Synthesis of 11-hydroxyl O-methylsterigmatocystin and the role of a cytochrome P-450 in the final step of aflatoxin biosynthesis

Article abstract of DOI:10.1021/ja012185v

The major skeletal rearrangements (anthraquinone → xanthone → coumarin) that occur in the complex biosynthesis of aflatoxin B₁ are mediated by cytochromes P-450. Previous experiments have suggested that two successive monooxygenase reactions are required to convert the xanthone O-methylsterigmatocystin (OMST) to aflatoxin, a process we demonstrate is mediated by a single P-450, OrdA, in Aspergillus parasiticus in accord with findings in A. flavus. The first oxidative cycle is proposed to result in the formation of 11-hydroxy O-methylsterigmatocystin (HOMST), while the second entails aryl ring cleavage, demethylation, dehydration, decarboxylation, and rearrangement to give aflatoxin - a remarkable sequence of transformations. To test this hypothesis, HOMST has been synthesized by an alkylnitrilium variant of the Houben-Hoesch reaction. The troublesome xanthone carbonyl was protected as a butylene to allow further elaboration of the molecule, and then the product xanthone was restored in a uniquely facile peracid deprotection. Methods were devised to construct the sensitive dihydrobisfuran and to maintain the oxidation state of the partially methylated hydroquinone. Expression of ordA in a yeast membrane preparation enabled the intermediacy of HOMST both to be detected in the conversion of OMST to aflatoxin and to be established directly in the biosynthesis of the mycotoxin. Having secured the role of HOMST in aflatoxin formation, the mechanism of the second oxidative cycle of this P-450 is considered.

Full text of DOI:10.1021/ja012185v

Published on Web 04/17/2002
Synthesis of 11-Hydroxyl O-Methylsterigmatocystin and the
Role of a Cytochrome P-450 in the Final Step of Aflatoxin
Biosynthesis
Daniel W. Udwary, Linda K. Casillas, and Craig A. Townsend*
Contribution from the Department of Chemistry, The Johns Hopkins UniVersity,
3400 North Charles Street, Baltimore, Maryland 21218
Received September 17, 2001
Abstract: The major skeletal rearrangements (anthraquinone f xanthone f coumarin) that occur in the
complex biosynthesis of aflatoxin B1 are mediated by cytochromes P-450. Previous experiments have
suggested that two successive monooxygenase reactions are required to convert the xanthone O-
methylsterigmatocystin (OMST) to aflatoxin, a process we demonstrate is mediated by a single P-450,
OrdA, in Aspergillus parasiticus in accord with findings in A. flavus. The first oxidative cycle is proposed to
result in the formation of 11-hydroxy O-methylsterigmatocystin (HOMST), while the second entails aryl
ring cleavage, demethylation, dehydration, decarboxylation, and rearrangement to give aflatoxin - a
remarkable sequence of transformations. To test this hypothesis, HOMST has been synthesized by an
alkylnitrilium variant of the Houben-Hoesch reaction. The troublesome xanthone carbonyl was protected
as a butylene to allow further elaboration of the molecule, and then the product xanthone was restored in
a uniquely facile peracid deprotection. Methods were devised to construct the sensitive dihydrobisfuran
and to maintain the oxidation state of the partially methylated hydroquinone. Expression of ordA in a yeast
membrane preparation enabled the intermediacy of HOMST both to be detected in the conversion of OMST
to aflatoxin and to be established directly in the biosynthesis of the mycotoxin. Having secured the role of
HOMST in aflatoxin formation, the mechanism of the second oxidative cycle of this P-450 is considered.
Introduction
oxidative cleavage of the xanthone A-ring, O-demethylation,
dehydration, decarboxylation, and rearrangement (vide infra and
Aflatoxin B1 (4, AFB1) is a widespread contaminant of
foodstuffs, especially in equatorial parts of the world where it
has been correlated to an increased incidence of liver and kidney
tumors.1-3 This environmental carcinogen is produced by strains
of the fungal genus Aspergillus and is notable for its complex
biosynthesis.4 In particular the post-polyketide synthase steps
are marked by skeletal rearrangements in which the first-formed
anthraquinone, norsolorinic acid (1), is transformed to the
dihydrobisfuran-containing versicolorin A (2), which then is
oxidatively rearranged to the xanthone demethylsterigmatocystin
(3, R ) R′ ) H).5 After stepwise O-methylations to O-meth-
ylsterigmatocystin (3, OMST; R ) R′ ) Me),4,6,7 recent genetic
evidence suggests that a single cytochrome P-450 converts this
penultimate intermediate to AFB1, a process involving net
Scheme 1).8
Insight into the latter coumarin-forming process can be
gathered from a diverse set of observations. First, deuterium
from isotopically labeled sodium acetate was found in expected
positions in the A-ring of sterigmatocystin (3, ST; R ) Me, R′
) H),9 but on adjacent carbons in the cyclopentenone ring of
AFB1 (4, Scheme 1).10 It can be deduced from these separate
experiments that epoxidation of the A-ring of 3 could lead to a
NIH shift and formation of the previously unknown 11-hydroxy
OMST (6, HOMST; Scheme 2). Second, consistent with
oxidative aromatic cleavage at this site in the A-ring of 3, 18O2
(O*) was found to be incorporated at C-1 and elsewhere in
AFB1.11 Third, the identity and oxidation state of the carbon
lost in the overall process of coumarin formation were shown
to be specifically C-11 (•, Scheme 1), which was trapped in
the correct stoichiometry exclusively as carbon dioxide.12 For
this oxidation state to be achieved, two cycles of monooxygenase
chemistry must be involved. Finally, while the number of
* To whom correspondence should be addressed. Tel: 410.516.7444.
Fax: 410.261.1233. E-mail: Townsend@jhunix.hcf.jhu.edu.
(1) Dickens, J. W. In Mycotoxins in Human and Animal Health; Rodricks, J.
V., Hesseltine, C. W., Mehlman, M. A., Eds.; Pathotox Publishers: Park
Forest South, IL, 1977; pp 99-105.
(2) Groopman, J. D.; Cain, L. G.; Kensler, T. W. Crit. ReV. Toxicol. 1988, 19,
113-145.
(3) Eaton, D. L.; Gallagher, E. P. Annu. ReV. Phamacol. Toxicol. 1994, 34,
135-172.
(8) Prieto, R.; Woloshuk, C. P. Appl. EnViron. Microbiol. 1997, 63, 1661-
1666.
(4) Minto, R. E.; Townsend, C. A. Chem. ReV. 1997, 97, 2537-2555.
(5) Watanabe, C. M. H.; Townsend, C. A. J. Am. Chem. Soc. 1998, 120, 6231-
6239.
(9) Sankawa, U.; Shimada, H.; Kobayashi, T.; Ebizuka, Y.; Yamamoto, Y.;
Noguchi, H.; Seto, H. Heterocycles 1982, 1053-1058.
(10) Simpson, T. J.; deJesus, A. E.; Steyn, P. S.; Vleggaar, R. J. J. Chem. Soc.,
Chem. Commun. 1983, 338-340.
(6) Keller, N. P.; Dischinger, H. C.; Bhatnagar, D.; Cleveland, T. E.; Ullah,
A. H. J. Appl. EnViron. Microbiol. 1993, 59, 479-484.
(7) Bhatnagar, D.; Ullah, A. H.; Cleveland, T. E. Prep. Biochem. 1988, 18,
321-349.
(11) Watanabe, C. M. H.; Townsend, C. A. J. Org. Chem. 1996, 61, 1990-
1993.
(12) Chatterjee, M.; Townsend, C. A. J. Org. Chem. 1994, 59, 4424-4429.
9
5294
J. AM. CHEM. SOC. 2002, 124, 5294-5303
10.1021/ja012185v CCC: $22.00 © 2002 American Chemical Society

Synthesis of 11-Hydroxyl O-Methylsterigmatocystin
A R T I C L E S
Scheme 1
Scheme 3
Scheme 2
mechanistic scheme. In this paper, we describe the total synthesis
of this putative intermediate by extension of our previously
described N-alkyl nitrilium route to highly substituted xantho-
nes.13 To provide a biochemical system to examine the multifold
reaction of this P-450 in isolation from the other proteins of
the aflatoxin biosynthetic pathway, we have cloned the ord1
analogue from A. parasiticus, ordA, and achieved its functional
expression in yeast.
Results
Synthesis of 11-Hydroxy O-Methylsterigmatocystin. Syn-
thesis of the fused dihydrobisfuran-containing intermediates of
AFB1 biosynthesis requires either early installation of the bicycle
in an aryl precusor prior to xanthone formation or its construc-
tion subsequently. The lability of the bisfuran (a masked
dialdehyde) to acidic and basic reaction conditions thwart the
former approach. Alternatively, assembly of the dihydrobisfuran
after xanthone formation must contend with the substantial
electrophilicity of the xanthone carbonyl.13 We have devised a
general solution to the preparation of highly substituted xan-
thones which first takes advantage of the reactivity of N-alkyl
nitrilium salts as 13 (Scheme 3) in a Friedel-Crafts sense to
give xanthones as 12.13 Second, protection of the xanthone
carbonyl as its corresponding butylene 11 can be simply
achieved by reaction with n-butyllithium. A wide variety of
transformations can then be carried out to elaborate the desired
substitution followed by a uniquely facile deprotection to yield
the xanthone product. The retrosynthesis of 11-hydroxy O-
methylsterigmatocystin (6, HOMST) is outlined in Scheme 3.
Considerable experimentation led to the choice of benzonitrile
18 (Scheme 4) as the nitrilium precursor, which was prepared
in five steps from commercially available acetophenone 15 in
44% overall yield. Baeyer-Villiger oxidation, hydrolysis of the
resulting acetate, and SEM protection afforded 16 (87% yield)
by adaptation of a published method.14 Directed metalation of
16 at low temperature occurred specifically at C-3 and allowed
the introduction of nitrilium precursors from, for example, an
alkyl isocyanate, or, finally, directly from phenylisocyanate.
