An attempted cascade radical-mediated cyclisation approach to ring-D aromatic steroids

Article abstract of DOI:10.1055/s-2006-951533

A synthetic approach to ring-D aromatic steroids, viz. 5, based on a cascade of radical-mediated cyclisations from the ortho-aryl-substituted iododienynones 12 and 13, instead led to the macrocyclic ketone 16 or to the novel bridged tricycle 17, respectively, as the major products. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-2006-951533

LETTER
3073
An Attempted Cascade Radical-Mediated Cyclisation Approach to Ring-D
Aromatic Steroids
Radical-Mediated
e
Cyclisation
r
A
pproa
a
chto Ring-
l
DArom
d
atic Steroids Pattenden,* Alexander J. Blake, L. Krishnakanth Reddy, Davey A. Stoker
School of Chemistry, The University of Nottingham, Nottingham, England, NG7 2RD, UK
Fax +44(115)9513535; E-mail: gp@nottingham.ac.uk
Received 18 May 2006
This Letter is dedicated to Richard Heck whose pioneering studies paved the way for the development of a wide range of Pd(0)-mediated
coupling reactions in synthesis, especially applications of ‘cascades’ of Heck reactions in the elaboration of polycyclic constructs inclu-
ding steroids.
O
H
O
Abstract: A synthetic approach to ring-D aromatic steroids, viz. 5,
based on a cascade of radical-mediated cyclisations from the ortho-
aryl-substituted iododienynones 12 and 13, instead led to the mac-
rocyclic ketone 16 or to the novel bridged tricycle 17, respectively,
as the major products.
Bu3SnH
AIBN
I
H
H
MeO
MeO
Key words: radical reactions, steroids, polycycles, cascade reac-
2
3
tions, nicandrenones
Equation 1 Cascade radical-mediated cyclisation of the iodo-
dienynone 2 to the oestrone 3
Ring-D aromatic steroids are found only rarely in nature,
but some of their number show interesting and unusual bi-
ological properties. One such family is the nicandrenones
(or ‘NICs’), e.g. 1 (Figure 1), isolated from the Peruvian
‘shoofly’ plant Nicandra physaloides, which have insect
antifeedant and repellent properties.1
Apart from studies by Corey et al.,2 limited attention has
been given to the total synthesis of these compounds, but
some detailed studies have been directed toward under-
standing the origin of ring-D aromatic steroids in vivo.3
additional methylene carbon in its top-side chain com-
pared to 2. By analogy with the conversion 2 → 3, we
anticipated that 4 would undergo a similar cascade of
radical cyclisations leading to the tetracycle 5, a possible
progenitor to the nicandrenones (Equation 2).
R
R
I
H
?
H
O
O
O
H
O
O
O
4
5
H
R = H or OMe
OH
OH
Equation 2 Proposed radical-mediated cascade cyclisation towards
the ring-D aromatic steroid 5
1
Figure 1 Nicandrenone (NIC 1) 1, insecticidal ring-D aromatic ste-
roid from Nicandra physaloides
We synthesised the E- and Z-isomers 12 and 13, respec-
tively, of the iododienynones 4, in a straightforward man-
ner using identical synthetic sequences, starting from
easily available starting materials (Scheme 1).5,6 Julia ole-
fination reactions7 between the benzothiazole sulfone 7
and the substituted aldehyde 6, in the presence of
NaHMDS in THF, first led to 2:3 mixtures of Z- and E-
isomers about the newly introduced trisubstituted double
bonds of the corresponding alkenes 8a. Following depro-
tection to the corresponding alcohols 8b, the Z- and E-iso-
mers were separated by routine chromatography. The pure
E- and Z-isomers of 8b were then converted separately
into the E- and Z-isomeric iododienynones 12 and 13,
respectively, following chlorination to 9a, reduction and
oxidation to 9b, addition of acetylene, leading to 10,
oxidation to 11 and, finally, chloride–iodide exchange
(Scheme 1).
