Total synthesis and antibacterial testing of the A54556 cyclic acyldepsipeptides isolated from streptomyces hawaiiensis

Article abstract of DOI:10.1021/np500158q

The first total synthesis of all six known A54556 acyldepsipeptide (ADEP) antibiotics from Streptomyces hawaiiensis is reported. This family of compounds has a unique mechanism of antibacterial action, acting as activators of caseinolytic protease (ClpP). Assembly of the 16-membered depsipeptide core was accomplished via a pentafluorophenyl ester-based macrolactamization strategy. Late stage amine deprotection was carried out under neutral conditions by employing a mild hydrogenolysis strategy, which avoids the formation of undesired ring-opened depsipeptide side products encountered during deprotection of acid-labile protecting groups. The free amines were found to be significantly more reactive toward late stage amide bond formation as compared to the corresponding ammonium salts, giving final products in excellent yields. A thorough NMR spectroscopic analysis of these compounds was carried out to formally assign the structures and to aid with the spectroscopic assignment of ADEP analogues. The identity of two of the structures was confirmed by comparison with biologically produced samples from S. hawaiiensis. An X-ray crystallographic analysis of an ADEP analogue reveals a conformation similar to that found in cocrystal structures of ADEPs with ClpP protease. The degree of antibacterial activity of the different compounds was evaluated in vitro using MIC assays employing both Gram-positive and Gram-negative strains and a fluorescence-based biochemical assay.

Full text of DOI:10.1021/np500158q

Article
pubs.acs.org/jnp
Total Synthesis and Antibacterial Testing of the A54556 Cyclic
Acyldepsipeptides Isolated from Streptomyces hawaiiensis
Jordan D. Goodreid,† Keith Wong,§ Elisa Leung,§ Shannon E. McCaw,‡ Scott D. Gray-Owen,‡
Alan Lough,† Walid A. Houry,§ and Robert A. Batey*,†
†Davenport Research Laboratories, Department of Chemistry, University of Toronto, 80 St. George Street, Toronto, ON M5S 3H6,
Canada
‡Department of Molecular Genetics, Medical Sciences Building, University of Toronto, Toronto, ON M5S 1A8, Canada
§Department of Biochemistry, Medical Sciences Building, University of Toronto, Toronto, ON M5S 1A8, Canada
S
* Supporting Information
ABSTRACT: The first total synthesis of all six known
A54556 acyldepsipeptide (ADEP) antibiotics from Streptomy-
ces hawaiiensis is reported. This family of compounds has a
unique mechanism of antibacterial action, acting as activators
of caseinolytic protease (ClpP). Assembly of the 16-membered
depsipeptide core was accomplished via a pentafluorophenyl
ester-based macrolactamization strategy. Late stage amine
deprotection was carried out under neutral conditions by
employing a mild hydrogenolysis strategy, which avoids the
formation of undesired ring-opened depsipeptide side
products encountered during deprotection of acid-labile
protecting groups. The free amines were found to be significantly more reactive toward late stage amide bond formation as
compared to the corresponding ammonium salts, giving final products in excellent yields. A thorough NMR spectroscopic
analysis of these compounds was carried out to formally assign the structures and to aid with the spectroscopic assignment of
ADEP analogues. The identity of two of the structures was confirmed by comparison with biologically produced samples from S.
hawaiiensis. An X-ray crystallographic analysis of an ADEP analogue reveals a conformation similar to that found in cocrystal
structures of ADEPs with ClpP protease. The degree of antibacterial activity of the different compounds was evaluated in vitro
using MIC assays employing both Gram-positive and Gram-negative strains and a fluorescence-based biochemical assay.
espite the achievements of modern antibiotic therapies, a
disturbing trend of increased bacterial resistance has
configuration and position of the methyl group at C-4 in the
trans-4-methyl-L-proline in 7, which for some of the A54556
factors was incorrectly assigned to C-17 with ambiguous
stereochemistry. Schmidt and co-workers reported the first
total synthesis of enopeptin B (8),8 featuring a pentafluor-
ophenyl ester based macrolactamization strategy.9
More recently, ADEPs have gained considerable attention in
the medicinal chemistry and biological communities owing to
their biologically distinct mechanism of action. Using reversed
genomics techniques,10 scientists at Bayer were the first to
identify the biological target of ADEPs as caseinolytic protease
(ClpP).11 ClpP comprises the core proteolytic unit in a large
family of bacterial serine protease complexes.12 There is a
wealth of crystal structure data available for ClpP from a large
variety of bacterial species.13 Moreover, the structural changes
underlying the ADEP-mediated ClpP activation mechanism
have been elucidated using X-ray crystallography, with two
cocrystal structures reported for Bacillus subtilis ClpP−ADEP14
and Escherichia coli (E. coli) ClpP−ADEP complexes.15 In
D
prompted a need for the development of novel, effective, and
safe antibiotics with new mechanisms of action as a means to
combat cross-resistance.1 Natural products from actinomycetes
(soil bacteria), particularly those of the Streptomyces genus, have
served as a promising source of new structural leads in this area,
a trend that will likely continue well into the future.2 The cyclic
depsipeptide class of antibiotics has provided several examples
of such compounds, including the antibiotic daptomycin.3 In
addition to their antibiotic activity, naturally occurring cyclic
depsipeptides display a broad range of biological activities, and
their complex chemical structures serve as challenging total
synthesis targets and as medicinal chemistry leads.4,5 In the
early 1980s, scientists from Eli Lilly and Company described
the isolation of eight acyldepsipeptides (ADEPs, previously
described as A54556 factors A−H) from Streptomyces
hawaiiensis NRRL 15010, some of which showed notable
antibiotic activity against Gram-positive bacteria.6 Of these
eight compounds, only six were structurally assigned (Figure 1,
1−6). In 1991, the group of Isono isolated the structurally
related enopeptins A (7) and B (8) from Streptomyces sp. RK-
1051.7 His group was the first to assign the correct
Received: February 18, 2014
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 1. Structures of A54556 factors (1−6) and enopeptins A (7) and B (8).
Scheme 1. Late-Stage Amide Coupling in Schmidt’s Total Synthesis of Enopeptin B (8)
Scheme 2. Synthesis of Conjugated Enoic Acids 14, 17, and 22
addition to mechanistic studies of ADEPs, QSAR studies have
also been reported.16 More recently, it has been shown that
coadministration of ADEPs with rifampicin leads to complete
eradication of highly resistant Staphylococcus aureus (Staph.
aureus) biofilms in vitro and in mouse models.17 Furthermore,
we and other research groups have discovered small-molecule
scaffolds, which have also been explored with respect to their
ability to both activate18 and inhibit19 ClpP.
Despite all of the literature surrounding the ADEP natural
products as structurally important lead compounds, there are to
the best of our knowledge no known reports of the total
synthesis of the originally assigned A54556 factors (1−6, Figure
1). In addition, the spectroscopic data6 that is available for these
natural products (and also many of the synthetic intermediates
reported in the literature) are limited. A full spectroscopic
analysis for the natural products will also facilitate the
B
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Scheme 3. Initial Strategy to Cyclic Acyldepsipeptide Natural Products 1 and 2
spectroscopic assignment of ADEP analogues and provide
information on their solution-phase conformations. The
synthetic challenges posed by these compounds include the
development of an efficient macrocyclization protocol (a
general problem for cyclic peptide/depsipeptide synthesis),20
as well as identifying an effective strategy for the attachment of
the exocyclic acylated amino acid group. The latter problem is
exemplified by the serious difficulties encountered in the late-
stage amide coupling in Schmidt’s total synthesis of enopeptin
B (8), which proceeded through the coupling of hydrobromide
salt 9 with acid 10 to generate the corresponding amide 11 in
only 17% yield (Scheme 1).8 Herein, we now report the total
synthesis of all six previously reported A54556 factors (1−6),
including a detailed spectroscopic assignment of these
compounds and comparison of 1 and 2 with biologically
produced samples from S. hawaiiensis. Also included is
measurement of the ClpP activation ability of compounds 1−
6 and their antibacterial effect on Gram-positive and Gram-
negative bacteria.
RESULTS AND DISCUSSION
■
The requisite trans-enoic acids (14, 17, and 22) for the final
amide couplings of ADEPs 1, 2, 4, and 6 were furnished
through a two-step sequence using Horner−Wadsworth−
Emmons (HWE) olefination21 followed by hydrolysis (Scheme
2). Ester 13 was obtained in 65% yield as an inseparable
mixture of geometric isomers (E:Z = 9:1) from the HWE
olefination of sorbic aldehyde 12. Basic hydrolysis of ester 13
followed by acidic workup afforded acid 14 in 92% yield
(geometric purity unchanged). Geometrically pure 14 was
obtained in 66% yield by recrystallization from hot chloroform
and minimal ethyl acetate. Diene 17 was synthesized in an
analogous fashion. The synthetic route to acid 22 utilized cis-2-
butene-1,4-diol 18 as a precursor for the synthesis of aldehyde
C
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Scheme 4. Revised Strategy to Compounds 1−6
intermediate 20.22,23 Subsequent HWE olefination of aldehyde
20 gave ester 21 in 81% yield as a single geometric isomer;
however basic hydrolysis of ester 21 required heating to obtain
full conversion, and after mild acidic workup with citric acid
(with no loss of the THP protecting group), acid 22 was
obtained in quantitative yield as a mixture of geometric isomers
(E:Z = 5:1).
easily separable.26 Phenacyl (Pac) deprotection of pentapep-
tides 26 and 27 using Zn/AcOH yielded the cyclization
precursors 28 and 29. The use of a pentafluorophenyl ester
based cyclization procedure9 for 28 and 29 afforded the N-Cbz-
protected macrolactams 30 and 31 as single diastereomers in
good overall yields. Palladium on carbon catalyzed hydro-
genolysis of the Cbz group in 30 and 31 required the addition
of a protic additive (i.e., HCl) to give hydrochloride salts 32
and 33. Attempts to remove the Cbz group in the absence of a
protic additive were unsuccessful due to catalyst poisoning by
the free amine, resulting in minimal conversion as indicated by
TLC analysis after 24 h.
