Total synthesis of ascospiroketal a through a agI-promoted cyclization cascade

Article abstract of DOI:10.1002/anie.201408905

The total synthesis of four candidate stereostructures for the marine octaketide ascospiroketal A have been achieved. These concise and highly stereocontrolled syntheses feature a unique Ag^I-promoted cyclization cascade involving an oxetanyl ketochlorohydrin to access the entire tricyclic core of the natural product in one step. These syntheses also establish the full stereochemistry for the ascospiroketal natural products.

Full text of DOI:10.1002/anie.201408905

Angewandte
Chemie
DOI: 10.1002/anie.201408905
Natural Product Synthesis
Total Synthesis of Ascospiroketal A Through a AgI-Promoted
Cyclization Cascade**
Stanley Chang, Soo Hur, and Robert Britton*
Abstract: The total synthesis of four candidate stereostructures
for the marine octaketide ascospiroketal A have been achieved.
These concise and highly stereocontrolled syntheses feature
a unique AgI-promoted cyclization cascade involving an
oxetanyl ketochlorohydrin to access the entire tricyclic core
of the natural product in one step. These syntheses also
establish the full stereochemistry for the ascospiroketal natural
products.
stereochemistry of the three remote stereocenters remained
undefined. Structurally, these compounds represent the most
complex members of a small family of tricyclic 5,5-spiroace-
tals[2] that includes the cephalosporolides,[3] penisporolides,[4]
and opaliferin,[5] for which several potentially useful biolog-
ical activities have been ascribed.[3–5] Herein we describe the
first total synthesis and full stereochemical assignment of
ascospiroketal A (1)[6] through a highly efficient AgI-pro-
moted cyclization cascade that assembles the entire tricyclic
core in one step from an acyclic precursor.
I
n 2007, Kçnig reported the isolation of ascospiroketals A (1)
and B (2) (Figure 1) as single spiroacetal epimers from the
marine-derived fungus Ascochyta salicorniae.[1] Following
extensive analysis of 1D and 2D NMR spectra, it was
proposed that 1 and 2 possess a rare octaketide tricyclic
We have recently reported that aldol adducts of a-
chloroaldehydes 3 (Scheme 1) undergo AgI-promoted cycli-
Figure 1. Naturally occurring octaketides ascospiroketals A and B.
core, in which the terminal ring is cyclized through an ether or
ester linkage and includes a quaternary stereogenic center at
C2 installed through geminal SAM methylation.[1] The
relative stereochemistry within the tricyclic core of the
ascospiroketals was established from analysis of NOESY
experiments and confirmed that these compounds are
anomeric spiroacetals. Unfortunately, no information could
be obtained regarding their absolute stereochemistry, the
configurational relationship between the tricyclic core and
side chain, or within the side chain itself. Thus, the relative
Scheme 1. A silver(I)-promoted spirocyclization strategy to access
ascospiroketal A.
zation to afford various spiroacetals 5,[7] including 5,5-
spiroacetals such as those embedded in the tricyclic core of
the ascospiroketals. This unique spirocyclization strategy
involves intramolecular alkylation of a hemiacetal by a AgI-
activated chloromethine (e.g., 4). In contemplating a total
synthesis of the ascospiroketals, we envisaged that ultimately
the side chain 6 (Scheme 1) would be appended to the fully
functionalized core 7 through a Sonogashira coupling.[8]
Importantly, this late-stage coupling would allow for the
rapid production of configurational isomers at the undefined
stereocenters C15, C2’, and C3’. In turn, a AgI-promoted
spirocyclization of the ketochlorohydrin 9 would provide the
spiroacetal 8 and the terminal ring would be accessed through
[*] S. Chang, S. Hur, Prof. R. Britton
Department of Chemistry, Simon Fraser University
Burnaby, British Columbia (Canada)
E-mail: rbritton@sfu.ca
[**] This work was supported by an NSERC Discovery Grant to R.B.,
a MSFHR Career Investigator Award to R.B., and an NSERC
Postgraduate Scholarship for S.C. The authors thank Milan
Bergeron-Brlek (SFU) for assistance with X-ray crystallographic
analysis.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201408905.
Angew. Chem. Int. Ed. 2015, 54, 211 –214
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
211

.