In recent years we have examined a range of complemen-
tary radical-mediated cascade cyclisations towards the
synthesis of steroids and polycyclic terpenoids.4 Thus, in
one such study we showed that treatment of the iododi-
enynone 2 with Bu3SnH–AIBN led to the oestrone 3, via
a cascade of 13-endo-dig, 5-exo-trig and 6-exo/endo radi-
cal cyclisations, in 40% yield (Equation 1).5
As part of a proof of principle investigation, we have now
examined the corresponding radical chemistry from the
structure 4, related to 2, as an approach to ring-D aromatic
steroids, similar to 1. The substrate 4 contains a trisub-
stituted, rather than a disubstituted, double bond and one
SYNLETT 2006, No. 18, pp 3073–3076
0
3
.1
1
.2
0
0
6
Advanced online publication: 25.10.2006
DOI: 10.1055/s-2006-951533; Art ID: S11706ST
© Georg Thieme Verlag Stuttgart · New York

3074
G. Pattenden et al.
LETTER
O
O
O
OEt
OEt
R'
i
iii
7
O
OR'
Cl
R
R
R
6
8
9
a R' = TBDPS
b R' = H
a R' = OEt
b R' = H
ii
iv, v
O
OH
vi
vii, viii
Cl
I
R
R
10
11
O
O
I
I
R
R
E
Z
12
13
a R = OMe
b R = H
Scheme 1 Reagents and conditions: (i) benzothiazole-SO2CHMe(CH2)3OTBDPS (7),6 NaHMDS, THF, –78 °C, 38%; (ii) TBAF, THF, 0 °C
→ 25 °C, 77%; (iii) NCS, K2CO3, PPh3, CH2Cl2, 0 °C, 97%; (iv) DIBAL-H, CH2Cl2, –78 °C, 86%; (v) MnO2, CH2Cl2, 89%; (vi) HC≡CMgBr,
THF, –78 °C → 25 °C, 98%; (vii) MnO2, CH2Cl2, 86%; (viii) NaI, K2CO3, butanone, 92%.
When a solution of the iododienynone 13b in refluxing mediate 15 could even allow for a facile intramolecular H-
benzene was treated with Bu3SnH in AIBN over 8 hours, radical transfer from this methyl to the vinyl radical via an
work-up and chromatography led to the isolation of a sin- eight-membered transition state.
1
gle product in approximately 35% yield. The H NMR
If the different outcome seen with the radical cyclisation
spectrum showed that both the E-styryl and the Z-trisub-
of 13, compared with 2, was not interesting enough, then
stituted double bonds in the starting material had been re-
more was to come when the E-isomer corresponding to
13a, i.e. 12a, was treated in a similar manner with
tained in the product, but that the terminal acetylene unit
in 13 had been replaced by a new E-disubstituted double
bond, i.e. dH 6.21 (1 H, d, J = 15.0 Hz), 6.58 (1 H, dt, J =
15.0, 7.5 Hz).
O
Further analysis of the 2D 1H NMR and 13C NMR spectra
supported the macrocyclic structure 16 for the product, re-
13
sulting from a 14-endo-dig cyclisation from the radical in-
termediate 14 produced from 13, followed by H-quench at
a R = OMe
b R = H
R
14
the resulting vinyl radical 15. A similar product was pro-
duced when the substrate 13a was treated likewise with
Bu3SnH–AIBN under identical conditions (Scheme 2).
14-endo dig
The single radical cyclisation of 13 to 16 in comparison
with the triple cyclisation of 2 to the oestrone 3 is surpris-
ing, bearing in mind that the precursors 2 and 13 differ
O
O
only by one carbon in the length of their top chains, and
that 13 has a trisubstituted rather than a disubstituted
double bond in the same carbon chain. Clearly, however,
these subtle changes have been sufficient to change the
conformational bias of the macrocyclic intermediate 15 to
preclude further radical transannular C → C bond-form-
ing reactions at the expense of H abstraction processes.
Indeed, the proximity of the vinyl radical centre to the me-
thyl group on the Z-trisubstituted double bond in the inter-
H quench
R
R
16
15
Scheme 2 Formation of the macrocycle 16 from the iododienynone
13.
