A unified strategy for the synthesis of phenanthroizidine alkaloids: Preparation of sterically congested pyridines

Article abstract of DOI:10.1021/ja962757p

Total syntheses of the representative phenanthroizidine alkaloids, tylophorine and antofine, and of their seco congeners, septicine and julandine, have been completed from sterically congested 3,4-diarylpyridines prepared by a modified Knoevenagel-Stobbe

Full text of DOI:10.1021/ja962757p

12082
J. Am. Chem. Soc. 1996, 118, 12082-12089
A Unified Strategy for the Synthesis of Phenanthroizidine
Alkaloids: Preparation of Sterically Congested Pyridines
Marco A. Ciufolini* and Frank Roschangar
Contribution from the Department of Chemistry, Rice UniVersity, P.O. Box 1892,
Houston, Texas 77251
ReceiVed August 7, 1996X
Abstract: Total syntheses of the representative phenanthroizidine alkaloids, tylophorine and antofine, and of their
seco congeners, septicine and julandine, have been completed from sterically congested 3,4-diarylpyridines prepared
by a modified Knoevenagel-Stobbe synthesis. Central to the success of this effort was the ability of R-dicarbonyl
enones to combine in a formal [4 + 2]-cycloaddition with sterically demanding vinyl ethers.
Introduction
Scheme 1
Phenanthroizidine alkaloids are found primarily in plants
belonging to the Asclepiadaceae family.1 Most members of
this small group of natural products display a phenanthroin-
dolizidine framework, e.g., tylophorine, 1, and antofine, 2,
though more rare phenanthroquinolizidines such as cryptopleu-
rine, 3, are also known. Seco congeners of 1-3, e.g. septicine,
4, and julandine, 5, occur in the same plant family and indeed,
they are the putative biogenetic precursors of phenanthroizidines
(Scheme 1).2 It is also well recognized that oxidative conversion
of the seco alkaloids to full-fledged phenanthroizidines may be
readily accomplished by chemical methods; therefore seco
substances may be also regarded as synthetic precursors to 1-3.
Phenanthroizidines exhibit interesting biological properties1
ranging from antitumor activity,2 seemingly through inhibition
of protein synthesis, to nerve growth stimulation, to unusual
cardiovascular and immunological effects, and to probable
antiinflammatory action,3 and as a consequence of their bioac-
tivity and unusual architecture they have engendered a great
deal of synthetic work.4
Scheme 2
Alkaloids 1-5 appear to be within the scope of our pyridine
synthesis, which can furnish end-products 9 with many permu-
tations of groups R1-R5 (Scheme 2).5 The retrosynthetic
X Abstract published in AdVance ACS Abstracts, November 1, 1996.
(1) Reviews: (a) Gellert, E. In Alkaloids: Chemical and Biological
PerspectiVes; Pelletier, S. W., Ed.; John Wiley & Sons: New York, New
York, 1987; Vol. 5, p 55. (b) Suffness, M.; Cordell, G. A. The Alkaloids;
Brossi, A., Ed.; Academic Press, Orlando, FL, 1985; Vol. 25, p 156; (c)
Bick, R. C.; Sinchai, W. In The Alkaloids; Rodrigo, R. G. A., Ed.; Academic
Press: New York, 1981; Vol. 19, p 193. (d) Govindachari, T. R.;
Viswanathan, N. Heterocycles 1978, 11, 587. (e) Govindachari, T. R. In
The Alkaloids; Manske, R. H. F., Ed.; Academic Press: New York, 1967;
Vol. 9, p 517.
(2) Detailed discussion: Suffness, M.; Cordell, G. A. In The Alkaloids;
Brossi, A., Ed.; Academic Press: Orlando, FL, 1985; Vol. 25, ch. 1, pp
3-355.
(3) Duan, J.; Yu, L. Zhongcaoyao 1991, 22, 316 (Chem. Abstr. 1992,
116, 386u).
(4) The literature cited in ref 2 contains extensive discussion of earlier
syntheses of these alkaloids. More recent reports containing new strategic
ideas: (a) Khatri, N. A.; Schmitthenner, H. F.; Shringarpure, J.; Weinreb,
S. M. J. Am. Chem. Soc. 1981, 103, 6387. (b) Cragg, J. E.; Hedges, S. H.;
Herbert, R. B. Tetrahedron Lett. 1981, 22, 2127. (c) Bremmer, M. I.; Khatri,
N. A.; Weinreb, S. M. J. Org. Chem. 1983, 48, 3661. (d) Iwao, M.;
Mahalanabis, K. K.; Watanabe, M.; de Silva, S. O.; Snieckus, V.
Tetrahedron 1983, 39, 1955. (e) Iida, H.; Watanabe, Y.; Tanaka, M.;
Kibayashi, C. J. Org. Chem. 1984, 49, 2412. (f) Iwasa, K.; Kamigauchi,
M.; Takao, N. J. Nat. Prod. 1988, 51, 172. (g) Grieco, P. A.; Parker, D. T.
J. Org. Chem. 1988, 53, 3325. (h) Ihara, M.; Takino, Y.; Tomotake, M.;
Fukumoto, K. J. Chem. Soc., Perkin Trans. 1 1990, 2287. (i) Snieckus, V.
Chem. ReV. 1990, 90, 879. (j) Snieckus, V. Pure Appl. Chem. 1990, 62,
2047. (k) Comins, D. L.; Morgan, L. A. Tetrahedron Lett. 1991, 32, 5919.
construct of Scheme 3 shows that a pyridine 12 suitable for the
creation of a phenanthroizidine 10 could arise through the
merger of enone 13 with ether 14, followed, in some appropriate
order, by izidine formation, controlled reduction of the pyridine,
and oxidative cyclization. Problems and solutions pertaining
to the present formulation are detailed herein.
Background
The hypothesis of Scheme 3 was initially explored by
attempting the cycloaddition6 of benzalacetone, 21, with ether
(l) Yerxa, B. R.; Yang, K.; Moore, H. W. Tetrahedron 1994, 50, 6173. (m)
Pearson, W. H.; Walavalkar, R. Tetrahedron 1994, 50, 12293. This last
work contains an extensive bibliography of synthetic activity in the
phenanthroizidine area.
(5) Cf. Ciufolini, M. A.; Shen, Y.-C.; Bishop, M. J. J. Am. Chem. Soc.
1995, 117, 12460, and references cited therein. See also ref 10.
S0002-7863(96)02757-6 CCC: $12.00 © 1996 American Chemical Society

Synthesis of Phenanthroizidine Alkaloids
J. Am. Chem. Soc., Vol. 118, No. 48, 1996 12083
Scheme 3
Scheme 5
Scheme 4
17, obtained as a mixture of geometric isomers in 90% distilled
yield by Wittig reaction of veratraldehyde with (methoxymeth-
ylene)triphenylphosphorane (Scheme 4). A similar protocol
furnished ether 18 from anisaldehyde. This second ether was
later to be used in the synthesis of alkaloids 2 and 5.
Disappointingly, addends 21 and 17 failed to combine under
our standard catalytic conditions.
In an effort to determine whether this might be due to
electronic problems, to steric effects, or to a combination of
both, we examined electronic aspects of the reaction at the
EHMO level on MM+ structures. Under standard approxima-
tions,7 the HOMO(ether)-LUMO(enone) energy gap is a good
predictor of relative reaction rates, barring adverse steric
interactions. Results of an estimate of molecular orbital energies
for relevant enones and vinyl ethers are summarized in Scheme
5.8 It is apparent that the failure of 17 to combine with 21
does not augur well for the feasibility of a cycloaddtion of the
ether with veratralacetone, 23, (greater HOMO-LUMO gap),
a computationally convenient analogue of an enone, wherein a
short chain suitable for completion of the izidine framework
would be present in lieu of a methyl group. However, concern
about unfavorable electronics (i.e., excessive HOMO-LUMO
gap) was dispelled by the experimental observation that ben-
zalacetone does react, albeit slowly, with ethyl vinyl ether (EVE,
19) under our standard cycloaddition protocol, even though the
HOMO of EVE is almost 1 eV lower in energy than that of 17.
It thus seemed that in all likelihood our initial attempt had failed
primarily because of unfavorable steric interactions between the
aryl groups of the cycloaddends at or near the transition state
for the reaction.
manner consonant with our ultimate synthetic goal. This
objective seemed attainable by buttressing the electron-
withdrawing power of the carbonyl subunit directly connected
to the olefin. To that end, we considered Knoevenagel-type
enone 22 and the isomeric R-diketone 20. Not only would
enone 20 ultimately be more suitable as a synthetic precursor
to 1-5, since its substitution pattern better coincides with that
of intermediate 12, but its LUMO energy turned out to
appreciably lower than that of 22.9 Our calculations also
revealed that the LUMO energy of 20 drops from -10.44 eV
to -10.94 eV when the dihedral angle between the carbonyls
tends to 0° (conformer 20a). Therefore, it may be speculated
that if activation of this enone by Yb(fod)3 in the course of the
cycloaddition were to involve a chelating interaction of the type
shown in structure 20b, the combined effects of conformation
and Lewis acid coordination would bring the LUMO of the
molecule to within less than 1 eV away from the HOMO of
ether 17. This should greatly favor the cycloaddition step.
