Quararibea metabolites. 4.1 total synthesis and conformational studies of (±)-funebrine and (±)-funebral

Article abstract of DOI:10.1021/jo981501u

Syntheses of racemic forms of the main secondary metabolites of Quararibea funebris, (±)-funebrine, (±)-funebral, and their biogenetic precursor (±)-(2S,3S,4R)-γ-hydroxyalloisoluecine lactone have been developed. In synthetic studies, a new variation of t

Full text of DOI:10.1021/jo981501u

J . Org. Chem. 1999, 64, 2657-2666
2657
Qu a r a r ibea Meta bolites. 4.1 Tota l Syn th esis a n d Con for m a tion a l
Stu d ies of (()-F u n ebr in e a n d (()-F u n ebr a l
Ying Dong, Niranjan N. Pai, Sheri L. Ablaza, Shao-Xia Yu, Simon Bolvig,
David A. Forsyth,* and Philip W. Le Quesne*
Department of Chemistry and Barnett Institute, Northeastern University,
Boston, Massachusetts 02115-5096
Received J uly 28, 1998
Syntheses of racemic forms of the main secondary metabolites of Quararibea funebris, (()-funebrine,
(()-funebral, and their biogenetic precursor (()-(2S,3S,4R)-γ-hydroxyalloisoluecine lactone have
been developed. In synthetic studies, a new variation of the Paal-Knorr condensation employing
titanium isopropoxide was utilized to construct the pyrrole lactone moiety. Two efficient synthetic
approaches to the key (()-γ-amino lactone have been developed, one based on Claisen chemistry
and the other on addition reactions to the butenolide ring of â-angelicalactone. The restricted rotation
around the C(sp3)-N(sp2) bond in the pyrrole lactone structures of (()-funebrine, (()-funebral,
and related aldehydes has been probed by conformational dynamic studies, and the barriers for
interconversion between conformations have been measured by full NMR line-shape analysis.
1
Molecular mechanics (MMX) and a H-1H NOE study indicate a distinct preferred conformation
for (()-funebrine.
Quararibea funebris (Llave) Vischer (Bombacaceae)2 is
a moderately large tree indigenous to southeastern
Mexico and Guatemala. It derives its specific name from
the fact that the Zapotec people of Oaxaca, Mexico,
conducted funeral rites beneath its branches. The strongly
odorous flowers of the tree have been used as a flavor
additive in chocolate drinks from pre-Columbian times,
and in local folk medicine the plant has served as a cough
remedy and antipyretic, to control menstrual disorders
and psychopathic fears, and as a hallucinogen.3 Our
interest in Q. funebris first arose from a report4 of
alkaloids in its flowers and from Schultes’s intriguing
observation3 that the dried flowers and leaves have a
characteristic spicy odor that persists even in long-stored
herbarium specimens. Our investigation of the secondary
metabolites of Q. funebris dried flowers yielded a family
of compounds, 1-4, from polar fractions. Enol lactone 3
and enamine 4 are responsible for the odor of the flowers.
The saturated lactone 2 and its parent amino acid,
(2S,3S,4R)-γ-hydroxyisoleucine, were obtained along with
the pyrrole alkaloid funebrine (1).5 The stereochemistry
shown in 1-4 is based on the presumed derivation of 1,
3, and 4 from 2.5 Subsequent investigations disclosed two
related metabolites, funebral (5)6 and funebradiol (6).7
All these Quararibea metabolites are structurally un-
usual. Indeed the only similar natural pyrrole isolated
thus far is the bisnorfunebral derivative 7. This com-
pound was obtained by Lynn8 from pea seedlings (Pisum
sativum L.), where it was identified as a regulator of
cellular cycles.
Funebrine (1) is a structurally unique alkaloid, and its
discovery also represents the first characterization of an
alkaloid in the family Bombaceae. Since alkaloids are
known to exhibit a wide range of physiological activity,9
and extraction of the flowers of Q. funebris yielded
insufficient funebrine (1) and funebral (5) for full biologi-
cal evaluation, we initiated a synthetic program.
The existence of an imine bond in funebrine (1)
suggests C14-N15 bond disconnection to give funebral
(5) and γ-hydroxyalloisoleucine lactone (2) (Scheme 1).10
The known lability of imines suggests this reaction
should be the last in the synthesis of funebrine.
Funebral (5) has a pyrrole ring with a N-substituted
lactone, the lactone being identical to lactone 2, and
formyl and hydroxymethyl substituents at positions 2
and 5 of the pyrrole, respectively. Various ways to
incorporate the N-, 2-, 5-substituents on a pyrrole ring
(1) For part 3, see Yu, S-X.; LeQuesne, P. W. Tetrahedron Lett. 1995,
36, 6205.
(2) For a thorough taxonomic discussion, see Alverson, W. S. Taxon
1989, 38, 377.
(3) Schultes, R. E. Bot. Mus. Leafl. 1957, 17, 247.
(4) Rosengarten, F. Bot. Mus. Leafl. 1977, 25, 184.
(5) Raffauf, R. F.; Zennie, T. M.; Onan, K. D.; Le Quesne, P. W. J .
Org. Chem. 1984, 49, 2714.
(6) Zennie, T. M.; Cassady, J . M.; Raffauf, R. F. J . Nat. Prod. 1986,
49, 695.
(7) Zennie, T. M.; Cassady, J . M. J . Nat. Prod. 1990, 53, 1611.
(8) Lynn, D. D.; J affe, K.; Cornwall, M.; Tramontano, W. J . Am.
Chem. Soc. 1987, 109, 5858.
(9) Dalton, D. R. The Alkaloids; Marcel Dekker Inc: New York,
Basel, 1979.
(10) The numbering system here originated from the earlier X-ray
result; see ref 5.
10.1021/jo981501u CCC: $18.00 © 1999 American Chemical Society
Published on Web 04/01/1999

2658 J . Org. Chem., Vol. 64, No. 8, 1999
Dong et al.
Sch em e 1
a convenient platform for arranging the substituents of
2 in the desired trans,trans-relationship.13 Treatment of
â-angelicalactone (19) first with lithium tris(thiophenyl)-
methide and then with trisyl azide gave 20, which was
reduced with Raney nickel to give 2. Despite the brevity
and stereospecificity of this route, the final reduction step
and workup were low-yielding, and the approach is at
present less efficient than that of Scheme 3, which now
proceeds in ca. 20% yield from t-BOC-glycine.
Our first approach to funebral (5) (Scheme 5), which
made use of the chemistry of Scheme 3 but employed
N-pyrrolylacetic acid instead of protected glycine, failed
because we did not find iodolactonization conditions to
which the pyrrole ring was unreactive. Esterification of
acid 21 with trans-crotyl alcohol proceeded normally to
give 22, which after rearrangement via 23 and methyl-
ation to give 24 was subjected unsuccessfully to a variety
of iodolactonization reaction conditions. We then turned
to pyrrole lactone 25, obtained from amino lactone
hydrochloride 15 and 2,5-dimethoxytetrahydrofuran
(Scheme 6). We hoped that this compound would under-
goVilsmeier-Haack formylation to give first the mono-
and then the diformylated derivative 26 and 16 (Scheme
6). Compound 26 was formed without difficulty, but its
1H NMR spectrum clearly showed restricted rotation
around the interannular N1-C6 bond. This finding
augured poorly for successful formylation at the remain-
ing R-carbon of the pyrrole. In fact, Vilsmeier-Haack
reaction of 26 gave only small quantities of the R,γ-
diformyl product. If compound 26 exists as a rotameric
pair at room temperature, funebral (5) itself should do
the same. The original sample of natural funebral
Sch em e 2
1
comprised only 0.54 mg, and its H NMR spectrum did
not clearly disclose doubled signals arising from re-
stricted rotation. A second isolation of funebral has now
given material whose 1H NMR spectrum showed the
existence of two rotamers in a 9:1 ratio in CDCl3.
