Effect of allylic and homoallylic substituents on cross metathesis: syntheses of prostaglandins F2α and J2

Article abstract of DOI:10.1016/j.tetlet.2006.06.140

We describe the effect of allylic (C15) and homoallylic (C11) substituents on cross metathesis reactions with Corey lactone derivatives. This strategy has led to the successful syntheses of PGF_2α and PGJ₂.

Full text of DOI:10.1016/j.tetlet.2006.06.140

Tetrahedron Letters 47 (2006) 6689–6693
Effect of allylic and homoallylic substituents on cross
metathesis: syntheses of prostaglandins F2a and J2
Neil A. Sheddan,a,* Vladimir B. Arionb and Johann Mulzera,*
aInstitut fu¨r Organische Chemie, Wa¨hringerstrasse 38, A-1090 Wien, Austria
bInstitut fu¨r Anorganische Chemie, Wa¨hringerstrasse 42, A-1090 Wien, Austria
Received 12 June 2006; revised 26 June 2006; accepted 27 June 2006
Available online 31 July 2006
Abstract—We describe the effect of allylic (C15) and homoallylic (C11) substituents on cross metathesis reactions with Corey lactone
derivatives. This strategy has led to the successful syntheses of PGF2a and PGJ2.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Prostaglandins (PGs) are present in minute amounts in
most animal cells and have demonstrated a wide range
of potent pharmacological properties.1 Their intriguing
structures have sparked the imagination of the organic
chemist, striving for a more efficient, convergent, and
naturally, where possible, more elegant synthesis; re-
flected by the scores of new PG syntheses over the years
(1–4).2 This also includes the synthesis of relatively new-
ly discovered epimeric PGs: isoprostanes and neuro-
prostanes (3) containing a cis-configuration for both
the a-and x-side chains (Fig. 1).3
As PGs are of great biological interest and unfortu-
nately only available in very low quantities, the need
for the development of efficient and flexible chemical
syntheses allowing direct access to not only naturally
occurring PGs, but also more active laboratory designed
4
analogues (Travoprost, 4) for biological screening is
strongly required.
Using a cross metathesis (CM)5 approach to install the
(E) ꢀ x-side chain double bond (C13–C14) would allow
a flexible and direct synthesis of prostaglandin type
compounds (Scheme 1).6
Commercially available Corey lactone 5 was oxidized to
the corresponding aldehyde using Dess–Martin period-
inane,7 which after treatment with one equivalent of
methyl Wittig ylide gave olefin 6 in rather a disappoint-
ing maximum yield of 30%. Addition of a further two
equivalents of methyl Wittig ylide increased the yield
to give a modest 69% (Scheme 2).
HO
CO2H
CO2H
HO
O
OH
OH
PGF2α (1)
PGJ2 (2)
HO
HO
CO2H
CO2Pr-
i
x-Side chains 8a/b were achieved starting from (R)-gly-
cidol. Protection as the trityl ether, followed by epoxide
opening with a suitable organocuprate species (generated
O
HO
HO
OH
OH
CF3
12-epi-PGF2α (3)
Travoprost (4)
O
O
Cross
Metathesis
Figure 1. Prostaglandins F2a and J2, Isoprostane F2a and Travoprost.
O
O
14
14
Keywords: Cross metathesis; Prostaglandin; Natural products; Total
synthesis.
13
13
*
Corresponding authors. Tel.: +43 1 4277 52190; fax: +43 1 4277
52189; e-mail addresses: neil.sheddan@univie.ac.at; johann. mulzer@
univie.ac.at
PG1O
PG1O
OPG2
OPG2
Scheme 1. Attachment of x-side chain via cross metathesis.
0040-4039/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.06.140

6690
N. A. Sheddan et al. / Tetrahedron Letters 47 (2006) 6689–6693
O
O
O
O
O
O
O
O
a,
a, b
OTBS
8a
66%
see Table 1
OH
BzO
PGO
BzO
OTBS
BzO
5
6
6
: PG = Bz
9b: PG = H
9a
Scheme 2. Reagents and conditions: (a) Dess–Martin periodinane,
NaHCO3, CH2Cl2, (b) MePh3P+Brꢀ, KOtBu, THF.
