Total solid phase syntheses of the quinazoline alkaloids: Verrucines A and B and anacine

Article abstract of DOI:10.1021/np010342y

The first total syntheses of verrucines A and B and anacine (revised structure) were accomplished on Sasrin resin. This work confirmed the structure of verrucine A and unambiguously showed verrucine B to be a derivative of D-phenylalanine and L-glutamine. The study also proved that anacine and its epimer are quinazoline alkaloids, not benzodiazepines as originally proposed. 1-Hydroxyverrucine B, derived from air oxidization of verrucine B, was also isolated and characterized.

Full text of DOI:10.1021/np010342y

J . Nat. Prod. 2001, 64, 1497-1501
1497
Tota l Solid P h a se Syn th eses of th e Qu in a zolin e Alk a loid s: Ver r u cin es A a n d B
a n d An a cin e
Haishan Wang and Mui Mui Sim*
Institute of Molecular and Cell Biology, 30 Medical Drive, Singapore 117609, Singapore
Received J uly 13, 2001
The first total syntheses of verrucines A and B and anacine (revised structure) were accomplished on
Sasrin resin. This work confirmed the structure of verrucine A and unambiguously showed verrucine B
to be a derivative of D-phenylalanine and L-glutamine. The study also proved that anacine and its epimer
are quinazoline alkaloids, not benzodiazepines as originally proposed. 1-Hydroxyverrucine B, derived
from air oxidization of verrucine B, was also isolated and characterized.
Fungi are capable of incorporating anthranilic acid and
amino acids into a variety of fungal quinazoline metabo-
lites.1 Examples are glyantrypine from Aspergillus clava-
tus,2 fumiquinazolines F and G from Aspergillus fumiga-
tus,3 and fiscalin B from Neosartorya fischeri and Corynascus
setosus.4 Recently Larsen et al. reported the isolation of
verrucines A (1) and B (2) as a major and minor metabolite,
respectively, from Penicillium verrucosum.5 Verrucine A
was derived from anthranilic acid, L-phenylalanine (Phe),
and L-glutamine (Gln). However, the stereochemistry of its
isomer, verrucine B, was not unambiguously assigned due
to its epimerization. In the paper, the authors also pointed
out that anacine, which was isolated from Penicillium
aurantiogriseum, should have a quinazoline structure (4)
rather than a benzodiazepine structure (3) as originally
proposed by Mantle and co-workers.6 The structure of
anacine was revised according to the similarity of its UV
and NMR spectra to those of verrucines A and B. To
unambiguously confirm their structures as well as to
establish their absolute configurations, we undertook the
challenge of synthesizing verrucines A and B and anacine
(revised structure).
alkaloids.8 This protocol was adapted for our syntheses of
verrucines A and B and anacine.
We foresaw that the steric hindrance of the 1,4-syn
disubstituted bulky side chains in verrucine A and anacine
could interfere with the cyclization step. Therefore, we
envisioned that solid phase synthesis would be the best
choice for the synthesis of these compounds for two
reasons: (1) workup at each step would be easy, as excess
reagents could be simply washed away; and (2) only the
cyclized compound would be released from the resin. The
primary amide in the glutamine side chain was protected
in order to minimize dehydration of the amide to nitrile.7b
Trityl (triphenylmethyl) was selected as a suitable protect-
ing group for glutamine amide because of its ease of
removal under acidic conditions.
Starting with Fmoc-L-Gln(Trt)-Sasrin-resin (5), the Fmoc
was removed using piperidine, and the resulting amino
resin was coupled with anthranilic acid to give peptide 6
(Scheme 1). The anthranilamide 6 was initially acylated
with Fmoc-L-phenylalanine acid chloride,9 added in two
portions (5.1 and 4.7 equiv). Due to the highly acid labile
property of Sasrin resin, the reaction solution was kept
neutral to minimize the possible cleavage of the resin. The
linear peptide 7a was dehydrated to give benzoxazine10 8a ,
which was further deprotected and transformed to the
amidine intermediate11 9a . Cyclization of 9a in refluxing
MeCN-Cl(CH2)2Cl provided N-trityl verrucine A (11a ) as
the major product in 12% overall yield from 5. A trace
amount (<1%) of the anti epimer 10a was also isolated.
Refluxing the resin for an additional 13.5 h provided only
minute amounts of 10a (0.4%) and 11a (1.5%). Apparently,
the additional reaction time did not improve the yield, but
increased the ratio of the epimerized product to the desired
product. To improve the yield of 11a , resin 6 was instead
subjected to a double acylation process (2 × 5 equiv).
Following the same cyclization protocol described above,
11a was obtained in 17.0% final yield.
The 1,4-syn configuration of 11a was studied and estab-
lished by NOE. A NOE correlation was observed between
one of the Phe methylene protons at 3.19 ppm and one of
the Gln side chain methylene protons at 1.93 ppm, thus
suggesting that the Phe and Gln side chains were on the
same side, i.e., a 1,4-syn configuration. Similarly, a NOE
correlation was observed between the methine of Phe of
10a at 4.88 ppm and that of Gln methylene protons at 2.22
ppm, thus indicating that the Phe and Gln side chains were
on the opposite side, i.e., a 1,4-anti configuration. The
results were in accordance with the conformations calcu-
lated by MM2 (see Supporting Information).
Resu lts a n d Discu ssion
A general protocol for quinazoline alkaloid synthesis has
been recently reported by Wang and Ganesan7 and since
utilized by several groups to synthesize other quinazoline
Verrucine A was obtained in 84% yield by deprotecting
11a with 40% TFA in CH2Cl2. The proton NMR of the crude
* To whom correspondence should be addressed. Tel: + 65 8741443.
Fax: + 65-7791117. E-mail: mcbsimmm@imcb.nus.edu.sg.
