Unconventional Fragment Usage Enables a Concise Total Synthesis of (-)-Callyspongiolide

Article abstract of DOI:10.1021/jacs.7b13591

An asymmetric synthesis of (-)-callyspongiolide is described. The route builds the macrolide domain atypically from a disaccharide and a monoterpene without passing through a seco-acid. Chiral iridium catalysis selectively joins fragments. Subsequent degradation of an imbedded butyrolactone via perhemiketal fragmentation affords a stereo- and regio-defined homoallylic alcohol that is engaged directly in a carbonylative macrolactonization. Further elaboration of the polyunsaturated appendage provides the natural product in a particularly direct and flexible manner.

Full text of DOI:10.1021/jacs.7b13591

Subscriber access provided by University of Rochester | River Campus & Miner Libraries
Communication
Unconventional fragment usage enables a
concise total synthesis of (-)-callyspongiolide
Francesco Manoni, Corentin Rumo, Liubo Li, and Patrick G Harran
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b13591 • Publication Date (Web): 13 Jan 2018
Downloaded from http://pubs.acs.org on January 13, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 5
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Unconventional fragment usage enables a concise total synthesis of
(−)-callyspongiolide
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Francesco Manoni, Corentin Rumo, Liubo Li and Patrick G. Harran*
Department of Chemistry and Biochemistry, University of California−Los Angeles, 607 Charles E. Young Drive East, Los Angeles,
California 90095-1569, United States
Supporting Information Placeholder
hol and while installing the trans disubstituted alkene with both posi-
tional and stereochemical control. It was known that fused bicyclic
perhemiketals of type 4 would fragment to macrocyclic homoallylic
esters 5 when treated with FeSO4 in the presence of Cu(OAc)2.7 Both
olefin geometry and regiochemistry were controlled – a result rational-
ized in terms of a transient alkoxy radical fragmenting the C-C bond
and the incipient carbon radical engaging Cu(II) to form an organome-
tallic that undergoes syn co-planar hydride elimination.7 Reasoning
by analogy, structure 3 would result from similar fragmentation of
butyrolactone-derived perhemiketal 6. To our knowledge, there was no
precedent for the ring system in 6, providing an occasion to explore
new structures while approaching callyspongiolide synthesis in a decid-
edly unconventional manner.
ABSTRACT: An asymmetric synthesis of (−)-callyspongiolide is
described. The route builds the macrolide domain atypically from a
disaccharide and a monoterpene without passing through a seco-acid.
Chiral iridium catalysis selectively joins fragments. Subsequent degra-
dation of an imbedded butyrolactone via perhemiketal fragmentation
affords a stereo and regio-defined homoallylic alcohol that is engaged
directly in a carbonylative macrolactonization. Further elaboration of
the polyunsaturated appendage provides the natural product in a par-
ticularly direct and flexible manner.
Polyketide natural products are a large, diverse family of oxygenated
hydrocarbons that continue to be valuable in research and medicine. In
2014 Proksch et al. described a new member of this group that was
named callyspongiolide (1, from Callyspongia sp., Figure 1).1 The
callyspongiolide macrolactone is reminiscent of those in oscillariolide
and the phormidolides,2 but its phenethylated polyene appendage is
unique. The natural product was purified as a cytotoxin. Interestingly,
in the cell types examined (Jurkat and Ramos B lymphocytes), its activ-
ity was reported to be caspase independent. This implies a non-
apoptotic mechanism for cell death. Because apoptotic signaling is
suppressed in many cancers, and this often underpins drug resistance in
clinical settings,3 compounds that kill cancer cells by non-apoptotic
mechanisms are potentially valuable tools to explore new therapeutic
strategies. With this in mind we set out to prepare the molecule. As
these studies progressed we benefitted from reports by Ye and Ghosh
whose impressive syntheses established absolute stereochemistry of the
natural product.4 These works assembled seco-acid precursors to the
macrolide using contemporary methods for acyclic diastereocontrol.
We had chosen to pursue the problem differently and were exploring
alternate means to build the macrolactone. We planned for its relative
stereochemical features to be established on medium ring sized inter-
mediates serving as chain synthons,5 and to circumvent a seco-acid
precursor. This led to us preparing a polyacetate-derived macrolide
from a carbohydrate and a monoterpene.
