Verrucine F, a quinazoline from Penicillium verrucosum

Article abstract of DOI:10.1021/np800105e

Verrucine F (3), a quinazoline similar to verruchtes A (1) and B (2), contains one anthranilic acid residue, one L-glutamine residue, and one α,β-unsaturated phenylalanine residue, as determined by NMR, MS, and chemical methods. Compounds 1 and 3, but not 2, were produced by Penicillium verrucosum J255 and eight additional P. verrucosum strains. Verrucines were typically more concentrated in the inner (older) parts of the colony, and 3 peaked 4-8 days after 1, but at 10-fold lower concentrations (～200 μg/g). Verrucine F (3) formed spontaneously from 1 in buffered water solutions, probably by oxidation at C-1 followed by water elimination.

Full text of DOI:10.1021/np800105e

J. Nat. Prod. 2008, 71, 1455–1457
1455
Verrucine F, a Quinazoline from Penicillium Werrucosum
Su-lin Leong,† Johan Schnu¨rer,† and Anders Broberg*,‡
Department of Microbiology, Swedish UniVersity of Agricultural Sciences, P.O. Box 7025, SE-750 07 Uppsala, Sweden, and Department of
Chemistry, Swedish UniVersity of Agricultural Sciences, P.O. Box 7015, SE-750 07 Uppsala, Sweden
ReceiVed February 15, 2008
Verrucine F (3), a quinazoline similar to verrucines A (1) and B (2), contains one anthranilic acid residue, one L-glutamine
residue, and one R,ꢀ-unsaturated phenylalanine residue, as determined by NMR, MS, and chemical methods. Compounds
1 and 3, but not 2, were produced by Penicillium Verrucosum J255 and eight additional P. Verrucosum strains. Verrucines
were typically more concentrated in the inner (older) parts of the colony, and 3 peaked 4-8 days after 1, but at 10-fold
lower concentrations (∼200 µg/g). Verrucine F (3) formed spontaneously from 1 in buffered water solutions, probably
by oxidation at C-1 followed by water elimination.
Mold spoilage of stored cereals is of major concern throughout
the world. Growth of spoilage fungi may be detrimental to the smell,
taste, or appearance of grains; more seriously, certain molds produce
mycotoxins, rendering the grains unsuitable for use as food or feed.
Penicillium Verrucosum (Dierckx), of the family Trichocomaceae,
is a common mold on stored cereals in temperate regions. This
species produces several mycotoxins including ochratoxin A,1 which
has recently gained significant attention as a problem in food and
feed. The European community has issued directives regarding the
maximum concentrations of ochratoxin A allowed in cereals, wine,
and coffee.2,3 At present, we are investigating some aspects of the
biosynthesis of ochratoxins A and B in P. Verrucosum, including
the search for key intermediates. In the course of this work, we
Figure 1. Relevant HMBC (single-headed arrows) and NOESY
(double-headed arrow) correlations for verrucine F (3).
have also come across a number of other secondary metabolites,
including verrucines A (1) and B (2).4 Larsen and co-workers, when
first describing 1 and 2, also indicated the existence of a series of
oxygenated verrucine analogues, verrucines C, D, and E.5 In the
present study, we describe the isolation and characterization of a
new related substance, with the proposed name verrucine F (3), as
well as its occurrence in different strains of P. Verrucosum and its
formation from verrucine A.
Mycelia from 50 agar cultures of P. Verrucosum were extracted
with aqueous MeOH and were subsequently fractionated by SPE
and preparative HPLC, which led to the isolation of 1 and 3.
Compound 1 was found to be partly isomerized to 2 when the
solvent (acidic aqueous MeCN) was evaporated under reduced
pressure. This epimerization has been observed in previous studies
of 1 and the related compound anacine (4).4,6 Verrucines A (1)
and B (2) were identified by NMR and MS data, aided by
comparison with literature data.4
The presence of a glutamine side chain was indicated by MSMS
data, which showed the sequential loss of NH3 and CO from 3,
4
1
just as was observed for 1 and 2. The major difference in the H
NMR data between 3 and both 1 and 2 was that the signals from
the phenylalanine-derived methine and methylene groups of 1 and
2 were absent in 3. Instead, a new olefinic proton singlet was present
at δ 7.30 (δC 116.6), along with the signals from the phenyl group
originating from phenylalanine. This indicated that 3 contained a
double bond in the phenylalanine-derived part of the molecule. This
was corroborated by cross-peaks in HMBC experiments, as
indicated in Figure 1. The presence of an extra double bond in 3
was also supported by MS data, showing a [M + H]+ ion at m/z
375 as compared to a [M + H]+ ion at m/z 377 for both 1 and 2.