Enone 20 necessary to test our prediction was prepared by
Wadsworth-Emmons reaction of 15 with phosphonate 24,10
followed by acid hydrolysis (Scheme 6). Indeed, diketone 20
combined slowly (2 days), but cleanly, with the mixture of
geometric isomers of 17 to furnish almost exclusively the all-
Such a problem may be remedied by decreasing the steric
demand of the addends, through replacement of one or both
aryl groups with smaller implements that would later allow
arylation of the intermediate pyridine. However, cures of this
type were unattractive, because they would substantially lengthen
the synthesis. A better approach was to augment the reactivity
of the enone through drastic lowering of its LUMO, but in a
(9) This may reflect the fact that in structure 20 the enone system is
essentially planar, though the dihedral angle between the vicinal carbonyl
groups is approximately 129°; therefore, excellent conjugations exists
throughout the π network involved in the cycloaddition. By contrast,
significant out-of-plane distortion of aryl, olefin, and carbonyl groups is
apparent in the MM+ structure of 22: conjugation is thus likely to be
disrupted. Indeed, the LUMO energy of 22 is identical to that of 21,
indicating that the activation provided by the second carbonyl is barely
sufficient to counterbalance the presence of two electron-releasing OMe
groups on the phenyl ring.
(6) The term “cycloaddition” is used throughout this paper only to
describe the gross outcome of reactions between R,â-unsaturated carbonyl
compounds and vinyl ethers. No mechanistic inferences should be drawn
from such usage, particularly with respect to concertedeness or other intimate
electronic details of the reactions.
(7) Because the reaction proceeds with inverse demand, the major frontier
orbital interaction is assumed to be the one between the HOMO of the
electron-rich ether and the LUMO of the electron-deficient enone.
(8) Calculations were carried out with the Hyperchem package (version
4.0) available from Hypercube, Inc., Ontario, Canada.
(10) Ciufolini, M. A.; Bishop, M. J. J. Chem. Soc., Chem. Commun.
1993, 1463.

12084 J. Am. Chem. Soc., Vol. 118, No. 48, 1996
Ciufolini and Roschangar
Scheme 6
Second, reaction of both 29a and 29b with hydroxylamine
hydrochloride produced small amounts (ca. 5% of total product)
of abnormal pyridines 37a and 37b, respectively. Their structure
was assigned on the basis of spectral evidence summarized in
Scheme 10.14 Our mechanistic rationale for the formation of
37 envision competition between simple deprotonation (path a,
major) and aryl migration/deprotonation (path b, minor) during
dehydrative aromatization of intermediate 36, which may arise
as indicated earlier in Scheme 9.15
Scheme 7
We conclude this section by pointing out that the introduction
of an R-dicarbonyl motif as a means to increase the reactivity
of an enone that is otherwise disinclined to undergo lanthanide-
mediated heterocycloaddition appears to be a general solution.
For instance, enone 38 refused to combine with 1-ethoxypropene
under our standard conditions, but compound 39 reacted
smoothly to form adduct 40, and thence pyridine 41 in good
overall yield (Scheme 11).16
Synthesis of Phenanthroizidines and Their Seco
Congeners
trans diastereomer of pyran 26. The stereochemistry of this
material rests on the magnitude of the homonuclear vicinal
coupling constants between the ring hydrogens (Scheme 7).
Relative to the “anomeric” methoxy group, compound 26 may
be regarded as the endo adduct of the enone with the (E)-isomer
of ether 17. The formation of an adduct arising from substan-
tially only one isomer of the ether reaffirms the principle that
facile isomerization of vinyl ethers may occur under lanthanide-
promoted cycloadditions and ene-like processes, and that often,
one particular regio- or stereoisomer of the ether is seen as
reacting considerably faster than the other(s).11 In sharp
contrast, reaction of 20 with ether 18 completed much faster (6
h), but produced pyran 27 as a mixture of diastereomers.
Fortunately, this was inconsequential, because conversion of
27 to a pyridine would ultimately eliminate all stereocenters.
With quantities of vinylpyridines 33a,b now secured, we set
out to reach some of the natural products. Access to phenan-
throindolizidines dictated the introduction of a C1 unit. To that
end, exposure of 33a to a mixture of LiCN in hot DMF
containing a small amount of AcOH (in situ formation of HCN)
induced smooth hydrocyanation to nitrile 42. Completion of
the molecular framework was accomplished in a straightforward
manner through acidic hydrolysis of the nitrile to an acid, LAH
reduction to the alcohol, and cyclization to pyridinium salt 45.
The precise nature of the anion in 45 was not defined, hence
X- may be either Cl- or MsO-. Indeed, this salt was not
characterized beyond a proton NMR spectrum; rather, it was
reduced with NaBH4 in hot ethanol to afford fully synthetic
(()-septicine, 4, which was converted to totally synthetic (()-
tylophorine, 1, in 74% recrystallized yield via a Liepa-type
oxidation with vanadium(V) oxyfluoride in the presence of
trifluoroacetic acid.17 We found that the success of the Liepa
protocol depended critically on the precise order of addition of
the reagents, an aspect of this chemistry that does not appear
to have been explicitly addressed in the literature. It was crucial
to introduce the TFA last, 5 to 10 min after VOF3 had been
added: reversal of the order of addition led to total destruction
of the 1,2-diarylethylene derivatives. In a like fashion and
yields, vinylpyridine 33b was advanced to compound 50, which
underwent oxidative cyclization to synthetic (()-antofine, 2,
in 51% recrystallized yield (Scheme 12).
A Common Intermediate for Phenanthroizidines Emerges
Pyran 26 formed pyridine oxime 34 when refluxed in a
suspension of hydroxylamine hydrochloride in acetonitrile. We
were unable to retrieve the more synthetically useful ketone 35
from 34 in a clean fashion; accordingly, reduction of 26/27 and
benzoylation of the intermediate alcohol were conducted prior
to pyridine formation, resulting in the ultimate creation of
heterocycles 30a (X ) OMe) and 30b (X ) H). Vinylpyridines
of the type 33 were soon identified as especially good
intermediates for side-chain elongation, due to the well-known
propensity of their olefinic linkage to undergo nucleophilic
addition.12 Either indolizidines or quinolizidines would there-
fore be made from 33 through addition of a C1 or a C2 unit,
respectively. The vinylpyridines emerged in good overall yield
from 30a,b through a sequence involving benzoate ester
cleavage, chloride formation by reaction of the intermediate
alcohol with SOCl2, and E2-type reaction with tBuOK as the
base (Scheme 8).
Addition of a C2 unit in the interest of creating an ultimate
quinolizidine was accomplished through addition of the enolate
of acetonitrile to 33b. This operation was best effected in
DMSO at room temperature with excess acetonitrile and tBuOK
as the base, and it furnished compound 51 in high yield. A
sequence identical to that shown in Scheme 12 above advanced
51 to fully synthetic (()-julandine, 5. Methods for the oxidative
cyclization of julandine to cryptopleurine are known,18 therefore
a synthesis of 5 amounts also to a formal synthesis of 3.19
Two aspects of the chemistry of Scheme 8 deserve further
comment. First, benzoylation of intermediates 28 was necessary
because substantial quantities of oxime 34 again resulted when
pyridine formation was attempted from the free alcohols. The
oxime probably forms through events that parallel those leading
to osazone formation in carbohydrate chemistry13 (Scheme 9).
(14) We thank Dr. Larry Alemany, of this department, for valuable
assistance with the NOE measurements.
(15) The abnormal products displayed the same chromatographic mobility
as the major pyridines under a variety of conditions, rendering their isolation
quite arduous, but they were found to undergo benzoate cleavage at a
considerably slower rate than “normal” pyridines 30a,b, probably for steric
reasons, and were thus separated from the major product at the stage of
alcohols 31.
(16) Bishop, M. J. Unpublished results from these laboratories.
(17) Liepa, A. J.; Summons, R. E. J. Chem. Soc., Chem. Commun. 1977,
826.
(11) Ciufolini, M. A. In AdVances in Heterocyclic Natural Product
Synthesis; Pearson. W. H., Ed.; JAI Press: Greenwich, CT, 1996; Vol. 3,
ch. 1, pp 1-55. See also ref 10.
(12) Cf., e.g., Danishefsky, S.; Cain, P. J. Am. Chem. Soc. 1975, 97,
5282.
