A selective C-H insertion/olefination protocol for the synthesis of α-methylene-γ-butyrolactone natural products

Article abstract of DOI:10.1039/c5ob02579f

A regio- and stereoselective one-pot C-H insertion/olefination protocol has been developed for the late stage installation of α-methylene-γ-butyrolactones into conformationally restricted cyclohexanol-derivatives. The method has been successfully applied in the total synthesis of eudesmanolide natural product frameworks, including α-cyclocostunolide.

Full text of DOI:10.1039/c5ob02579f

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
A selective C–H insertion/olefination protocol
for the synthesis of α-methylene-γ-butyrolactone
natural products†
Cite this: DOI: 10.1039/c5ob02579f
Matthew G. Lloyd,a Mariantonietta D’Acunto,b Richard J. K. Taylor*a and
William P. Unsworth*a
Received 16th December 2015,
Accepted 17th December 2015
A regio- and stereoselective one-pot C–H insertion/olefination protocol has been developed for the late
stage installation of α-methylene-γ-butyrolactones into conformationally restricted cyclohexanol-deriva-
tives. The method has been successfully applied in the total synthesis of eudesmanolide natural product
frameworks, including α-cyclocostunolide.
DOI: 10.1039/c5ob02579f
www.rsc.org/obc
The selective functionalization of sp3 centres via the activation
of unfunctionalised C–H bonds is of much current interest,1
given that it facilitates the synthesis of complex molecular
architectures from relatively simple precursors. Over the last
two decades, rhodium(II)-catalysed C–H insertions have
become a mainstay in this field; an array of useful reaction
modes with well-established reactivity patterns have been deve-
loped, including asymmetric variants, based on the C–H inser-
tion of rhodium-stabilised carbenoids.2 In particular, the
donor/acceptor carbenoid systems popularised by Davies have
proved to be especially valuable.2d
Fig. 1 C–H
γ-butyrolactones.
insertion/olefination
approach
to
α-methylene-
The acceptor/acceptor carbenoid class has received less
attention in comparison,3 although there are prominent excep-
tions.4 A useful feature of carbenoids of this type is the fact
that the additional acceptor substituent may be used as a
handle for further chemical modification; this is exemplified
by work published by our own group, in which a one-pot
pared to those examined previously, as there is no electronic
bias to direct the C–H insertion, but the products 4 are argu-
ably more important given that a huge number of bioactive
cyclohexane-based α-methylene-γ-butyrolactone natural pro-
ducts have been isolated.6 Our aim (Fig. 1b) was to design the
cyclohexane precursors so that C–H insertion (and subsequent
olefination) occurs exclusively into equatorial C–H bonds,7 to
selectively form fused γ-lactones 4. The success of this
approach, and its application in the total synthesis of three
natural product targets and one isomeric analogue, are
described.
Our only previous attempt at performing a C–H insertion/
olefination of this type was not encouraging; when the diazo-
phosphonate derivative of cyclohexanol (i.e. compound 3, with
R = H) was treated under the standard reaction conditions
[Rh2(oct)4 (2 mol%), CH2Cl2, 45 °C, 16 h; (ii) KOBu-t, THF, −78
– 0 °C, 1 h; (iii) (CH2O)n, 0 °C – rt, 1 h]5 a diastereomeric
mixture of γ-lactones, as well as some β-lactone product, was
obtained (corresponding to insertion into all three of the high-
lighted C–H bonds in 3) and the overall yield was low.5b It was
postulated that the poor selectivity in this case may be related
rhodium(II)-catalysed
C–H
insertion/Horner–Wadsworth–
Emmons olefination (HWE) sequence for the conversion of
α-diazo-α-(diethoxy)phosphoryl acetates 1 into α-methylene-
γ-butyrolactones 2 was reported (Fig. 1a).5 This research
focused primarily on substrates with electron rich C–H bonds
(e.g. benzylic reaction system 1) that are well-suited to react
with electrophilic carbenoids. The work described herein con-
cerns the extension of this method to cyclohexanol derivatives
(3, Fig. 1b). These are much more challenging substrates com-
aDepartment of Chemistry, University of York, Heslington, York, YO10 5DD, UK.
