A cell wall-active lipopeptide from the fungus Pochonia bulbillosa

Article abstract of DOI:10.1021/np800341u

Bioassay-directed fractionation of a fermentation of Pochonia bulbinosa, culture 38G272, led to the isolation of a series of structurally novel, prospective cell wall-active lipopeptides. The main component of this suite is 1, a linear hexapeptide with a δ-hydroxymyristic acid amide substituted N-terminus. The structure was deduced using high-field microsample NMR, Fourier transform mass spectrometry, and microscale chemical degradation. The potent cell wall activity and synthetically accessible structure of 1 make it a potential lead for further investigation.

Full text of DOI:10.1021/np800341u

J. Nat. Prod. 2008, 71, 2045–2048
2045
A Cell Wall-Active Lipopeptide from the Fungus Pochonia bulbillosa
Frank E. Koehn,*,† Donald R. Kirsch,‡ Xidong Feng,† Jeffrey Janso,† and Mairead Young†
Natural Products DiscoVery and DiscoVery Analytical Chemistry, Chemical and Screening Sciences, Wyeth Research, Pearl RiVer,
New York 10965, and Cambria Biosciences, Woburn, Massachusetts 01801
ReceiVed June 9, 2008
Bioassay-directed fractionation of a fermentation of Pochonia bulbinosa, culture 38G272, led to the isolation of a series
of structurally novel, prospective cell wall-active lipopeptides. The main component of this suite is 1, a linear hexapeptide
with a δ-hydroxymyristic acid amide substituted N-terminus. The structure was deduced using high-field microsample
NMR, Fourier transform mass spectrometry, and microscale chemical degradation. The potent cell wall activity and
synthetically accessible structure of 1 make it a potential lead for further investigation.
Fungal pathogens are eukaryotes and show significant biochemi-
cal similarity to their hosts. Therefore, a major prerequisite for the
discovery of nontoxic antifungal therapeutics is the identification
of novel fungal-selective biochemical targets and compounds that
selectively inhibit these targets. Fungal cell wall biosynthesis has
been an area of long-term interest, since the production of a cell
wall is essential for fungal viability and human cells lack both a
cell wall and the enzymes that synthesize the fungal cell wall
polysaccharide components.1 Our efforts to identify novel antifungal
cell wall-acting agents have in part focused upon the use of an
assay designed to identify general inhibitors of cell wall biosyn-
thesis.2 This assay exploits a Neurospora crassa os-1 mutation,
which allows cell growth in the absence of a cell wall.3 In the
presence of concentrations of cell wall-acting antifungal agents that
inhibit the growth of wild-type strains, strains carrying an os-1
mutation will grow as wall-less protoplast cells. Most compounds
acting on non-cell wall targets have nearly identical inhibitory
actions on os-1 and wild-type strains, and thus differential growth
with the production of protoplasts can be used to identify
compounds with this desired mechanism of action. We have
reported the identification of a novel cyclic lipopeptide with this
screen: 15G256γ.4 This compound showed antifungal activity and
induced the production of protoplasts in a number of fungal species,
yet inhibited neither chitin synthase nor ꢀ-1,3-glucan synthetase,
and may thus be acting through a cell wall biosynthesis regulatory
mechanism.5 The interesting mechanism and challenging structure
of this molecule prompted us to pursue additional screening in order
to identify synthetically accessible molecules that could serve as
the basis for medicinal chemistry optimization.
weight from 700 amu to over 1200 amu. Repeated reversed-phase
HPLC furnished 0.7 mg of the major component 1 (Chart 1), which
gave a 20 mm zone of protoplast production in the os-1 assay at
600 ng per disk. Since 1 was the most abundant of the active
components, we focused our attention on its structure elucidation.
