Discovery, synthesis, and insecticidal activity of cycloaspeptide E

Article abstract of DOI:10.1021/np060219c

Several Penicillia and one Tricothecium strain produced a new, insecticidally active member of the cycloaspeptide family, with the proposed name cycloaspeptide E (1). The structure, which was determined on the basis of spectroscopic (NMR, UV, MS) data and Marfey amino acid analysis, was the tyrosine desoxy version of cycloaspeptide A (2). Two synthetic routes to compound 1 were developed: one a partial synthesis from 2 and the other a total synthesis from methyl alaninate hydrochloride. Cycloaspeptide E, the first member of this series not to contain a tyrosine moiety, is also the first to be reported with insecticidal activity.

Full text of DOI:10.1021/np060219c

1506
J. Nat. Prod. 2006, 69, 1506-1510
Discovery, Synthesis, and Insecticidal Activity of Cycloaspeptide E
Paul Lewer,*,† Paul R. Graupner,† Donald R. Hahn,† Laura L. Karr,† Dennis O. Duebelbeis,† Justin M. Lira,†
Peter B. Anzeveno,† Stephen C. Fields,† Jeffrey R. Gilbert,† and Cedric Pearce‡
Dow AgroSciences, 9330 ZionsVille Road, Indianapolis, Indiana 46268, and Mycosynthetix, Inc., 4905 Pine Cone DriVe, Suite 6,
Durham, North Carolina 27707
ReceiVed May 12, 2006
Several Penicillia and one Tricothecium strain produced a new, insecticidally active member of the cycloaspeptide
family, with the proposed name cycloaspeptide E (1). The structure, which was determined on the basis of spectroscopic
(NMR, UV, MS) data and Marfey amino acid analysis, was the tyrosine desoxy version of cycloaspeptide A (2). Two
synthetic routes to compound 1 were developed: one a partial synthesis from 2 and the other a total synthesis from
methyl alaninate hydrochloride. Cycloaspeptide E, the first member of this series not to contain a tyrosine moiety, is
also the first to be reported with insecticidal activity.
Table 1. NMR Data for Cycloaspeptide E (1) at 600 MHz (150
Fungi continue to be an exciting source of novel active molecules
that add to chemical diversity and contribute screening inputs to
both pharmaceutical and agrochemical lead generation programs.
As part of our ongoing search for natural products with potential
value as agrochemicals, a set of fungi was collected over several
years from leaf litter, wood, and twigs. Initial screening of small
fermentations of these fungi revealed several strains (Table S1;
Supporting Information) that were insecticidally active. LC-MS-
bioassay profiling showed the presence of the same unidentified,
low-abundance insecticidal compound in each strain. These obser-
vations prompted us to perform bioassay-directed isolation of the
active molecules from large-scale fermentations of several of these
strains. Three molecules were isolated and identified as a result:
penitrem A,1 a known insecticidal metabolite, was identified from
one strain (DA051320); a small amount of a novel, bioactive cyclic
peptide (1) related to cycloaspeptide A (2) and a larger amount
(approximately 60× the level of 1) of cycloaspeptide A2,3 itself
were present in all strains examined. Interestingly, 2 was inactive,
although it was helpful in the structure elucidation of 1 and in
enhancing our understanding of the structure-activity relationship
for this series. In order to further understand the potential for this
class of peptides as insecticidal agents, the synthesis of cycloaspep-
tides was investigated, and two successful routes to these com-
pounds are reported here.
MHz for 13C) in CDCl3
residue
position
δC
δH
NH
CR
6.60 (1H, d, J ) 7.0 Hz)
4.41 (1H, dq, J ) 7.0, 6.6 Hz)
0.43 (1H, d, J ) 6.6 Hz)
44.4
16.8
173.9
30.4
63.7
34.4
Ala
Câ
CO
N-Me
CR
2.89 (3H, s)
5.21 (1H, dd, J ) 11.8, 3.0 Hz)
3.49 (1H, dd, J ) 14.3, 2.6 Hz)
3.05 (1H, dd, J ) 14.9, 12.3 Hz)
Câ
Phe(1)
C1
C2-5
C6
139.1
129
127
7.36-7.28 (4H, m)
7.35 (1H, m)
CO
NH
CR
168.3
7.14 (1H, d, J ) 8.7 Hz)
4.72 (1H, ddd, J ) 8.5, 8.5, 5.6)
1.88 (1H, m)
48.8
41.7
Câ
1.38 (1H, m)
Leu
Cγ
25.2
23.7
22.4
170.2
39.5
70.0
33.4
1.71 (1H, m)
1.02 (3H, d, J ) 6.7 Hz)
1.01 (3H, d, J ) 6.7 Hz)
Cδ
Cꢀ
CO
N-Me
CR
Câ
2.70 (3H, s)
3.87 (1H, dd, J ) 10.8, 3.7 Hz)
3.62 (1H, dd, J ) 13.8, 4.1 Hz)
3.47 (1H, dd, J ) 14.3, 11.3 Hz)
Phe(2)
C1
C2-5
C6
138.1
129
127
7.36-7.28 (4H, m)
7.35 (1H, m)
CO
NH
C1
C2
C3
C4
C5
C6
CO
171.0
12.03 (1H, s)
115.4
142.0
121.4
134.7
122.7
127.2
170.0
8.95 (1H, d, J ) 8.7 Hz)
ABA
7.51 (1H, dd, J ) 8.7, 8.2 Hz)
7.02 (H, dd, J ) 8.2, 8.2 Hz)
7.45 (1H, d, J ) 8.2 Hz)
molecular weight of 1 to be 625, i.e., 16 units less than for 2,
suggesting 1 to be a desoxy analogue of 2. High-resolution ESIMS
supported this proposal, with an experimentally determined accurate
molecular weight for 1 of 625.3286, in good agreement with the
calculated weight for the desoxy analogue of 2 [C36H43N5O6 )
625.3264]. 1H NMR revealed that the tyrosine residue present in 2
had been replaced with a phenylalanine group in compound 1;
otherwise the amino acid constituents were the same (Table 1).