A similar unsaturated anacine analogue (5) was recently isolated
from P. aurantiogriseum.7 Analysis of 3 by HRFABMS yielded a
[M + H]+ ion at m/z 375.1489, in accordance with the molecular
formula C21H18N4O3. NOESY experiments showed that the con-
figuration of the double bond was Z, by a cross-peak between the
phenyl ring CH signal at δ 7.67 and the amide proton signal at δ
10.59 (Figure 1). The corresponding E-isomer was not detected in
any sample. The absolute configuration of the glutamine residue
was determined by acidic hydrolysis followed by esterification with
(S)-2-butanol, amide formation with perfluoropropanoic anhydride,
and GC-MS analysis. Comparison with reference samples showed
a 10:1 ratio between the derivatives of L- and D-glutamic acid in
the sample, where the D-glutamic acid derivative was assumed to
be formed by racemization during sample derivatization, as
previously observed.8 The structural difference between 1/2 and 3
was also reflected by the corresponding UV spectra. The UV
spectrum of 3 in EtOH displayed peaks at 209, 230, 257, and 337
nm. Compounds 1 and 2 displayed UV peaks at similar wave-
lengths, but with overall lower log ε values.4 The UV result can
be explained by the extended conjugated double-bond system in 3,
Verrucine F (3) yielded 1H NMR data similar to 1 and 2,
suggesting the presence of anthranilic acid and glutamine moieties.
* Corresponding author. Phone: +46 18 672217. Fax: +46 18 673476.
E-mail: Anders.Broberg@kemi.slu.se.
† Department of Microbiology.
‡ Department of Chemistry.
10.1021/np800105e CCC: $40.75
2008 American Chemical Society and American Society of Pharmacognosy
Published on Web 06/26/2008

1456 Journal of Natural Products, 2008, Vol. 71, No. 8
Notes
To discriminate between these two modes of formation, verrucine
A (1) was incubated with whole cell extracts of P. Verrucosum, at
pH 7 and 9, in the presence of NAD+, as well as without cell
extracts (pH between 4 and 9). LC-MS analyses showed that a
fraction of 1 was transformed to 3 in the absence of cell extracts
and that the addition of cell extracts did not enhance the formation.
The formation of 3 was the highest at pH 4 and 9, which is
consistent with acid- and base-catalyzed, respectively, water
elimination from a C-1 hydroxylated verrucine A (Scheme 1);
however, the conversion was also observed at neutral pH. Moreover,
in all P. Verrucosum samples containing 3, as well as in all samples
from incubation of 1, a possible C-1 hydroxylated 1 was detected
by LC-MS. This substance displayed a [M + H]+ ion at m/z 393,
from which water was lost to produce a daughter ion at m/z 375.
These results indicate that the formation of 3 is not catalyzed by
enzymes.
Figure 2. Verrucines A (1) and F (3) in inner and outer parts of
colonies of Penicillium Verrucosum J255 grown on yeast extract
sucrose agar at 25 °C.
Experimental Section
General Experimental Procedures. H and 13C NMR data were
Scheme 1
1
acquired on a Bruker DRX600 NMR spectrometer equipped with a
2.5 mm SEI microprobe (1H/13C). All NMR experiments were recorded
at 30 °C in DMSO-d6, and the chemical shifts were determined relative
to the residual signal of the solvent (δC 39.5; δH 2.50). For structure
1
elucidation, 1D H NMR, COSY, NOESY (300 ms), DEPT-HSQC,
and HMBC (65 ms) were applied, and the pulse sequences were used
as provided by the manufacturer. HRFABMS was performed on a four-
sector tandem mass spectrometer (Jeol SX/SX102A), equipped with a
FAB ion-source (Xe), using glycerol as matrix and PEG as internal
standard. LC-MS and LC-MSMS were performed on an Agilent 1100
HPLC system connected to a Bruker Esquire-LC ion-trap mass
spectrometer (electrospray ionization, ESI) operated in the positive ion
mode, and GC-MS on a HP5890/5970 GC-MS (Hewlett-Packard) using
He as carrier gas. Preparative HPLC was run on a Gilson system. The
mobile phase for LC-MS and HPLC comprised MeCN (HPLC isocratic
grade) and deionized filtered water. SPE was performed with 10 g
prepacked columns [Isolute C18 (EC), International Sorbent Technol-
ogy, Hengoed, UK].