Peptide coupling of the hydrochloride salts 32 and 33 with
N-Boc-L-phenylalanine gave N-Boc-protected hexadepsipepti-
des 34 and 35 in good yields. In Schmidt’s total synthesis of
enopeptin B (8), unusually harsh acidic conditions (HBr in
AcOH) were required for Boc deprotection.8 Other conditions
that have been reported in the literature for this deprotection
validate these findings (TFA:H2O, 9:1).16c In our hands, the
latter conditions led to mixtures of the desired cyclic Boc-
deprotected TFA salt accompanied by minor amounts of the
corresponding depsipeptide ring-opened TFA salt (seco acid).
Similarly, attempts to obtain the analogous ammonium
hydrochloride salts from Boc-protected hexadepsipeptides
through the use of HCl in dioxane were not successful. It
was found however that the undesired ring-opened byproducts
from Boc-protected hexadepsipeptides 34 and 35 could be
The pentapeptide precursors required for the key macro-
lactamization were assembled from didepsipeptide and
tripeptide fragments using standard solution-phase peptide
chemistry (Scheme 3).24 Pentadepsipeptides 26 and 27 were
obtained in good yields by the amide coupling of didepsipep-
tide hydrochloride salts 24 and 25 with tripeptide 23 using the
uronium-based coupling reagent TPTU.25,26 Although this
particular reagent class, similar to aminium-based reagents (e.g.,
HATU27), has found widespread use in peptide synthesis26 and
has been employed for late-stage amide formations in previous
ADEP syntheses,8,16,24 we found that separation of the
N,N,N′,N′-tetramethyl urea byproduct from the pentadepsipep-
tide products is a challenge due to their comparable polarity on
silica gel (the same is true for subsequent hexadepsipeptide
intermediates). Similar purification issues occurred using polar
aprotic solvents for these reactions (e.g., DMF and NMP). To
circumvent these issues, further peptide coupling reactions were
carried out in halogenated solvents (e.g., CH2Cl2 or CHCl3)
using phosphonium-based coupling reagents (e.g., PyBOP and
PyAOP), as the corresponding phosphoramide byproduct is
D
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 2. X-ray crystal structure of N-Cbz-protected hexadepsipeptide 38 showing intramolecular H-bonding (methanol solvate was omitted for
clarity). Left: ORTEP diagram (showing 30% probability ellipsoids for the non-H atoms). Right: Structural representation.
avoided by using strictly anhydrous TFA in CH2Cl2, giving
TFA salts 36 and 37 in quantitative yields. It is plausible that
depsipeptide ring-opening may have occurred as a result of
adventitious water during this step in Schmidt’s synthesis
(resulting in a poor amide yield in the next step, Scheme 1),
since we have found that even small quantities of water present
in the TFA can still promote the formation of this side product.
Amide coupling of the trifluoroacetic acid salts 36 and 37 with
trienoic acid 14 using PyAOP furnished the natural products 1
and 2 in only modest yields despite prolonged reaction time
(48 h). Our initial attempts to optimize this coupling step
revealed that adding more equivalents of the acid component or
complexes.14,15 Each of the dihedral angles (ψ and ϕ) of the
amide backbone of 38 fall within the normal range associated
with the amino acid residues found in peptides.29
It is plausible to propose that, based on the crystal structure
data, the decreased reactivity of the TFA salts 36 and 37
(Scheme 3) toward amide coupling reactions may be attributed
to the strong conformational preference of the depsipeptide
ring. The resultant intramolecular H-bonding interaction
+
between the alanine CO with the phenylalanine-NH3
group in 36 and 37 could lead to the formation of a tight
ion pair, resistant to deprotonation by Hunig’s base.
̈
1
Spectroscopic Analysis. The assignment of the H NMR
spectrum of 1 was accomplished using a combination of 2D
NMR techniques including COSY, TOCSY, HSQC, and
ROESY30 (Table 1). The assignment of the 13C NMR
spectrum of 1 utilized a combination of HSQC (for carbons
with attached hydrogens) and HMBC30 spectra (for carbonyl
carbons) (Table 1 and Figure 3). Selected NOE correlations
observed in the ROESY of 1 (Figure 3) (i) provided
unambiguous assignment of relative stereochemistry for all
hydrogens in both proline residues and (ii) further support a
molecular conformation in solution that is consistent with the
conformation for the X-ray crystal structure of Cbz-protected
hexapeptide 38, based upon the transannular interactions
observed across the macrocyclic ring. Prominent interactions
include (i) NH(1)−H-19 (distance in X-ray crystal structure of
38 = 2.66 Å), (ii) H-2−H-7 (2.19 Å), (iii) H-8−H-12 (C-8−C-
12 = 4.25 Å), (iv) H-10−H-14 (1.95 Å), (v) H-11−H-15b (C-
11−H-15b = 3.41 Å), and (vi) H-14−H-19 (2.15 Å). Weak
transannular NOE correlations (not shown) were also observed
for NH(2)−H-5a (2.51 Å) and H-14−H-20a (2.74 Å).
Hunig’s base did not increase yields appreciably. Interestingly,
̈
the corresponding free amines of compounds 36 and 37,
initially obtained in low yields by extraction from NaHCO3
(sat. aq) and CH2Cl2 (likely due to appreciable water
solubility), underwent the same amide formations in excellent
yields under identical reaction conditions.
Problems with acidic deprotection of the Boc groups and the
increased amide coupling yields observed with free amines
prompted the use of an alternative amine protecting group
strategy that could allow for neutral cleavage under mild
reaction conditions, providing direct access to the free amines
from the cyclic hexadepsipeptides. Accordingly, it was found
that the analogous N-Cbz-protected hexadepsipeptides 38 and
39 could be deprotected in quantitative yields under neutral
conditions without the need for protic additives, yielding the
free amines 40 and 41 (Scheme 4). PyAOP-based amide
coupling of amines 40 and 41 with the acid 14 or 17 or with
sorbyl chloride resulted in excellent yields of the depsipeptide
natural products 1−5. Similarly, peptide coupling of 40 with
THP-protected acid 22 gave ADEP precursor 42 in 90% yield.
Subsequent deprotection using catalytic PPTS in EtOH at 60
°C gave natural product 6 in 87% yield, without any observed
ring-opening or O → N-acyl transfer side products. The triene-
containing ADEPs 1 and 2 were observed to be both thermally
and photochemically unstable compounds upon isolation, and
special precautions were necessary in their handling and storage
to avoid decomposition.28
An X-ray crystal structure of N-Cbz-protected hexadepsipep-
tide 38 was obtained confirming the stereochemical integrity of
the synthetic sequence (Figure 2). The conformation of 38
shows two stabilizing intramolecular hydrogen bonds: (i) the
phenylalanine CO with the alanine NH (1.98 Å) and (ii) the
alanine CO with the phenylalanine NH (2.00 Å), which is
Cbz-protected. These intramolecular interactions impose a
significant amount of conformational bias and give ADEP
molecules their unique conformation. The conformation of 38
is the same as that observed with the known ClpP−ADEP
Many of the same relative stereochemical relationships for
the proline residues in 2 are similar to 1; hence the majority of
assignments for 2 (Table 2) were made by comparison of the
1H, 13C, HSQC, and ROESY31 spectra with those of 1. The
only structural difference between 1 and 2 is at C-4, where the
methyl group (4-Me) present in 1 is absent in 2. The
1
methylene hydrogens in 2 were not resolved at all in the H
NMR spectrum and, hence, could not be individually assigned.
Similarly, the diastereotopic methylene H-23 hydrogens were
not sufficiently resolved in the 1H NMR spectra to be
individually assigned for any of the final compounds.
Spectroscopic assignment for the remainder of compounds
3−6 was made by analogy depending on whether the C-4
position was methylated (related to 1) or not (related to 2),
each compound having only minor structural differences in
their acyl side chains.
The concentration effect on spectral appearance can be quite
pronounced for ADEP compounds due to their high degree of
E
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
a
intramolecular H-bonding in solution. As the concentration is
increased, an increased propensity for aggregation32 in solution
occurs due to competing intermolecular H-bonds. This effect is
shown at various compound concentrations (c = 7−35 mg/
Table 1. NMR Data (600 MHz, CDCl3) of Synthetic 1
b
no.