10.1021/np010342y CCC: $20.00
© 2001 American Chemical Society and American Society of Pharmacognosy
Published on Web 11/08/2001

1498 J ournal of Natural Products, 2001, Vol. 64, No. 12
Wang and Sim
Sch em e 1. Total Solid Phase Syntheses of Quinazoline Alkaloids: Verrucines A and B and Anacinea
a
Reagents and conditions: (a) 20% piperidine in DMF, 15 min × 2. (b) EDC (13.4 equiv), anthranilic acid (12.1 equiv), DMF (method A) or NMP (method
B), rt 19 h. For 7a -11a : (c) pyridine (7.9 equiv), Fmoc-L-Phe-Cl (5.1 equiv), CH2Cl2, rt 13 h, workup; repeat once in DMF. (d) Ph3P (12.0 equiv), I2 (11.1
equiv), Et(i-Pr)2N (25.0 equiv), CH2Cl2, rt overnight (15 h). (e) 20% piperidine in CH2Cl2, rt 30 min. (f) MeCN/(CH2Cl)2 (1:1), reflux overnight. (g) TFA/
Et3SiH/CH2Cl2 (2:2:1), rt 15 min. Please refer to the Experimental Section for detailed conditions for 7b-11b and 7c-11c.
product showed that ca. 3% 1,4-anti epimer was also form-
ed. The H-1 and H-4 signals of 1 were well separated in
CDCl3 solution. The NOE data reconfirmed the 1,4-syn
configuration. Its NMR data in DMSO-d6 were identical
to those reported for verrucine A. The synthetic verrucine
A was found to be optically purer than the one reported
{[R]30 +57 ( 12° (c 0.25, EtOH), lit.5 [R]22 +37° (c 0.1,
those reported for anacine. The optical rotation data of
synthetic 4 {[R]27 +79 ( 13° (c 0.26, EtOH)} showed that
D
it has the same absolute configuration as the natural one
{[R]27 +81 ( 5° (c 0.60, EtOH)}.13 The absolute configu-
D
ration of 14 was not assigned in this work. However,
identical proton NMR data were obtained for both com-
pound 14 and the natural diastereoisomer of anacine. This
confirmed the relative 1,4-anti configuration of the natural
diastereoisomer of anacine.
D
D
EtOH)}.
Total synthesis of verrucine B was achieved in a similar
manner using Fmoc-D-Phe-Cl in the double acylation steps
(Scheme 1). 10b and 11b were obtained in 20.0% and 2.2%
yields, respectively. The 1,4-syn epimer (11b) had compa-
rable optical data {[R]31D +63 ( 27° (c 0.12, CHCl3)} to the
The overall yields for 11a , 10b, and 11c were relatively
low as compared to the previously reported less bulky
pyrazino[2,1-b]quinazoline-3,6(1H,4H)-diones on Wang
resins.7c The bulky Sasrin resin and the bulky trityl-
protected Gln may have an important role in lowering the
yields. To troubleshoot the problem, the Sasrin resins (6,
7a , 7c, 9a , 9c, post-refluxed 9a and 9c) were treated
separately with 5% TFA in CH2Cl2 for 30 min at room
temperature. Cleavage of resin 9a provided the desired
amidine 9a and unreacted linear peptide (7a , Fmoc ) H)
in a ratio of 93:7 as estimated by MS (the MS peak
intensities of the products were calibrated according to
their respective concentrations). Similarly, cleavage of resin
9c provided the desired amidine 9c and unreacted linear
peptide (7c, Fmoc ) H) in a ratio of 89:11 (by MS). Cleavage
of the post-MeCN-(CH2Cl)2 refluxed resin 9a provided the
linear peptide and some other unidentified products. This
might account for the small amount of products obtained
in the second cyclization. There was ca. 7% of amidine
obtained from the cleavage of post-MeCN-(CH2Cl)2 refluxed
resin 9c. Cleavage of 6 and 7 gave 16a and 17a as the only
byproducts. This clearly indicated that complete acylation
previous major product (11a ) {[R]31 +60 ( 4° (c 0.62,
D
CHCl3)}. The epimerization was observed to occur mainly
at the C-1 position.
Deprotection of 10b provided verrucine B (2, 73%) as well
as the 1,4-syn epimer (13, 27%).12 The NOE data of 2
showed that H-1 (4.96 ppm) and one of the Gln methylene
protons at 2.27 ppm were on the same side. The optical
rotation data of 2 showed that it had the same configura-
tion but was optically purer {[R]29 +183° (c 0.19, EtOH),
D
lit.5 [R]22 +124° (c 0.08, EtOH)} than the natural one.
D
Verrucine B was confirmed unambiguously as the deriva-
tive of D-phenylalanine and L-glutamine.
The synthesis of anacine started with Fmoc-L-Leu-Cl and
was accomplished using the same synthetic route. 1,4-syn
compound 11c and 1,4-anti epimer 10c were obtained in
12.5% and 0.4% yields, respectively. Deprotection of 11c
provided anacine (4, 74%) and the 1,4-anti epimer (14,
25%). The NMR data in DMSO-d6 of 4 were identical to

Syntheses of the Quinazoline Alkaloids
J ournal of Natural Products, 2001, Vol. 64, No. 12 1499
of 6 with amino acid chloride was achieved even though
the yield of isolated Fmoc derivative was only 70% from 5.
The major byproduct, formamidine 16a , was confirmed by
MS and NMR. Apparently the Fmoc-deprotected 5 and 6
condensed readily with solvent DMF under dehydrative
conditions to give formamidines 16a and 17a , respectively.
The latter was obtained at about 14% of 16a as estimated
by MS. A trace amount (0-0.5% of 15a as estimated by
MS) of the dimer 18a was also obtained. The ratio of 15a :
compounds cyclize-released from the resins. Besides con-
firming the absolute configuration of verrucine A, verrucine
B was shown unambiguously to be the derivative of
D-phenylalanine and L-glutamine. Anacine and its natural
diastereoisomer were confirmed as 1,4-syn and anti quinazo-
lines (4 and 14), respectively, rather than benzodiazepines.
