Total synthesis and structural reassignment of (+)-dictyosphaeric acid a: A tandem intramolecular michael addition/alkene migration approach

Article abstract of DOI:10.1002/anie.201002416

(Figure Presented) The acid test? The synthetic route to the title compound features a Z-selective ring-closing metathesis reaction and an E-selective tandem intramolecular Michael addition/alkene migration sequence as the key transformations. The stereochemical configuration of the product was reassigned based on NMR studies.

Full text of DOI:10.1002/anie.201002416

Communications
DOI: 10.1002/anie.201002416
Natural Product Synthesis
Total Synthesis and Structural Reassignment of (+)-Dictyosphaeric
Acid A: A Tandem Intramolecular Michael Addition/Alkene
Migration Approach**
Alan R. Burns, Graeme D. McAllister, Stephen E. Shanahan, and Richard J. K. Taylor*
Given the recent increase in bacterial resistance, the search
for new and more effective antibiotics is an ever-pressing
concern. The polyketide-derived natural products (+)-dic-
tyosphaeric acid A (1) and (+)-dictyosphaeric acid B (2) were
isolated by Ireland and co-workers in 2004 from a fungal
isolate (F01V25) obtained from the green alga Dictyosphaeria
versluyii.[1] Interestingly, when screened for biological activity,
(+)-dictyosphaeric acid A (1) exhibited antibacterial activity
against methicillin-resistant Staphylococcus aureus (MRSA),
vancomycin-resistant Enterococcus faecium, and Candida
albicans. In marked contrast, (+)-dictyosphaeric acid B (2)
did not exhibit any significant antibacterial activity, which
presumably indicates the importance of the a,b-unsaturated
ketone in the pharmacophore of dictyosphaeric acid A (1). As
Scheme 1. Dictyosphaeric acid A (1) and B (2), colletofragarones A1
(3) and A2 (4), and key nOe and rOe correlations supporting the
assignment of the relative configuration of 1.
shown in Scheme 1, both natural products are based on a
highly oxygenated decalactone core and triene carboxylic
acid side-chain, with the only difference being that the enone
double bond of 1 is hydrated in 2. Dictyosphaeric acid A (1)
has five stereocentres, four of which are contiguous, whereas
dictyosphaeric acid B (2) has six stereocentres, five of which
are contiguous. The relative stereochemical configuration of
both molecules was determined by extensive 1D and two-
dimensional (2D) NMR studies[1] [key nOe and rotating-
frame Overhauser effect (rOe)[2] correlations shown in
Scheme 1]. However, assignment of the relative configuration
at C6 relied on a single correlation (not conclusive in a
conformationally mobile system) and the absolute configu-
ration was not elucidated.[1] These natural products pose a
significant synthetic challenge, not only due to their structural
complexity, but also due to the need for a flexible and
convergent route, which would give easy access to analogues
for biological screening and SAR (structure–activity relation-
ship) studies.
The only other natural products isolated to date, which
have the same tricyclic decalactone-based carbon skeleton as
the dictyosphaeric acids, are the fungal metabolites colleto-
fragarone A1 (3) and colletofragarone A2 (4).[3] To date,
there have been no total syntheses reported for either pair of
natural products, although our group published a route to
novel dictyosphaeric acid analogues in 2008,[4] and Harwood
et al. recently disclosed model studies in the colletofragarone
area.[5,6] Herein, we report the first total synthesis of
dictyosphaeric acid A (1) which also serves to clarify the
uncertainties of relative and absolute configuration referred
to above.
[*] A. R. Burns, Dr. G. D. McAllister, Prof. R. J. K. Taylor
Department of Chemistry, University of York
Heslington, York, YO10 5DD (UK)
Fax: (+44)1904-434523
E-mail: rjkt1@york.ac.uk
Homepage: http://www.york.ac.uk/res/rjkt/
Dr. S. E. Shanahan
GlaxoSmithKline, Chemical Development
Old Powder Mills, Tonbridge, Kent TN11 9AN (UK)
The retrosynthetic analysis adopted in the current study is
shown in Scheme 2. Initial disconnection would give the
activated enol 5 and an organometallic triene such as
stannane 6, with a view to utilising a Pd-mediated cross-
coupling approach. Then, we envisaged introducing the 1,2-
diol portion of the decalactone ring through dihydroxylation
or epoxidation/hydrolysis of exocyclic enone 7. We proposed
to form conjugated alkene 7 by isomerisation of alkene 8
which, in turn, would be accessed from b-keto ester 9 using a
doubly tethered intramolecular Michael addition (IMA) of
the type developed in earlier model studies.[4] We anticipated
preparing macrocyclic alkene 9 by a ring-closing metathesis
[**] We thank the BBSRC for studentship funding (A.R.B.), Glaxo-
SmithKline for CASE support, and the University of York for post-
doctoral funding (G.D.M.). We are also grateful to Dr. A. C.
Whitwood and R. Thatcher (University of York) for X-ray crystallog-
raphy studies and to Dr. T. Dransfield (University of York) for
invaluable assistance with mass spectrometry. We would also like to
thank Prof. C. M. Ireland (University of Utah) for providing us with
copies of 1H and 13C NMR spectra for compounds 1 and 2, and for
helpful discussions. We also thank Dr. C. W. Barfoot for carrying out
relevant model studies (to be reported in a full paper) and
Archimica for a kind gift of T3P.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201002416.
5574
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 5574 –5577