We then considered Lynn’s method for the synthesis
of N-substituted-R,R′-diformylpyrroles, i.e., ozonolysis of
a bisisobutenyl compound such as 17. Such pyrroles can
be prepared by Paal-Knorr reactions between amines
and 2,9-dimethyldeca-2,8-diene-4,7-dione (18),27 which
itself was obtained by Cu(II)-mediated oxidative dimer-
ization of mesityl oxide. Attempted Paal-Knorr reactions
between amino lactone 2 and 18 employing various
literature procedures either returned starting material
or led to decomposition. Molecular models suggested that
the target pyrrole 17 should be sterically crowded,
probably also having restricted rotation at room temper-
ature. In this case its formation in the Paal-Knorr
reaction should be relatively slow. We therefore sought
a catalyst for the Paal-Knorr reaction that would
strongly coordinate the oxygen atoms of the diketone and
yield neutral reaction products, which would not catalyze
polymerization of the pyrrole. Titanium(IV) isopropoxide
suggested itself as a suitable catalyst, and a series of
model reactions shown in Table 1 demonstrated its
utility.14
were investigated (see below), but eventually Lynn’s
strategy8 of chemoselective reduction of one of the alde-
hyde groups in the 2,5-diformyl-N-substituted pyrrole 16
was adopted. Dialdehyde 16 was obtained by oxidation
of the diene 17, which was conveniently synthesized by
Paal-Knorr condensation between γ-diketone 18 and
lactone 2 (Scheme 2).
Synthesis of (()-amino lactone 2 was achieved by two
routes. The first, shown in Scheme 3, is based on highly
stereoselective ester enolate Claisen chemistry developed
by Bartlett and co-workers.11 The key rearrangement of
ester 8 took place with >20:1 stereoselectivity in favor
of the desired isomer, based on peak area ratio of the
methyl doublets in proton NMR. Direct iodolactonization
of acid 9 gave a cyclic carbamate as the major product,
since the tert-butoxycarbonyl protecting group of 9 pref-
erentially participated in the cyclization. We therefore
adjusted the protecting group of 9 to phthaloyl by
hydrolysis with trifluoroacetic acid to yield the amine salt
10, which was subsequently refluxed with phthalic
anhydride in the presence of triethylamine. The methyl
ester derivative 12 was used instead of acid 11 in the
iodolactonization step because high stereoselectivity was
achieved under thermodynamic control.12 Reduction of
the iodomethyl compound 13 with tributylstannane gave,
after acid hydrolysis of the phthaloyl group, the hydro-
chloride 15 of the desired amino lactone 2.
Condensation of amino lactone 2 with 18 in the
presence of titanium(IV) isopropoxide gave pyrrole 17
1
(Scheme 7). H and 13C NMR spectra of 17, as expected,
In a second approach (Scheme 4), we considered that
the butenolide ring of â-angelicalactone (19) would offer
showed significant restricted rotation around the inter-
annular N1-C6 bond. Treatment of this compound with
(11) Bartlett, P. A.; Richardson, D. P.; Myerson, J . Tetrahedron 1984,
40, 2317.
(12) Bartlett, P. A.; Myerson, J . J . Am. Chem. Soc. 1978, 100, 3950.
(13) Chen, S.-Y.; J oullie´, M. M. J . Org. Chem. 1984, 49, 2168.
(14) Yu, S-X. M.S. Thesis, Northeastern University, Boston, 1995.

Quararibea Metabolites
J . Org. Chem., Vol. 64, No. 8, 1999 2659
Sch em e 3
Sch em e 4
Sch em e 6
Sch em e 5
Ta ble 1. P a a l-Kn or r Syn th esis of N,2,5-Tr isu bstitu ted
P yr r oles Usin g Tita n iu m (IV) Isop r op oxid e
osmium tetraoxide and sodium metaperiodate gave di-
aldehyde 16, which, on reduction with sodium cyanoboro-
hydride, gave funebral (5) having identical spectral
characteristics to the natural product.
Synthesis of (()-funebrine (1) from (()-funebral (5)
presented two challenges. First, reaction of (()-lactone
2 with (()-funebral (5) could potentially yield (()-
funebrine and its diastereomer; secondly, formation of
the imine bond might be slow on account of steric
crowding. To understand better the chemistry of the
imine bond present in funebrine (1), the reaction between
the model compound N-methyl-2-formylpyrrole (28) and
the easily obtainable (SRR,RSS)-amino lactone hydro-

2660 J . Org. Chem., Vol. 64, No. 8, 1999
Dong et al.
Sch em e 7
Sch em e 8
Sch em e 9
Sch em e 10
chloride 2928 was investigated (Scheme 8). Attempted
condensation of 28 and 29 in benzene with molecular
sieves and triethylamine under reflux15 gave an insepa-
1
rable mixture, the H NMR spectrum of which showed
no evidence of condensation having taken place. Using a
procedure similar to Olsson’s,16 hydrochloride 29, after
neutralization with dipotassium hydrogen phosphate,
was treated with sodium cyanoborohydride in methanol
in the presence of aldehyde 28. A ninhydrin- and Ehrlich-
positive compound was formed after 48 h of stirring and
workup, which, after separation of components, gave the
reductive amination product 30.
The free (()-amino lactone 2 was generated from the
hydrochloride without opening the sensitive lactone ring
by adding the solid 15 to mixed 1:1 chloroform and
saturated sodium bicarbonate solution. When aldehyde
28, in a slurry of molecular sieves and triethylamine in
anhydrous chloroform,17 was treated with the free base
of (()-amino lactone 2, an unstable imine 32 was formed
after stirring overnight at room temperature. Under
similar conditions, 2 condensed with 5 to give 1 (60%)
after refluxing in chloroform for four days (Scheme 9).
The identity of the synthetic (()-funebrine was confirmed
by mass spectral analysis, 13C NMR data, and comparison
of the 1H NMR spectrum with that of the naturally
derived sample.
mixture of (()-funebrine and a diastereomer, “(()-Ψ-
funebrine”, in which the absolute configurations of the
component lactone rings are opposite rather than the
same. Careful search of the condensation reaction prod-
ucts revealed only (()-funebrine, recovered starting
1
materials, and traces of a compound which from its H
NMR spectrum was clearly a decomposition product.
Molecular mechanics calculations (vide infra) of funebrine
and “Ψ-funebrine” showed that the most stable confor-
mation of the latter is less stable by 2 kcal/mol than the
most stable conformation of funebrine. Since imine
formation is a reversible reaction, formation of the
thermodynamically more stable funebrine would be
favored.
Incidentally, titanium tetraisopropoxide was observed
to catalyze the formation of an imine bond between
aldehyde 28 and benzylamine (Scheme 10). The Schiff’s
base 31 was formed in 83% yield after 12 h at 20 °C.
However, titanium tetraisopropoxide did not induce
imine bond formation between (()-funebral and amino
lactone 2, probably because of complexation between
titanium and the lactone oxygens.
As mentioned above, condensation between 2 and (()-
funebral (5) might have been expected to produce a
(15) Brunner, H.; Reiter, R.; Riepl, G. Chem. Ber. 1984, 17, 1330.
(16) Olsson, K. Pernemalm, P. Acta Chem. Scand. 1979, B33, 125.
(17) Scott, M. K.; Martin, G. E.; DiStefano, D. L.; Fedde, C. L.;
Kukla, M. J .; Barrett, D. L.; Baldy, W. J .; Elgin, R. J ., J r.; Kesslick, J .
M.; Mathiasen, J . R.; Shank, R. P.; Vaught, J . L. J . Med. Chem. 1992,
49, 2168.
The pyrrole lactone structure central to funebrine,
funebral, and their synthetic precursors presents a
structural situation in which crowding introduced by

Quararibea Metabolites
J . Org. Chem., Vol. 64, No. 8, 1999 2661
F igu r e 1. Experimental 300 MHz 1H NMR spectra of the
aldehyde signal from 26 at various temperatures are in the
right column, and DNMR5 simulations at different exchange
rates are in the left column.