Exp er im en ta l Section
Gen er a l Exp er im en ta l P r oced u r es. Please see previous
papers (ref 7) for details. All the 1D and 2D NMR experiments
for 1H (400.13 MHz), 13C (100.61 MHz), and 15N (40.55 MHz)
nuclei were obtained on a Bruker AVANCE-400 digital NMR
spectrometer. For solid phase, workup means filtered, washed
with DMF (×5), 10% MeOH/CH2Cl2 (×4), MeOH (×4), 10%
MeOH/CH2Cl2 (×4), and MeOH (×4), dried under vacuum.
Cleavage means the Sasrin resin was shaken with TFA-Et3-
SiH-CH2Cl2 (5:5:90) at room temperature for 30 min, then
filtered and washed with 10% MeOH/CH2Cl2 and MeOH, and
the filtrate was evaporated to dryness.
1
16a :17a was 78.4:19.1:2.6 as estimated by HPLC and H
NMR. The same study was carried out for 15b-d on Wang
resins (Ala, Phe, and Leu).7c All the cleaved (TFA-Et3SiH-
CH2Cl2, 50:5:45, rt, 1.5 h) products showed the respective
peaks of 15b-d through 18b-d by MS, although the
amount was less when compared to those obtained from
cleaving 6. The purities of 15b, 15c, and 15d were
estimated as 90%, 92%, and 88% (respectively) by 1H NMR.
N-(2-Am in oben zoyl)-L-Gln (Tr t)-Sasr in Resin (6). Meth od
A. Fmoc-L-Gln(Trt)-Sasrin-resin (5, Feinchemikallen AG, load-
ing 0.345 mmol/g, 1.919 g) was converted to resin 6 (1.738 g,
new loading of 0.381 mmol/g) according to ref 7c.
The quality of resin 5 was checked by cleaving the depro-
tected resin 5, and the cleaved L-Gln(Trt)-OH gave m/z 389.15
as the only peak. Resin 6 was cleaved, and the product was
analyzed by MS, HPLC, and NMR. MS (ESI, positive mode):
m/z (relative intensity) 508.2307 (M + H, 100%, 15a , calcd for
C
31H29N3O4+H, 508.2236), 444.2352 (M + H, 41%, 16a , calcd
for C27H29N3O3+H, 444.2287), 563.2809 (M + H, 5%, 17a , calcd
for C34H34N4O4+H, 563.2658), 627.25 (M + H, 1%, 18a , calcd
for C38H34N4O5+H, 627.26); the formamidine 16a is more
sensitive than 15a . Analytical HPLC (C18, 2.1 × 200 mm, 5
µm, flow rate 0.3 mL/min, linear gradient: 30-100% MeCN/
H2O + 0.1% TFA in 40 min) showed 16a (tR 14.9 min) and
15a (tR 17.3 min) in a ratio of 20:80 (peak area at 220 nm).
The crude product was also purified by preparative HPLC (C18,
25 × 210 mm, 6 µm, flow rate 20 mL/min, linear gradient:
40-100% MeCN/H2O + 0.04% TFA in 20 min) to afford 15a
and 16a . The formamidine part was confirmed by 1D and 2D
NMR (acetone-d6). Two methyl groups at δH 2.85 (δC 37.1) and
3.46 (δC 44.2) were correlated with the quaternary carbon at
158.2 ppm by HMQC and HMBC. The purity of resin 6 was
estimated as 78%.
To avoid the formation of undesired 16 and 17, the
coupling reaction was carried out in 1-methyl-2-pyrrolidi-
none (NMP). The cleaved product 15a was cleaner, con-
taining only a trace amount of 18a (2.4% as estimated by
MS) as the only byproduct.
Compounds 2 and 4 in DMSO-d6 and MeOH were found
to degrade when they were kept at room temperature and
exposed to air for several weeks. Purification of the
decomposed mixture of 2 provided a polar major product,
19. The proton NMR showed that H-1 has disappeared
Resin-bound 15b-d were also cleaved with 50% TFA for
1.5 h. The formation of formamidines 16b-d was also con-
firmed by MS and NMR. For example, the formamidine part
1
of 16b was identified at δH 8.04 (s, 1H, CH, δC 156.0, J C,H
)
196 Hz), 3.11 (s, 3H, Me, δC 36.5), and 3.27 (s, 3H, Me, δC 43.8);
all these protons were correlated with the carbon at 156.0 ppm
by HMQC and HMBC (CD3OD). Its purity was 90%, as
1
estimated by H NMR.
Meth od B. This method is similar to method A except NMP
was used instead of DMF. The cleavage product showed only
trace amounts of 18a (2.4% by MS).
(1S,4S)-1,3,4,6-Tetr a h yd r o-3,6-d ioxo-1-(p h en ylm eth yl)-
2H -p y r a zin o [2,1-b ]q u in a zo lin e -4-(N -t r it y l)p r o p a n a -
m id e (N-Tr ityl-ver r u cin e A, 11a ). Resin 6 (0.584 g, equal
to 0.223 mmol of 5, prepared by method A, purity 71%) was
acylated with Fmoc-L-Phe-Cl (5.1 + 4.7 equiv), then dehy-
drated, deprotected, and cyclized according to ref 7c. The
process provided 11a (8.8 mg, 12.0% from 5) and trace amounts
of epimer 10a (<1%). The resin was further reacted for 13.5 h
to give additional 10a (0.3 mg, 0.4%) and 11a (1.1 mg, 1.5%)
after purification.