(13) Discussion: Jones, J. K. N. In Chemistry of Carbon Coumpounds;
Rodd, E. H., Ed.; Elsevier: Amsterdam, The Netherlands, 1952; Vol. I,
Part B, Ch. 19; see especially p 1238.
(18) Cragg, J. E.; Herbert, R. B. J. Chem. Soc., Perkin Trans. 1 1992,
2487.

Synthesis of Phenanthroizidine Alkaloids
J. Am. Chem. Soc., Vol. 118, No. 48, 1996 12085
Scheme 8
Scheme 9
Scheme 11
Scheme 10
a unified approach to a number of phenanthroizidine alkaloids
and their seco relatives. The scope of the methodology, the
value of which has already been demonstrated in connection
with the total synthesis of diverse heterocyclic natural products,20
is thus further expanded.
Experimental Section21
Vinyl Ether 17. tert-Butyllithium (1.7 M in pentane, 104 mL, 176
mmol) was added to a cold (-78 °C) suspension of Ph3P+CH2OMe
Cl- (61.7 g, 180.1 mmol) in THF (100 mL) and the mixture was
warmed to 0 °C. After 2 h, veratraldehyde (14.3 g, 85.5 mmol) in
THF (30 mL) was added. The mixture was refluxed until the reaction
was completed (TLC, ca. 12 h), and then it was cooled to room temp,
quenched (H2O), and extracted (hexanes). The combined extracts were
sequentially washed with H2O, saturated aqueous NaHCO3, H2O, and
saturated aqueous NaCl, dried (Na2SO4), and concentrated. Chroma-
tography with 1% f 20% EtOAc/hexanes followed by concentration
and Kugelrohr distillation gave 15.0 g (90%) of 17, colorless oil. 1H
(major peaks only, cis:trans ∼ 76:24): 7.25-7.24 (d, J ) 1.7 Hz),
(20) (a) Ciufolini, M. A.; Byrne, N. E. J. Am. Chem. Soc. 1991, 113,
8016 (cystodytins). (b) Bishop, M. J.; Ciufolini, M. A. J. Am. Chem. Soc.
1992, 114, 10081 (dercitins and kuanoniamines). (c) Ciufolini, M. A.; Shen,
Y.-C. Tetrahedron Lett. 1995, 36, 4709 (diplamine and shermilamine), as
well as refs 5, 10 (lavendamycin methyl ester), and 11.
In summary, we have shown that our pyridine-forming
reaction may be used to prepare sterically congested 3,4-
diarylpyridine, and therefore it can play a central role in defining
(21) Experimental Protocols. Melting points (uncorrected) were measured
on a Fischer-Johns hot stage apparatus. NMR spectra (ppm on the δ scale)
(19) It was not possible to compare the spectra of our synthetic final
products with spectra of authentic natural materials. However, all of these
alkaloids have been synthesized several times, and physical and spectral
data obtained by us were in complete agreement with those reported by
previous workers. As further detailed in ref 22, controversy exists regarding
the melting point (but not the spectral properties) of (()-tylophorine,
literature values for which span almost 30 °C.
1
were recorded at room temperature at 250 MHz for H and 62.5 MHz for
13C in CDCl3 solutions. Analytical and preparative TLC employed Merck
SiO2 60 plates with fluorescent indicator. THF was distilled from Na/Ph2-
CO; DCE, DME, pyridine, and CH2Cl2 were distilled from CaH2; DMSO
was vacuum distilled from CaH2; MeOH was dried over 4 Å mol sieves.
All other reagents and solvents were used as received.

12086 J. Am. Chem. Soc., Vol. 118, No. 48, 1996
Ciufolini and Roschangar
Scheme 12
NaCl, dried (Na2SO4), and concentrated. Chromatography (20%
EtOAc/hexanes) provided 10.9 g (94%) of 25, yellow oil. 1H: 7.76-
7.70 (d, 1H, J ) 15.9 Hz), 7.20-7.16 (dd, 1H, J1 ) 8.3 Hz, J2 ) 1.9
Hz), 7.12-7.05 (d, 1H, J ) 15.9 Hz), 7.10-7.09 (d, 1H, J ) 2.0 Hz),
6.87-6.83 (d, 1H, J ) 8.3 Hz), 3.91 (s, 3H), 3.89 (s, 3H), 1.44 (s,
3H). 13C: 196.6, 151.5, 149.1, 145.1, 127.5, 123.6, 118.0, 110.9, 110.0,
102.3, 55.9, 49.7, 20.1. IR: 1693, 1593, 1511, 1463, 1265, 1140, 1042.
MS: 280 (M+), 249, 221, 191, 89 (100%). HRMS: Calcd for
C15H20O5: 280.1310 (M+). Found: 280.1313 (M+).
Scheme 13
Diketone 20. A mixture of 25 (1.135 g, 4.05 mmol), CHCl3 (20
mL) and 70% aqueous trifluoroacetic acid (20 mL) was stirred at 0 °C
for 2 h and then poured into saturated aqueous NaCl and extracted
with CHCl3. The extracts were sequentially washed with H2O and
saturated aqueous NaCl, dried (Na2SO4), and concentrated. The residue
was recrystallized from toluene/hexanes to provide 718 mg (76%) of
20, bright yellow solid, mp 124 °C. 1H: 7.75-7.69 (d, 2H, J ) 16.3
Hz), 7.20-7.09 (m, 3H), 6.84-6.81 (d, 1H, J ) 8.3 Hz), 3.88 (s, 3H),
3.87 (s, 3H), 2.39 (s, 3H). 13C: 199.1, 186.3, 152.1, 149.2, 147.7,
127.3, 124.3, 115.4, 110.9, 109.8, 55.8 (2 signals), 24.3: IR: 1711,
1671, 1576, 1510, 1265, 1244. MS: 234 (M+), 191 (100%). HRMS:
Calcd for C13H14O4: 234.0892 (M+). Found: 234.0891 (M+) EA:
(Calcd) C: 66.52 (66.66); H: 6.13 (6.02).
7.09-7.05 (dd, J1 ) 8.3 Hz, J2 ) 1.8 Hz), 6.97-6.92 (d, J ) 12.9
Hz), 6.81 (s) 6.78 (s), 6.07-6.05 (d, J ) 7.0 Hz), 5.80-5.75 (d, J )
13.0 Hz), 5.18-5.15 (d, J ) 7.0 Hz) 3.88 (s), 3.86 (s), 3.85 (s), 3.76
(s), 3.66 (s). 13C (major peaks only): 149.0, 148.4, 147.7, 147.3, 147.1,
146.5, 129.3, 129.0, 120.7, 117.5, 111.5, 110.9, 108.3, 105.4, 104.8,
60.5, 56.4, 55.9, 55.8, 55.7. IR: 1655, 1534, 1453, 1259. MS: 194
(M+, 100%), 179, 151. HRMS: Calcd for C11H14O3: 194.0943 (M+).
Found: 194.0942 (M+). EA (Calcd): C: 67.46 (68.02); H: 7.20 (7.27).
Vinyl Ether 18. tert-Butyllithium (1.7 M in pentane, 34 mL, 58
mmol) was added to a cold (-78 °C) suspension of Ph3P+CH2OMe
Cl- (19.8 g, 57.8 mmol) in THF (80 mL), and the mixture was warmed
to 0 °C. After 2 h, p-anisaldehyde (4.7 mL, 38.5 mmol) in THF (16
mL) was added and the mixture handled as described above for 17 to
furnish 4.5 g (71%) of 18, a colorless oil, which solidifies upon storage
at -20 °C. 1H (major peaks only, cis:trans ∼19:81): 7.55-7.51 (d, J
) 8.8 Hz), 7.18-7.14 (m), 6.97-6.92 (d, J ) 13.0 Hz), 6.87-6.81
(m) 6.78 (s), 6.07-6.05 (d, J ) 7.0 Hz), 5.82-5.77 (d, J ) 13.0 Hz),
5.20-5.17 (d, J ) 7.0 Hz) 3.80 (s), 3.79 (s), 3.77 (s), 3.75 (s), 3.67
(s). 13C: 157.8, 147.5, 146.3, 129.3, 128.3, 126.1, 114.0, 113.5, 105.2,
104.5, 60.2, 56.4, 55.2. IR: 1642, 1511, 1463, 1246. MS: 164 (M+,
100%), 149, 121. HRMS: Calcd for C10H12O2: 194.0837 (M+).
Found: 164.0837 (M+). EA (Calcd): C: 73.53 (73.15); H: 7.29 (7.37).
