Synthesis and stereochemical elucidation of a 14-membered ring phosphonate

Article abstract of DOI:10.1021/ol006422v

We report the synthesis and stereochemical elucidation of a 14-membered ring phosphonate. The key step in the synthesis of the macrolide phosphonate was the cyclization of the acyclic precursor using the Mitsunobu reaction, a mild reaction for the prepara

Full text of DOI:10.1021/ol006422v

ORGANIC
LETTERS
2001
Vol. 3, No. 5
643-646
Synthesis and Stereochemical
Elucidation of a 14-Membered Ring
Phosphonate
,†
Michael D. Pungente* and Larry Weiler‡
Faculty of Pharmaceutical Sciences, UniVersity of British Columbia, 2146 East Mall,
VancouVer, British Columbia V6T 1Z3, Canada, and Department of Chemistry,
UniVersity of British Columbia, 2036 Main Mall, VancouVer,
British Columbia V6T 1Z1, Canada
mpungent@interchange.ubc.ca
Received August 4, 2000
ABSTRACT
We report the synthesis and stereochemical elucidation of a 14-membered ring phosphonate. The key step in the synthesis of the macrolide
phosphonate was the cyclization of the acyclic precursor using the Mitsunobu reaction, a mild reaction for the preparation of mixed phosphonates.
Phosphonate esters have been used extensively as haptens
for the production of catalytic antibodies, and for this reason,
we set out to synthesize the macrocyclic phosphonate 10.
used with carboxylic acids, phenols, phosphates, and phos-
phonic acids as the nucleophilic component. Thus we
designed the synthesis of phosphonate 10 such as to employ
the Mitsunobu reaction in the key ring-forming transforma-
tion of the monophosphonic acid 9 into the cyclic phospho-
nate 10.
The synthesis of macrocyclic phosphonate 10 (Scheme 1)
began with the Grignard reaction of undecylenic aldehyde
and methylmagnesium bromide to give the hydroxy alkene
1 in 90% yield. The hydroxyl group of 1 was subsequently
protected as the THP ether 2 in 84% yield. The two
diastereomers of 2 were not separated but were treated
together in the following sequences. Hydroboration of 2
followed by oxidation gave 3 in 96% yield. The alcohol 3
was subsequently oxidized under Swern conditions to give
the aldehyde 4 in 87% yield. Aldehyde 4 was then reacted
with the stabilized Wittig reagent4 5 in refluxing toluene to
give the vinyl phosphonate 6 in a moderate yield of 52%
with a 4:1 ratio of trans to cis isomers. The alkene 6 was
Macrocyclic phosphonates, like macrocyclic lactones,
present a challenge to the synthetic chemist because of the
difficulty in carrying out the ring-forming step. The use of
the Mitsunobu protocol by Smith et al.1 in the key lactone-
forming step in the total synthesis of the latrunculins, a class
of compounds containing a macrocyclic lactone ring, caught
our attention. Over the past two decades, the Mitsunobu
reaction has found extensive use in organic synthesis,
particularly for the inversion of the stereochemistry of
alcohols via an esterification/hydrolysis procedure.2 The
Mitsunobu reaction is a mild coupling reaction for the
preparation of mixed phosphonates from phosphonic acid
monoesters and primary or secondary alcohols.3 It has been
† Faculty of Pharmaceutical Sciences.
‡ Department of Chemistry; deceased, April 28, 1999.
(1) Smith, A. B., III; Leahy, J. W.; Noda, I.; Remiszewski, S. W.;
Liverton, N. J.; Zibuck, R. J. Am. Chem. Soc. 1992, 114, 2995.
(2) Mitsunobu, O. Synthesis 1981, 1.
(4) Jones, G. H.; Hamamura, E. K.; Moffatt, J. G. Tetrahedron Lett. 1968,
(3) Campbell, D. A. J. Org. Chem. 1992, 57, 6331.
5731.
10.1021/ol006422v CCC: $20.00 © 2001 American Chemical Society
Published on Web 02/10/2001