Dou b le Acyla t ion . To a mixture of 6 (0.651 g, equal to
0.256 mmol of 5, prepared by method A, purity 78%), CH2Cl2
(6.5 mL), and pyridine (2.03 mmol, 7.9 equiv to 5) was added
solid Fmoc-L-Phe-Cl (0.526 g, 1.30 mmol, 5.1 equiv). The
reaction mixture was shaken at room temperature for 13 h
and worked up. To the resin was added DMF (4 mL), pyridine
(1.98 mmol, 7.7 equiv), and solid Fmoc-L-Phe-Cl (0.521 g, 1.28
mmol, 5.0 equiv). The mixture was shaken at room tempera-
while a new carbon appeared at 81.97 ppm in the 13C NMR.
With additional information from MS, 19 was assigned as
an oxidized product of 2 at the C-1 position. The stereo-
chemistry was established by NOE. Since natural products
such as fumiquinazolines C and E contain oxygen at the
C-1 positions,3 it is therefore likely that an oxidation
reaction will occur at the C-1 position of verrucine B.
In summary, (+)-verrucines A and B and anacine were
synthesized in seven steps in overall yields of 14.3%, 14.5%,
and 9.3%, respectively.14 Although the yields obtained were
relatively low, their purities were high,7c as only the desired

1500 J ournal of Natural Products, 2001, Vol. 64, No. 12
Wang and Sim
ture overnight, followed by a workup to give resin 7a . It was
processed as above to give dibenzofulvene-piperidine adducts
(69%), 10a (Rf 0.57, 1.1 mg, 0.8% from 5), and 11a (Rf 0.31,
24.1 mg, 17.0% from 5) after purification by preparative TLC
(EtOAc/hexane, 2:1).
) 14.5, 3.7) and 2.95 (dd, 1H, J ) 14.6, 10.4, Phe-CH2), 2.59
(m, 1H) and 2.52 (m, 1H, Gln-CH2CONH), 2.27 (m, 1H) and
2.22 (m, 1H, Gln-CH2CNH); 13C NMR (CDCl3) δ 170.10, 167.95,
160.68 (C-6), 149.82 (C-11a), 146.82 (C-10a), 144.51 (aromatic
Cq of Ph3C), 135.28 (Cq of Phe), 134.85 (C-9), 129.46 (CH ×
2), 129.32 (CH × 2), 128.67 (CH × 6), 127.89 (CH × 6), 127.81
(d), 127.63 (d), 127.54 (d), 126.99 (CH × 3), 126.95 (d), 120.35
(s), 70.61 (Ph3CN), 55.44 (C-4), 53.56 (C-1), 37.55 (Phe-CH2),
33.58 (Gln-CH2CONH), 26.89 (Gln-CH2CHN); NOESY and
ROESY δH 4.88 with 2.22; 15N NMR (obtained from 1H-15N
HSQC and HMBC in CDCl3) δN 112.0 (N-2, 1J NH ) 90.5), 139.0
(CONHTrt, 1J NH ) 85.8), 165.9 (N-5) and 237.1 (N-11); HRMS
(ESI, positive mode) m/z 619.2725 ([M + H]+, calcd for
Com p ou n d 11a : white solid, mp 134 °C (dec); [R]28 +60
D
( 4° (c 0.62, CHCl3); 1H NMR (CDCl3) δ 8.29 (dd, 1H, J ) 8.0,
1.0, H-7), 7.80 (td, 1H, J ) 7.8, 1.4, H-9), 7.70 (d, 1H, J ) 8.1,
H-10), 7.52 (t, 1H, J ) 7.1, H-8), 7.30-7.21 (m, 20H, Ph +
Ph3C), 7.12 (s, 1H, Ph3CNH), 6.18 (d, 1H, J ) 3.7, H-2), 5.26
(dd, 1H, J ) 9.9, 5.7, H-4), 4.71 (dt, 1H, J ) 10.7, 3.8, H-1),
3.43 (dd, 1H, J ) 13.4, 3.5) and 3.19 (dd, 1H, J ) 13.4, 10.7,
Phe-CH2), 2.69 (dd, 1H, J ) 15.2, 6.6) and 2.63 (dd, 1H, J )
15.2, 7.4, Gln-CH2CONH), 2.33 (m, 1H) and 1.93 (m, 1H, Gln-
CH2CNH); 13C NMR (CDCl3) δ 170.12, 167.22, 160.53 (C-6),
149.77 (C-11a), 147.12 (C-10a), 144.73 (aromatic Cq of Ph3C),
135.56 (Cq of Phe), 134.94 (C-9), 129.66 (CH × 2), 129.14 (CH
× 2), 128.81 (CH × 6), 127.89 (CH × 6), 127.59 (d), 127.26 (d),
126.98 (d), 126.97 (d), 126.94 (CH × 3), 120.10 (s), 70.63
(Ph3CN), 58.50 (C-1), 54.72 (C-4), 43.89 (Phe-CH2), 34.15 (Gln-
CH2CONH), 29.50 (Gln-CH2CHN); NOE δH 3.19 (3.43, 1.93),
1.93 (5.26, 3.19); HRMS (ESI, positive mode) m/z 619.2705 ([M
+ H]+, calcd for C40H34N4O3+H, 619.2709).
C
40H34N4O3+H, 619.2709); m/z 641.2526 ([M + Na]+, calcd for
C40H34N4O3Na, 641.2529).
(1R,4S)-(+)-1,3,4,6-Tetr a h yd r o-3,6-d ioxo-1-(p h en ylm e-
th yl)-2H-pyr azin o[2,1-b]qu in azolin e-4-pr opan am ide (Ver -
r u cin e B, 2). Compound 10b (19.8 mg, 0.0320 mmol) was dis-
solved in CH2Cl2 (1 mL) and reacted with triethylsilane (0.2
mL) and TFA (0.8 mL) at room temperature for 30 min. The
solution was evaporated, and the residue was purified by
preparative TLC (EtOAc/MeOH/CH2Cl2, 50:5:45) to give ver-
rucine B 2 (Rf 0.34, 8.7 mg, 73%) and the syn epimer 13 (Rf
(1S,4S)-1,3,4,6-Tetr a h yd r o-3,6-d ioxo-1-(p h en ylm eth yl)-
2H-p yr a zin o[2,1-b]qu in a zolin e-4-p r op a n a m id e (Ver r u -
cin e A, 1). Compound 11a (9.1 mg, 0.0147 mmol) was
dissolved in CH2Cl2 (0.2 mL) and reacted with TFA (0.2 mL)
and triethylsilane (0.1 mL) at room temperature for 15 min.