Analysis of an HMBC spectrum indicated that the amino acids in
1 were connected in the same order as for 2. Finally, a small amount
of 1 was hydrolyzed, derivatized using Marfey’s reagent,4 and ana-
lyzed by LC-MS. This confirmed the presence of N-methylphen-
Both 1 and 2 showed λmax ) 256 and 307 nm in their diode
array UV spectra during HPLC. Further, ESIMS determined the
* To whom correspondence should be addressed. Tel: 317-337-3671.
Fax: 317-337-3546. E-mail: plewer@dow.com.
† Dow AgroSciences.
‡ Mycosynthetix.
10.1021/np060219c CCC: $33.50
© 2006 American Chemical Society and American Society of Pharmacognosy
Published on Web 09/23/2006

Notes
Journal of Natural Products, 2006, Vol. 69, No. 10 1507
assembled to the pentapeptide (5) in good overall yield with no
aminobenzoate cleavage (Scheme 1). Macrocyclization of 5 with
DPPA yielded 1 in low yield (10-15%), accompanied by formation
of a major byproduct. Evidence from NMR suggested that this
byproduct was the acylated aminal (6) resulting from azide attack
on the phosphoryl-activated alanine carboxyl group, followed by
Curtius-type rearrangement (Scheme 2). However, this product was
not exhaustively characterized. This side reaction was circumvented
by use of the cyclic anhydride of propane-1-phosphonic acid, which
yielded 1 in a more respectable 67% yield from 5.
In the present work, several Penicillia and one Tricothecium
strain were shown to produce the insecticidal pentapeptide cyclo-
aspeptide E (1),8 which was isolated chromatographically guided
by its BAW activity. Compound 1 is an analogue of the known
cycloaspeptide A, which was also produced by these fungi at much
(60×) higher levels, but shown to be insecticidally inactive.
Previously, members of this structural class were reported to be
produced by Penicillium algidum,2 P. jamesonlandense,9 P. ribeum,9
P. soppii,9 and P. lanosum9 (cycloaspeptide A), P. ribeum5
(cycloaspeptide D), and Aspergillus sp. NE-452 (cycloaspeptides
A, B, and C).
Cycloaspeptide E (1) was moderately active against beet army-
worm following ingestion from artificial diet, showing an MIC of
approximately 10 ppm, whereas, under the same conditions,
cycloaspeptide A (2) was inactive, with MIC > 500 ppm (Table
S2; Supporting Information). By comparison, spinosyn A, the major
factor in the commercial product Spinosad, had MIC ) 0.02 ppm
under these conditions. Cycloaspeptide E showed activity following
topical treatment directly to both beet armyworm and cabbage
looper, as well as following consumption of treated leaves. Again,
under the same conditions, 2 was essentially inactive. Finally,
following injection into beet armyworms, test conditions that
remove intrinsic barriers to activity (gut hydrolysis, cuticle penetra-
tion, etc.), 1 was active, whereas 2 showed only, at best, fleeting
initial activity. A characteristic symptomology of larvae injected
with 1 was full-body tremors, consistent with a neurotoxic mode
of action.
Compound 1 is the first member of the cycloaspeptide series
not containing a tyrosine residue, as well as being the first-reported
insecticidal example. The total synthesis reported here is the first
such route reported for this class. The discovery of compound 1
and its insecticidal activity represents another example of the
contribution fungi can play in the generation of novel agrochemical
leads via studies of natural products.
Figure 1. 2D representation of the structure of the anthranilic acid
portion of the cycloaspeptides. The single-headed arrow indicates
a key HMBC correlation, whereas the double-headed arrow denotes
a key NOE signal determining the assignments of H-6 and H-3.
Note the proximity of H-3 to the Phe(2) carbonyl, causing a large
downfield shift for H-3.
ylalanine, alanine, and leucine as the constituent amino acids. This
analysis also showed that each amino acid was present in the
L-configuration, as had been observed for previous members of this
series.2,3,5 Thus the structure of 1 was confirmed to be the desoxy
analogue of 2. Large-scale fermentations of strains DA087002,
DA087027, DA087041, and DA087044 gave similar relative
production levels of 1 and 2 without production of penitrem A.