Attempted generation of the nitrilium salt 13 (Scheme 3) from
imidoyl chloride 1915 and reaction with 14 proceeded poorly.
Similarly, von Braun dehydration16 of 20 failed presumably
enzymes required for the conversion of OMST (3, R ) R′ )
Me) to AFB1 (4) was initially thought to be larger, the
remarkable finding of Prieto and Woloshuk8 in the related
organism A. flaVus clearly pointed to a single cytochrome P-450
carrying out this multistep oxidative cleavage and molecular
reorganization. To accommodate all of these observations, we
have proposed the following mechanism to account for this
remarkable sequence of events as shown in Scheme 2.4,11
The first cycle of P-450 mediated oxidation is proposed to
give aryl epoxide 5, which would give the previously suggested
NIH shift to 11-hydroxy OMST (6).10 The second cycle of
oxidation can be visualized to occur in several ways. One of
these (Scheme 2) involves further epoxidation to 7 and
rearrangement to the oxepin 8, which would rapidly tautomerize
to the hydrolytically unstable 7-membered lactone 9. Opening
of the lactone, demethylation, reclosure, and decarboxylation
in some unspecified order can be invoked to give AFB1 (4).
The intermediacy of HOMST (6) is central to this proposed
(13) Casillas, L. K.; Townsend, C. A. J. Org. Chem. 1999, 64, 4050-4059.
(14) Cantrell, A. S.; Engelhardt, P.; Hogberg, M.; Jaskunas, S. R.; Johansson,
N. G.; Jordan, C. L.; Kangasmetsa, J.; Kinnick, M. D.; Lind, P.; Morin, J.
M., Jr.; Muesing, M. A.; Noree´n, R.; Oberg, B.; Pranc, P.; Sahlberg, C.;
Ternansky, R. J.; Vasileff, R. T.; Vrang, L.; West, S. J.; Zhang, H. J. Med.
Chem. 1996, 39, 4261-4274.
(15) Mills, J. E.; Cosgrove, R. M.; Shah, R. D.; Maryanoff, C. A.; Paragamian,
V. J. Org. Chem. 1984, 49, 546-547.
(16) Ketcha, D. M.; Gribble, G. W. J. Org. Chem. 1985, 50, 5451-5457.
9
J. AM. CHEM. SOC. VOL. 124, NO. 19, 2002 5295

A R T I C L E S
Udwary et al.
Scheme 4
Scheme 6
Scheme 5
However, both phenols could be readily silylated, and, owing
to participation by the adjacent carbomethoxy group, the
TBDMS at the C-3 hydroxyl could be rapidly and specifically
removed by treatment with HF (95% yield). The C-3 hydroxyl
was methoxymethylated to 23 in essentially quantitative yield.
Prior to construction of the dihydrobisfuran, the xanthone
carbonyl was masked as an isomeric mixture of butylenes 24
by the low-temperature addition of butyllithium and dehydration
induced during silica gel chromatography.13
With the substituted xanthone completely protected, the
methyl ester 24 was converted to the corresponding aldehyde
25 by low-temperature DIBAL reduction followed by TPAP/
NMO oxidation20 in excellent yield over the two steps. The
Martin geminal disubstitution protocol21,22 is well-suited for
assembly of the required carbon skeleton and proceeded in
75-80% yield to 26. Treatment of 26 with TIPS triflate
according to the method of Whittamore et al.22,23 smoothly
afforded the mixed acetal 27, which was directly reduced to
aldehyde 28 with DIBAL (90% yield).
In the presence of the unprotected aldehyde, 28 was treated
with m-CPBA to selectively restore the xanthone carbonyl in
29 as the only detectable product, attesting to the usefulness of
the butylene protection/deprotection tactic. As anticipated,22
fluoride-mediated desilylation afforded the hemiacetals 30,
which were converted to the thiophenyl acetals 31.24 Oxidation
with m-CPBA at -45 °C gave a diastereotopic mixture of sul-
foxides, which was pyrolyzed in a preheated bath (110-120
°C) to generate (()-HOMST (6). The total synthesis of 6
required 14 linear steps from arylnitrile 18 and was completed
in >13% overall yield (Scheme 6).
owing to decomposition of the SEM group. Metallation of 16
followed by treatment with phenylcyanate17 and introduction
of the pivaloyl group, however, gave benzonitrile 18 in >75%
yield. These reactions were readily scalable to allow 10 g
quantities of this key starting material to be prepared.
Pivaloyl protection rendered the phenol group more electron-
withdrawing and, therefore, enhanced the electrophilicity of the
nitrilium salt 13, which was most efficiently prepared by SbCl5
reaction with the aryl nitrile 18 and 2-chloropropane. When the
nitrilium salt of 18 (13) and 14 (Scheme 5) were reacted in a
1:1 ratio, isopropylimine 21 was obtained in about 50% yield
accompanied by approximately 10% of this product lacking the
pivaloyl group. Reasoning that loss of this protecting group
would slow the rate of electrophilic aromatic substitution by
the nitrilium ion and itself lead to side reactions, the ratio of
the reaction partner 14 was increased (14:18, 2.5:1), and the
time of reaction was halved to now afford 21 in 90% yield and
<5% of the depivaloyated material. Alkaline hydrolysis readily
afforded the xanthone 12 (Scheme 5).18,19
Expression of ordA in Yeast. The A. parasiticus ordA gene,
identified previously by Bhatnagar,25 and independently in our
group (Minto, R. E.; Townsend, C. A., unpublished), was
(20) Ley, S. V.; Norman, J.; Griffith, W. P.; Marsden, S. P. Synthesis 1994,
639-666.
(21) Phillips, G. W.; Martin, S. F.; Puckette, T. A.; Colapret, J. A. J. Am. Chem.
Soc. 1980, 102, 5866-5872.
The sparingly soluble 12 could not be selectively O-protected
with MOMCl at C-3 or by TBDMSCl at either C-3, or C-5.
(22) Graybill, T. L.; Casillas, E. G.; Pal, K.; Townsend, C. A. J. Am. Chem.
Soc. 1999, 121, 7729-7746.
(23) Townsend, C. A.; Whittamore, P. R. O.; Brobst, S. W. J. Chem. Soc., Chem.
Commun. 1988, 726-728.
(17) Murray, R. E.; Zweifel, G. Synthesis 1980, 150-151.
(18) Schmittling, E. A.; Sawyer, J. S. Tetrahedron Lett. 1991, 32, 7207-7210.
(19) Schmittling, E. A.; Sawyer, J. S. J. Org. Chem. 1993, 58, 3229-3230.
(24) Ley, S. V.; Santafianos, D.; Blaney, W. M.; Simmonds, M. S. J. Tetrahedron
Lett. 1987, 28, 221-224.
9
5296 J. AM. CHEM. SOC. VOL. 124, NO. 19, 2002

Synthesis of 11-Hydroxyl O-Methylsterigmatocystin
A R T I C L E S
amplified by PCR from wild-type A. parasiticus cDNA, and
ligated into the multiple cloning site of the commercially
available S. cereVisiae expression vector pYES2 (Invitrogen),
yielding the final ordA expression vector pYCapoKX. Trans-
formation into a suitable host, Invitrogen INVSc2 S. cereVisiae
cells, was confirmed by reextracting plasmid DNA from yeast
transformants, and retransforming and screening E. coli. The
yeast transformants were best maintained on minimal medium
plates supplemented with glucose, which represses expression
of ordA under the control of the pYES2 GAL1 promoter,26,27
but conditionally selects for uracil utilization by complemen-
tation with the pYES2 ura3 selection marker.28
shortened form of the gene, which resulted from deletion of
amino acids 3-20, was amplified by PCR and ligated into
pYES2. When induced under identical conditions to those
described above for wild-type OrdA, we found that the modified
protein continued to catalyze the conversion of OMST to AFB1,
but the enzyme activity was exclusively membrane associated,
presumably owing to other membrane-binding residues, which
has since been documented.31
OrdA-Mediated Conversion of OMST and 11-Hydroxy
OMST to AFB1. Small-scale expression was used to screen
pYCapoKX-transformed yeast colonies, or otherwise verify
OrdA activity before large-scale expressions. Larger scale cell-
free extracts were used for the purification or analysis of
potential intermediates in the OMST to AFB1 conversion. We
found that by using this method we were able to retrieve more
of the final product and untransformed starting compound than
by simply administering the synthetic compounds to cultures
of whole cells.
As expected, OrdA appeared firmly membrane-bound. When
induced yeast cells transformed with pYCapoKX were disrupted,
activity could only be found in the membrane fraction. To rule
out the necessity of insoluble coenzymes, inactive yeast
membranes from pYES2 transformants were mixed with the
soluble fraction from a cell-free preparation. If OrdA were
soluble or partially soluble, it could function with the hypotheti-
cal membrane-bound yeast coenzyme. Only the induced and
transformed yeast cell membranes exhibited the ability to
convert OMST to AFB1, and activity was relatively high as
compared to that of whole cells, although only for 24-48 h.
The addition of NADPH to the cell-free system did not
observably improve enzyme activity. This implies that all
necessary coenzymes (specifically cytochrome P450 reductase39)
and cofactors (NADPH, FAD, and FMN) are sufficiently bound
to the yeast cell membrane or to other membrane-bound
enzymes after lysis to sustain the observed chemistry.
By using this cell-free extract, we were able to duplicate in
A. parasiticus the observed production of AFB1 from OMST
in A. flaVus,8 and we have extended it to monitor AFB1
production from synthetic HOMST. Moreover, with the added
sensitivity of the cell-free system, we have been able to monitor
transient production of HOMST from OMST. These results are
shown in Figure 1. To control for degradative side products
that might occur during the incubation, we placed synthetic
standards of OMST, HOMST, and AFB1 in buffer under
conditions identical to the yeast incubations (lanes 1, 2, and 5).