Organisms and Cultivation. The following strains of Penicillium
Verrucosum were grown as single-point cultures on yeast extract sucrose
agar13 at 25 °C: J255 (Culture Collection, Dept of Microbiology,
Swedish University of Agricultural Sciences, Uppsala), OTA11 (Na-
tional Food Administration, Uppsala, Sweden), BFE489, BFE490,
BFE491, BFE492, BFE493, BFE495, and BFE505 (Max Rubner
Institute, Institute of Nutrition and Food, Institute for Safety and Quality
of Fruits and Vegetables, Karlsruhe, Germany).
Isolation of Verrucines A (1), B (2), and F (3). Mycelia from 50
cultures of P. Verrucosum J255 (14 days) were ground in liquid
nitrogen and subsequently extracted with 1 L of aqueous 40% MeOH
for 10 min in an ultrasonic bath, followed by 1 h with stirring at
room temperature. The extract was filtered through a textile cloth,
then sequentially through two different filter papers (3 and 00H,
respectively, Munktell, Sweden), and subsequently diluted with H2O
(3 L). The resulting solution (4 L) was loaded onto five 10 g SPE
columns. Each SPE column was washed with 50 mL of H2O and
eluted with 50 mL of aqueous 95% MeCN. The combined 95%
MeCN extract was dried under vacuum, redissolved in 4 mL of
aqueous 40% MeCN, and injected onto a preparative C-18 column
(20 × 100 mm with a 20 × 30 mm guard column, 5 µm, Dr. A.
Maisch High Performance LC GmbH, Germany). The column was
eluted with a gradient of MeCN in H2O (with 0.1% TFA), running
from 40% to 70% MeCN in 10 min, followed by 70% for 10 min,
at 10 mL/min. The eluate was monitored at 332 nm, and fractions
were collected in deep-well plates (2 mL). Selected fractions were
analyzed by ESI-MS, and the fractions containing putative verru-
cines, as guided by m/z values, were pooled and dried under reduced
pressure. The pooled fractions were fractionated further on the same
preparative HPLC column (40% aqueous MeCN with 0.1% TFA,
at 10 mL/min). The fractions containing 3 were dried under reduced
pressure, yielding 1 mg of pure substance. Verrucine A (1) was
isolated in parallel, but during evaporation of the solvent (40% to
70% MeCN gradient with 0.1% TFA) under reduced pressure, 1
was partly isomerized to 2. Pure 1 and 2 were obtained by
but not in 1 or 2. This difference was also apparent when analyzing
samples containing 1 and 3 using LC-MS with simultaneous UV
detection at 332 nm. Samples giving similar peak areas for 1 and
3 in the corresponding extracted ion chromatograms always yielded
much larger peaks for 3 in the UV chromatogram.
The presence of 1, 2, and 3 in a number of P. Verrucosum strains
was tested by extracting mycelia from agar cultures with MeOH
and analyzing the extracts by LC-MS. Compounds 1 and 3 were
detected in all cultures, whereas 2 was not detected in any culture.
For 1 the concentrations ranged from 0.3 to 4 mg/g agar plug, and
for 3 from 5 to 100 µg/g agar plug. The production of 1 and 3 in
P. Verrucosum J255, a strain that also produces ochratoxin A and
citrinin, was monitored by harvesting cultures at intervals over 30
days, followed by MeOH extraction and HPLC analysis. Verrucine
A (1) attained maximum concentrations of ∼1800 µg/g mycelium
in both the inner and outer parts of the colony after 14-18 days
(Figure 2). Initially, the concentration of 1 was approximately 2-fold
greater in the inner (older) than the outer part of the colony, although
after 18 days, concentrations in both parts were similar. Verrucine
F (3) attained maximum concentrations of ∼200 µg/g mycelium
after 24-26 days and was typically more concentrated in the inner
than the outer parts of the colony. The concentrations of both 1
and 3 decreased slightly after 26 days. This is consistent with a
steady enzymatic or spontaneous formation of 3 from 1 or from 2
after epimerization. Enzymatic activity introducing a double bond
in phenylalanine residues has previously been identified in other
members of the Penicillium subgenus Penicillium section Viridicata,
as part of the machinery forming the alkaloids cyclopenin and
cyclopenol.9 This enzymatic activity required the presence of NAD+
and was optimum at pH 9.1.10 Spontaneous formation of 3 could
occur through oxidation of 1 or 2 at C-1, forming a hydroxylated
analogue (6), a process that previously has been observed,5,11,12
followed by elimination of H2O to form 3 (Scheme 1). A similar
C-1 hydroxylated anacine analogue has also been reported.7 As
only the Z-isomer of 3 was observed, this could indicate involve-
ment of an enzymatic process. However, if formation of 3 is
spontaneous, via a C-1 oxidized intermediate, subsequent elimina-
tion of H2O would still be expected to lead mainly to the Z-isomer
due to steric interactions between the phenyl ring and the quinazo-
line ring system in the E-isomer, as well as during the water
elimination process.