One noteworthy feature of the NMR spectra of the cycloaspep-
tides is the far downfield chemical shift of the anthranilate H-3 in
all solvents. In earlier assignments of the structure2,5 the doublet
signal at δ 8.5 was assigned to H-6. However in this work, this
peak did not show a cross-peak to the anthranilate carbonyl atom
in any HMBC experiments; rather, the doublet at δ 7.45 did.
Additionally, in gsNOE experiments, irradiation of Ala-NH gave
an enhancement to this doublet, assigning it to H-6, and H-3 to the
signal at δ 8.95. The reason for this low-field shift for H-3 may be
attributed to a strong hydrogen bond between the anthranilic acid
NH and carbonyl. With a small cyclic peptide, this forces the
carbonyl of the phenylalanine (or tyrosine in cycloaspeptides A-D)
toward the plane of the anthranilic acid phenyl ring, thus strongly
deshielding H-3 and also bringing Ala-NH close to H-6 (Figure
1). These observations were consistent between different solvents
(DMSO, CDCl3, and MeOD).
The observation of lepidopteran activity for compound 1 was
novel within this series of natural products, and the activity
differential between 1 and its OH-analogue 2 prompted us to further
investigate this series. While milligram amounts of 1 and gram
amounts of 2 were subsequently obtained by scale-up fermentation
and isolation, this was not considered a sustainable source to provide
starting material for synthesis, especially if analogues were needed.
Since structure 1 looked synthetically approachable either by total
synthesis or via partial synthesis from naturally occurring 2, we
decided to investigate both routes.
Partial synthesis of 1 from naturally occurring 2 was accom-
plished via reduction of the triflate (3). While standard methods of
aryl triflate reductions either did not work or gave significant
hydrolysis, the method of Kotsuki6 using triethylsilane as the
hydrogen source was effective. Milligram-scale formation of the
triflate followed by reduction gave 73% purified (HPLC) yield of
1. Our total synthesis route was modeled on the published route7
for the related avellanins, using the Fmoc protocol. In the synthesis
of the avellanins, a linear peptide was generated starting from
methyl-2-aminobenzoate, with macrocyclization being accomplished
in 29% yield using diphenyl phosphoramidate (DPPA). It rapidly
became apparent that synthesis of the cycloaspeptides could not
be similarly accomplished. Thus, deprotection of Fmoc or BOC
moieties using standard conditions from a suitable tri- or tetrapeptide
precursor to 1 resulted in significant hydrolysis of the benzoate
moiety, decreasing yields substantially. Success was accomplished,
however, by starting with alanine already linked to the amino
benzoate (4) and protected as the methyl ester, which could be
Experimental Section
General Experimental Procedures. All NMR spectra were recorded
on a Bruker DRX600 spectrometer operating at 600.13 MHz (1H) and
150.62 MHz (13C). HPLC separations were achieved using a Hewlett-
Packard 1100 LC system. LC-MS was performed on a Micromass
Platform single-quadrupole mass spectrometer in both positive elec-
trospray (+ESI) and negative electrospray (-ESI) modes. Accurate
LC-MS and LC-MS/MS experiments were conducted on a Micromass
hybrid quadrupole time-of-flight (Q-Tof) instrument.
General Synthetic Procedures. A: Peptide Coupling. Oxalyl
chloride (5 equiv) was added in one portion to a stirred solution of
Fmoc-protected amino acid (1 equiv) in DCM. DMF (1 drop) was then
added and the solution stirred under nitrogen for 1 h. The solvent and
excess oxalyl chloride were then removed at reduced pressure. The
residue was redissolved in DCM, and this solvent was then removed,
affording the crude acid chloride. This crude acid chloride was then
dissolved in DCM, and this solution was added in one portion to a
well-stirred solution of the deprotected peptide (1 equiv) in DCM. To
this solution saturated aqueous NaHCO3 solution was immediately
added, and this mixture was well stirred for 1 h. The reaction mixture
was then diluted with DCM and H2O. The aqueous phase was separated
and the organic phase washed successively with saturated NaHCO3, 1
N HCl, saturated NaHCO3, and brine, then dried (MgSO4). Concentra-
tion left a crude coupled product, which was purified by flash

1508 Journal of Natural Products, 2006, Vol. 69, No. 10
Scheme 1. Synthesis of Acyclic Precursor 5
Notes
Scheme 2. Cyclization of 5 to Synthetic 1
chromatography over 130 mL of silica eluted with 5-9% acetone in
DCM to afford pure coupled Fmoc-protected peptide.