We observed little decomposition of OMST and AFB1, while
HOMST was less stable. After incubation of OMST with ordA-
expressing yeast (lane 3), the unmistakable bright blue fluo-
rescence corresponding to AFB1 appeared at the appropriate Rf.
Moreover, we observed the formation of a second compound
corresponding to HOMST, strongly supporting its intermediacy
in this conversion. When HOMST was incubated with ordA-
expressing yeast (lane 4), again production of AFB1 was seen.
Conversely, when OMST and HOMST were incubated with
pYES2 (vector alone) transformants (lanes 6 and 7), no AFB1
production was observed.40
Expression of ordA was achieved by growth of transformants
in YNB medium containing 2% galactose as the sole carbon
source to induce the GAL1 promoter. We attempted to quantify
the amount of OrdA produced by two methods. Induced transfor-
mants were disrupted with glass beads, and after analysis of
the soluble extract by SDS-PAGE we could see no induction
of the 60 kD OrdA. Eukaryotic P-450s are usually tightly mem-
brane bound29 owing to an N-terminal membrane anchor se-
quence (also present in OrdA), and so we assumed that the pro-
tein remained bound to the membrane in its yeast host. Second,
we attempted a carbon monoxide binding assay. This is a well-
known technique for detection of P-450s,30 whereby CO gas is
bubbled through a protein extract or live cell culture, binding
irreversibly to the heme iron and producing a characteristic
absorption at 450 nm, which correlates with P-450 concentration.
However, P-450 expression could not be detected above
background levels of absorption in our host by this technique.
The sole means of detection of OrdA was by catalytic activity.
Incubation of small amounts of OMST with live yeast trans-
formants grown in galactose produced aflatoxin B1, the blue
fluorescent signature of which was unambiguously detectable
by TLC. The same host cells transformed with empty pYES2
vector did not produce aflatoxins, nor did either of these
transformants when grown in a glucose-containing medium,
repressing expression under control of the GAL1 promoter.
To facilitate purification of OrdA we attempted to overpro-
duce an N-terminally truncated and soluble protein, an approach
that has been used successfully in the past.31,32 The location
and extent of the N-terminal membrane anchor in OrdA was
identified by sequence alignment with other P-450s for which
this information is known,33-36 and by comparing secondary
structure predictions37,38 of eukaryotic P-450 N-termini. A
(25) Yu, J.; Chang, P.; Ehrlich, K. C.; Cary, J. W.; Montalbano, B.; Dyer, J.
M.; Bhatnagar, D.; Cleveland, D. E. Appl. EnViron. Microbiol. 1998, 64,
4834-4841.
(26) Giniger, E.; Barnum, S. M.; Ptashe, M. Cell 1985, 40, 767-774.
(27) West, R. W. J.; Yocum, R. R.; Ptashe, M. Mol. Cell. Biol. 1984, 4, 2467-
2478.
(28) Guthrie, C.; Fink, G. R. Guide to Yeast Genetics and Molecular Biology;
Academic Press: San Diego, CA, 1991; Vol. 194.
(29) Ortiz de Montellano, P. R., Ed. Cytochrome P450: Structure, Mechanism,
and Biochemistry, 2nd ed.; Plenum Press: New York, 1995.
(30) Stansfield, I.; Kelly, S. L. Purification and Quantification of Saccharomyces
cereVisiae Cytochrome P450; Humana Press Inc.: Totowa, NJ, 1996; Vol.
53.
(31) Cosme, J.; Johnson, E. F. J. Biol. Chem. 2000, 275, 2545-2553.
(32) Peng, H. M.; Raner, G. M.; Vaz, A. D.; Coon, M. J. Arch. Biochem. Biophys.
1995, 318, 333-339.
(33) Williams, P. A.; Cosme, J.; Sridhar, V.; Johnson, E. F.; McRee, D. E. Mol.
Cell. 2000, 5, 121-131.
(34) Gillam, E. M.; Baba, T.; Kim, B. R.; Ohmori, S.; Guengerich, F. P. Arch.
Biochem. Biophys. 1993, 305, 123-131.
(37) Kneller, D. G.; Cohen, F. E.; Langridge, R. J. Mol. Biol. 1990, 214, 171-
182.
(38) Rost, B. Methods Enzymol. 1996, 266, 525-539.
(39) van den Brink, H. J. M.; van Gorcom, R. F. M.; van den Hondel, C. A. M.
J. J.; Punt, P. J. Fungal Genet. Biol. 1998, 23, 1-17.
(35) Gillam, E. M. J.; Guo, Z.; Guengerich, F. P. Arch. Biochem. Biophys. 1994,
312, 59-66.
(36) Pernecky, S. J.; Olken, N. M.; Bestervelt, L. L.; Coon, M. J. Arch. Biochem.
Biophys. 1995, 318, 446-456.
9
J. AM. CHEM. SOC. VOL. 124, NO. 19, 2002 5297

A R T I C L E S
Udwary et al.
Figure 1. Thin-layer chromatographic analysis of the conversion of OMST and HOMST into AFB1 under (A) long wavelength UV light, and (B) short
wavelength UV light. Lane: (1) OMST standard; (2) HOMST standard; (3) OMST incubated with pYCapoKX-transformant cell free extract; (4) HOMST
incubated with pYCapoKX-transformant cell free extract; (5) AFB1 standard; (6) OMST incubated with pYES2-transformant cell free extract; (7) HOMST
incubated with pYES2-transformant cell free extract.
These findings were confirmed by HPLC, ELISA, or MALDI-
MS experiments. Small amounts of AFB1 produced from OMST
or HOMST were detectable and separable from yeast metabo-
lites by reverse phase HPLC, and the characteristic AFB1 UV
signature was observed. In addition, the putative AFB1-
containing fractions were collected, lyophilized, and confirmed
by TLC. While readily detectable by fluorescence quenching
on TLC, HOMST could not be observed as an intermediate in
the OMST to AFB1 reaction by HPLC, presumably because the
transient amounts produced were simply below the limits of
UV detection. It was detectable by MALDI-MS, however, after
partial purification by preparative TLC to give an M + H ion
at m/z 355.24, while an authentic sample of HOMST gave the
corresponding ion at m/z 355.23. The results were further
confirmed with a commercially available competitive-direct
enzyme linked immunosorbent assay (CD-ELISA). We found
the test kit to work well at aflatoxin levels as low as 10 ppb,
and it uses a simple color test that did not significantly cross-
react with OMST, HOMST, or other aflatoxin biosynthetic
precursors available to us. Fluorescence, however, remains the
best technique for sensitive detection of AFB1, with an analytical
limit on TLC in our hands as low as 0.3 ng.
Substrate Specificity of OrdA. Bhatnagar, Cleveland, and
Kingston have previously shown that a partially purified enzyme
complex was able to convert both OMST (3, R ) R′ ) Me,
Scheme 1) and dihydro OMST (32, DHOMST) to AFB1 (4)
and AFB2 (33), respectively, and that OMST was the preferred
substrate for the reaction.41 With our washed membrane system
Figure 2. Conversion of DHOMST 32 and DHHOMST 34 to AFB2 33.
(A) TLC analysis of chloroform extracts of OrdA-active and -inactive
yeast cell extracts. Lane: (1) Inactive yeast, no metabolites added;40 (2)
active yeast, no metabolites added;40 (3) inactive yeast incubated with
AFB1; (4) active yeast incubated with AFB1; (5) inactive yeast incu-
bated with ST; (6) active yeast incubated with ST; (7) inactive yeast
incubated with DHHOMST; (8) active yeast incubated with DHHOMST
(AFB2 has a slightly lower Rf than does AFB;43 (9) inactive yeast incu-
bated with DHOMST; (10) active yeast incubated with DHOMST. (B) Agri-
Tox CD-ELISA screening kit results. Blue color indicates no aflatoxins;
pink indicates the presence of aflatoxins. Lanes correspond to those in (A),
and where control lanes are: (a) no aflatoxin; (b) 100 microliter 20 ppb
AFB1.
we had the opportunity to evaluate whether the single expressed
P-450 was able to accept both dihydro- and tetrahydrobisfuran
substrates, or if another gene product of the biosynthetic cluster
was necessary for the reactions to proceed.
Dihydro-O-methylsterigmatocystin (32, DHOMST) and di-
hydro-11-hydroxy-O-methylsterigmatocystin (34, DHHOMST)
were prepared from synthetic OMST and HOMST, respectively,
by hydrogenation over Pd/C.42 After incubation with cultures
of pYCapoKX transformants grown in galactose, we found that
AFB2 was produced in a yield suitable for analysis by TLC
and ELISA (Figure 2, lanes 8 and 10). While the amount of
DHHOMST (34) available was limited, conversion to AFB2 by
the active cells was clearly detectable relative to the inactive
control (lanes 7 and 8) and particularly in the CD-ELISA test
(cf. lanes 7, 8, and b). As expected, sterigmatocystin (3, R )
Me, R′ ) H) was not observably consumed and produced no
aflatoxins (lanes 5 and 6). Furthermore, after identical incuba-
tion, AFB1 was not reduced to AFB2 by either the P-450 or the
(40) It should be noted that the pale spots seen in the photograph at an Rf similar
to AFB1 were clearly visible to the eye as pale yellow metabolites that
appeared in every yeast/chloroform extract, but were only visible when
not masked by the bright fluorescence of AFB1.
(41) Bhatnagar, D.; Clevelend, T. E.; Kingston, D. G. I. Biochemistry 1991,
30, 4343-4350.
(42) Chen, P. N.; Kingston, D. G. I.; Vercellotti, J. R. J. Org. Chem. 1977, 42,
359-3605.
