Stereocontrolled synthesis of cyclopropanol amino acids from allylic sulfones: Conformationally restricted building blocks

Article abstract of DOI:10.1021/ol035451d

(Matrix presented) A new amino acid methyl ester with a cyclopropanol has been synthesized starting from the allyl sulfone 10. The starting material, 10, could be obtained in both enantiomeric forms. The stereoselectivity of the cyclopropane formation has

Full text of DOI:10.1021/ol035451d

ORGANIC
LETTERS
2003
Vol. 5, No. 20
3687-3690
Stereocontrolled Synthesis of
Cyclopropanol Amino Acids from Allylic
Sulfones: Conformationally Restricted
Building Blocks
David D´ıez,* P. Garcia, Isidro S. Marcos, N. M. Garrido, P. Basabe,
Howard B. Broughton,† and Julio G. Urones
Departamento de Qu´ımica Orga´nica, UniVersidad de Salamanca,
Plaza de los Ca´ıdos 1-5, 37008 Salamanca, Spain, and Lilly S.A.,
AVenida de la Industria, 30, 28108 Alcobendas, Madrid, Spain
ddm@usal.es
Received July 31, 2003
ABSTRACT
A new amino acid methyl ester with a cyclopropanol has been synthesized starting from the allyl sulfone 10. The starting material, 10, could
be obtained in both enantiomeric forms. The stereoselectivity of the cyclopropane formation has been studied by molecular modeling.
The cyclopropane ring is present in a wide variety of
naturally occurring compounds, many of these with interest-
ing biological activities.1 Although there are not many
naturally occurring amino acids containing a cyclopropyl
group,2 a recently isolated natural product, belactosin A, 1
(Figure 1), contains a (2S,1′R,2′S)-3-(trans-2-aminocyclo-
propyl)alanine amino acid 2 as its core feature. This
interesting natural product modulates cell-cycle progression
via inhibition of cyclin-cdk complexes, showing weak
antitumor activity.3
energy minimum conformation of 3 (as determined by a 15°-
resolution full-circle torsion-driving conformational search
of all sp3-sp3 rotatable bonds in MacroModel5 with the GB/
SA water solvation model) and the pharmacologically
relevant extended conformation of glutamate can be seen
from the overlay in Figure 2. The cyclopropanol OH group
is almost perfectly superimposed on the δ acid oxygen of
glutamate.
In recent years, we have published a methodology for the
synthesis of cyclopropanols from allylic sulfones. These
Up to the present time, only two syntheses of this amino
acid 2 have been reported, by the groups of de Meijere and
recently by Armstrong.4 Routes to analogues of belactosin
A are clearly of interest, with a view to improving its weak
antitumor activity. Conformationally restricted analogues of
important amino acids such as glutamate or homoserine are
also of general interest. The close analogy between the global
† Lilly S.A.
(1) Salau¨n, J.; Baird, M. S. Curr. Med. Chem. 1995, 2, 511-542.
(2) Salau¨n, J. Top. Curr. Chem. 2000, 207, 1-67.
(3) (a) Asai, A.; Mizukami, T.; Yamashita, Y.; Akinaga, S.; Ikeda, S.;
Kanda, Y. Eur. Patent 1166781, 2002; Chem Abstr. 2002, 133, 120677. (b)
Asai, A.; Hasegawa, A.; Ochiai, K.; Yamashita, Y.; Mizukami, T. J. Antibiot.
2000, 53, 81-83
Figure 1.
10.1021/ol035451d CCC: $25.00
© 2003 American Chemical Society
Published on Web 09/11/2003

Scheme 2
Figure 2. Overlay of Glutamic acid with 3.
cyclopropanols also contain a vinyl sulfone unit whose
reactivity could be exploited further.6 In this paper, we
describe the enantioselective synthesis of the methyl ester
of 3, which could be used as a core structure for the
construction of belactosin A analogues.
we were able to increase the trans/cis ratio of the product
cyclopropanes to 95:5, independent of the protecting group
used (Scheme 3).
We have previously obtained cyclopropanols by treatment
of allylic sulfones such as 4 or 5 with LDA, with a
diastereoselectivity of 70:30 or 30:70 for the trans:cis
cyclopropanols, depending on the protecting group (for
example, see Scheme 1).
Scheme 3
Scheme 1
When compounds 14, 15 and 16 were treated under the
conditions described for the cyclopropanol formation, vinyl
cyclopropanols 6, 8, and 17, respectively, were obtained with
excellent diastereoselectivity (Scheme 3).
The stereochemistry of compound 17 was established by
X-ray crystallography (Figure 3).
Using the same methodology used previously7 (Scheme
2), we were able to obtain compounds 14 and 15, which
differ with respect to compounds 4 and 5 only in the
stereochemistry of the allyl sulfone, and the protecting group
variant 16. Here we report that by means of this stereochem-
ical change, in combination with a change in the base used,
(4) (a) Brandl, M.; Kozhushkov, S. I.; Loscha, K.; Kokoreva, O. V.;
Yufit, D. S.; Howard, J. A. K.; de Meijere, A. Synlett 2000, 1741-1744.
(b) Larionov, O. V.; Savel’eva, T. F.; Kochetkov, K. A.; Ikonnokov, N. S.;
Kozhushkov, S. I.; Yufit, D. S.; Howard, J. A. K.; Khrustalev, V. N.;
Belokon, Y. N.; de Meijere, A. Eur. J. Org. Chem. 2003, 869-877. (c)
Armstrong, A.; Scutt, J. M. Org. Lett. 2003, 5, 2331-2334.
(5) MacroModel V8.1.031; Schrodinger, LLC: Portland, OR.
(6) (a) Diez, D.; Garc´ıa, P.; Pacheco, M. P.; Marcos, I. S.; Garrido, N.
M.; Basabe, P. and Urones, J. G. Synlett 2002, 355. (b) Diez, D.; Garc´ıa,
P.; Marcos, I. S.; Garrido, N. M.; Basabe, P., Urones, J. G. Synthesis 2003,
53
Figure 3. Single-crystal X-ray structure of 17.
Since our initial attempts to try to understand why the
stereoselectivity was so much better for the cis-allyl sulfone
(7) D´ıez, D.; San Feliciano, S. G.; Marcos, I. S.; Basabe, P.; Garrido, N.
M.; Urones, J. G. Synthesis 2001, 1069.
3688
Org. Lett., Vol. 5, No. 20, 2003