The low-resolution positive ion ESMS of 1 gave a nominal mass
of m/z 821 for the M + H, and high-resolution positive ion
nanoelectrospray FTMS showed a sodium adduct of the molecular
ion at m/z 843.48364 [M + Na]+ in positive mode and a
deprotonated molecular ion at m/z 819.48767 [M - H]- in negative
mode, corresponding to a molecular formula of C41H68N6O11 (∆
mmu ) -0.19, 0.34, respectively). The 1H NMR spectrum in
DMSO-d6, first at 300 MHz and then at 500 MHz, displayed a
number of distinctive features. Among these were resonances for
six NH amide protons (determined by methanol-d4 exchange), four
of which were contained in a strongly overlapped multiplet at δ
7.85, a phenolic OH at δ 9.13, six methyl doublets clustered in
groups of three at δ 0.9 and 1.1, an overlapping muliplet of
R-protons centered at δ 4.1, and a large envelope of CH2 protons
at δ 1.2. These features, taken together with the molecular formula,
suggested the molecule was a hexapeptide consisting of single
residues of threonine, valine, and tyrosine, three alanine residues,
and an oxygenated aliphatic C-14 acid or amide unit. The strongly
overlapping alanine signals and the molecular formula taken
together with the general lack of 1H NMR spectral dispersion
suggested a linear molecule.
The 500 MHz 2D TOCSY and multiplicity-edited HMQC data
confirmed the presence and allowed assignment of the constituent
amino acid residues as listed in Table 1. Readily assigned via 2D
TOCSY experiments were the 1H spin systems for the threonine at
δ 7.87 (NH), 4.18 (HR), 3.74 (Hꢀ), and 1.07 (Hγ), the tyrosine at
δ 7.99 (NH), 4.61 (HR), and 2.89, 2.61 (Hꢀ), the aromatic protons
at δ 7.03 and 6.61, and the valine residue at δ 8.18 (NH), 4.15
(HR), 2.03 (Hꢀ), and 0.829, 0.83 (Hγ). The three individual alanine
spin systems showed considerable overlap in the 1D and 2D
homonuclear spectra, with resonances assignable for alanine-1 at
δ 8.26 (NH), 4.17 (HR), and 1.21 (Hꢀ), alanine-2 at δ 7.87 (NH),
4.22 (HR), and 1.14 (Hꢀ), and alanine-3 at δ 7.81 (NH), 4.24 (HR),
and 0.98 (Hꢀ). The 2D TOCSY spectrum contained additional
cross-peaks showing coupling of a proton at δ 3.31 to a methylene
at δ 1.59, 1.45 and a methylene at δ 2.11; the latter was assumed
to be adjacent to the amide carbonyl by virtue of proton chemical
shift. The δ 3.31 was also coupled into the methylene envelope at
δ 1.28. These data indicated the presence of a hydroxy group located
somewhere along the C-14 chain. 13C NMR chemical shifts were
assigned by means of a multiplicity-edited HMQC experiment and
are listed in Table 1.
Fungi of the genus Pochonia sp. are producers of resorcylic acid
lactone metabolites such as pochonins, monocillins, radicicol, and
various analogues, and they also produce the unique heterospiro-
cycle pseurotin A.6,7 These compounds show antiviral and anti-
parasitic activity and HSP90 inhibition.8 The broad biosynthetic
capabilities of this genus along with our interest in simple antifungal
natural products that might serve as medicinal chemistry leads
prompted us to examine in detail a culture of Pochonia bulbillosa.
The n-butanol broth extract of a strongly os-1-active 1 L
fermentation of culture 38G272 was chromatographed on a column
of LH-20 in MeOH, and the subsequent active fractions were
subjected to repeated semipreparative reversed-phase HPLC in
MeOH/H2O/TFA and then MeCN/H2O/TFA mixtures. MS and 1H
NMR examination of active fractions showed that the os-1 activity
was due to a complex suite of lipopeptides ranging in molecular
* To whom correspondence should be addressed. Tel: 845-602-4094.
Fax: 845-602-4094. E-mail: koehnf@wyeth.com.
† Wyeth Research.
We turned to g-HMBC and ROESY spectra to provide sequence
information. The carbonyl region of the g-HMBC spectrum
‡ Cambria Biosciences.