Enone Ketal 25. Phosphonate 24 (12.9 g, 54.0 mmol) in 1,2-
dimethoxyethane (DME, 15 mL) was cannulated into a cold (0 °C)
solution of tBuOK (6.4 g, 54.0 mmol) in DME (47 mL) in a flame-
dried flask, and the mixture was stirred at 0 °C for 30 min. The ice-
bath was removed, and veratraldehyde (6.9 g, 41.5 mmol) in DME
(15 mL) was added. The mixture was refluxed until the reaction was
completed (TLC), and then it was cooled, poured into saturated aqueous
NaHCO3, and extracted with ether. The extracts were sequentially
washed with saturated aqueous NaHCO3, H2O, and saturated aqueous
Cycloadduct 26. A solution of 20 (4.8 g, 20.5 mmol), 17 (12.0 g,
61.8 mmol), and Yb(fod)3 (2.2 g, 2.0 mmol) in 1,2-dichloroethane
(DCE, 45 mL) was gently refluxed until the reaction completed
(TLC, 36 h). The cooled mixture was poured into saturated aqueous
NaHCO3 and extracted with CHCl3. The extracts were sequentially
washed with saturated aqueous NaHCO3, H2O and saturated aqueous
NaCl, dried (Na2SO4), and concentrated. Chromatography (50%
EtOAc/hexanes) provided 26 (7.479 g, 91%), yellow foam. 1H: 6.80-
6.15 (m), 5.28-5.11 (m), 3.96-3.45 (m), 2.91-2.84 (t, J ) 8.7 Hz),
2.42-2.39 (m). 13C: 194.0, 148.7, 148.5 (2 signals), 147.8, 133.5,
131.8, 120.0 (2 signals), 113.2, 111.5, 111.2, 111.0, 110.8, 103.7, 56.6,
55.7, 55.3, 47.4, 25.7. IR: 1696, 1643, 1521, 1267. MS: 428 (M+),
194 (100%). HRMS: Calcd for C24H28O7: 428.1835 (M+). Found:
428.1833 (M+).
Cycloadduct 27. A solution of 20 (2.2 g, 9.4 mmol), 18 (4.7 g,
28.5 mmol), and Yb(fod)3 (1.0 g, 0.9 mmol) in DCE (20 mL) was
gently refluxed until the reaction completed (TLC, 7 h). Workup as
described above for 26 and chromatography with 40% EtOAc/hexanes
1
provided 27 (3.7 g, 99%), yellow foam, mixture of diastereomers. H
(major peaks only): 7.13-6.06 (m), 5.21-5.06 (m), 4.12-3.20 (m),
2.90-2.83 (t, J ) 8.6 Hz), 2.40-2.35 (m). 13C (major peaks only):
193.7, 158.3, 147.7, 147.6, 132.2, 131.0, 130.1, 128.8, 128.7, 122.3,
120.9, 119.8, 114.1, 113.8, 113.3, 113.2, 113.0, 112.3, 111.1, 109.9,
101.2, 55.8, 55.6, 55.5, 55.0, 45.9, 39.0, 25.8, 25.7, 25.6. IR: 1687,-
1632, 1514, 1246. MS: 398 (M+), 310, 164 (100%). HRMS: Calcd
for C23H26O6: 398.1729 (M+). Found: 398.1728 (M+).
Alcohol 28a. Diisobutylaluminum hydride (DIBAL, 1.5 M in
toluene, 13.1 mL, 19.6 mmol) was added at a slow dropwise rate into

Synthesis of Phenanthroizidine Alkaloids
J. Am. Chem. Soc., Vol. 118, No. 48, 1996 12087
a cold (-78 °C) solution of 26 (6.5 g, 15.1 mmol) in CH2Cl2 (50 mL).
The mixture was stirred at -78 °C for 10 min (TLC: reaction complete)
and then treated with cold MeOH (-78 °C, 3.5 mL) and poured into
cold (0 °C) saturated aqueous NaHCO3 with vigorous swirling over
15 min. The resulting slurry was extracted with CHCl3. The extracts
were washed with saturated aqueous NaCl, dried (Na2SO4), and
concentrated to give 6.4 g (99%) of 28a, pale yellow foam, mixture of
diastereomers, which was used in crude form. 1H (major peaks only):
6.92-6.23 (m), 5.14-5.02 (m), 4.41-4.32 (br q, J ) 5.9 Hz), 3.96-
3.38 (m), 2.85-2.76 (m), 2.5-2.1 (br s, chemical shift depending on
concentration), 1.51-1.44 (b). 13C (major peaks only): 154.1, 154.0,
148.5, 148.3, 147.5, 147.4, 135.3, 132.4, 119.8, 119.7, 111.5, 111.0,
110.6, 103.6, 100.3, 67.4, 56.3, 55.6, 55.5, 52.3, 52.2, 45.8, 45.7, 21.1,
20.7. IR: 1687,1632, 1514, 1246. MS: 430 (M+), 194 (100%).
HRMS Calcd for C24H30O7: 430.1992. Found: 430.1992 (M+).
Alcohol 28b. DIBAL (1.5 M solution in toluene, 7.1 mL, 10.6
mmol) was added at a slow dropwise rate into a cold (-78 °C) solution
of 27 (3.8 g, 9.6 mmol) in CH2Cl2 (32 mL). The same procedure
detailed above for 28a furnished 3.8 g (99%) of 29b, pale yellow foam,
mixture of diastereomers, which was used in crude form. 1H (major
peaks only): 7.00-6.11 (m), 5.15-4.99 (m), 4.37-4.30 (br q, J )
5.8 Hz, 3.86-3.19 (m), 2.85-2.76 (m), 2.5-2.1 (br s, chemical shift
depends on concentration), 1.51-1.15 (m). 13C (major peaks only):
158.1, 154.1, 154.0, 148.2, 147.3, 135.2, 131.8, 131.1, 128.9, 119.5,
113.6, 111.0, 110.6, 103.8, 100.4, 67.7, 67.5, 56.3, 55.6, 55.5, 55.0,
52.0, 51.9, 45.9, 45.7, 21.1, 20.7. IR: 3480 (br), 1680, 1611, 1514,
1464, 1248. MS: 400 (M+), 386, 270, 164 (100%). HRMS: Calcd
for C23H28O6: 400.1886 (M+). Found: 400.1886 (M+).
Benzoate 29a. Benzoyl chloride (BzCl, 3.5 mL, 30.2 mmol) was
added at room temp to a solution of 28a (6.5 g, 15.1 mmol) and pyridine
(4.9 mL, 60.4 mmol) in CH2Cl2 (36 mL). The mixture was stirred at
room temp for 30 min and then added to saturated aqueous NaHCO3
and extracted with CHCl3. The extracts were washed with H2O and
with saturated aqueous NaCl, dried (Na2SO4), and concentrated.
Chromatography (50% EtOAc/hexanes) gave 6.3 g (78%) of 29a,
yellow foam, mixture of diastereomers. 1H (major peaks only): 8.16-
8.01 (m), 7.60-7.15 (m), 6.79-6.40 (m), 5.61-5.53 (q, J ) 6.6 Hz),
4.85-4.81 (d, J ) 7.7 Hz), 3.91-3.32 (m), 2.85-2.71 (m), 2.23-
2.02 (m), 1.69-1.27 (m). 13C (major peaks only): 165.3, 148.4, 148.2,
147.1, 135.0, 132.8, 131.7, 130.0, 129.4, 128.1, 119.9, 119.4, 112.0,
110.8, 101.7, 101.4, 100.4, 100.2, 70.7, 69.9, 56.3, 56.0, 55.5, 52.7,
52.3, 41.5, 36.3, 15.4, 15.3, 14.3. IR: 1723, 1522, 1270. MS: 534
(M+), 300 (100%), 194. HRMS: Calcd for C31H34O8: 534.2253 (M+).
Found: 534.2252 (M+).
Pyridine 30b. A mixture of 29b (3.7 g, 7.4 mmol) and hydroxy-
lamine hydrochloride (3.1 g, 44.7 mmol) in MeCN (75 mL) was treated
as detailed above for 30a. Chromatography (30% EtOAc/hexanes) gave
2.0 g (59%) of 30b, pale yellow foam. 1H: 8.58 (s, 1H), 8.15-8.11
(d, 2H, J ) 7.0 Hz), 7.55-7.36 (m, 5H), 7.09-7.06 (d, 2H, J ) 8.7
Hz), 6.85-6.77 (m, 5H), 6.55-6.54 (d, J ) 1.5 Hz) 6.29-6.21 (q,
1H, J ) 6.6 Hz), 3.85 (s, 3H), 3.82 (s, 3H), 3.53 (s, 3H), 1.82-1.80
(d, 3H, J ) 6.6 Hz). 13C: 165.8, 159.0, 158.8, 150.4, 148.7, 148.2,
147.9, 134.3, 132.9, 131.1, 130.7, 130.1, 129.9, 129.6, 128.3, 128.1,
121.6, 120.8, 114.4, 113.8, 112.7, 110.8, 73.5, 55.7, 55.4, 55.1, 20.7.