(such as 16) than it had been for trans-allyl sulfones (such
as 4) using Newman projections or Dreiding models failed
to provide a convincing explanation, we used a simple
molecular modeling approach to help guide us in the
formulation of a hypothesis. It has been suggested8 that there
are a number of accessible structures for solvated allyl
sulfone lithium species, including some in which the lithium
is remote from the reacting site. The metal ion and solvation
shell are thus less likely to dominate the reaction profile,
justifying a highly simplified model in which the cation and
solvent were ignored. Further support for this simplification
came from our experiments, which showed that the stereo-
chemical outcome of the cis-allyl sulfone cyclization was
essentially independent of the cation. Without detailed
knowledge of the solvation state and the location of the
cation, the conclusions from these studies must be seen as a
starting point for future, more detailed theoretical and
experimental investigations, but nonetheless provide some
support for an initial hypothesis as to the origin of the
observed stereocontrol. We initially considered the possibility
that the transition state would be starting-material-like and
studied the likely conformational preferences of the anion
of 16 and of the trans double-bond isomer of 16. A grid
search of the conformational energy was carried out on the
anionic species using the MMFF94S force field and charge
model in Sybyl.9 The C2-C1 bond was rotated 360° in 15°
increments and the C1-O bond rotated 360° in 60° incre-
ments. However, the conformational profiles of both cis- and
trans-allyl sulfone anions were very similar, with the lowest-
energy conformation in each case being one that would lead
to the cis-cyclopropane, with conformations that would give
rise to the trans-cyclopropane, higher in energy by ca. 2 kcal/
mol. Simplified structures (methyl instead of phenyl, and
methoxy instead of MOM) based on the local energy minima
located by MMFF were subjected to geometry optimization
at the ab initio HF/3-21G* level (using Spartan10) and showed
slightly decreased energy differences but the same general
trend (preference for the cis-cyclopropane precursor confor-
mation of 1.56 and 0.65 kcal/mol for the cis and trans double
bond, respectively, clearly not in line with observation). We
therefore considered the possibility that the transition state
was product-like and examined the conformations of the
product cyclopropanes 17 and 18. We constrained the C7-
C6-C2-C1 torsion to the eclipsed syn and anti positions,
modeling the expected initial product of the cis- and trans-
allyl sulfone cyclizations, respectively. The top two structures
in Figure 4 are models for product-like transition states of
trans-allyl sulfone cyclization, while those in the lower half
are models for the cyclization of the cis allyl sulfones. The
structures on the left are trans-cyclopropanes (conformers
of 17), and those on the right are cis-cyclopropanes (con-
formers of 18). In each case, the trans-cyclopropane is
slightly lower in MMFF94s energy. The difference is larger
for the bottom two structures in Figure 4, based on the cis-
allyl sulfone, than for the top two structures, albeit by only
Figure 4. Models for product-like transition states for trans- (top)
and cis- (bottom) allylsulfone cyclizations to trans- (left side) or
cis- (right side) cyclopropanes.
0.4 kcal/mol; this is consistent with a stronger preference
for trans-cyclopropane formation from the cis double-bond
precursor than from the trans double-bond precursor. As a
difference between energy differences, the 0.4 kcal/mol value
should benefit from cancellation of error and hence be a
useful indicator of the trend, despite its magnitude. Repetition
of the calculation at the ab initio HF/6-31G* level in Spartan
suggests that cyclization of the trans isomer of 16 should
preferentially give rise to the cis-cyclopropane (difference
of 0.87 kcal/mol), while cyclization of the cis compound 16
should preferentially give rise to the trans-cyclopropane
(difference 0.93 kcal/mol); hence, the difference between
energy differences at the HF/6-31G* level is widened to 1.8
kcal/mol. This appears to arise from an unfavorable interac-
tion between the proton on the C7 carbon and the oxygen
atom of the MOM group, as can be seen in the lower right
structure in Figure 4. This interaction would indeed be
expected to be significant in a late transition state for the
cyclization of 16 to the cis-cyclopropane 18 but is absent
for the trans product 17 or for trans-allyl sulfone starting
materials. Thus, it seems reasonable to suggest that the
stereochemical outcome of this reaction is due to a fairly
late transition state in which the unfavorable C7-O interac-
tion is avoided to give the trans-cyclopropane 17 as the major
product.
The versatility of vinyl sulfones in organic chemistry11 is
well-known, in particular as Michael acceptors and in
stabilization of an anion at the R-position.12 These com-
pounds seemed to us to be excellent starting materials to
explore the utility of lithiated Scho¨llkopf’s bislactim ether13
in Michael addition with vinyl sulfones. This methodology
has been widely employed in aldol reactions and Michael-
type reactions with ester, nitro, R,â-unsaturated 2,4-penta-
dienoates and 1,3-butadienylphosphonates.14
(11) (a) Ba¨ckvall, J.-E.; Juntunen, S. K. J. Org. Chem. 1988, 53, 2398.
(b) Fuchs, P. L.; Hentemann, M. F. Org. Lett. 1999, 1, 1355. Iradier, F.;
Arrayas, R. G.; Carretero, J. C. Org. Lett. 2001, 3, 2957. (c) Padwa, A.;
Murphree, S. S.; Ni, Z. J. J. Org. Chem. 1996, 61, 3829. (d) Ba¨ckvall,
J.-E.; Chinchilla, R.; Na´jera, C.; Yus, M. Chem. ReV. 1998, 98, 2291.(e)
Iradier, F.; Arrayas, R. G.; Carretero, J. C. Org. Lett. 2001, 3, 2957. (f)
Ba¨ckwall, J.-E.; Chinchilla, R.; Na´jera, C.; Yus, M. Chem. ReV. 1998, 98,
2291
(8) Piffl, M.; Weston, J.; Gu¨nther, W.; Anders, E. J. Org. Chem. 2000,
65, 5942
(9) Sybyl, version 6.9; Tripos, Inc.: St. Louis, MO.
(10) Spartan 02 Windows; Wavefunction, Inc.: Irvine, CA.
Org. Lett., Vol. 5, No. 20, 2003
3689