Callyspongiolide was dissected into two fragments, separating its poly-
ene from the macrocycle at C15 (Figure 1). The C7 oxygenation pre-
sent on the latter as a urethane was thought accessible via ring-opening
of the pyran in dioxobicyclic precursor 2. It was unclear how to control
acrylate geometry by this method. However, because the ring system in
2 is found in other bioactive natural products,6 the risk was offset by
opportunities. To construct 2, the C2-C3 linkage was initially designat-
ed the ring-closure bond, such that the ansa-bridge might be installed
via reaction between a pyran derived oxonium ion and an enolized
form of the acetate in 3. This left us to consider how to build the
homoallylic acetate segment of 3, ideally without acylating a C13 alco-
Figure 1. Retrosynthetic analysis of the (−)-callyspongiolide macrolac-
tone.
Optically active butenolide 7 is commercially available and can be
prepared on scale from cellulose powder.8 Silylation of this material,
treatment with Me2CuLi and alkylation of the resultant conjugate addi-
tion product with 3-chloro-2-(iodomethyl)-1-propene (9)9 gave allylic
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 5
chloride 10.10 Both carbon-carbon bond forming steps in the sequence
occurred with diastereoselectivity greater than 95:5 (Scheme 1).
Compound 14 was ozonolyzed to generate a C3 aldehyde. When that
molecule was dissolved in MeCN containing catalytic amounts of
PPTS, the system smoothly rearranged to linked bis-heterocycle 15. It
was then possible to selectively exchange the furanyl methoxy group
with peroxide under mildly acidic conditions (Scheme 3).
1
2
3
4
Scheme 1. Elaboration of a cellulose-derived butenolide.
5
6
7
Scheme 3. Equilibrating heterocycles en route to a seco-acetate via
perhemiketal fragmentation.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We next employed a remarkable method for catalyzed asymmetric allyl
transfer to join 10 with citronellol derivative 11.11 Warming a stoichio-
metric mixture of 10 and 11 with 5 mol% of Krische’s iridium complex
1212 gave a single diastereomer of homoallylic alcohol 13.13 While there
was room for improving reaction rate and catalyst turnover, the out-
come was striking. The reaction forges a carbon-carbon bond between
a primary alcohol and an allylic chloride via net loss of HCl, and does
so with exquisite stereocontrol (dr >95:5).14 With 13 in hand, four of
five chiral centers in the target macrocycle had been established. What
remained was to saturate the methylidene to set C9 stereochemistry
(Scheme 2).
The resultant perhemiketal 16 was dissolved in MeOH and treated
with Cu(OAc)2 and FeSO4·7H2O at room temperature. This caused
rapid ring-opening to generate trans homoallylic acetate 17 in 20%
overall isolated yield from 14. Consistent with the mechanistic hypoth-
esis described earlier, both Fe(II) and Cu(II) were required, and sig-
nals indicative of alkene regioisomers were undetectable in 1H NMR
spectra of crude reaction mixtures.
Scheme 2. Iridium catalyzed diastereoselective fragment coupling.
Scheme 4. Pursuing a modified path to the target macrocycle.
This result validated a central tenant in our approach and positioned us
one bond away from target 2 (Figure 1). The intent was to close the
large ring via net displacement of the methoxy group with an acetate-
derived nucleophile. While we could form silyl ketene acetal derivatives
of 17, as well as various metal enolates, unfortunately none of these
species could be coaxed to cyclize in a desired manner. The relative
inertness of the methoxy pyran was the problem, yet preparing activat-
ed variants via additional functional group manipulations was not an
attractive solution. Eventually we found an alternative that did not
require added steps. Instead of rearranging the ozonolysis product of
14, it was reacted with Ph3PCH2 to afford 18. That material was treated
with aq. H2O2 / cat. PPTS in MeCN and the resultant perhemiketal
degraded with FeSO4·7H2O in the presence of Cu(OAc)2. The
While selective hydrogenation of the 1,1-disubstituted alkene (vs the
1,2-disubstituted olefin) was achieved readily, controlling facial selec-
tivity in the reduction was not. After much effort, it became clear we
needed to impart a facial bias in the substrate by reducing conforma-
tional freedom. Since the plan was to convert the lactone in 13 to a
perhemiketal of type 6 (Figure 1), we started down this path by adding
MeLi to 13, which provided a corresponding lactol in high yield. Desic-
cation of that material with dry PPTS then gave an intermediate cis-
fused dioxaperhydroazulene whose topology allowed for selective hy-
drogenation using Wilkinson’s complex as catalyst to afford a single
isomer of bicycle 14 (Scheme 2).