Angewandte
Communications
rearrangement of the hydroxy oxetane function (i.e., 8!7).[9]
Molecular models suggested that Lewis acid activation of the
oxetane function in 8 could involve a secondary coordination
to the central ring oxygen (Scheme 1). Importantly, this
bidentate chelate structure would enforce the desired sense of
diastereoselectivity on this critical rearrangement and secure
the correct configuration at the all-carbon quaternary center
C2. While without precedent, the potential for these two
distinct cyclization reactions to be promoted in tandem by
a AgI salt was a particularly appealing aspect of this route.
Building on our experiences in aldol reactions of a-chloro-
aldehydes,[10] we expected the ketochlorohydrin 9 to be
readily available from the union of suitably functionalized
aldol coupling partners 10 and 11.
In an attempt to address stereochemical uncertainties
regarding the ascospiroketal A side chain,[1] we first targeted
the four potential diastereomeric truncated side chains 16–19
(Scheme 2). Thus, readily available TBS-protected (2S,3R)-3-
Scheme 3. Synthesis of the oxetanyl ketochlorohydrin 9. a) Et2O,
ꢀ408C, 1.5 h, 80%; b) Ti(OiPr)4, (ꢀ)-DIPT, TBHP, CH2Cl2, ꢀ208C,
38 h, 40% (ee 98%); c) 1) PTSA, H2O/acetone, 658C, 2.5 h; 2) Et3N,
TMSCl, THF, 08C to RT, 91% over two steps; d) MsCl, Et3N, CH2Cl2,
08C, 1 h; e) KCN, DMF, 808C, 16 h, 76% over two steps; f) DIBAL,
CH2Cl2, ꢀ58C, 1 h; g) NCS, 26 (20 mol%), CH2Cl2, 58C, 24 h, 64%
over two steps; h) 23, LDA, then 10, THF, ꢀ788C, 46% (72% based
on recovered 23, d.r.=12:1); i) PTSA, H2O/acetone, 5 min, 83%.
(ꢀ)-DIPT=(ꢀ)-diisopropyl d-tartrate, TBHP=tert-butyl hydroperoxide,
PTSA=p-toluenesulfonic acid, TMSCl=chlorotrimethylsilane,
DMF=N,N-dimethylformamide, DIBAL=diisobutylaluminum hydride,
NCS=N-chlorosuccinimide, LDA=lithium diisopropylamide.
comparison of their spectral data to that reported for the
natural product.
Synthesis of the tricyclic core of ascospiroketal A was
initiated with the addition of vinyl lithium reagent 21 to the
known aldehyde 20 (Scheme 3).[14] Sharpless asymmetric
epoxidation[15] of 22 afforded the corresponding epoxide
(not shown) along with recovered alcohol (+)-22 in high
enantiomeric purity (98% ee) at 60% conversion. Removal of
the acetal protecting group and protection of the secondary
alcohol function in (+)-22 then yielded the methyl ketone 23.
The preparation of a-chloroaldehyde 10 involved a one-
carbon homologation of commercially available alcohol 24 by
displacement of the corresponding mesylate with cyanide and
subsequent reduction. The organocatalytic asymmetric a-
chlorination[10a] was explored using the conditions reported by
MacMillan,[16] Jørgensen,[17] and Christmann.[18] After some
experimentation with this unusual substrate, we found that
a combination of MacMillan’s catalyst 26[16] and NCS[18] gave
a-chloroaldehyde 10 in optimal enantiomeric purity (85%
ee). Finally, coupling of the lithium enolate derived from
methyl ketone 23 with the a-chloroaldehyde 10 provided the
aldol adduct 9 in good yield and excellent diastereoselectivity
(d.r. = 12:1).[10f]
Scheme 2. Synthesis of diastereomeric ascospiroketal A side chains
and comparison of their 1H NMR spectral data to 1. a) TMS-acetylene,
nBuLi, then 12, BF3·OEt2, Et2O, ꢀ788C, 69%; b) DIC, DMAP, CH2Cl2,
20 h, 96%; c) TBAF, THF, 83%. DIC=N,N’-diisopropylcarbodiimide,
DMAP=4-(dimethylamino)pyridine, TBAF=tetrabutylammonium fluo-
ride, TBS=tert-butyldimethylsilyl, TMS=trimethylsilyl.
