntosh, J. M.; Newman, D. J.; Potts, B. C.; Shuster, D.
E. The odyssey of marine pharmaceuticals: a current pipeline
perspective. Trends Pharmacol. Sci. 2010, 31, 255−265.
(5) Taori, K.; Paul, V. J.; Luesch, H. Structure and activity of
largazole, a potent antiproliferative agent from the Floridian marine
cyanobacterium Symploca sp. J. Am. Chem. Soc. 2008, 130, 1806−1807.
(6) Villa, F. A.; Gerwick, L. Marine natural product drug discovery:
Leads for treatment of inflammation, cancer, infections, and
neurological disorders. Immunopharm. Immunotox. 2010, 32, 228−237.
(7) Plaza, A.; Keffer, J. L.; Bifulco, G.; Lloyd, J. R.; Bewley, C. A.
Chrysophaentins A-H, antibacterial bisdiarylbutene macrocycles that
inhibit the bacterial cell division protein FtsZ. J. Am. Chem. Soc. 2010,
132, 9069−9077.
(8) Plaza, A.; Gustchina, E.; Baker, H. L.; Kelly, M.; Bewley, C. A.
Mirabamides A-D, depsipeptides from the sponge Siliquariaspongia
mirabilis that inhibit HIV-1 fusion. J. Nat. Prod. 2007, 70, 1753−1760.
(9) Plaza, A.; Baker, H. L.; Bewley, C. A. Mirabalin, [corrected] an
antitumor macrolide lactam from the marine sponge Siliquariaspongia
mirabilis. J. Nat. Prod. 2008, 71, 473−477.
(10) Matsunaga, S.; Fusetani, N.; Kato, Y.; Hirota, H. Aurantosides A
and B: cytotoxic tetramic acid glycosides from the marine sponge
Theonella sp. J. Am. Chem. Soc. 1991, 113, 9690−9692.
(11) Harada, K.; Fujii, K.; Hayashi, K.; Suzuki, M.; Ikai, Y.; Oka, H.
Application of D,L-FDLA derivatization to determination of absolute
configuration of constituent amino acids in peptide by advanced
Marfey’s method. Tetrahedron Lett. 1996, 37, 3001−3004.
(12) Rouhiainen, L.; Jokela, J.; Fewer, D. P.; Urmann, M.; Sivonen,
K. Two alternative starter modules for the non-ribosomal biosynthesis
of specific anabaenopeptin variants in Anabaena (Cyanobacteria).
Chem. Biol. 2010, 17, 265−273.
(13) Walther, T.; Arndt, H. D.; Waldmann, H. Solid-support based
total synthesis and stereochemical correction of brunsvicamide A. Org.
Lett. 2008, 10, 3199−3202.
(14) Kobayashi, J.; Sato, M.; Ishibashi, M.; Shigemori, H.; Nakamura,
T.; Ohizumi, Y. Keramamide A, a novel peptide from the Okinawan
marine sponge Theonella sp. J. Chem. Soc., Perkin Trans. 1991, 1,
2609−2611.
(15) Kobayashi, J.; Sato, M.; Murayama, T.; Ishibashi, M.; Walchi, M.
R.; Kanai, M.; Shoji, J.; Ohizumi, Y. Konbamide, a novel peptide with
ASSOCIATED CONTENT
■
S
* Supporting Information
General experimental procedures, 1H and 2D NMR spectra for
synthetic and natural products, HPLC data for compounds
used in biological assays, tables with binding energies and grid
boxes used in the calculations, and docking of 1 to CPU. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: bifulco@unisa.it, caroleb@mail.nih.gov.
741
dx.doi.org/10.1021/jm201238p | J. Med. Chem. 2012, 55, 735−742

Journal of Medicinal Chemistry
Article
calmodulin antagonistic activity from the Okinawan marine sponge
Theonella sp. J. Chem. Soc., Chem. Commun. 1991, 3, 1050−1052.
(16) Schmidt, E. W.; Harper, M. K.; Faulkner, D. J. Mozamides A
and B, cyclic peptides from a theonellid sponge from mozambique. J.
Nat. Prod. 1997, 60, 779−782.
(17) Uemoto, H.; Yahiro, Y.; Shigemori, H.; Tsuda, M.; Takao, T.;
Shimonishi, Y.; Kobayashi, J. Keramamides K and L, new cyclic
peptides containing unusual tryptophan residue from Theonella
sponge. Tetrahedron 1998, 54, 6719−6724.
(18) Marsh, I. R.; Bradley, M.; Teague, S. J. Solid-phase total
synthesis of oscillamide Y and analogues. J. Org. Chem. 1997, 62,
6199−6203.