E-mail: richard.taylor@york.ac.uk, william.unsworth@york.ac.uk
bUniversity of Salerno, Department of Chemistry and Biology, Via Giovanni Paolo II,
132, Salerno, 84084 Fisciano, Italy
†Electronic supplementary information (ESI) available: Experimental, spectral
and crystallographic data. CCDC 1421154, 1421158 and 1421164. For ESI and
crystallographic data in CIF or other electronic format see DOI: 10.1039/
c5ob02579f
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
to the flexible nature of the substrate, and that by restricting
In view of the excellent regio- and stereoselectivity observed
its conformation, the regio- and stereoselectivity may be in these reactions, attention turned to their application in
improved. natural product synthesis. Sesquiterpene lactones are the most
To test this idea, 4-tert-butyl cyclohexyl derivatives 5a and common class of α-methylene-γ-butyrolactone found in
5b were formed and reacted under the standard one-pot C–H Nature6e,8 and selected compounds from a sub-class of this
insertion/olefination conditions (Fig. 2). The expectation was family collectively known as the eudesmanolides, are shown in
that the tert-butyl group would lock the cyclohexane scaffold in Fig. 3 (8–13).
a chair conformation to better distinguish the two γ-C–H inser-
First, compounds 8 and 9 (labelled morifolins A and B,
tion sites, and improve the diastereoselectivity. Greater control Fig. 3) were targeted. As the C–H insertion procedure had not
was indeed observed in both cases; γ-lactones 6a and 6b were previously been tested on a cis-decalin framework (which can
each isolated as single diastereoisomers, with the stereochemi- potentially ring flip), they were considered to be an interesting
cal outcome consistent with insertion into the equatorial C–H challenge to the methodology. An additional reason for
bonds. However, in both reactions the overall yield was rela- performing the synthesis of the morifolins was to clear up con-
tively low, which is partly explained by the formation of fusion that exists in the literature about their structural assign-
β-lactone side-products 7a and 7b (not shown, see ESI†). Pleas- ments. In 1985, Dominguez and co-workers isolated a series of
ingly, by moving the tert-butyl group closer to the C–H inser- sesquiterpene lactones, two of which were named morifolin
tion site, a significant improvement was observed; 2-tert-butyl A and B and assigned the structures 8 and 9 above.9 However,
substrates 5c and 5d furnished α-methylene-γ-butyrolactones in a 2004 publication,10 Herz suggested that these products
6c and 6d respectively, with complete diastereoselectivity and a had been assigned incorrectly, and proposed that they were in
significant increase in isolated yield (76% and 90%), with no fact identical to isocritonilide 10 and critonilide 11, respecti-
β-lactone side-products being formed in either case. The tert- vely, described by Bohlmann and co-workers in 1983.11 The
butyl group is likely to be playing two roles in these substrates, spectral data in the Dominguez publication were insufficient
both fixing the conformation of the cyclohexane ring and pro- to draw a definitive conclusion, and hence it was decided to
viding a steric barrier to competing β-insertion reactions. To complete the total syntheses of lactones 8 and 9 to clarify the
further probe this cooperative effect, other 2-subtituted confor- anomaly.
mationally restricted systems based on menthol and decalinol
The synthesis began with a lithium naphthalenide
(5e, 5f and 5g) were treated under the standard conditions and mediated 1,2-addition of chloride 14 into cyclohexenone,12
all afforded the expected γ-lactone products selectively (6e, 6f which was followed by oxidative rearrangement with PCC,
and 6g). It is noteworthy that in all cases, there is complete furnishing α,β-unsaturated ketone 16 (Scheme 1). Next, the
selectivity for equatorial C–H insertion, irrespective of whether 1,4-addition of methylmagnesium chloride under Gilman-type
the diazoester substituent itself has an equatorial (6a, 6c, 6e, conditions afforded doubly-masked keto-aldehyde 17. This was
6f) or axial (6b, 6d, 6g) configuration.
followed by an intramolecular aldol reaction under acidic con-
ditions, which furnished cis-β-hydroxyketone 18 as a single dia-
stereoisomer, as reported in the literature.13 Next, silyl
protection of the alcohol, vinyl triflate formation and iron-
catalysed cross-coupling14 with methylmagnesium chloride
furnished alkene 21 in excellent overall yield. Desilylation
using TBAF followed by a T3P-mediated acylation and Regitz
diazo transfer reaction, generated the key diazo substrate 24.