containing fungus from a malt agar slant, was transferred into 7 mL of
YESD broth [soy peptone (Sigma P-1265; 2% w/v), dextrose (Sigma
G-8270; 2%), yeast extract (Becton Dickinson 4311929; 1%)] in a 50
mL centrifuge tube. This was incubated for 7 days at 22 °C on an
orbital shaker at 200 rpm with a 1 in. throw. For 2 L or larger
fermentations, the contents of the 7 mL starter culture above were
transferred to a 250 mL flask containing YESD medium (75 mL). This
was incubated a further 7 days at 22 °C. An aliquot (3 mL) from this
secondary culture was used to inoculate either a Nunc bioassay dish
containing solid medium (500 mL) or a 2 L Fernbach flask containing
liquid medium, and incubating at 22 °C for 11 days. The media11 used
for these scale-up fermentations (Table S1) were as follows: (A)
mannitol (0.5%), soy grits (0.2%); (B) glucose (1%), yeast extract
(0.1%), casein (0.2%), agar (1.8%); (C) oatmeal agar; (D) Czapek agar;
(E) mycological agar, low pH; (F) mannitol (0.5%), soy grits (0.2%),
agar (1.8%); (G) potato dextrose agar.
To prepare the completed fermentations for extraction, the plates
were frozen at -75 °C for 24 h, then freeze-dried until completely
dry. After freeze drying, the dried residue was transferred to a 5 L
graduated beaker. One liter of methanol (MeOH) was added to the
container and allowed to stand for approximately 10 h. The extract
was filtered through 25 cm, grade 417, qualitative filter paper into an
Erlenmeyer graduated flask. After filtering, the residue plus paper was
rinsed with MeOH (500 mL) and the two MeOH phases were combined,
then dried under vacuum.
Insect Screening. Lepidopteran Diet Assay. Diet was dispensed
into the wells of a 96-well microtiter plate (100 µL/well) and allowed
to cool and dry. Test samples were dried in the wells of a second 96-
well plate and then dissolved in acetone-water (50:50; 50 µL) with
sonication. The test sample solutions were transferred to the plate
containing insect diet and allowed to dry on the surface of the diet.
Test compounds were placed thus on the surface of the diet at a series
of rates such that their final concentrations (expressed in ppm of test
compound per total weight of diet per well) were 0.98-500 ppm, in
2-fold steps. Each well was then infested with 8-10 beet armyworm
(BAW) eggs. Test plates were covered with a layer of sterile cotton
and then the plate lid. The effects of the test compounds on the
B: Deprotection of Fmoc Group. To a solution of protected peptide
(1 equiv.) in DCM was added 4-(aminomethyl)piperidine (excess) in
one portion at ambient temperature. The mixture was then stirred under
nitrogen for 1 h. After dilution with more DCM the organic phase was
washed successively with water, pH 5.4-5.5 buffer and brine and then
dried (Na2SO4). Concentration left deprotected peptide which was
usually pure enough to be used without further purification.
Fungal Collection and Fermentation Conditions. Fungal strains
examined in this study are listed in Table S1 (Supporting Information).
These strains are maintained in the Mycosynthetix fungal repository
in Durham, NC. Compounds 1 and 2 were isolated from fermentation
of strains DA051320, DA087002, DA087027, DA087041, and
DA087044, and their structures were determined spectroscopically
(NMR, MS, UV). Compounds 1 and 2 were detected in fermentation
broths from the remaining strains by LC-MS analysis only. Strain
DA051320 did not form aerial mycelia or conidiophores on any media
tested; therefore, morphological identification could not be completed.
A partial DNA sequence of the large subunit ribosomal RNA gene
(28S) was determined for strain DA05132 at Midi Labs (Newark, DE).
The DNA sequence was compared to the GenBank database using the
BLASTN algorithm (http//www.ncbi.nlm.nih.gov/BLAST/). The stron-
gest BLAST match for DA058132 was Trichothecium roseum (U69891).
The four remaining strains from which cycloaspeptides were isolated
all produced abundant aerial mycelium on a variety of media with a
dark green pigment and conidiophores characteristic of Penicillium spp.
A partial DNA sequence of the small subunit ribosomal RNA gene
(18S) was determined10 for strains DA087041 and DA087044. The
strongest BLAST matches for each of these two strains were to various
Penicillium species. The partial rRNA gene sequences were deposited
in GenBank under the following accession numbers: Tricothecium sp.
DA051320 (DQ497012), Penicillium sp. DA087041 (DQ443729), and
Penicillium sp. DA087044 (DQ443730).
Fungi were cultured on a small scale in multiple media for initial
screening purposes. The following protocol was used to prepare larger
amounts of broth for scale-up metabolite isolation. A small agar plug,

Notes
Journal of Natural Products, 2006, Vol. 69, No. 10 1509
development of the insects were evaluated after a 6-day incubation
period at 28 °C. The insecticidal potency of the test compounds was
reported as the minimum concentration of compound (MIC, in ppm)
required to inhibit insect growth compared with an untreated control.
Lepidopteran Topical Assay. The compounds were dissolved in
acetone at 5 mg/mL. One-microliter treatments of the solutions of each
compound were applied along the dorsa of each of six third-instar BAW
and cabbage looper (CL) (Trichoplusia ni) using a microapplicator,
resulting in a treatment rate of 5 µg/larva. Solvent-only and untreated
insects were prepared as controls. Following topical application, insects
were placed individually into the wells of six-well tissue culture plates.