δC, type
172.7, C
δH (J in Hz)
COSY
HMBC
1
2
1
1
60.2, CH
4.45, d (8.5)
H-3b
1, 3, 4, 5
1, 2, 4, 5
mL) in the H NMR spectra of 6 (Figure 4). The H NMR
signal of the solvent-exposed amide NH(3) shifts dramatically
downfield (Δδ +0.46 for c = 7 and 35 mg/mL) with increasing
concentration in comparison with NH(1) (not shown, Δδ
+0.01) and NH(2) (Δδ +0.02), which are involved in
intramolecular H-bonding. Similarly, the rate of deuterium
exchange for NH(3) is much faster than for NH(1) and
NH(2), which is apparent by the significant broadening of this
signal at low concentration. The change in spectroscopic
appearance with concentration is not limited to the amide NH
signals, but also is apparent for the α-hydrogen H-22 signal
(close proximity to intramolecular hydrogen bond), which
undergoes a noticeable downfield shift with increasing
concentration (Δδ +0.11), resulting in eventual overlap with
the H-10 signal at c = 29 mg/mL. More subtle upfield shifts at
higher concentrations were also observed for H-8 (not shown,
Δδ −0.04), H-12 (not shown, Δδ −0.03), H-20a (Δδ −0.02),
and the aromatic phenylalanine H-25, H-26, and H-27 signals
(Δδ −0.02, −0.03, and −0.04, respectively), whereas H-17b
was slightly downfield shifted (not shown, Δδ +0.05). All other
hydrogen signals from the molecule did not change appreciably
in terms of chemical shift with changing concentration. These
effects made initial spectroscopic comparison of the synthetic
natural products with biologically produced samples33 from S.
hawaiiensis slightly more challenging. However, comparable
NMR data sets for 1 and 2 with the isolated natural products
were collected at similar concentrations and using the same
frequency spectrometer.34
Biological Data. To assess and compare the antibacterial
activity of 1−6, two types of biological assays were conducted.
A fluorometric assay revealed that E. coli ClpP in the presence
of compounds 1 and 4, and to a lesser extent compound 2 (at 5
μM concentration), was able to effectively degrade a fluorescein
isothiocyanate-labeled casein substrate (casein-FITC), demon-
strating the ability of the compounds to act as ClpP
activators.35 In contrast, compounds 3, 5, and 6, incorporating
shorter side chains, were not effective. Thus, the length and
degree of hydrophobicity of the acyl side chains appear to play
a pivotal role in E. coli ClpP activation, with longer hydrophobic
side chains tending to be more far more active than shorter
chains. This is perhaps not surprising given that this moiety is
surrounded largely by hydrophobic residues in the reported E.
coli ClpP−1 cocrystal structure.15 Interestingly, compounds 1
and 4 incorporating the trans-4-methyl-proline residue are more
effective E. coli ClpP activators than the corresponding proline-
based compound 2. This could be rationalized based on subtle
differences in their binding affinities for ClpP. For compounds
1 and 2, we have previously measured their dissociation
constants (Kd) by isothermal titration calorimetry, which were
found to be 0.3 0.1 and 0.7 0.2 μM, respectively.18
The antibacterial activities of ADEPs 1−6 are summarized in
Table 5. The activity of the compounds was initially confirmed
by testing them against previously susceptible Gram-positive
organisms, namely, Staphylococcus aureus and Streptococcus
pneumoniae.6,11a For these bacterial species, the glycopeptide
antibiotic vancomycin was used as a positive control.
Compounds 1−5 all showed good to moderate activity against
Staph. aureus (MICs in the range 0.5−16 μg/mL), while
compound 6, which incorporates a more hydrophilic acyl side
3a
38.90, CH2
2.06, dd (13.0, H-3b
7.0)
c
3b
1.77, m
H-2, H-3a
1, 2, 4, 4-
Me
c
4
29.6, CH
18.4, CH3
54.5, CH2
2.36, m
4-Me
H-4
4-Me
5a
0.98, d (6.5)
3, 4, 5
3, 4
3.49, dd (12.0, H-5b
9.0)
5b
3.08, dd (12.0, H-5a
8.5)
3, 4, 4-Me
6
7
170.00, C
47.9, CH
4.89, dq (9.5,
6.5)
NH(1), H-8
6, 8
6, 7
8
17.7, CH3
170.07, C
56.4, CH
1.37, d (6.5)
H-7
9
10
4.76, q (6.5)
1.51, d (6.5)
H-11
H-10
9, 11, 12,
13
11
12
13
14
15.8, CH3
9, 10
31.05, N-CH3 2.83, s
172.3, C
10, 13
56.5, CH
5.13, dd (8.0,
H-15a
15, 16, 17
13, 16
3.0)
c
15a
15b
16a
16b
17a
31.06, CH2
2.34, m
H-14, H-15b
H-15a, H-16a
H-16b
c
1.96, m
23.3, CH2
46.7, CH2
2.16, m
15, 17
15
c
1.95, m
H-16a
3.74, ddd (12.0, H-17b
15, 16
8.0, 5.0)
c
17b
18
3.54, m
H-17a
14, 15, 16
164.7, C
51.3, CH
19
4.51, ddd (9.0, NH(3), H-
9.0, 1.5) 20a, H-20b
18, 20, 21
1, 19
20a
65.0, CH2
4.84, dd (12.0, H-19, H-20b
1.5)
c
20b
21
3.52, m
H-19, H-20a
1, 19
171.6, C
55.3, CH
22
4.69, m
NH(2), H-23 21, 23, 24,
28
d
23
38.93, CH2
2.97,
m
H-22
21, 22, 24,
25
24
25
26
27
28
29
30
31
136.3, C
129.5, CH
128.7, CH
126.9, CH
166.3, C
7.15, d (7.0)
H-26, H-27
H-25, H-27
H-25, H-26
23, 24, 27
24, 25
c
7.27, m
7.19, t (7.0)
25, 26
123.1, CH
141.5, CH
128.1, CH
6.27, d (15.0)
H-30
28, 31
c
7.26, m
H-29, H-31
28, 31, 32
29, 30, 33
6.25, dd (14.5, H-30, H-32
11.5)
32
139.9, CH
6.50, dd (14.5, H-31, H-33
11.0)
30, 33, 34
33
34
131.5, CH
134.0, CH
6.13, m
H-32, H-34
31, 35
32, 35
5.89, dq (14.0, H-33, H-35
7.0)
35
18.6, CH3
1.81, m
H-34, H-33
H-7
32, 33, 34
NH(1)
NH(2)
NH(3)
8.54, d (9.5)
6.96, d (8.0)
9
H-22
21, 22, 28
21
6.83, br d (9.0) H-19
a
b
c = 16 mg/mL. HMBC correlations, optimized for 8.0 Hz, are from
c
proton(s) stated to the indicated carbon. Denotes values were taken
from correlations observed in a H−13C HSQC NMR experiment.
1
d
Protons are not resolved.
F
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 3. Left: Selected NOE correlations observed in the ROESY. Right: Selected HMBCs for 1.
chain, was found to be largely inactive (MIC 128 μg/mL).
Interestingly, compound 4 (MIC 0.5 μg/mL) was found to be
8 times more potent against Staph. aureus than the next most
active ADEPs 1 and 3 (MIC 4 μg/mL). This increased activity
is somewhat suprising given that 1 and 4 have identical chain
lengths and differ only in the degree of unsaturation of the acyl
group. Unlike the case with Staph. aureus, all of the ADEPs 1−6
were found to be effective against S. pneumoniae (MICs in the
range <0.0625 to 0.5 μg/mL). In particular, both compounds 1
and 4 showed the highest levels of activity (MICs <0.0625 μg/
mL) and were found to be at least 32 times more potent than
vancomycin. Similar to the trends observed for Staph. aureus,
compounds 1−5 showed good activity against the Gram-
negative species Neisseria meningitidis (MICs in the range
<0.0625 to 4 μg/mL), with compounds 4 and 1 having the
highest activities, while compound 6 was found to have more
modest activity (MIC 64 μg/mL). For this species, the
aminoglycoside kanamycin A was used as a positive control.
In contrast to related studies,6,36 compounds 1−6 were found
to be sufficiently active against this Gram-negative strain and
did not necessitate the use of an external membrane
permeabilizing agent in the culture broth.37 In general, the
observed antibacterial trends suggest that compounds lacking a
C-4 methyl group (i.e., 2 and 5) show lower activity than their
methylated counterparts 1 and 3, while compounds with longer
acyl side chains (i.e., 1, 2, and 4) show higher activity than
comparable compounds with shorter side chains (i.e., 3 and 5).
Conclusions. In conclusion, we have completed the first
total syntheses of all of the originally isolated ADEPs 1−6 from
S. hawaiiensis, compounds that are increasingly attracting
significant attention due to their unique mode of antibacterial
activity. Several practical challenges were encountered during
the course of our synthetic studies. These problems were
addressed, aiding future studies toward the synthesis of related
natural products and ADEP analogues. Notably, the use of a
late stage Cbz-deprotection strategy was found to offer
significant advantages in terms of the general synthesis of
these compounds. Thus, quantitative yields of free amine
species can be obtained under mild conditions using hydro-
genolysis, avoiding formation of depsipeptide ring-opened side
products (sometimes encountered under strongly acidic
deprotection conditions). The free amines, in contrast with
the corresponding ammonium salts, reliably undergo the final
amide formations in excellent yields.
observed in the two known ClpP−ADEP cocrystal structures.