9
5298 J. AM. CHEM. SOC. VOL. 124, NO. 19, 2002

Synthesis of 11-Hydroxyl O-Methylsterigmatocystin
A R T I C L E S
Scheme 7
yeast host cells (lanes 3 and 4). These results confirm previous
findings that the latter stages of the aflatoxin biosynthetic
pathway, specifically the steps beyond the “branch point” at
which the tetrahydrobisfuran versicolorin B is oxidized to
versicolorin A (2),41,43,44 are controlled by the same enzymes,
which only discriminate between the dihydro- and tetrahydro-
bisfuran-containing late intermediates to the extent of small
kinetic differences in reaction rate.
Discussion
The formation of aflatoxin takes place through an unusually
long and complex sequence of biochemical events. In particular,
the nuclear reorganization of the initially formed anthraquinone
(see Scheme 1) to xanthone intermediates and, finally, the
coumarin of aflatoxin itself are all oxidative. Cytochrome P-450
enzymes mediate these rearrangements, which are exemplified
in the biosynthesis of other fungal acetogenins suggesting a
wider synthetic role for proteins of this superfamily.
As discussed at the outset, evidence for an apparent NIH shift
and the stoichiometric loss of C-11 from OMST (3, R ) R′ )
Me; Scheme 2) as carbon dioxide point to two sequential
monooxygenase steps in the formation of aflatoxin. We have
demonstrated that this extensive structural change is catalyzed
by OrdA, a single P-450 isolated from the A. parasiticus
aflatoxin biosynthetic gene cluster in keeping with experiments
in A. flaVus.8 We have proposed the intermediacy of HOMST
(6, Scheme 2) as the product of the first monooxygenase
reaction.4 It is likely formed through the proposed epoxide 5 to
account for the evident NIH shift.10
To establish the role of this proposed intermediate in the final
steps to AFB1, HOMST was prepared by total synthesis. This
molecule contains a sensitive dihydrobisfuran, functionally a
masked dialdehyde, and a partially methylated hydroquinone.
These reactive structural elements were set in place in good
overall yield using a general method for the synthesis of highly
substituted xanthones recently developed in this laboratory
(Schemes 4-6).13 Specifically, conversion of an arylnitrile to
its N-isopropylnitrilium salt with SbCl5 and 2-chloropropane
followed by Houben-Hoesch reaction with an appropriately
substituted aryl partner proceeded in high yield to a xanthone
intermediate, whose carbonyl was protected as a butylene by
addition of n-butyllithium. With the xanthone nucleus stabilized,
the skeleton of the dihydrobisfuran was constructed, and the
carbonyl was then restored by exceedingly facile peracid
oxidation. With a sample of HOMST (6) in hand, a yeast
membrane preparation containing expressed OrdA was found
to convert OMST (3, R ) R′ ) Me) to AFB1 (4) and to
accumulate transient amounts of HOMST as secured by
chromatographic comparisons and MALDI-MS. Similarly,
HOMST could be independently converted to AFB1 in parallel
with OMST itself. As anticipated, the tetrahydrobisfuran-
containing analogues of these metabolites 32 and 34, respec-
tively, were converted to AFB2 (33, Scheme 7).
Scheme 8
hydroxyl radical (Scheme 8). Recent experiments with radical
clocks and the intervention of cationic rearrangements, however,
have complicated the mechanistic interpretation of these oxida-
tions.45,46 In parallel to the electrophilic mechanisms that have
been advanced to account for these findings, evidence has been
gathered to suggest that aldehydes can undergo nucleophilic
reaction with iron-peroxy intermediates to lead on to deformy-
lation reactions during product formation.47-49
The overall mechanism suggested above in Scheme 2
hypothesizes two electrophilic reactions (epoxidation) mediated
by OrdA to cleave the xanthone A-ring. C-11 is ultimately lost
in a subsequent ionic step in this mechanism as carbon dioxide.
The demonstrated intermediacy of HOMST (6) limits the
possible mechanisms of the second oxidation to initiate the
cleavage and rearrangement to AFB1 (cf. Scheme 2). One might
object to the necessity in Scheme 2 to epoxidation in two
regiochemical senses in the A-ring to achieve proper 18O
labeling and loss of C-11. An alternative suggestion can be
advanced (but unfortunately not proved) that, after the first
electrophilic oxidation to 5 and NIH shift to HOMST (6),
nucleophilic attack by a iron-peroxy species 35 on the keto
tautomer of 6 would involve little motion of the substrate relative
to the iron center and lead directly to Baeyer-Villiger-like
rearrangement to the seven-membered lactone 9 (Scheme 9).
Moreover, by avoiding electrophilic reaction to an epoxide, the
propensity to quinone formation can be circumvented. Opening
of the lactone by water present in the active site, or by ferric
hydroxyl/water formed in the reaction, or conceivably by an
enzyme X-group to 37 (X ) OH, X) can be visualized.
Decarboxylation yields a stabilized anion, which can close to
38. â-Elimination from the enol would afford 39, which upon
loss of methanol from the labile vinyl ether can be hypothesized
For quite some time a consensus view of the cytochrome
P-450 hydroxylation reaction has invoked hydrogen abstraction
from substrate by a heme iron-oxo species and “rebound” to
product by recombination of the alkyl radical and iron-bound
(45) Newcomb, M.; Shen, R.; Choi, S.; Toy, P.; Hollenberg, P.; Vaz, A.; Coon,
M. J. Am. Chem. Soc. 2000, 122, 2677-2686.
(46) Collman, J. P.; Chien, A. S.; Eberspacher, T. A.; Brauman, J. I. J. Am.
Chem. Soc. 2000, 122, 11098-11100.
(43) McGuire, S. M.; Brobst, S. W.; Graybill, T. L.; Pal, K.; Townsend, C. A.
J. Am. Chem. Soc. 1989, 111, 8308-8309.
(44) Yabe, K.; Ando, Y.; Hamasaki, T. Agric. Biol. Chem. 1991, 55, 1907-
1911.
(47) Cole, P. A.; Robinson, C. H. J. Am. Chem. Soc. 1988, 110, 1284-1285.
(48) Aktar, M.; Wright, J. N. Nat. Prod. Rep. 1991, 8, 527-555.
(49) Kuo, C.-L.; Raner, G. M.; Vaz, A. D. N.; Coon, M. J. Biochemistry 1999,
38, 10511-10518.
9
J. AM. CHEM. SOC. VOL. 124, NO. 19, 2002 5299

A R T I C L E S
Udwary et al.
Scheme 9
29.9 ppm), or CD2Cl2 (5.32 and 54.0 ppm) as indicated by the indi-
vidual experiments. High- and low-resolution mass spectra were
recorded using a VG Instruments 70-S 250 GC/MS at 70 eV in
EI+, CI+, or FAB operating modes. MALDI-TOF data were ob-
tained on a Bruker BIFLEX III. Infrared spectra were recorded on a
Perkin-Elmer 1600 Series FTIR spectrophotometer. Combustion analy-
ses were performed by Atlantic Microlab, Inc. (Norcross, GA). Flash
chromatography was performed on EM silica gel 60 (230-400 mesh).
A ratio of 25-100:1 silica gel/crude product by weight and flow rates
of 1-2 in/min were normally employed for flash columns. Thin-layer
chromatography was performed on Analtech Uniplate glass plates
containing fluorescent indicator, visualized by UV and 20% ethanolic
solution of phosphomolybdic acid reagent and photographed with a
Kodak DC290 digital camera, or by film photography by Johns Hopkins
University staff. Reagents and AFB1 and AFB2 standards were sup-
plied by Aldrich Chemical Co. (Milwaukee, WI), and all media
components were from Difco (Detroit, MI) or Fisher (Pittsburgh, PA).
All nonaqueous reactions were performed in flame-dried glassware
under a positive atmosphere of dry N2 or Ar. All solvents were dried
and distilled immediately prior to use (THF and Et2O were distilled
from sodium/benzophenone, and CH2Cl2 and CH3CN were distilled
from CaH2). HPLC was performed with a Perkin-Elmer 235C diode
array detector and Series 410 LC pump using a Spherex 5 C18 analytical
column from Phenomenex (Torrance, CA). PCR primers were pur-
chased from Sigma-Genosys (The Woodlands, TX), and sequencing
reactions were performed at the Biosynthesis and Sequencing Facility
at the Johns Hopkins Medical Institute. The pYES2 yeast expression
vector was purchased from Invitrogen (Carlsbad, CA), while the
pT7Blue vector and Perfectly Blunt cloning kit were purchased from
Novagen (Madison, WI). All restriction enzymes were from Stratagene
(La Jolla, CA) or Life Technologies (Grand Island, NY). DNA, mRNA,
and PCR products were purified using kits purchased from Qiagen
(Valencia, CA).
Scheme 10
Fungal and Yeast Strains and Media. The A. parasiticus wild-
type strain SU-1 (ATCC 56775) and the ST producer A. Versicolor
(ATCC 28286) were purchased from the American Type Culture
Collection (Manassas, VA). A. parasiticus and A. Versicolor strains
were plated on potato dextrose agar (PDA) plates,5 which contained
yeast extract, 2.5 g; Bacto-agar, 2.5 g; potato dextrose agar, 19 g; and
distilled water. Liquid cultures of A. parasiticus (500 mL) were grown
in Adye and Mateles (AM) growth medium,50 which contained per L:
sucrose, 50 g; potassium phosphate monobasic, 10 g; ammonium sulfate,
3 g; magnesium sulfate (anhydrous), 1 g; and trace metals, 2 mL. A.