O
O
b,
see Table 1
OTBS
8a
Mixture of
Compounds
from n-butylmagnesium bromide), and TBS protection
of the resulting secondary alcohol, 7 was obtained in
74% yield (over three steps). Chemoselective trityl ether
deprotection was effected with Et2AlCl, to give the corre-
sponding primary alcohol, which, after Swern oxidation
and Wittig reactions, gave allylic alcohol 8a. Desilylation
to give 8b was achieved with TBAF (Scheme 3).
PGO
OTBS
9a: PG = Bz
9b: PG = H
Scheme 4. Reagents and conditions: (a) 6 mol % Grubbs’ 2nd gen. cat.
(via syringe pump 8 h/12 h), CH2Cl2; (b) 6 mol % Hoveyda–Grubbs’
2nd gen. cat. (via syringe pump 8 h/12 h), CH2Cl2.
With both bicyclic olefin 6 and allylic alcohol 8a in
hand, we were now in a position to try the CM reaction.
Treatment of olefin 6 (1 equiv) and allylic alcohol 8a
(2 equiv) with 6 mol % of Grubbs’ 2nd generation cata-
lyst over 8 h gave CM product 9a, with a modest E:Z
ratio (7:1) and 40% yield. In addition to the desired CM
product 9a, C11 deprotected benzoate CM compound
9b was also isolated, also with modest E:Z ratio (6:1)
and 25% yield. Unfortunately, the combined yield was
disappointing (65%) due to a large percentage of both
bicyclic olefin and x-side chain olefin undergoing homo-
dimerization. On changing to the Hoveyda–Grubbs’ 2nd
generation catalyst, the desired CM product 9a was ob-
tained, again, in poor yield. A complex mixture of com-
pounds was also isolated, which again included
deprotected benzoate CM compound 9b (Scheme 4,
Table 1).
Although the initial results for this CM reaction were
positive, the benzoate protecting group at the homo-
allylic C11 was obviously unstable under these conditions
and therefore unsuitable. Furthermore, as protecting
groups for allylic and homoallylic alcohols have previ-
ously been documented to have an influence on meta-
thesis reactions,8 in an analogous example (albeit allylic,
not homoallylic) catalyst deactivation was observed,
which could account for our observed low yields9
(Scheme 5). We therefore chose to investigate the effect
of both a free hydroxyl group and a silyl protected
(TBS ether) at both the C11 homoallylic, and at the
C15 allylic positions.
a, b, c
HO
) -glycidol
TrO
O
A simple interconversion of the C11 benzoate to the
TBS ether was imagined. Cleavage of the benzoate pro-
tecting group with K2CO3 proceeded without incident
but unfortunately, TBS protection proved to be more
challenging than first expected as treatment with both
TBSCl and TBSOTf only delivered disappointingly
low yields of TBS ether 13a (Scheme 6). As the C11 ben-
zoate protecting group had been problematic over sev-
eral steps, resulting in low yield, it was decided that
installation of this TBS ether would be more efficient
at an earlier stage (Scheme 6). Corey lactone 5 was pro-
74%
OTBS
(
R
7
d, e, f
88%
8a: R = TBS
8b: R = H
g quant.
OR
Scheme 3. Reagents and conditions: (a) TrCl, NEt3, DMAP, CH2Cl2,
rt, 5 h, (b) CuI, n-butylmagnesium bromide, THF, ꢀ30 ꢁC, (c) TBSCl,
imid., DMF, rt, (d) Et2AlCl, CH2Cl2, ꢀ50 ꢁC, 2 h, (e) (COCl)2,
DMSO, NEt3, CH2Cl2, ꢀ78 ꢁC, (f) MePh3P+Brꢀ, KOtBu, THF, rt, (g)
1 M TBAF, THF.
Table 1. Results from CM reactions from Scheme 4
Entrya
Bicyclic olefin
PG
x-Side chain olefinb
Product
E:Z
Yield (%)c
40
9a
7:1
1
6
Bz
9b
6:1
25
OTBS
2
6
Bz
9a
9b
7:1
6:1
25
15
OTBS
a Catalyst added over 8 h or 12 h in CH2Cl2 at 40 ꢁC.
b 2 equiv of x-side chain used where up to 0.5 equiv was recovered.
c Isolated yields.