F igu r e 2. Experimental 300 MHz 1H NMR spectra of the C12-
CH3 signal of 26 are in the right column, and DNMR5
simulations are in the left column.
substituents at the 2- and 5-positions of the pyrrole
results in restricted conformational motion about the
C6-N1 bond. While the sixfold barrier for an unhindered
rotation about a bond linking tetrahedral and trigonal
planar centers is ordinarily very small, the conformations
available to the pyrrole lactones are restricted and the
barrier for rotation about the C(sp3)-N(sp2) bond is
For the dialdehyde 16, two aldehyde 1H signals of equal
area are seen in the ambient temperature 300 MHz
spectrum, corresponding to nonequivalence of the alde-
hyde groups induced by slow rotation about the C-N
bond linking the pyrrole and lactone portions. Indeed,
the equal-area signals, along with the lack of a second
set of signals for lactone resonances, establish firmly that
the dynamic NMR observations are not due to hindered
rotation about the bond linking the formyl group to the
pyrrole ring. Coalescence of the aldehyde signals occurs
at elevated temperatures in spectra of a solution of 16
in tetrachloroethane-d2. Line-shape analysis yields a
barrier of ∆Gq(300 K), 17.2 ( 0.3 kcal/mol, with a ∆Hq of
13.3 kcal/mol and a ∆Sq of -13.1 eu.
1
substantial. We find evidence in variable temperature H
NMR studies for the existence of distinct conformations
for the pyrrole lactones 26, 16, 5, and 1, and we have
measured the barriers for interconversion between the
conformations by full NMR line-shape analysis. Molec-
ular mechanics (MMX) studies18 provide a further basis
for defining these conformations.
For the monoaldehyde 26 just one signal is seen for
each 1H nucleus in the ambient temperature 300 MHz
spectrum. Upon cooling, signals broaden and decoalesce
eventually into two sets of signals for two unequally
populated conformers. At 226 K, the ratio is about 3:2 in
CDCl3. The exchange process is most easily analyzed
from the singlet aldehyde signal that becomes two
singlets, and in the C12-CH3 doublet signal that becomes
two doublets at low temperature. DNMR5 lineshape
analyses19 of these two signals are shown in Figures 1
and 2. (The numbering scheme for all atoms discussed
in regard to NMR data is given in Scheme 1.) The
dynamical information from the two signals agrees very
well. The rates are most likely to be determined ac-
curately from spectra in which there is significant line-
broadening, and, accordingly, the rate constants agree
within 10% for the intermediate temperatures from 226
K and 266 K. Agreement for the narrower spectra at
higher and lower temperatures is still good. Combining
the data from both sources in an Eyring plot of log k/T
vs 1/T gives the activation parameters for the conversion
from the major conformer to the minor conformer: ∆Gq-
(300 K), 13.1 ( 0.2 kcal/mol; ∆Hq 9.4 kcal/mol; and ∆Sq,
-12 eu.
Funebral (5) shows two sets of NMR signals in an
approximately 8:1 ratio in ambient temperature 300 MHz
1H spectra in a tetrachloroethane-d2 solution, indicating
two different conformations. Raising the temperature
leads to coalescence of all doubled signals. Complete line-
shape analysis of the aldehyde signal gives activation
parameters for the conversion from the more stable
conformer to the less stable conformer: ∆Gq(300 K), 17.2
( 0.2 kcal/mol; ∆Hq 12.3 kcal/mol; and ∆Sq -16.4 eu.
Funebrine (1) appears to be similar to 5 and 16 in the
magnitude of the rotational barrier about the pyrrole
lactone C-N bond, but it shows an even more pronounced
conformational preference than for funebral (5). In
acetone-d6, the ratio of the major to the minor conforma-
tion is 13.5:1 at ambient temperature. In CDCl3, the
minor conformer was not detected in a low-noise 1H NMR
spectrum, and so the ratio was estimated to be g50:1.
Signal broadening was evident in a 1H spectrum in
acetone-d6 when the temperature was raised to 45 °C,
but the dynamics were not pursued further because of
the low boiling point of acetone and because the lopsided
conformational preference would render questionable the
accuracy of the line-shape analysis.
The conformational preferences for funebrine and
Ψ-funebrine were studied via molecular mechanics using
the MMX forcefield. The searches for global minima were
carried out by systematic variation of the dihedral angles
involving rotation about the N1-C6, C5-C24, C2-C14,
C14-N15, and N15-C16 bonds. The dihedral driver
routines were not used because it was discovered that
(18) Gajewski, J . J .; Gilbert, K. E.; McKelvie, H. In Advances in
Molecular Modelling; Liotta, D., Ed.; J AI Press: Greenwich, CT, 1990;
Vol. 2. PCMODEL, V. 6.0, Serena Software: Box 3076, Bloomington,
IN.
(19) A locally modified version of the DNMR5 (Stephenson, D. S.;
Binsch, G.) program further adapted by C. B. LeMaster, C. L.
LeMaster, and N. S. True. Program no. QCMP059, Quantum Chem-
istry Program Exchange, Indiana University, Bloomington, IN 47405.

2662 J . Org. Chem., Vol. 64, No. 8, 1999
Dong et al.
there are relatively few low-lying minima in this crowded
molecule. Rotation through 360° for most of the bonds
produce just two minima. Further, the lactone ring
conformation is well-defined because of the trans-trans
alignment of the three substituents on the five-membered
ring. The lowest energy conformation for funebrine is
depicted in Figure 3. This conformation is 2.3 kcal/mol
lower in energy than the lowest found for the diastere-
omeric Ψ-funebrine. It is also lower by nearly 3.0 kcal/
mol than any other conformers for funebrine that involve
different placements of the imine bond or lactone rings.
There are a few additional low energy conformers involv-
ing rotations of the CH2OH group, but the predicted
orientation of the hydroxyl oxygen, and indeed the entire
structure, correspond precisely to the conformation found
for the solid state in an X-ray crystal structure.5
As shown in Figure 3, the conformation appears to
have a fairly close approach of the two lactone rings.
However, the lactone shapes are complementary and not
particularly crowded since the methyls of the pyrrole
lactone are above the plane of the pyrrole ring as shown
while the methyls of the imine lactone are below the
plane. The closest approach is 3.33 Å between the
methine proton at C20 in the imine lactone and the C12
methyl carbon attached to the pyrrole lactone. (See
Scheme 1 for atom numbers.) Dipole-dipole interactions
probably play an important role in determing the con-
formational preference. In particular, the imine nitrogen
N15 is predicted to be only 2.84 Å from the C7 carbonyl
carbon of the pyrrole lactone, its closest approach to any
of the carbons or oxygens of that lactone.
F igu r e 3. The lowest energy conformer of funebrine predicted
in MMX calculations.
Strong support for the actual existence of funebrine
in the conformation shown in Figure 3 comes from
steady-state, difference NOE measurements and from the
chemical shift of the C6-H methine signal. Preirradiation
(for 10 s) of the imine hydrogen C14-H produced a
positive NOE of 6.8% at C16-H and 6.4% at C3-H in a
CDCl3 solution (no other NOE’s were detected in this
experiment). As shown in Figure 3, the imine C14-H is
predicted to be at the middle of a nearly planar “double-
U” segment with distances of 2.37 Å to C16-H of the
imine lactone and 2.64 Å to the pyrrole C3-H. The high
NOE’s are certainly consistent with the “double-U”
alignment and are otherwise difficult to explain. At-
tempts to detect an inter-ring NOE between the C20-H
of the imine lactone and C12 methyl yielded ambiguous
results because of the overlap of C12 and C22 methyl
signals.
F igu r e 4. The chemical shifts at C6-H and conformations of
pyrrole aldehydes related to funebrine.