Each well also contained a 1 cm3 portion of artificial lepidopteran diet.
Insects were held under ambient laboratory conditions and were
observed for symptoms of intoxication and mortality at 24 and 120 h.
Lepidopteran Limited Ingestion Exposure Assay. To measure the
impact of limited ingestion exposure, 0.25 cm2 leaf disks were treated
with 1 µL applications of a 5 mg/mL acetone solution of the test
materials. Six cabbage disks (for CL) and six cotton disks (for BAW)
were treated and then permitted to dry. Solvent control and untreated
disks were also prepared. Disks were then offered to six of each species
of lepidopteran for a 24 h period. Larvae and leaf disks were held
individually in the wells of six-well tissue culture plates, under ambient
laboratory conditions. After 24 h, the larvae were offered a 1 cm3
portion of artificial lepidopteran diet, and mortality rate, symptoms of
intoxication, and amount of the leaf disks consumed were noted.
BAW Injection Assay. The compounds were dissolved in acetone
at a concentration of 20 mg/mL. To assess the intrinsic activity of the
cycloaspeptides, six fourth-instar BAW were each injected with a 0.5
µL dose of test solution (equivalent to 10 µg per insect). Solvent-only
and untreated insects were prepared as controls. After treatment, insects
were placed individually into the wells of a six-well tissue-culture plate.
Each well contained a 1 cm3 portion of artificial lepidopteran diet.
Treated insects were held under ambient laboratory conditions and were
observed for symptoms of intoxication at 1, 24, and 48 h, and for
mortality at 48 h.
Metabolite Isolation. The dried MeOH extract from 3 L of
DA051320 broth (1.9 g) was dissolved in water-dichloromethane
(1:1; 400 mL), shaken to partition, and separated. The aqueous phase
was extracted with two further aliquots (2 × 200 mL) of DCM and
the organic phase dried on a rotary evaporator. The dried sample was
dissolved in MeOH (1 mL), and the components were separated by
semipreparative HPLC using a Hypersil-C8-BDS column (250 × 10
mm; 8 µm) with gradient elution from 0% to 100% acetonitrile in 10
mM NH4OAc. This yielded 2 (3.4 mg) and a mixed fraction of 1 and
penitrem A. This mixed fraction was dissolved in MeOH (500 µL)
and further separated by HPLC on a Hypersil-C8-BDS column (250 ×
4.6 mm; 5 µm) eluting isocratically with 10 mM NH4OAc-acetonitrile
(1:1) to yield 1 and penitrem A. On the basis of analytical HPLC peak
area-matching using 2 from above as a response standard, the amount
of 1 isolated was estimated to be approximately 50 µg. NMR data are
summarized in Table 1. Fermentations of strains DA087002, DA087027,
DA087041, and DA087044 were performed on a scale of tens of liters
and worked up using the same methodology to provide larger quantities
of 1 and 2.
and the mixture was concentrated to a light brown oil in vacuo. The
triflate was found to be stable to both normal and reversed-phase
chromatography, aqueous bicarbonate, and dilute aqueous HCl. The
crude product was used in the next step without further purification:
MS (-ES) m/z for C37H42N5O8SF3 ) 772.3 [M - 1]-, 132.9 (base
peak); 1H NMR (CDCl3) 0.43 (d, J ) 7 Hz), 1.00 (d, J ) 7 Hz), 1.01
(d, J ) 7 Hz), 1.41-1.35 (m), 1.68-1.63 (m), 1.91-1.84 (m), 2.81
(s), 2.90 (s), 3.01-2.96 (m), 3.54-3.48 (m), 3.69-3.66 (m), 1.98 (br
dd, J ) 11 Hz, 5 Hz), 4.45-4.41 (m), 4.74 (dt, J ) 9 Hz, 6 Hz), 5.23
(dd, J ) 9 Hz, 3 Hz), 6.91 (br d, J ) 7 Hz), 7.05 (t, J ) 8 Hz), 7.36-
7.18 (m), 7.48 (d, J ) 8 Hz), 7.51 (t, J ) 9 Hz), 8.90 (d, J ) 9 Hz),
12.07 (s).
Reduction of the Triflate (3).6 The crude triflate was dissolved in
anhydrous DMF (2 mL), and to it was added 1 spatula tip of Pd(OAc)2
and 1 spatula tip of DPPF (more than catalytic amounts). The mixture
was heated to 60 °C, during which time it darkened in color. Et3SiH
(0.5 mL) was added and the mixture immediately turned black. The
temperature was raised to 100 °C and held there for 10 min, after which
time HPLC showed 30% conversion (method as above; tR starting
material ) 5.52 min, tR product ) 5.00 min). Further aliquots of the
reagents described above were added in the same order, and after
heating 10 more min conversion was 65%. Repeating the sequence a
third time gave 100% conversion, as determined by HPLC. The solution
was cooled, and MeOH (5 mL) was added. Suspended materials were
removed by filtration through a plug of Celite using MeOH as eluent.