N. A. Sheddan et al. / Tetrahedron Letters 47 (2006) 6689–6693
6691
O
O
O
O
O
O
O
O
O
Ru
a
-
Ru
see Table 2
O
Ru
O
PGO
PGO
OR
O
13a, PG = TBS
13b, PG = H
14a: PG = TBS, R = TBS
14b: PG = H, R = H
14c: PG = H, R = TBS
Ph
Ph
6
10
Scheme 5. Possible deactivation of catalyst by C11 benzoate group.
O
O
b
OTBS
OTBS
O
O
OTBS
O
O
TBSO
a, b
15
O
69%
O
BzO
PGO
16
6
13a: PG = TBS
13b: PG = H
g
Scheme 7. CM reaction of bicyclic olefins. Reagents and conditions:
(a) Olefin 8a/b, 6 mol % Grubbs’ 2nd gen. cat. (via syringe pump 8 h/
12 h), CH2Cl2, (b) olefin 8a/b, 6 mol % Hoveyda–Grubbs’ 2nd gen. cat.
(via syringe pump 8 h/12 h), CH2Cl2.
d, e, f
83%
O
O
O
O
c
a, b
95%
Treatment of bicyclic olefins 13a/b and allylic x-side
chain olefins 8a/b with Grubbs’ 2nd generation catalyst
over 8 or 12 h gave to our delight CM product 14a with
moderate yield and E:Z selectivity (Scheme 7, Table 2).
Bicyclic olefin 13a, with a C11 TBS group, gave an
immediate increase in yield and similar E:Z selectivity,
when compared with its benzoate analogue 6. On
increasing the catalyst addition time (8 ! 12 h), reaction
yields increased dramatically from 60% to 70% to over
80% (Table 2, entry 2). Furthermore, deprotection of
either the C11 TBS ether or C15 TBS protected alcohols
did not have a significant effect where only a mild sacri-
fice in both yield and E:Z selectivity was observed. The
observed drop in yields was due to the respective com-
pound with a free hydroxyl group at either C11 or
C15 undergoing more rapid homodimerization. A
change from Grubbs’ 2nd generation catalyst to the
Hoveyda–Grubbs’ 2nd generation catalyst favoured
homodimerization, of both bicyclic olefin 13a and x-side
chain 8a, over a CM reaction, resulting in x-side chain
dimer 15 and bicyclic dimer 16 (Scheme 7).
5
95%
OTr
OTr
BzO
TBSO
11
12
Scheme 6. Reagents and conditions: (a) K2CO3, MeOH; (b) TBSCl,
imid. DMF or TBSOTf, 2,6-lutidine, CH2Cl2, (c) TrCl, DMAP, NEt3,
CH2Cl2, (d) Et2AlCl, CH2Cl2, ꢀ50 ꢁC, 2 h, (e) Dess–Martin period-
inane, NaHCO3, CH2Cl2, (f) Ph3P+MeBrꢀ, KOtBu, THF (g) 1 M
TBAF, THF.
tected as the trityl ether, which after deprotection of the
C11 benzoate and subsequent re-protection as the TBS
ether gave 12. Trityl ether deprotection, with Et2AlCl,
followed by Dess–Martin oxidation and Wittig olefin-
ation reactions delivered desired olefin 13a in a much
improved yield of 75% (over six steps) compared with
a previous yield of 46% (over four steps).
The results of X-ray diffraction studies of both benzoate
11 and TBS ether 12 as single crystals, obtained by slow
diffusion of hexane into the corresponding chloroform
solutions, are shown in Figure 2.