The orientation of the pyrrole lactone ring in funebrine
is established by the chemical shift of the C6-H signal
that exhibits a remarkable 2.25 ppm shift difference
between the major and minor conformations. The C6-H
signal for the predominant conformer occurs at a rather
ordinary δ 5.18 in acetone-d6 (δ 5.03 in CDCl3), while
C6-H for the minor conformer is strongly deshielded at
δ 7.43. These shifts correspond to having the C6-H
approximately anti to C2 and its attached imine group
in the major conformer (Figure 3), and approximately syn
in the minor conformer. The syn alignment brings C6-H
close to the imine nitrogen and results in strong deshield-
ing. These assignments are based on comparison with
similar chemical shift behavior in the aldehydes 5, 16,
and 26.
F igu r e 5. MMX structures of the observed major (left) and
minor (right) conformations of 26.
16, the C6-H methine must be near one of the two
carbonyls, and the nucleus is deshielded to δ 6.46. That
the carbonyl oxygens of 16 are oriented syn with respect
to the pyrrole nitrogen was established by steady state,
difference NOE measurements on a degassed CDCl3
solution: preirradiation (10 s) of the δ 9.84 aldehyde
signal gave a 14.7% enhancement to C3-H at δ 7.08, and
irradiation at δ 9.75 gave a 12.6% effect to C4-H at δ
7.15. (Saturation of the closely spaced aldehyde signals
was not completely selective, but the qualitative conclu-
sion is clear. Saturation of both aldehyde signals resulted
in a positive NOE of only 2.2% at C6-H, most likely due
The C6-H signal is diagnostic of the conformation of
the pyrrole lactone with respect to the carbonyl group in
5, 16, and 26, as shown in Figure 4. In the dialdehyde

Quararibea Metabolites
J . Org. Chem., Vol. 64, No. 8, 1999 2663
(SR,RS)-3-Meth yl-2-p h th a lim idopen t-4-en oic Acid (11).
Compound 11 (mixed with <5% of its diastereomer) was
prepared from the trifluoroacetate salt of 10 (6.24 g, 25.6
mmol) by using the procedure of ref 22. It was obtained as a
colorless solid (4.79 g, 73%): mp 129-132 °C; IR (KBr) cm-1
3500 (br), 1775, 1725; 1H NMR (CDCl3) δ 1.04 (d, J ) 7.0 Hz,
3H), 3.34 (m, 1H), 4.81 (d, J ) 8.2 Hz, 1H), 5.13 (m, 2H), 5.90
(m, 1H), 7.71∼7.86 (m, 4H), 10.7 (bs, 1H); minor isomer: 1.25
(d, J ) 6.7 Hz), 3.12 (m), 4.66 (d, J ) 9.5 Hz), 4.8-4.98 (m),
5.51-5.63 (m); 13C NMR (CDCl3) δ 16.6, 37.3, 55.9, 115.7,
123.6, 131.4, 134.2, 139.7, 167.6, 173.5; minor isomer: 18.0,
38.3, 45.0, 116.9, 139.0, 167.3; MS: 259 [M+].
Meth yl (SR,RS)-2-P h th a lim id o-3-m eth ylp en t-4-en ola te
(12). Compound 11 (0.65 g, 2.5 mmol) in dry Et2O was added
to excess ethereal diazomethane (16.6 mmol) and stirred at 0
°C for 2 h. Excess diazomethane was destroyed by the dropwise
addition of acetic acid. The ethereal solution was washed with
sat. NaHCO3 solution (2 × 5 mL), dried (Na2SO4), and
evaporated to give 12 (0.56 g, 76%) as a chromatographically
homogeneous amber oil which could be used in the next step
without further characterization: IR (CHCl3) cm-1 3010, 2920,
1750; 1H NMR (CDCl3) δ 1.05 (d, J ) 7.0 Hz, 3H), 3.35 (m,
1H), 3.89 (s, 3H), 4.48 (d, J ) 8.8 Hz, 1H), 5.05-5.2 (m, 2H),
5.85-6.05 (m, 1H), 7.60-7.86 (m, 4H); minor isomer: 1.25 (d,
J ) 6.7 Hz), 3.34 (m), 3.82 (s), 4.68 (d), 4.85-5.0 (m), 5.55-
5.70 (m); 13C NMR (CDCl3) δ 16.6, 37.3, 55.9, 115.7, 123.6,
131.4, 134.2, 139.7, 167.6, 173.5; minor isomer: 18.0, 38.3,
45.0, 116.9, 139.0, 167.3; MS: 273 [M+], 241 [M+ - 32], 218
[M+ - 45], 190 [M+ - 83], 126 [M+ - 147].
(SSS,RRR)-3,4-Dihydro-2-(iodomethyl)-3-methyl-4-phthal-
im id ofu r a n -2[5H]-on e (13). Compound 13 (mixed with <5%
of its diastereomer) was prepared from 12 (12 g, 44 mmol) by
using the procedure of ref 12. The crude product was recrystal-
lized (ethyl acetate/hexane) to yield pure 13 (12.22 g, 72%):
mp 187-190 °C; IR (KBr) cm-1 2960, 2940, 1775, 1720; 1H
NMR (CDCl3) δ 1.3 (d, J ) 7.2 Hz, 3H), 2.87 (m, 1H), 3.47 (m
1H), 3.58 (m, 1H), 4.18 (m, 1H), 4.81 (d, J ) 11.6 Hz, 1H),
7.75-7.95 (m, 4H); minor isomer: 1.0 (d), 4.35 (m); 13C NMR
(CDCl3) δ 13.5, 15.5, 40.9, 55.4, 82.2, 123.8, 131.5, 134.6, 166.8,
169.7; minor isomer: 12.0, 37.0, 79.3; MS: 385 [M+], 258 [M+
- 127].
to rotation through minor conformers during the long
preirradiation.) For 26, the minor conformer has C6-H
at δ 6.37 and thus is assigned as having C6-H syn to the
aldehyde based on similarity to the shift in the dialde-
hyde. The C6-H occurs at δ 4.58 in the major conformer.
Similarly, for funebral (5), C6-H occurs at δ 6.63 in the
minor conformer but at only δ 4.92 in the major con-
former.
The geometry assignments deduced above from NMR
data are consistent with MMX calculations that predict
approximately syn and anti alignments for C6-H in the
two conformations. For example, the major and minor
conformations predicted from 26 are shown in Figure 5.
The MMX structures for 26 are nearly isoenergetic, with
the observed minor conformer favored by 0.1 kcal/mol in
the calculations. In either structure, rotation of the
aldehyde group by about 180° raises the energy by about
three kcal/mol.
In summary, total syntheses of (()-funebral and (()-
funebrine have been achieved that involve a new varia-
tion of the Paal-Knorr condensation to construct the
pyrrole lactone moiety. Condensation between (()-fune-
bral and the racemic amino lactone 2 produced only (()-
funebrine and not the possible diastereomeric (()-Ψ-
funebrine. Molecular mechanics calculations indicate that
funebrine is more stable than Ψ-funebrine and also
predict that the most stable conformation for funebrine
is the same as the one found in the solid state. NMR
chemical shift and NOE measurements support the
conclusion that this predicted conformation is strongly
favored in CDCl3 solution, although less than 10% of a
minor conformer can be detected in acetone-d6. The
barrier to rotation about the pyrrole lactone C-N bond
in funebral (5) is 17.2 kcal/mol and funebrine (1) has a
similarly high barrier for this rotation.