IR: 1717, 1606, 1514, 1253. MS: 469 (M+), 364 (100%). HRMS:
Calcd for C29H27NO5: 469.1889 (M+). Found: 469.1890 (M+).
Alcohol 31a. A mixture of 30a (2.0 g, 4.1 mmol), MeOH (41 mL),
and 1% aqueous NaOH (14 mL) was stirred at room temperature for
1.5 h and then poured into saturated aqueous NaHCO3 and extracted
with CHCl3. Chromatography (0.5% MeOH/CHCl3) gave 1.6 g (99%)
of 32a, pale yellow foam. 1H: 8.52 (s, 1H), 7.32 (s, 1H), 6.83-6.73
(m, 4H), 6.60-6.58 (m, 2H), 5.00-4.92 (q, 1H, J ) 6.5 Hz), 3.85 (s,
6H), 3.61 (s, 3H), 3.59 (s, 3H), 1.58-1.55 (d, 3H, J ) 6.6 Hz). 13C:
161.9, 149.1, 148.8, 148.5, 148.4, 148.3, 148.2, 134.2, 131.2, 130.2,
121.9, 121.6, 120.2, 113.1, 112.6, 111.0, 110.9, 68.9, 55.8, 55.6, 24.2.
IR: 3450 (br), 1600, 1524, 1252. MS: 395 (M+, 100%), 380. HRMS
Calcd for C23H25NO5: 395.1733 (M+). Found: 395.1733 (M+).
Alcohol 31b. A mixture of 30b (2.0 g, 4.30 mmol), MeOH (43
mL), and 1% aqueous NaOH (14 mL) was processed as described above
for 31a to give 1.5 g (94%) of 31b, pale yellow foam. 1H: 8.49 (s,
1H), 7.32 (s, 1H), 7.08-7.05 (d, 2H, J ) 8.7 Hz), 6.86-6.78 (m, 4H),
6.56-6.55 (d, 1H, J ) 1.4 Hz), 5.01-4.93 (q, 1H, J ) 6.5 Hz), 3.87
(s, 3H), 3.78 (s, 3H), 3.57 (s, 3H), 1.59-1.56 (d, 3H, J ) 6.5 Hz). 13C:
161.8, 158.9, 149.3, 148.8, 148.3, 148.1, 134.1, 131.2, 130.8, 130.0,
121.6, 120.2, 113.8, 112.8, 110.9, 68.9, 55.7, 55.6, 55.2, 24.2. IR:
3377 (br), 1607, 1515, 1253. MS: 365 (M+), 350 (100%). HRMS
Calcd for C22H23NO4: 365.1627 (M+). Found: 365.1626 (M+).
Chloride 32a. A mixture of SOCl2 (256 µL, 3.5 mmol), 31a (697
mg, 1.8 mmol), pyridine (284 µL, 3.5 mmol), and benzene (18 mL)
was refluxed for 20 min and then poured into saturated aqueous
NaHCO3 and extracted with CHCl3. The extracts were washed with
H2O and with saturated aqueous NaCl, dried (Na2SO4), and concen-
trated. Chromatography (50% EtOAc/hexanes) gave 670 mg (92%)
of 33a, yellow foam. 1H: 8.59 (s, 1H), 7.51 (s, 1H), 6.86-6.77 (m,
4H), 6.62-6.61 (m, 2H), 5.27-5.19 (q, 1H, J ) 6.8 Hz), 3.88 (s, 6H),
3.64 (s, 3H), 3.62 (s, 3H), 1.98-1.95 (d, 3H, J ) 6.9 Hz). 13C: 159.4,
150.2, 148.8, 148.6, 148.5, 148.4, 148.3, 134.8, 131.0, 130.0, 121.9,
121.7, 121.6, 113.0, 112.6, 111.0, 110.9, 58.7, 55.8, 55.6, 24.8. IR:
1522, 1251. MS: 413 (M+, 100%), 378. HRMS Calcd for C23H24-
ClNO4 413.1394 (M+). Found: 413.1393 (M+).
Benzoate 29b. BzCl (2.2 mL, 19.1 mmol) was added at room temp
to a solution of 29b (3.8 g, 9.5 mmol) and pyridine (3.1 mL, 38.2
mmol) in CH2Cl2 (32 mL), and the mixture was handled as described
above for 29a to give 3.7 g (85%) of 29b, yellow foam, mixture of
diastereomers. 1H (major peaks only): 8.12-8.01 (m), 7.56-7.41 (m),
6.99-6.43 (m), 5.75-5.60 (m), 5.19-5.06 (m), 3.80-3.31 (m), 2.93-
2.80 (m), 1.67-1.29 (m). 13C (major peaks only): 165.6, 165.5, 158.2,
151.0, 150.7, 148.3, 147.3, 135.3, 135.2, 133.4, 132.9, 132.0, 131.9,
130.0, 129.5, 128.9, 128.3, 119.7, 113.7, 113.4, 111.0, 110.8, 110.5,
103.7, 103.5, 102.8, 102.1, 70.1, 70.0, 56.0, 55.8, 55.6, 55.4, 55.0, 52.0,
45.8, 45.6, 18.5, 18.0. IR: 1719, 1514, 1269. MS: 504 (M+), 270,
164 (100%). HRMS: Calcd for C30H32O7: 504.2148 (M+). Found:
504.2149 (M+).
Chloride 32b. A mixture of SOCl2 (572 µL, 7.9 mmol), 31b (1.4
g, 3.9 mmol), pyridine (637 µL, 7.9 mmol) and benzene (40 mL) was
processed as discussed above for 32a. Chromatography (30% EtOAc/
hexanes) gave 1.4 g (91%) of 32b, yellow foam. 1H: 8.53 (s, 1H),
7.48 (s, 1H), 7.08-7.05 (d, 2H, J ) 8.7 Hz), 6.87-6.78 (m, 4H), 6.55-
6.54 (d, 1H, J ) 1.7 Hz), 5.24-5.16 (q, 1H, J ) 6.8 Hz), 3.85 (s, 3H),
3.76 (s, 3H), 3.55 (s, 3H), 1.95-1.92 (d, 3H, J ) 6.8 Hz). 13C: 159.2,
158.9, 150.2, 148.7, 148.3, 148.2, 134.7, 130.9, 130.7, 129.7, 121.7,
121.6, 113.8, 112.6, 110.9, 58.7, 55.7, 55.5, 55.1, 24.8. IR: 1608,
1514, 1249. MS: 383 (M+, 100%), 348. HRMS Calcd for C22H22-
ClNO3 383.1288 (M+). Found: 383.1288 (M+).
Vinylpyridine 33a. A solution of 32a (670 mg, 1.6 mmol) and
tBuOK (1.9 g, 16.2 mmol) in THF (16 mL) was refluxed for 10 min,
and then it was cooled and poured into saturated aqueous NaHCO3
and extracted with CHCl3. The extracts were washed with H2O and
with saturated aqueous NaCl, dried (Na2SO4), and concentrated.
Chromatography (50% EtOAc/hexanes) gave 431 mg (71%) of 33a,
yellow foam. 1H: 8.60 (s, 1H), 7.38 (s, 1H), 6.95-6.86 (dd, 1H, J1 )
17.5 Hz, J2 ) 10.3 Hz), 6.91-6.77 (m, 4H), 6.63-6.61 (d, 2H, J )
6.8 Hz), 6.30-6.23 (d, 1H, J ) 17.5 Hz), 5.54-5.50 (d, 1H, J ) 10.7
Hz), 3.88 (s, 6H), 3.62 (s, 6H). 13C: 154.4, 150.5, 148.7, 148.5, 148.3,
147.6, 136.5, 134.2, 131.3, 130.3, 121.9, 121.6, 118.1, 113.1, 112.6,
111.0, 110.9, 55.8, 55.6. IR: 1517, 1252. MS: 377 (M+, 100%).
HRMS Calcd for C23H23NO4: 377.1627 (M+). Found: 377.1629 (M+).
