Chemoselectivity of the Ruthenium-Catalyzed Hydrative Diyne Cyclization: Total Synthesis of (+)-Cylindricine C, D, and E

Article abstract of DOI:10.1021/ol035752n

(Matrix presented) The chemoselectivity of the ruthenium-catalyzed hydrative diyne cylization is explored In an expeditious synthesis of the tricyclic alkaloids cylindricine C, D, and E.

Full text of DOI:10.1021/ol035752n

ORGANIC
LETTERS
2003
Vol. 5, No. 24
4599-4602
Chemoselectivity of the
Ruthenium-Catalyzed Hydrative Diyne
Cyclization: Total Synthesis of
(+)-Cylindricine C, D, and E
Barry M. Trost* and Michael T. Rudd
Department of Chemistry, Stanford UniVersity, Stanford, California 94305-5080
bmtrost@stanford.edu
Received September 11, 2003
ABSTRACT
The chemoselectivity of the ruthenium-catalyzed hydrative diyne cylization is explored in an expeditious synthesis of the tricyclic alkaloids
cylindricine C, D, and E.
As part of an ongoing study of the ability to effect a hydrative
cyclization of diynes to acylcycloalkenes,1 the question of
the chemoselectivity with unsymmetrical diynes arises (eq
1). In addressing this important issue, the impact of conjuga-
cylindricines A-J. These tricyclic alkaloids were isolated
by Blackman5 from ClaVelina cylindrical, a marine benthic
invertebrate (ascidian) found off the coast of Tasmania.
Cylindricine C, D, and E appeared to represent an attractive
context in which to explore this question.
tion, wherein one of the R groups would be vinyl, was
particularly intriguing because electronic versus steric effects
might be anticipated to operate in opposing directions.
Among a group of structurally interesting secondary
metabolite alkaloids isolated from various ascidians, includ-
ing clavepictines,2 fasicularin,3 and lepadiformine,4 stand
Crude cylindrincine extracts have shown some biological
activity in a brine shrimp bioassay,5 whereas other members
of this family have activity against a DNA-repair-deficient
yeast strain3 and inhibit growth of murine leukemia and
human solid tumor cell lines.2 The absolute configurations
of natural cylindricine C, D, and E are unknown, and the
optical rotations were not reported. Cylindricine C has been
(1) Trost, B. M.; Rudd, M. T. J. Am. Chem. Soc. 2003, 125, 11516.
(2) Raub, M. F.; Cardelina, J. H., II; Choudhary, M. I.; Ni, C.-Z.; Clardy,
J.; Alley, M. C. J. Am. Chem. Soc. 1991, 113, 3178.
(3) Patil, A. D.; Freyer, A. J.; Reichwein, R.; Carte, B.; Killmer, L. B.;
Faucette, L.; Johnson, R. K.; Faulkner, D. J. Tetrahedron Lett. 1997, 38,
363.
(5) (a) Blackman, A. J.; Li, C.; Hockless, D. C. R.; Skelton, B. H.
Tetrahedron 1993, 49, 8645. (b) Blackman, A. J.; Li, C. Aust. J. Chem.
1994, 47, 1355. (c) Blackman A. J.; Li, C. Aust. J. Chem. 1995, 48, 955.
(4) Biard, J. F.; Guyot, S.; Roussakis, C.; Verbist, J. F.; Vercauternen,
J.; Weber, J. F.; Boukef, K. Tetrahedron Lett. 1994, 34, 686.
10.1021/ol035752n CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/25/2003