Exp er im en ta l Section
(SSR,RRS)-3,4-Dih yd r o-4,5-d im eth yl-3-p h th a lim id ofu -
r a n -2[5H]-on e (14). Compound 13 (1.3 g, 3.38 mmol) was
dissolved in 20 mL of dry THF. Tri-n-butyltin hydride (5.41 g,
18.6 mmol) was added via syringe, and the reaction mixture
was stirred at 20 °C for 20 h. The solvent was then evaporated,
and the residue was dissolved in 15 mL of CH3CN. The
solution was extracted with hexane (8 × 5 mL) until TLC and
NMR showed no tin compound present. The CH3CN phase was
evaporated at reduced pressure to give a pale yellow solid as
crude product which upon recrystallization from ethyl acetate/
hexane gave a colorless solid (0.806 g, 93%): mp 175-177 °C;
IR (KBr) cm-1 2920, 1775, 1720; 1H NMR (CDCl3) δ 1.15 (d, J
) 6.60 Hz, 3H), 1.53 (d, J ) 6.2 Hz, 3H), 2.83 (m, 1H), 4.25
(m, 1H), 4.71 (d, J ) 11.8 Hz, 1H), 7.73-7.86 (m, 4H); minor
isomer: 0.95 (d), 1.40 (d), 2.72 (m), 5.02 (d); 13C NMR (CDCl3)
δ 14.0, 18.7, 41.9, 55.6, 80.5, 123.7, 131.6, 134.4, 166.5, 170.9;
MS: 259 [M+], 220 [M+ - 39], 200 [M+ - 59]; Anal. calcd for
C14H13NO4: C, 64.86; H, 5.02; N, 5.41. Found: C, 64.67; H,
5.00; N, 5.30.
Ma t er ia ls a n d Met h od s. Microanalyses were by Spang
Microanalytical Laboratory, Eagle Harbor, MI. Melting points
were determined on a capillary melting point apparatus and
are uncorrected. IR spectra were recorded using KBr pellets
or in CCl4. Chromatography refers to flash chromatography
on silica gel according to the method of Still.20 trans-Crotyl
N-t-Boc-glycinate (8),21 2-(N-t-Boc)-3-methylpent-4-enoic acid
(9),22 (SSS,RRR)-3,4-dihydro-2-(iodomethyl)-3-methyl-4-phthal-
imidofuran-5[2H]-one (13)12 and 2,4,6-triisopropylbenzene-
sulfonyl azide23 were prepared by procedures previously
reported. Ethereal diazomethane was prepared by an estab-
lished procedure.24 Toluene and Et3N were distilled from
calcium hydride and stored over 4 Å molecular sieves. THF
and Et2O were distilled from sodium/benzophenone ketyl prior
to use. DMF and CHCl3 were dried over 4 Å molecular sieves
in several batches. All other reagents were used as purchased.
Tr iflu or oa ceta te Sa lt of (SR,RS)-2-Am in o-2-(N-t-Boc-
)-3-m eth ylp en t-4-en oic Acid (10). Compound 10 was pre-
pared from (SR,RS)-2-amino-2-(N-t-Boc-)-3-methylpent-4-enoic
acid 922 (10.96 g, 47.97 mmol) by using the procedure of ref
22. The crude trifluoroacetate salt was purified by recrystal-
lization from methanol/water yielding a colorless solid (10.7
g, 92%): mp 154-156 °C; IR (KBr) cm-1 3450,3300, 2250, 1600;
1H NMR (D2O) δ 1.0 (d, J ) 7.0 Hz, 3H), 2.75 (m, 1H), 3.84 (d,
J ) 4.2 Hz, 1H), 5.13 (m, 2H), 5.65 (m, 1H).
(SSR,RRS)-3,4-Dih yd r o-3-a m in o-4,5-d im eth ylfu r a n -2-
[5H]-on e Hyd r och lor id e (15). Compound 14 (0.806 g, 3.31
mmol) was refluxed with 6 N HCl (10 mL) at 110 °C for 14 h.
The reaction mixture was cooled to 20 °C and extracted with
Et2O (4 × 25 mL). The aqueous layer was then evaporated to
dryness. The crude product was further purified by recrystal-
lization from CH3OH/Et2O and dried in vacuo over KOH for
24 h to yield a colorless solid (0.417 g, 81%) identical in all
respects except optical rotation with the hydrochloride of the
(20) Still, C. W.; Kahn, M.; Mitra, A. J . Org. Chem. 1978, 50, 2923.
(21) Bartlett, P. A.; Tanzella, D. J .; Barstow, J . F. Tetrahedron Lett.
1982, 23, 619.
(22) Bartlet, P. A.; Barstow, J . F. J . Org. Chem. 1982, 47, 3933.
(23) Harmon, R. E.; Wellman, G.; Gupta, S. K. J . Org. Chem. 1973,
38, 11.
1
natural amino lactone 2: mp 205-207 °C; H NMR (D2O) δ
1.21 (d, J ) 6.6 Hz, 3H), 1.42 (d, J ) 6.2 Hz, 3H), 2.25-2.40
(m, 1H), 4.12 (d, J ) 11.7 Hz, 1H), 4.33-4.45 (dq, J ) 9.6, 6.2
Hz, 1H); minor isomer: 1.10 (d), 1.40 (d), 2.72 (m), 4.60 (d);
13C NMR (D2O) δ 15.1, 20.0, 45.5, 58.5, 85.0, 176.2. Anal. Calcd
(24) Hudlicky, M. J . Org. Chem. 1980, 45, 5377.

2664 J . Org. Chem., Vol. 64, No. 8, 1999
Dong et al.
for C6H12NO2Cl‚1/2 H2O: C, 41.26; H, 7.45; N, 8.02. Found:
three times and suspended in 10 mL of THF was added to the
flask. The mixture was stirred mechanically for 4 h at reflux
temperature (60-65 °C) under a hydrogen atmosphere. The
reaction was cooled to 20 °C, and 3 mL of concd NH4OH was
added with stirring. The clear THF solution was transferred
to a new flask, and the Raney nickel particles were washed
three times with a mixture of THF (22 mL) and concd NH4-
OH (3 mL). The combined THF-NH4OH solution was evapo-
rated under reduced pressure. The residue was dissolved in
20 mL of diethyl ether and extracted with water (3 × 10 mL).
The organic layer was treated with concd HCl (2 mL) and
extracted with H2O (3 × 25 mL). The combined aqueous phase
was evaporated to dryness under reduced pressure, giving
compound 15 (16 mg, 15%) identical to that prepared above.
(E)-2-Bu t en yl-1-p yr r olyl Acet a t e (22). 1-Pyrrolylacetic
acid 2126 (0.254 g, 2.03 mmol), was mixed in dry ether (30 mL)
with trans-crotyl alcohol (0.146 g, 2.03 mmol), dicyclohexyl-
carbodiimide (0.419 g, 2.03 mmol), and a catalytic amount of
p-(N,N-dimethylamino)pyridine (0.5 mg). The mixture was
stirred for 24 h at room temperature. After removal of the
precipitated dicyclohexylurea, the solution was washed with
saturated sodium bicarbonate (2 × 10 mL) solution, dried
(MgSO4), filtered, and evaporated to give 0.431 g of crude
product. This was flash chromatographed on silica gel (ether:
petroleum ether 1/20, 1/11, ether 100%) to give compound 22
as a yellow oil (0.248 g, 68%): IR (CHCl3) cm-1 3100, 2980,
C, 40.99; H, 7.37; N, 8.32.
1-{(SSR,RRS)-3,4-Dih yd r o-3-a m in o-4,5-d im eth ylfu r a n -
2[5H]-on e}-2,5-d ifor m yl P yr r ole (16). Osmium(VIII) oxide
(7.2 mg, 0.028 mmol) and then, in several portions over a 30-
min period, sodium metaperiodate (363.7 mg, 1.70 mmol) were
added at 20 °C to a stirred solution of 17 (122 mg, 0.425 mmol)
in a mixture of dioxane (6.5 mL) and water (3.5 mL). The
reaction was allowed to proceed for 20 h, after which TLC
analysis indicated complete consumption of the starting mate-
rial. The reaction mixture was filtered and extracted thor-
oughly with diethyl ether (50 mL). The combined organic
layers were washed, dried (Na2SO4), and concentrated under
vacuum. Chromatography (2:1; hexane/ethyl acetate, v/v) of
the crude product gave pure 16 as colorless needles (44 mg,
1
44%): mp 121.5-122.5 °C (from hexane); H NMR (CDCl3) δ
1.14 (d, J ) 6.5 Hz, 3H), 1.61 (d, J ) 6.1 Hz, 3H), 2.66 (ddq,
J ) 11.5, 9.5, 6.5 Hz, 1H), 4.30 (dq, J ) 9.5, 6.1, 1H), 6.46 (d,
J ) 11.6 Hz, 1H), 7.08 (d, J ) 4.2 Hz, 1H), 7.15 (d, J ) 4.2 Hz,
1H), 9.75 (s, 1H), 9.84 (s, 1H); 13C NMR (CDCl3) δ 14.3, 18.5,
44.3, 62.0, 80.5, 123.9, 124.4, 136.0, 136.3, 171.2, 181.6, 182.9.