The solution was evaporated, and the residue was purified by
preparative TLC (EtOAc/MeOH/CH2Cl2, 50:5:45) to give re-
covered 11a (1.2 mg, 13%), verrucine A (1) (4.6 mg, 84%), and
a trace amount of anti epimer 12 (ca. 3%, estimated by proton
0.29, 3.2 mg, 27%). Verrucine B 2: white solid; [R]29 +183 (
D
1
13° (c 0.19, EtOH), lit.5 [R]22 +124° (c 0.08, EtOH); H NMR
D
(CDCl3) δ 8.28 (dd, 1H, J ) 8.3, 1.0, H-7), 7.80 (td, 1H, J )
7.6, 1.5, H-9), 7.74 (d, 1H, J ) 7.8, H-10), 7.53 (td, 1H, J )
7.5, 1.2, H-8), 7.41 (m, 2H, Ph), 7.37-7.32 (m, 3H, Ph), 6.47
(s, 1H, H-2), 5.78 (s, 1H) and 5.50 (s, 1H, CONH2), 5.39 (t, 1H,
J ) 8.1, H-4), 4.96 (dd, 1H, J ) 10.2, 3.6, H-1), 4.15 (dd, 1H,
J ) 14.6, 3.5) and 2.97 (dd, 1H, J ) 14.6, 10.3, Phe-CH2), 2.45
(t, 2H, J ) 6.7, Gln-CH2CONH2), 2.31 (m, 1H) and 2.25 (m,
1H, Gln-CH2CNH); 13C NMR (CDCl3) δ 173.71 (CONH2),
168.25 (C-3), 160.81 (C-6), 149.99 (C-11a), 146.86 (C-10a),
135.72 (Cq of Phe), 134.86 (C-9), 129.53 (CH × 2), 129.37 (CH
× 2), 127.68 (CH), 127.66 (C-10), 127.52 (C-8), 126.88 (C-7),
120.32 (C-6a), 55.39 (C-4), 53.98 (C-1), 37.57 (Phe-CH2), 31.35
(Gln-CH2CONH2), 26.10 (Gln-CH2CHN); NOESY and ROESY
δH 4.96 with 2.27; HRMS (ESI, positive mode) m/z 377.1628
([M + H]+, calcd for C21H20N4O3+H, 377.1614); m/z 399.1418
([M + Na]+, calcd for C21H20N4O3Na, 399.1433); syn epimer
13: MS (ESI, positive mode) m/z 377.159 ([M + H]+) and
399.135 ([M + Na]+).
NMR). Verrucine A (1): white solid; [R]30 +57 ( 12° (c 0.25,
D
EtOH), lit.5 [R]22 +37° (c 0.1, EtOH); 1H NMR (CDCl3) δ 8.28
D
(dd, 1H, J ) 8.0, 1.0, H-7), 7.81 (ddd, 1H, J ) 8.3, 7.0, 1.4,
H-9), 7.71 (d, 1H, J ) 8.2, H-10), 7.53 (td, 1H, J ) 7.6, 1.0,
H-8), 7.39-7.28 (m, 5H, Ph), 6.40 (d, 1H, J ) 3.6, H-2), 6.07
(br s, 1H) and 5.54 (br s, 1H, CONH2), 5.20 (dd, 1H, J ) 10.0,
5.4, H-4), 4.79 (dt, 1H, J ) 10.5, 3.8, H-1), 3.51 (dd, 1H, J )
13.5, 3.7) and 3.27 (dd, 1H, J ) 13.5, 10.4, Phe-CH2), 2.64 (dt,
1H, J ) 15.8, 7.4) and 2.61 (dt, 1H, J ) 15.8, 6.8, Gln-CH2-
CONH2), 2.31 (m, 1H) and 1.91 (m, 1H, Gln-CH2CNH); 13C
NMR (CDCl3) δ 173.61 (CONH2), 167.25 (C-3), 160.73 (C-6),
149.71 (C-11a), 147.14 (C-10a), 135.58 (Cq of Phe), 135.02 (C-
9), 129.68 (CH × 2), 129.20 (CH × 2), 127.69 (CH × 6), 127.33
(C-8), 127.02 (C-10 or C-7), 126.92 (C-7 or C-10), 120.03 (C-
6a), 58.42 (C-1), 54.65 (C-4), 43.96 (Phe-CH2), 32.26 (Gln-CH2-
CONH2), 29.21 (Gln-CH2CHN); NOESY and ROESY δH 3.27
with 1.91; HRMS (ESI, positive mode) m/z 377.1601 ([M + H]+,
calcd for C21H20N4O3+H, 377.1614); m/z 399.1418 ([M + Na]+,
calcd for C21H20N4O3Na, 399.1433). anti epimer 12: HRMS
(ESI, positive mode) m/z 377.1608 ([M + H]+, calcd for
(1S,4S)-1,3,4,6-Tetr ah ydr o-3,6-dioxo-1-(2-m eth ylpr opyl)-
2H -p y r a zin o [2,1-b ]q u in a zo lin e -4-(N -t r it y l)p r o p a n a -
m id e (N-Tr ityl-a n a cin e, r evised str u ctu r e, 11c). The pro-
cedure was similar to the one described for 11a (double
acylation), except that resin 6 (0.662 g, equal to 0.260 mmol
of 5, prepared by method A, purity 78%) was acylated twice
with Fmoc-L-Leu-Cl (total 5.0 equiv × 2). The process provided
dibenzofulvene-piperidine adducts (66%), 10c (Rf 0.63, 0.5 mg,
0.4% from 5), and 11c (Rf 0.40, 16.7 mg, 12.5% from 5) after
purification by preparative TLC (EtOAc/hexane, 2:1). anti
epimer 10c: HRMS (ESI, positive mode) m/z 585.2892 ([M +
H]+, calcd for C37H36N4O3+H, 585.2866).
