Further studies of the Daphniphyllum alkaloid polycyclization cascade

Article abstract of DOI:10.1021/jo001145r

The scope of the 2-azadiene intramolecular Diels - Alder cyclization, previously employed for synthesis of the Daphniphyllum alkaloids, has been further investigated. Through a series of 1,5-diol cyclization precursors the substitution pattern of both the dienophile and the 2-azadiene were examined. From these studies it was shown that the cascade reaction is tolerant toward a variety of alkyl-substituted dienophiles. However, it was also demonstrated that this reaction is very sensitive to the substitution pattern of the 2-azadiene. Alterations made to the structure of the 2-azadiene cause either competing side reactions or complete failure of the reaction cascade.

Full text of DOI:10.1021/jo001145r

450
J . Org. Chem. 2001, 66, 450-454
F u r th er Stu d ies of th e Da ph n iph yllu m Alk a loid P olycycliza tion
Ca sca d e
Grier A. Wallace and Clayton H. Heathcock*
Center for New Directions in Organic Synthesis,† Department of Chemistry, University of California,
Berkeley, California 94720
heathcock@cchem.berkeley.edu
Received J uly 28, 2000
The scope of the 2-azadiene intramolecular Diels-Alder cyclization, previously employed for
synthesis of the Daphniphyllum alkaloids, has been further investigated. Through a series of 1,5-
diol cyclization precursors the substitution pattern of both the dienophile and the 2-azadiene were
examined. From these studies it was shown that the cascade reaction is tolerant toward a variety
of alkyl-substituted dienophiles. However, it was also demonstrated that this reaction is very
sensitive to the substitution pattern of the 2-azadiene. Alterations made to the structure of the
2-azadiene cause either competing side reactions or complete failure of the reaction cascade.
Sch em e 1
The Daphniphyllum alkaloids are a group of polycyclic
natural products first isolated from the deciduous tree
Yuzuriha (Daphniphyllum macropodum) in 1909.1 Since
then, over 30 Daphniphyllum alkaloids have been iso-
lated and structurally characterized.2 Methyl homodaph-
niphyllate (1) and methyl homosecodaphniphyllate (2) are
representative members of this group of natural products
and illustrate two of the pentacyclic core structures that
are found.
During the 1980s, we developed the biomimetic ap-
proach to these alkaloids that is illustrated in Scheme
1.3,4 This one-pot procedure begins with the oxidation of
a 1,5-diol (3) to a dialdehyde (4). Treatment of the crude
oxidation mixture with ammonia, followed by acetic acid
and ammonium acetate, leads to the formation of an
azadiene (6), which undergoes an intramolecular Diels-
Alder cyclization to form imine 7. Heating the acetic acid
solution of imine 7 facilitates an intramolecular aza-Prins
cyclization to provide pentacyclic amine 8. This remark-
able process, which forms three carbon-carbon bonds
and two nitrogen-carbon bonds and establishes six
stereocenters, has been used as the key step in the
synthesis of five of the Daphniphyllum alkaloids.3,5,6
In this paper, we report further studies that explore
the scope and generality of the intramolecular 2-azadiene
Diels-Alder cyclization.7 By studying the cyclizations of
diols 9-16 we hoped to examine the effect of both the
† The Center for New Directions in Organic Synthesis is supported
by Bristol-Myers Squibb as Sponsoring Member.
(1) Yagim S. Kyoto Igaku Zasshi. 1909, 6, 208.
(2) For reviews on the Daphniphyllum alkaloids see: (a) Yamamura,
S.; Hirata, Y. In The Alkaloids; Manske, R. H. F., Ed.; Academic
Press: New York, 1975; Vol. 15, p 41. (b) Yamamura, S.; Hirata, Y.
Int. Rev. Sci.: Org. Chem., Ser. Two 1976, 9, 161. (c) Yamamura, S.
