Synthesis of the bicyclo[4.3.1]decenone core of CP-225,917 and CP-263,114

Article abstract of DOI:10.1055/s-1998-1699

The bicyclo[4.3.1]decenone core of the ras-farnesyl protein transferee inhibitors CP-225,917 and CP-263,114 is prepared in 6 steps from cyclohexanone. The key step is an intramolecular Mukaiyama aldol reaction.

Full text of DOI:10.1055/s-1998-1699

552
LETTERS
SYNLETT
Synthesis of the Bicyclo[4.3.1]decenone Core of CP-225,917 and CP-263,114
a
a
b
Alan Armstrong,* Trevor J. Critchley and Andrew A. Mortlock
a
Department of Chemistry, University of Nottingham, Nottingham NG7 2RD, UK
Fax +44 115 9513564; E-mail Alan.Armstrong@nottingham.ac.uk
b
Zeneca Pharmaceuticals, Mereside, Alderley Park, Macclesfield, Cheshire, SK10 4TG, UK
Received 22 February 1998
Abstract: The bicyclo[4.3.1]decenone core of the ras-farnesyl protein
transferase inhibitors CP-225,917 and CP-263,114 is prepared in 6 steps
from cyclohexanone. The key step is an intramolecular Mukaiyama
aldol reaction.
The closely related natural products CP-225,917 and CP-263,114 were
1
recently isolated by workers at Pfizer from an unidentified fungus. The
2
compounds are inhibitors of ras-farnesyl protein transferase, and are
3
therefore of interest in the search for novel anticancer agents.
Structurally, they are related to the nonadride family of natural
4
products in that they contain a nine-membered ring fused to a maleic
anhydride moiety. Interestingly, they also display an anti-Bredt olefin.
The combination of interesting biological activity and intriguing
structure has made these compounds attractive targets for total synthesis
Scheme 1
cyclohexanone with diethylketomalonate gave 1 (79%). Conversion of 1
5
6
7
and the groups of Davies, Nicolaou and Clive have recently
described syntheses of models of the bicyclic core. These recent reports
prompt us to disclose our own preliminary results which have led to the
development of a short (6 step) synthesis of the bicyclo[4.3.1]decenone
core from cyclohexanone.
into 2 was achieved using a two step procedure: acetylation (Ac O, cat.
2
8
TMSOTf,
93% yield) followed by treatment with 1,4-
diazabicyclo[2.2.2]octane (Dabco ) to effect acetate elimination and
alkene isomerisation (75% yield). We were pleased to find that 2
™
9
underwent clean and regioselective alkylation with commercially
available 3-bromopropanal dimethylacetal, thus creating the C13-C14
bond. Conversion of 3 into the trimethylsilyl enol ether 4 then set the
10
stage for the key ring closure step. Pleasingly, treatment of 4 with
11
TiCl in CH Cl
2
provided the desired bicyclic product 5 (52%) as a
4
2
mixture of diastereomers which could be separated by flash
12
chromatography to give 5a (12%) and 5b (40%). We have briefly
examined alternative Lewis acids: boron trifluoride etherate gave
similar overall yield (20% 5a and 28% 5b), while none of the cyclised
product was observed when trimethylsilyl triflate was employed.
A first indication that the desired cyclisation had occurred was the
upfield shift observed for H16 in the NMR spectrum (ca. 6.4 ppm in 5
compared to 6.95 ppm in 3), suggesting that the enone system was not
13
now able to adopt planarity. That the desired cyclic structure had
Our strategy (Scheme 1) has envisaged introduction of the sensitive
anhydride unit late in the synthesis. The next key feature of our
approach relies on the fact that, since it is present in the natural product,
cyclisation of a pro-S carboxylic acid at C14 onto the C26 carbonyl is
expected to be facile. This is presumably due to relief of ring strain in
going from an sp centre at C26 to an sp centre. It is therefore hoped
that group differentiation of a symmetrical 1,1-diester at C14 can be
used to create the stereocentre at this position: cyclisation of the pro-S
carboxylate would leave the pro-R acid derivative free for subsequent
homologation. This strategy would also enable the anion-stabilising
properties of the C14-diester to be utilised in construction of the C13-
C14 bond. Further disconnection at C10-C11, using one of several
conceivable methods for bond formation α- to the C26 carbonyl, would
then lead to a simple cyclohexenone derivative and a suitable bis-
electrophile. Here we report the realisation of the first major objective in
this plan, the synthesis of the carbobicyclic core by bond formation at
C13-C14 and C10-C11.
indeed been obtained was proven by HMBC and HMQC experiments
on 5b: in particular, 2-bond correlations between C10 and H11, and
between C11 and H10, proved that cyclisation had occurred. The major
diastereomer 5b is tentatively assigned the relative configuration shown
based on NOE studies in conjunction with molecular mechanics
2
3
14
calculations.
In summary, we have developed a short synthesis (6 steps from
cyclohexanone) of the bicyclic core of the natural products CP-225,917
and CP-263,114 utilising a Mukaiyama aldol reaction as the key step.
We are currently investigating the use of alternative bis-electrophile
equivalents and the introduction of the C14 stereocentre via
lactonisation onto the C26 carbonyl, as well as methods for the
formation of the anhydride unit.
Acknowledgements. We thank the EPSRC and Zeneca
Pharmaceuticals (CASE award to TJC) for their support of this work.
We also thank Dr. J.A. Ballantine and the EPSRC Mass Spectrometry
Service for accurate mass measurements. We wish to acknowledge the
In order to establish methodology for the synthesis of the carbobicyclic
core, our initial synthetic studies (Scheme 2) have omitted the alkyl
1
2
15
sidechains (i.e. R = R = H). Thus, reaction of the lithium enolate of
use of the EPSRC’s Chemical Database Service at Daresbury.