Versicolor strains were grown in CZAPEK medium,5 which contained
per L: sucrose, 30 g; sodium nitrate, 3 g; potassium phosphate
monobasic, 1 g; magnesium sulfate (anhydrous), 0.5 g; potassium
chloride, 0.5 g; and ferrous sulfate, 0.01 g.
to give AFB1 (4). In the steps past lactone 9 we invoked
spontaneous chemical reactions. Interestingly, sterigmatocystin
(3, R ) Me, R′ ) OH) lacking the A-ring methoxy group is
not a substrate for the enzyme. It may be that the presence of
this methoxy group enforces the hydroquinone oxidation state
in ring-A and maintains the syn-geometry in hypothetical
intermediate 37 to favor five-membered ring formation in the
proposed decarboxylative aldol reaction.
Alternatively, the last step after lactone 9 could take place in
a quite different manner in which tautomerization of 37 would
give the stabilized 40 (a doubly vinylogous ester, Scheme 10),
where anion formation adjacent to the carboxylate and intramo-
lecular condensation with loss of water and methanol would
afford 42. Here decarboxylation to 4 would be expected to be
especially facile to give an anion vinylogously delocalized to
both the ketone and the lactone. These speculations await more
detailed examination at the level of purified enzyme to further
dissect the mechanism of this unusual P-450-mediated trans-
formation.
The S. cereVisiae strains INVSc1 and INVSc2 (Invitrogen, Carlsbad,
CA) were used interchangeably as yeast expression hosts. Prior to
transformation, all S. cereVisiae host strains were grown in YPD
medium,51 which contained per L: yeast extract, 10 g; bacto-peptone,
20 g; and dextrose, 20 g. Once transformed, the yeast were grown in
a selective medium containing per L: 1.7 g of yeast nitrogen base [not
containing amino acids or (NH4)2SO4], 5 g of (NH4)2SO4, 20 g of
glucose, and 30 mg each of histidine, tryptophan, and leucine. The
yeast expression medium was similar, but glucose was replaced with
20 g of galactose. Selective plates for yeast transformants consisted of
growth medium supplemented with 15 g/L of bacto-agar.
Preparation of Sterigmatocystin (3, ST; R ) Me, R′ ) OH) and
O-Methylsterigmatocystin (3, OMST; R ) R′ ) OMe). ST was
purified from an A. Versicolor liquid culture on the basis of a method
Experimental Section
General Materials and Methods. Melting points were determined
with a Thomas-Hoover oil bath apparatus in open capillaries and
(50) Adye, J.; Mateles, R. I. Biochim. Biophys. Acta 1964, 86, 418-420.
(51) Ausubel, F. M.; Brent, R.; Kingston, R. E.; Moore, D. D.; Seidman, J. G.;
Smith, J. A.; Struhl, K.; Albright, L. M.; Coen, D. M.; Varki, A. Current
Protocols in Molecular Biology; John Wiley & Sons: New York, 1994;
Vols. 1 and 2.
1
are uncorrected. H and 13C NMR spectra were recorded on a Bruker
AMX-300 (300 MHz) or Varian UnityPlus 400 MHz spectrometer and
are referenced to CDCl3 (7.26 and 77.0 ppm), acetone-d6 (2.04 and
9
5300 J. AM. CHEM. SOC. VOL. 124, NO. 19, 2002

Synthesis of 11-Hydroxyl O-Methylsterigmatocystin
A R T I C L E S
previously reported.5 A fresh stock of A. Versicolor was plated onto
PDA plates and grown in the dark for 5 days until spore growth was
heavy. Spores were collected by suspension in 9 mL of 0.05% Tween
80, and 4 mL was used to inoculate 4 × 1 L CZAPEK medium. The
cultures were grown in the dark at 28 °C on a shaker at 175 rpm for
19 days. Cells were collected by filtration through cheesecloth, frozen
in liquid nitrogen, and steeped in acetone. The flow through was pooled
and extracted with chloroform. All organic extracts were pooled, filtered
through Celite 545 (Fisher), and dried over anhydrous MgSO4. Solvent
was evaporated, the solid residue redissolved in 99:1 CHCl3:MeOH,
and ST was purified by silica column chromatography. OMST was
then quickly and easily prepared from ST by methylation using a
technique previously reported.12
Reverse Transcription of Fungal mRNA. Wild-type A. parasiticus
was grown for 60 h from frozen spore stocks, and collected by filtration,
frozen in liquid nitrogen, and ground to a powder. Total RNA was
purifed from the cells using a Qiagen RNeasy Plant kit, followed by
mRNA purification using a Qiagen Oligotex mRNA kit according to
the manufacturer’s instructions. Reverse transcription was achieved by
combining in a microfuge tube: 5 µL of 10× MMLV RT buffer and
1 µL of RNAse Block (Stratagene, La Jolla, CA), 12 µL of 10 mM
dNTPs (Applied Biosystems, Foster City, CA), 2 µL of poly-dT primer
(Roche Molecular Biochemicals, Indianapolis, IN), 20 µL of mRNA,
and 8.5 µL of distilled deionized water. The mixture was incubated at
room temperature to allow the primer to bind for 15 min, then 1.5 µL
of Stratagene MMLV reverse transcriptase was added, and the mixture
was placed at 37 °C for 1 h. Following the reaction, the mixture was
treated with 1 µL of RNAse (10 mg/mL) for 20 min at room
temperature, and could then be used as template for PCR.
Construction of Expression Vectors. The cDNA prepared above
was used in a PCR with primers DUapoK-1 (5′-GGGGTACCCCAT-
GATTTATAGCATAATTATTTGT-3′; bold ) KpnI restriction site)
and DUapoX-2 (5′-GCTCTAGAGCTCAAATCATCTGATTTCTGGC-
3′; bold ) XbaI restriction site). The 1.5 kb PCR product was purified
with a Qiagen QIAquick gel extraction kit, and ligated into the cloning
vector pT7Blue-3 using a Perfectly Blunt Cloning Kit, yielding the
vector pT7CapoKX. The insert was retrieved by excision with KpnI
and XbaI and ligated into the multiple cloning site of the yeast
expression vector pYES2 (Invitrogen) to construct the vector pYCa-
poKX. To build the expression vector for the N-terminally modified
ordA, pT7CapoKX was used as template with a new 5′ primer,
DUapoDK (5′-GGTACCATGATTCTGCTGGCGCCCAAAGAC-3′;
bold ) KpnI restriction site) and DUapoX-2. The PCR product was
then cloned into pT7Blue-3, the insert was retrieved by digestion with
KpnI and XbaI, and ligated into pYES2, yielding the yeast expression
vector pYCapoDKX.
Expression of ordA in Yeast. Yeast transformation was achieved
using a lithium acetate transformation protocol,52 and was plated onto
selective medium plates. Yeast transformants were picked from plates,
and a seed culture was grown to saturation (A600 ≈ 1.5) in 5 mL of
growth medium over 2 days at 30 °C. Two different production scales
were used to express ordA in yeast. The small-scale expression was
used to screen transformed yeast colonies, or otherwise verify ordA
activity before large-scale expressions. In this method 500 µL of seed
culture was used to inoculate 5 mL of expression medium, supplemented
with 50 µg of OMST or HOMST in a minimal volume of acetone.
The culture was allowed to shake 24 h at 28 °C and 300 rpm, and then
was extracted with CHCl3 and analyzed by TLC for aflatoxin
production. For large-scale expressions 100 mL of expression media
was inoculated with 1 mL of seed culture, fed with 1 mg of OMST or
HOMST in a minimal volume of acetone, and allowed to shake for
24-48 h at 28 °C and at 300 rpm before extraction with CHCl3 and
analysis by HPLC, TLC, or MS.
expression medium (defined above). The cells were collected by
centrifugation, washed with water, and resuspended in a buffer
consisting of 30% glycerol, 50 mM KxPO4 pH 7.4, 100 µM benzami-
dine, and 1 mM EDTA. After disruption with glass beads using a
standard method51 the lysed cells were centrifuged 1 h at 12 000g, the
supernatant was decanted and saved, and the membrane pellet was
washed once and resuspended in buffer.
Synthesis. Full experimental details for intermediates and reagents
shown in Schemes 4-6 not described below can be found in the
Supporting Information.
2-Fluoro-6-methoxy-3-pivaloylbenzonitrile (18). To a suspension
of 2-fluoro-3-hydroxy-6-methoxybenzonitrile (5.63 g, 33.71 mmol, see
Supporting Information) in CH2Cl2 (170 mL, 0.2 M) under argon was
added pivaloyl chloride (6.6 mL, 53.94 mmol) and pyridine (4.9
mL, 60.68 mmol). The resulting solution was stirred 2 h and then
quenched by the addition of saturated NH4Cl. The mixture was ex-
tracted with saturated CuSO4, cold 5% HCl, 5% NaHCO3, and brine.
The organic portion was dried (MgSO4), filtered, and concentrated to
an orange oil, which was triturated with ether and hexane to provide 3
g of pure product. The mother liquor was purified by column
chromatography (10 f 30% EtOAc/hexanes) to give an additional
4.86 g (93% total yield) of 18, mp 73-74°. Rf ) 0.45 (20% EtOAc/
1
hexane). H NMR (400 MHz, CDCl3): δ 7.25 (dd, J ) 9.2, 8.4 Hz,
1H), 6.72 (dd, J ) 9.2, 1.4 Hz, 1H), 3.90 (s, 3H), 1.32 (s, 9H). 13C
NMR (100 MHz, CDCl3): δ 175.9, 159.3 (d, 3.6), 155.3 (d, 259.7),
131.9 (d, 11.4), 129.0 (d, 3.2), 110.7, 106.6 (d, 3.7), 92.6 (d, 15.4),
56.7, 56.6, 39.0, 26.9. IR (CHCl3): 3018, 2972, 2237, 1760, 1493,
1254, 1103 cm-1. MS (EI) m/e (rel inten): 251 (M+, 5), 167 (100),
152 (9). Anal. Calcd for C13H14FNO3: C, 62.15; H, 5.58. Found: C,
62.14; H, 5.56.