In The Alkaloids; Brossi, A., Ed.; Academic Press: New York, 1986;
Vol. 29, p 265.
(3) Heathcock, C. H.; Hansen, M. M.; Ruggeri, R. B.; Kath, J . C. J .
Org. Chem. 1992, 57, 2544.
(4) We have also reported a classical approach to these alkaloids:
Heathcock, C. H.; Davidsen, S. K.; Mills, S. G.; Sanner, M. A. J . Org.
Chem. 1992, 57, 2531.
(5) Heathcock, C. H.; Stafford, J . J . Org. Chem. 1992, 57, 2566.
(6) Heathcock, C. H.; Rugerri, R. B.; McClure, K. F. J . Org. Chem.
1992, 57, 2585.
(7) For references on 2-aza dienes, see: (a) Eddaif, A.; Mison, P.;
Laurent, A.; Pellissier, N.; Carrupt, P.-A.; Vogel, P. J . Org. Chem. 1987,
52, 5548. (b) Cheng, Y.-S.; Ho, E.; Mariano, P. S.; Ammon, H. L. J .
Org. Chem. 1985, 50, 5678. (c) Boger, D. L.; Weinreb, S. M. In Hetero
Diels-Alder Methodology in Organic Synthesis; Academic Press: San
Diego, CA, 1987; pp 255-260. (d) Boger, D. L. Tetrahedron 1983, 39,
2869.
10.1021/jo001145r CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/23/2000

Studies of the Daphniphyllum Alkaloid Polycyclization Cascade
J . Org. Chem., Vol. 66, No. 2, 2001 451
Sch em e 2
Sch em e 3
Sch em e 4
substitution pattern of the 2-azadiene as well as the
pendent dienophile. In addition, the dienophile tether
was extended by one carbon, such that six-membered
rings would be formed during the Diels-Alder cycliza-
tions.
ment of the resulting adduct with iodide 20 provided the
desired adduct 30 in 92% yield (Scheme 3).
Resu lts a n d Discu ssion
Syn th esis of Diols 9-16. The synthesis of diols 9-13
began with the three-component coupling of amides 17
and 188 with enoate 199 and iodides 20-23 (Scheme 2).6
The lithium enolate of amide 17 or 18 was treated with
enoate 19, and the corresponding Michael addition ad-
duct was trapped with iodides 20-23.10 The desired
products could be isolated in good to moderate yields in
all cases except for ester-amide 28. The stereochemical
assignment of the ester-amides 24-27 is based on
literature precedent of similar Michael reactions of amide
enolates.9,11
The failure to obtain ester-amide 28 by this method
was not unexpected, in light of earlier work in these
laboratories.9 Although none of the desired ester-amide
28 was isolated from the reaction of the lithium enolate
of 18 with enoate 19 followed by trapping with iodide 20,
products resulting from 1,2-addition were observed.
However, the lithium enolate of thioamide 29,12 a softer
nucleophile,4,13 reacts smoothly with enoate 19. Treat-
The conversion of amides 24-27 and 30 to their
corresponding diols is shown in Scheme 4. Lithium
triethylborohydride chemoselectively reduced the ester
moiety of these substrates to provide alcohols 31-35 in
good yields. Acid-catalyzed lactonization followed by
lithium aluminum hydride reduction efficiently provided
the desired diols 9-13.
Attempts to prepare diols 14-16 through the three-
component-coupling method described above lead to
intractable mixtures of products. Rather than pursue this
method, we developed the alternate approaches to these
diols shown in Schemes 5 and 6. The synthesis of diols
14 and 15 began with the alkylation of the lithium
enolate of tert-butylpropionate (41) with iodide 20 to
provide ester 42 in high yield. Subsequent allylation of
the lithium enolate of ester 42 with either 2-methylallyl
bromide or allyl bromide provided esters 45 and 46,
respectively. Selective hydroboration of the terminal
olefins in esters 45 and 46, followed by oxidation lead to
(8) Oare, D. A.; Henderson, M. A.; Sanner, M. A.; Heathcock, C. H.