Vinylpyridine 33b. A solution of 32b (1.343 g, 3.50 mmol) and
tBuOK (2.1 g, 17.5 mmol) in THF (35 mL) was treated as discussed
Pyridine 30a. A mixture of 29a (6.2 g, 11.7 mmol) and hydroxy-
lamine hydrochloride (4.9 g, 70.3 mmol) in acetonitrile (MeCN, 117
mL) was refluxed for 1.5 h and then cooled to room temp, quenched
(saturated aqueous NaHCO3), and extracted with CHCl3. The combined
extracts were washed with H2O and with saturated aqueous NaCl, dried
(Na2SO4), and concentrated. Chromatography (50% EtOAc/hexanes)
gave 4.7 g (80%) of 30a, pale yellow foam. 1H: 8.58 (s, 1H), 8.11-
8.08 (d, 2H, J ) 7.5 Hz), 7.54-7.36 (m, 5H), 6.81-6.73 (m, 5H, J1 )
8.4 Hz, J2 ) 1.5 Hz), 6.25-6.17 (q, 1H, J ) 6.6 Hz), 3.82 (s, 3H),
3.81 (s, 3H), 3.58 (s, 3H), 3.53 (s, 3H), 1.79-1.77 (d, 3H, J ) 6.6
Hz). 13C: 165.6, 159.0, 150.2, 148.6, 148.4, 148.3, 148.2, 147.9, 134.3,
132.8, 131.1, 130.1, 130.0, 129.5, 128.2, 121.7, 121.5, 120.7, 113.0,
112.6, 110.9, 110.8, 73.3, 55.6, 55.5, 55.4, 20.6. IR: 1717, 1601, 1515,
1252. MS: 499 (M+), 394, 105 (100%). HRMS: Calcd for C30H29-
NO6: 499.1995 (M+). Found: 499.1995 (M+).

12088 J. Am. Chem. Soc., Vol. 118, No. 48, 1996
Ciufolini and Roschangar
above for 33a. Chromatography (30% EtOAc/hexanes) gave 933 mg
(77%) of 33b, yellow foam. 1H: 8.54 (s, 1H), 7.34 (s, 1H), 7.26-
7.08 (d, 2H, J ) 8.7 Hz), 6.92-6.83 (dd, 1H, J1 ) 17.4 Hz, J2 ) 10.8
Hz), 6.88-6.77 (m, 4H), 6.56-6.56 (d, 1H, J ) 1.4 Hz), 6.28-6.21
(dd, J1 ) 17.5 Hz, J2 ) 1.4 Hz), 5.51-5.46 (dd, J1 ) 10.8 Hz, J2 )
1.2 Hz), 3.85 (s, 3H), 3.76 (s, 3H), 3.57 (s, 3H). 13C: 158.8, 154.3,
150.6, 148.6, 148.3, 147.5, 136.5, 134.0, 131.2, 130.7, 130.0, 121.8,
121.5, 117.9, 113.7, 112.7, 110.8, 55.7, 55.5, 55.1. IR: 1513, 1253.
MS: 347 (M+, 100%). HRMS Calcd for C22H21NO3: 347.1521 (M+).
Found: 347.1521 (M+).
Nitrile 42. AcOH (352 µL, 6.7 mmol) was added to a solution of
33a (580 mg, 1.5 mmol) in commercial LiCN in DMF (Aldrich, 0.5
M, 31 mL, 15.5 mmol CAUTION: highly toxic). The mixture was
heated at 140 °C for 18 h and then cooled, poured into saturated aqueous
NaHCO3, and extracted with CHCl3. The extracts were washed with
H2O and with saturated aqueous NaCl, dried (Na2SO4), and concen-
trated. Chromatography (50% EtOAc/hexanes) gave 620 mg (99%)
of 42, yellow foam. 1H: 8.48 (s, 1H), 7.19 (s, 1H), 6.77-6.52 (m,
4H), 6.54-6.52 (d, 2H, J ) 5.7 Hz), 3.79 (s, 6H), 3.55 (s, 3H), 3.53
(s, 3H), 3.13-3.08 (t, 2H, J ) 7.2 Hz), 2.86-2.81 (t, 2H, J ) 7.1 Hz).
13C: 155.7, 150.5, 148.7, 148.5, 148.4, 148.2, 147.9, 133.9, 130.8,
130.0, 123.5, 121.8, 121.5, 119.4, 113.0, 112.5, 111.0, 110.8, 55.7,
55.6, 32.9, 16.6. IR: 1597, 1515, 1251, 1141, 1025. MS: 404 (M+,
100%). HRMS Calcd for C24H24N2O4: 404.1736 (M+). Found:
404.1737 (M+).
Nitrile 46. AcOH (201 µL, 3.5 mmol) was added to a solution of
33b (307 mg, 0.9 mmol) in commercial LiCN in DMF (Aldrich, 0.5
M, 18 mL, 8.8 mmol CAUTION: highly toxic). The mixture was
heated at 140 °C for 12 h. Workup and chromatography as detailed
above for 42 gave 244 mg (74%) of 46, yellow foam. 1H: 8.50 (s,
1H), 7.25 (s, 1H), 7.07-7.03 (d, 2H, J ) 8.7 Hz), 6.84-6.79 (m, 4H),
6.55-6.54 (d, 1H, J ) 1.1 Hz), 3.84 (s, 3H), 3.76 (s, 3H), 3.55 (s,
3H), 3.19-3.13 (t, 2H, J ) 7.2 Hz), 2.92-2.86 (t, 2H, J ) 7.2 Hz).
13C: 158.8, 155.6, 150.5, 148.7, 148.2, 147.9, 133.8, 130.7, 129.8,
123.5, 121.6, 119.4, 113.7, 112.6, 110.8, 55.6, 55.4, 55.1, 32.9, 16.6.
IR: 1609, 1515, 1251, 1026. MS: 374 (M+, 100%). HRMS Calcd
for C23H22N2O3: 374.1630 (M+). Found: 374.1631 (M+).
Nitrile 51. A solution of 33b (634 mg, 1.83 mmol) and MeCN
(191 µL, 3.66 mmol) in DMSO (18 mL) was stirred at room temp for
5 min after addition of tBuOK (647 mg, 5.49 mmol), and then it was
poured into saturated aqueous NaHCO3 and extracted with CHCl3. The
extracts were washed with H2O and with saturated aqueous NaCl, dried
(Na2SO4), and concentrated. Chromatography (40% EtOAc/hexanes)
gave 623 mg (88%) of 51, orange foam. 1H: 8.50 (s, 1H), 7.22 (s,
1H), 7.10-7.08 (d, 2H, J ) 8.8 Hz), 6.85-6.79 (m, 4H), 6.56 (s, 1H),
3.90 (s, 3H), 3.79 (s, 3H), 3.58 (s, 3H), 3.03-2.97 (t, 2H, J ) 7.3 Hz),
2.50-2.44 (t, 2H, J ) 7.1 Hz), 2.26-2.14 (quintet, 2H, J ) 7.2 Hz).
13C: 158.8, 158.1, 150.6, 148.8, 148.4, 147.8, 133.4, 131.0, 130.8,
130.0, 123.6, 121.6, 119.5, 113.8, 112.7, 110.9, 55.8, 55.6, 55.2, 36.0,
25.0, 16.6. IR: 2318 (weak), 1608, 1515, 1254, 1026. MS: 388 (M+),
335 (100%). HRMS Calcd for C24H24N2O3: 388.1787 (M+). Found:
388.1789 (M+).
6.4 Hz), 6.87-6.77 (m, 4H), 6.54 (s, 1H), 3.82 (s, 3H), 3.74 (s, 3H),
3.51 (s, 3H), 3.34 (br s, 2H), 2.95 (br s, 2H). 13C (CDCl3): 174.5,
159.5, 155.0, 153.5, 149.9, 148.5, 143.7, 135.4, 130.5, 128.7, 127.3,
126.0, 122.2, 114.1, 112.4, 111.0, 55.7, 55.5, 55.2, 33.5, 29.3. IR:
3511 (br), 1724, 1609, 1516, 1257. MS: 393 (M+), 348 (100%).
HRMS Calcd for C23H23NO5: 393.1576 (M+). Found: 393.1576 (M+).
Carboxylic Acid 52. A solution of 51 (235 mg, 0.6 mmol) in concd
HCl (2.4 mL)/dioxane (9.6 mL) was heated at 40 °C for 2 days.
Workup as described above for 43 afforded acid 52 (240 mg, 99%),
orange foam, which was advanced to the next step without purification.
1H: 9.05 (br s, 1H), 8.53 (br s, 1H), 7.43 (s, 1H), 7.06-7.04 (d, 2H,
J ) 6.4 Hz), 6.87-6.77 (m, 4H), 6.55 (s, 1H), 3.86 (s, 3H), 3.77 (s,
3H), 3.55 (s, 3H), 3.10 (br s, 2H), 2.48 (br s, 2H), 2.18 (br s, 2H). 13C:
175.9, 159.8, 155.1, 154.7, 150.2, 148.6, 142.8, 135.8, 130.6, 128.4,
127.0, 126.1, 122.3, 114.3, 112.4, 111.2, 55.8, 55.6, 55.2, 33.2, 32.8,
24.6. IR: 3504 (br), 1723 (medium intensity), 1609, 1517, 1258. MS:
407 (M+), 348, 335 (100%). HRMS Calcd for C24H25NO5: 407.1733
(M+). Found: 407.1731 (M+).
Alcohol 44. A solution of 43 (332 mg, 0.8 mmol) in THF (16 mL)
was treated with lithium aluminum hydride (LAH, 187 mg, 4.7 mmol).