2,4-Dihydroxy-6,6′-dimethoxy-2′-fluoro-3-methoxycarbonyl-3′-
pivaloyl-benzophenone-N-(2-methylethyl)ketimine (21). To
a
CH2Cl2 solution (460 mL, 0.06 M) of benzonitrile (18, 7.0 g, 27.89
mmol) and 2-chloropropane (25.5 mL, 278.89 mmol) was added SbCl5
(27.9 mL, 36.26 mmol, 1 M in CH2Cl2) over 8 min. After 40 min,
methyl 2,6-dihydroxy-4-methoxybenzoate (14, 13.81 g, 69.72 mmol)
was added in 80 mL of CH2Cl2 by cannula, and the reaction mixture
was stirred for 2 h at room temperature. The volume was reduced to
200 mL, and the reaction mixture was filtered through a short silica
pad and washed with 5% CH3OH/CH2Cl2 to remove the antimony salts.
Column chromatography (40% f 60% EtOAc/hexanes f 20% CH3-
CN/EtOAc) provided 12.98 g (95%) of 21 as a yellow oil. Rf ) 0.6
(2%MeOH/CH2Cl2). 1H NMR (400 MHz, CDCl3): δ 17.00 (brs, 1H),
13.22 (s, 1H), 7.08 (dd, J ) 8.8, 8.4 Hz, 1H), 6.69 (dd, J ) 9.2, 1.2
Hz, 1H), 5.54 (s, 1H), 4.00 (s, 3H), 3.75 (s, 3H), 3.46 (m, 1H), 3.36 (s,
3H), 1.35 (s, 9H), 1.23 (m, 6H). 13C NMR (100 MHz, CDCl3): δ 178.2,
176.3, 173.3, 172.5, 166.5, 160.8, 153.5 (d, 6.4), 150.0 (d, 247.0), 132.3
(d, 13.7), 123.7, 114.5 (d, 17.8), 105.6 (d, 3.2), 101.6, 98.1, 88.0, 56.2,
55.4, 52.2, 48.3, 39.0, 27.0, 23.3, 23.0. IR (CHCl3): 3032, 2984, 1752,
1630, 1596, 1571, 1550, 1500, 1450, 1109, 1089 cm-1. MS (EI) m/e
(rel inten): 491 (M+, 40), 459 (100), 428 (13), 375 (15), 360 (15), 298
(17). Anal. Calcd for C25H30FNO8: C, 61.10; H, 6.11. Found: C, 59.12;
H, 6.16.
3,5-Dihydroxy-1,8-dimethoxy-4-methoxycarbonyl-9H-xanthen-9-
one (12). A suspension of 21 (1 g, 2.04 mmol), K2CO3 (5.6 g, 40.36
mmol), KF‚Al2O318 (0.33 g, 2.04 mmol), and 18-crown-6 (0.05 g, 0.2
mmol) in 200 mL of dry CH3CN was heated to reflux and stirred for
6 h. A 1:1 mixture of CH3OH/H2O (200 mL) was added, and the
resulting solution was stirred at reflux for 1 h, then overnight at room
temperature. The reaction mixture was concentrated to an orange residue
and dissolved in 100 mL of 1 N HCl and 300 mL of CH2Cl2, and
extracted 3× with 100 mL portions of CH2Cl2. The organic extracts
were dried (MgSO4), filtered, and concentrated under vacuum. Purifica-
tion by flash chromatography (2.5% CH3OH/10% CH3CN/CH2Cl2)
afforded 0.53 g (75%) of 12 as a white solid, mp 244-245 °C (decomp).
Disruption of Yeast Cells. One hundred milliliter cultures of ordA-
expressing yeast and untransformed yeast were grown to saturation in
1
(52) Chen, D. C.; Yang, B. C.; Kuo, T. T. Curr. Genet. 1992, 21, 83-84.
Rf ) 0.2 (2% MeOH/CH2Cl2). H NMR (300 MHz, CDCl3): δ 12.39
9
J. AM. CHEM. SOC. VOL. 124, NO. 19, 2002 5301

A R T I C L E S
Udwary et al.
(s, 1H), 7.18 (d, J ) 8.9 Hz, 1H), 6.70 (d, J ) 8.9 Hz, 1H), 6.38 (s,
1H), 5.81 (brs, 1H), 4.10 (s, 3H), 3.99 (s, 3H), 3.92 (s, 3H). 13C NMR
(100 MHz, DMSO-d6/pyridine-d5): δ 173.6, 168.1, 164.7, 163.7, 156.7,
151.6, 145.3, 139.8, 119.7, 114.4, 107.2, 106.7, 98.1, 95.3, 56.2, 55.8,
52.0. IR (CHCl3): 3534, 3009, 2934, 1652, 1582, 1483, 1291, 1263
cm-1. MS (EI) m/e (rel inten): 346 (M+, 75), 314 (100), 299 (78), 285
(13). Anal. Calcd for C17H14O8: C, 58.96; H, 4.05. Found: C, 58.72;
H, 4.08.
(1.25 g, 2.43 mmol) in 6 mL of dry THF was added to the resulting
lithium benziminomethyl-phosphonate53,54 by cannula over 5 min. After
50 min, the mixture was allowed to stir at room temperature for 1.5 h
and then was cooled to -78 °C, and nBuLi (2.32 mL, 1.47 M, 3.41
mmol) was added over 3 min. After 2.5 h, bromoethyl acetate (1.35
mL, 12.17 mmol) was added to the -78 °C mixture, which was then
allowed to warm to room temperature over 1.5 h. The reaction was
quenched by the addition of 7 mL of 1 M aqueous tartaric acid and
stirred for 1 h. The mixture was extracted with cold 5% HCl (2×), 5%
NaHCO3 (2×), and brine, dried (MgSO4), filtered, and concentrated
under vacuum. The residue was purified by flash chromatography (2
w 40% EtOAc/hexane) to give 1.18 g (79%) of 26 as a yellow oil and
a mixture of E and Z isomers. Rf ) 0.39 (20% EtOAc/hexane). 1H NMR
(300 MHz, CDCl3): δ 9.72 (s), 9.69 (s, 1H), 9.62 (s, 1H), 6.77 (d, J
) 8.8 Hz, 1H), 6.66 (d, J ) 8.8 Hz, 1H), 6.58 (s, 1H), 6.55 (s, 1H),
6.52 (d, J ) 8.8 Hz, 2H), 6.20 (m, 1H), 6.10 (m, 1H), 5.14 (m, 4H),
4.10 (q, J ) 7.2 Hz, 4H), 4.06 (m, 2H), 3.84 (s, 6H), 3.79 (s, 6H),
3.45 (s, 3H), 3.43 (s, 3H), 3.24 (m, 2H), 2.69 (m, 2H), 2.30 (m, 2H),
1.86 (m, 2H), 1.45 (m, 4H), 1.24 (t, J ) 7.1 Hz, 6H), 0.99 (s, 18H),
0.88 (t, J ) 7.3 Hz, 6H), 0.30 (s, 3H), 0.28 (s, 3H), 0.23 (s, 3H), 0.22
(s, 3H). 13C NMR (100 MHz, CDCl3): δ 199.8, 199.7, 171.9, 171.7,
155.9, 155.8, 155.7, 154.8, 154.6, 153.9, 152.5, 152.3, 149.5, 149.4,
145.6, 144.3, 137.8, 137.6, 137.4, 133.6, 133.5, 118.8, 118.6, 118.5,
118.3, 117.3, 117.0, 116.1, 115.6, 111.6, 108.7, 105.9, 105.6, 105.3,
105.2, 94.5, 94.4, 94.0, 93.1, 60.2, 56.2, 56.1, 55.9, 55.7, 55.2, 44.2,
33.2, 33.0, 32.9, 32.8, 25.8, 22.5, 18.3, 14.0, 13.8, -4.1, -4.2, -4.4.
IR (CHCl3): 3012, 2958, 2922, 2851, 1724, 1573, 1492, 1465, 1256,
1099, 1055 cm-1. MS (EI) m/e (rel inten): 614 (M+, 48), 583 (48),
557 (100). HRMS (EI) calcd for C33H46O9Si (M+), 614.2911; found,
614.2918.
1,8-Dimethoxy-5-(t-butyldimethylsilyloxy)-3-hydroxy-4-methoxy-
carbonyl-9H-xanthen-9-one (22). To a suspension of xanthone 12
(1.74 g, 5.03 mmol) in CH2Cl2/CH3CN (3:2, 125 mL, 0.04 M) was
added TBDMSCl (2.27 g, 15.09 mmol) and imidazole (2 g, 29.1 mmol).
After 1 h the reaction mixture was quenched by the addition of cold
5% HCl and extracted with 5% NaHCO3 and brine. The dried (MgSO4)
filtrate was concentrated to ca. 15 mL of CH3CN and treated with 20
drops of 49% aqueous HF. After 30 min, the reaction mixture contained
only the mono-TBS product. The mixture was diluted with CH2Cl2
and extracted with 5% NaHCO3 and brine. After drying (MgSO4) and
filtering, the organic portion was concentrated and purified by column
chromatography (2.5% CH3OH/10% CH3CN/CH2Cl2) to provide 1.77
g (77%) of the product as a tan solid. Recrystallization from hot EtOAc/
CH2Cl2 and hexane afforded 22 as colorless crystals, mp 202-203 °C.
1
Rf ) 0.39 (2.5% CH3OH/10% CH3CN/CH2Cl2). H NMR (400 MHz,
CDCl3): δ 12.24 (s, 1H), 7.08 (d, J ) 8.8 Hz, 1H), 6.63 (d, J ) 9.2
Hz, 1H), 6.35 (s, 1H), 4.02 (s, 3H), 3.96 (s, 3H), 3.90 (s, 3H), 0.97 (s,
9H), 0.21 (s, 6H). 13C NMR (100 MHz, CDCl3): δ 174.9, 170.5, 167.7,
165.3, 158.0, 153.7, 147.8, 137.5, 124.0, 114.7, 108.2, 105.8, 95.6,
95.0, 56.5, 53.1, 23.8, 18.7, -3.9. IR (CHCl3): 3006, 2955, 2925, 2854,
1652, 1581, 1490, 1333, 1111, 1075 cm-1. MS (EI) m/e (rel inten):
460 (M+, 11), 403 (100), 373 (77). Anal. Calcd for C23H28O8Si: C,
60.00; H, 6.09. Found: C, 59.83; H, 6.16.