Angewandte
Chemie
Scheme 2. Retrosynthetic analysis.
Scheme 3. Conversion of iodide 11 into exocyclic enone 17. Reagents
and conditions: a) 14 (1 equiv), [(NBS)Pd(PPh3)2] (15) (3.4 mol%),
3m K3PO4 (aq.), THF, 608C, 16 h, 82%; b) TBAF (1.1 equiv), THF,
08C, 2.5 h; c) 13 (1 equiv), T3P (1.3 equiv), DIPEA (1.95 equiv), PhMe,
08C to RT, 5 h, 73% (over 2 steps); d) Hoveyda–Grubbs catalyst, 2nd
generation (8 mol%), CH2Cl2, 408C, 24 h, 82%; e) (optimum condi-
tions) piperidine (0.5 equiv), Bu4NHSO4 (0.1 equiv), MeCN, 858C,
24 h, 65% 9 + 23% 8; f) piperidine (0.5 equiv), THF, 708C, 24 h,
62% 17, 26% 8. NBS=N-bromosuccinimide, TBAF=tetra-butylammo-
nium fluoride, DIPEA=N,N-di-isopropyl(ethyl)amine, T3P=propane
phosphonic acid anhydride.
(RCM) reaction from precursor diene 10, which could be
further simplified to functionalized cyclohexenone 11, pro-
penyl organometallic reagent 12 and carboxylic acid 13.
The starting point in the synthesis was the known
enantiomerically enriched (> 99% ee) iodide (À)-11.[4,7]
This was expediently functionalized using a Suzuki–Miyaura
coupling with the commercially available propenyl MIDA
boronate 14[8] and trans-bromo[N-succinimidyl-bis(triphenyl-
phosphine)]palladium(II) (15), developed within our own
group,[9] as catalyst. Notably, the corresponding vinyl deriv-
ative (as opposed to the propenyl variant) could also be
prepared, but proved to be prone to decomposition, even
when stored at À208C. Straightforward elaboration to RCM
precursor 10 was achieved by desilylation and esterification
with carboxylic acid 13,[10] utilizing T3P as the coupling
reagent (Scheme 3).[11,12] Treating substrate 10 with the
Hoveyda–Grubbs second generation catalyst[13] furnished
the 13-membered macrocycle 9 in excellent yield (82%)
and, interestingly, with the exclusive formation of the (Z)-
alkene (3J = 11.5 Hz).
8 (23%) which was readily converted into 17 using piperidine
in THF. The requirement for a phase transfer catalyst in the
IMA using organic amines is fascinating (no IMA occurs in its
absence) but not yet fully understood.
Interestingly, the isomerized enone product was formed as
a single diastereomer and as a single alkene geometric
isomer; X-ray crystallography[14] confirmed that this was the
(E)-enone 17, as depicted in Scheme 4, and not as the (Z)-
enone 7 as proposed in the retrosynthetic analysis (Scheme 2).
The stage was then set for one of the key reaction
sequences of the synthesis: IMA of 9 and then alkene
isomerization to give the exocyclic enone 17. Initially, the
IMA was achieved with either NaH or Et3N/Bu4NCl (see
Table in Scheme 3, entries 1 and 2), and then the resulting
alkene 8 was isomerized using DBU to give the conjugated
alkene. However, the isomerization step was particularly low-
yielding (ca. 30%). Therefore, as both the Michael addition
and the alkene isomerization reaction could be performed
with an amine base, a tandem “two-step, one-pot” process was
investigated (entries 3 and 4). Pleasingly, and after extensive
optimization, treatment of enone 9 with sub-stoichiometric
piperidine (0.5 equiv) and tetra-butylammonium hydrogen-
sulfate (0.1 equiv) in acetonitrile facilitated the tandem IMA/
alkene isomerization reaction to furnish the exocyclic alkene
17 (65%), along with the un-reacted Michael addition adduct
Scheme 4. X-ray crystal structure[14] of (E)-enone 17 depicted using
Mercury 2.2.
Clearly, this observation concerning the 5,6-alkene configu-
ration had implications with respect to the strategy for
installing the 5S,6S-diol of dictyosphaeric acid A (1). Dihy-
droxylation of compound 17 from the more accessible top
face of the olefin would produce the 5S,6R-diastereomer
whereas nucleophilic epoxidation followed by a regioselective
epoxide opening with H2O would lead to the correct 5S,6S-
Angew. Chem. Int. Ed. 2010, 49, 5574 –5577
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5575