C
C
21H20N4O3+H, 377.1614); m/z 399.1417 ([M + Na]+, calcd for
21H20N4O3Na, 399.1433).
(1R,4S)-1,3,4,6-Tetr a h yd r o-3,6-d ioxo-1-(p h en ylm eth yl)-
2H -p y r a zin o [2,1-b ]q u in a zo lin e -4-(N -t r it y l)p r o p a n a -
m id e (N-Tr ityl-ver r u cin e B, 10b). The procedure was
similar to the one described for 11a (double acylation), except
that resin 6 (0.508 g, equal to 0.193 mmol of 5, prepared by
method A, purity 71%) was acylated twice with Fmoc-D-Leu-
Cl (total 5.7 equiv × 2). The process provided dibenzofulvene-
piperidine adducts (40.0 mg, 79%), 10b [Rf 0.55, 23.9 mg
(20.0%), and 1.1 mg (0.9%, 2nd batch)], and the syn epimer
11b [Rf 0.27, 2.6 mg (2.2%) and 0.6 mg (0.5%, 2nd batch)] after
purification by preparative TLC (MeOH/CH2Cl2/EtOAc/hexane,
syn 11c: white solid; mp 131 °C (dec); [R]29 +94 ( 7° (c
D
1
0.53, CHCl3); H NMR (CDCl3) δ 8.27 (dd, 1H, J ) 8.0, 1.0,
H-7), 7.77 (ddd, 1H, J ) 8.3, 7.0, 1.4, H-9), 7.64 (dd, 1H, J )
8.2, 0.5, H-10), 7.49 (ddd, 1H, J ) 8.0, 7.1, 1.0, H-8), 7.30-
7.26 (m, 6H) and 7.26-7.20 (m, 9H, Ph3C), 7.13 (s, 1H, Ph3-
CNH), 6.91 (d, 1H, J ) 4.2, H-2), 5.27 (dd, 1H, J ) 9.6, 5.8,
H-4), 4.52 (ddd, 1H, J ) 9.6, 5.3, 2.7, H-1), 2.76 (ddd, 1H, J )
16.7, 7.1, 4.7) and 2.68 (dd, 1H, J ) 15.5, 7.6, Gln-CH2CONH),
2.38 (m, 1H) and 2.19 (m, 1H, Gln-CH2CNH); 1.86-1.80 (m,
2H, Leu-CH2), 1.77 (m, 1H, Leu-CHMe2), 0.98 (d, 6H, J ) 6.5,
Leu-CHMe2); 13C NMR (CDCl3) δ 170.19 (CONH2), 168.00 (C-
3), 160.73 (C-6), 150.79 (C-11a), 147.19 (C-10a), 144.71 (aro-
matic Cq of Ph3C), 134.83 (C-9), 128.80 (CH × 6), 127.87 (CH
× 6), 127.09 (C-8 or C-10), 127.02 (C10 or C-8), 126.93 (CH ×
3), 126.85 (C-7), 120.00 (s), 70.58 (Ph3CN), 54.98 (C-1), 54.77
(C-4), 46.87 (Leu-CH2), 34.37 (Gln-CH2CONH), 29.91 (Gln-CH2-
CHN), 24.65 (Leu-CHMe2), 23.08 and 21.15 (Leu-Me2); NOESY
and ROESY δH 2.20 with 1.84; 15N NMR (obtained from 1H-
15N HSQC and HMBC in CDCl3) δN 112.8 (N-2), 138.6
2.5:47.5:25:25). syn epimer 11b: [R]31 +63 ( 27° (c 0.12,
D
CHCl3); HRMS (ESI, positive mode) m/z 619.2743 ([M + H]+,
calcd for C40H34N4O3+H, 619.2709).
Com p ou n d 10b: white solid; [R]31 +149 ( 13° (c 0.19,
D
1
CHCl3); H NMR (CDCl3) δ 8.28 (dd, 1H, J ) 8.0, 1.0, H-7),
7.79 (td, 1H, J ) 7.6, 1.4, H-9), 7.73 (d, 1H, J ) 8.0, H-10),
7.50 (td, 1H, J ) 7.5, 1.1, H-8), 7.39-7.31 (m, 3H, Ph), 7.29 (d
like, 2H, Ph), 7.26-7.18 (m, 9H, Ph3C), 7.17-7.14 (m, 6H,
Ph3C), 6.92 (s, 1H, Ph3CNH), 5.82 (s, 1H, H-2), 5.43 (t, 1H, J
) 8.0, H-4), 4.88 (dd, 1H, J ) 10.3, 3.8, H-1), 4.06 (dd, 1H, J
1
(CONHTrt, J NH ) 86), 166.1 (N-5) and 236.4 (N-11); HRMS

Syntheses of the Quinazoline Alkaloids
J ournal of Natural Products, 2001, Vol. 64, No. 12 1501
(ESI, positive mode) m/z 585.2861 ([M + H]+, calcd for
44.07 (Phe-CH2), 30.39 (Gln-CH2CONH2), 26.37 (Gln-CH2-
CHN); NOESY δH 6.46 with 2.49; 1H-13C HMBC δH (δC) 7.96
(169.28, 150.00, 81.97, 54.79), 6.47 (81.97, 44.07); 1H-15N
HSQC δH (δN) 7.96 (113.4), 6.62 and 5.42 (85.6); HRMS (ESI,
positive mode) m/z 393.1545 ([M + H]+, calcd for C21H20N4O4+H,
393.1563); m/z 415.1378 ([M + Na]+, calcd for C21H20N4O4Na,
415.1382); m/z 375.1425 ([M + H - H2O]+, calcd for C21H19N4O3,
375.1457). Minor isomer: HRMS (ESI, positive mode) m/z
393.1559 ([M + H]+, calcd for C21H20N4O4+H, 393.1563).