This product was immediately reduced as described below.
1-{(SSR,RRS)-3,4-Dih yd r o-3-a m in o-4,5-d im eth ylfu r a n -
2[5H]-on e}-2,5-bis(isobu ten yl)p yr r ole (17). To a suspen-
sion of 15 (360.4 mg, 2.18 mmol) in 8 mL of dry toluene was
added Ba(OH)2‚8H2O powder (186.5 mg, 1.09 mmol). The
mixture was stirred at 20 °C for 2 h. In a separate flask, a
solution of 1827 (422.5 mg, 2.18 mmol) and titanium tetraiso-
propoxide (631.6 mg, 2.18 mmol) in 8 mL of dry toluene was
stirred at 20 °C. The two solutions were then mixed and heated
under reflux for 24 h. Water was removed by a Dean-Stark
trap. The reaction mixture was cooled to 20 °C and solvent
evaporated under reduced pressure. The brown solid residue
was subjected to chromatography (8% ethyl acetate/hexane,
v/v), to give crude 17 which recrystallized from hexane as a
1
2960, 1725; H NMR (CDCl3) δ 1.70 (d, J ) 7.0 Hz, 3H), 4.55
(d, J ) 6.9 Hz, 2H), 4.60 (s, 2H), 5.58 (m, 1H), 5.79 (m, 1H),
6.19 (d, J ) 2.1 Hz, 2H), 6.65 (d, J ) 2.1 Hz, 2H); 13C NMR
(CDCl3) δ 17.7, 50.6, 66.1, 108.9, 121.6, 124.3, 132.3, 168.5;
MS: 179 [M+].
(SR,RS)-2-(1-P yr r olyl)-3-m eth ylp en t-4-en oic Acid (23).
n-Butyllithium (2.0 M in hexane, 0.747 mL, 1.49 mmol) was
added to diisopropylamine (0.151 g, 0.209 mL, 1.49 mmol) in
dry THF (8 mL) at 0 °C, stirred for 5 min, and cooled to -78
°C, after which ester 22 (0.223 g, 1.25 mmol) in THF (2 mL)
was added dropwise with stirring. After the mixture had been
stirred for 10 min, trimethylsilyl chloride (0.150 g, 0.175 mL,
1.40 mmol) was added, and the solution was stirred for 5 min
before being allowed to warm to 21 °C over 30 min. The
mixture was then warmed to 55-60 °C for 1 h, cooled, dilute
with ether (10 mL), and extracted with 2 N NaOH (4 × 10
mL). The aqueous layer was acidified to pH 2 with concen-
trated hydrochloride acid and extracted with chloroform (4 ×
15 mL). The chloroform extract was dried (MgSO4), and the
solvent was removed under reduced pressure to give 0.144 g
(64%) of 23 as an oily mixture of diastereomers (>95% of 23)
which solidified to a colorless wax on standing: 1H NMR
(CDCl3) δ 1.10 (d, J ) 6.8 Hz, 3H), 3.08 (m, 1H), 4.38 (d, J )
9.1 Hz, 1H), 4.96-5.10 (m, 1H), 6.15 (t, J ) 2.1 Hz, 2H), 6.74
(t, J ) 2.1 Hz, 2H), 11.60 (br s, 1H); minor isomer: 0.87 (d, J
) 6.8 Hz, 3H), 4.28 (d, J ) 10.4 Hz, 1H), 5.08-5.24 (m, 2H),
5.64-5.76 (m, 1H), 6.19 (t, J ) 2.1 Hz, 2H), 6.79 (t, J ) 2.1
Hz, 2H); 13C NMR (CDCl3) δ 16.7, 40.4, 67.0, 108.6, 116.8,
120.7, 137.5, 176.0; minor isomer: 16.4, 41.7, 67.0, 108.9,
117.4, 120.5, 138.0, 175.9; MS: 179 [M+].
(SR,RS)-2-(1-P yr r olyl)-3-m eth ylp en t-4-en oa te (24). Acid
23 (0.134 g, 0.749 mmol) was treated with excess diaz-
omethane dissolved in ether (20 mL) at 0 °C. After 30 min of
stirring, the mixture was kept at room temperature, when
excess of diazomethane was allowed to evaporate. The reaction
mixture (ether layer) was washed with saturated sodium
bicarbonate solution (2 × 15 mL), dried (MgSO4), and evapo-
rated to give 80 mg of 24 (62.5%, based on 15 mg of recovered
acid 23 from the aqueous phase) as a yellow oily mixture of
diastereomers which was relatively unstable and immediately
utilized in attempted iodolactonization: 1H NMR (CDCl3) δ
1.09 (d, J ) 6.6 Hz, 3H), 3.03 (m, 1H), 3.79 (s, 3H), 4.37 (d, J
1
pale yellow solid (148 mg, 24%): mp 71.5-72.5 °C; H NMR
(CDCl3) δ 1.03 (d, J ) 6.6 Hz, 3H), 1.46 (d, J ) 6.2 Hz, 3H),
1.84 (s, 6H), 1.86 (s, 3H), 1.89 (s, 3H), 2.48 (m, 1H), 4.16 (m,
1H), 4.66 (d, J ) 11.8 Hz, 1H), 5.80 (s, 1H), 5.93 (s, 1H), 6.03
(br, 1H), 6.09 (br, 1H); 13C NMR (CDCl3) δ 14.4, 19.3, 20.0,
26.3, 26.4, 44.8, 61.2, 80.0, 108.7, 110.5, 114.4, 115.6, 129.1,
131.7, 136.6, 139.2, 173.6. This was immediately subjected to
osmium tetraoxide-periodate cleavage.
(RSR,SRS)-3,5-Dih yd r o-3-a zid o-5-m eth yl-4-[tr is(p h en -
ylth io)m eth yl]fu r a n -2[5H]-on e (20). To a solution of tris-
(phenylthio)methane (1.56 g, 4.59 mmol) in freshly distilled
THF (10 mL) at -78 °C, was added n-butyllithium (1.96 mL
of 2.5 M solution in hexane). After 30 min, â-angelicalactone25
(0.375 g, 0.343 mL, 3.82 mmol) in THF (5 mL) was added
dropwise during 30 min at -78 °C, and the mixture was stirred
for an additional 4 h at the same temperature. A solution of
trisyl azide (1.53 g, 4.90 mmol) in THF (15 mL) was added to
the lactone solution at -78 °C via a cannula. After 20 min,
the reaction was quenched by the addition of acetic acid (0.88
mL, 15.3 mmol) followed by warming on a water bath to 20
°C over 30 min. After addition of ethyl acetate (25 mL) and
washing with satd sodium bicarbonate (3 × 30 mL), the
organic layer was dried (Na2SO4) and evaporated at reduced
pressure to give a crude pale yellow solid. Flash chromatog-
raphy (ethyl acetate/hexane, 0-50%) yielded (0.819 g, 45%)
compound 20 as a chromatographically pure colorless powder
which was carried to the next step immediately: mp 128-
130 °C; IR (KBr) cm-1 3080, 2990, 2110, 1780, 1475, 1440; 1H
NMR (CDCl3) δ 1.15 (d, J ) 6.45 Hz, 3H), 2.55 (t, 1H), 4.67 (d,
J ) 2.7 Hz, 1H), 4.97 (dq, J ) 2.8, 6.5 Hz, 1H), 7.39 (m, 9H),
7.66 (m, 6H); 13C NMR (CDCl3) δ 22.8, 57.6, 61.3, 77.2, 78.1,
128.9, 130.0, 130.1, 136.3, 172.6.