Notes
Journal of Natural Products, 2008, Vol. 71, No. 8 1457
preparative HPLC on the same column using aqueous 40% MeCN
without TFA.
Agilent Technologies), eluted with a gradient of MeCN in 2% acetic
acid: held initially at 20% MeCN for 5 min, then from 20% to 70%
MeCN in 20 min, from 70% to 100% in 3 min, maintained at 100%
for 6 min, and re-equilibrated at 20% for 4 min, at 0.8 mL/min.
Verrucines A (1) and F (3) were quantified by UV detection as
previously described.
Verrucine F (3): white powder; UV λmax (EtOH) nm (log ε) 209
1
(4.52), 230 (4.35), 257 (4.03), 337 (4.27); H NMR (DMSO-d6, 600
MHz) δ 10.59 (1H, s, NH), 8.17 (1H, d, J ) 8 Hz, H-7), 7.88 (1H, t,
J ) 8 Hz, H-9), 7.77 (1H, d, J ) 8 Hz, H-10), 7.67 (2H, d, J ) 8 Hz,
Ph-2/6), 7.56 (1H, t, J ) 8 Hz, H-8), 7.46 (2H, t, J ) 8 Hz, Ph-3/5),
7.36 (1H, t, J ) 8 Hz, Ph-4), 7.30 (1H, s, CH-Ph), 7.25 (1H, s) and
6.70 (1H, s, CONH2), 5.25 (1H, t, J ) 7 Hz, H-4), 2.19 (2H, m, CH2-
CONH2), 2.13 (1H, m) and 2.07 (1H, m, CH2-CH2-CONH2); 13C NMR
(DMSO-d6, 150 MHz) δ 172.2 (CONH2), 165.7 (CO, C-3), 160.1 (CO,
C-6), 146.9 (C, C-10a), 146.6 (C, C-11a), 134.6 (CH, C-9), 133.4 (C,
Ph-1), 129.0 (CH, Ph-2/6), 128.1 (CH, Ph-3/5), 127.8 (CH, Ph-4), 126.8
(CH, C-8), 126.6 (CH, C-10), 126.6 (C, C-1), 125.9 (CH, C-7), 119.6
(C, C-6a), 116.6 (CH, CH-Ph), 54.6 (CH, C-4), 30.2 (CH2, CH2-
CONH2), 28.2 (CH2, CH2-CH2-CONH2); HRFABMS m/z 375.1489 [M
+ H]+ (calcd for C21H19N4O3 375.1452).
Incubation of 1 with Cell Extracts. Mycelia of P. Verrucosum J255
were ground using a magnetic mill cooled with liquid nitrogen (Spex
6700). Ground material (1.0 g) was thawed and suspended in Tris-
HCl buffer (2.0 mL, 50 mM, pH 7.0 or 9.0), and the resulting cell
suspension used for incubation of 1. To 2 mL HPLC vials containing
200 µg of dried 1 (duplicate samples) were added cell suspension (200
µL), H2O (200 µL), and NAD+ (18 mM, 50 µL). Samples without 1
and samples with buffer instead of cell suspension were included as
controls. All samples were stirred at 30 °C, and after 1 and 6 h, 150
µL was withdrawn and stored at -18 °C. Samples were subsequently
analyzed by LC-MS [same column as above, eluted with aqueous 40%
MeCN (0.1% TFA) at 0.1 mL/min and with UV detection at 210 nm].
Prior to LC-MS analysis, the samples were thawed and centrifuged
(13 000 rpm, 5 min) and the supernatants transferred to HPLC vials.
All samples without cell suspension were also analyzed after 22 days.
In addition, samples of 1 were treated at pH 4.1 (100 mM acetate
buffer), pH 6.4 (100 mM bicarbonate buffer), and pH 7.6 (100 mM
phosphate buffer), also at 30 °C, and were repeatedly analyzed by LC-
MS over a period of 6 weeks.
Determination of Absolute Configuration of Verrucine F (3).
Verrucine F (20 µg) was hydrolyzed in 1 mL of 6 M HCl in a sealed
test tube at 110 °C for 18 h. The sample was dried under reduced
pressure and was treated with 200 µL of (S)-2-butanol/acetyl chloride
(10:1) in a closed vial at 100 °C for 45 min. After drying under a
stream of nitrogen, the sample was reacted with 200 µL of perfluoro-
propionic anhydride in a closed test tube at 100 °C for 40 min. The
reagent was removed with a stream of nitrogen and the sample dissolved
in EtOAc (150 µL) and analyzed by GC-MS (HP-5MS, 30 m × 0.25
mm, 0.25 µm, Agilent Technologies) with the oven held at 150 °C for
5 min and then raised to 200 °C at 5°/min. The temperature of both
the injector and the transfer line to the mass spectrometer was 240 °C.