noester 9 was reacted with DEAD and triphenylphosphine
in benzene under high dilution conditions to produce the
cyclic phosphonates 10a and 10b in 82% yield as a 5:1
mixture of diastereomers.6
Scheme 1
In addition to the two 14-membered ring diastereomers
10a and 10b, a cyclic dimer side product 11 was isolated as
rod-shaped crystals on purification of 10a and 10b. The
structure of dimer 11 was determined by single-crystal X-ray
diffraction (Figure 1).7 Dimer 11, a centrosymmetric 28-
Figure 1. Stereoview of an ORTEP representation of dimer 11
showing 33% probability ellipsoids. Hydrogen atoms have been
given arbitrary thermal parameters for clarity.
membered ring compound with two parallel chains bridged
at both ends by the phosphonate group, contains four
stereogenic centers with R*R* stereochemistry on one bridge
and the mirror image S*S* stereochemistry at the opposite
then reduced using H2 and 10% Pd/C in EtOAc to give the
saturated diphenyl phosphonate 7 in 99% yield. Phosphonate
7 was subsequently reacted with p-toluenesulfonic acid to
liberate the secondary hydroxyl of compound 8 in 86% yield.
Phosphonic acid monoester 9 was produced in 76% yield
by base hydrolysis of 8 in refluxing potassium hydroxide/
THF.
(6) Selected data for compounds are as follows. 10a: Rf ) 0.71 (silica
gel, petroleum ether/ethyl acetate 1:1); FTIR (neat) 2927, 2859, 1593, 1490,
1
1458, 1380, 1253, 1210, 1163, 1071, 991, 920, 766 cm-1; H NMR (400
MHz, CDCl3) δ 7.05-7.38 (m, 5H), 4.75 (m, 1H), 1.24-1.88 (m, 22H),
1.22 (d, J ) 6.3 Hz, 3H); 31P NMR (81 MHz, CDCl3) δ 29.6; HRMS calcd
for C19H31O3P 338.2011, found 338.2002. Anal. Calcd for C19H31O3P: C,
67.42; H, 9.24. Found: C, 67.20; H, 9.20. 10b: mp 97-99 °C.; Rf ) 0.57
(silica gel, petroleum ether/ethyl acetate 1:1); FTIR (CDCl3) 2932, 2860,
1593, 1492, 1456, 1223, 1011 cm-1; 1H NMR (400 MHz, CDCl3) δ 7.05-
7.38 (m, 5H), 4.65 (m, 1H), 1.93 (m, 2H), 1.20-1.73 (m, 20H), 1.43 (d, J
) 6.2 Hz, 3H); 31P NMR (81 MHz, CDCl3) δ 27.8; HRMS calcd for
C19H31O3P 338.2011, found 338.2009. Anal. Calcd for C19H31O3P: C, 67.42;
H, 9.24. Found: C, 67.15; H, 9.39. 11: mp 103-105 °C; Rf ) 0.62 (silica
gel, petroleum ether/ethyl acetate 1:1); FTIR (CDCl3) 2930, 2856, 1593,
The key step5 in the synthesis of the macrocyclic phos-
phonate 10 was the cyclization of 9 to give the 14-membered
ring phosphonate 10. The hydroxy phosphonic acid mo-
(5) To a stirred solution of compound 9 (125 mg, 0.351 mmol) in 150
mL of dry PhH at room temperature under N2 was added Ph3P (368 mg,
1.40 mmol), followed by DEAD (221 µL, 1.40 mmol). After 2 h, the solvent
was removed under reduced pressure, and the resulting crude mixture was
purified by flash chromatography using 2:1 hexane and ethyl acetate to
afford the separated diastereomers, 10a and 10b, in a 5:1 ratio based on
recovered yields and a combined yield of 82%. In addition, 6 mg of dimer
11 was isolated as a colorless, crystalline solid.
1491, 1229, 1012 cm-1 1H NMR (400 MHz, CDCl3) δ 7.10-7.33 (m,
;
10H), 4.66 (m, 2H), 1.84 (m, 4H), 1.68 (m, 4H), 1.56 (m, 4H), 1.20-1.45
(m, 40H), 1.15 (d, J ) 6.2 Hz, 6H); 31P NMR (81 MHz, CDCl3) δ 29.0;
HRMS calcd for C38H62O6P2 676.4022, found 676.4002. Anal. Calcd for
C38H62O6P2: C, 67.42; H, 9.24. Found: C, 67.51; H, 9.28.
(7) Crystal data for 11: C38H62O6P2; monoclinic; P21/n; a ) 7.841(1)
Å; b ) 5.664(2) Å; c ) 43.378(1) Å; â ) 90.259(8)°; V ) 1926.4(5) Å3;
Z ) 2; colorless; T ) 294 K; R ) 0.030; GOF ) 2.27.
644
Org. Lett., Vol. 3, No. 5, 2001

