Synthesis of 16-desmethylepothilone B: Improved methodology for the rapid, highly selective and convergent construction of epothilone B and analogues

Article abstract of DOI:10.1039/a809954e

During a synthesis of 16-desmethylepothilone B new methods for the convergent and highly stereoselective synthesis of epothilone B and analogues were developed.

Full text of DOI:10.1039/a809954e

Synthesis of 16-desmethylepothilone B: improved methodology for the rapid,
highly selective and convergent construction of epothilone B and analogues
K. C. Nicolaou,* David Hepworth, M. Ray V. Finlay, N. Paul King, Barbara Werschkun and Antony Bigot
Department of Chemistry and The Skaggs Institute for Chemical Biology, The Scripps Research Institute, 10550
North Torrey Pines Road, La Jolla, California 92037, USA and Department of Chemistry and Biochemistry,
University of California San Diego, 9500 Gilman Drive, La Jolla, California 92093, USA. E-mail: kcn@scripps.edu
Received (in Corvallis, OR, USA) 21st December 1998, Accepted 4th February 1999
S
N
S
N
i
During a synthesis of 16-desmethylepothilone B new meth-
ods for the convergent and highly stereoselective synthesis of
epothilone B and analogues were developed.
O
2
O
3
ii
Due to their great potential as anticancer drugs with a paclitaxel-
like mechanism of action, the epothilones have recently been at
the focus of widespread scientific investigations.1 In particular,
significant attention has concerned chemical synthesis of the
epothilone natural products, which we and others have
successfully achieved.1–3 In connection with on-going studies
in the epothilone area we were interested in preparing analogues
of epothilone B containing less conformationally restrained
heteroaromatic side-chains, in particular the 16-desmethyl
analogue 1 (Fig 1). Since this endeavour required a complete
resynthesis of the epothilone macrocycle we took this opportu-
nity to develop new methods for the swift assembly of
epothilone B analogues with improved selectivities and
yields.
Our aims were to combine the benefits of our highly modular
and convergent approach to epothilone A and analogues via
olefin metathesis,1 which allowed for extremely rapid macro-
cycle construction, with the improved stereoselectivities in-
herent in our macrolactonisation strategy to epothilone B.1 We
envisioned that our approach to epothilone B analogues could
be made more convergent by employing a fully functionalised
C7–C12 fragment in a highly selective olefination reaction to
construct the C12–C13 bond. This approach would also allow
access to C26 modified analogues as previously described.1,3a
Furthermore, we wished to retain a hydroxy group at the C26
position until a late stage in the synthesis and thus profit from a
chemo-, regio- and stereo-selective Sharpless epoxidation.1,3a
In adopting this route we were aware that a method for
subsequent deoxygenation at the C26 position (adjacent to the
epoxide) would have to be developed. Finally, we hoped to
improve the pivotal aldol coupling step which provides the
stereocentres at C6 and C7 of the macrocycle. Previously we
have found this reaction to be somewhat capricious.
O
S
N
S
N
iv, v
OTBDMS
6a
OR
4 R = H
iii
5 R = TBDMS
Scheme 1 Reagents and conditions: Ph3PNCHCHO, CH2Cl2, reflux, 12 h,
82%; ii, (+)-Ipc2B(allyl), CH2Cl2, Et2O, pentane, 2100 °C, 2 h, 100%; iii,
TBDMSCl, imidazole, DMF, 0 ? 25 °C, 2 h, 99%; iv, OsO4 (cat.), NMO,
THF, ButOH, H2O, 25 °C, 12 h, 68%; v, NaIO4, MeOH, H2O, 0 ? 25 °C,
1 h, 89% (Ipc = isopinocampheyl).