10.1021/np800341u CCC: $40.75
2008 American Chemical Society and American Society of Pharmacognosy
Published on Web 11/18/2008

2046 Journal of Natural Products, 2008, Vol. 71, No. 12
Notes
Chart 1
Table 1. 1H (500 MHz) and 13C (75 MHz) NMR Data for Compound 1 in DMSO-d6 at 30 °Ca
residue
13C (75 MHz)
1H (500 MHz, mult, J in Hz)
HMBC correlation
Thr
CdO
NH
R
170.6
7.87 (d, 1H, J ) 8.7)
4.18
3.74 (q, 1H, J ) 7.4)
1.07 (d, 3H, J ) 6.7)
172.1
170.6, 67.2
170.6
59.19
67.43
20.28
ꢀ
γ
67.2, 58.2
Ala 1
CdO
NH
R
171.6
8.26 (d, 1H, J ) 7.3)
4.17 (m, 1H)
1.21 (d, 3H, J ) 7.2)
170.6
170.6
171.6, 48.5
57.0
17.52
ꢀ
Ala 2
CdO
NH
R
171.2
7.87 (d, 1H, J ) 7.2)
4.22 (m, 1H)
1.14 (d, 3H, J ) 6.7)
48.47*
17.77
ꢀ
171.2, 48.0
Ala 3
CdO
NH
R
171.2
7.81 (d, 1H, J ) 7.4)
4.24 (m, 1H)
0.98 (d, 3H, J ) 6.7)
171.3
171.6
171.6, 48.0
47.90*
18.30
ꢀ
Tyr
CdO
NH
R
171.31
7.99 (d, 1H, J ) 8.9)
4.61
171.7
171.2, 38.2
171.2, 127.6, 130.1
171.2, 127.6,130.1
38.15, 114.6, 130.1
38.15, 114.6, 130.1
53.81
38.15
ꢀ
2.89 (dd 1H, J )15.0, 4.0)
2.61 (dd, 1H, J ) 15.0, 11.0)
7.03, (dt, 2H, J ) 8.4)
6.61 (dt, 2H, J ) 8.4)
9.13 (OH)
other
Val
CdO
NH
R
172.8
8.18 (br, 1H)
4.15 (br, 1H))
2.03 (m, 1H)
0.83 (d, 3H, J ) 6.7)
0.83 (d, 3H, J ) 6.7)
171.2
57.0
30.0
19.0
19.0
ꢀ
172.2, 56.99
30.04, 56.99
γ
HMA
CdO
172.2
35.10
21.6
36.5
69.2
R
2.11 (m, 2H)
1.60 (m, 2H)
1.24 (m, 2H)
3.31 (m, 1H)
1.27 (m, 2H)
1.24 (m, 2H)
1.24 (br envelope)
0.86 (t, 3H)
ꢀ
γ
δ
e
37.2
φ
25.2
29.8-22.7
13.9
C8-C13
C14
a HMA: δ-hydroxymyrstic acid amide. *Assignment may be interchanged due to overlap.
contained limited information for linking the residues, with the
carbonyl signals for all three alanine moieties and the tyrosine all
located within 0.5 ppm of δ 171.3. Correlation peaks from HR and
Hꢀ in the C14 unit at δ 2.11/172.3 and δ 1.59,1.45/172.3,
respectively, allowed assignment of the N-terminal amide carbonyl
chemical shift. The remaining correlations in the carbonyl region
of the HMBC spectrum were intraresidue NH/CdO for tyrosine at
δ 7.99/171.7 and alanine-3 at δ 7.81/171.3. An additional signal at

Notes
Journal of Natural Products, 2008, Vol. 71, No. 12 2047
Under consideration of the work of Anderson et al., it was
hypothesized that given the peptide sequence, differential rates of
hydrolysis under mild acid hydrolysis conditions should liberate
the tripeptide Ala-Ala-Ala as an early product in the hydrolysis.10
Compound 1 was subjected to mild acid hydrolysis in 2:1 MeCN/1
N HCl ·H2O, and the reaction mixture was monitored by LCMS.
After 3 h the mixture showed the presence of the diagnostic D-Ala-
L-Ala-L-Ala tripeptide with [M + H]+ m/z ) 232, by comparison
with an authentic standard. Fragment partial hydrolysis products
were also observed for [M + H]+ m/z ) 595, 496, 352, and 333
for TAAAYV,TAAAY, AYV, and TAAA, respectively. These
results complete the sequence assignment and structure elucidation.