This conformation is locked by the presence of hydrogen bonds
that occur between two of the cyclic depsipeptide ring amide
residues and the two amide linkages of the phenylalanine side
1
chain. The first formal assignment of the H and 13C NMR
spectra for the ADEPs using a combination of 2D NMR
techniques (COSY, TOCSY, HSQC, HMBC, and ROESY) and
a direct comparison of biologically produced and synthetic 1
and 2 from S. hawaiiensis provided sufficient spectroscopic
evidence for structural confirmation. A pronounced concen-
1
tration effect for the H NMR spectra was also observed,
presumably reflecting the breakdown of the intramolecular H-
bond network and/or increased aggregation at higher
concentrations. Similar effects and conformational preferences
are likely for related ADEP analogues currently being
investigated as part of medicinal chemistry studies. Finally,
the antibacterial activities obtained in vitro confirm activity
trends of the compounds from two well-studied Gram-positive
bacterial strains (Staph. aureus and S. pneumoniae), as well as
demonstrating good levels of activity against a previously
untested and clinically relevant Gram-negative bacterial species,
N. meningitidis.
EXPERIMENTAL SECTION
■
General Experimental Procedures. All reactions were per-
formed under nitrogen in flame-dried glassware. Tetrahydrofuran was
freshly distilled from sodium/benzophenone ketyl under nitrogen.
Dichloromethane was freshly distilled from calcium hydride under
nitrogen. Anhydrous dimethylformamide and methanol were obtained
as ≥99.9% pure and stored under argon. All other solvents were ACS
grade or better from commercial suppliers and used as received. Flash
chromatography on silica gel (60 Å, 230−400 mesh) was performed
with reagent grade solvents. Analytical thin-layer chromatography
(TLC) was performed on precoated silica gel plates and visualized
with a UV254 lamp. Solvent ratios for chromatography and Rf values are
reported as v/v ratios. Melting points are uncorrected and obtained on
>95% pure compounds without any further recrystallization. All 1D
(1H, 13C) and 2D (COSY, TOCSY, HSQC, HMBC, and ROESY)
NMR spectra were obtained on 300, 400, 400, 500, and 600 MHz
spectrometers as solutions in deuterated solvents. Chemical shifts are
reported in δ ppm values. 1H chemical shifts were internally referenced
to tetramethylsilane (δ 0.00) for CDCl3 or to the residual proton
resonance in CD3OD (δ 3.31) and DMSO-d6 (δ 2.49). Carbon
chemical shifts were internally referenced to the solvent resonances in
CDCl3 (δ 77.16 ppm), CD3OD (δ 49.15 ppm), or DMSO-d6 (δ 39.51
ppm). Peak multiplicities are designated by the following abbrevia-
tions: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br,
broad; J, coupling constant in Hz and rounded to the nearest 0.5 Hz.
Exact mass measurements were performed on a time-of-flight mass
An X-ray crystal structure analysis of one of the synthetic
intermediates revealed that the conformation of both the cyclic
depsipeptide ring and the side chain is the same as that
G
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
a
spectrometer utilizing electrospray ionization (ESI), direct analysis in
real time ionization (DART), and electron impact ionization (EI).
General Procedure A: Synthesis of Amides Using Coupling
Reagents. To an ice bath-cooled solution of amine, acid, and
Table 2. NMR Data (500 MHz, CDCl3) of Synthetic 2
b
no.
δC, type
172.8, C
δH (J in Hz)
COSY
HMBC
1
2
coupling reagent in CH Cl was added Hunig’s base dropwise. The
̈
2
2
59.3, CH
30.8, CH2
4.44, d (8.0)
H-3b
1, 3, 4, 5
1, 4, 5
1, 2, 4
3
reaction was stirred overnight with slow warming to room temperature
under nitrogen. The reaction mixture was diluted with water and
extracted with three portions of CH2Cl2. The combined organic layers
were then dried over MgSO4 and concentrated in vacuo, and the crude
product was then purified using flash chromatography to give the
corresponding amide.
c
3a
3b
4
2.00, m
H-3b
c
2.16, m
H-2, H-3a
cd
,
21.3, CH2
47.1, CH2
1.88,
m
c
5a
5b
6
3.60, m
3.30, m
H-5b
H-5a
4
3, 4
170.06, C
48.0, CH
General Procedure B: Synthesis of Amides Using Acid
Chlorides. To an ice bath-cooled solution of amine in CH2Cl2 was
added via cannula transfer the acid chloride in CH2Cl2, followed by the
7
4.89, dq (9.5,
6.5)
NH(1), H-8
H-7
6, 8
6, 7
dropwise addition of Hunig’s base. The reaction was stirred overnight
̈
8
17.6, CH3
170.08, C
56.4, CH
1.36, d (6.5)
with slow warming to room temperature under nitrogen. The reaction
mixture was then diluted with water and extracted with three portions
of CH2Cl2. The combined organic layers were then dried over MgSO4
and concentrated in vacuo, and the crude product was then purified
using flash chromatography to give the corresponding amide.
For ADEP compounds 1−6, analytically pure material (used for
characterization and biological testing) was obtained by semi-
preparative reversed-phase HPLC of the product previously isolated
by flash chromatography. HPLC fractions that contained the product
by LRMS (ESI+) were combined into a lyophilization flask, frozen
solid in a dry ice−acetone bath, and then lyophilized overnight,
yielding solid compounds.
9
10
11
12
13
14
4.78, q (7.0)
1.52, d (7.0)
H-11
H-10
9, 11, 12
9, 10
15.8, CH3
e
31.0, N-CH3 2.82, s
172.2, C
10, 13
56.6, CH
5.15, dd (8.0,
3.0)
H-15a, H-15b 15, 16, 17
e
15a
15b
16a
16b
17a
31.0, CH2
2.35, m
H-14, H-15b
H-15a
13, 14, 16
c
1.97, m
c
23.3, CH2
46.7, CH2
2.16, m
H-16b
15, 17
15
c
1.97, m
H-16a
A55456 Factor A (1).6 This was synthesized according to general
procedure A using amine 40 (99.3 mg, 0.166 mmol, 1.0 equiv), acid 14
(30.4 mg, 0.220 mmol, 1.3 equiv), PyAOP (122.0 mg, 0.234 mmol, 1.3
3.74, ddd (12.0, H-17b
15, 16
8.0, 5.0)
c
17b
18
3.58, m
H-17a
14, 15, 16
164.8, C
51.2, CH
equiv), CH Cl (5 mL), and Hunig’s base (90.0 μL, 0.517 mmol, 3.1
̈
2
2
equiv), 19 h. Note: The reaction flask was covered in tin foil to protect
the coupling reagent and product from light. Flash chromatography
(19:1 EtOAc−MeOH) of the crude organic extracts gave 110.3 mg of
1 in 92% yield. Semipreparative reversed-phase HPLC of the
aforementioned silica gel purified product gave 1 as a white solid:
19
4.51, ddd (9.5,
9.5, 1.5)
NH(3), H-
20a, H-20b
18, 20, 21
1, 19
20a
20b
65.1, CH2
4.82, dd (11.0,
1.5)
H-19, H-20b
3.53, dd (11.0,
9.5)
H-19, H-20a
1, 19
mp 175−176 °C (CH2Cl2); Rf 0.55 (19:1 EtOAc−MeOH); [α]24
D
21
22
171.6, C
55.3, CH
−48.0 (c 0.24, CHCl3); IR (thin film in CH2Cl2) νmax 3500 (br), 3293
1
(br), 2960, 2933, 2876, 1650, 1437, 1271, 1111, 1011 cm−1; H and
4.72, m
NH(2), H-23 21, 23, 24,
28
13C NMR see Table 1; LRMS (ESI+) m/z (rel intensity) 741.4 (97)
[M + Na]+, 719.4 (100) [M + H]+, 452.3 (18); HRMS (ESI+) m/z
calcd for C38H51N6O8 [M + H]+ 719.3762, found 719.3774; analytical
HPLC (C18 column, 5 μm, 4.6 × 150 mm, gradient elution with
water−acetonitrile, flow rate 2.5 mL/min, column temp 23 °C,
detector λ 280 nm, tR 6.48 min).