ACS Paragon Plus Environment

Page 3 of 5
Journal of the American Chemical Society
homoallylic acetate produced was saponified in situ to afford diol 19 in
NMR spectrum (500 MHz, CDCl3) of aliquots of photolysis reaction.
C) Time-course plot of integral ratio for B.
1
21% isolated overall yield from 14. We then used Dai’s cascade variant
of a Semmelhack cyclization15 to directly form the target bicycle 20.16
2
3
4
5
6
7
8
9
Interestingly, 21 was isolated exclusively as an E,E geometric isomer.
However, when a dilute acetone solution of that material was irradiated
in a Rayonet apparatus (300 nm bulb set) in a Pyrex vessel, a photo
equilibrium was established wherein the Z acrylate was favored (Figure
2). As depicted in Figure 2, the photolysis was efficient. Performing the
reaction in acetone, wherein substrate excitation likely occurred via
energy transfer from the solvent,17 prevented secondary photochemis-
try of the Z-enoate that led to deconjugation.18 This was a major com-
peting process in hexane solvent.
Scheme 5. Sequenced carbonylative macrolactonization and pyran
ring-opening.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To complete the synthesis, the geometric isomers of 21 were separated
by chromatography. This provided pure Z-21 in 55% yield along with
recovered E-21 which could be resubjected to the photolysis. Z-21 was
oxidized using the Swern protocol. The resulting aldehyde was homol-
ogated with trimethylsilyl triphenyl(prop-2-yn-1-ylidene)-5-
phosphane (22),19 and the product desilylated in situ to afford enyne
23 (Scheme 6).20 This material was then joined with vinyl iodide (R)-
24 (>99.5% ee)4e via Sonogashira coupling4b to afford (−)-
callyspongiolide (1).21 Spectroscopic data for 1 was identical to that
reported in the literature.1
Slow addition of 19 to a suspension of CuCl2 in dichloroethane con-
taining 10 mol% Pd(OAc)2 at 40 °C under an atmosphere of CO af-
forded macrolactone 20. The reaction produced a mixture of C3 epi-
mers (dr 6:4), but this was not consequential. Both diastereomers
rapidly ring-opened at -78 °C when treated with LDA. Derivatization of
the resultant alcohol with ClSO2NCO, diluting the reaction with 20%
aq. THF and subsequent addition of HF pyridine complex gave hy-
droxy urethane 21 in good overall yield (Scheme 5).
Scheme 6. Completion of the synthesis.
The longest linear sequence in the above synthesis is 18 steps. It builds
the target from three segments while exploiting unique intermediate
structures. The design leverages the chiral pool for raw materials and
contemporary catalysis to join fragments and form rings. The route
makes available callyspongiolide segments, isomers and derivatives on
scales useful for driving research into their molecular biology and
pharmacology. We have prepared ample amounts (>100 mg) of the
natural product 22 and have considerable intermediate supplies for these
experiments. In preliminary studies we have confirmed the activity of 1
in Jurkat B lymphoctyes (IC50 = 43 nM) and find that its ∆2,3-E isomer
is comparably potent (IC50 = 115 nM). In contrast, neither macrolac-
tone E-21 nor the unsaturated side chain (see SI) possess activity at 3
M in this cell line. Less than 3% of callyspongiolide toxicity towards
Jurkat cells is blocked by pre-treatment with 100 M Z-VAD-FMK, a
pan-caspase inhibitor. Notably, 1 strongly inhibits the growth of mu-
Figure 2. A) Photoisomerization: h (300 nm, Pyrex, Rayonet), ace-
tone (3 mM), rt, 20 min., 55% isolated Z-21 (90% brsm). B) Partial 1H
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 5
Acta 1996, 79, 51-60. (c) Zampella, A.; D’Auria, M. V.; Minale, L.; Debitus, C.;
Roussakis, C. J. Am. Chem. Soc. 1996, 118, 11085-11088.
(7) (a) Schreiber, S. L. J. Am. Chem. Soc. 1980, 102, 6163-6165. See also: (b)
Becker, J.; Ohloff, G. Helv. Chim. Acta 1971, 54, 2889-2895.
tant Saccharomyces cerevisiae lacking the PDR5 drug transporter. In
liquid culture, 1 also potently inhibits growth of a yeast ERG3 mutant.
This suggests there is a yeast receptor(s) for the compound and opens
the possibility of using genome wide mutagenesis to identify it. Exper-
iments to test this exciting prospect are ongoing as are attempts to
adapt the synthesis to preparations of biochemical probes and related
bioactive natural products.