The filtrate was concentrated in vacuo to give a brown oil, which was
chromatographed by reversed-phase HPLC, giving 9.0 mg (73% isolated
yield over two steps) of pure 1 as a crystalline solid, which was identical
by NMR to the natural material from fermentation.
Synthesis of Cycloaspeptide E from Methyl Alaninate Hydro-
chloride. 2-Nitrobenzoyl-L-Ala-OMe. To a cold (0-5 °C), well-stirred,
nitrogen-blanketed suspension of methyl alaninate‚HCl (1.4 g, 0.01 mol)
in a mixture of triethylamine (3.0 mL, 0.022 mol) and DCM (25 mL)
containing DMAP (∼50 mg) was added dropwise a solution of
2-nitrobenzoyl chloride (1.9 g, 0.01 mol) in DCM during 10 min,
maintaining the reaction temperature at 0-5 °C. This mixture was
stirred in the cold for 2.5 h, then diluted with DCM (40 mL) and water
(25 mL). The organic phase was separated and washed successively
with 1 N HCl (20 mL), saturated NaHCO3 (25 mL), and brine (25
mL) and then dried (Na2SO4). Concentration left 2.3 g (99%) of the
title compound as a colorless foam, which was pure enough to be used
without further purification: 1H NMR (CDCl3) δ 1.69 (d, 3, J ) 7
Hz), 3.92 (s,3), 4.94 (m, 1), 6.65 (d, 2, J ) 6 Hz), 7.64-7.74 (m, 2),
7.82 (m, 1), 8.20 (d, 1, J ) 8 Hz).
H-Ant-L-Ala-OMe. A solution of 2-nitrobenzyl-L-ala-OMe (1.0 g,
4.23 mmol) in ethanol (50 mL) was hydrogenated at 45 psi over 20%
Pd(OH)2/C catalyst for 4 h. The catalyst was removed by filtration
through a bed of Celite and washed with ethanol (25 mL). The
combined filtrate and wash was concentrated, leaving the title compound
(1.0 g, quant.) as a colorless oil: 1H NMR (CDCl3) δ 1.50 (d, 3, J )
7 Hz), 3.78 (s, 3), 4.75 (m, 1), 5.51 (br s, 2), 6.64-6.69 (m, 3), 7.15-
7.27 (m, 2), 7.39 (d, 2, J ) 7 Hz).
Fmoc-L-MePhe-Ant-L-Ala-OMe. This compound was prepared
using procedure A, but with a coupling reaction time of 2 h. From 1.7
g (4.25 mmol) of Fmoc-L-MePhe and 1.0 g (4.23 mmol) of Ant-L-ala-
OMe, 1.5 g (60%) of the title compound was obtained as a colorless
foam after chromatographic purification with 4% acetone in DCM as
eluent: 1H NMR (CDCl3) δ 1.34 (d, 3, J ) 7 Hz), 2.91 (s), 2.98 (s),
3.10 (m), 3.50 (m), 3.67 (s), 3.754 (s), 4.21-4.38 (m), 4.40-4.60 (m),
4.90 (m), 5.25 (m), 6.72 (m), 7.10-7.60 (m), 7.77 (m), 8.70 (m), 11.65
and 11.68 (s, total 1H).
Marfey Analysis. Compound 1 was dissolved in MeOH (2
mg/mL). An aliquot (50 µL; approximately 0.15 µmol) was placed in
a 1.8 mL amber glass HPLC autosampler vial and dried (Speedvac),
and the vial was flushed with nitrogen for 1 min. After adding 6 N
HCl (200 µL) and capping, the vial was heated at 110 °C for 15 h.
After cooling to room temperature, the sample was dried under vacuum.
The residue was dissolved in 1 M NaHCO3 (500 µL) and divided into
two equal aliquots. Each aliquot was treated with Marfey reagent (120
µL of 10 mM acetone solution) and allowed to react at 40 °C for 90
min. After cooling, the samples were quenched with 2 N HCl (150
µL), with gentle agitation, then allowed to sit for 5 min before being
dried under vacuum. The solid residues were dissolved in 50% aqueous
DMSO with sonication for 10 min, then analyzed by LC-MS.
Preparation of Triflate (3) from Natural 2. To a stirred solution
of 2 (12.5 mg) in CHCl3 (2 mL) at room temperature was added 2
drops of Et3N (excess) followed by 2 drops of triflic anhydride (excess).
The reaction was stirred for 10 min and monitored by HPLC (Kromasil-
H-L-MePhe-Ant-L-Ala-OMe. This compound was prepared using
procedure B. From 0.6 g (1.01 mmol) of the protected tripeptide, 0.4
g (quant.) of the title compound as a colorless oil was obtained. This
was used without further purification: 1H NMR (CDCl3) δ 1.52 (d, 3,
J ) 7 Hz), 2.17 (s), 2.34 (s), 2.76 (dd, J ) 13, 9 Hz), 3.30 (m,2), 3.81
(s, 3), 4.80 (m), 6.74 (d, 2, J ) 7 Hz), 7.12 (m, 1), 7.20-7.37 (m),
7.53 (m), 7.70 (m), 8.65 (d, 1, J ) 8 Hz), 11.65 (s, 1).