d
23
39.0, CH2
2.97,
m
H-22
21, 22, 24,
25
24
25
26
27
28
29
30
136.5, C
129.6, CH
128.7, CH
126.8, CH
166.3, C
7.15, m
H-26, H-27
H-25, H-27
H-25, H-26
23, 24, 27
24, 25
c
7.27, m
A55456 Factor B (2).6 This was synthesized according to general
procedure A using amine 41 (96.9 mg, 0.166 mmol, 1.0 equiv), acid 14
(31.1 mg, 0.225 mmol, 1.3 equiv), PyAOP (127.0 mg, 0.243 mmol, 1.5
7.18, m
25, 26
123.1, CH
141.5, CH
6.29, d (14.5)
H-30
28, 31, 32
28, 29, 32
equiv), CH Cl (4 mL), and Hunig’s base (80.0 μL, 0.459 mmol, 2.8
̈
2
2
7.30, dd (14.5,
11.0)
H-29, H-31
equiv), 24 h. Note: The reaction flask was covered in tin foil to protect
the coupling reagent and product from light. Flash chromatography
(19:1 EtOAc−MeOH) of the crude organic extracts gave 109.8 mg of
2 in 94% yield. Semipreparative reversed-phase HPLC of the
aforementioned silica gel purified product gave 2 as a white solid:
31
32
33
34
35
128.3, CH
139.8, CH
131.5, CH
134.0, CH
18.6, CH3
6.25, dd (14.5,
11.0)
H-30, H-32
H-31, H-33
29, 30, 33
6.50, dd (14.5,
11.0)
29, 30, 33,
34
mp 187−188 °C (CH2Cl2); Rf 0.25 (19:1 EtOAc−MeOH); [α]24
6.12, m
H-32, H-34,
H-35
31, 32, 35
D
−46.2 (c 0.37, CHCl3); IR (thin film in CH2Cl2) νmax 3440 (br), 3304
(br), 3061, 3026, 2982, 2936, 2887, 1650 (br), 1437, 1267, 1007 cm−1;
1H and 13C NMR see Table 2; LRMS (ESI+) m/z (rel intensity) 727.4
(8) [M + Na]+, 705.4 (100) [M + H]+; HRMS (ESI+) m/z calcd for
C37H49N6O8 [M + H]+ 705.3603, found 705.3606; analytical HPLC
(C18 column, 5 μm, 4.6 × 150 mm, gradient elution with water−
acetonitrile, flow rate 2.5 mL/min, column temp 23 °C, detector λ 280
nm, tR 6.04 min).
5.89, dq (14.0,
7.0)
H-33, H-35
32, 35
1.81, m
H-34, H-33
31, 32, 33,
34
NH(1)
NH(2)
NH(3)
8.56, d (9.5)
H-7
9
f
7.01, m
H-22
H-19
22, 28
19, 21
f
7.00, m
A55456 Factor C (3).6 This was synthesized according to general
procedure B using amine 40 (77.4 mg, 0.129 mmol, 1.0 equiv) in
CH2Cl2 (2 mL), sorbyl chloride (36.0 mg, 0.276 mmol, 2.1 equiv) in
a
b
c = 26 mg/mL. HMBC correlations, optimized for 8.0 Hz, are from
c
proton(s) stated to the indicated carbon. Denotes values were taken
from correlations observed in a H−13C HSQC NMR experiment.
CH Cl (2 mL), and Hunig’s base (50.0 μL, 0.287 mmol, 2.2 equiv),
1
̈
2
2
d
e
f
Protons are not resolved. Carbon signals are overlapping. Denotes
28 h. Flash chromatography (19:1 EtOAc−MeOH) of the crude
organic extracts gave 65.1 mg of 3 in 73% yield. Semipreparative
reversed-phase HPLC of the aforementioned silica gel purified product
values were taken from correlations observed in a H−13C HMBC
1
NMR experiment.
H
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 4. Effect of concentration on spectroscopic appearance for compound 6.
a
Table 5. In Vitro Antibacterial Activity of A54556 Factors (1−6) against Gram-Positive and Gram-Negative Pathogens
MIC (μg/mL)
Compound
Staph. aureus (ATCC 29213)
S. pneumoniae (ATCC 46919)
N. meningitidis (H44/76)
1
4
16
4
<0.0625
0.5
0.125
2
0.5
3
0.125
<0.0625
0.25
2
4
0.5
16
128
2
<0.0625
5
4
6
0.25
64
b
vancomycin
2
−
kanamycin A
−
−
64
MIC values shown have been determined by standard broth-microdilution techniques.38 Symbol “−” indicates not determined.
a
b
gave 3 as a white solid: decomposition temp 113 °C (CH2Cl2); Rf 0.32
(19:1 EtOAc−MeOH); [α]24D −39.1 (c 0.18, CHCl3); IR (thin film in
CH2Cl2) νmax 3500 (br), 3284 (br), 2961, 2934, 2876, 1736, 1650
C-23), 38.84 (CH2, C-3), 31.06 (CH2, C-15), 31.03 (N-CH3, C-12),
29.6 (CH, C-4), 23.3 (CH2, C-16), 18.7 (CH3, C-33), 18.4 (CH3, 4-
Me), 17.6 (CH3, C-8), 15.8 (CH3, C-11); LRMS (ESI+) m/z (rel
intensity) 715.4 (72) [M + Na]+, 693.4 (100) [M + H]+, 452.3 (13);
HRMS (ESI+) m/z calcd for C36H49N6O8 [M + H]+ 693.3606, found
693.3631; analytical HPLC (C18 column, 5 μm, 4.6 × 150 mm,
gradient elution with water−acetonitrile, flow rate 2.5 mL/min,
column temp 23 °C, detector λ 280 nm, tR 5.62 min).
1
(br), 1441, 1111, 1001 cm−1; H NMR (CDCl3, 400 MHz, c 20 mg/
mL) δ 8.55 (1H, d, J = 9.5 Hz, NH(1)), 7.32−7.21 (3H, m, H-26, H-
30), 7.20−7.12 (3H, H-25, H-27), 6.99 (1H, d, J = 9.5 Hz, NH(3)),
6.95 (1H, d, J = 8.5 Hz, NH(2)), 6.28−6.15 (2H, m, H-29, H-31),
6.14−6.04 (1H, m, H-32), 5.13 (1H, dd, J = 9.0, 3.0 Hz, H-14), 4.88
(1H, dq, J = 9.5, 6.5 Hz, H-7), 4.83 (1H, dd, J = 12.0, 1.5 Hz, H-20a),
4.80−4.66 (2H, m, H-10, H-22), 4.51 (1H, ddd, J = 9.5, 9.5, 1.5 Hz,
H-19), 4.44 (1H, d, J = 8.0 Hz, H-2), 3.73 (1H, ddd, J = 11.5, 8.0, 5.0
Hz, H-17a), 3.62−3.45 (3H, m, H-5a, H-17b, H-20b), 3.08 (1H, dd, J
= 12.0, 8.5 Hz, H-5b), 3.03−2.92 (2H, m, H-23), 2.82 (3H, s, N−CH3
(H-12)), 2.44−2.28 (2H, m, H-4, H-15a), 2.22−2.10 (1H, m, H-16a),
2.06 (1H, dd, J = 13.0, 7.0 Hz, H-3a), 2.02−1.90 (2H, m, H-15b, H-
16b), 1.88−1.74 (4H, m, H-33, H-3b), 1.51 (3H, d, J = 7.0 Hz, H-11),
1.35 (3H, d, J = 6.5 Hz, H-8), 0.99 (3H, d, J = 7.0 Hz, 4-Me); 13C
NMR (CDCl3, 100 MHz, c 20 mg/mL) δ 172.7 (C, C-1), 172.2 (C, C-
13), 171.6 (C, C-21), 170.07 (C, C-9), 169.96 (C, C-6), 166.5 (C, C-
28), 164.8 (C, C-18), 141.6 (CH, C-30), 137.7 (CH, C-32), 136.4 (C,
C-24), 130.1 (CH, C-31), 129.6 (CH, C-25), 128.7 (CH, C-26), 126.9
(CH, C-27), 122.0 (CH, C-29), 65.1 (CH2, C-20), 60.2 (CH, C-2),
56.5 (CH, C-14), 56.4 (CH, C-10), 55.2 (CH, C-22), 54.5 (CH2, C-
5), 51.3 (CH, C-19), 47.9 (CH, C-7), 46.7 (CH2, C-17), 38.89 (CH2,
A55456 Factor D (4).6 This was synthesized according to general
procedure A using amine 40 (85.9 mg, 0.143 mmol, 1.0 equiv), acid 17
(26.2 mg, 0.187 mmol, 1.3 equiv), PyAOP (102.7 mg, 0.197 mmol, 1.4
equiv), CH Cl (4 mL), and Hunig’s base (80.0 μL, 0.459 mmol, 3.2
̈
2
2
equiv), 32 h. Note: The reaction flask was covered in tin foil to protect
the coupling reagent from light. Flash chromatography (19:1 EtOAc−
MeOH) of the crude organic extracts gave 91.5 mg of 4 in 89% yield.