The reaction sequence includes an efficient cuprate coupling,4
and a chloroformate quench of an unstabilised ylide5 as the key
steps (Scheme 2). Control of temperature during the latter
reaction is critical to obtain phosphorane of good quality.§
The all-important Wittig coupling of 14 and 6a proceeded
smoothly with excellent yield and selectivity (E:Z > 30:1 in all
cases) to provide the coupled product 15a, which was processed
to 21a (Scheme 3). We have since extended this approach to the
natural epothilone B series 23b, the 26-hydroxyepothilone B
series 21b and the highly flexible vinyl iodide series 21c, which
we have shown to be invaluable for producing heterocycle-
modified epothilones.1,3b,c This route also enabled us to
approach the highly epimerisation-prone aldehydes 21a-c and
23b through stereochemically ‘safe’ alcohol oxidation proce-
dures.¶
We reasoned that the moderate and variable stereoselectiv-
ities obtained in the coupling of 24 with chiral aldehydes such
as 21a–c and 23b may be due, at least in part, to the reversible
nature of aldol reactions using ketone-derived lithium enolates.7
In an attempt to alleviate such problems we have performed the
reaction with an excess of enolate,∑ very short reaction time and
a rapid low temperature quench (AcOH). Gratifyingly, the
The thiazole aldehyde fragment 6a† was prepared in an
analogous fashion to the related epothilone B fragment with
minor changes in experimental conditions (Scheme 1). An
asymmetric allylboration was the key step used to introduce the
stereogenic centre at C15.‡ The regioselectivity of the dihy-
droxylation was lower than in the epothilone B series which
slightly impaired the overall yield for this fragment. The fully
functionalised phosphorane 14 was prepared in short order from
commercially available 7 without the need for chiral auxiliaries.
X
iii
RO
7 R = H, X = Br
TBDMSO
i
10
8 R = TBDMS, X = Br
9 R = TBDMS, X = I
ii
iv
CO2Me
vii
PPh3
X
TBDMSO
TBDMSO
11 X = OH
12 X = I
13 X = P+Ph3 I–
14
v
Deoxygenation
vi
26
O
Wittig olefination
Sharpless epoxidation
S
12
Scheme 2 Reagents and conditions: TBDMSCl, imidazole, DMF, 0 ? 25
°C, h; ii, NaI, acetone, reflux, 12 h, 99% (2 steps); iii,
13
1
HO
N
16
7
6
CH2NCHCH2CH2MgBr, Li2CuCl4 (cat.), THF, 0 °C, 1 h, 96%; iv, (a) O3,
CH2Cl2, 278 °C, then PPh3; (b) NaBH4, EtOH, 0 °C, 30 min., 91% (2
steps); v, I2, PPh3, CH3CN, Et2O, 0 ? 25 °C, 1 h, 100%; vi, PPh3, neat, 100
°C, 2 h; vii, KHMDS, THF, 0 °C, 30 min, then MeOCOCl, 278 °C, 3 h (ca.
100%, 2 steps, unpurified).
1
O
Yamaguchi
macrolactonisation
Aldol coupling
O
O
OH
Fig. 1 Numbering and bond disconnections for 1.
Chem. Commun., 1999, 519–520
519

OR2
OTr
Y
O
R1
R1
i
iii
R2
X
R1O
Th
R2
TBDMSO
Th
OR
OTBDMS
O
OTBDMS
X
6a R1 = H, R2 = Th
b R1 = Me, R2 = Th
c R1 = Me, R2 = I
15a,b,c X = TBDMSO, H; Y = CO2Me
16a,b,c X = TBDMSO, H; Y = CH2OH
17a,b,c X = TBDMSO, H; Y = CH2OTr
18b X = TBDMSO, H; Y = CH2Cl
19b X = TBDMSO, H; Y = CH3
20a,b,c X = OH, H; Y = CH2OTr
21a,b,c X = O; Y = CH2OTr
O
OTBDMS
R1O
O
O
ii
28 R1 = TBDMS; R2 = Tr
29 R1 = R2 = H
iii
25a R = TBDMS; X = CH2OTBDMS
26 R = TBDMS; X= CH2OH
27 R = H; X = CO2H
iv
v
i
iv
ii
vi
S
v
X
Th
=
viii
vii
N
O
O
R1
R1
22b X = OH, H; Y = CH3
viii
vii
HO
Th
23b X = O; Y = CH3
HO
Th
O
O
O
Scheme 3 Reagents and conditions: i, 14 (1.3 equiv. based on 12), C6H6,
reflux, 18 h, 84–93%; ii, DIBAL-H, THF, 278 °C, 3 h, 71–95%; iii, TrCl,
DMAP, DMF, 70 °C, 18 h, 82–95%; iv, CCl4, PPh3, reflux, 18 h, 80%; v,
LiEt3BH, THF, 278 °C, 1 h, 92%; vi, HF•Py.Py, THF, 25 °C, 4 h, 66–73%;
vii, CSA, MeOH, 25 °C, 1 h, 95%; viii, SO3•Py, DMSO, Et3N, THF, 0 °C,
1 h.