Fmoc-L-Leu-L-MePhe-Ant-L-Ala-OMe. This compound was pre-
pared using procedure A, but with a coupling time of 2.5 h. From 0.36
g (1.00 mmol) of Fmoc-L-Leu and 0.39 g (1.0 mmol) of l-MePhe-Ant-
L-Ala-OMe, 0.4 g (56%) of the title compound as a colorless foam
was obtained after chromatographic purification with 5% acetone in
DCM as eluent: 1H NMR (CDCl3) δ 0.69 (d, J ) 6 Hz), 0.93 (d, J )
6 Hz), 0.98 (d, J ) 6 Hz), 1.47 (d, J ) 7 Hz), 1.30-1.60 (m), 2.95 (s),
C
18 4.6 × 150 mm column, linear gradient from 50% to 100% MeCN
in H2O-TFA, 99.95:0.05; pH 2) over 5 min, then hold at 100% MeCN,
1.2 mL/min flow rate; tR starting material ) 3.5 min, tR product )
5.52 min). EtOH (1 mL) was added when no starting material remained,

1510 Journal of Natural Products, 2006, Vol. 69, No. 10
Notes
3.00 (m), 3.57 (dd, J ) 14, 6 Hz), 3.75 (s), 4.18 (m), 4.35 (d, J ) 6
Hz), 4.60 (m), 5.12 (d, J ) 9 Hz), 5.47 (d, J ) 7 Hz), 7.00-7.63 (m),
7.78 (d, J ) 8 Hz), 8.64 (d, J ) 8 Hz), 11.43 (s, 1).
Cycloaspeptide E (1) from Cyclization of L-MePhe-L-Leu-L-
MePhe-Ant-L-Ala-OH (5) with 1-Propanephosphoric Acid Cyclic
Anhydride. A solution of 133 mg (0.20 mmol) of l-MePhe-L-Leu-L-
MePhe-Ant-L-Ala-OH in DCM (30 mL) was added dropwise over 6 h
under nitrogen to a stirred solution of triethylamine (0.73 mL, 5.24
mmol), 1-propanephosphonic acid cyclic anhydride (T3P, 0.62 mL of
a 50% solution in EtOAc, ∼1.0 mmol), and DMAP (∼2.5 mg) in DCM
(125 mL). This solution was stirred for 48 h at ambient temperature,
then concentrated to dryness at reduced pressure. The residue (340 mg)
was chromatographed over 100 mL of flash silica using 3% MeOH in
DCM as eluent to give 84 mg (67%) of 1. LC-MS analysis of this
showed 3-5% of a racemized product. Recrystallization from EtOAc-
petroleum ether (1:4) gave pure product as colorless microcrystals,
identical by 1H NMR and LC-MS analysis to the natural material: mp
H-L-Leu-L-MePhe-Ant-L-Ala-OMe. This compound was prepared
using procedure B, with a reaction time of 1 h 30 min. From 0.53 g
(0.74 mmol) of Fmoc-L-Leu-L-MePhe-Ant-L-Ala-OMe, 0.38 g (quant.)
of the title compound as a colorless foam was isolated and used without
further purification: 1H NMR (CDCl3) δ 0.69 (d, J ) 6 Hz), 0.93 (d,
J ) 6 Hz), 0.98 (d, J ) 6 Hz), 1.40-1.60 (m), 2.95 (s), 3.00-3.20
(m), 3.50-3.65 (m), 3.75 (s), 4.70 (m), 6.78 (d, J ) 7 Hz), 7.10-7.30
(m), 7.60 (m), 8.60 (d, J ) 8 Hz), 11.42 (s), 11.62 (s); LRMS (CI) m/z
497 (M + H, 45), 275 (100), 190 (40).
Fmoc-L-MePhe-L-Leu-L-MePhe-Ant-L-Ala-OMe. This compound
was prepared using procedure A, but with a coupling time of 3.5 h.
From 0.29 g (0.72 mmol) of Fmoc-L-MePhe and 0.36 g (0.72 mmol)
of L-Leu-L-MePhe-Ant-L-Ala-OMe, 0.3 g (47%) of the title compound
as a colorless foam was obtained after chromatographic purification
over silica with 5% acetone in DCM as eluent: 1H NMR (CDCl3) δ
0.60 (d, J ) 6 Hz), 0.82 (d, J ) 6 Hz), 0.93 (d, J ) 6 Hz), 1.45 (d, J
) 7 Hz), 1.50 (m), 2.78 (s), 2.83 (s), 2.88 (s), 2.92 (m), 3.20 (m), 3.55
(dd, J ) 14, 6 Hz), 3.75 (s), 4.05-4.40 (m), 4.70 (m), 4.90 (m), 5.53
(m), 6.10 (d, J ) 8 Hz), 6.70 (m), 7.00-7.58 (m), 7.78 (d, J ) 8 Hz),
8.60 (d, J ) 8 Hz), 11.22 (s), 11.40 (s); LRMS (CI) m/z 880 (M + H,
20), 658 (100).