This was then primed to undergo the one-pot C–H insertion/
olefination sequence, which was performed under the
Fig. 2 C–H insertion/olefination sequence for conformationally
restricted cyclohexane derivatives 5a–g. Reaction conditions: (i)
Rh2(oct)4 (2 mol%), CH2Cl2, 45 °C, 16 h; (ii) KOBu-t, THF, −78 – 0 °C, 1 h;
a
(iii) (CH2O)n, 0 °C – rt, 1 h. Yields of isolated product.
β-lactone
b
product 7a, (12%) was also isolated; β-lactone product 7b, (19%) was
also isolated.
Fig. 3 Eudesmanolide natural products 8–13.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 2 Synthesis of α-methylene-γ-butyrolactone 9. Reagents and
conditions: (a) TMSCH2Li, THF, −78 °C, 84%; (b) NaH, THF, 65 °C, 100%;
(c) TBAF, THF, 65 °C, 81%; (d) diethylphosphonoacetic acid, DIPEA, T3P,
PhMe, rt, 82%; (e) LHMDS, p-ABSA, THF, −78 °C – rt, 78%; (f) (1)
Rh2(OAc)4 (2 mol%), CH2Cl2, 45 °C, 16 h; (2) KOBu-t, THF, 0 °C, 1 h; (3)
(CH2O)n, −78 °C – rt, 1 h, 45% (9) and 11% (30).
with a small amount of a cyclopropane side-product 30 (not
shown, see ESI†). The 1H NMR data of synthetic material 9
were again significantly different to those published for the
natural product,9 confirming that morifolin B was also in-
correctly assigned in the literature. Therefore, similarly to
morifolin A, it again seems most likely that the isolated
material named morifolin B 9 is in fact the same as critonilide
11, again as suggested previously by Herz.10
Scheme 1 Synthesis of α-methylene-γ-butyrolactone 8. Reagents and
conditions: (a) (1) 14, Li, naphthalene, THF, −78 °C; (2) 2-cyclohexen-1-
one, −78 °C – rt, 100%; (b) PCC, Al2O3, CH2Cl2, 0 °C – rt, 52%; (c) CuI,
LiCl, TMSCl, MeMgCl, THF, −78 °C – rt, 98%; (d) 10% aq. HCl, MeOH,
80 °C, 65%; (e) TBSOTf, 2,6-lutidine, CH2Cl2, 0 °C – rt, 99%; (f) LHMDS,
Tf2O, THF, −40 °C, 80%; (g) Fe(acac)3, MeMgCl, THF : NMP (1 : 3), 82%;
(h) TBAF, THF, 65 °C, 88%; (i) diethylphosphonoacetic acid, DIPEA, T3P,
PhMe, rt, 100%; ( j) LHMDS, p-ABSA, THF, −78 °C – rt, 89%; (k) (1)
Rh2(oct)4 (2 mol%), CH2Cl2, 45 °C, 16 h; (2) KOBu-t, THF, 0 °C, 1 h; (3)
(CH2O)n, −78 °C – rt, 1 h, 64%.