O
O
O
OH
OH
30 R1 = H; X = OH
31 R1 = H; X = I
32 R1 = Me; X=I
1 R1 = H
33 R1 = Me
vi
Scheme 5 Reagents and conditions: i, HF•Py.Py, THF, 25 °C, 3 h, 87%
(after 1 recycle); ii, (a) (COCl)2, DMSO, CH2Cl2, 278 °C, 30 min, then
Et3N, 278 ? 0 °C, 30 min; (b) NaClO2, Me2CNCHMe, NaH2PO4, ButOH–
H2O, 25 °C, 2 h; (c) TBAF, THF, 25 °C, 12 h, ca. 100% (3 steps); iii,
2,4,6-Cl3C6H2COCl, Et3N, THF, 0 °C, 1 h, then add to DMAP in toluene,
75 °C, 1 h, 73%; iv, HF•Py, THF, 0 ? 25 °C, 24 h, 78%; v, (+)-diethyl l-
tartrate, Ti(PriO)4, ButOOH, CH2Cl2, 4 Å MS, 230 °C, 2 h, 80%; vi, (a)
TsCl, Et3N, DMAP, CH2Cl2, 0 ? 25 °C, 1 h; (b) NaI, acetone, 25 °C, 15 h,
91% (2 steps); vii, NaBH3CN, HMPA, 45 °C, 40 h, 67–70%.
application of these conditions to aldehydes 21a–c and 23b lead
to significant amelioration of stereoselectivity which was
naturally accompanied by an improved yield of the desired aldol
stereoisomer in each case (Scheme 4, Table 1). Intermediates
25a–d were obtained in good overall yields after subsequent
TBDMS-protection. Under these closely defined conditions the
aldol reaction is highly reproducible and applicable to multi-
gram quantities. These results, in terms of diastereoselectivities
and yields, are at least as good as those obtained by Schinzer and
co-workers in their epothilone B syntheses.2b Since our
approach features a rather convenient protecting group strategy,
we believe this development renders our modified route the
most effective solution to date for epothilone B.
The TBDMS-protected aldol product 25a was processed
through to 26-hydroxy-16-desmethylepothilone B 30 in the
same fashion as our published route to 26-hydroxyepothilone B
(Scheme 5).3a As projected, the Sharpless epoxidation pro-
ceeded with complete stereo- and regio-control at the C12–C13
position. It should be noted that for this analogue series,
competitive side-chain epoxidation would likely have been
problematic upon use of conventional oxidants. In order to
complete the synthesis, removal of the C26 hydroxy group was
required. This was achieved by initial conversion to iodide 31
followed by reductive deiodination with NaBH3CN7 to provide
X
our desired analogue 1 in good overall yield. Furthermore, this
deoxygenation sequence has also been demonstrated for the
production of epothilone B (33) itself from our previously
reported iodide 32. Thus, all of the synthetic methodology
described herein is applicable to epothilone B, which renders
our approach now highly selective at every step, and sig-
nificantly increases the speed of access to this and related
structures. Work is currently underway to assess the biological
activity of 1 and related analogues of epothilone B.
We thank D.-H. Huang and Gary Siuzdak for NMR and MS
studies, respectively. This research was financially supported
by the National Institutes of Health USA and The Skaggs
Institute for Chemical Biology, from the George Hewitt
Foundation (to N. P. K.), the Deutsche Forschungsgemeinschaft
(to B. W.), and grants from CaPCURE and Novartis.
Notes and references
† All new compounds exhibited satisfactory spectral and exact mass data.