After gas evolution ceased, the mixture was refluxed for 15 min and
then cooled, poured into saturated aqueous NaHCO3 (CAUTION:
vigorous H2 evolution), and extracted with CHCl3. The extracts were
washed with saturated aqueous NaCl, dried (Na2SO4), and concentrated
to afford highly polar 44 (yellow foam; 280 mg, 87%), which was not
further purified. 1H: 8.47 (s, 1H), 7.21 (s, 1H), 6.80-6.71 (m, 4H),
6.57-6.56 (d, 2H, J ) 5.9 Hz), 3.84 (s, 6H), 3.77-3.73 (t, 2H, J )
6.9 Hz), 3.60 (s, 3H), 3.58 (s, 3H), 3.03-2.98 (t, 2H, J ) 6.9 Hz),
2.08-1.98 (t, 2H, J ) 6.5 Hz). 13C: 160.1, 149.6, 148.7, 148.5, 148.4,
148.2, 148.0, 133.0, 131.7, 130.2, 123.5, 121.8, 121.6, 113.1, 112.6,
111.0, 110.8, 62.2, 55.7, 55.6, 34.9, 31.7. IR 3359 (br), 1601, 1514,
1251, 1026. MS: 409 (M+), 365 (100%). HRMS Calcd for C24H27-
NO5: 409.1889 (M+). Found: 409.1889 (M+).
Alcohol 48. A solution of 47 (207 mg, 0.5 mmol) in THF (11 mL)
was treated with LAH (127 mg, 3.2 mmol). The same procedure
described above for 44 delivered highly polar 48 (orange foam; 177
mg, 89%), which was not further purified. 1H: 8.47 (s, 1H), 7.22 (s,
1H), 7.07-7.03 (d, 2H, J ) 8.7 Hz), 6.82-6.78 (m, 4H), 6.54 (s, 1H),
3.86 (s, 3H), 3.79-3.74 (t, 2H, J ) 6.7 Hz), 3.77 (s, 3H), 3.56 (s, 3H),
3.05-2.99 (t, 2H, J ) 6.7 Hz), 2.09-1.99 (quintet, 2H, J ) 6.4 Hz).
13C: 160.0, 158.8, 149.7, 148.7, 148.3, 148.0, 133.0, 131.2, 130.7,
130.0, 123.6, 121.6, 113.8, 112.7, 110.9, 62.3, 55.7, 55.5, 55.2, 34.9,
31.7. IR: 3387 (br), 1608, 1515, 1252. MS: 379 (M+), 335 (100%).
HRMS Calcd for C23H25NO4: 379.1784 (M+). Found: 379.1780 (M+).
Alcohol 53. A solution of 52 (132 mg, 0.4 mmol) in THF (8 mL)
was treated with LAH (94 mg, 2.3 mmol). The same procedure
described above for 44 delivered highly polar 53 (thick yellow oil;
113 mg, 89%), which was not further purified. 1H: 8.46 (s, 1H), 7.53
(br s, 1H), 7.19 (s, 1H), 7.07-7.04 (d, 2H, J ) 7.7 Hz), 6.84-6.74
(m, 4H), 6.54 (s, 1H), 3.86 (s, 3H), 3.77 (s, 3H), 3.73-3.68 (t, 2H, J
) 6.3 Hz), 3.55 (s, 3H), 2.92-2.86 (t, 2H, J ) 6.8 Hz), 1.95-1.83
(quintet, 2H, J ) 6.6 Hz), 1.74-1.64 (quintet, 2H, J ) 6.5 Hz). 13C:
160.6, 158.7, 150.0, 148.6, 148.3, 147.7, 132.8, 131.3, 131.1, 130.8,
130.2, 123.4, 121.6, 113.8, 112.8, 112.6, 110.8, 62.2, 56.1, 55.7, 55.5,
55.2, 37.1, 32.2, 25.9. IR: 3401 (br), 1607, 1515, 1251. MS: 393
(M+), 335 (100%). HRMS Calcd for C24H27NO4: 393.1940 (M+).
Found: 393.1937 (M+).
Carboxylic Acid 43. A solution of 42 (365 mg, 0.90 mmol) in
concd HCl (5 mL) and dioxane (20 mL) was heated at 40 °C for
18 h. The cooled mixture was poured into saturated aqueous NaHCO3
(CAUTION: vigorous foaming) and extracted with ether. The extracts
were discarded and the aqueous mixture was acidified to pH 2, saturated
with solid NaCl, and repeatedly extracted with EtOAc. The extracts
were dried (Na2SO4) and concentrated to afford 43 (332 mg, 87%) as
a greenish foam that was advanced to the next step without purification.
1H: 12.39 (br s, 1H), 8.51 (br s, 1H), 7.48 (s, 1H), 6.89-6.72 (m,
4H), 6.60-6.58 (d, 2H, J ) 6.3 Hz), 3.85 (s, 6H), 3.62 (s, 3H), 3.58
(s, 3H), 3.31 (br s, 2H), 2.96 (br s, 2H). 13C: 174.7, 156.1, 152.0,
149.5, 148.8, 148.7, 148.5, 145.2, 134.9, 129.3, 128.2, 125.3, 121.9,
121.8, 112.6, 112.3, 111.1, 110.9, 55.7, 33.7, 30.0. IR: 3590 (br),
1723, 1601, 1516, 1465, 1254, 1024. MS: 423 (M+), 378 (100%).
HRMS Calcd for C24H25NO6: 423.1682 (M+). Found: 423.1680 (M+).
Carboxylic Acid 47. A solution of 46 (205 mg, 0.55 mmol) in
concd HCl (2.2 mL)/dioxane (8.8 mL) was heated at 40 °C for 20 h.
Workup as described above for 43 afforded 47 (207 mg, 96%), yellow
foam, which was advanced to the next step without purification. 1H:
11.08 (br s, 1H), 8.47 (br s, 1H), 7.57 (s, 1H), 7.04-7.01 (d, 2H, J )
Pyridinium Salt 45. Methanesulfonyl chloride (MsCl, 102 µL, 1.3
mmol) was added to a cold (0 °C) mixture of crude 45 (270 mg, 0.7
mmol) and Et
3N (275 µL, 2.0 mmol) in CH2Cl2 (13 mL) with good
stirring. After 30 min (TLC: reaction complete), the mixture was
poured into saturated aqueous NaCl and extracted with CHCl
3. The
extracts were dried (Na2SO4) and concentrated to afford 45 (245 mg),
orange-brown foam, which was advanced to the next step without
purification or characterization beyond a 1
H NMR spectrum. The nature
of the counterion was not determined. 1H: 9.21 (s, 1H), 7.75 (s, 1H),
6.96-6.55 (m, 6H), 5.16-5.11 (t, 2H, J ) 7.3 Hz), 3.89-3.49 (m,
14H), 2.56-2.45 (quintet, 2H, J ) 7.3 Hz).
Pyridinium Salt 49. MsCl (70 µL, 0.9 mmol) was added to a cold
(0 °C) mixture of crude 48 (172 mg, 0.4 mmol) and Et3N (187 µL, 1.3
mmol) in CH2Cl2 (9 mL). The same protocol detailed above for 45
furnished 49 (191 mg), orange-brown foam, which was advanced to

Synthesis of Phenanthroizidine Alkaloids
J. Am. Chem. Soc., Vol. 118, No. 48, 1996 12089
the next step without purification or characterization beyond a 1H NMR
spectrum. The nature of the counterion was not determined 1H: 9.26
(s, 1H), 7.78 (s, 1H), 7.26-6.81 (m, 6H), 6.58 (d, 1H, J ) 1.3 Hz),
5.29-5.24 (t, 2H, J ) 7.1 Hz), 3.87-3.53 (m, 11H), 2.66-2.51 (quintet,
2H, J ) 7.2 Hz).
Pyridinium Salt 54. MsCl (46 µL, 0.6 mmol) was added to a cold
(0 °C) mixture of crude 54 (113 mg, 0.3 mmol) and Et3N (121 µL, 0.9
mmol) in CH2Cl2 (6 mL). The same protocol detailed above for 45
furnished 54 (127 mg), orange-green foam, which was advanced to
the next step without purification or characterization beyond a 1H NMR
spectrum. The nature of the counterion was not determined. 1H: 9.05
(s, 1H), 7.67 (br s, 1H), 7.31-7.26 (m, 2H), 6.92-6.82 (m, 4H), 6.61
(s, 1H), 4.98 (br s, 2H), 3.87-3.54 (m, 11H), 3.31 (br s, 2H), 2.22 (br
s, 2H), 2.14-2.02 (br m, 2H).