25
161-164 °C; [R]D - 203 (c 0.1 MeOH).
Supporting Information Available: Representative NMR spectra
(1H and HSQC) of 1; positive ionization electrospray mass spectrum
of 1; ultraviolet spectra of 1 and 2; summary of cycloaspeptide-
producing strains used in this study (Table S1); and comparison of
insecticidal activity of 1 and 2 (Table S2). This material is available
free of charge via the Internet at http://pubs.acs.org.
H-L-MePhe-L-Leu-L-MePhe-Ant-L-Ala-OMe. This compound was
prepared using procedure B with a reaction time of 2 h. From 0.3 g
(0.34 mmol) of Fmoc-L-MePhe-L-Leu-L-MePhe-Ant-L-Ala-OMe, 0.20
g (88%) of the title compound as a colorless foam was obtained after
chromatographic purification with 4% MeOH in DCM as eluent: 1H
NMR (CDCl3) δ 0.68 (d, J ) 6 Hz), 0.92 (d, J ) 6 Hz), 0.98 (d, J )
6 Hz), 1.48 (d, J ) 6 Hz), 1.42-1.70 (m), 2.06 (s), 2.65 (m), 2.90-
3.16 (m), 3.05 (s), 3.58 (dd, J ) 14, 6 Hz), 3.80 (s), 4.74 (t, J ) 7 Hz),
4.98 (m), 5.52 (dd, J ) 10, 6 Hz), 6.80 (d, J ) 7 Hz), 7.05-7.38 (m),
7.46-7.56 (m), 8.61 (d, J ) 7 Hz), 11.35 (s); LRMS (CI) m/z 658 (M
+ H, 90), 436 (100).
H-L-MePhe-L-Leu-L-MePhe-Ant-L-Ala-OH. To a stirred solution
of 0.195 g (0.29 mmol) of l-MePhe-L-Leu-L-MePhe-Ant-L-Ala-OMe
in MeOH (4 mL), 0.34 mL of 1 N NaOH was added in one portion,
and this mixture was stirred overnight at ambient temperature under
nitrogen. HCl (1 N, 0.36 mL) was then added, and this mixture was
stirred for 5 min. The solvent was evaporated to near dryness at reduced
pressure and the residue diluted with DCM (50 mL) and water (5 mL).
The organic phase was separated and washed with brine (6 mL) and
then dried (Na2SO4). Concentration left 0.185 g (96%) of the title
compound as a white powder, which was used as isolated: 1H NMR
(CD3OD) δ 0.74 (d, J ) 6), 0.92 (m), 1.34-1.64 (m), 2.42 (s), 2.78
(m), 2.90-3.10 (m), 3.00 (s), 3.05 (s), 3.20 (m), 3.45 (m), 3.80 (m),
4.58 (m), 4.90 (m), 7.00 (d, J ) 8 Hz), 7.15-7.35 (m), 7.45 (m), 7.80
(d, J ) 8 Hz), 8.22 (d, J ) 8 Hz), 8.52 (d, J ) 8 Hz).
References and Notes
(1) De Jesus, A. E.; Steyn, P. S.; Van Heerden, F. R.; Vleggaar, R.;
Wessels, P. L.; Hull, W. E. J. Chem. Soc. Perkin Trans. 1 1983, 8,
1847-1856.
(2) Kobayashi, R.; Samejima, Y.; Nakajima, S.; Kawai, K.; Udagawa,
S. Chem. Pharm. Bull. 1987, 35, 1347-1352.
(3) Dalsgaard, P. W.; Larsen, T. O.; Christophersen, C. J. Antibiot. 2005,
58, 141-144.
(4) Marfey, P. Carlsberg Res. Commun. 1984, 49, 591-596.
(5) Dalsgaard, P. W.; Larsen, T. O.; Frydenvang, K.; Christophersen,
C. J. Nat. Prod. 2004, 67, 878-881.
(6) Kotsuki, H.; Datta, P. K.; Hayakawa, H.; Suenaga, H. Synthesis 1995,
11, 1348-1350.
(7) Nakao, K.; Hamada, Y.; Shioiri, T. Chem. Pharm Bull. 1989, 37,
930-932.
(8) Lewer, P.; Graupner, P. R.; Gilbert, J. R.; Karr, L. L.; Chapin, E.;
Duebelbeis, D. O.; Martin, L.; Fields, S.; Pearce, C. Poster 248,
American Society of Pharmacognosy Meeting, Corvallis, OR, July
24-27, 2005.
(9) Frisvad, J. C.; Larsen, T. O.; Dalsgaard, P. W.; Seifert, K. A.; Louis-
Seize, G.; Lyhne, E. K.; Jarvis, B. B.; Fettinger, J. C.; Overy, D. P.
Int. J. Syst. EVol. Microbiol. 2006, 56, 1427-1437.
(10) Gargas, A.; DePriest, P. T. Mycologia 1996, 88, 745-748.
(11) Atlas, R. M. Handbook of Microbiological Media, 2nd ed.; CRS Press
Inc.: Boca Raton, FL, 1997.
NP060219C