Recent efforts to develop novel, clinically useful antifungal agents
have focused on compounds of the echinocandin class.11 These
lipopeptide compounds prevent fungal cell wall glucan synthesis
by inhibiting the enzyme ꢀ-1,3-D-glucan synthase, a highly selective
target since mammals do not synthesize ꢀ-1,3-D-glucans, providing
such compounds with high intrinsic potential for selectivity. The
clinically employed echinocandins are based upon natural product
leads optimized via semisynthetic chemical approaches. As an
example, caspofungin, the first clinically approved agent in this
class, is an aza-substituted bisamine derivative of the antibiotic
pneumocandin Bo.12 While several echinocandins have clinically
useful antifungal potency and spectrum (e.g., caspofungin, mico-
fungin, and anidulafungin), their poor absorption after oral admin-
istration limits their dosing to the intravenous route.13,14 The
unattractive pharmacological properties of these compounds coupled
with significant limitations for synthetic modification have stimu-
lated the search for alternate structural types that will inhibit cell
wall synthesis.1 The putative cell wall synthesis inhibitory activity
of 1 coupled with its potential for synthetic elaboration could make
this compound a valuable anti-infective lead.
Figure 1. Negative mode nanoelectrospray FTMS mass spectrum
of 1. (a) Multi-CHEF isolation of the [M - H]- precursor ion. (b)
Multi-CHEF SORI-CID mass spectrum of the molecular ion [M
- H]- with inset showing the sequence coverage. Internal reference
peaks are labeled with stars [*] at the predicted values of m/z
431.98233, 601.97898, and 1033.98812 from Agilent ES tuning
mix. HMA: hydroxymyristic acid amide substituted N-terminus.
δ 3.31/69.2 in the multiplicity-edited HSQC was assigned to an
OH-bearing carbon located along the chain of the N-terminal C-14
amide. The remaining peaks in the HMBC spectrum confirmed the
intraresidue assignments made by TOCSY and HSQC (Table 1).
The 2D ROESY spectrum contained additional information to allow
partial sequence assignment. The presence of the C-14 amide on
the threonine nitrogen could be ascertained by a ROESY cross-
peak from the threonine NH at δ 7.87 to the HR methylene at δ
2.11. Furthermore, the tyrosine NH at δ 7.99 showed a ROESY
cross-peak to one alanine HR at δ 4.23, and the tyrosine aromatic
protons showed ROESY cross-peaks to the Hꢀ methyl group of
Ala-1 as well, suggesting an Ala-Tyr partial structure.
Experimental Section
General Experimental Procedures. HPLC analysis was performed
on an HPLC system from Agilent Technologies (Wilmington, DE)
consisting of an Agilent 1100 pump with an online degasser, a diode-
array detector (which was set at a 215 nm wavelength), and an
autosampler. For instrument control, data acquisition, and processing
an Agilent Chemstation data system was applied (Agilent Chemstation
B.1.03).
The position of the OH group in the C-14 chain was assigned to
the γ-position by using 2D TOCSY and HSQC-TOCSY experi-
ments at 20, 55, and 70 ms mixing times to establish the proton
and 13C NMR chemical shifts of methylene units in stepwise fashion
out from the OH group. At a mixing time of 20 ms, correlations
were observed from protons at δ 3.31 to 1.27 to 13C NMR signals
at δ 37.2 and 36.5, indicating the directly adjacent vicinal protons
and associated carbons. Increasing the mixing time to 55 ms and
then to 70 ms showed additional stepwise increasing correlations
to δ 1.59/21.6 and then to δ 2.11/35.1, indicating these methylene
units were located at successively removed positions. Previous
assignment of the δ 2.11 signal to the R-position by g-HMBC
therefore placed the OH at the δ-position. Comparison of observed
and calculated 13C NMR chemical shifts agreed closely with the δ
OH assignment.
NMR experiments were performed on a Bruker DRX 500 equipped
with a 5 mm TXI Cryoprobe, or a Bruker AMX 300 spectrometer.