(()-Septicine (1). NaBH4 (182 mg, 4.6 mmol) was added at room
temp to a suspension of crude 45 (164 mg) in EtOH (9 mL). Strong
gas evolution ensued. When gas evolution subsided, the mixture was
refluxed for 30 min to complete reduction (TLC), and then it was
cooled, poured into 10% aqueous NaOH, and extracted with CH2Cl2.
The extracts were dried (Na2SO4) and concentrated. Chromatography
(2% MeOH/CHCl3) gave 106 mg (62% from 44) of (()-septicine, 1,
mp 135-137 °C. 1H: 6.71-6.65 (m, 4H), 6.57-6.53 (d, 2H, J ) 5.0
Hz), 3.93-3.87 (d, 1H, J ) 15.9 Hz), 3.80 (s, 6H), 3.59 (s, 3H), 3.57
(s, 3H), 3.36-3.29 (t, 1H, J ) 6.8 Hz), 3.14-3.08 (d, 1H, J ) 16.1
Hz), 2.78-2.71 (d, 1H, J ) 12.6 Hz) 2.46-1.84 (m, 6H), 1.65-1.52
(m, 1H). 13C: 148.2, 148.1, 147.5, 134.9, 133.5, 132.9, 132.3, 121.0,
120.7, 113.0, 112.8, 110.6 (2 signals), 60.4, 57.3, 55.7, 55.6, 54.1, 38.2,
30.7, 29.7, 21.5. IR: 1513, 1261, 1027. MS: 395 (M+), 326, 295
(100%), 264. HRMS Calcd for C24H29NO4: 395.2097 (M+). Found:
395.2097 (M+).
(()-Seco-Antofine (50). NaBH4 (192 mg, 4.8 mmol) was added at
room temp to a suspension of crude 50 (180 mg) in EtOH (10 mL).
Strong gas evolution ensued. The mixture was handled as detailed
above for 1 to give 139 mg (84% from 48) of (()-seco-antofine 50,
pale orange foam. 1H: 6.98-6.94 (d, 2H, J ) 6.9 Hz), 6.69-6.63
(m, 4H), 6.46 (s, 1H), 6.57-6.53 (d, 2H, J ) 5.0 Hz), 3.90-3.83 (d,
1H, J ) 15.9 Hz), 3.80 (s, 3H), 3.72 (s, 3H), 3.54 (s, 3H), 3.34-3.27
(dt, 1H, J1 ) 8.3 Hz, J2 ) 2.0 Hz), 3.11-3.05 (d, 1H, J ) 15.9 Hz),
2.81-2.65 (m, 1H) 2.44-1.80 (m, 6H), 1.65-1.48 (m, 1H). 13C
(CDCl3): 157.9, 147.8, 147.1, 135.0, 133.5, 132.6, 130.1, 120.6, 113.3,
113.0, 110.4, 60.4, 57.8, 55.6, 55.4, 55.1, 54.2, 38.4, 30.8, 29.7, 21.5.
IR: 1509, 1247. MS: 365 (M+), 296, 265 (100%). HRMS Calcd for
C23H27NO3: 365.1991 (M+). Found: 365.1990 (M+).
(()-Tylophorine (1). A cold (0 °C) solution of (()-septicine, 4
(54 mg, 0.14 mmol) and VOF3 (84 mg, 0.6 mmol) in CH2Cl2 (3
mL) was stirred for 15 min, and then trifluoroacetic acid (136 µL) was
slowly introduced and stirring was continued at 0 °C for an additional
15 min. The mixture was poured into 10% aqueous NaOH and
extracted with CH2Cl2. The extracts were dried (Na2SO4) and
concentrated. Recrystallization of the crude product from acetonitrile
gave 40 mg (74%) of beige (()-tylophorine, 1, mp 272-274 °C dec.
(lit. mp’s range from 263-265 °C to 292-294 °C).22 1H: 7.83 (s,
2H), 7.32 (s, 1H), 7.16 (s, 1H), 4.66-4.60 (d, 1H, J ) 14.5 Hz), 4.12
(s, 6H), 4.06 (two s, each 3H), 3.57 (s, 3H), 3.70-3.64 (d, 1H, J )
14.6 Hz), 3.51-3.45 (t, 1H, J ) 7.4 Hz), 3.42-3.34 (dd, 1H, J1 ) 5.7
Hz, J2 ) 2.4 Hz), 2.97-2.86 (t, 1H, J ) 13.0 Hz), 2.56-2.41 (m,
2H), 2.31-2.19 (m, 1H), 2.09-1.70 (m, 3H). 13C: 148.7, 148.5, 126.3,
126.0, 125.8, 124.3, 123.6, 123.4, 104.0, 103.4, 103.3, 103.1, 60.2,
56.0, 55.9, 55.2, 54.1, 33.8, 31.3, 21.6. IR: 1513, 1248. MS: 393
(M+), 324(100%). HRMS Calcd for C24H27NO4: 393.1940 (M+).
Found: 393.1939 (M+).
(()-Antofine (2). A cold (0 °C) solution of (()-seco-antofine (50)
(83 mg, 0.23 mmol) and VOF3 (143 mg, 1.1 mmol) in CH2Cl2 (5 mL)
was stirred for 15 min, and then trifluoroacetic acid (230 µL) was slowly
introduced and stirring was continued at 0 °C for an additional 15 min.
The same workup described above for 1 and recrystallization from
acetonitrile furnished 42 mg (51%) of pale yellow (()-antofine, 2, mp
205-207 °C (dec; lit. 213-215 °C, dec.).23 1H: 7.91-7.89 (m, 2H),
7.83-7.80 (d, 1H, J ) 9.1 Hz), 7.31 (s, 1H), 7.22-7.18 (dd, 1H, J1 )
9.1 Hz, J2 ) 2.5 Hz), 4.72-4.66 (d, 1H, J ) 14.9 Hz), 4.11 (s, 3H),
4.06 (s, 3H), 4.02 (s, 3H), 3.57 (s, 3H), 3.72-3.66 (d, 1H, J ) 14.9
Hz), 3.49-3.42 (dt, 1H, J1 ) 7.6 Hz, J2 ) 1.9 Hz), 3.36-3.30 (dd,
1H, J1 ) 5.7 Hz, J2 ) 2.3 Hz), 2.94-2.84 (t, 1H, J ) 12.9 Hz), 2.50-
2.39 (m, 2H), 2.26-2.18 (m, 1H), 2.08-1.71 (m, 3H). 13C: 157.4,
149.4, 148.3, 130.1, 127.1, 126.7, 125.5, 124.2, 124.1, 123.5, 114.8,
104.6, 104.0, 103.8, 60.2, 56.0, 55.9, 55.5, 55.1, 53.9, 33.7, 31.3, 29.7,
21.6. IR: 3402 (br), 1616, 1512, 1469, 1257, 1205. MS: 363 (M+),
294 (100%). HRMS Calcd for C23H25NO3:363.1834 (M+). Found:
363.1834 (M+).
Acknowledgment. We gratefully acknowledge the NIH
(CA-55268), the NSF (CHE 95-26183), and the R. A. Welch
Foundation (C-1007) for their generous support of our research
program. MAC is a Fellow of the Alfred P. Sloan Foundation,
1994-1996.
(()-Julandine (5). NaBH4 (116 mg, 2.9 mmol) was added at room
temp to a suspension of crude 54 (118 mg) in EtOH (6 mL). Strong
gas evolution ensued. The mixture was handled as detailed above for
1 to give 91 mg (86% from 53) of pale yellow (()-julandine, 5, mp
135-137 °C. 1H: 7.00-6.95 (dd, 2H, J1 ) 8.6 Hz, J2 ) 1.9 Hz),
6.71-6.63 (m, 4H), 6.46 (s, 1H), 3.81 (s, 3H), 3.73-3.54 (m, 10 H, of
which 3 s with 3H each), 3.21-3.09 (m, 2H), 2.54-1.25 (m, 10H).
13C: 158.0, 147.8, 147.2, 134.3, 133.1, 131.2, 130.1, 129.4, 129.3,
120.5, 113.4, 112.9, 110.4, 60.2, 57.9, 55.9, 55.6, 55.4, 55.1, 39.3, 33.1,
25.7, 24.2. IR: 3502 (br), 1606, 1511, 1248. MS: 379 (M+, 100%),
296, 265. HRMS Calcd for C24H29NO4: 379.2147 (M+). Found:
379.2147 (M+).
Supporting Information Available: Hardcopy 1H and 13
C
spectra of selected compounds (30 pages). See any current
masthead page for ordering and Internet access instructions.
JA962757P
(22) Discord exists in the literature regarding the melting point of (()-
1, e.g.: 292-294 °C (ref 4k); 287 °C (ref 4m); 273 °C (ref 17); 263-265
°C (ref 4f).
(23) Govindachari, T. R.; Ragade, I. S.; Viswanathan, N. J. Chem. Soc.
1962, 1356.