bridge. Dimer 11 has a center of symmetry as a result of the
R*R* and S*S* stereochemistry at the bridge ends.
The 1H and 31P NMR spectra of the 28-membered
macrocyclic phosphonate dimer 11, together with the struc-
tural information from the X-ray structure of 11, have
allowed us to infer structural information about monomers
10a and 10b.
bridge was created by inverting the stereochemistry at the
S*S* bridge.
The stereoview of the ORTEP representation of the
macrocyclic dimer 11 (Figure 1) reveals that the doubly
bonded oxygen on the phosphorus atom is “syn” to the C-13
methine proton, and the C-1 methyl group is “syn” to the
oxygen of the phenol moiety (Figure 2). Analysis of the
This R*R*R*R* model retained one R*R* bridge common
to 11, where the aryloxy moiety attached to the phosphorus
atom occupies a pseudoequatorial position, projecting away
from the ring while the doubly bonded oxygen atom projects
into the ring (as in Figure 1). However, the R*R* stereo-
chemistry at the opposite bridge of the R*R*R*R* model
forces the methyl and aryloxy moieties to project in toward
the ring, as illustrated below. MACROMODEL calculations8
on the R*R*S*S* and R*R*R*R* dimers using the MM2
force field9 gave a lower energy conformation for the
R*R*S*S* dimer, 92.40 kJ/mol, compared to 109.23 kJ/mol
for the R*R*R*R* dimer.
Figure 2. The relative stereochemistries of the OPh and C-1 methyl
groups of compound 11.
These calculations suggest that the R*R*S*S* dimer is of
lower energy and may provide a rationale for why only this
dimer was isolated on cyclization of compound 9.
transition states for ring closure reveals that the pathway
leading to R*R* product has fewer steric interactions
compared to cyclization to S*R* product (Figure 3).
If the transition state leading to the R*R* stereochemistry
is extended to the monomers 10a and 10b, it is reasonable
to suggest that the major monomer 10a might also exhibit
the R*R* relative stereochemistry. We are suggesting,
therefore, that the major isomer 10a has the R*R* relative
stereochemistry between the C-1 methyl and the phosphorus
atom consistent with that of compound 11. In addition, we
believe that the C-1 methyl group and the OPh group on
phosphorus are mainly in a conformation in which both are
pseudoequatorial in isomer 10a as found in the X-ray
structure of 11 (Figure 2). The suggested relative stereo-
chemical assignments, therefore, between the OPh and C-1
methyl groups of macrocyclic phosphonates 10a, 10b, and
11 are as shown in Figure 4.
Figure 3. Suggested transition states A and B leading to the R*R*
and S*R* products, respectively.
The condensation of two molecules of 9 via the Mitsunobu
methodology resulted in the R*R* stereochemistry. It follows
that closure of the opposite end of the acyclic dimer would
be governed by the same steric control, resulting in either
the SS or RR stereochemistry for the second phosphonate
bridge. Why is the R*R*S*S* dimer 11 the only dimer
molecule isolated on cyclization of 9? Molecular models of
the R*R*S*S* and R*R*R*R* diastereomers of dimer 11
were calculated using MACROMODEL. The R*R*S*S*
model was calculated according to the X-ray structure
obtained for dimer 11. The R*R*R*R* model was calculated
from the same conformation as 11, where the second R*R*
Figure 4. The relative stereochemistries of the macrocyclic
phosphonates 10a, 10b, and 11.
Org. Lett., Vol. 3, No. 5, 2001
645