Synlett 2006, No. 18, 3073–3076 © Thieme Stuttgart · New York

LETTER
Radical-Mediated Cyclisation Approach to Ring-D Aromatic Steroids
3075
Bu3SnH–AIBN in refluxing benzene.8 This reaction pro- tramolecular carbon-to-carbon bond-forming processes
duced a number of products from which an oily mixture involving C-1 and C-11, C-5 and C-10, and C-4 and C-13,
of two diastereoisomers of a chemically homogenous in addition to an intermolecular radical coupling between
polycyclic product was separated by chromatography in C-14 and the benzene solvent (see structure 17). Although
approximately 30% yield. Initially, we deluded ourselves the formation of radicals from, and the addition of carbon
into believing that this product was a mixture of diaste- centred radicals to, benzene and other aryls, is precedent-
reoisomers of the hoped for ring-D steroid structure 5 ed,11 we have no satisfactory understanding or rational ex-
(R = OMe), even though there were some inconsistencies planation for the formation of the novel bridged tricycle
i.e. unaccounted for additional absorptions, in the dH = 17 from 12a under the stated radical conditions. Some
1
7.2–7.5 ppm region of the H NMR spectrum. The oily radical reactions are known to be promiscuous, and many
mixture of purified diastereoisomers of the polycyclic lead to unusual structures not available by more conven-
phenol methyl ether, resulting from treatment of 12a with tional synthetic methods. The conversion 12a → 17 is,
Bu3SnH–AIBN, was demethylated using BBr3 leading to without doubt, novel and unprecedented, and the unex-
a mixture of diastereoisomers of the corresponding poly- pected outcome has made us further evaluate our strategy
cyclic phenol. The major diastereoisomer was purified by toward ring-D aromatic steroids involving similar radical
crystallisation and its structure was analysed by X-ray cascade processes.
crystallography.
Acknowledgment
We thank the EPSRC and AstraZeneca for their support of this work
via a studentship (to DAS).
References
(1) (a) Begley, M. J.; Crombie, L.; Ham, P. J.; Whiting, D. A. J.
Chem. Soc., Perkin Trans. 1 1976, 304; and references cited
therein. See also: (b) Bates, R. B.; Morehead, S. R. J. Chem.
Soc., Chem. Commun. 1974, 125. (c) Bates, R. B.; Eckert,
D. J. J. Am. Chem. Soc. 1972, 94, 8258.
(2) Stoltz, B. M.; Kano, T.; Corey, E. J. J. Am. Chem. Soc. 2000,
122, 9044.
Figure 2 X-ray crystal structure of 17b
(3) See: Green, S. P.; Whiting, D. A. J. Chem. Soc., Perkin
Trans. 1 1996, 1027; and references cited therein.
(4) For example: (a) De Boeck, B.; Harrington-Frost, N. M.;
Pattenden, G. Org. Biomol. Chem. 2005, 3, 340.
(b) Boehm, H. M.; Handa, S.; Pattenden, G.; Roberts, L.;
Blake, A. J.; Li, W.-S. J. Chem. Soc., Perkin Trans. 1 2000,
3522. (c) Handa, S.; Nair, P. S.; Pattenden, G. Helv. Chim.
Acta 2000, 83, 2629. (d) Jones, P.; Li, W.-S.; Hitchcock, S.
A.; Houldsworth, S. J.; Pattenden, G.; Pryde, D. C.;
Thomson, N. M.; Blake, A. J. J. Chem. Soc., Perkin Trans. 1
1998, 3181. (e) Pattenden, G.; Thomson, N. M. Tetrahedron
Lett. 1997, 38, 9069.
To our amazement the crystal structure analysis
(Figure 2)9 showed that we not only had produced the
bridged tricyclic structure 17 rather than the angular 6,6,6-
ring system 5, from the radical cascade but that we had si-
multaneously incorporated a new benzene ring at the ter-
minus of the ynone electrophore in the original starting
material 12a. This new ‘benzene addition’ structure
(Figure 2) then allowed us to rationalise the inconsisten-
cies we were earlier confused by in the 1H NMR spectrum
of the product (Equation 3).10
(5) Pattenden, G.; Gonzalez, M. A.; McCulloch, S.; Walter, A.;
Woodhead, S. J. Proc. Natl. Acad. Sci. U.S.A. 2004, 101,
12024.