J . Org. Chem. 1990, 55, 132-157.
(12) (a) Oare, D. A. Doctoral Thesis, University of California at
Berkeley, 1988. (b) Thomsen, I.; Clausen, K.; Scheibye, S.; Lawesson,
S. O. Org. Synth. 1984, 62, 158.
(13) (a) Pearson, R. G. J . Chem. Educ. 1968, 45, 581, 643. (b)
Pearson, R. G. J . Am. Chem. Soc. 1963, 85, 3533. (c) Deschamps, B.;
Anh, N. T.; Seyden-Penne, J . Tetrahedron Lett. 1973, 14, 527.
(9) Ruggeri, R. B.; Hansen, M. M.; Heathcock, C. H. J . Am. Chem.
Soc. 1988, 110, 8734.
(10) Millar, J . G.; Underhill, E. W. J . Org. Chem. 1986, 51, 4726.
(11) Yamaguchi, M. Yuki Gosei Kagaka. 1986, 44, 405.

452 J . Org. Chem., Vol. 66, No. 2, 2001
Wallace and Heathcock
Sch em e 5
Sch em e 7
Sch em e 6
Sch em e 8
the imine products were isolated by column chromatog-
raphy on silica gel. The cyclization sequences of diols
9-12 are shown in Scheme 7.
Oxidation of diols 9-12 led smoothly to the 1,5-
dialdehydes 52-55. These delicate molecules can be
observed in the crude form by 1H NMR, following an
aqueous workup of the Moffatt-Swern oxidation. How-
ever, as noted above, the reaction mixtures were more
routinely treated directly with ammonia, followed by
solvent exchange to acetic acid. Following an aqueous
workup, imines 60-63 were isolated in yields ranging
from 69 to 81% by column chromatography on silica gel
that had been pretreated with triethylamine. The struc-
tures of imines 60-63 were determined through 1H
NMR, 13C NMR, 13C DEPT, IR, and elemental analysis.
In addition to the spectral and analytical evidence that
supported the assigned structures of tetracyclic imines
60-63, the structure of 62 was further confirmed by
conversion into derivative 64, the structure of which was
rigorously determined by X-ray crystallography (Scheme
8).15
the primary alcohols 47 and 48. Reduction of the ester
function in 47 and 48 gave access to the desired diols 14
and 15.
The synthesis of diol 16 was carried out in two steps
as shown in Scheme 6. Alkylation of the lithium enolate
of δ-valerolactone with iodide 20 provided lactone 50 in
low and variable yields (25-50%). We believe that ring-
opening of the lactone 50 and oligomerization to polyester
51 was responsible for the low and variable yields of 50.
Support for this hypothesis came from the observation
that reduction of the crude reaction mixture with LAH
provided the desired diol 16 in a reproducible 69% yield.
Exp lor a tion of th e Aza d ien e Diels-Ald er Cyliza -
tion . With diols 9-16 in hand, we were prepared to
commence with our studies of the scope and generality
of the Daphniphyllum alkaloid azadiene Diels-Alder
cyclization. From the onset of this project, we planned
to use the biomimetic reaction protocol developed during
the Daphniphyllum alkaloid synthesis, rather than at-
tempting to optimize the reaction conditions for each
substrate. This protocol involves Moffatt-Swern14 oxida-
tion of the 1,5-diol to the dialdehyde, treatment of the
crude methylene chloride solution with ammonia followed
by solvent exchange from methylene chloride to a buff-
ered acetic acid solution. Following an aqueous workup,
In the course of performing the cyclizations of diols
9-12, we noted a definite trend in the reaction rates of
these substrates (imines 60 and 61 can be formed at room
temperature, whereas imines 62 and 63 require heating
at 80 °C). Qualitatively, the rates of these azadiene
(15) Crystallographic data (excluding structure factors) for the
structures in this paper have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication nos. CCDC
147538. Copies of the data can be obtained, free of charge, on
application to CCDC, 12 Union Road, Cambridge CB2 1EZ, U.K. (fax:
+44 1223 336033 or e-mail: deposit@ccdc.cam.ac.uk).