May 1998
SYNLETT
553
CH(CO Et) ), 37.5 (t, CH ), 25.8 (t, CH ), 22.2 (t, CH ), 13.7 (q,
2
2
2
2
2
+
2 x Me); HRMS (EI) Found M , 254.1162. C
254.1154.
H O requires
13 18 5
(10) A similar Mukaiyama aldol reaction has been used to make the
corresponding saturated ring system, bicyclo[4.3.1]decanone:
Kellner, D.; Loewenthal, H.J.E. Tetrahedron Lett. 1983, 24, 3397-
3400.
(11) Mukaiyama, T. Org. React. 1982, 28, 203-331.
(12) TiCl (240 µl of a 1M solution in CH Cl , 0.24 mmol) was added
4
2
2
rapidly to a cooled (-78 °C) solution of 4 (100mg, 0.23 mmol) in
CH Cl (2ml) under N . The mixture was stirred at -78°C for 2
2
2
2
hours then allowed to warm slowly to room temperature over 13
hours. It was then cooled to 0 °C, quenched with saturated
aqueous NaHCO (10 ml) and extracted with ether (3 x 10 ml).
3
The combined organic extracts were washed with H O (10 ml),
2
dried (MgSO ) and concentrated in vacuo. Chromatography on
4
silica (20% EtOAc-petrol) yielded 5a (9 mg, 12%) and 5b (30 mg,
40%) as oils.
Data for 5a (minor): R 0.28 (25% EtOAc-petrol); ν
(film) /
max
f
-1
cm 1736, 1708, 1451, 1245, 1088; δ (400MHz, CDCl ) 6.46
H
3
(1H, dd, J 4.3, 7.8, HC=), 4.31-4.20 (4H, m, 2CH O), 3.29 (3H, s,
2
MeO), 3.28-3.15 (2H, m, CHCO and CHOMe), 2.36-2.21 (2H, m,
CH ), 2.09-1.56 (6H, m, 3 x CH ), 1.28 (6H, t, J 7.1, 2 x Me); δ
C
2
2
(100 MHz, CDCl ) 204.1, 169.3, 169.2, 137.6, 135.4, 82.2, 61.8,
3
61.8, 59.1, 56.6, 50.6, 30.7, 27.7, 22.3, 19.4, 14.0, 13.9; m/z (EI)
+
+
+
324 (M , 0.14%), 279 (M -EtO, 1), 251 (M -CO Et, 5); Found
2
+
References and Notes
M , 324.1554. C
H O requires 324.1573.
17 24 6
Data for 5b (major): R 0.22 (25% EtOAc-petrol); ν
(film) /
max
(1) Dabrah, T.T.; Kaneko, T.; Massefski Jr., T.; Whipple, E.B. J. Am.
Chem. Soc. 1997, 119, 1594-1598.
f
-1
cm 1731, 1715, 1449, 1243, 1201; δ (500MHz, CDCl ) 6.34
H
3
(1H, dd, J 4.2 and 7.7, HC=), 4.28-4.19 (4H, m, 2 x CH O), 3.30
2
(2) Dabrah, T.T.; Harwood Jr., H.J.; Huang, L.H.; Jankovich, N.D.;
Kaneko, T.; Li, J-C.; Lindsey, S.; Moshier, P.M.; Subashi, T.A.;
Therrien, M.; Watts, P.C. J. Antibiot. 1997, 50, 1-7.
(3H, s, CH O), 3.15 (1H, m, CHOMe), 2.96 (1H, m - resolution
3
enhanced to ddd, J 2.6, 5.2, 8.3, CHCO), 2.33-1.90 (6H, m,
3CH ), 1.85 (1H, m, CH), 1.36-1.21 (7H, m, 2Me + CH ); δ (70
2
ax
C
(3) For a review of ras-farnesyl protein transferase inhibition, see:
Buss, J.E.; Morsters, J.C. Chem. Biol. 1995, 2, 787-791.