The formation of the bridged tricycle 17 from the iododi-
enynone 12a under radical conditions requires three in-
(6) The benzothiazole sulfone 7 was prepared from g-valero-
lactone in four steps: i) reduction using LiAlH4, ii) protec-
tion as the TBDPS ether, iii) coupling to 2-mercaptobenzo-
thiazole, iv) oxidation of the sulphide to the sulphone using
MCPBA. See also: Baudin, J. B.; Hareau, G.; Julia, S. A.;
Lorne, R.; Ruel, O. Bull. Soc. Chim. Fr. 1993, 130, 856.
(7) For a review, see: Blakemore, P. R. J. Chem. Soc., Perkin
Trans. 1 2002, 2563.
O
O
14
Bu3SnH
13
11
11
H
13
1
14
AIBN, 80 °C,
benzene
4
10
10
(8) Typical Experimental Procedure for the Synthesis of
Polycycles 17a and 17b.
5
H
H
I
4
RO
RO
Polycycle 17a.
1
5
A solution of tri-n-butyltin hydride (170 mL, 0.61 mmol) and
2,2¢-azobis(isobutyronitrile) (AIBN; 25 mg, 0.15 mmol) in
degassed benzene (20 mL) was added dropwise over 8 h by
syringe pump to a stirred solution of the iodide 12a (200 mg,
0.51 mmol) and AIBN (50 mg, 0.30 mmol) in degassed
benzene (200 mL) under reflux and an argon atmosphere.
The mixture was heated under reflux for a further 12 h, then
17
12a
a R = Me
b R = H
Equation 3 Formation of the bridged tricycle 17a from the iodo-
dienyone 12a
Synlett 2006, No. 18, 3073–3076 © Thieme Stuttgart · New York

3076
G. Pattenden et al.
LETTER
allowed to cool to r.t. and concentrated in vacuo. The residue
was purified by column chromatography on silica, using a
gradient of 2–10% Et2O in light PE (bp 40–60 °C) as eluent,
to give the bridged tricyclic ketone 17a (45 mg, 30%) as an
inseparable mixture of diastereoisomers in a 2:1 ratio as a
colourless oil.
1.95–2.13 (3 H, m), 2.79 (1 H, app d, J = 10.2 Hz, C=OCH),
2.93–2.99 (2 H, m, ArCH2), 3.30 (1 H, dd, J = 10.2, 1.9 Hz,
ArCH), 5.25 (1 H, br s, OH), 6.59 (1 H, d, J = 2.6 Hz,
HOCCHC), 6.65 (1 H, dd, J = 8.4, 2.6 Hz, HOCCHCH),
6.71 (1 H, s, PhCH), 6.92 (1 H, d, J = 8.4 Hz, HOCCHCH),
7.25–7.39 (5 H, m, PhH). 13C NMR (100.6 MHz): d (major
diastereoisomer) = 21.4 (t), 24.7 (t), 24.9 (t), 27.6 (q), 31.0
(t), 36.1 (d), 37.0 (t), 43.5 (s), 45.7 (d), 52.6 (d), 113.7 (d),
114.8 (d), 127.8 (d), 127.9 (d), 128.7 (d), 128.8 (s), 129.0 (d),
135.4 (d), 137.0 (s), 141.0 (s), 144.0 (s), 153.3 (s), 205.9 (s);
d (minor diastereoisomer) = 21.8 (t), 23.8 (t), 24.3 (q), 25.9
(t), 28.9 (t), 40.7 (s), 43.0 (d), 46.0 (t), 46.8 (t), 47.3 (d),
112.3 (d), 115.7 (d), 127.7 (s), 127.8 (d), 128.0 (d), 128.7 (d),
128.9 (d), 135.3 (d), 138.1 (s), 139.3 (s), 144.5 (s), 154.1 (s),
206.7 (s); ESI-MS: m/z calcd for C25H27O2: 359.2006; found:
359.1999 [MH+].