CM product 14a was reduced with DIBAL-H to give its
corresponding lactol (as a mixture of epimers) which
was immediately subjected to Wittig conditions with
ylide 17, to give exclusively the C5–C6 (Z)-double bond
(Scheme 8). Further treatment with a 0.5 N HCl solu-
tion for two days yielded PGF2a (1).13 Complementary,
14c, with a free C11 hydroxyl group, which also under-
went CM, could be converted to PGJ2 (2), following the
procedure of Zanoni et al.14
In conclusion, we have shown from the results in Tables
1 and 2 that protection of the C11 homoallylic hydroxyl
group clearly has an inhibiting effect on homodimeriza-
tion. However, a careful choice of protecting group is
also critical for good yield and E:Z selectivity for the
cross metathesis reaction. This strategy has allowed
the syntheses of prostaglandins F2a and J2; both
Figure 2. ORTEP10 representations of (a) benzoate 11 and (b) TBS
ether 12 (ellipsoids at 50% probability).11

6692
N. A. Sheddan et al. / Tetrahedron Letters 47 (2006) 6689–6693
Table 2. Results from CM reactions from Scheme 612
Entrya
1
Bicyclic olefin
PG
x-Side chain olefinb
Product
E:Z
Yield (%)c
76
13a
TBS
14a
6:1
OTBS
OTBS
OH
2d
3
13a
13a
13b
TBS
TBS
H
14a
14b
14c
17:1
12:1
14:1
84
60e
70
4
OTBS
a Catalyst added over 8 h in CH2Cl2 at 40 ꢁC.
b 2 equiv of x-side chain used where up to 0.5 equiv was recovered.
c Isolated yields (not optomized).
d Catalyst added over 12 h at 40 ꢁC.
e 30% homodimerized x-side chain obtained.
O
Maruyama, T.; Ohuchida, S.; Nakai, H.; Hashimoto, S.;
Kawamura, M.; Kondo, K.; Toda, M. Bioorg. Med.
Chem. 2002, 10, 2103–2110; (e) Komoto, J.; Yamada, T.;
Watanabe, K.; Woodward, D. F.; Takusagawa, F.
Biochemistry 2006, 45, 1987–1996, and references therein.
2. For a recent review, see: Collins, P. W.; Djuric, S. W.
Chem. Rev. 1993, 93, 1533–1564.
O
HO
CO2H
a, b, c
70%
6
5
PGO
HO
OTBS
OH
PGF2α (1)
14a: PG = TBS
14c: PG = H
3. (a) Morrow, J. D.; Awad, J. A.; Boss, H. J.; Blair, I. A.;
Roberts, L. J., II. Proc. Natl. Acad. Sci. U.S.A. 1992, 89,
10721–10725; (b) Roberts, L. J., II; Montine, T. J.;
Markesbery, W. R.; Tapper, A. R.; Hardy, P.; Chemtob,
S.; Dettbarn, W. D.; Morrow, J. D. J. Biol. Chem. 1998,
273, 13605–13612, and references therein.
CO2H
Ref. 14
6
5
O
OH
4. Klimko, P. G.; Bishop, J.; Desantis Jr., L.; Sallee, V. L.,
EP 639563 A2 950222, US 93-101598 930803.
PGJ2 (2)
5. For recent reviews on cross metathesis, see: (a) Chatterjee,
A. K.; Choi, T.-L.; Sanders, D. P.; Grubbs, R. H. J. Am.
Chem. Soc. 2003, 125, 11360–11370; (b) Van de Weghe, P.;
Eustache, J.; Cossy, J. Curr. Topics Med. Chem. 2005, 5,
1461–1472; (c) Prunet, J. Curr. Topics Med. Chem. 2005, 5,
1559–1577.
Scheme 8. Prostaglandins F2a and J2. Reagents and conditions: (a)
DIBAL-H, CH2Cl2, ꢀ78 ꢁC, (b) HOOC(CH2)4Ph3P+Brꢀ (17), KOtBu,
THF, (c) 0.5 N HCl, THF.
6. (a) Sheddan, N. A.; Mulzer, J. Org. Lett. 2006, 8, 3101–
3104; (b) Jacobo, S. H.; Chang, C.-T.; Lee, G.-J.; Lawson,
J. A.; Powell, W. S.; Pratico, D.; FitzGerald, G. A.;
Rokach, J. J. Org. Chem. 2006, 71, 1370–1379.