C
37H36N4O3+H, 585.2866); m/z 607.2644 ([M + Na]+, calcd for
C37H36N4O3Na, 607.2685).
(1S,4S)-1,3,4,6-Tetr ah ydr o-3,6-dioxo-1-(2-m eth ylpr opyl)-
2H-p yr a zin o[2,1-b]qu in a zolin e-4-p r op a n a m id e (An a cin e,
r evised str u ctu r e, 4) a n d Ep im er of An a cin e (14). Com-
pound 11c (10.5 mg, 0.0180 mmol) was dissolved in CH2Cl2 (1
mL) and reacted with triethylsilane (0.2 mL) and TFA (0.8
mL) at room temperature for 30 min. The solution was
evaporated and purified by preparative TLC (EtOAc/MeOH/
CH2Cl2, 50:5:45) to give anacine 4 (Rf 0.29, 4.5 mg, 74%) and
the anti epimer 14 (Rf 0.36, 1.5 mg, 25%). Anacine 4: white
solid; [R]27D +79 ( 13° (c 0.26, EtOH); 1H NMR (CDCl3) δ 8.26
(dd, 1H, J ) 8.0, 1.2, H-7), 7.76 (td, 1H, J ) 7.7, 1.3, H-9),
7.65 (d, 1H, J ) 8.1, H-10), 7.49 (td, 1H, J ) 7.6, 1.0, H-8),
7.22 (d, 1H, J ) 4.0, H-2), 6.16 (s, 1H) and 5.74 (s, 1H, CONH2),
5.23 (dd, 1H, J ) 9.9, 5.5, H-4), 4.60 (m, 1H, H-1), 2.68 (t, 2H,
J ) 7.1, Gln-CH2CONH2), 2.40 (m, 1H) and 2.23 (m, 1H, Gln-
CH2CNH), 1.97 (overlapped m, 1H, Leu-CHMe2), 1.94 (d, 2H,
J ) 5.4, Leu-CH2), 1.08 (t, 3H, J ) 6.5) and 1.06 (t, 3H, J )
6.5, Leu-CHMe2); 13C NMR (CDCl3) δ 173.85 (s), 168.04 (s),
160.93 (s), 150.83 (s), 147.23 (s), 134.90 (C-9), 127.15 (C-8 or
C-10), 127.07 (C-10 or C-8), 126.81 (C-7), 119.92 (s), 54.98 (C-
1), 54.79 (C-4), 47.05 (Leu-CH2), 32.35 (Gln-CH2CONH2), 29.45
(Gln-CH2CHN), 24.73 (Leu-CHMe2), 23.20 and 21.17 (Leu-
Me2); ROESY δH 2.68 with 1.07, 2.23 with 1.94; HRMS (ESI,
positive mode) m/z 343.1777 ([M + H]+, calcd for C18H22N4O3+H,
343.1770); m/z 365.1577 ([M + Na]+, calcd for C18H22N4O3Na,
365.1590).
Ack n ow led gm en t. The authors thank Professor Peter G.
Mantle (Imperial College of Science, Technology and Medicine,
London SW7 2AY, UK) for providing an authentic sample of
anacine. This work was supported by grants from the National
Science and Technology Board of Singapore.
Su p p or tin g In for m a tion Ava ila ble: 1H and/or 13C NMR spectra
for 1, 2, 4, 10b, 11a , 11c, 14, and 19 in CDCl3. List of 1H and 13C
NMR data for 1, 2, 4, and 14 in DMSO-d6. Conformations of 1, 2, 4,
10b, 11a , 10c, 11c, and 14 calculated by MM2. This material is
available free of charge via the Internet at http://pubs.acs.org.
Refer en ces a n d Notes
(1) For recent reviews, see: (a) Michael, J . P. Nat. Prod. Rep. 2000, 17,
603-620. (b) J ohne, S. In Supplements to the 2nd Edition of Rodd’s
Chemistry of Carbon Compounds; Ansell, M. F., Ed.; Elsevier:
Amsterdam, 1995, Vol. IV I/J , pp 223-240.
(2) Penn, J .; Mantle, P. G.; Bilton, J . N.; Sheppard, R. N. J . Chem. Soc.,
Perkin Trans. 1 1992, 1495-1496.
(3) (a) Numata, A.; Takahashi, C.; Matsushita, T.; Miyamoto, T.; Kawai,
K.; Usami, Y.; Matsumura, E.; Inoue, M.; Ohishi, H.; Shingu, T.
Tetrahedron Lett. 1992, 33, 1621-1624. (b) Takahashi, C.; Mat-
sushita, T.; Doi, M.; Minoura, K.; Shingu, T.; Kumeda, Y.; Numata,
A. J . Chem. Soc., Perkin Trans. 1 1995, 2345-2353.
(4) (a) Wong, S.-M.; Musza, L. L.; Kydd, G. C.; Kullnig, R.; Gillum, A.
M.; Cooper, R. J . Antibiot. 1993, 46, 545-553. (b) Fujimoto, H.;
Negishi, E.; Yamaguchi, K.; Nishi, N.; Yamazaki, M. Chem. Pharm.