The HRMS spectra of lipopeptide 38G272-820 were measured using
a Bruker Daltonics APEXII FTICR mass spectrometer equipped with
an actively shielded 9.4 T magnet, an external Bruker APOLLO ESI
source, and a Synrad 50W CO2 CW laser. Typically, 5 µL samples
were loaded into the nanoelectrospray tip, and a high voltage, about
800 V, was applied between the nanoelectrospray tip and the capillary.
Data reported here are based on internal calibration using a HP tuning
mix. The ESI-FTMS/MS experiments were conducted on either the
deprotonated molecular ion at m/z 819 for negative mode or the sodium
adduct molecular ion at m/z 843 for positive mode, using sustained
off-resonance irradiation with collision-induced dissociation (SORI-
CID) or an infrared multi-photon dissociation (IRMPD). Data reported
here are from SORI-CID experiments where the precursor ion of m/z
819 was isolated together with three other HP tuning mix ions (predicted
values of m/z 431.98233, 601.97898, and 1033.98812) using a multi-
CHEF (correlated harmonic excitation fields) sweep (Figure 1). The
parent ions were activated by an rf pulse, off resonance from the parent
ions by 1000 Hz. Argon gas was pulsed into the ICR cell to collide
with the activated parent ions to yield fragment ions. The SORI-CID
mass spectrum was internally calibrated using HP tuning mix peaks
specified above.
Since NMR methods did not yield sufficient information for the
complete sequence assignment, we resorted to FTMS.9 The negative
mode nanoelectrospray multi-CHEF SORI-CID mass spectra of 1
showed all y sequence ions necessary to complete the sequence
assignment (Figure 1, Table 2). The positive mode IRMPD mass
spectrum of the [M + Na]+ molecular ion provided equally
thorough sequence information (Supporting Information).
Acid hydrolysis followed by derivatization to produce the
n-propyl pentafluoropropionyl esters and chiral GC analysis showed
that the peptide portion of the molecule consisted of one residue
each of D-Tyr, L-Val, L-Thr, and D-Ala and two residues of L-Ala.
Given three alanine units, the question remained as to the sequence-
specific configuration assignment. The linear nature of the molecule
and lack of dispersion of the NMR spectrum, particularly the alanine
CHR and methyl resonances, precluded reliable assignment by
NMR. Therefore we resorted to microscale chemical degradation
methods with the goal of generating alanine-bearing peptides that
could be chromatographically compared with authentic standards.
Fungal Material. P. bulbillosa 38G272 was isolated from unknown
material collected from oak forests in San Jose Province, Costa Rica,
in 1992. The nuclear internal transcribed spacer region (ITS) of 38G272
(Cyan308) was searched against the GenBank database via BLASTN
2.2.115 and found to be 100% similar to P. bulbillosa16 (CBS 145.70).
In a study comparing ITS sequences of Verticillium sect. Prostrata,
Zare et al.17 found that Verticillium bulbillosum formed a clade with

2048 Journal of Natural Products, 2008, Vol. 71, No. 12
Notes
a
Table 2. Proposed Analysis of FTMS/MS of m/z 819.48767 [(HMA)TAAAYV(OH) - H]-
exptl mass
pred mass
∆ (mmu)b
rel abun
elemental formula
structure assignments
ion assignments
-
819.48767
593.29466
492.24649
421.20918
350.17234
279.13500
116.07164
819.48733
593.29405
492.24637
421.20926
350.17214
279.13503
116.07170
0.34
0.61
0.12
-0.08
0.20
-0.03
-0.06
87%
3%
22%
8%
2%
2%
C41H67N6O11
C27H41N6O9
C23H34N5O7
C20H29N4O6
[M - H]-
precursor
-
[TAAAYV(OH)]-
[AAAYV(OH)]-
[AAYV(OH)-
[AYV(OH)]-
[YV(OH)]-
y6
y5
y4
y3
y2
y1
-
-
-
C17H24N3O5
-
C14H19N2O4
-
6%
C5H10NO2
[V(OH)]-
a HMA: ꢁ-hydroxymyristic acid amide substituted N-terminus. b Errors (mmu) calculated as experimental value - predicted value.
other Verticillium spp., which parasitize nematode eggs or cysts and
form dictyochlamydospores. V. sect. Prostrata was revised at the generic
level,18 resulting in the renaming of V. bulbillosum to Pochonia
bulbillosa.18 The ITS sequence for fungus 38G272 was deposited in
GenBank under accession number EU999952.
Culture 38G272 was grown on cornmeal agar (Oxoid) at room
temperature (about 24 °C) for morphological analyses. The agar-grown
culture was examined microscopically on the seventh and 24th days
of incubation. By the seventh day of growth, lunate or crescent-shaped
conidia were visible on the aerial mycelia in slimy heads. Conidia were
produced by solitary phialides or 2-3 phialides per node. By the 24th
day of growth, thick-walled dictychlamydospores were present. The
morphology of 38G272 concurs with the ITS identification of 38G272
as P. bulbillosa. Fermentation of 38G272 for chemical studies was
carried out in 1 L shake flasks in modified Sabouraud dextrose broth,
substituting maltose for dextrose.