IR (film): 1693, 1612 cm–1. 1H NMR (400 MHz): d (major
diastereoisomer) = 1.39 (3 H, s, CH3), 1.47–1.63 (3 H, m),
1.69–1.85 (3 H, m), 2.01–2.19 (3 H, m), 2.67 (1 H, app td,
J = 15.5, 2.9 Hz, ArCHaHb), 2.89 (1 H, app dt, J = 15.5, 3.4
Hz, ArCHaHb), 3.21 (1 H, ddd, J = 8.1, 3.0, 2.4 Hz,
C=OCH), 3.32 (1 H, dd, J = 10.1, 3.0 Hz, ArCH), 3.80 (3 H,
s, OCH3), 6.69 (1 H, d, J = 2.7 Hz, MeOCCHC), 6.77 (1 H,
dd, J = 8.6, 2.7 Hz, MeOCCHCH), 6.87 (1 H, s, PhCH), 7.17
(1 H, d, J = 8.6 Hz, MeOCCHCH), 7.26–7.44 (5 H, m, PhH);
d (minor diastereoisomer) = 1.27 (3 H, s, CH3), 1.51–1.82 (6
H, m), 1.96–2.19 (3 H,m), 2.81 (1 H, app d, J = 8.7 Hz,
C=OCH), 3.00 (2 H, app t, J = 8.4 Hz, ArCH2), 3.29 (1 H,
dd, J = 8.7, 1.6 Hz, ArCH), 3.79 (3 H, s, OCH3), 6.68 (1 H,
d, J = 2.9 Hz, MeOCCHC), 6.71 (1 H, s, PhCH), 6.78 (1 H,
dd, J = 8.4 and 2.9 Hz, MeOCCHCH), 7.00 (1 H, d, J = 8.4
Hz, MeOCCHCH), 7.26–7.38 (5 H, m, PhH). 13C NMR
(100.6 MHz): d (major diastereoisomer) = 21.4 (t), 24.7 (t),
24.8 (t), 27.6 (q), 31.2 (t), 36.1 (d), 36.9 (t), 43.4 (s), 45.7 (d),
52.6 (d), 55.2 (q), 112.4 (d), 113.3 (d), 127.8 (2 C d), 127.9
(d), 128.5 (d), 128.7 (s), 129.0 (2 C d), 135.3 (d), 136.9 (s),
140.7 (s), 144.0 (s), 157.4 (s), 205.8 (s); d (minor
(9) Crystal Data.
C25H26O2, M = 358.46, monoclinic, a = 8.7699 (9),
b = 22.255 (2), c = 9.7402 (10) Å, b = 95.633 (2)°,
V = 1891.9 (5) Å3, T = 150 (2) K, space group ‘Ia’ (No. 9),
Z = 4, Dcalcd = 1.259 g cm–3, m(Mo-Ka) = 0.078 mm–1, 2181
unique reflections measured and used in all calculations.
Final R1 [2066 F > 4s(F)] = 0.0401 and wR2 [all F2] was
0.0994. Data have been deposited with the Cambridge
Crystallographic Data Centre as CCDC 601625; they are
available free of charge via www.ccdc.cam.ac.uk/
data_request/cif.
diastereoisomer) = 21.8 (t), 23.7 (t), 24.2 (q), 25.9 (t), 29.1
(t), 40.7 (s), 42.9 (d), 46.0 (t), 46.7 (d), 47.3 (d), 55.2 (q),
110.6 (d), 114.4 (d), 124.2 (d), 127.7 (2 C d), 127.8 (d),
129.0 (2 C d), 131.6 (s), 135.0 (d), 136.6 (s), 138.7 (s), 142.6
(s), 158.0 (s), 206.1 (s). ESI-MS: m/z calcd for: 373.2162;
found: 373.2155 [MH+].
(10) An identical 2:1 mixture of diastereoisomers of desmethoxy
bridged tricycle, corresponding to 17a, was obtained (ca.
30%) when the E-isomer 12b was treated similarly with
Bu3SnH–AIBN.
Spectroscopic Data for Desmethoxy 17a.