(14) Mancusco, A. J .; Swern, D. Synthesis 1981, 165.

Studies of the Daphniphyllum Alkaloid Polycyclization Cascade
J . Org. Chem., Vol. 66, No. 2, 2001 453
Sch em e 9
Sch em e 10
Sch em e 11
cyclizations follow the order 56 > 57 > 58 > 59. This
observation supports a rate-limiting inverse-electron-
demand Diels-Alder reaction mechanism in which the
more electron-rich dienophiles provide faster reaction
rates. However, it was also noted that azadiene 57
cyclizes at a faster rate than azadiene 58, even though
both substrates have disubstituted dienophiles. We be-
lieve the steric congestion associated with cyclizing the
Z-olefin of azadiene 58 is the cause of this marked
decrease in reaction rate. In addition, it was noted that
the olefin geometry of diols 10 and 11 is conserved during
the reaction providing imines 61 and 62 respectively,
further supporting a concerted mechanism for the Diels-
Alder step of the cascade. These experiments also dem-
onstrate that the cascade cyclization works well when
the tether length is such that a six-membered ring is
formed (the examples previously demonstrated in our
Daphniphyllum alkaloid syntheses all give rise to five-
membered rings). These cyclizations also show that the
cascade succeeds with substrates having various alkyl
substitution patterns on the dienophile, ranging from
mono- to trisubstituted. At this point we turned our
attention to the structure of the aza-diene intermediate.
The cyclization sequence of diol 13 is shown in Scheme
9. Treatment of diol 13 under the standard cyclization
conditions provided imine 67 in a rather disappointing
23% yield. This was intriguing because the only differ-
ence between diol 13 and previous diols is the methyl
stereocenter alpha to one of the aldehydes. At this point
attempts were then made to determine at which stage
the reaction cascade was faltering. Oxidation of diol 13
under Moffatt-Swern conditions followed by an aqueous
work up provided the crude 1,5-dialdehyde 65, verifying
the efficiency of this step of the cascade. Repeating the
cyclization protocol and stopping after treatment of
dialdehyde 65 with ammonia provided a complex mixture
of products that has been assigned as the various
bisaminal and bishemiaminal structural isomers related
to compound 66. Treatment of this mixture with acetic
acid and ammonium acetate again provided imine 67 in
low yield, along with intractable polymeric material. To
check the stability of imine 67 to the reaction conditions
it was taken up in D-4 acetic acid and heated for 50 h at
80 °C, at which time there was no sign of decomposition
as judged by 1H NMR. From these data we believe
inefficient azadiene formation is responsible for the low
yield in the cyclization of diol 13.
We next turned our attention to the cyclization of diols
14-16, in which the five-membered ring of diols 9-13
is absent. Treatment of diol 14 under the standard
cyclization conditions provided the desired tricyclic imine
69 in 51% yield along with bicyclic imine-hydroperoxide
70 in 10-20% yield (Scheme 10). The structure of imine-
hydroperoxide 70 was tentatively assigned on the basis
1
of H NMR, IR, and mass spectroscopy. This compound
was very sensitive; partially decomposing when exposed
to silica gel chromatography as well as on storage at 0
°C under a nitrogen atmosphere. We believe that imine-
hydroperoxide 70 arises from autoxidation of the corre-
sponding imine on exposure to air during the workup of
the reaction.16 This cyclization was particularly interest-
ing because we had not previously been able to isolate
and identify any side products from the reaction cascade.