hydroxy-2-methylbutyric acid 14[11] was coupled with the
homopropargyl alcohol 13 derived from TMS-acetylene
addition to (ꢀ)-propylene oxide (12).[12] Deprotection of the
resulting silyloxy ester 15 gave the hydroxy ester 16. Repeat-
ing this sequence of reactions separately with (+)-propylene
oxide and/or (2S,3S)-3-hydroxy-2-methylbutyric acid[13] (see
the Supporting Information (SI) for full details) afforded the
corresponding esters 17–19. With these materials in hand,
comparison of their 1H and 13C NMR spectral data (Scheme 2
and SI) with that reported for the equivalent portion of
ascospiroketal A[1] suggested that the natural products pos-
sess a (2’S*,3’R*) relative configuration as depicted for esters
16 and 19. Unfortunately, we were not able to confidently
assign the relative configuration at C15 using these model
compounds. Considering this uncertainty, the complete ste-
reochemical assignment of ascospiroketal A would ultimately
require the synthesis of four candidate stereostructures using
the side-chain precursors 16, 19, and ent-16, ent-19, and
Having established a concise, 6-step synthesis of keto-
chlorohydrin 9 we next explored the key spirocyclization
reaction. Using our optimized conditions[7] for the formation
of simple spiroacetals, we were delighted to find that over-
night reaction of 9 with AgOTf and Ag2O proceeded
smoothly to afford the anomeric spiroacetals 27 in good
combined yield. Pleasingly, the only by-products produced in
212
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 211 –214

Angewandte
Chemie
Scheme 5. Synthesis of candidate stereostructure 34 and ORTEP
representation of 32. a) DMP, NaHCO3, CH2Cl2, 1 h, 79%; b) NaClO2,
NaH2PO4·H2O, 2-methyl-2-butene, H2O:tBuOH, 45 min, 87%; c) NIS,
HFIP, 08C, 30 min, 67%; d) 16, PdCl2(PPh3)2, CuI, DIPA, THF, 14 h;
e) H2 (balloon), Lindlar catalyst, quinoline, EtOH, 46% over two steps.
DMP=Dess–Martin periodinane, NIS=N-iodosuccinimide,
HFIP=1,1,1,3,3,3-hexafluoro-2-propanol, DIPA=diisopropylamine.
Ellipsoids are given at 50% probability.
Scheme 4. AgI-promoted cascade cyclization of 9. a) Ag2O, AgBF4,
THF, 208C to 508C, 82% (29/30 ~1:1); b) ZnCl2, MgO, CH2Cl2, 6.5 h,
29/30 (2:1).
only 12 steps from the readily available aldehyde 20.
Repetition of the final two reactions using the stereochem-
ically unique side chains 19, ent-16, and ent-19 gave the
candidate stereostructures 35, 36, and 37, respectively
(Figure 2). Whereas the 1H NMR spectra derived from
these synthetic materials were similar, the resonances for
the diastereotopic protons at C14 proved to be diagnostic for
the C15-(S) and C15-(R) series. Notably, only the spectra
derived from the C15-(S) diastereomers 34 and 37 contained
H14a/H14b resonances characteristic of ascospiroketal A
(see truncated spectra, Figure 2).[25] Furthermore, whereas
any appreciable quantity were the tricycles 29 and 30. Further
optimization of the formation of these tricycles involved
a brief screen of AgI salts, whereupon a combination of
AgBF4 and Ag2O[7] was identified as optimal, delivering 29
and 30 in a combined isolated yield of 82%. As highlighted in
Scheme 4, the complete diastereoselectivity observed in the
opening of the prochiral oxetane may be attributed to
chelation of the central ring oxygen and that of the oxetane
to the AgI salt (see 27). This bidentate chelation[19] is not
possible in the pro-S transition structure 28. It is notable that
the AgI-promoted cyclization cascade effects the formation of
three heterocyclic rings and a spiroacetal center, and secures
the relative stereochemistry at the all-carbon quaternary
center required for ascospiroketal A. To improve the overall
efficiency of this reaction, the undesired spiroactetal 29 was
readily epimerized using Dudley’s conditions (ZnCl2,
MgO)[3a,20] to provide a mixture of the anomeric spiroacetals
29 and 30 and a means to recycle the former material.
Completion of the total synthesis of the candidate
stereostructures for ascospiroketal A is depicted in
Scheme 5. A two-step oxidation[21,22] of the primary alcohol
function in tricycle 30 provided the carboxylic acid 32 without
epimerization of the spiroacetal center. At this point, suitable
crystals were obtained for X-ray crystallographic analysis (see
ORTEP, Scheme 5), which confirmed our stereochemical
assignment of the tricyclic core. Moreover, the spectral data
(1H and 13C NMR) derived from 32 were in close agreement
with that reported for the same region of ascospiroketal A,
supporting the assigned stereochemistry for the natural
product. Following silicon to iodine exchange (32!33),[23]
Sonogashira coupling[8] of the side chain 16 with vinyl iodide
33 provided the full carbon skeleton of ascospiroketal A.