The lithium salt of 19 rapidly adds to compound 17 at
low temperatures to afford the addition product 20 in
excellent yield with nearly complete diastereoselectivity. This
stereoselectivity could be understood by reference to transi-
tion states similar to those proposed by Scho¨llkopf for the
addition of the lithiated bislactim ether to R,â-unsaturated
ester or nitro compounds.13,14b The stereochemistry of the
addition product 20 was established by study of the NMR
spectra and NOE experiments (see Scheme 4.)
Scheme 5
Scheme 4
unfavorable interactions between C7 and the protected
secondary alcohol are minimized. These compounds have
been used for the synthesis of a chiral unnatural amino acid
of restricted conformation that could be used for the synthesis
of analogues of belactosin A or other peptides. The synthesis
makes use of Scho¨llkopf lithiated bislactim ether methodol-
ogy with vinyl sulfones, a methodology that has not been
used before. This opens the way for the synthesis of a large
variety of cyclopropanol amino acids due to the simplicity,
high yield, and high diastereocontrol of the method. All
compounds could be obtained in both enantiomeric forms
by choosing the appropriate starting material, easily available
in both enantiomeric forms.6
Finally, deprotection first of the sulfone group (under the
usual conditions) and then successive (via 22) or one-pot
hydrolysis of the protecting groups led to the cyclopropanol
amino acid methyl ester 23 (Scheme 5).
Acknowledgment. The authors thank the CICYT (BQU
2001-1034), Junta Castilla y Leo´n, for financial support (SA
13-00B), Universidad de Salamanca for a doctoral fellowship
to P.G, and Dr. F. Sanz for X-ray structure determination.
In conclusion, we have obtained cyclopropanols substituted
with a vinyl sulfone with high diastereoselectivity, which
we attribute to a product-like transition state in which
Supporting Information Available: Experimental pro-
cedures and complete characterization for all new com-
pounds, as well as X-ray crystalographic analysis data for
17. This material is available free of charge via the Internet
at http://pubs.acs.org.
(12) Simpkins, N. S. Sulphones in Organic Synthesis. In Tetrahedron
Organic Series; Pergamon: New York, 1993.
(13) Scho¨llkopf, U.; Pettig, D.; Busse U. Synthesis 1986, 737-740.
(14) (a) Williams, R. M. Synthesis of Optically Active R-Amino Acids.
In Tetrahedron Organic Series; Pergamon: New York, 1989. (b) Scho¨llkopf,
U.; Ku¨hnle, W; Egert, E.; Dyrbusxch, M. Angew. Chem., Int. Ed. 1987,
26, 480-482. (c) Ojea, V.; Conde, S.; Ruiz, M.; Ferna´ndez, M. C.; Quintela,
J. M. Tetrahedron Lett. 1997, 38, 4311-4314.
OL035451D
3690
Org. Lett., Vol. 5, No. 20, 2003