Bull. 1996, 44, 1843-1848.
a n ti ep im er 14: 1H NMR (CDCl3) δ 8.26 (d, 1H, J ) 7.9,
H-7), 7.79 (t, 1H, J ) 7.1, H-9), 7.70 (d, 1H, J ) 8.0, H-10),
7.51 (t, 1H, J ) 7.4, H-8), 6.13 (s, 1H, H-2), 5.78 (br s, 1H, one
of CONH2), 5.47 (t, 1H, J ) 8.2, H-4), 5.40 (br s, 1H, one of
CONH2), 4.74 (dd, 1H, J ) 9.4, 3.6, H-1), 2.61-2.49 (m, 3H,
Gln-CH2CONH2 and one of Leu-CH2), 2.36-2.30 (m, 2H, Gln-
CH2CNH), 1.87 (m, 1H, Leu-CHMe2), 1.75 (ddd, 1H, J ) 14.3,
9.5, 4.7, one of Leu-CH2), 1.11 (d, 3H, J ) 6.7) and 1.09 (d,
3H, J ) 6.5, Leu-CHMe2); 13C NMR (CDCl3) δ 173.38 (CONH2),
168.54 (C-3), 160.97 (C-6), 150.65 (C-11a), 146.89 (C-10a),
134.75 (C-9), 127.70 (C-10), 127.41 (C-8), 126.82 (C-7), 120.22
(C-6a), 55.47 (C-4), 51.16 (C-1), 39.55 (Leu-CH2), 31.48 (Gln-
CH2CONH2), 25.98 (Gln-CH2CHN), 24.69 (Leu-CHMe2), 23.62
and 21.29 (Leu-CHMe2); NOESY δH 4.74 with 2.34; HRMS
(ESI, positive mode) m/z 343.1781 ([M + H]+, calcd for
(5) Larsen, T. O.; Franzyk, H.; J ensen, S. R. J . Nat. Prod. 1999, 62, 1578-
1580.
(6) Boyes-Korkis, J .; Gurney, K. A.; Penn, J .; Mantle, P. G.; Bilton, J .
N.; Sheppard, R. N. J . Nat. Prod. 1993, 56, 1707-1717.
(7) (a) Wang, H.; Ganesan, A. J . Org. Chem. 1998, 63, 2432-2433. (b)
Wang, H.; Ganesan, A. J . Org. Chem. 2000, 65, 1022-1030. (c) Wang,
H.; Ganesan, A. J . Comb. Chem. 2000, 2, 186-194.
C
C
18H22N4O3+H, 343.1770); m/z 365.1604 ([M + Na]+, calcd for
18H22N4O3Na, 365.1590).
(8) (a) Hart, D. J .; Magomedov, N. Tetrahedron Lett. 1999, 40, 5429-
5432. (b) Snider, B. B.; Zeng, H. Org. Lett. 2000, 2, 4103-4106.
(9) Carpino, L. A.; Cohen, B. J .; Stephens, K. E., J r.; Sadat-Aalaee, S.
Y.; Tien, J .-H.; Langridge, D. C. J . Org. Chem. 1986, 51, 3732-3734.
(10) Mazurkiewicz, R. Monatsh. Chem. 1989, 120, 973-980.
(11) He, F.; Snider, B. B. J . Org. Chem. 1999, 64, 1397-1399.
(12) Epimerization may be due to the strongly acidic condition. Numata
et al. (ref 3) pointed out that the fumiquinazolines epimerized at both
C-1 and C-4 positions under strongly basic conditions, but epimerized
at C-1 only under strongly acidic conditions. C-6-benzene ring-N-
11-C-11a-C-1 is a highly conjugated unsaturated ketone-like system.
The interchangeable ketone and enol forms are the driving force for
epimerization at the C-1 position.
(1S,4S)-1,3,4,6-Tetr ah ydr o-3,6-dioxo-1-h ydr oxyl-1-(ph en -
ylm eth yl)-2H-p yr a zin o[2,1-b]qu in a zolin e-4-pr op a n a m id e
(1-Hyd r oxyver r u cin e B, 19). Oxid a tion of Ver r u cin e B.
The NMR sample of 2 in DMSO-d6 was transferred into a flask.
The verrucine B in DMSO-d6 and MeOH solution was kept at
room temperature for several weeks. After removal of the
solvent under high vacuum, the oxidized mixture was purified
by preparative TLC (EtOAc/MeOH/CH2Cl2, 50:5:45). Two polar
compounds were isolated. The major one was assigned as
1-hydroxyverrucine B, while the minor one might be its isomer.
Oxidized verrucine B (19): white solid; 1H NMR (CDCl3) δ 8.28
(d, 1H, J ) 7.9, H-7), 7.96 (s, 1H, H-2), 7.85 (strong coupled
dd, 1H, J ) 8.0, 1.4, H-10), 7.82 (td, 1H, J ) 8.3, 1.2, H-9),
7.54 (td, 1H, J ) 7.5, 1.2, H-8), 7.45 (dd, 2H, J ) 8.0, 0.8, Ph),
7.37-7.28 (m, 3H, Ph), 6.62 (br s, 1H, one of CONH2), 6.47 (s,
1H, OH-1), 5.42 (s, 1H, one of CONH2), 5.24 (dd 1H, J ) 9.4,
7.7, H-4), 3.90 (d, 1H, J ) 14.1) and 3.63 (d, 1H, J ) 14.1,
Phe-CH2), 2.60 (m, 1H) and 2.49 (m, 1H, Gln-CH2CONH2), 2.58
(m, 1H) and 2.50 (m, 1H, Gln-CH2CNH); 13C NMR (CDCl3) δ
176.06 (CONH2), 169.28 (C-3), 160.76 (C-6), 150.00 (C-11a),
146.78 (C-10a), 134.82 (C-9), 134.39 (Cq of Phe), 131.70 (CH
× 2), 128.55 (CH × 2), 128.08 (C-10), 127.67 (C-8), 127.34
(PhH), 126.71 (C-7), 120.45 (C-6a), 81.97 (C-1), 54.79 (C-4),
(13) Optical rotation data of anacine and its natural diastereoisomer were
not reported in the original paper (ref 6). The value reported here
was obtained from the authentic sample provided by Professor
Mantle.
(14) The authentic samples of verrucines A and B were not available.
NP010342Y