Next, attention turned to the synthesis of α-cyclocostunolide
12,17
a
cytotoxic18 trans-decalin eudesmanolide natural
product with anti-trypanosomal19 and anti-coagulant activity.20
Its synthesis began with a two-step epimerisation of cis-
β-hydroxyketone 18 via an oxidation–reduction sequence,
which provided the desired trans-β-hydroxyketone 32,
standard conditions, affording α-methylene-γ-butyrolactone 8 in addition to diastereoisomeric cis-β-hydroxyketone 33
in 64% yield with complete stereo- and regiocontrol. The (Scheme 3).21
relative configuration of 8 was found to be in line with that
Then, the same sequence shown in Scheme 2 was per-
observed in the model studies and was assigned unambi- formed on each of these β-hydroxyketones, affording diazo
guously by X-ray crystallography.15 With the identity of our syn- substrates 34 and 35 without complication. We were then
thetic sample 8 confirmed, its spectral data were then pleased to isolate α-cyclocostunolide 12 as the sole product
compared with those reported for the natural product mori- from the reaction of diazophosphonate 34 under the standard
folin A; significant differences between the 1H NMR data of one-pot C–H insertion/olefination conditions, with its spectral
cis-decalin 8 and the natural product were clearly observed, data fully matching those of natural α-cyclocostunolide.17 In
thus confirming that morifolin A had indeed been incorrectly addition, diastereomeric lactone 36 was also isolated from dia-
assigned. Thus it appears most likely that the natural isolated zophosphonate 35, and was again formed in good yield, via
material named morifolin A 8 is in fact the same as isocritoni- selective equatorial C–H insertion.22
lide 10, as suggested previously by Herz.10
In summary, a highly regio- and stereoselective one-pot
The synthesis of the proposed structure of morifolin B 9 C–H insertion/olefination protocol for the late-stage functiona-
started with a common intermediate from the morifolin A lisation of conformationally restricted cyclohexanol-derivatives
route, ketone 19, which was treated with TMSCH2Li to form has been developed. Exclusive formation of γ-lactones via
alcohol 25 (Scheme 2). This was followed by a base-mediated insertion into equatorial C–H bonds was observed and the
Peterson olefination to generate exocyclic alkene 26, and sub- method was validated in natural product synthesis. Eudesma-
sequent desilylation, acylation and diazotization as before, to nolide sesquiterpene natural product α-cyclocostunolide 12
generate diazo substrate 29 in excellent overall yield. Then, was synthesised in racemic form in high yield using this proto-
Rh2(OAc)4 catalysed C–H insertion16 and olefination in the col. In addition, structures 8 and 9, originally assigned to the
usual way, furnished lactone 9 as a single diastereoisomer natural products morifolin B and morifolin A, were prepared
(which was again verified by X-ray crystallography),15 along and it was demonstrated unambiguously that the original
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
M. M. Pearson, J. Am. Chem. Soc., 1993, 115, 958;
(c) T. Yakura, D. Yoshida, A. Ueki, K. Nakao and M. Ikeda,
Chem. Pharm. Bull., 1997, 45, 651; (d) C. N. Slattery and
A. R. Maguire, Tetrahedron Lett., 2013, 54, 2799;
(e) T. Takahashi, H. Tsutsui, M. Tamura, S. Kitagaki,
M. Nakajima and S. Hashimoto, Chem. Commun., 2001,
1604; (f) C. S. Jungong and A. V. Novikov, Tetrahedron:
Asymmetry, 2013, 24, 151.
5 (a) M. G. Lloyd, R. J. K. Taylor and W. P. Unsworth, Org.
Lett., 2014, 16, 2772; (b) M. G. Lloyd, M. D’Acunto,
R. J. K. Taylor and W. P. Unsworth, Tetrahedron, 2015, 71,
7107.
6 (a) V. Saroglou, A. Karioti, J. Heilmann, Z. Kypriotakis and
H. Skaltsa, Helv. Chim. Acta, 2007, 90, 171; (b) Y.-Y. Chou
and C.-C. Liao, Org. Lett., 2013, 15, 1584; (c) L. Kraut,
R. Mues and M. Sim-Sim, Phytochemistry, 1994, 37, 1337;
(d) J. Triana, J. L. Eiroa, M. Morales, F. J. Pérez, I. Brouard,
M. T. Marrero, S. Estévez, J. Quintana, F. Estévez,
Q. A. Castillo and F. León, Phytochemistry, 2013, 92, 87;
(e) R. R. A. Kitson, A. Millemaggi and R. J. K. Taylor,
Angew. Chem., Int. Ed., 2009, 48, 9426 and references
therein.