‡ The ee was shown to be !97% by chiral HPLC (Chiralcel OD-H column)
by comparison with the racemate.
R1
O
OTBDMS
OTBDMS
24
TBDMSO
R2
i, ii
§ Due to high polarity and complex NMR spectra 14 was not purified as
characterized. We observed full mass return for 12 ? 14 and used the
material crude using quantities assuming purity.
¶ The aldehydes were freshly prepared and not purified prior to use.
∑The unreacted ketone 24 is easily recovered by chromatography.
1 For a review of the literature to Feb. 1998, see: K. C. Nicolaou, F.
Roschangar and D. Vourloumis, Angew. Chem., Int. Ed., 1998, 37,
2014.
X
OTBDMS
+
O
OTBDMS
OTBDMS
R1
R2
O
OTBDMS
21a R1 = H, R2 = Th, X = OTr
25a
R
R
R
1 = H, R2 = Th, X = OTr
1 = Me, R2 = Th, X = OTr
1 = Me, R2 = I, X = OTr
R
1 = Me, R2 = Th, X = H
b
c
R
R
R
1 = Me, R2 = Th, X = OTr
1 = Me, R2 = I, X = OTr
1 = Me, R2 = Th, X = H
b
c
d
23b
2 For recent syntheses see: (a) S. A. May and P. Grieco, Chem. Commun.,
1998, 1597; (b) D. Schinzer, A. Bauer and J. Scheiber, Synlett, 1998, 861;
(c) A. Balog, C. Harris, K. Savin, X.-G. Zhang, T. C. Chou and S. J.
Danishefsky, Angew. Chem., Int. Ed., 1998, 37, 2675; (d) J. Mulzer, A.
Mantoulidis and E. Öhler, Tetrahedron Lett., 1998, 39, 8633; (e) S. C.
Sinha, C. F. Barbas III and R. A. Lerner, Proc. Natl. Acad. Sci. U.S.A.,
1998, 95, 14603.
Scheme 4 Reagents and conditions: i, LDA (2.4 equiv.), 24 (2.3 equiv.),
THF, 278 ? 240 °C, 1 h, then add 21 or 23, THF 278 °C, 2 min, then
AcOH, 278 ? 0 °C; ii, TBDMSOTf, 2,6-lutidine, THF, 278 ? 0 °C,
1 h.
Table 1 Yields and stereoselectivities for aldol reactions
3 (a) K. C. Nicolaou, M. R. V. Finlay, S. Ninkovic and F. Sarabia,
Tetrahedron, 1998, 54, 7127; (b) K. C. Nicolaou, M. R. V. Finlay, S.
Ninkovic, N. P. King, Y. He, T. Li, F. Sarabia and D. Vourloumis, Chem.
Biol., 1998, 5, 365; (c) K. C. Nicolaou, N. P. King, M. R. V. Finlay, Y.
He, F. Roschangar, D. Vourloumis, H. Vallberg, F. Sarabia, S. Ninkovic
and D. Hepworth, Bioorg. Med. Chem., 1998, in the press.
4 M. Tamura and J. Kochi, Synthesis, 1971, 303.
5 Y. Morimoto and H. Shirahama, Tetrahedron, 1996, 52, 10631.
6 C. H. Heathcock, in Comprehensive Organic Synthesis, ed. B. M. Trost
and I Fleming, Pergamon, Oxford, 1991, vol. 2, pp. 181–235.
7 R. O. Hutchins, D. Kandasamy, C. A. Maryanoff, D. Masilamani and
B. E. Maryanoff, J. Org. Chem., 1977, 42, 82.
Aldol product
Aldehyde Selectivitya (%)b
25 (%) (2 steps)c
21a
21b
21c
23b
!10:1
!15:1
!10:1
!10:1
76
77
79
71
72
73
75
67
a
Conservative estimates based upon mass recovery of the major isomer
b
relative to all other polar impurities. Often contaminated with small
c
amounts ( = 5%) of starting aldehyde. Yields are compensated. Single
stereoisomer free from traces of aldehyde and other impurities.
Communication 8/09954E
520
Chem. Commun., 1999, 519–520