IR (film): 1694, 1614 cm–1. 1H NMR (400 MHz): d (major
diastereoisomer) = 1.40 (3 H, s, CH3), 1.48–1.89 (6 H, m),
1.98–2.24 (3 H, m), 2.68 (1 H, app td, J = 15.6, 2.6 Hz,
ArCHaHb), 2.94 (1 H, app dt, J = 15.6, 3.5 Hz, ArCHaHb),
3.28 (1 H, app dt, J = 8.0, 3.0 Hz, C=OCH), 3.38 (1 H, dd,
J = 10.0, 3.0 Hz, ArCH), 6.89 (1 H, s, PhCH), 7.08–7.43 (9
H, m, ArH); d (minor diastereoisomer) = 1.27 (3 H, s, CH3),
1.48–1.89 (6 H, m), 1.98–2.24 (3 H, m), 2.86 (1 H, app d,
J = 10.3 Hz, C=OCH), 2.99–3.05 (2 H, m, ArCH2), 3.34 (1
H, app dd, J = 10.3, 1.6 Hz, ArCH), 6.71 (1 H, s, PhCH),
7.08–7.43 (9 H, m, ArH). 13C NMR (100.6 MHz): d (major
diastereoisomer) = 21.4 (t), 24.7 (t), 25.0 (t), 27.6 (q), 30.9
(t), 36.8 (d), 37.1 (t), 43.5 (s), 45.7 (d), 52.4 (d), 55.2 (q),
125.7 (d), 126.4 (d), 127.5 (d), 127.9 (2 C d), 128.0 (d),
129.0 (d), 129.1 (d), 135.4 (d), 136.8 (s), 137.0 (s), 139.6 (s),
144.1 (s), 205.7 (s); d (minor diastereoisomer) = 21.8 (t),
24.0 (t), 24.3 (q), 25.8 (t), 28.7 (t), 40.7 (s), 43.6 (d), 46.0 (t),
46.5 (d), 47.2 (d), 123.3 (d), 125.4 (d), 126.2 (d), 127.8 (2 C
d), 128.6 (2 C d), 128.9 (d), 135.1 (d), 136.6 (s), 137.4 (s),
139.5 (s), 142.6 (s), 206.1 (s). ESI-MS: m/z calcd for
C25H27O: 343.2056; found: 343.2053 [MH+].
Polycycle 17b.
Boron tribromide (50 mL, 0.53 mmol) was added dropwise
to a stirred solution of the tricycle 17a (50 mg, 0.13 mmol)
in anhyd CH2Cl2 (10 mL), at –78 °C under a nitrogen
atmosphere. The solution was warmed to r.t. slowly over 13
h, and then quenched with H2O (50 mL). The separated
aqueous phase was extracted with CH2Cl2 (3 × 50 mL) and
the combined organic extracts were dried over MgSO4 and
concentrated in vacuo. The residue was purified by column
chromatography, using 10% Et2O in light PE (bp 40–60 °C)
as eluent, to give a 2:1 mixture of diastereoisomers of the
phenol 17b (23 mg, 48%) as a viscous liquid solid.
Crystallisation from Et2O and pentane gave the major
diastereoisomer as colourless crystals; mp 195–196 °C. IR
(film): 3597, 1693, 1608 cm–1. 1H NMR (400 MHz): d
(major diastereoisomer) = 1.39 (3 H, s, CH3), 1.46–1.62 (3
H, m), 1.69–1.85 (3 H, m), 2.01–2.20 (3 H, m), 2.63 (1 H,
app td, J = 15.6, 3.4 Hz, ArCHaHb), 2.85 (1 H, app dt,
J = 15.6, 3.2 Hz, ArCHaHb), 3.19 (1 H, app d, J = 10.1 Hz,
C=OCH), 3.31 (1 H, dd, J = 10.1, 3.0 Hz, ArCH), 4.71 (1 H,
br s, OH), 6.62 (1 H, d, J = 2.5 Hz, HOCCHC), 6.66 (1 H,
dd, J = 8.3, 2.5 Hz, HOCCHCH), 6.89 (1 H, s, PhCH), 7.11
(1 H, d, J = 8.3 Hz, HOCCHCH), 7.27–7.36 (3 H, m, PhH),
7.42 (2 H, app d, J = 7.3 Hz, PhH); d (minor
(11) (a) Ohno, H.; Iwasaki, H.; Eguchi, T.; Tananka, T. Chem.
Commun. 2004, 2228. (b) Benati, L.; Leardini, R.; Minozzi,
M.; Nanni, D.; Spagnolo, P.; Zanardi, G. J. Org. Chem.
2000, 65, 8669.
diastereoisomer) = 1.26 (3 H, s, CH3), 1.47–1.81 (6 H, m),
Synlett 2006, No. 18, 3073–3076 © Thieme Stuttgart · New York