MHz, CDCl ) 202.8 (s, C=O), 169.6 (s, CO R), 169.2 (s, CO R),
3
2
2
138.2 (s, R C=), 133.4 (d, HC=), 81.0 (d, CHOMe), 61.8 (t,
2
(4) Barton, D.H.R.; Sutherland, J.K.; J. Chem. Soc. 1965, 1769-1772,
CH O), 57.3 (s, C(CO R) ), 56.3 (q, CH O), 51.8 (d, CHCO),
2
2
2
3
and subsequent papers.
27.4 (t, CH ), 24.9 (t, CH ), 24.5 (t, CH ), 21.7 (t, CH ), 14.0 (q,
2
2
2
2
+
+
+
(5) Davies, H.M.L.; Calvo, R.; Ahmed, G. Tetrahedron Lett. 1997,
38, 1737-1740.
Me); m/z (FAB) 347 (MNa , 36%), 325 (MH , 18), 293 (M -
OMe, 10) 279 (M -OEt, 15); Found MH , 325.1640. C
requires 325.1651.
+
+
H O
17 25 6
(6) Nicolaou, K.C.; Härter, M.W.; Boulton, L.; Jandeleit, B. Angew.
Chem., Int. Ed. Engl. 1997, 36, 1194-1196. Nicolaou, K.C.;
Postema, M.H.D.; Miller, N.D.; Yang, G. Angew. Chem., Int. Ed.
Engl. 1997, 36, 2821-2823.
(13) Also in accord with this is the carbonyl stretching frequency in the
-1
IR spectrum: the observed values of 1708 cm for 5a and 1715
-1
-1
cm for 5b are similar to that reported (1710 cm ) for
bicyclo[4.3.1]dec-1-en-10-one itself: Cordiner, B.G.; Vegar,
M.R.; Wells, R. Tetrahedron Lett. 1970, 2285-2286.
(7) Sgarbi, P.W.M.; Clive, D.L.J. Chem. Commun. 1997, 2157-2158.
(8) Procopiou, P.A.; Baugh, S.P.D.; Flack, S.S.; Inglis, G.G.A. Chem.
Commun. 1996, 2625-2626.
(14) Compound 5b displayed an NOE to H11 on irradiation of H10;
this interaction was not evident in the NOESY spectrum of 5a.
MM2* calculations in MacroModel 5.5 (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, 440-467) suggest that an NOE between H10 and H11 is
unlikely in the preferred conformations of the compound with
stereochemistry 5a, whereas the H10-H11 distance is relatively
short in 5b.
™
(9) Far better yields were obtained with Dabco than with other
bases (e.g. DBU); we believe that the alkene isomerisation step
proceeds via a conjugate addition - elimination mechanism.
Further details will be reported in due course. Data for 2:
colourless oil, R 0.38 (20% EtOAc-petrol); ν
(film) 1732,
max
f
1681, 1640 (weak); δ (250 MHz, CDCl ) 7.02 (1H, t, J 4.1,
H
3
HC=), 4.69 (1H, d, J 1.0, HC(CO Et) ), 4.24-4.13 (4H, qd, J 7.1,
2
2
1.6, 2CH O), 2.51-2.43 (4H, m), 2.08-2.00 (2H, m), 1.25 (6H, t, J
2
7.1, 2 x Me); δ (100 MHz, CDCl ) 196.4 (s, C=O), 167.7 (s, 2 x
(15) Fletcher, D.A.; McMeeking, R.F.; Parkin, D. J. Chem. Inf.
Comput. Sci. 1996, 36, 746-749.
C
3
CO R), 148.4 (d, HC=), 132.6 (s, C=), 61.4 (t, 2 x CH O), 50.4 (d,
2
2