7 For examples of related equatorial C–H insertions, see:
(a) D. F. Taber and R. E. Ruckle, J. Am. Chem. Soc., 1986,
108, 7686; (b) M. P. Doyle, A. B. Dyatkin, G. H. P. Roos,
F. Canas, D. A. Pierson, A. Vanbasten, P. Muller and
P. Polleux, J. Am. Chem. Soc., 1994, 116, 4507;
(c) M. P. Doyle, A. V. Kalinin and D. G. Ene, J. Am. Chem.
Soc., 1996, 118, 8837; (d) M. P. Doyle, S. B. Davies and
E. J. May, J. Org. Chem., 2001, 66, 8112; (e) M. P. Doyle,
V. Bagheri, M. M. Pearson and J. D. Edwards, Tetrahedron
Lett., 1989, 30, 7001.
Scheme 3 Synthesis of ( )-α-cyclocostunolide 12 and α-methylene-
γ-butyrolactone 36. Reagents and conditions: (a) DMP, CH2Cl2, 0 °C,
80%; (b) NaBH4, MeOH, 0 °C, 10% (32), 16% (33); (c) TBSOTf, 2,6-lutidine,
CH2Cl2, 0 °C – rt, 100%; (d) Tf2O, DTBMP, CH2Cl2, rt, 67%; (e) Fe(acac)3,
MeMgCl, THF : NMP (1 : 3), 92%; (f) TBAF, THF, 65 °C, 92%; (g) diethyl-
phosphonoacetic acid, DIPEA, T3P, PhMe, rt, 99%; (h) LHMDS, p-ABSA,
THF, −78 °C – rt, 88%; (i) TBSOTf, 2,6-lutidine, CH2Cl2, 0 °C – rt, 96%;
( j) LHMDS, Tf2O, THF, −40 °C, 62%; (k) Fe(acac)3, MeMgCl, THF : NMP
(1 : 3), 59%; (l) TBAF, THF, 65 °C, 71%; (m) diethylphosphonoacetic acid,
DIPEA, T3P, PhMe, rt, 63%; (n) LHMDS, p-ABSA, THF, −78 °C – rt, 75%;
(o) (1) Rh2(oct)4 (2 mol%), CH2Cl2, 45 °C, 16 h; (2) KOBu-t, THF, 0 °C, 1 h;
(3) (CH2O)n, −78 °C – rt, 1 h, 52% (12 from 34) and 66% (36 from 35).
structural assignments were in error. Finally, a fourth isomeric
α-methylene-γ-butyrolactone 36, which is apparently novel, was
also prepared.22 This chemistry is expected to be applicable to
a variety of synthetic targets possessing the α-methylene-
γ-butyrolactone motif.
8 M. Chadwick, H. Trewin, F. Gawthrop and C. Wagstaff,
Int. J. Mol. Sci., 2013, 14, 12780.
The authors wish to thank Elsevier Foundation (M. G. L.),
the University of Salerno (M. D.) and the University of York
(W. P. U.) for funding, Dr Adrian C. Whitwood (University of
York) for X-ray crystallography and Euticals for providing T3P.
9 A. G. Gonzalez, J. B. Barrera, A. Y. Hernandez, F. E. Rosas
and X. A. Dominguez, Heterocycles, 1985, 23, 1601.
10 W. Herz, Biochem. Syst. Ecol., 2004, 32, 1159.
11 (a) F. Bohlmann, J. Jakupovic and M. Lonitz, Chem. Ber.,
1977, 110, 301; (b) F. Bohlmann, J. Jakupovic and
A. Schuster, Phytochemistry, 1983, 22, 1637.
12 For related lithium naphthalenide reactions, see:
(a) M. Yus and D. J. Ramon, J. Chem. Soc., Chem. Commun.,
1991, 398; (b) B. H. Lipshutz, B. Amorelli and J. B. Unger,
J. Am. Chem. Soc., 2008, 130, 14378; (c) W. P. Unsworth,
K. A. Gallagher, M. Jean, J. P. Schmidt, L. J. Diorazio and
R. J. K. Taylor, Org. Lett., 2013, 15, 262.
Notes and references
1 (a) H. M. L. Davies and J. R. Manning, Nature, 2008, 451,
417; (b) H. M. L. Davies, Beilstein J. Org. Chem., 2012, 8,
1552.