As references, samples of D- and L-Glu were derivatized and analyzed
using the same protocol (D-Glu, 18.83 min; L-Glu, 18.96 min).
Identification and Quantification of 3 in Cultures. P. Verrucosum
strains were grown for 14 days, after which two agar plugs (5 mm
diameter) from the inner and outer parts of the colony were removed
from each culture, pooled, and extracted with MeOH (1 mL) by
vortexing. The mixtures were held at room temperature for 1 h, then
were vortexed again, after which the plugs were discarded. The extracts
were centrifuged (13 000 rpm, 10 min), and the supernatant was
transferred to HPLC vials for LC-MS analysis. LC-MS analysis was
performed on a C-18 column (2.1 × 125 mm, 3 µm, Dr. A. Maisch
High Performance LC GmbH) eluted with a gradient of MeCN in H2O
(0.1% TFA), running from 20% to 70% MeCN in 18 min, from 70%
to 100% in 1 min, maintained at 100% for 5 min, and re-equilibrated
at 20% for 8 min, at 0.2 mL/min and with UV detection at 332 nm.
MS data were used for supporting the identification of the compounds,
whereas UV data were used for quantification. As references for
identification and quantification, samples containing known amounts
of 1, 2, and 3 were used. For quantification of 1 and 3 during growth
on YES agar, P. Verrucosum J255 cultures were frozen at -20 °C 6
days after inoculation and then every 4 days for a total of 30 days.
After thawing, mycelia were divided at half the radius into inner and
outer portions, and excess agar was scraped from the reverse with a
spatula. Each portion was weighed and extracted into 10 times volume
MeOH by vortexing, followed by sonication for 10 min, and vortexing
again. Extracts were centrifuged in preparation for HPLC analysis.
Triplicate samples were analyzed at each time point, and analyses were
performed on a C-18 column (4.6 × 150 mm, 5 µm, Zorbax SB 300,
Acknowledgment. We thank Mr. Suresh Gohil for performing the
HRFABMS measurement. This work was supported by the Foundation
for Strategic Environmental Research (MISTRA).
References and Notes
(1) Frisvad, J. C.; Lund, F.; Elmholt, E. J. Appl. Microbiol. 2005, 98,
684–692.
(2) Commission Regulation (EC) No. 123/2005 of January 26, 2005,
amending regulation (EC) No. 466/2001 as regards ochratoxin A. Off.
J. Eur. Union, 2005, L 25, 3-5.
(3) Commission Recommendation of August 17, 2006, on the presence
of deoxynivalenol, zearalenone, ochratoxin A, T-2, and HT-2, and
fumonisins in products intended for animal feeding. Off. J. Eur. Union,
2006, L 229, 7-9.
(4) Larsen, T. O.; Franzyk, H.; Jensen, S. R. J. Nat. Prod. 1999, 62, 1578–
1580.
(5) Larsen, T. O.; Svendsen, A.; Smedsgaard, J. Appl. EnViron. Microbiol.
2001, 67, 3630–3635.
(6) Boyes-Korkis, J. M.; Guerney, K. A.; Penn, J.; Mantle, P. G.; Bilton,
J. N.; Sheppard, R. N. J. Nat. Prod. 1993, 56, 1707–1717.
(7) Hong Xin, Z.; Fang, Y.; Du, L.; Zhu, T.; Duan, L.; Chen, J.; Gu,
Q.-Q.; Zhu, W.-M. J. Nat. Prod. 2007, 70, 853–855.
(8) Broberg, A.; Menkis, A.; Vasiliauskas, R. J. Nat. Prod. 2006, 69, 97–
102.
(9) Framm, J.; Nover, L.; El Azzouny, A.; Richter, H.; Winter, K.; Werner,
S.; Luckner, M. Eur. J. Biochem. 1973, 37, 78–85.
(10) Aboutabl, S. Pharmazie 1974, 29, 73.
(11) Wang, H.; Sim, M. M. J. Nat. Prod. 2001, 64, 1497–1501.
(12) Larsen, T. O. Personal communication.
(13) Hocking, A. D.; Pitt, J. I.; Samson, R. A.; Thrane, U. AdVances in
Food Mycology; Springer: New York, 2006; p 356.
NP800105E