A preliminary evaluation of a metathesis approach to bryostatins

Article abstract of DOI:10.1016/j.tetlet.2006.01.097

Preliminary investigations into the synthesis of bryostatins using ring-closing metathesis to form the C(16)-C(17) double bond led to a synthesis of the bryostatin analogue 51; precursors 26 and 52, which possess the geminal dimethyl group at C-18, did no

Full text of DOI:10.1016/j.tetlet.2006.01.097

Tetrahedron Letters 47 (2006) 2223–2227
A preliminary evaluation of a metathesis
approach to bryostatins
Matthew Ball, Benjamin J. Bradshaw, Raphae¨l Dumeunier, Thomas J. Gregson,
Somhairle MacCormick, Hiroki Omori and Eric J. Thomas*
The School of Chemistry, The University of Manchester, Manchester M13 9PL, UK
Received 22 December 2005; accepted 18 January 2006
Available online 10 February 2006
Abstract—Preliminary investigations into the synthesis of bryostatins using ring-closing metathesis to form the C(16)–C(17) double
bond led to a synthesis of the bryostatin analogue 51; precursors 26 and 52, which possess the geminal dimethyl group at C-18, did
not undergo the required ring-closing metathesis.
Ó 2006 Elsevier Ltd. All rights reserved.
The bryostatins are marine macrolides with potent anti-
neoplastic activity. They act by modulation of the activ-
ity of protein kinase Cs and, in conjunction with other
chemotherapies, are in clinical trials for the treatment of
cancer.1 Three total syntheses of bryostatins have been
reported to date,2 a novel series of advanced acetal-con-
taining analogues has been prepared,3 and aquaculture
techniques are able to provide 100 g quantities of bryo-
statin 1 (1) per annum for clinical trials. Nevertheless,
there remains a need for better access to bryostatins
and analogues for further biological investigations.
ration and macrolactonisation.2 However, the efficiency
of this Julia reaction, which involves deprotonation a to
a quaternary centre, is very substrate dependent,1 and
can result in low yields or in syntheses which lack con-
vergency. It is therefore of interest to develop alternative
strategies for assembly of the bryostatin nucleus.
We have developed synthetic approaches to the
C(1)–C(16) and the C(17)–C(27) fragments of the 20-
deoxybryostatin, bryostatin 11 (2).4,5 We now report
preliminary studies into the assembly of the bryostatin
nucleus using ring-closing metathesis to form the
C(16)–C(17) double bond.6
Me Me
H
HO
OAc
MeO2C
For a preliminary evaluation of the metathesis strategy
for the preparation of bryostatin 11 (2), the C(1)–
C(16) fragment 12 was prepared via a modification of
our earlier route4 as shown in Scheme 1. Condensation
of the aldehyde 3 and phosphonate 4 using barium
hydroxide as base gave the a,b-unsaturated ketone 5,
which was treated with the pyridine–hydrogen fluoride
complex to give the diol 6. This cyclised regio- and stereo-
selectively when reacted with a catalytic amount of
potassium tert-butoxide in tetrahydrofuran to give the
4-methylenetetrahydropyran 7 as a single diastereoiso-
mer. A Dess–Martin oxidation followed by a Wittig con-
densation then furnished the alkene 8. This, on treatment,
with trimethyl orthoformate in methanol containing a
catalytic amount of pyridinium toluene p-sulfonate, gave
the acetal 9, in which the exo-cyclic allylic alcohol had
also been deprotected. Protection of both the primary
and secondary alcohols as trimethylsilylethoxymethyl
H
H
H
O
O
OH
1
O
17
OH
16
O
O
Me
Me
27
Me
X
HO
CO2Me
1 X = nPrCH=CH-CH=CH-CO2
2 X = H
In the three bryostatin syntheses completed to date, the
macrolide was assembled using a Julia reaction to form
the C(16)–C(17) double bond followed by further elabo-
*
Corresponding author. Tel.: +44 161 275 4614; fax: +44 161 275
4939; e-mail: e.j.thomas@manchester.ac.uk
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.01.097

2224
M. Ball et al. / Tetrahedron Letters 47 (2006) 2223–2227
Me Me
Me Me
TBSO
CHO
i, ii
(MeO)2P(O)
OTMS
+
TBSO
OR1
OR2
O
OBn
O
O
OTBDPS
O
OBn
O
O
OTBDPS
Me Me
OTES
Me Me
3
5 R1 = TMS, R2 = TES
6 R1 = R2 = H
4
iii, iv
Me Me
MeO
Me Me
H
H
O
H
O
MeO
OBn
OBn
Me Me
H
H
H
H
viii, ix
R1O
O
v - vii
SEMO
O
H
R
TBSO
O
O
OBn
O
O
OTBDPS
CO2H
Me Me
R1O
OR2
SEMO
7 R = CH2OH
8 R = CH= CH2
9 R1 = H, R2 = TBDPS
12
10 R1 = SEM, R2 = TBDPS
11 R1 = SEM, R2 = H
Scheme 1. Reagents and conditions: (i) Ba(OH)2, THF, H2O, rt; (ii), HFÆpy., rt, 12 min; (iii), tBuOK, THF, rt, 15 min (7, 48% from 4); (iv) (a) Dess–
t
i
Martin periodinane; (b) Ph3P+MeBrÀ, BuLi (60% from 7); (v) HC(OMe)3, MeOH, PPTS (58%); (vi) SEMCl, Pr2NEt, CH2Cl2, rt (79%); (vii)
nBu4NF, THF, rt (85%); (viii) Dess–Martin periodinane; (ix) NaClO2, NaH2PO4, 2-methylbut-2-ene, BuOH.