10-(t-Butyldimethylsilyloxy)-5,7-dimethoxy-1-(ethoxycarbonyl-
methyl)-2-[tris(1-methylethyl)silyloxy)]-6H-furo[2,3-c]xanthene-∆6,δ
-
1,8-Dimethoxy-5-(t-butyldimethylsilyloxy)-4-methoxycarbonyl-3-
O-methoxymethyl-9H-xanthene-∆9,δ-butane (24). To a -82 °C THF
solution (50 mL, 0.04 M) of protected xanthone 23 (0.96 g, 1.91 mmol)
was added 1.9 mL of n-BuLi (2.5 mmol, 1.3 M). After 1 h, 20 mL of
CH3OH and 20 mL of 1 M tartaric acid were added, and the resulting
mixture was stirred for 45 min and warmed to room temperature. The
mixture was diluted with EtOAc, and the two phases were separated.
The aqueous portion was back-extracted with CHCl3 until the aqueous
layer was colorless, and the combined organic layers were washed with
5% NaHCO3 and brine. After drying (MgSO), filtering, and concentrat-
ing to an orange oil, the residue was loaded onto a silica gel column
and eluted with three column volumes of 1% HOAc/CHCl3 to facilitate
the dehydration of the tertiary alcohol. Finally, eluting with 1% HOAc/
5% acetone/CHCl3 provided the pure 24 as a yellow oil (0.95 g, 92%),
as a mixture of (E)- and (Z)-isomers. Rf ) 0.4 (20% EtOAc/hexane).
1H NMR (300 MHz, CDCl3): δ 6.75 (d, J ) 8.9 Hz, 1H), 6.64 (d, J
) 8.9 Hz, 1H), 6.53 (s, 1H), 6.52 (s, 1H), 6.52 (d, J ) 8.8 Hz, 2H),
6.18 (m, 1H), 6.09 (m, 1H), 5.20 (s, 2H), 5.17 (s, 2H), 3.93 (s, 3H),
3.92 (s, 3H), 3.85 (s, 6H), 3.79 (s, 6H), 3.49 (s, 3H), 3.47 (s, 3H), 2.25
(m, 2H), 1.81 (m, 2H), 1.38 (m, 4H), 1.01 (s, 18H), 0.89 (t, J ) 7.3
Hz, 3H), 0.85 (t, J ) 7.3 Hz, 3H), 0.24 (s, 6H), 0.19 (s, 6). 13C NMR
(100 MHz, CDCl3): δ 165.6, 157.1, 157.0, 154.9, 153.6, 152.7, 151.2,
149.6, 149.4, 145.6, 144.3, 137.8, 137.5, 133.9, 133.8, 119.0, 118.3,
118.1, 118.0, 117.6, 115.5, 111.7, 108.8, 107.0, 106.7, 106.0, 105.4,
95.2, 95.1, 94.8, 93.9, 56.3, 56.0, 55.9, 55.5, 55.4, 52.4, 33.3, 33.1,
25.9, 22.7, 18.4, 14.1, 13.9, -4.3, -4.5. IR (CHCl3): 3006, 2956, 2847,
1730, 1575, 1495, 1252, 1084, 1049 cm-1. MS (EI) m/e (rel inten):
544 (M+, 15), 487 (100), 447 (15), 441 (31). HRMS (EI) calcd for
C29H40O8Si (M+), 544.2492; found, 544.2498.
butane (27). Aldehyde 26 (1.03 g, 1.67 mmol) and triethylamine (0.37
mL, 2.68 mmol) were combined in 21 mL of dry THF and cooled to
0 °C. Triisopropylsilyl trifluoromethylsulfonate (0.54 mL, 2.0 mmol)
was added dropwise, and the mixture was stirred for 3 h and warmed
to room temperature. The reaction was quenched by the addition of
N,N-dimethylethanolamine (0.5 mL, 5.02 mmol) and extracted with
cold 5% HCl, 5% NaHCO3, brine, and dried (MgSO4), filtered, and
concentrated under vacuum. The residue was purified by flash chro-
matography (5% ethyl acetate/hexane) to provide 1.06 g of 27 (87%)
as a colorless oil, which was a mixture of four isomers: (E)- and (Z)-
and cis- and trans-acetals, mp 51-54 °C. Rf ) 0.55 (10% EtOAc/
1
hexane). H NMR (400 MHz, CDCl3): δ 6.78 (d, J ) 8.8 Hz, 1H),
6.67 (d, J ) 8.8 Hz, 1H), 6.54 (d, J ) 8.8 Hz, 2H), 6.29 (s, 1H), 6.27
(s, 1H), 6.15 (m, 2H), 5.93 (m, 2H), 4.14 (m, 4H), 3.83 (s, 6H), 3.81
(s, 6H), 3.76 (m, 4H), 3.12 (ddd, J ) 19.6, 16.8, 3.2 Hz, 1H), 2.99
(dd, J ) 16.8, 2.8 Hz, 1H), 2.47 (m, 2H), 2.31 (m, 2H), 1.89 (m, 2H),
1.44 (m, 4H), 1.27 (t, J ) 7.2 Hz, 3H), 1.22 (t, J ) 7.2 Hz, 3H), 1.09
(m, 42H), 1.05 (s, 9H), 1.04 (s, 9H), 0.90 (m, 6H), 0.30 (s, 3H), 0.25
(s, 3H), 0.24 (s, 3H), 0.20 (s, 3H). 13C NMR (400 MHz, CDCl3): δ
171.5, 171.3, 159.3, 159.1, 158.0, 157.9, 157.0, 156.8, 156.7, 151.8,
151.6, 150.3, 149.9, 149.7, 146.0, 145.7, 144.7, 144.4, 137.8, 137.5,
137.2, 132.9, 132.8, 132.7, 132.6, 119.0, 118.9, 118.8, 118.5, 117.6,
117.2, 116.2, 110.4, 107.5, 106.9, 106.8, 106.6, 106.0, 105.9, 105.8,
105.7, 105.2, 105.1, 90.6, 90.4, 89.9, 89.6, 60.4, 56.0, 55.9, 55.5, 55.4,
46.2, 46.1, 45.9, 45.8, 36.0, 35.9, 35.2, 35.1, 33.4, 33.3, 33.2, 25.7,
22.7, 18.2, 18.1, 17.8, 17.7, 14.1, 14.0, 13.9, 12.1, -4.3, -4.4, -4.5,
-4.7. IR (CHCl3): 3006, 2950, 2857, 1726, 1633, 1605, 1488, 1460,
1090 cm-1. MS (EI) m/e (rel inten): 726 (M+, 52), 695 (37), 669
(100). Anal. Calcd for C40H62O8Si2: C, 66.12; H, 8.54. Found: C, 66.10;
H, 8.57.
5-(t-Butyldimethylsilyloxy)-1,8-dimethoxy-4-(ethyl-3′-formylpro-
panoate)-3-O-methoxymethyl-9H-xanthen-∆9,δ-butane (26). To a
cooled (-78 °C) solution of 1-(N-benzylideneamino)aminometh-
ylphosphonate (0.66 g, 2.9 mmol) in 15 mL of dry THF was added
n-butyllithium (2.0 mL, 1.47 M in hexanes, 2.9 mmol) over 5 min.
This solution was stirred at -78 °C for 1 h. A solution of aldehyde 25
(53) Yamauchi, K.; Mitsuda, Y.; Kinoshita, M. Bull. Chem. Soc. Jpn. 1975, 48,
3285-3286.
(54) Ratcliffe, R. W.; Christensen, B. G. Tetrahedron Lett. 1973, 46, 4645-
4648.
9
5302 J. AM. CHEM. SOC. VOL. 124, NO. 19, 2002

Synthesis of 11-Hydroxyl O-Methylsterigmatocystin
A R T I C L E S
1-(Acetaldehyde)-10-(t-butyldimethylsilyoxy)-5,7-dimethoxy-2-
[tris(1-methylethyl)silyloxy)]-6H-furo[2,3-c]xanthen-6-one (29). To
a 3 mL CH2Cl2 solution (0.01 M) of butylene aldehyde 28 (0.18 g,
0.26 mmol) and 4 Å molecular sieves (0.02) was added ∼70% m-CPBA
(0.02 g, 0.10 mmol). After 10 min the reaction was quenched by the
addition of saturated NaHCO3 (2 mL) and diluted with CH2Cl2. The
CH2Cl2 layer was separated and washed with brine, dried (MgSO4),
filtered, and concentrated to a yellow solid. Column chromatography
(2.5% CH3OH/10% CH3CN/CH2Cl2) provided 0.15 g (88%) of the
desired 29 as a white solid, which was recrystallized from hot EtOAc/
CH2Cl2 and hexane, mp 172-174 °C. Rf ) 0.62 (2.5% CH3OH/10%
144.6, 139.1, 139.0, 135.5, 134.9, 133.5, 130.8, 130.0, 129.8, 129.2,
129.0, 128.9, 127.5, 126.7, 126.5, 119.6, 119.5, 114.1, 114.0, 113.9,
113.8, 112.5, 107.9, 107.7, 107.1, 106.8, 106.7, 106.6, 105.7, 104.8,
104.4, 100.2, 99.9, 96.2, 95.6, 94.3, 90.6, 90.2, 87.5, 85.7, 56.4, 56.3,
56.1, 43.7, 43.6, 36.3. IR (CHCl3): 3591, 3176, 3011, 3007, 2939,
1639, 1657, 1594, 1578, 1487, 1083 cm-1. MS (EI) m/e (rel inten):
464 (M+, 40), 355 (100), 110 (38). HRMS (EI) calcd for C25H20O7S
(M+), 464.0930; found, 464.0937.