1
2
3
4
5
6
7
8
(8) (a) Koseki, K.; Ebata, T.; Kawakami, H.; Matsushita, H.; Itoh, K.; Naoi, Y.
Method of preparing (S)-γ-hydroxymethyl-α, β-butenolide. US 4994585 A,
February 19, 1991. (b) Cao, F.; Schwartz, T. J.; McClelland, D. J.; Krishna, S.
H.; Dumesic, J. A.; Huber, G. W. Energy Environ. Sci. 2015, 8, 1808-1815.
(9) Li, C.-J.; Chan, T.-H. Organometallics 1991, 10, 2548-2549.
(10) A small amount (ca. 5%) of the iodo variant of 10 is produced during this
reaction. Because the iodide interferes with subsequent iridium catalysis (see
ref. 13), it was converted to 10 during work-up via Finkelstein exchange (LiCl,
acetone, rt).
(11) Taneja, S. C.; Sethi, V. K.; Andotra, S. S.; Koul, S.; Qazi, G. N. Synth.
Commun. 2005, 35, 2297-2303.
(12) (a) Hassan, A.; Townsend, I. A.; Krische, M. J. Chem. Comm. 2011, 47,
10028-10030. See also: (b) Shin, I.; Hong, S.; Krische, M. J. J. Am. Chem. Soc.
2016, 138, 14246-14249. (c) Feng, J.; Garza, V. J.; Krische, M. J. J. Am. Chem.
Soc. 2014, 136, 8911-8914.
ASSOCIATED CONTENT
Supporting Information
9
Supporting Information is available free of charge on the ACS Publica-
tions website.
Experimental procedures and spectral data (PDF).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Author
* harran@chem.ucla.edu
(13) Interestingly, both the bromo and iodo congeners of 10 gave different
results. The former was transformed into a mixture (ca. 1:1) of desired product
13 and symmetric dimer 25 (see SI). The latter appeared to poison 12. Intri-
guingly 25 was also formed when 10 was contaminated with only minor
amounts of its iodo variant (<5%).
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENT
Funding provided by the D. J. & J. M. Cram Endowment, the Human
Frontier Science Program (postdoctoral fellowship to FM,
LT001006/2015-C), the Chinese Scholarship Council (predoctoral
fellowship to LL, 201606205095) and NSF equipment grant (CHE-
1048804). CR was a visiting scholar on leave from the University of
Fribourg. We also thank the group of Professor Hosea Nelson (UCLA)
for the use of their CSP-HPLC instrument. PGH is indebted to Profes-
sor Jared Rutter (HHMI, University of Utah) for hosting his recent
stay in Salt Lake City, where Dr. Sarah Fogarty provided invaluable
guidance on cell culture, live cell microscopy and yeast assays.
(14) Alternatively, 13 could be synthesized from the iodo variant of 10 and the
aldehyde derived from 11 using stoichiometric indium metal (THF, rt, 85-
90%). However diastereoselectivity was poor (ca. 1:1) and separating the iso-
mers was impractical on scale.
(15) (a) Semmelhack, M. F.; Bodurow, C. J. Am. Chem. Soc. 1984, 106, 1496-
1498. (b) Semmelhack, M. F.; Zhang, N. J. Org. Chem. 1989, 54, 4483-4485.
(c) Semmelhack, M. F.; Kim, C.; Zhang, N.; Bodurow, C.; Sanner, M.; Dobler,
W.; Meier, M. Pure & Appl. Chem. 1990, 62, 2035-2040.
(16) (a) Bai, Y.; Davis, D. C.; Dai, M. Angew. Chem. Int. Ed. 2014, 53, 6519-
6522. (b) Bai, Y.; Davis, D. C.; Dai, M. J. Org. Chem. 2017, 82, 2319-2328.
(17) Barltrop, J. A.; Wills, J. Tetrahedron Lett. 1968, 48, 4987-4990.
(18) (a) Jorgenson, M. J. Chem. Commun. 1965, 137-138. (b) Jorgenson, M. J.
J. Am. Chem. Soc. 1969, 91, 198-200.
REFERENCES
(1) (a) Pham, C.-D.; Hartmann, R.; Böhler, P.; Stork, B.; Wesselborg, S.; Lin,
W.; Lai, D.; Proksch, P. Org. Lett. 2014, 16, 266-269.