t
(SEM) ethers followed by selective cleavage of the tert-
butyldiphenylsilyl ether gave the alcohol 11, which was
oxidised to the acid 12 in two steps. This acid corre-
sponds to the required C(1)–C(16)-fragment of the
bryostatins 1 and 2.
protection giving the SEM-ether 15, ring opening of
the lactone gave the Weinreb amide 16 which was pro-
tected as its tert-butyldimethylsilyl ether 17 and reduced
to the aldehyde 18. Addition of allenylzinc bromide to
18 was expected5 to proceed with chelation control but
a mixture of epimers 19 and 20 was obtained. However,
these could be separated and, following inversion of the
unwanted epimer 20 via a Mitsunobu reaction and
saponification, the required alcohol 19 was obtained in
an overall yield of 70%.8
A preparation of the C(17)–C(27) fragment 23 based on
our earlier work,5 is outlined in Scheme 2. Asymmetric
dihydroxylation of methyl (3E)-pent-3-enoate 13 using
AD-mix-b gave the lactone 14.7 Following hydroxyl
O
O
OSEM
OR1
iii - v
i, ii
MeO2C
Me
O
R2
Me
RO
Me
16 R1 = H, R2 = NMeOMe
17 R1 = TBS, R2 = NMeOMe
18 R1 = TBS, R2 = H
13
14 R = H
15 R = SEM
vi
vii, viii
TBSO
Br
OSEM
OTBS
HO
OSEM
OTBS
HO
OSEM
OTBS
ix
+
Me
Me
20
Me
21
19
OTBDPS
x, xi
Me
Me
O
OTBS
OTBS
Me
Me
O
OR
OSEM
OTBS
Me
+
OAc
Me
Me
Me
OTBS
Me
25
22 R = SEM
23 R = H
24
OTBDPS
OTBDPS
Scheme 2. Reagents and conditions: (i) AD-mix-b, tBuOH, H2O (59%); (ii) SEMCl, iPr2NEt, DMAP, 3 days (95%); (iii) MeNHOMe, HCl, AlMe3;
(iv) TBSCl, imid. (66% from 15); (v) DIBAL-H, THF, À78 °C (87%); (vi) propargyl bromide, Zn powder, À15 to À78 °C (78%; 19:20 = 50:50); (vii)
Cl2CHCO2H, PPh3, DIAD, THF; (viii) NaOH, MeOH (70% of 19 based on 18); (ix) (a) TBSOTf, 2,6-lutidine (94%); (b) nBuLi, MeOCOCl, À78 °C
(92%); (c) (Bu3Sn)2, nBuLi, CuBrÆDMS, THF, À50 °C (88%); (d) DIBAL-H, DCM, À78 °C (83%); (e) TBDPSCl, imid. DCM (100%); (f) NBS,
DCM, rt (97%); (x) PdCl2Ædppe, Bu3SnOMe, tol., 25, 120 °C (22, 43%; 24, 38%); (xi) MgBr2, nBuSH, K2CO3 (64%).

M. Ball et al. / Tetrahedron Letters 47 (2006) 2223–2227
2225
OPMB
OPMB
The alcohol 19 was taken through to the vinyl bromide
CHO
iv
ii, iii
21 via protection, methoxycarbonylation, stereoselec-
tive addition of a tin cuprate, reduction of the ester, fur-
ther protection and substitution of the vinyl stannane
using N-bromosuccinimide. Palladium(0) catalysed
coupling of the vinylic bromide 21 with the enol acetate
25 then gave a mixture of the required non-conjugated
ketone 22 and the Heck product 24,9 but the ketone
22 could be isolated in a yield of 43% which was suffi-
cient to enable the metathesis to be evaluated. Selective
removal of the SEM group gave the alcohol 23, which
corresponds to the C(17)–C(27) fragment of bryostatin
11 (2).