Semipreparative reversed-phase HPLC of the aforementioned silica gel
purified product gave 4 as a white solid: mp 140−142 °C (CH2Cl2); Rf
0.51 (19:1 EtOAc−MeOH); [α]24 −51.9 (c 0.18, CHCl3); IR (thin
D
film in CH2Cl2) νmax 3500 (br), 3287 (br), 2959, 2932, 2874, 1732,
1650 (br), 1439, 1271, 1111, 1001 cm−1; 1H NMR (CDCl3, 400 MHz,
c 17 mg/mL) δ 8.54 (1H, d, J = 9.5 Hz, NH(1)), 7.32−7.22 (3H, m,
H-26, H-30), 7.21−7.10 (3H, m, H-25, H-27), 6.93 (1H, d, J = 8.0 Hz,
NH(2)), 6.87 (1H, d, J = 9.5 Hz, NH(3)), 6.26−6.16 (2H, m, H-29,
H-31), 6.12−6.03 (1H, m, H-32), 5.13 (1H, dd, J = 9.0, 3.0 Hz, H-14),
I
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
4.89 (1H, dq, J = 9.5, 6.5 Hz, H-7), 4.84 (1H, dd, J = 12.0, 1.5 Hz, H-
20a), 4.76 (1H, q, J = 7.0 Hz, H-10), 4.73−4.66 (1H, m, H-22), 4.51
(1H, ddd, J = 9.5, 9.5, 1.5 Hz, H-19), 4.44 (1H, d, J = 8.0 Hz, H-2),
3.78−3.69 (1H, m, H-17a), 3.62−3.46 (3H, m, H-5a, H-17b, H-20b),
3.09 (1H, dd, J = 12.0, 8.5 Hz, H-5b), 3.03−2.92 (2H, m, H-23), 2.82
(3H, s, N-CH3 (H-12)), 2.44−2.29 (2H, m, H-4, H-15a), 2.22−2.03
(4H, m, H-3a, H-16a, H-33), 2.02−1.91 (2H, m, H-15b, H-16b),
1.86−1.74 (1H, m, H-3b), 1.51 (3H, d, J = 7.0 Hz, H-11), 1.44 (2H,
sextet, J = 7.5 Hz, H-34), 1.36 (3H, d, J = 6.5 Hz, H-8), 0.98 (3H, d, J
= 6.5 Hz, 4-Me), 0.90 (3H, t, J = 7.5 Hz, H-35); 13C NMR (CDCl3,
100 MHz, c 17 mg/mL) δ 172.8 (C, C-1), 172.3 (C, C-13), 171.6 (C,
C-21), 170.1 (C, C-9), 170.0 (C, C-6), 166.5 (C, C-28), 164.8 (C, C-
18), 143.0 (CH, C-30), 141.8 (CH, C-32), 136.4 (C, C-24), 129.6
(CH, C-25), 128.8 (CH, C-31), 128.7 (CH, C-26), 126.9 (CH, C-27),
122.1 (CH, C-29), 65.0 (CH2, C-20), 60.2 (CH, C-2), 56.5 (CH, C-
14), 56.4 (CH, C-10), 55.2 (CH, C-22), 54.5 (CH2, C-5), 51.3 (CH,
C-19), 47.9 (CH, C-7), 46.7 (CH2, C-17), 38.89 (CH2, C-23), 38.83
(CH2, C-3), 35.1 (CH2, C-33), 31.06 (CH2, C-15), 31.03 (N-CH3, C-
12), 29.6 (CH, C-4), 23.3 (CH2, C-16), 22.1 (CH2, C-34), 18.4 (CH3,
4-Me), 17.6 (CH3, C-8), 16.0 (CH3, C-11), 13.8 (CH3, C-35); LRMS
(ESI+) m/z (rel intensity) 743.4 (100) [M + Na]+, 721.4 (81) [M +
H]+, 452.3 (14); HRMS (ESI+) m/z calcd for C38H53N6O8 [M + H]+
721.3919, found 721.3888; analytical HPLC (C18 column, 5 μm, 4.6 ×
150 mm, gradient elution with water−acetonitrile, flow rate 2.5 mL/
min, column temp 23 °C, detector λ 280 nm, tR 6.94 min).
and concentrated in vacuo. Flash chromatography (19:1 EtOAc−
MeOH) of the crude organic extracts gave 6 in 87% yield.
Semipreparative reversed-phase HPLC of the aforementioned silica
gel purified product gave 6 as a white solid: decomposition temp 168
°C (CH2Cl2); Rf 0.35 (9:1 EtOAc−MeOH); [α]24 −47.3 (c 0.32,
D
CHCl3); IR (thin film in CH2Cl2) νmax 3439 (br), 3289 (br), 2961,
2934, 2876, 1732, 1650 (br), 1441, 1271, 1111, 1007 cm−1; 1H NMR
(CDCl3, 700 MHz, c 20 mg/mL) δ 8.53 (1H, d, J = 9.5 Hz, NH(1)),
7.31 (1H, dd, J = 15.5, 11.0 Hz, H-30), 7.29−7.23 (2H, m, H-26), 7.18
(1H, t, J = 7.5 Hz, H-27), 7.14 (2H, d, J = 7.5 Hz, H-25), 7.06 (1H, d,
J = 8.0 Hz, NH(2)), 6.98 (1H, d, J = 9.5 Hz, NH(3)), 6.48−6.43 (1H,
m, H-31), 6.34 (1H, d, J = 15.5 Hz, H-29), 6.20 (1H, dt, J = 15.5, 5.0
Hz, H-32), 5.13 (1H, dd, J = 8.5, 3.0 Hz, H-14), 4.89 (1H, dq, J = 9.5,
6.5 Hz, H-7), 4.82 (1H, dd, J = 12.0, 1.5 Hz, H-20a), 4.76 (1H, q, J =
7.0 Hz, H-10), 4.74−4.69 (1H, m, H-22), 4.51 (1H, ddd, J = 9.5, 9.5,
1.5 Hz, H-19), 4.44 (1H, d, J = 8.5 Hz, H-2), 4.29−4.25 (2H, br m, H-
33), 3.76−3.71 (1H, m, H-17a), 3.57 (1H, ddd, J = 11.5, 7.0, 7.0 Hz,
H-17b), 3.53 (1H, dd, J = 12.0, 9.5 Hz, H-20b), 3.48 (1H, dd, J = 12.0,
9.0 Hz, H-5a), 3.09 (1H, dd, J = 12.0, 9.0 Hz, H-5b), 3.01−2.94 (2H,
m, H-23), 2.82 (3H, s, N-CH3 (H-12)), 2.44−2.32 (2H, m, H-4, H-
15a), 2.20−2.13 (1H, m, H-16a), 2.07 (1H, dd, J = 13.0, 7.0 Hz, H-
3a), 2.00−1.93 (2H, m, H-15b, H-16b), 1.88 (1H, br s, OH(1)),
1.83−1.76 (1H, m, H-3b), 1.41 (3H, d, J = 7.0 Hz, H-11), 1.34 (3H, d,
J = 6.5 Hz, H-8), 0.99 (3H, d, J = 6.5 Hz, 4-Me); 13C NMR (CDCl3,
175 MHz, c 20 mg/mL) δ 172.7 (C, C-1), 172.2 (C, C-13), 171.5 (C,
C-21), 170.0 (2C; C, C-6 and C, C-9), 166.1 (C, C-28), 164.7 (C, C-
18), 140.4 (CH, C-30), 139.7 (CH, C-32), 136.2 (C, C-24), 129.6
(CH, C-25), 128.7 (CH, C-26), 128.5 (CH, C-31), 126.9 (CH, C-27),
124.4 (CH, C-29), 65.1 (CH2, C-20), 63.0 (CH2, C-33), 60.2 (CH, C-
2), 56.5 (CH, C-14), 56.4 (CH, C-10), 55.4 (CH, C-22), 54.4 (CH2,
C-5), 51.3 (CH, C-19), 47.9 (CH, C-7), 46.7 (CH2, C-17), 38.92
(CH2, C-23), 38.87 (CH2, C-3), 31.0 (2C; N-CH3, C-12 and CH2, C-
15), 29.6 (CH, C-4), 23.3 (CH2, C-16), 18.5 (CH3, 4-Me), 17.6 (CH3,
C-8), 15.8 (CH3, C-11); LRMS (ESI+) m/z (rel intensity) 731.4 (99)
[M + Na]+, 709.4 (100) [M + H]+; HRMS (ESI+) m/z calcd for
C36H49N6O9 [M + H]+ 709.3555, found 709.3553; analytical HPLC
(C18 column, 5 μm, 4.6 × 150 mm, gradient elution with water−
acetonitrile, flow rate 2.5 mL/min, column temp 23 °C, detector λ 280
nm, tR 3.71 min).
Antibacterial Activity Evaluation. MIC values for compounds
1−6 were determined by standard protocols for broth microdilution.38
The Gram-positive strains Staph. aureus (ATCC 29213) and S.
pneumoniae (ATCC 49619) were provided by Dr. Don Low, Mount
Sinai Hospital, Toronto. The Gram-negative strain N. meningitidis
(H44/76) was provided by Dr. Andy Gorringe, University of Bath,
UK. Staph. aureus cells were grown on LB agar (Invitrogen), and all
the other strains were grown on GC agar (BD Biosciences)
supplemented with 1% IsoVitaleX (BD Biosciences) for 16−18 h at
37 °C in a moist atmosphere containing 5% CO2. Cells were harvested
and diluted in brain heart infusion broth (BHI, BD Biosciences)
supplemented with 1% IsoVitaleX. The optical density of each
inoculum (λ = 600 nm for Staph. aureus and S. pneumoniae, λ = 550
nm for N. meningitidis) was measured, and the inoculum was adjusted
to ∼106 cfu/mL. DMSO stock solutions were made for compounds
1−6 at 5 mg/mL. Each compound was then diluted in BHI broth
supplemented with 1% IsoVitaleX to 256 μg/mL (dilute 51.2 μL of
DMSO stock solution into 948.8 μL of BHI + 1% IsoVitaleX, DMSO
at 5.12% after dilution). A serial dilution at 0.5-fold per step, with 256
μg/mL being the highest concentration and 0.125 μg/mL being the
lowest, was prepared for each compound. The analogous serial
dilutions were also done for the antibiotics vancomycin (BioShop) and
kanamycin A (BioShop) and were used as positive controls. The serial
dilutions were then transferred onto sterile round-bottom 96-well
microtiter plates (Corning) at 50 μL per well. Bacterial cells were then
applied at 50 μL per well using the ∼106 cfu/mL inoculum prepared.