made once previously6 starting from a chiral triol, and the
racemic synthesis of cylindricines D and E has also been
completed.7
may function well in the reaction on the basis of experience
gained in the methodological development of the reaction.1
The resulting methyl ketone could then be readily trans-
formed into B through an aldol condensation reaction. The
diastereoselective double conjugate addition would then give
the core of the cylindricines with the existing chiral center
controlling the selectivity of the other three.
As illustrated in Scheme 1, it was envisioned that a double
Scheme 1. Cylindricine Retrosynthesis
In the forward direction, the cyclization precursor A1 was
prepared from serine iodide 19 and enediyne 210 utilizing
the alkyl copper-zinc/haloalkyne chemistry developed by
Knochel11 (eq 2).
With compound 3 in hand, the ruthenium-catalyzed
hydrative diyne cyclization was carried out under standard
conditions. Interestingly, a single cyclization product formed
that proved to be the isomer, which would not be productive
for the cylindricine synthesis (eq 3).
conjugate addition of a pendant amine onto an appropriately
situated divinyl ketone would allow expeditious entry into
the tricyclic core of the cylindricines. Although a similar
endgame was utilized by Molander,6 we hoped to exploit
the versatility of the novel hydrative diyne cyclization
recently developed in our laboratory1 to access the function-
alized amino divinyl ketone core in a rapid fashion.
The key step in the synthesis is the ruthenium-catalyzed
hydrative cyclization of intermediate A to B. The proposed
mechanism of this reaction is shown below and involves
water addition to a ruthenacyclopentadiene or -cyclopen-
tatriene (Scheme 2). Intermediate B could be made directly
if A1 (R′′ ) heptenyl) could be chemoselectively cyclized
with water attack occurring only at the top and presumably
least hindered alkyne.8 As an alternative, A2 (R′′ ) methyl)
Although previous data1 indicated that steric effects were
the chief factor in determining chemoselectivity and a vinyl
group is generally viewed as smaller than a â-branched alkyl
chain,12 it appears other factors can influence the selectivity
as well. It is possible that disruption of conjugation developed
in the proposed ruthenacyclopentadiene or cyclopentatriene
(C, Scheme 3) going to adduct D disfavors attack at the
alkyne adjacent to the vinyl group. Thus, the addition of
water favors formation of adduct E, which is the precursor
of the observed product.
Scheme 2. Mechanistic Proposal
As indicated above, the alternative strategy was to perform
the hydrative cyclization on the methyl alkyne, followed by
an aldol condensation. This example competes the differential
steric effects versus the prospect of amide participation.
Previously, we have noted a cyclization involving a tethered
(6) Molander, G. A.; Ronn, M. J. Org. Chem. 1999, 64, 5183.
(7) Snider, B. B.; Liu, T. J. Org. Chem. 1997, 62, 5630.
(8) Previous development indicated that the attack of water occurs at
the least hindered alkyne.
(9) Compound 1 is prepared in three steps (methyl ester formation, Boc
protection, iodination) from serine in 60% overall yield (see Supporting
Information).
(10) Compound 2 is prepared from 1,7-octadiyne and the vinyl iodide
derived from 1-octyne (DIBAL-H, I2) using standard Sonagashira coupling
followed by iodination with n-BuLi and I2 in 40% overall yield (see
Supporting Information).
(11) Yeh, M. C. P.; Knochel, P. Tetrahedron Lett. 1989, 30, 4799.
(12) (a) Allinger, N. L.; Freiberg, L. A. J. Org. Chem. 1966, 31, 894.
(b) Eliel E. L.; Manoharan, M. J. Org. Chem. 1981, 46, 1959.
4600
Org. Lett., Vol. 5, No. 24, 2003