Lindlar reduction[24] of the resulting enyne completed the
total synthesis of candidate stereostructure 34, which required
1
the H NMR spectrum of 37 acquired at 500 MHz closely
matched that of the natural product, several subtle differences
in the chemical shift (Dd > 0.01 ppm) and/or shape of
resonances for H10, H13, H14, and H3’ existed between the
spectra of 34 and ascospiroketal A.[25] Thus, based on the
distinct 1H NMR spectra of the candidate stereostructures
34–37, the relative configuration for ascospiroketal A was
confidently assigned as that depicted for 37. The specific
rotation for 37 ([a]2D0 =+ 5 (c 0.20 in MeOH)) was also
consistent in sign with that reported for the natural product
([a]2D0 + 20 (c 0.45 in MeOH)),[1] confirming the absolute
stereochemistry of ascospiroketal A as shown for 37. Inter-
estingly, the structurally related octaketide cephalosporoli-
de H,[2c] also isolated from a marine-derived fungus, possesses
the same absolute stereochemistry within its tricyclic spiro-
acetal core.
In summary, exploiting a AgI-promoted cyclization cas-
cade, concise (14-step) syntheses of four candidate stereo-
structures of the naturally occurring polyketide ascospiroke-
tal A were realized. Comparison of their spectral data with
that reported for 1 allowed for the confident assignment of
the relative and absolute stereochemistry for the natural
product as (2R, 3R, 4R, 6R, 9S, 15S, 2’R, 3’S). Considering the
similarities in structure between the ascospiroketals A (1) and
Angew. Chem. Int. Ed. 2015, 54, 211 –214
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
213

.
Angewandte
Communications
Fernandes, M. B. Halle, Asian J. Org. Chem. 2013, 2, 593 – 599;
g) C. N. Kona, C. V. Ramana, Tetrahedron 2014, 70, 3653 – 3656.
[4] X. Li, I. Sattler, W. Lin, J. Antibiot. 2007, 60, 191 – 195.
[5] A. Grudniewska, S. Hayashi, M. Shimizu, M. Kato, M. Suenaga,
H. Imagawa, T. Ito, Y. Asakawa, S. Ban, T. Kumada, T.
Hashimoto, A. Umeyama, Org. Lett. 2014, 16, 4695 – 4697.
[6] For a synthetic approach to ascospiroketal B, see: J. W. Kim,
H. W. Lee, D.-H. Lee, Bull. Korean Chem. Soc. 2011, 32, 2877 –
2878.
[7] S. Chang, R. Britton, Org. Lett. 2012, 14, 5844 – 5847.
[8] K. Sonogashira, Y. Tohda, N. Hagihara, Tetrahedron Lett. 1975,
16, 4467 – 4470.
[9] R. N. Loy, E. N. Jacobsen, J. Am. Chem. Soc. 2009, 131, 2786 –
2787.
[10] a) B. Kang, J. Mowat, T. Pinter, R. Britton, Org. Lett. 2009, 11,
1717 – 1720; b) B. Kang, S. Chang, S. Dekker, R. Britton, Org.
Lett. 2010, 12, 1716 – 1719; c) J. Draper, R. Britton, Org. Lett.
2010, 12, 4034 – 4037; d) S. Halperin, B. Kang, R. Britton,
Synthesis 2011, 12, 1946 – 1953; e) M. T. Holmes, R. Britton,
Chem. Eur. J. 2013, 19, 12649 – 12652; f) R. Britton, B. Kang, Nat.
Prod. Rep. 2013, 30, 227 – 236.
[11] C. E. O’Connell, K. A. Salvato, Z. Meng, B. A. Littlefield, C. E.
Schwartz, Bioorg. Med. Chem. Lett. 1999, 9, 1541 – 1546.
[12] J. D. White, T. C. Somers, G. N. Reddy, J. Org. Chem. 1992, 57,
4991 – 4998.
[13] P. Renaud, D. Seebach, Helv. Chim. Acta 1986, 69, 1704 – 1710.
[14] A. P. Ramirez, A. M. Thomas, K. A. Woerpel, Org. Lett. 2009,
11, 507 – 510.
[15] a) T. Katsuki, K. B. Sharpless, J. Am. Chem. Soc. 1980, 102,
5974 – 5976; b) V. S. Martin, S. S. Woodard, T. Katsuki, Y.