Support for the pseudoequatorial assignment of the OPh
group on the phosphorus atom of the 14-membered ring
phosphonate 10a comes from correlations between the H
and 31P NMR chemical shift data (see Table 1). It has been
might reveal a trend. Table 1 reveals that the 13C NMR shifts
for the C-1 methyl carbons of compounds 11, 10a, and 10b
are essentially the same. However, the 13C NMR shifts for
the C-13 carbons of compounds 11, 10a, and 10b support
our correlation between compounds 11 and 10a, as the carbon
chemical shifts for 11 and 10a are consistent (see Table 1).
Finally, the relative stereochemical assignments for iso-
mers 10a and 10b were confirmed through NOE difference
experiments performed on macrocycles 11, 10a, and 10b.
These NOE experiments revealed the same positive NOE
enhancements between the C-1 methyl protons and those
associated with the OPh group for compounds 11 and 10a;
however, the NOE experiment for isomer 10b did not reveal
these enhancements. These NOE results further confirm our
assignment for isomers 10a and 10b.
1
1
Table 1. Key H, 13C, and 31P NMR Chemical Shifts (ppm )
for the Cyclic Phosphonates
δ 1H for δ 1H for
δ
13C for
CH3
δ
13C for
C-13 methine
carbon
CH3
methine
proton
compound protons
carbon
δ
31P
10a
10b
11
1.22
1.43
1.15
4.75
4.65
4.66
22.4
22.3
22.1
74.5
75.3
74.7
29.6
27.8
29.0
The macrocyclic phosphonate, 10, was synthesized in nine
steps with an overall yield of 17%. The key step in the
synthesis of 10 was the use of the Mitsunobu reaction in the
ring-closing step, which proceeded in an 82% yield with a
5:1 ratio of the two isomers 10a and 10b. The structural
information from the X-ray structure of 11, together with
observed that the 31P NMR signal for the equatorial P-OR
isomer is approximately 2 ppm lower field than the corre-
sponding axial isomer in the cyclic phosphonates of hexo-
pyranoses10 and in other six-membered ring phosphonates.11
This trend supports our correlation between compounds 11
and 10a, since the 31P NMR shift for isomer 10a is nearly 2
ppm downfield from that of isomer 10b.
As further support for our relative stereochemical assign-
ments for isomers 10a and 10b, 13C NMR data was obtained
for all three compounds, 11, 10a, and 10b. It was anticipated
that the 13C NMR shifts of the C-1 methyl carbons and/or
the C-13 carbon (bearing the methyl group; see Figure 1)
1
the H, 13C, and 31P NMR spectra of compounds 11, 10a,
and 10b, allowed us to make structural assignments for
diastereomers 10a and 10b. Finally, comparison of NOE
difference experiments performed on the three compounds
allowed us to confirm our relative stereochemical assign-
ments for isomers 10a and 10b.
We shall report the use of 10 as a hapten for the generation
of monoclonal antibodies for the purpose of catalyzing a
macrolactonization reaction elsewhere.
(8) Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.; Lipton,
M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W. C. J. Comput. Chem.
1990, 11, 440.
(9) Burkert, U.; Allinger, N. L. ACS Monograph 177; American Chemical
Society: Washington, DC, 1982.
(10) (a) Darrow, J. W.; Drueckhammer, D. G. J. Org. Chem. 1994, 59,
2976. (b) Darrow, J. W.; Drueckhammer, D. G. Bioorg. Med. Chem. 1996,
4, 1341. (c) Hanessian, S.; Gale´otti, N.; Rosen, P.; Oliva, G.; Babu, S.
Bioorg. Med. Chem. Lett. 1994, 4, 2763. (d) Harvey, T. C.; Simiand, C.;
Weiler, L.; Withers, S. G. J. Org. Chem. 1997, 62, 6722.
(11) (a) Gorenstein, D. G. Chem. ReV. 1987, 87, 1047. (b) Chang, J.
W.; Gorenstein, D. G. Tetrahedron 1987, 43, 5187.
Acknowledgment. We are grateful to Dr. S. Rettig and
Professor J. Trotter for the X-ray crystallographic data and
analysis. We also thank Dr. Gregory Dake and Professor John
Scheffer for their helpful suggestions during the preparation
of this manuscript and the Natural Sciences and Engineering
Research Council of Canada for financial support.
OL006422V
646
Org. Lett., Vol. 3, No. 5, 2001