OR
OTES
O
31
29 R = H
30 R = TES
X
27 X = OH
28 X = I
v, vi
i
Me Me
OR
O
OBn
O
O
OTBDPS
Me Me
32 R = TES 33 R = H
vii
Me Me
H
O
Esterification of the acid 12 using the alcohol 23 gave the
ester 26, but attempts to form the bryostatin macrocycle
using ring-closing metathesis using the Grubbs 2 cata-
lyst6 were unsuccessful with a complex mixture of pro-
ducts being obtained.
H
O
OBn
O
O
OTBDPS
Me Me
34
viii, ix
Me Me
Me Me
H
Me Me
H
O
MeO
H
O
MeO
H
O
MeO
OBn
OBn
OBn
x
H
H
H
H
H
SEMO
O
O
O
2,4,6-trichlorobenzoyl
chloride
OSEM
O
12
+
23
CO2H
DMAP, tol.
58%
OTBS
19
Me
Me
O
O
SEMO
OR
OTBDPS
37
35 R = H 36 R = SEM
Me
OTBS
Scheme 3. Reagents and conditions: (i) I2, PPh3, imid. (89%); (ii)
tBuLi, THF, À78 °C, 15 min; (iii) TESCl, imid., rt, 2 h (99% from 28);
(iv) (a) DDQ, DCM, pH 7 buffer (70%); (b) Dess–Martin periodinane;
(v) Ba(OH)2, THF, rt, 18 h (86% over two steps); (vi) HFÆpy, THF, rt,
15 min (100%); (vii) BuOK, THF, rt (90%); (viii) HC(OMe)3, PPTS,
MeOH, rt (86%); (ix) SEMCl, iPr2NEt, DMAP (80%); (x) (a)
26
OTBDPS
t
During these attempted metatheses, the double bond
attached to the C(1)–C(16) fragment was lost, but the
more hindered double bond attached to the (C17)–
C(27) fragment remained unchanged. These problems
in accessing a bryostatin directly by ring-closing metathe-
sis of the alkene 26 were disappointing but were not
totally unexpected. Ring-closing metatheses of terminal
alkenes with geminal allylic methyl groups are known
but would appear to be very sensitive to minor struc-
tural modifications.10 In the case of the metathesis pre-
cursor 26, it was not clear whether the geminal
dimethyl group at C(18), the presence of a ketone at
C(19), the presence of chelating groups in the vicinity
of the alkenes involved in the metathesis, for example,
the SEM group, or the flexible open-chain structure of
the C(17)–C(23) fragment, was the major factor respon-
sible for the difficulties in the ring-closing metathesis. It
was therefore decided to study the ring-closing metathe-
sis using simpler systems to attempt to establish the
major factors which might be involved.
TBAFÆTHF; (b) Dess–Martin periodinane; (c) NaClO2, NaH2PO4,
2-methylbut-2-ene, BuOH (94%).
t
TES deprotection, base-induced cyclisation gave tetra-
hydropyran 34, which was converted to acetal 35 by
treatment with trimethyl orthoformate in acidic metha-
nol. Following protection of the secondary alcohol as
its SEM-ether 36, desilylation and oxidation gave acid
37.
Simpler C(17)–C(27)-fragments with and without the
geminal dimethyl group at C(18) (bryostatin numbering)
were prepared as outlined in Schemes 4 and 5. Thus,
reduction of the epoxide 38, prepared by Sharpless
epoxidation of the corresponding (E)-alkene,7 gave diol
39, which was differentially protected to give the bis-silyl
ether 40. This was taken through to the bc-unsaturated
ketone 41 and acetalisation and desilylation gave a mix-
ture of the inseparable epimeric acetals 42 (Scheme 4).
The C(1)–C(16) fragment 37 lacking the exocyclic alk-
oxymethylene group was prepared as outlined in Scheme
3. The hydroxy-epoxide 27 prepared by Sharpless epoxi-
dation of the corresponding (E)-alkene7 was converted
into iodide 28, which gave allylic alcohol 29 on treat-
ment with tert-butyllithium. Following protection of
the secondary alcohol as its triethylsilyl (TES) derivative
30, selective removal of the p-methoxybenzyl ether and
oxidation gave aldehyde 31. This was condensed with
keto-phosphonate 4 to give enone 32. After selective
The second modified C(17)–C(27) fragment 49 was pre-
pared by copper(I) catalysed reaction of epoxide 4311
with allylmagnesium bromide followed by protection
to give the tert-butyldimethylsilyl ether 44 (Scheme 5).