This yielded an inoculum at ∼5 × 105 cfu/mL per well and the
following concentrations of compounds: 128, 64, 32, 16, 8, 4, 2, 1, 0.5,
0.25, 0.125, and 0.0625 μg/mL. Negative control wells (3), which
contained only DMSO instead of the test compound/antibiotic, and
blank control wells (3), which contained only the growth media but
A55456 Factor E (5).6 This was synthesized according to general
procedure B using amine 41 (74.6 mg, 0.128 mmol, 1.0 equiv) in
CH2Cl2 (4 mL), sorbyl chloride (29.0 mg, 0.222 mmol, 1.7 equiv) in
CH Cl (2 mL), and Hunig’s base (50.0 μL, 0.287 mmol, 2.2 equiv),
̈
2
2
20 h. Flash chromatography (19:1 EtOAc−MeOH) of the crude
organic extracts gave 75.4 mg of 5 in 87% yield. Semipreparative
reversed-phase HPLC of the aforementioned silica gel purified product
gave 5 as a white solid: decomposition temp 145 °C (CH2Cl2); Rf 0.28
(19:1 EtOAc−MeOH); [α]24D −49.1 (c 0.35, CHCl3); IR (thin film in
CH2Cl2) νmax 3500 (br), 3281 (br), 2982, 2936, 1732, 1650 (br),
1437, 1265, 1159, 1111, 1003 cm−1; 1H NMR (CDCl3, 400 MHz, c 16
mg/mL) δ 8.56 (1H, d, J = 9.5 Hz, NH(1)), 7.32−7.22 (3H, m, H-26,
H-30), 7.21−7.13 (3H, m, H-25, H-27), 7.13−6.96 (2H, m, NH(2),
NH(3)), 6.27−6.17 (2H, m, H-29, H-31), 6.14−6.04 (1H, m, H-32),
5.15 (1H, dd, J = 8.5, 3.0 Hz, H-14), 4.89 (1H, dq, J = 9.5, 6.5 Hz, H-
7), 4.82 (1H, dd, J = 11.5, 1.5 Hz, H-20a), 4.78 (1H, q, J = 7.0 Hz, H-
10), 4.73−4.65 (1H, m, H-22), 4.51 (1H, ddd, J = 9.5, 9.5, 1.5 Hz, H-
19), 4.44 (1H, d, J = 8.0 Hz, H-2), 3.73 (1H, ddd, J = 11.5, 8.0, 5.0 Hz,
H-17a), 3.66−3.50 (3H, m, H-5a, H-17b, H-20b), 3.34−3.26 (1H, m,
H-5b), 3.07−2.90 (2H, m, H-23), 2.82 (3H, s, N-CH3 (H-12)), 2.41−
2.29 (1H, m, H-15a), 2.22−2.10 (2H, m, H-3b, H-16a), 2.03−1.92
(3H, m, H-3a, H-15b, H-16b), 1.91−1.78 (5H, m, H-4, H-33), 1.52
(3H, d, J = 7.0 Hz, H-11), 1.36 (3H, d, J = 6.5 Hz, H-8); 13C NMR
(CDCl3, 100 MHz, c 16 mg/mL) δ 172.9 (C, C-1), 172.3 (C, C-13),
171.6 (C, C-21), 170.1 (C, C-9), 170.0 (C, C-6), 166.5 (C, C-28),
164.8 (C, C-18), 141.6 (CH, C-30), 137.7 (CH, C-32), 136.4 (C, C-
24), 130.2 (CH, C-31), 129.6 (CH, C-25), 128.7 (CH, C-26), 126.9
(CH, C-27), 121.8 (CH, C-29), 65.1 (CH2, C-20), 59.3 (CH, C-2),
56.6 (CH, C-14), 56.4 (CH, C-10), 55.3 (CH, C-22), 51.2 (CH, C-
19), 48.1 (CH, C-7), 47.1 (CH2, C-5), 46.7 (CH2, C-17), 39.0 (CH2,
C-23), 31.0 (2C; N-CH3, C-12 and CH2, C-15), 30.8 (CH2, C-3), 23.3
(CH2, C-16), 21.3 (CH2, C-4), 18.7 (CH3, C-33), 17.6 (CH3, C-8),
15.8 (CH3, C-11); LRMS (ESI+) m/z (rel intensity) 701.4 (100) [M +
Na]+, 679.4 (27) [M + H]+; HRMS (ESI+) m/z calcd for C35H47N6O8
[M + H]+ 679.3449, found 679.3462; analytical HPLC (C18 column, 5
μm, 4.6 × 150 mm, gradient elution with water−acetonitrile, flow rate
2.5 mL/min, column temp 23 °C, detector λ 280 nm, tR 5.14 min).
A55456 Factor H (6).6 To a solution of THP-protected
depsipeptide 42 (62.2 mg, 0.784 mmol, 1.0 equiv) in EtOH (4 mL)
was added PPTS (8.7 mg, 0.035 mmol, 44 mol %), and the reaction
mixture was heated in a 60 °C oil bath under nitrogen (flask fitted with
a water-cooled reflux condenser). Complete consumption of the
starting material was observed by TLC after 5 h, and the reaction
mixture was diluted with water (20 mL) and extracted with CH2Cl2 (3
× 20 mL). The combined organic layers were then dried over MgSO4
J
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
without bacteria, were set in the plate as well. A total of three
independently constructed plates were prepared for each bacterial
strain. All plates were sealed with a Breath-Easy sealing membrane
(Sigma-Aldrich), and they were incubated for 16−18 h at 37 °C in a
moist atmosphere containing 5% CO2. Cell growth was determined by
visually examining for the presence of cell pellets or turbidity of the
wells. The lowest concentration of compound used that did not exhibit
any visible cell growth was taken as the MIC value.
Ganesan, A. Org. Lett. 2011, 13, 6334−6337. (e) Sleebs, M. M.;
Scanlon, D.; Karas, J.; Maharani, R.; Hughes, A. B. J. Org. Chem. 2011,
76, 6686−6693. (f) Narayanaswamy, V. K.; Albericio, F.; Coovadia, Y.
M.; Kruger, H. G.; Maguire, G. E. M.; Pillay, M.; Govender, T. J. Pept.
Sci. 2011, 17, 683−689. (g) Wang, D.; Jia, X.; Zhang, A. Org. Biomol.
Chem. 2012, 10, 7027−7030. (h) Cochrane, J. R.; Yoon, D. H.;
McErlean, C. S. P.; Jolliffe, K. A. Beilstein J. Org. Chem. 2012, 8, 1344−
1351. (i) Lam, H. Y.; Zhang, Y.; Liu, H.; Xu, J.; Wong, C. T. T.; Xu,
C.; Li, X. J. Am. Chem. Soc. 2013, 135, 6272−6279.
ASSOCIATED CONTENT
* Supporting Information
(5) For a review of the total synthesis of selected antimicrobial and
antitumor cyclic depsipeptides, see: Li, W.; Schlecker, A.; Ma, D.
Chem. Commun. 2010, 46, 5403−5420.
■
S
Additional experimental procedures and compound data
pertaining to other synthesized intermediates, copies of 1D
NMR spectra (1H and 13C) for all synthesized numbered
compounds, 2D NMR spectra (COSY, TOCSY, HSQC,
HMBC, and ROESY) for selected compounds 1−6, compar-
(6) Michel, K. H.; Kastner, R. E. A54556 Antibiotics and Process for
Production Thereof. U.S. Patent 4,492,650, January 8, 1985.
(7) Koshino, H.; Osada, H.; Yano, T.; Uzawa, J.; Isono, K.
Tetrahedron Lett. 1991, 32, 7707−7710.
(8) Schmidt, U.; Neumann, K.; Schumacher, A.; Weinbrenner, S.
Angew. Chem., Int. Ed. Engl. 1997, 36, 1110−1112.
1
ison H and 13C NMR spectra of synthetic and isolated for
compounds 1 and 2, analytical HPLC chromatograms for
compounds 1−6, semipreparative HPLC chromatograms for
compounds 1−6 and 26−29, ORTEP of X-ray crystal structure
and CIF for compound 38, Ramachandran plot for compound
38, and fluorometric assay data for compounds 1−6. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(9) Schmidt, U.; Langner, J. J. Pept. Res. 1997, 49, 67−73.
(10) Hardy, S.; Legagneux, V.; Audic, Y.; Paillard, L. Biol. Cell 2010,
102, 561−580.
(11) (a) Brotz-Oesterhelt, H.; Beyer, D.; Kroll, H.-P.; Endermann,
̈
R.; Ladel, C.; Schroeder, W.; Hinzen, B.; Raddatz, S.; Paulsen, H.;
Henninger, K.; Bandow, J. E.; Sahl, H.-G.; Labischinski, H. Nat. Med.
2005, 11, 1082−1087. (b) Sass, P.; Josten, M.; Famulla, K.; Schiffer,
G.; Sahl, H.-G.; Hamoen, L.; Brotz-Oesterhelt, H. Proc. Natl. Acad. Sci.
̈
U.S.A. 2011, 108, 17474−17479.
AUTHOR INFORMATION
Corresponding Author
*Tel: 416-978-5059. E-mail: rbatey@chem.utoronto.ca.