TIC detection. After 24 h no significant products were observed and
the reaction mixture was heated to 90 °C. LCMS using absorbance at
215 nm single-ion monitoring at m/z ) 232 showed the presence of
D-Ala-L-Ala-L-Ala (tR ) 2.44 min). Additional major products observed
during the hydrolysis were [M + H]+ m/z ) 595, 496, 352, and 333
for TAAAYV, TAAAY, AYV, and TAAA, respectively. Minor
amounts of presumably racemized trialanine were also observed.
Antifungal Assays. Assays were performed as described in ref 2.
Briefly, os-1 strain spores were inoculated into growth media and test
samples applied to the agar surface. The assay plates were scored after
36-48 h of incubation at 37 °C. Active samples were identified by the
appearance of characteristic growth zones surrounding the sample.
Activity was confirmed by microscopic examination of the activity
zones to determine whether mycelial growth (inactive) or protoplasts
(active) were present.
Acknowledgment. The authors gratefully acknowledge V. Bernan
for microbiological contributions, S. Silverman and D. Alabaugh for
biological assays, and M. Siegel and K. Tabei for mass spectral
measurements.
Extraction and Isolation of 1. The whole fermentation broth from
a 1 L fermentation of culture 38G272 was centrifuged, the pellet
removed, and the supernatant extracted twice with equal volumes of
n-BuOH. A small portion of the crude BuOH-soluble material was
chromatographed on a 15 cm Rainin microsorb C18 HPLC column with
60% 0.1 N TFA/MeCN, and fractions were assayed for os-1 activity,
which appeared to elute in a broad peak spanning 4-15 min. On the
basis of this result the bulk of the crude BuOH-soluble material was
loaded onto a 25 mm × 60 cm column of Sephadex LH-20 and eluted
with MeOH in 10 mL fractions. The main peak of os-1 activity eluted
in fractions 10 through 18, which were combined (36 mg) and
subsequently chromatographed on a 4.6 mm × 15 cm Rainin C18
microsorb column with 67% MeOH/0.1 N TFA. Compounds that failed
to elute were washed from the column with MeOH to give 5.9 mg of
material, which by HPLC showed the presence of 1 as a major
component. This material was dissolved in MeOH and chromatographed
on a 1.0 cm × 25 cm Rainin microsorb C18 column in 71% MeOH/0.1
N TFA. Fractions were collected on the basis of absorbance at 225
nm, giving 0.7 mg of 1 along with smaller amounts of related
compounds.
Determination of Amino Acid Configuration. Approximately 10
µg of 1 was dissolved in 6 N constant boiling HCl and heated at 110
°C for 22 h. The solution was diluted with H2O and lyophilized. The
residue was dissolved in HCl-1-propanol (prepared by additional of
acetyl chloride, 10%, to 1-propanol) and heated at 90 °C for 1 h and
then evaporated under an N2 stream. The residue was taken up in 10%
pentafluoropropionic anhydride in CH2Cl2 and heated at 50-60 °C for
4 h. The solvent was removed under an N2 stream and the residue
redissolved in CH2Cl2 and analyzed by GC-MS with selective ion
monitoring on a Chirasil capillary column. Comparison and co-injection
with n-propyl pentafluoropropionyl derivatives of authentic L and D
amino acid standards showed the hydrolysis product to be composed
of D-Tyr, L-Val, L-Thr, D-Ala, and L-Ala. The D,L-Ala components were
present in a 1:2 ratio, respectively.
Note Added after ASAP Publication: The structure of compound
1 was corrected.
Supporting Information Available: This material is available free
of charge via the Internet at http://pubs.acs.org.
References and Notes
(1) Selitrennikoff, C. P.; Nakata, M. Curr. Opin. InVest. Drugs 2003, 4
(2), 200–205.
(2) Kirsch, D. R.; Lai, M. H. J. Antibiot. 1986, 39 (11), 1620–2.
(3) Schumacher, M. M.; Enderlin, C. S.; Selitrennikoff, C. P. Curr. Opin.