water/acetone at 60 °C to give R,â-unsaturated ketone 12 in
90% yield (Scheme 5). Only a trace amount of the undesired
Scheme 3. Rationale for Formation of Undesired Isomer
Scheme 5. Hydrative Diyne Cyclization and Completion of
Cylindricine Ca
hydroxyl nucleophile.1 The synthesis of the desired cycliza-
tion precursor was carried out in a straightforward manner
starting from readily available compounds (Scheme 4).
a (a) 5% 11, 10% water/acetone, 60 °C, 2 h; (b) (i) LDA,
heptanal, (ii) MsCl, Et3N; (c) (i) TFA, DCM, (ii) K2CO3, PhMe
reflux, (iii) TBAF.
Scheme 4. Synthesis of Cyclization Precursora
isomer was detected. Diyne 8 also participated in the
cyclization to deliver a single product of the same chemo-
selectivity as 10. However, it was desirable to reduce the
ester prior to cyclization in order to avoid potential ketone
reduction. Catalyst 11 has previously been shown to deprotect
unhindered TBS groups; however, no sign of TBDPS
cleavage was observed in the cyclization of 10. A standard
LDA-mediated aldol reaction was then carried out between
methyl ketone 12 and heptanal. The crude reaction mixture
was directly submitted to elimination (MsCl, Et3N) without
prior purification to give divinyl ketone 13 in 83% yield.
The double conjugate addition occurred readily upon cleav-
age of the Boc group with TFA/DCM and free-basing
(NaOH/DCM) the TFA-salt of the resulting primary amine.
On a small scale the double cyclization occurred directly to
give a moderate yield of the TBDPS-protected cylindricine
C along with the monocyclized adduct.
a (a) LHMDS, TMS-Cl; (b) n-BuLi, CH3I; (c) 10% AgNO3, NBS,
DMF; (d) (i) activated Zn + 1, (ii) CuCN/2 LiCl, (iii) 7; (e) LiBH4;
(f) TBDPS-Cl, imidazole, DMF.
Monosilylation of 1,7-octadiyne proceeded in moderate
yield with the remainder of mass being bisilylated product
and starting material. This intermediate was then methylated
in quantitative yield, and the alkynyl-TMS group was
converted into an alkynyl bromide13 in one step to give the
diyne coupling precursor 7 in 44% yield over three steps.
Serine-derived iodide 1 was then coupled to 7 using the
Knochel copper-zinc reagent11 in moderate yield. The
remainder of the mass was predominately the serine-bromide
and alkyne resulting from halogen-metal exchange. Use of
the alkynyl iodide led to a much lower yield because of
increased exchange, whereas the alkynyl chloride did not
react. Other attempts to optimize this reaction did not lead
to any increase in yield. The ester group was then selectively
reduced with LiBH4, and the primary alcohol was protected
as the TBDPS ether to produce the cyclization precursor 10
in six steps.
On a larger scale, refluxing the mixture of amines with
potassium carbonate in toluene at reflux for 68 h14 led to
nearly complete conversion to the protected cyclindricine C,
which could be isolated in 90% yield. The TBDPS group
was then removed with TBAF in THF to give a quantitative
yield of cylindricine C, whose spectra matched the natural
product. The rotation ([R]25 +61 (c ) 0.4, CH2Cl2)) was
D
opposite in sign of that reported by Molander6 ([R]25 -64
D
(c ) 0.2, CH2Cl2)) as expected.
Cylindricine C could then be transformed into cyclindri-
cine D and E in a straightforward manner. Acylation
proceeded in quantitative yield to give cyclindricine E,
whereas methylation produced cyclindricine D in 90% yield
(Scheme 6).
The ruthenium-catalyzed hydrative diyne cyclization was
carried out with 5% [CpRu(CH3CN)3]PF6 (11) in 10 vol %
In conclusion, we have shown the utility of the ruthenium-
catalyzed hydrative diyne cyclization to transform unsym-
(13) Nishikawa, T.; Shibuya, S.; Hosokawa, S.; Isobe, M. Synlett 1994
485.
(14) The time required for cyclization seemed to vary somewhat, with
larger scale reactions taking longer times.
Org. Lett., Vol. 5, No. 24, 2003
4601

play significant roles in determining the chemoselectivty. The
vinyl ketone intermediate was then transformed into cylin-
dricine C in 9 steps, and cylindricine D and E in 10 steps
from commercially available 1,7-octadiyne.
Scheme 6. Conversion of Cylindricine C into D and Ea
Acknowledgment. We thank Prof. Adrian Blackman for
copies of the authentic cylindricine spectral data. We thank
the National Institutes of Health, General Medical Science
(GM 13598), and the National Science Foundation for their
generous support of our programs. Mass spectra were
provided by the Mass Spectrometry Regional Center of the
University of California-San Francisco supported by the NIH
Division of Research Resouces.
Supporting Information Available: Experimental pro-
cedures for the preparation of new compounds and charac-
terization data. This material is available free of charge via
the Internet at http://pubs.acs.org.
a (a) Ac2O, Et3N, DMAP, DCM; (b) MeI, Ag2O, CH3CN, 3 d.
metrical diynes into a single regioisomeric vinyl ketone
product. It appears the both steric and conjugation factors
OL035752N
4602
Org. Lett., Vol. 5, No. 24, 2003