Yamada, M. Ikeda, K. B. Sharpless, J. Am. Chem. Soc. 1981,
103, 6237 – 6240.
[16] M. Amatore, T. D. Beeson, S. P. Brown, D. W. C. MacMillan,
Angew. Chem. Int. Ed. 2009, 48, 5121 – 5124; Angew. Chem.
2009, 121, 5223 – 5226.
[17] N. Halland, A. Braunton, S. Bachmann, M. Marigo, K. A.
Jørgensen, J. Am. Chem. Soc. 2004, 126, 4790 – 4791.
[18] P. Winter, J. Swatschek, M. Willot, L. Radtke, T. Olbrisch, A.
Schꢀfer, M. Christmann, Chem. Commun. 2011, 47, 12200 –
12202.
[19] a) A. Yanagisawa, Y. Matsumoto, K. Asakawa, H. Yamamoto,
Tetrahedron 2002, 58, 8331 – 8339; b) A. M. Szpilman, E. M.
Carreira in Silver in Organic Chemistry (Ed.: M. Harmata),
Wiley, New Jersey, 2010, pp. 46 – 73.
[20] Treating compound 29 or a structurally related model system
with the following Lewis or Brønsted acids also resulted in the
production of ~ 1:1 mixtures of spiroacetals: pyridinium p-
toluenesulfonate, BF3·OEt2, camphorsulfonic acid, Sc(OTf)3,
InCl3.
Figure 2. Comparison of the 1H NMR spectra ([D6]acetone) of candi-
date stereostructures 34–37 to 1 and enlargement of the H14
resonances from the 1H NMR spectra of 1[25] and 34–37.
B (2), the relative and absolute stereochemistry of ascospiro-
ketal B should also be revised accordingly.
Received: September 8, 2014
Published online: November 6, 2014
Keywords: natural products · polyketides ·
.
stereochemical assignment · tandem cyclization · total synthesis
[21] D. B. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4155 – 4156.
[22] B. S. Bal, W. E. Childers, Jr., H. W. Pinnick, Tetrahedron 1981,
37, 2091 – 2096.
[1] S. F. Seibert, A. Krick, E. Eguereva, S. Kehraus, G. M. Kçnig,
Org. Lett. 2007, 9, 239 – 242.
[23] a) D. P. Stamos, A. G. Taylor, Y. Kishi, Tetrahedron Lett. 1996,
37, 8647 – 8650; b) E. A. Ilardi, C. E. Stivala, A. Zakarian, Org.
Lett. 2008, 10, 1727 – 1730.
[2] a) M. J. Ackland, J. R. Hanson, P. B. Hitchcock, J. Chem. Soc.
Perkin Trans. 1 1985, 843 – 847; b) V. Rukachaisirikul, S. Pramjit,
C. Pakawatchai, M. Isaka, S. Supothina, J. Nat. Prod. 2004, 67,
1953 – 1955; c) X. Li, Y. Yao, Y. Zheng, I. Sattler, W. Lin, Arch.
Pharmacal Res. 2007, 30, 812 – 815.
[3] For recent cephalosporolide syntheses, see: a) S. F. Tlais, G. B.
Dudley, Org. Lett. 2010, 12, 4698 – 4701; b) M. A. Brimble, O. C.
Finch, A. M. Heapy, J. D. Fraser, D. P. Furkert, P. D. O’Connor,
Tetrahedron 2011, 67, 995 – 1001; c) S. F. Tlais, G. B. Dudley,
Beilstein J. Org. Chem. 2011, 7, 570 – 577; d) S. F. Tlais, G. B.
Dudley, Beilstein J. Org. Chem. 2012, 8, 1287 – 1292; e) L. Song,
Y. Liu, R. Tong, Org. Lett. 2013, 15, 5850 – 5853; f) R. A.
[24] For Lindlar reduction of 1,3-enynes, see : a) K. C. Nicolaou, S. E.
Webber, J. Am. Chem. Soc. 1984, 106, 5734 – 5736; b) G. A.
Molander, F. Dehmel, J. Am. Chem. Soc. 2004, 126, 10313 –
10318; c) M. Bock, R. Dehn, A. Kirschning, Angew. Chem. Int.
Ed. 2008, 47, 9134 – 9137; Angew. Chem. 2008, 120, 9274 – 9277.
[25] The authors thank Professor Gabrielle Kçnig for providing
1
a data file of the 500 MHz H NMR spectrum of ascospiroke-
tal A (500 MHz, [D6]acetone).
214
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 211 –214