Communications
diol configuration with respect to the proposed structures of
the natural products.
using tandem oxidation chemistry[16]). This Pd0-catalyzed
coupling delivered adduct 22 in a 60% yield over the two-
step process. Saegusa–Ito oxidation[17] of ketone 22, through
the corresponding silyl enol ether, then proceeded as
expected to give enone 23.
Finally, treatment of compound 23 with aqueous TFA
removed both the tert-butyl ester and acetonide groups to
generate the acid 24. We were now in a position to determine
whether natural dictyosphaeric acid possessed the 5S,6S-diol
configuration (as reported in the literature)[1] or the 5S,6R-
diol configuration, which also seemed compatible with the
published 2D NMR experiments. Interestingly, the NMR data
for acid 24 was essentially identical to the published[1] NMR
data, with key comparisons being shown in Table 1 (see the
Frustratingly, various attempts at nucleophilic epoxida-
tion of the exocyclic enone 17 failed. Given the uncertainty
concerning the C6 configuration referred to earlier, we
therefore turned our attention to dihydroxylation procedures.
To our delight, utilising the mildly acidic osmylation con-
ditions developed by Sharpless et al.,[15] dihydroxylation of
enone 17 proceeded in high yield and with complete
diastereoselectivity to produce diol 18 (Scheme 5). Protection
of the diol group proved to be far from trivial (e.g. PMB-
acetal, acetate, TBS- and TMS-ether formation all failed).
Ultimately, acetonide protection was successful giving adduct
19, albeit with the unavoidable formation of the mixed acetal
20. However, compound 20 could be hydrolysed back to diol
18 (98%) with 2m HCl and thus recycled with reasonable
efficiency.
1
Table 1: Comparison of key H and 13C NMR data for synthetic (+)-24
and natural (+)-dictyosphaeric acid A (1).[a,b]
With the acetonide 19 in hand, a triflation/Stille sequence
was carried out using trienylstannane 6 (readily prepared
Synthetic (+)-24
Lit. (+)-1[1]
13C[c] 1H[d] mult.,
J [Hz]
Position 13C[c] 1H[d]
mult.,
J [Hz]
2
3
141.3 6.59–6.57
129.1 6.09
m
141.2 6.58 dd, 10.5, 3.3
dd, 10.2, 1.0 129.1 6.09 dd, 10.5, 0.8
4
5
202.9
78.1
–
–
202.9
78.0
–
–
6
10
14
74.1 3.79
75.3 4.86–4.82
52.3 4.22
d, 8.8
m
d, 8.8
74.1 3.80 d, 8.9
75.3 4.84
52.2 4.23 d, 8.7
m
[a] For a full comparison, as well as copies of spectra, see the Supporting
Information. [b] Solvent: CDCl3/CD3OD. [c] 125 MHz. [d] 500 MHz.
Supporting Information for a full comparison). In addition,
the optical rotation of acid 24 ([a]D = + 116.9, c = 0.12,
MeOH) was very close to the published value for dictyos-
phaeric acid A (lit.[1] + 126, c = 0.22, MeOH). We therefore
concluded that (+)-dictyosphaeric acid A actually possesses
the 5S,6R-24 structure and not that with the 5S,6S-diol
configuration originally reported.[18]
In summary, we have developed a concise and convergent
synthesis of the antibacterial natural product (+)-dictyos-
phaeric acid A (24), which proceeds in twelve steps from
known iodide (À)-11 in an overall yield of 10%. Through
synthesis, we were able to correct the published (minor)
structural mis-assignment, and to confirm the absolute
configuration of dictyosphaeric acid A (+)-24. We are
currently extending this route to prepare dictyosphaeric
acid B and the colletofragarones. We are also exploiting the
efficiency and diversification potential of the synthetic route
to prepare a small library of dictyosphaeric acid analogues for
biological screening. These results, along with a full account of
this work, will be published in due course.
Scheme 5. The elaboration of enone 17. Reagents and conditions:
a) K2[OsO2(OH)4] (1 mol%), NMO (1.1 equiv), citric acid (2 equiv),
MeCN/tBuOH/H2O (1:1:1), RT, 7 d, 95%; b) 2,2-dimethoxypropane
(10 equiv), acetone (20 equiv), (Æ)-CSA (0.1 equiv), PhMe, 808C, 18 h,
63% 19 + 30% 20; c) KHMDS (2.0 equiv), Tf2O (1.2 equiv), 1,2-DME,
À408C to RT, 1.5 h; d) stannane 6 (1.2 equiv), [Pd2dba3] (10 mol%),
LiCl (3 equiv), (2-furyl)3P (0.3 equiv), 1,2-DME, 858C, 1 h, 508C 16 h,
60% (2 steps); e) LiHMDS (2 equiv), Et3N (3 equiv), TMSCl (3 equiv),
THF, À788C, 1.5 h; f) Pd(OAc)2 (4.4 equiv), MeCN, RT, 4 d, 59% (2
steps, 83% brsm); g) 80% TFA (aq), CH2Cl2, RT, 28 h, 100%.
CSA=camphorsulfonic acid, dba=dibenzylideneacetone, DME=di-
methoxyethane, HMDS=hexamethyldisilazide, NMO=N-methylmor-
pholine-N-oxide, TfO=trifluoromethylsulfonate, TFA=trifluoroacetic
acid, TMS=trimethylsilyl.
Received: April 23, 2010
Published online: June 25, 2010
Keywords: dictyosphaeric acid · metathesis · Michael addition ·
.
tandem reactions · total synthesis
5576
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 5574 –5577