̌
■
(12) Kress, W.; Maglica, Z; Weber-Ban, E. Res. Microbiol. 2009, 160,
618−628 and references therein.
(13) Yu, A. Y. H.; Houry, W. A. FEBS Lett. 2007, 581, 3749−3757.
(14) Lee, B.-G.; Park, E. Y.; Lee, K.-E.; Jeon, H.; Sung, K. H.;
Notes
Paulsen, H.; Rubsamen-Schaeff, H.; Brotz-Oesterhelt, H.; Song, H. K.
̈
̈
The authors declare no competing financial interest.
Nat. Struct. Mol. Biol. 2010, 17, 471−478.
(15) Li, D. H. S.; Chung, Y. S.; Gloyd, M.; Joseph, E.; Ghirlando, R.;
ACKNOWLEDGMENTS
Wright, G. D.; Cheng, Y.-Q.; Maurizi, M. R.; Guarne,
Chem. Biol. 2010, 17, 959−969.
(16) (a) Hinzen, B.; Raddatz, S.; Paulsen, H.; Lampe, T.;
Schumacher, A.; Habich, D.; Hellwig, V.; Benet-Buchholz, J.;
́
A.; Ortega, J.
■
The authors are grateful for financial support through the
Natural Science and Engineering Research Council (NSERC)
of Canada via a Discovery Grant for R.A.B. and a PGS-D
scholarship to J.D.G. This work was supported in part by the
Canadian Institutes of Health Research Emerging Team Grants
from the Institute of Infection and Immunity (XNE-86945) to
S.G.-O., R.A.B., and W.A.H. The authors also wish to
acknowledge the Canadian Foundation for Innovation (CFI
project number 19119) and the Ontario Research Fund for
funding of the Centre for Spectroscopic Investigation of
Complex Organic Molecules and Polymers (CSICOMP). Dr.
Y.-Q. Cheng (University of Wisconsin−Milwaukee) and Dr. M.
Siegel (Eli Lilly and Company) are thanked for providing
authentic samples of natural products 1 and 2. J.G. wishes to
thank Dr. T. Burrow, Dr. D. C. Burns, and Mr. D. Pichugin of
the CSICOMP facility for their assistance with NMR
experiments. J.G. also wishes to thank C. J. White for assistance
with purification of ADEPs via reversed-phase HPLC and
discussions on peptide chemistry.
̈
Endermann, R.; Labischinski, H.; Brotz-Oesterhelt, H. ChemMedChem
̈
2006, 1, 689−693. (b) Socha, A. M.; Tan, N. Y.; LaPlante, K. L.; Sello,
J. K. Bioorg. Med. Chem. 2010, 18, 7193−7202. (c) Carney, D. W.;
Schmitz, K. R.; Truong, J. V.; Sauer, R. T.; Sello, J. K. J. Am. Chem. Soc.
2014, 136, 1922−1929.
(17) Conlon, B.; Nakayasu, E.; Fleck, L.; LaFleur, M.; Isabella, V.;
Coleman, K.; Leonard, S.; Smith, R.; Adkins, J.; Lewis, K. Nature 2013,
503, 365−370.
(18) Leung, E.; Datti, A.; Cossette, M.; Goodreid, J.; McCaw, S. E.;
Mah, M.; Nakhamchik, A.; Ogata, K.; El Bakkouri, M.; Cheng, Y.-Q.;
Wodak, S. J.; Eger, B. T.; Pai, E. F.; Liu, J.; Gray-Owen, S.; Batey, R.
A.; Houry, W. A. Chem. Biol. 2011, 18, 1−12.
(19) Gersch, M.; Gut, F.; Korotkov, V. S.; Lehmann, J.; Bottcher, T.;
̈
Rusch, M.; Hedberg, C.; Waldmann, H.; Klebe, G.; Sieber, S. A. Angew.
Chem., Int. Ed. 2013, 52, 3009−3014 and references therein.
(20) White, C. J.; Yudin, A. K. Nat. Chem. 2011, 3, 509−524.
(21) Ley, S. V.; Smith, S. C.; Woodward, P. R. Tetrahedron 1992, 48,
1145−1174.
(22) Deslongchamps, P.; Lamothe, S.; Lin, H.-S. Can. J. Chem. 1987,
65, 1298−1307.
REFERENCES
■
(23) Corey, E. J.; Suggs, J. W. Tetrahedron Lett. 1975, 2647−2650.
(1) von Nussbaum, F.; Brands, M.; Hinzen, B.; Weigand, S.; Habich,
̈
(24) Hinzen, B.; Brotz-Oesterhelt, H.; Endermann, R.; Henninger,
̈
D. Angew. Chem., Int. Ed. 2006, 45, 5072−5129 and references therein.
(2) Clardy, J.; Fischbach, M. A.; Walsh, C. T. Nat. Biotechnol. 2006,
24, 1541−1550 and references therein.
(3) Baltz, R. H.; Maio, V.; Wrigley, S. K. Nat. Prod. Rep. 2005, 22,
717−741.
K.; Paulsen, H.; Raddatz, S.; Lampe, T.; Hellwig, V.; Schumacher, A.
Antibacterial Macrocycles. U.S. Patent 0,107,288 A1, May 19, 2005.
(25) TPTU = (2-(2-pyridon-1-yl)-1,1,3,3-tetramethyluronium tetra-
fluoroborate.
(26) For comparative studies on coupling reagents, see: (a) Davies, J.
S.; Howe, J.; Le Breton, M. J. Chem. Soc. Perkin Trans. 2 1995, 2335−
2339. (b) Albericio, F.; Bofill, J. M.; El-Faham, A.; Kates, S. A. J. Org.
Chem. 1998, 63, 9678−9683. (c) Pon, R. T.; Yu, S.; Sanghvi, Y. S.
Bioconjugate Chem. 1999, 10, 1051−1057. (d) Hachmann, J.; Lebl, M.
(4) For selected recent total syntheses of cyclic depsipeptide natural
products, see: (a) Li, W.; Gan, J.; Ma, D. Angew. Chem. 2009, 121,
9053−9057. (b) Seo, H.; Lim, D. J. Org. Chem. 2009, 74, 906−909.
(c) Schmauder, A.; Sibley, L. D.; Maier, M. E. Chem.Eur. J. 2010, 16,
4328−4336. (d) Benelkebir, H.; Donlevy, A. M.; Packham, G.;
K
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Biopolymers (Pept. Sci.) 2006, 84, 340−347. For recent reviews on
coupling reagents, see: (e) Han, S.-Y.; Kim, Y.-A. Tetrahedron 2004,
60, 2447−2467. (f) Montalbetti, C. A. G. N.; Falque, V. Tetrahedron
2005, 61, 10827−10852. (g) Valeur, E.; Bradley, M. Chem. Soc. Rev.
2009, 38, 606−631. (h) El-Faham, A.; Albericio, F. Chem. Rev. 2011,
111, 6557−6602. (i) Pattabiraman, V. R.; Bode, J. W. Nature 2011,
480, 471−479.
(27) HATU = 1-[(dimethylamino)(dimethyliminium)methyl]-1H-
1,2,3-triazolo[4,5-b]pyridine-3-oxide hexafluorophosphate.
(28) In the final amide coupling step for compounds 1 and 2, the
reaction was protected from light by wrapping the vessel in aluminum
foil. Following aqueous workup, the compounds were immediately
purified by flash chromatography and concentrated into amber vials.
To minimize decomposition, the compounds should be stored as dry
solids in a cold freezer (−10 °C or colder) when not in use, as they are
not stable at room temperature for prolonged periods of time.
(29) For a Ramachandran plot for compound 38, see the Supporting
Information.
(30) For a detailed labeled ROESY spectrum of 1 showing all 1H−1H
NOE correlations pertinent to stereochemical assignment and for a
labeled HMBC spectrum showing all 1H−13C interactions pertinent to
the assignment of carbonyl carbons, see the Supporting Information.
(31) NOE correlations observed for H-2 in the ROESY of 2 showed
1
that H-3b is significantly shifted downfield in the H NMR (δ 2.16)
when compared with the analogous signal in 1 (δ 1.77). For a detailed
1
labeled ROESY spectrum of 2 showing all H−1H NOE correlations
pertinent to stereochemical assignment see the Supporting Informa-
tion.
(32) Balaram, P. Proc. Indian Acad. Sci., Chem. Sci. 1985, 95, 21−38.
(33) For experimental details on the isolation of 1 and 2 from S.
hawaiiensis, see the Supporting Information.
(34) For a comparison of the 1H and 13C NMR data of 1 and 2 with
the isolated natural products from S. hawaiiensis, see the Supporting
Information.
(35) For a more detailed explanation of the fluorometric assay, see
the Supporting Information.
(36) Osada, H.; Yano, T.; Koshino, H.; Isono, K. J. Antibiot. 1991, 44,
1463−1466.
(37) Addition of the membrane permeabilizing agent polymyxin B
nonapeptide to the culture broth had no measureable beneficial effect
on the observed antibacterial activity.
(38) Wiegand, I.; Hilpert, K.; Hancock, R. E. W. Nat. Protoc. 2008, 3,
163−175.
L
dx.doi.org/10.1021/np500158q | J. Nat. Prod. XXXX, XXX, XXX−XXX