(19) Coste, J.; Le-Nguyen, D.; Castro, B. PyBOP: A new peptide
coupling reagent devoid of toxic by-product. Tetrahedron Lett. 1990,
31, 205−208.
(20) Harada, K.; Fujii, K.; Shimada, T.; Suzuki, M.; Sano, H.; Adachi,
K.; Carmichael, W. W. 2 cyclic-peptides, anabaenopeptins, a 3rd group
of bioactive compounds from the cyanobacterium Anabaena-flos-aquae
Nrc-525−17. Tetrahedron Lett. 1995, 36, 1511−1514.
(21) Bubik, A.; Sedmak, B.; Novinec, M.; Lenarcic, B.; Lah, T. T.
Cytotoxic and peptidase inhibitory activities of selected non-
hepatotoxic cyclic peptides from cyanobacteria. Biol. Chem. 2008,
389, 1339−1346.
(22) Gesner-Apter, S.; Carmeli, S. Protease inhibitors from a water
bloom of the cyanobacterium Microcystis aeruginosa. J. Nat. Prod.
2009, 72, 1429−1436.
(23) Van Wagoner, R. M.; Drummond, A. K.; Wright, J. L. Biogenetic
diversity of cyanobacterial metabolites. Adv. Appl. Microbiol. 2007, 61,
89−217.
(34) Breneman, C. M.; Wiberg, K. B. Determining atom-centered
monopoles from molecular electrostatic potentials. The need for high
sampling density in formamide conformational analysis. J. Comput.
Chem. 1990, 11, 361−373.
(35) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hawegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Dapprish, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 03;
Gaussian, Inc.: Wallingford, CT, 2009.
(24) Walther, T.; Renner, S.; Waldmann, H.; Arndt, H. D. Synthesis
and structure-activity correlation of a brunsvicamide-inspired cyclo-
peptide collection. ChemBioChem 2009, 10, 1153−1162.
(25) Bunnage, M. E.; Blagg, J.; Steele, J.; Owen, D. R.; Allerton, C.;
McElroy, A. B.; Miller, D.; Ringer, T.; Butcher, K.; Beaumont, K.;
Evans, K.; Gray, A. J.; Holland, S. J.; Feeder, N.; Moore, R. S.; Brown,
D. G. Discovery of potent & selective inhibitors of activated thrombin-
activatable fibrinolysis inhibitor for the treatment of thrombosis. J.
Med. Chem. 2007, 50, 6095−6103.
(26) Rees, D. C.; Lipscomb, W. N. Binding of ligands to the active
site of carboxypeptidase A. Proc. Natl. Acad. Sci. U.S.A. 1981, 78,
5455−5459.
(27) Lauro, G.; Romano, A.; Riccio, R.; Bifulco, G. Inverse virtual
screening of antitumor targets: pilot study on a small database of
natural bioactive compounds. J. Nat. Prod. 2011, 74, 1401−1407.
(28) Trott, O.; Olson, A. J. AutoDock Vina: improving the speed and
accuracy of docking with a new scoring function, efficient
optimization, and multithreading. J. Comput. Chem. 2010, 31, 455−
461.
(29) The top two hits corresponded to galectin 7 and calmodulin
(CaM). Inspection of the docked models showed 1 to be located in
the galactose binding site of galectin 7, anchored through the exocyclic
Phe residue, and in an extensive hydrophobic channel in CaM, making
only partial contacts with this site. Because these interactions were not
optimal relative to the known ligands, we did not pursue them further.
(30) Morris, G. M.; Huey, R.; Lindstrom, W.; Sanner, M. F.; Belew,
R. K.; Goodsell, D. S.; Olson, A. J. AutoDock4 and AutoDockTools4:
Automated docking with selective receptor flexibility. J. Comput. Chem.
2009, 30, 2785−2791.
(31) Rees, D. C.; Lipscomb, W. N. Refined crystal structure of the
potato inhibitor complex of carboxypeptidase A at 2.5 A resolution. J.
Mol. Biol. 1982, 160, 475−498.
(32) Roos, E.; Bernabe, P.; Hiemstra, H.; Speckamp, N.; Kaptein, B.;
Boesten, W. H. J. Palladium-catalyzed transprotection of allylox-
ycarbonyl-protected amines: efficient one-pot formation of amides and
dipeptides. J. Org. Chem. 1995, 60, 1733−1741.
(33) Mock, W. L.; Liu, Y.; Stanford, D. J. Arazoformyl peptide
surrogates as spectrophotometric kinetic assay substrates for
carboxypeptidase A. Anal. Biochem. 1996, 239, 218−222.
742
dx.doi.org/10.1021/jm201238p | J. Med. Chem. 2012, 55, 735−742