(SSR,RRS)-3,4-Dih yd r o-3-a m in o-4,5-d im eth ylfu r a n -2-
[5H]-on e (2). In a 50 mL three-necked flask was placed a
solution of 20 (400 mg, 840 mmol) in 15 mL of THF. Raney
nickel (2 g, in 50% water slurry) washed with absolute ethanol
(26) Clemo, G. R.; Ramage, G. R. J . Chem. Soc. 1931, 49.
(27) Ito, Y.; Konoike, T.; Saegusa, T. J . Am. Chem. Soc. 1975, B33,
125.
(28) Ben-Ishai, D.; Altman, J . A.; Bernstein, Z.; Peled, N. Tetrahe-
dron 1978, 34, 467. Pai, N.N. Ph.D Dissertation, Northeastern
University, Boston, 1989.
(25) Thiele, J .; Tischbein, R.; Lossow, E. J ustus Liebigs Ann. Chem.
1901, 319, 191.

Quararibea Metabolites
J . Org. Chem., Vol. 64, No. 8, 1999 2665
) 10.1 Hz, 1H), 4.92-5.10 (m, 1H), 5.47-5.62 (m, 1H), 6.17
(t, J ) 2.1 Hz, 2H), 6.79 (t, J ) 2.1 Hz, 2H); minor isomer:
0.86 (d, J ) 6.8 Hz, 3H), 3.70 (s, 3H), 4.30 (d, J ) 10.6 Hz,
1H), 5.10-5.24 (m, 2H), 5.64-5.77 (m, 1H), 6.20 (t, J ) 2.1
Hz, 2H), 6.83 (t, J ) 2.1 Hz, 2H); 13C NMR (CDCl3) δ 16.6,
40.8, 52.4, 67.1, 108.4, 116.4, 120.6, 137.8, 170.3; minor
isomer: 16.4, 42.1, 52.1, 67.2, 108.6, 116.8, 120.3, 138.4; MS:
193 [M+].
C12H18N2O2: C, 64.84; H, 8.16; N, 12.6. Found: C, 64.60; H,
8.20; N, 12.45.
1-Met h yl-2-(N-b en zyl for m im id oyl)p yr r ole (31). To a
slurry of aldehyde 28 (200 mg, 1.833 mmol) and 5 Å molecular
sieves (0.7 g) in dry CHCl3 (1.8 mL) was added titanium(IV)
isopropoxide (260 mg, 0.916 mmol) dropwise. After stirring for
10 min at 20 °C, benzylamine (196 mg, 1.833 mmol) in dry
CHCl3 (1.8 mL) was introduced to the mixture. The reaction
mixture was stirred at 20 °C for 18 h and filtered through
Celite. The solid was washed with CHCl3 thoroughly (10 mL),
and the chloroform solution was concentrated under reduced
pressure. The residue was purified by column chromatography
on silica gel (ethyl acetate/hexane, 2:1, v/v) to give the mixed
isomers of 31 (301 mg, 83%) as an orange semisolid: mp 28-
1-(SSR,RRS)-3,4-Dih yd r o-3-a m in o-4,5-d im eth ylfu r a n -
2[5H]-on e P yr r ole (25). A mixture of 15 (0.77 g, 4.65 mmol),
sodium acetate (3.480 g, 25.6 mmol), and 2,5-dimethoxytet-
rahydrofuran (0.60 mL, 4.65 mmol) in acetic acid (5 mL) was
refluxed for 1.5 h. After cooling to room temperature, the
reaction mixture was diluted with water (50 mL). The resulting
solution was extracted with ethyl acetate (50 mL, 3 × 25 mL)
and washed with saturated aqueous sodium bicarbonate
solution (3 × 30 mL) and water (25 mL). After drying (Na2-
SO4) the combined organic extracts, evaporation of the solvent
under reduced pressure gave a dark brown oil. Flash column
chromatography (ethyl acetate in hexane, 20-60%) yielded
compound 25 as a pale yellow oil (0.724 g, 87%): IR (CHCl3)
1
30 °C; H NMR (CDCl3) δ 3.96 (s, 3H), 4.72 (s, 2H), 6.18 (dd,
J ) 4.0, 4.8 Hz, 1H), 6.52 (dd, J ) 4.4, 4.8 Hz, 1H), 6.70 (br,
1H), 7.21-7.45 (m, 5H), 8.24 (s, 1H); minor conformer: 3.62,
4.83, 7.80, 8.40; 13C NMR (CDCl3) δ 36.8, 65.5, 108.1, 116.7,
126.8, 127.7, 128.0, 128.4, 129.9, 140.4, 153.4.
1{(SSR,RRS)-3,4-Dih yd r o-3-a m in o-4,5-d im eth ylfu r a n -
2[5H]-on e}-2-for m yl-5-(h yd r oxym eth yl)p yr r ole (5). Com-
pound 16 (24 mg, 0.10 mmol) was dissolved in 1 mL of dioxane
and mixed with 100 µL of HCOOH and 100 mL of H2O. Then
NaBH3CN (6.4 mg, 0.10 mmol) was added, and the mixture
was stirred at 20 °C for 4 h. After neutralization to pH 7.0 by
the addition of NaHCO3 (219 mg), the mixture was extracted
thoroughly with diethyl ether (50 mL). The combined organic
layer was washed with H2O and brine and dried (Na2SO4). The
dried ether solution was evaporated under reduced pressure
to give a yellow solid as crude product (20.4 mg, 84%) which
was further purified by chromatography (1:1, ethyl acetate:
hexane, v/v) to give pure 5 (13 mg, 54%) as colorless crystals,
identical except for optical rotation with the natural product:
1
cm-1 3100, 2950, 1785, 1500, 1200; H NMR (CDCl3) δ 1.09
(d, J ) 6.5 Hz, 3H), 1.43 (d, J ) 6.2 Hz, 3H), 2.25 (m, 1H),
4.11 (dq, J ) 6.5, 3.3 Hz, 1H), 4.43 (d, J ) 12.1 Hz, 1H), 6.20
(t, J ) 2.1 Hz, 2H), 6.66 (t, J ) 2.1 Hz, 2H); 13C NMR (CDCl3)
δ 13.3, 18.2, 46.4, 65.0, 79.2, 109.2, 119.8, 172.3. Anal. Calcd
for C10H12NO2: C, 67.42; H, 6.74; N, 7.86. Found: C, 67.94;
H, 7.46; N, 7.58.
1-{(SSR,RRS)-3,4-Dih yd r o-3-a m in o-4,5-d im eth ylfu r a n -
2[5H]-on e}-2-for m yl P yr r ole (26). To an ice-cooled flask
containing N,N-dimethylformamide (0.85 mL, 11.0 mmol) was
added phosphorus oxychloride (1.0 mL, 11.0 mmol) dropwise.