Angewandte
Chemie
[11] M. Wedel, A. Walter, F.-P. Montforts, Eur. J. Org. Chem. 2001,
[1] T. S. Bugni, J. E. Janso, R. T. Williamson, X. Feng, V. S. Bernan,
M. Greenstein, G. T. Carter, W. M. Maiese, C. M. Ireland, J. Nat.
Prod. 2004, 67, 1396.
1681.
[12] H. Wissmann, H. J. Kleiner, Angew. Chem. 1980, 92, 129; Angew.
Chem. Int. Ed. Engl. 1980, 19, 133.
[2] M. Ivancic, V. L. Hsu, Biopolymers 2000, 54, 35.
[3] M. Inoue, H. Takenaka, T. Tsurushima, H. Miyagawa, T. Ueno,
Tetrahedron Lett. 1996, 37, 5731.
[4] C. W. Barfoot, A. R. Burns, M. G. Edwards, M. N. Kenworthy,
M. Ahmed, S. E. Shanahan, R. J. K. Taylor, Org. Lett. 2008, 10,
353.
[13] S. B. Garber, J. S. Kingsbury, B. L. Gray, A. H. Hoveyda, J. Am.
Chem. Soc. 2000, 122, 8168.
[14] CCDC 764933 (17) contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
from the Cambridge Crystallographic Data Centre at www.ccdc.
cam.ac.uk/data_request/cif.
[5] M. G. B. Drew, A. Jahans, L. M. Harwood, S. A. B. H. Apoux,
Eur. J. Org. Chem. 2002, 3589.
[15] P. Dupau, R. Epple, A. A. Thomas, V. V. Fokin, K. B. Sharpless,
Adv. Synth. Catal. 2002, 344, 421.
[6] J. G. Marrero, L. M. Harwood, Tetrahedron Lett. 2009, 50, 3574.
[7] M. T. Barros, C. D. Maycock, M. R. Ventura, J. Chem. Soc.
Perkin Trans. 1 2001, 166.
[8] D. M. Knapp, E. P. Gillis, M. D. Burke, J. Am. Chem. Soc. 2009,
131, 6961.
[9] M. J. Burns, I. J. S. Fairlamb, A. R. Kapdi, R. J. K. Taylor, Org.
Lett. 2007, 9, 5397.
[10] S-(+)-Hex-5-en-2-ol (99.3% optical purity) is commercially
available.
[16] R. J. K. Taylor, M. Reid, J. Foot, S. A. Raw, Acc. Chem. Res.
2005, 38, 851.
[17] Y. Ito, T. Hirao, T. Saegusa, J. Org. Chem. 1978, 43, 1011.
[18] After the preparation of this manuscript, a pre-print was sent to
Professor C. M. Ireland, University of Utah, who replied (email
correspondence dated April 21, 2010) stating that “The hydroxy
at C6 is up and the stereochemistry R. It would appear that there
was an error made in translating the stereoview to a flat
drawing.”
Angew. Chem. Int. Ed. 2010, 49, 5574 –5577
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5577