7. Fleming, I.; Winter, S. B. D. J. Chem. Soc., Perkin Trans.
1 1998, 2687–2700.
accessed using a highly selective cross metathesis reac-
tion, as the key step, for introduction of the x-side chain
moitey. This flexible strategy could allow the syntheses
of a large number of structurally interesting prostaglan-
din analogues for further biological screening.
8. For the effects of allylic and homoallylic substituents on
metathesis see (and references cited therein); (a) Castoldi,
Acknowledgements
´
D.; Caggiano, L.; Bayon, P.; Costa, A. M.; Cappella, P.;
Sharon, O.; Gennari, C. Tetrahedron 2005, 61, 2123–2139;
´
(b) Caggiano, L.; Castoldi, D.; Beumer, R.; Bayon, P.;
We would like to thank Schering A. G., Berlin, for
financial support for N.A.S. and also the generous gift
of Corey lactone 5. We also thank Hanspeter Ka¨hlig
for NMR.
Telser, J.; Gennari, C. Tetrahedron Lett. 2003, 44, 7913–
7919; (c) McNaughton, B. R.; Bucholtz, K. M.; Camaano-
˜
Moure, A.; Miller, B. L. Org. Lett. 2005, 7, 733–736; (d)
Kawaguchi, T.; Funamori, N.; Matsuya, Y.; Nemoto, H.
J. Org. Chem. 2004, 69, 505–509.
9. (a) BouzBouz, S.; Cossy, J. Org. Lett. 2001, 3, 1451–1454;
(b) BouzBouz, S.; Simmons, R.; Cossy, J. Org. Lett. 2004,
6, 3465–3467.
References and notes
1. For lead references and recent reviews see: (a) Noyori, R.;
Suzuki, M. Science 1993, 259, 44–45; (b) Sugrue, M. F. J.
Med. Chem. 1997, 40, 2793–2809; (c) Clark, A. F.
Emerging Drugs 1999, 4, 333–353; (d) Maruyama, T.;
Kuwabe, S.-I.; Kawanaka, Y.; Shiraishi, T.; Shinagawa,
Y.; Sakata, K.; Seki, A.; Kishida, Y.; Yoshida, H.;
10. Farrugia, L. J. J. Appl. Cryst. 1997, 30, 565.
11. Crystal structure data for (3aR,4S,5R,6aS)-2-oxo-4-
(trityloxymethyl)-hexahydro-2H-cyclopenta(b)furan-5-yl
benzoate (11) at 100 K: C34H30O5, Mr = 518.60, ortho-
rhombic, space group P212121, a = 10.1115(7), b =
3
˚
˚
15.8912(13), c = 16.4928(14) A, V = 2650.1(4) A , F(000) =

N. A. Sheddan et al. / Tetrahedron Letters 47 (2006) 6689–6693
1096, Dc (Z = 4) = 1.300 g cmꢀ3, l(MoKa) = 0.86 cmꢀ1
6693
,
flash column chromatography using a hexanes:ethyl ace-
tate (10:1) mixture as the mobile phase to afford the
corresponding cross metathesis product 14a. 1H NMR
(600 MHz, CDCl3) 5.50 (1H, ddd, J 15.4, 5.8, 1.0,
HC@CH), 5.37 (1H, ddd, J 15.4, 7.7, 1.0, HC@CH),
4.94 (1H, ddd, J 7.1, 7.1, 2.1, HCOC@O), 4.04 (1H, ddd, J
5.9, 5.7, 5.7, C@CHCHOTBS), 3.98 (1H, ddd, J 5.3, 5.3,
5.3, CHOTBS), 2.75 (1H, dd, J 18.1, 10.4, OC@OCHH),
2.67–2.61 (1H, m, HC@CHCHCH), 2.49 (1H, dd, J 18.1,
2.8, OC@OCHH), 2.44 (dd, J 12.9, 6.0, H2C@CHCH),
2.24 (1H, ddd, J 14.8, 7.0, 5.8, CHHCHOTBS), 1.98 (1H,