(2) Oscillariolide: (a) Murakami, M.; Matsuda, H.; Makabe, K.; Yamaguchi, K.
Tetrahedron Lett. 1991, 32, 2391-2394. Phormidolides (b) Williamson, R. T.;
Boulanger, A.; Vulpanovici, A.; Roberts, M.; Gerwick, W. H. J. Org. Chem.
2002, 67, 7927-7936.
(3) (a) Pommier, Y.; Sordet, O.; Antony, S.; Hayward, R. L.; Kohn, K. W. On-
cogene 2004, 23, 2934–2949. (b) Chang, Y.-H.; Yang, Y.-L.; Chen, C.-M.;
Chen, H.-Y. Am. J. Cancer. Res. 2015, 5, 1844-1853.
(4) For the first total syntheses of (−)-callyspongiolide: (a) Zhou, J.; Gao, B.;
Xu, Z.; Ye, T. J. Am. Chem. Soc. 2016, 138, 6948-6951. (b) Ghosh, A. K.;
Kassekert, L. A.; Bungard, J. D. Org. Biomol. Chem. 2016, 14, 11357-11370. For
the synthesis of the antipode see: (c) Ghosh, A. K.; Kassekert, L. A. Org. Lett.
2016, 13, 3274-3277. For fragment studies see: (d) Athe, S.; Sharma, A.; Maru-
mudi, K.; Ghosh, S. Org. Biomol. Chem. 2016, 14, 6769-6779. (e) Matoušova,
E.; Koukal, P.; Formánek, B.; Kotora, M. Org. Lett. 2016, 18, 5656-5659.
(5) (a) Woodward, R. B.; Logusch, E.; Nambiar, K. P.; Sakan, K.; Ward, D. E.;
Au-Yeung, B.-W.; Balaram, P.; Browne, L. J.; Card, P. J.; Chen, C. H.; Chenev-
ert, R. B.; Fliri, A.; Frobel, K.; Gais, H.-J.; Garratt, D. G.; ̂ Hayakawa, K.; Heg-
gie, W.; Hesson, D. P.; Hoppe, D.; Hoppe, I.; Hyatt, J. A.; Ikeda, D.; Jacobi, P.
A.; Kim, K. S.; Kobuke, Y.; Kojima, K.; Krowicki, K.; Lee, V. J.; Leutert, T.;
Malchenko, S.; Martens, J.; Matthews, R. S.; Ong, B. S.; Press, J. B.; Rajan Babu,
T. V.; Rousseau, G.; Sauter, H. M.; Suzuki, M.; Tatsuta, K.; Tolbert, L. M.;
Truesdale, E. A.; Uchida, I.; Ueda, Y.; Uyehara, T.; Vasella, A. T.; Vladuchick,
W. C.; Wade, P. A.; Williams, R. M.; Wong, H. N.-C. J. Am. Chem. Soc. 1981,
103, 3210-3213. (b) Stork, G.; Nair, V. J. Am. Chem. Soc. 1979, 101, 1315–
1316.
(19) Gibson, A. W.; Humphrey, G. R.; Kennedy, D. J.; Wright, S. H. B. Synthe-
sis 1991, 414-416.
(20) The small amount of cis olefin produced in the enyne preparation allowed
us to isolate callyspongiolide isomer 26 following Sonogashira coupling of 23
with (R)-24. Brief photolysis of 26 under conditions identical to those used to
interconvert E- and Z-21 resulted in roughly a 1:1 mixture of 26 and 1 contain-
ing lesser amounts of isomeric byproducts.
(21) Following the protocols reported, the ∆2,3-E isomer of (−)-
callyspongiolide can be easily accessed from the E isomer of 21 (see SI).
(22) Synthetic (−)-callyspongiolide (1) was handled following the procedure
reported in: Hensgen, M. I.; Stump, B. Methods Mol. Biol. 2013, 1045, 133-143.
(6) (a) Wright, A. E.; Botelho, J. C.; Guzmán, E.; Harmody, D.; Linley, P.;
McCarthy, P. J.; Pitts, T. P.; Pomponi, S. A.; Reed, J. K. J. Nat. Prod. 2007, 70,
412-416. (b) D'Ambrosio, M.; Guerriero, A.; Debitus, C.; Pietra, F. Helv. Chim.
ACS Paragon Plus Environment

Page 5 of 5
Journal of the American Chemical Society
Insert Table of Contents artwork here
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
ACS Paragon Plus Environment