2 (a) H. M. L. Davies and R. E. J. Beckwith, Chem. Rev., 2003,
103, 2861; (b) M. P. Doyle, R. Duffy, M. Ratnikov and 13 S. M. Abdur Rahman, H. Ohno, H. Yoshino, N. Satoh,
L. Zhou, Chem. Rev., 2010, 110, 704; (c) M. P. Doyle and
D. C. Forbes, Chem. Rev., 1998, 98, 911; (d) H. M. L. Davies
and D. Morton, Chem. Soc. Rev., 2011, 40, 1857.
M. Tsukaguchi, K. Murakami, C. Iwata, N. Maezaki and
T. Tanaka, Tetrahedron, 2001, 57, 127.
14 B. Scheiper, M. Bonnekessel, H. Krause and A. Fürstner,
J. Org. Chem., 2004, 69, 3943.
3 Possibly due to their relatively high electrophilicity, which
leads to a decrease in chemo- and stereoselectivity in many 15 CCDC 1421158(8) and 1421154(9) contain the supplemen-
processes. tary crystallographic data for this paper.
4 (a) L. Eun, W. J. Kyung and S. K. Yong, Tetrahedron Lett., 16 Rh2(OAc)4 was used instead of Rh2(oct)4 in this example, as
1990, 31, 1023; (b) M. P. Doyle, L. J. Westrum,
W. N. E. Wolthuis, M. M. See, W. P. Boone, V. Bagheri and
it led to small improvement in the ratio of products 9
and 30.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
17 For information relating to α-cyclocostunolide, see: 18 H.-S. Chang, S.-J. Lee, C.-W. Yang and I.-S. Chen, Chem. Bio-
(a) G. H. Kulkarni, G. R. Kelkar and S. C. Bhattacharyya, diversity, 2010, 7, 2737.
Tetrahedron, 1964, 20, 2639; (b) T. C. Jain and 19 K. Otoguro, M. Iwatsuki, A. Ishiyama, M. Namatame,
J. E. McCloskey, Tetrahedron Lett., 1969, 10, 2917;
(c) J.-W. de Kraker, M. C. R. Franssen, M. C. F. Dalm, A. de
Groot and H. J. Bouwmeester, Plant Physiol., 2001, 125,
A. Nishihara-Tukashima, H. Kiyohara, T. Hashimoto,
Y. Asakawa, S. Ōmura and H. Yamada, Phytochemistry,
2011, 72, 2024.
1930; (d) T. C. B. Tomassini and B. Gilbert, Phytochemistry, 20 B.-G. Wang, X. Hong, L. Li, J. Zhou and X.-J. Hao, Planta
1972, 11, 1177; (e) T. C. Jain and J. E. McCloskey, Tetra- Med., 2000, 66, 511.
hedron Lett., 1971, 12, 1415; (f) Y. Asakawa, G. Ourisson 21 The relative stereochemistry of compounds 32 and 33 were
and T. Aratani, Tetrahedron Lett., 1975, 16, 3957;
(g) J. D. Connolly and I. M. S. Thornton, Phytochemistry,
1973, 12, 631; (h) A. G. González, J. B. Barrera, A. C. Yanes
assigned retrospectively based on an X-ray crystal structure
of a TBS-protected derivative of 32 (S8); see CCDC 1421164
and the ESI† for details.
Hernández, F. E. Rosas and X. A. Dominguez, Phytochemi- 22 To the best of our knowledge, isomer 36 is novel and has
stry, 1985, 24, 1847; (i) T. C. Jain and J. E. McCloskey, Tetra-
hedron, 1975, 31, 2211; ( j) R. W. Doskotch and F. S. El-
Feraly, J. Org. Chem., 1970, 35, 1928.
not yet been found in Nature, although in view of the high
number of related natural products known, it is not incon-
ceivable that it occurs naturally.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.