1
exo-Sulfide. H NMR (400 MHz, CDCl3): δ 7.51 (dd, J ) 8, 1.6
Hz, 1H, Hpara), 7.41 (dd, J ) 8, 1.6 Hz, 2H, Hortho), 7.29 (m, 2H, Hmeta),
7.18 (d, J ) 8.8 Hz, 1H, H-10), 6.68 (d, J ) 8.8 Hz, 1H, H-9), 6.54
(d, J ) 6 Hz, 1H, H-3a), 6.34 (s, 1H, H-5), 5.32 (dd, J ) 10.4, 4.4 Hz,
1H, H-2), 5.14 (s, 1H, OH), 4.27 (m, 1H, H-12c), 3.91 (s, 3H, OCH3),
3.90 (s, 3H, OCH3), 2.69 (dd, J ) 136, 4.4 Hz, 1H, H-1a), 2.34 (m,
1H, H-1b).
1
CH3CN/CH2Cl2). H NMR (400 MHz, CDCl3): δ 9.81 (s, 1H), 7.00
(d, J ) 9.2 Hz, 1H), 6.59 (d, J ) 9.2 Hz, 1H), 6.31 (s, 1H), 5.85 (s,
1H2), 3.90 (s, 3H), 3.88 (s, 3H), 3.84 (dd, J ) 11 Hz, 1H), 3.12 (dd,
J ) 18, 3.2 Hz, 1H), 2.63 (dd, J ) 18, 11 Hz, 1H), 1.05 (m, 21H,
TIPS), 0.99 (s, 9H), 0.20 (s, 3H), 0.19 (s, 3H). 13C NMR (100 MHz,
CDCl3): δ 199.9, 175.4, 163.3, 162.9, 154.0, 153.5, 147.8, 137.0, 123.0,
115.9, 108.6, 106.8, 105.6, 105.3, 90.6, 56.6, 56.4, 44.6, 43.7, 25.5,
10.1, 17.7, 17.6, 11.9, -4.4, -4.6. IR (CHCl3): 3002, 2940, 2868,
1721, 1654, 1597, 1577, 1484, 1091 cm-1. MS (EI) m/e (rel inten):
642 (M+, 100), 627 (3), 599 (15), 440 (51), 411 (21). Anal. Calcd for
C34H50O8Si2: C, 63.55; H, 7.79. Found: C, 63.39; H, 7.79.
2,11-Dihydroxy-6,8-dimethoxy-7H-1,2-dihydrofuro[3′,2′:4,5]furo-
[2,3-c]xanthen-7-one (30). TEA‚(HF)3 (0.25 mL, 1.56 mmol) was
added to a suspension of aldehyde 29 (0.090 g, 0.14 mmol) in CH3CN
(8 mL, 0.02 M). The reaction mixture clarified somewhat over the
period of 1 h, at which point it was filtered through a small pad of
silica and washed with copious amounts of 5% CH3OH/CH2Cl2 to
provide 0.038 g (73%) of 30 as a white solid, mp 253 °C (dec). Rf )
0.13 (5% CH3OH/CH2Cl2). 1H NMR (300 MHz, DMSO-d6; a complex
mixture of open and closed exo and endo forms giving the following
signals): δ 9.15 (s, 1H), 9.08 (s), 9.04 (s), 7.07 (d), 7.06 (d, J ) 8 Hz,
1H), 6.88 (d), 6.79 (d), 6.59 (d, J ) 8.7 Hz, 1H), 6.41 (d, J ) 6 Hz,
1H), 6.39 (d), 6.30 (s), 6.27 (s, 1H), 6.20 (s), 6.16 (m), 5.85 (m, 1H),
5.60 (m), 5.42 (m), 4.22 (m, 1H), 3.82 (s, 3H), 3.78 (s, 3H), 2.33 (m),
2.10 (m), 1.93 (d, J ) 12.3 Hz, 1H). IR (CHCl3): 3296, 3004, 2979,
2866, 1730, 1643, 1590, 1484, 1098 cm-1. MS (EI) m/e (rel inten):
372 (M+, 41), 354 (100), 339 (53), 325 (42). HRMS (EI) calcd for
C19H16O8 (M+), 372.0845; found, 372.0850.
6,8-Dimethoxy-11-hydroxy-2-thiophenyl-7H-1,2-dihydrofuro-
[3′,2′:4,5]furo[2,3-c]xanthen-7-one (endo-31). A suspension of hemi-
acetal 30 (0.05 g, 0.13 mmol), thiophenol (0.1 mL, 0.97 mmol), 4 Å
molecular sieves (0.1 g), and Amberlyst 15 resin (0.05 g) was heated
to reflux in 10 mL of CH3CN (0.01 M) for 2 h. The resulting solution
was concentrated to 3 mL, filtered, and then purified by column
chromatography (20 f 30% CH3CN/CH2Cl2) to provide 0.062 g (93%)
of an inseparable mixture of endo- and exo-sulfides as a yellow solid,
mp 273 °C (dec). Rf ) 0.41 (5% CH3OH/CH2Cl2). 1H NMR (400 MHz,
CDCl3): δ 7.51 (dd, J ) 8, 1.6 Hz, 1H, Hpara), 7.41 (dd, J ) 8, 1.6 Hz,
2H, Hortho), 7.29 (m, 2H, Hmeta), 7.19 (d, J ) 8.8 Hz, 1H, H-10), 6.70
(d, J ) 8.8 Hz, 1H, H-9), 6.56 (d, J ) 6 Hz, 1H, H-3a), 6.39 (s, 1H,
H-5), 5.81 (d, J ) 7.2 Hz, 1H, H-2), 5.14 (s, 1H, OH), 4.32 (dd, J )
8.4, 6 Hz, 1H, H-12c), 3.95 (s, 3H, OCH3), 3.92 (s, 3H, OCH3), 2.88
(m, 1H, H-1a), 2.60 (d, J ) 13.6 Hz, 1H, H-1b). 13C NMR (100 MHz,
CDCl3): δ endo/exo, 173.8, 173.5, 173.4, 163.9, 163.4, 162.3, 162.1,
160.3, 160.2, 156.7, 155.7, 153.2, 152.9, 152.7, 151.5, 151.4, 144.7,
6,8-Dimethoxy-11-hydroxy-7H-furo[3′,2′:4,5]furo[2,3-c]xanthen-
7-one, (11-Hydroxy-O-methylsterigmatocystin, 6). To a -45 °C
CH2Cl2 suspension (4 mL, 0.01 M) of phenylsulfides 31 (0.017 g, 0.037
mmol) was added 70% dry m-CPBA (0.11 g, 0.064 mmol). After 40
min, dimethyl sulfide (0.01 mL) and 10% Na2SO3 (2 mL) were added,
and the mixture was warmed to room temperature. The resulting
biphasic mixture was separated, and the aqueous portion was back-
extracted twice with CH2Cl2. The combined organic layers were
extracted with saturated NaHCO3 and brine, dried over K2CO3/MgSO4
(1:1), filtered (washing solids well with 5% CH3OH/CH2Cl2), and
concentrated to a tan solid. Purification by column chromatography
(5% CH3OH/CH2Cl2) provided the sulfoxides as a mixture of four
diastereomers, which were pyrolyzed directly. The sulfoxides and
diisopropylethylamine (two drops) were dissolved in toluene (3.5 mL)
and DMSO (0.75 mL), placed in a preheated oil bath at 120 °C, and
stirred for 1 h. The mixture was cooled to room temperature, diluted
with CH2Cl2, and extracted with saturated NaHCO3 to remove the
sulfinic acid and brine. The organic portion was dried over K2CO3/
MgSO4 (3:1), filtered, concentrated to ca. 2 mL, and loaded onto a
column of silica gel. Elution with 2 f 5% CH3OH/CH2Cl2 provided
pure 6 (8 mg, 62% yield from the phenylsulfides) as a white solid, mp
280 °C. Rf ) 0.30 (5% CH3OH/CH2Cl2). 1H NMR (400 MHz, CDCl3/
DMSO-d6): δ 8.54 (s, 1H), 6.70 (d, J ) 8.8 Hz, 1H), 6.38 (d, J ) 7.2
Hz, 1H), 6.21 (d, J ) 8.8 Hz, 1H), 6.05 (t, J ) 2.8 Hz, 1H), 5.98 (s,
1H), 5.15 (dd, J ) 2.8, 2.4 Hz, 1H), 4.51 (dt, J ) 7.2, 2.4 Hz, 1H),
3.47 (s, 3H), 3.43 (s, 3H). 13C NMR (100 MHz, CDCl3/DMSO-d6): δ
174.1, 161.8, 161.4, 151.8, 151.3, 144.1, 143.6, 138.2, 118.8, 113.4,
112.0, 107.4, 105.5, 105.2, 102.3, 89.3, 55.7, 55.5, 47.2. IR (CHCl3):
3180, 3007, 2961, 2852, 1638, 1581, 1487, 1465, 1271, 1263, 1137,
1103, 1082, 980 cm-1. UV-vis (c ) 8 µg/mL, CH3OH): λmax 316
(11 800), 239 (31 800). MS (CI) m/e (rel inten): 355 ([M + H]+, 28),
341 (12), 102 (56). HRMS (EI) calcd for C19H14O7 (M+), 354.0740;
found, 354.0739.
Supporting Information Available: Full experimental details
for intermediates and reagents shown in Schemes 4-6 (PDF).
This material is available free of charge via the Internet at
http://pubs.acs.org.
JA012185V
9
J. AM. CHEM. SOC. VOL. 124, NO. 19, 2002 5303