Hydroboration/oxidation followed by a Swern oxida-
tion, a zinc-mediated reaction with 3,3-dimethylallyl
bromide and further oxidation gave the ketone 45. Desil-
ylation and cyclisation then gave enol ether 46,12 which
gave a mixture of the hydroxyacetals 47 on oxidation

2226
M. Ball et al. / Tetrahedron Letters 47 (2006) 2223–2227
OR2
obtained in which the less hindered terminal double
i, ii
O
bond attached to the tetrahydropyran seemed to have
been destroyed whilst the more hindered terminal dou-
ble bond with geminal allylic methyl groups was mainly
unchanged. A small amount of a dimeric material was
detected by MS.
OR1
OH
38
39 R1 = R2 = H
40 R1 = TIPS, R2 = TES
iii
iv
OMe
O
OTES
H
Me Me
Me Me
O
OTIPS
OH
H
O
H
O
MeO
MeO
OBn
OBn
H
H
H
H
Grubbs 2
25-30%
O
O
41
42
OSEM
OSEM
O
O
Scheme 4. Reagents and conditions: (i) Red-Al, THF, 0 °C (99%); (ii)
(a) TIPSCl, imid., rt, 2 h (99%); (b) TESCl, imid., rt, 2 h (99%); (iii) (a)
O3, DCM, then PPh3, À78 °C; (b) allylmagnesium bromide, THF (68%
over two steps); (c) Dess–Martin periodinane (80%); (iv) (a)
HC(OMe)3, PPTS, MeOH, rt (68%); (b) TBAF, THF, rt (85%).
OMe
O
OMe
O
H
H
O
O
50
51
Me Me
H
Me Me
H
H
O
H
O
MeO
MeO
OBn
OBn
H
H
O
O
TBSO
OBn
X
OBn
i
OSEM
OSEM
ii
OTBS
O
O
O
O
O
H
Me
Me
OMe
O
OMe
O
H
H
Me
Me
OBn
O
O
43
Me
Me
44
45
Me
Me
iii
53
52
MeO
MeO
OMe
O
H
H
iv
Me
Me
Me
HO
These results indicate that the ring-closing metathesis
can be carried out on systems which lack the geminal
dimethyl groups at C(18), for example, the successful
metathesis of 50, although it would appear that ring-
closing metatheses of substrates in which these methyl
groups are present, for example, the dienes 26 and 52,
are more difficult to carry out. Diene 52 has the unnat-
ural configuration at C(20), cf. bryostatin 1 (1), and the
C(17)–C(27) fragment of intermediate 26 lacks the con-
formational constraint of the six-membered cyclic hemi-
acetal present in bryostatins. It may be that these factors
are important in affecting the ring-closing metathesis
and so clearly more work is required to delineate fully
all the factors involved in this system. However, ring-
closing metathesis does provide access to bryostatin
analogues, which may well be biologically active, albeit
lacking the geminal dimethyl groups at C(20). Finally,
this work confirms that advanced intermediates for
bryostatin synthesis are available using the chemistry
described herein, so that different strategies for assembly
of the bryostatin system can now be evaluated.
OBn
OBn
OH
47
48
46
v
OMe
O
H
OMe
O
vi
Me
Me
Me
OBn
HO
MeO
49
Scheme 5. Reagents and conditions: (i) (a) allylmagnesium bromide,
CuI (83%); (b) TBSCl, imid. (93%); (ii) (a) BH3ÆTHF then H2O2/
NaOH (86%); (b) (COCl)2, DMSO, then Et3N; (c) Me2C@CHCH2Br,
Zn dust (67% over two steps); (d) (COCl)2, DMSO, then Et3N; (iii) (a)
TBAF (85% over two steps); (b) camphor sulfonic acid, benzene, MS
4A (99%); (iv) mCPBA, MeOH; (v) (a) Dess–Martin periodinane (86%
from 46); (b) LiBHEt3 (88%); (vi) (a) NaH, MeI, DMF (87%); (b) Na,
liq. NH3, THF (96%).
using m-chloroperoxybenzoic acid. Further oxidation
gave the corresponding ketone, which was reduced
stereoselectively using lithium triethylborohydride to
give the alcohol 48. O-Methylation and debenzylation
then gave the primary alcohol 49.