Because of the sensitive nature of imine-hydroperoxide
70 we decided to modify the cyclization protocol in order
to produce a more stable product. Attempts to treat 70
with reducing agents such as trimethyl phosphite or
triphenylphosphine in order to reduce the peroxide
moiety were unsuccessful, resulting in either no reaction
or decomposition of the starting material. Based on these
results, we decided to investigate the alternative ap-
proach shown in Scheme 11. The idea was that it may
be possible to intercept the precursor to 70 by acylation
of the imine nitrogen,17 thus producing an enamide that
would be less prone to autoxidation. In the event,
treatment of the crude cyclization reaction mixture with
(16) Kleinman, E. F. Doctoral Thesis, University of California at
Berkeley, 1980.
(17) Meth-Cohn, O.; Westwood, K. T. J . Chem. Soc., Perkin Trans.
1 1984, 6, 1173-1182.

454 J . Org. Chem., Vol. 66, No. 2, 2001
Wallace and Heathcock
Sch em e 12
an excess of acetic anhydride provided a mixture of
amides 72 (40%) and 73 (13%). As expected, enamide 73
was not prone to autoxidation and was fully character-
ized. The relative stereochemistry of 73 was established
by 1D and 2D (NOESY) 1H NMR spectroscopy. In
addition, amide 72 is a crystalline solid and its structure
was unambiguously determined by X-ray crystallogra-
phy.18
With the structures of 72 and 73 firmly established
we considered the factors governing their competing
formation. Initial experiments were performed to deter-
mine whether imines 69 and 71 can be interconverted.
Because imine 69 was isolable it was easily resubjected
to the reaction conditions and shown to be stable. As
noted above, imine 71 was not easily isolable. To circum-
vent this difficulty the cyclization reaction was carried
out for various times and the ratio of amides 72 and 73
was measured. Extending the reaction time had little
effect on either the yield or ratio of amides 72 and 73.
From these data we conclude that imines 69 and 71 are
both kinetic products of this reaction. Interestingly,
removal of the five-membered ring of diol 9 generates a
system in which the aza-Diels-Alder cyclization is only
slightly more favorable than the aza-Prins cyclization.
Through our efforts it has been possible to observe the
1,5-dialdehyde 68, but we have been unable to observe
the corresponding azadiene intermediate in this reaction.
Therefore, it is presently unclear at which point in the
cyclization cascade the paths leading to imines 69 and
71 diverge.
that inefficient aza-diene formation is responsible for the
failure of these reactions.
Con clu sion
From the above studies we have been able to expand
the scope of the Daphniphyllum alkaloids cyclization,
while defining some of the limitations of this cascade
reaction. The cyclization is very permissive to various
alkyl substitution patterns of the dienophile, but a
marked decrease in rate is noted when relatively electron
deficient dienophiles are employed. In addition, it has
been demonstrated that the structure of the 2-azadiene
is crucial. The cyclopentyl ring, quaternary carbon and
tertiary carbon centers in the diol starting material all
play a role in providing a selective and high-yielding
cyclization.
We also attempted the cascade cyclization with diols
15 and 16 (Scheme 12). Both of these substrates failed
to produce any tractable products. This was not unex-
pected based on the results gained from diol 13. Again,
aldehydes 74 and 75 could be observed in their crude
form, confirming the efficiency of the oxidation step in
the cascade reaction. As previously discussed, we believe
Ack n ow led gm en t. This work was supported by a
research grant from the United States Public Health
Service Institute of General Medical Studies (GM46057).
We are also grateful to Elf Atochem for a predoctoral
fellowship supporting G.A.W.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and characterization for all new compounds reported
in this manuscript. This material is available free of charge
via the Internet at http://pubs.acs.org.
(18) Crystallographic data (excluding structure factors) for the
structures in this paper have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication nos. CCDC
147539. Copies of the data can be obtained, free of charge, on
application to CCDC, 12 Union Road, Cambridge CB2 1EZ, U.K. (fax:
+44 1223 336033 or e-mail: deposit@ccdc.cam.ac.uk).
J O001145R