Stirring was continued at ambient temperature for 30 min,
followed by the addition of a small portion of dichloroethane
(3 mL). The reaction mixture was cooled to 0 °C, and the
pyrrole 25 (0.565 g, 3.15 mmol) in dichloroethane (3 mL) was
added slowly. Following the addition, the reaction mixture was
refluxed for 1 h and cooled to 20 °C. A 10 % aqueous sodium
acetate solution (25 mL) was added dropwise, and the reaction
solution was refluxed for an additional 30 min. After cooling
to 20 °C, the lower layer was removed, and the remaining
aqueous solution was extracted with saturated sodium carbon-
ate solution (4 × 20 mL) and water (15 mL) and dried (Na2-
SO4). Evaporation of the solvent under aspirator pressure gave
a pale brown oil which upon purification by flash column
chromatography (ethyl acetate in hexane, 20-50%) yielded
compound 26 as a pale yellow crystalline solid (0.354 g,
54.2%): mp 77-79 °C; IR (CHCl3) cm-1 3025, 2980, 1790, 1660,
1220; 1H NMR (CDCl3) δ 0.72 (d, J ) 6.5 Hz, 3H), 1.53 (d, J )
6.2 Hz, 3H), 2.74 (m, 1H), 4.41 (dq, J ) 6.5, 3.3 Hz, 1H), 6.33
(dd, J ) 4.0, 2.7 Hz, 1H), 7.03 (dd, J ) 4.0, 1.5 Hz, 1H), 7.06
(m, 1H), 9.52 (s, 1H); minor isomer: 0.62 (d), 1.41 (d), 3.05
(m), 4.81 (m); 13C NMR (CDCl3) δ 13.6, 18.4, 40.8, 59.2, 81.9,
1
mp 129-130 °C (from diethyl ether); H NMR (CDCl3) δ 1.14
(d, J ) 6.6 Hz, 3H), 1.60 (d, J ) 6.6 Hz, 3H), 2.01 (br, 1H),
2.70 (m, 1H), 4.20 (dq, J ) 10.8, 6.8 Hz, 1H), 4.62 (d, J ) 14.2
Hz, 1H), 4.70 (d, J ) 14.2 Hz, 1H), 5.09 (d, J ) 12.0 Hz, 1H),
6.27 (d, J ) 4.0 Hz, 1H), 7.01 (d, J ) 4.0 Hz, 1H), 9.41 (s, 1H);
minor conformer: δ 1.06, 2.53, 4.18, 6.36, 6.63, 6.94, 9.48; 13
C
NMR (CDCl3) δ 14.8, 18.8, 43.6, 56.7, 62.8, 81.0, 111.2, 126.4,
132.3, 142.7, 172.3, 179.0; minor conformer: 14.0, 17.4, 46.7,
61.6, 113.4, 180.5.
(SSR,RRS)-Dih yd r o-3-{2-(h yd r oxym eth yl)-5-[N-[(SSR-
,RRS)-tetr ah ydr o-4,5-dim eth yl-2-oxo-3-fu r yl]for m im idoyl]-
p yr r ol-1-yl}-4,5-d im eth yl-2-(3H)-fu r a n on e (1). A slurry of
5 (98 mg, 0.4135 mmol) and 5 Å molecular sieves (3.8 g) in 2
mL of dry CHCl3 was treated with a solution of 2 (220 mg,
1.705 mmol) in 3 mL of dry CHCl3. After 5 min, Et3N (128 µL,
0.4142 mmol) was added to the mixture. The reaction mixture
was then held at 60-65 °C with stirring for 4 days. After
cooling to 20 °C, the mixture was filtered through diatoma-
ceous earth. The solid was thoroughly washed with CHCl3 (30
mL), and the chloroform solution was concentrated under
reduced pressure. The residue was then subjected to column
chromatography on silica gel (5% methanol in chloroform, v/v)
to give 1 (86.5 mg, 60%), identical in every respect except
optical rotation with funebrine: mp 168-169 °C; 1H NMR
(CDCl3) δ 1.08 (d, J ) 6.5 Hz, 6H), 1.45 (d, J ) 6.3 Hz, 3H),
1.48 (d, J ) 6.2 Hz, 3H), 1.70 (br, 1H), 2.50 (m, 1H), 3.17 (m,
1H), 3.60 (d, J ) 10.8 Hz, 1H), 4.16 (m, 2H), 4.60 (d, J ) 13.8
Hz, 1H), 4.69 (d, J ) 13.7 Hz, 1H), 5.03 (d, J ) 11.7 Hz, 1H),
6.22 (d, J ) 3.8 Hz, 1H), 6.61 (d, J ) 3.8 Hz, 1H), 7.98 (s, 1H);
1H NMR (acetone-d6) δ 1.07 (d, J ) 6.8 Hz, 3H), 1.12 (d, J )
6.8 Hz, 3H), 1.40 (d, J ) 6.8 Hz, 3H), 1.43 (d, J ) 6.8 Hz, 3H),
2.43 (m, 1H), 3.78 (d, J ) 10.9 Hz, 1H), 4.23 (m, 2H), 4.33 (t,
J ) 5.45 Hz, 1H), 4.65 (d, J ) 5.45 Hz, 1H), 5.18 (d, J ) 10.8
Hz, 1H), 6.20 (d, J ) 3.9 Hz, 1H), 6.64 (d, J ) 3.9 Hz, 1H),
8.05 (s, 1H); minor conformer: 1.16 (d), 1.45 (d), 4.00 (d), 6.26
(d), 6.60 (d), 7.43 (d), 8.25 (s); 13C NMR (CDCl3) δ 13.7, 14.6,
18.2, 18.8, 42.7, 46.2, 57.0, 57.1, 63.0, 76.3, 80.7, 80.9, 110.4,
120.5, 130.4, 139.4, 155.3, 172.7, 174.8; ESI-MS (m/z): 349.6
(78.0%) [M + H]+, 331.6 (100%) [M + H - H2O]+; daughter
ions of 349.6:349.1, 331.2, 258.9, 231.3, 220.2, 192.3, 174.4,
160.0, 147.8, 146.4, 132.6, 119.5, 112.9, 94.1, 69.1, 57.0, 43.6,
26.3, 14.9.
110.9, 126.1, 130.8, 131.6, 171.7, 179.4. Anal. Calcd for C11H13
-
NO3: C, 63.76; H, 6.32; N, 6.76. Found: C, 63.60; H, 6.46; N,
6.62.
3-[2′-(1′-Meth ylpyr r ol)m eth ylam in o]-4,5-dim eth ylfu r an -
2[5H]-on e (30). A solution of amino lactone‚HCl 2928 (100 mg,
0.60 mmol) in water (5.0 mL) was neutralized to pH 7 with
K2HPO4, and NaBH3CN (31 mg, 0.50 mmol) was added,
followed by aldehyde 28 (56 mg, 0.50 mmol) dissolved in
methanol (25 mL). The mixture was stirred at 20 °C for 48 h,
diluted with water (10 mL), carefully saturated with NaHCO3,
and immediately extracted with CH2Cl2 (3 × 60 mL). The CH2-
Cl2 solution was dried (Na2SO4), filtered, and evaporated at
aspirator pressure to give 124 mg of crude product. Preparative
TLC (silica gel, ethyl acetate/hexane, 3:7, v/v) of the crude
product gave 9 mg of aldehyde 28 and product 30 (58 mg, 51%)
as a yellowish brown oil: 1H NMR (CDCl3) δ 0.80 (d, J ) 7.0
Hz, 3H), 1.32 (d, J ) 6.8 Hz, 3H), 1.75 (br, 1H), 2.44 (m, 1H),
3.55 (d, J ) 6.8 Hz, 2H), 3.68 (s, 3H), 3.89 (q, J ) 8.2 Hz, 1H),
4.50 (dq, J ) 6.8, 4.2 Hz, 1H), 6.05 (d, J ) 3.0 Hz, 2H), 6.60 (t,
J ) 2.5 Hz, 1H); 13C NMR (CDCl3) δ 7.2 (q), 15.4 (q), 33.7 (q),
38.6 (d), 44.2 (t), 61.5 (d), 76.7 (d), 106.4 (d), 108.7 (d), 122.8
(d), 129.7 (s), 177.2 (s); MS: 222 [M+]. Anal. Calcd for

2666 J . Org. Chem., Vol. 64, No. 8, 1999
Dong et al.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectra of compound 11-17, 20, 22-24, 26, 30-31, (()-1, (()-
5, and 1H NMR spectra of compound 10, (()-2. Dynamic
lineshape analysis data are given for compounds 5 and 16.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Ack n ow led gm en t. We are grateful to Dr. Thomas
Zennie for informing us of his recent study of the 1H
NMR spectrum of natural funebral. We thank the
Barnett Innovative Research Fund and Telor Oph-
thalmic Pharmaceuticals Inc. for financial support,
Professor Paul Vouros and Drs. J ianmei Ding, Meg
Annan, and J im Kyranos for mass spectra, and Profes-
sor David J ebaratnam for discussions.
J O981501U