Acknowledgements
Using dicyclohexycarbodiimide (DCC), esterification of
the acid 37 using the alcohol 42 gave the ester 50 (88%),
as a mixture of diastereoisomers. This on treatment
with the Grubbs 2 catalyst underwent stereoselective
ring-closing metathesis to give the macrolide 51, a
macrocyclic analogue of the bryostatins, in which the
C(16)–C(17) double bond had the (E)-configuration. In
contrast, the ester 52 prepared from the acid 37 and
the alcohol 49, (DCC, 85%) did not undergo ring clos-
ing metathesis with the Grubbs 2, Hoveyda or Schrock
catalysts under a variety of conditions.13 As observed
for the ester 26, complex mixtures of products were
We thank the EPSRC for support (to B.B. and R.D.).
References and notes
1. Mutter, R.; Wills, M. Bioorg. Med. Chem. 2000, 8, 1841;
Hale, K. J.; Hummersone, M. G.; Manaviazar, S.;
Frigero, M. Nat. Prod. Reports 2002, 19, 413.
2. Masamune, S. Pure Appl. Chem. 1988, 60, 1587; Kage-
yama, M.; Tamura, T.; Nantz, M. H.; Roberts, J. C.;
Somfrai, P.; Whitenour, D. C.; Masamune, S. J. Am.
Chem. Soc. 1990, 112, 7407; Evans, D. A.; Carter, P. H.;

M. Ball et al. / Tetrahedron Letters 47 (2006) 2223–2227
2227
Carreira, E. M.; Prunet, J. A.; Charette, A. B.; Lautens,
M. Angew. Chem., Int. Ed. 1998, 37, 2354; Evans, D. A.;
Carter, P. H.; Carreira, E. M.; Prunet, J. A.; Charette, A.
B.; Lautens, M. J. Am. Chem. Soc. 1999, 121, 7540;
Ohmori, K.; Ogawa, Y.; Obitsu, T.; Ishikawa, Y.;
Nishiyama, S.; Yamamura, S. Angew. Chem., Int. Ed.
2000, 39, 2290.
8. Alternative procedures for the stereoselective conversion
of aldehyde 18 into the alcohol 19 using enantiomerically
enriched reagents were investigated. Full details will be
reported in a full paper.
9. An alternative sequence for the synthesis of the ketone 23
based on the stereoselective addition of an allylic Grignard
reagent to a conjugated acetylenic ketone has also been
developed (cf. Ref. 4). Full details will be reported
elsewhere.
10. Michalak, M.; Wicha, J. Synlett 2005, 2277; Kozmin, S.
A.; Iwama, T.; Huang, Y.; Rawal, V. H. J. Am. Chem.
Soc. 2002, 124, 4629.
11. Frick, J. A.; Klassen, J. B.; Bathe, A.; Abramson, J. M.;
Rapoport, H. Synthesis 1992, 7, 621.
3. Wender, P. A.; De Brabander, J.; Harran, P. G.; Jimenez,
J.-M.; Koehler, M. F. T.; Lippa, B.; Park, C.-M.;
Shiozaki, M. J. Am. Chem. Soc. 1998, 120, 4534; Wender,
P. A.; De Brabander, J.; Harran, P. G.; Hinkle, K. W.;
Lippa, B.; Pettit, G. R. Tetrahedron Lett. 1998, 39, 8625;
Wender, P. A.; Lippa, B. Tetrahedron Lett. 2000, 41,
1007.
4. Ball, M.; Baron, A.; Bradshaw, B.; Omori, H.; MacCor-
mick, S.; Thomas, E. J. Tetrahedron Lett. 2004, 45, 8737.
5. Almendros, P.; Rae, A.; Thomas, E. J. Tetrahedron Lett.
2000, 41, 9565; Gracia, J.; Thomas, E. J. J. Chem. Soc.,
Perkin Trans. 1 1998, 2853.
6. Nicolaou, K. C.; Bulger, P. G.; Sarlah, D. Angew. Chem.,
Int. Ed. 2005, 44, 4490.
7. The ee’s of the hydroxylactone 14 and the hydroxyepox-
ides 27 and 38 were shown to be P90% and their absolute
configurations as shown using the Mosher’s method. Full
details will be reported in a full paper.
12. Attempts to convert the hydroxyketone formed on desily-
lation of ketone 45 into the corresponding methoxy acetal
were unsuccessful; the enol ether 46 was the only product
which could be isolated.
13. The linear ketone 34 was also desilylated, oxidised to the
corresponding carboxylic acid, and esterified using
the alcohols 42 and 49. The ester of alcohol 42 lacking
the geminal dimethyl groups at C(18) underwent ring-
closing metathesis, whereas the ester derived from alcohol
49 did not. These results are consistent with those reported
here and will be described elsewhere.