Microbiol. 1997, 34 (6), 340–347.
(4) Schlingmann, G.; Lisa, M.; Williams, D. R.; Carter, G. T. J. Antibiot.
1998, 51 (3), 303–316.
(5) Albaugh, D.; Albert, G.; Bradford, P.; Cotter, V.; Froyd, J.; Gaughran,
J.; Kirsch, D. R.; Lai, M.; Rehnig, A.; Sieverding, E.; Silverman, S.
J. Antibiot. 1998, 51 (3), 317–322.
(6) Hellwig, V.; Mayer-Bartschmid, A.; Mueller, H.; Greif, G.; Kleymann,
G.; Zitzmann, W.; Tichy, H.-V.; Stadler, M. J. Nat. Prod. 2003, 66
(6), 829–837.
(7) Bloch, P.; Tamm, C. HeIV. Chim. Acta 1981, 64 (1), 304–15.
(8) Moulin, E.; Zoete, V.; Barluenga, S.; Karplus, M.; Winssinger, N.
J. Am. Chem. Soc. 2005, 127 (19), 6999–7004.
(9) Feng, X. S., M. M. Anal. Bioanal.Chem. 2007, 389 (5), 1341–1363.
(10) Barsby, T.; Warabi, K.; Sorensen, D.; Zimmerman, W. T.; Kelly, M. T.;
Andersen, R. J. J. Org. Chem. 2006, 71 (16), 6031–6037.
(11) Denning, D. W. Lancet 2003, 362 (9390), 1142–1151.
(12) Bouffard, F. A.; Zambias, R. A.; Dropinski, J. F.; Balkovec, J. M.;
Hammond, M. L.; Abruzzo, G. K.; Bartizal, K. F.; Marrinan, J. A.;
Kurtz, M. B. J. Med. Chem. 1994, 37 (2), 222–5.
(13) Wagner, C.; Graninger, W.; Presterl, E.; Joukhadar, C. Pharmacology
2006, 78 (4), 161–77.
Partial Hydrolysis of 1: Determination of Sequence Position of
D- and L-Alanine. A UV-HPLC method to separate the three forms of
the alanine tripeptides DLL, LDL, and LLD was developed using the
following chromatographic conditions: a YMC PackPro AM (5 µm,
150 mm × 4.6 mm i.d.) column from Waters (Milford, MA); mobile
phase composed of (A) 0.02% TFA in H2O and (B) 0.02% TFA in
MeCN. Gradient elution was performed as follows: 0-6 min 5% B,
6-10 min 5-95% B, and 10-15 min 95% B at a flow rate of 1 mL/
min. The retention of authentic alanine tripeptide standards (Bachem)
were as follows: D-Ala-L-Ala-L-Ala at 2.5 min, L-Ala-L-Ala-D-Ala at
3.1 min, and L-Ala-D-Ala-L-Ala at 4.2 min. For analysis, a sample of
1 (40 µg) was dissolved in 1.0 mL of MeCN/1 N HCl H2O (3:2) in a
sealed vial at 30 °C and monitored by LCMS with single-ion and full
(14) Onishi, J.; Meinz, M.; Thompson, J.; Curotto, J.; Dreikorn, S.;
Rosenbach, M.; Douglas, C.; Abruzzo, G.; Flattery, A.; Kong, L.;
Cabello, A.; Vicente, F.; Pelaez, F.; Diez, M. T.; Martin, I.; Bills, G.;
Giacobbe, R.; Dombrowski, A.; Schwartz, R.; Morris, S.; Harris, G.;
Tsipouras, A.; Wilson, K.; Kurtz, M. B. Antimicrob. Agents Chemoth-
er. 2000, 44 (2), 368–377.
(15) Altschul, S. F.; Madden, T. L.; Schaffer, A. A.; Zhang, J.; Zhang, Z.;
Miller, W.; Lipman, D. J. Nucleic Acids Res. 1997, 25 (17), 3389–
402.
(16) Gams, W., Z. R. NoVa Hedwigia 2001, 72, 329–337.
(17) Zare, R., G. W.; Culham, A. NoVa Hedwigia 2000, 71, 465–480.
(18) Zare, R., G. W.; Evans, H. C. NoVa Hedwigia 2001, 73, 51–86.
NP800341U