ddd, J 14.8, 4.9, 1.9, CHHCHOTBS), 1.51–1.44 (1H, m,
CHOTBSCHH), 1.44–1.37 (1H, m, CHOTBSCHH), 1.34–
1.20 (6H, m, 3 · alkyl CH2), 0.90–0.88 (3H, m, alkyl CH3),
0.88 (9H, s, SiC(CH3)3), 0.87 (9H, s, SiC(CH3)3), 0.05 (3H,
s, SiCH3), 0.04 (3H, s, SiCH3), 0.03 (3H, s, SiCH3), 0.00
(3H, s, SiCH3); 13C NMR (100 MHz, CDCl3) 177.0 (q),
135.8 (t), 128.4 (t), 83.4 (t), 79.0 (t), 72.9 (t), 56.1 (t), 42.3
(t), 40.7 (s), 38.4 (s), 35.0 (s), 31.8 (s), 25.9 (p), 25.7 (p),
25.0 (s), 22.6 (s), 18.2 (q), 18.0 (q), 14.0 (p), ꢀ4.3 (p), ꢀ4.7
(p), ꢀ4.8 (p), ꢀ4.9 (p); HRMS (160 ꢁC 70 eV): m/z calcd
for C27H52O4Si2: 496.3404; found 496.3395; ꢀ19.2
(c = 0.43, acetone).
crystal dimensions 0.18 · 0.14 · 0.12 mm, 2hmax 60ꢁ, wR
(all 7845 data) 0.1058, conventional (6630 data
with I > 2r(I)) 0.042. Crystal structure data
for (3aR,4S,5R,6aS)-5-(tert-butyldimethylsilyloxy)-4-(tri-
tyloxymethyl)- hexahydrocyclopenta(b)furan-2-one (12)
at 100 K:C33H40O4Si1, Mr = 528.75, orthorhombic, space
R
group P212121, a = 9.3904(6), b = 15.6368(9), c =
3
˚
˚
20.6870(11) A, V = 2650.1(4) A , F(000) = 1136, Dc
(Z = 4) = 1.156 g cmꢀ3
, , crystal
l(MoKa) = 1.11 cmꢀ1
dimensions 0.30 · 0.20 · 0.20 mm, 2hmax 50ꢁ, wR (all
3038 data) 0.1898, conventional R (2521 data with
I > 2r(I)) 0.0631. Crystallographic data (excluding struc-
ture factors) for the structures in this paper have been
deposited with the Cambridge Crystallographic Data
Centre as supplementary publication numbers CCDC
602481 and 602482, respectively. Copies of the data can be
obtained, free of charge, on application to CCDC, 12
Union Road, Cambridge CB2 1EZ, UK [fax: +44(0)-
1223-336033 or e-mail: deposit@ccdc.cam.ac.uk].
12. Representative procedure for cross metathesis (Table 2,
entry 2): To a solution of bicyclic olefin 13a (0.098 g,
0.38 mmol) and x-side chain olefin 8a (0.168 g, 0.69 mmol)
in anhydrous degassed CH2Cl2 (7 mL) was added a
solution of Grubbs’ 2nd generation catalyst (6 mol %,
0.018 g, 0.021 mmol) in anhydrous degassed CH2Cl2
(2 mL) via a syringe pump over 12 h at 40 ꢁC. After the
reaction was complete (monitored by TLC), air was
bubbled through the solution to destroy the catalyst.
The mixture was concentrated in vacuo and purified by
13. 1H, 13C and optical rotation are concurrent with literature
values (a) Parve, O.; Jarving, I.; Martin, I.; Metsala, A.;
Vallikivi, I.; Aidnik, M.; Pehk, T.; Samel, N. Bioorg. Med.
Chem. Lett. 1999, 9, 1853–1858; (b) Sato, Y.; Takimoto,
M.; Mori, M. Chem. Pharm. Bull. 2000, 48, 1753–1760.
14. Zanoni, G.; Porta, A.; De Toma, Q.; Castronova, F.;
Vidari, G. J. Org. Chem. 2003, 68, 6437–6439.