Two Ene-Yne Metathesis Approaches to the Total Synthesis of Amphidinolide P

Article abstract of DOI:10.1021/acs.orglett.5b01601

The total synthesis of amphidinolide P was achieved through two different ene-yne metathesis approaches. In each approach, the metathesis step was performed at late stages in the synthesis with all other functionality present. By forging two successful pa

Full text of DOI:10.1021/acs.orglett.5b01601

Letter
pubs.acs.org/OrgLett
Two Ene−Yne Metathesis Approaches to the Total Synthesis of
Amphidinolide P
Edgars Jecs and Steven T. Diver*
Department of Chemistry, University at Buffalo, The State University of New York, Amherst, New York 14260, United States
*
S Supporting Information
ABSTRACT: The total synthesis of amphidinolide P was
achieved through two different ene−yne metathesis ap-
proaches. In each approach, the metathesis step was performed
at late stages in the synthesis with all other functionality
present. By forging two successful pathways to the synthesis of
1
, some of the strengths and weaknesses of metathesis-
intensive synthetic strategies were identified.
mphidinolide P (1) is a cytotoxic macrolide natural
product found in the marine dinoflagellate Amphidinium
sp. It shows mild cytotoxicity against the murine leukemia
L1210 cancer cell line and epidermoid carcinoma KB cells with
IC ’s of 1.6 and 5.8 μg/mL, respectively. The density of π
featuring late stage metathesis reactions: an unprecedented low
temperature cross ene−yne metathesis and a ring-closing
metathesis pathway.
4
A
1
Late stage ene−yne metathesis is rare in complex molecule
synthesis. Catalytic bond coupling can be problematic in
advanced stages of total synthesis due to a high density of
multiple functional groups. Specifically, in ene−yne metathesis
of terminal alkynes, alkyne oligomerization is a problem that
makes ene−yne metathesis more challenging than widely used
ene−ene RCM reactions. Though rare, ene−yne metathesis
approaches have been used in the early stages of the synthesis
50
bonds, oxygen functionality, and seven stereogenic centers
make 1 a challenging synthetic target. To date, three total
2
2g
syntheses and one formal synthesis have been reported but
none use an ene−yne metathesis approach. Due to the
presence of the 1,3-diene, we envisioned that 1 could be
accessed by ruthenium carbene-catalyzed ene−yne metathesis
at the last or penultimate step, by either an intramolecular, ring-
closing (RCM) enyne metathesis or an intermolecular (cross)
5,6
of two related amphidinolides, E and K.
Additionally,
functional groups such as alcohols may trigger catalyst
decomposition. We were inspired by two syntheses that feature
3
ene−yne metathesis (EYM) (Scheme 1). In this letter, we
̈
successful late stage ene−yne metathesis. Furstner et al. used an
describe the total synthesis of 1 by two different approaches
ethylene−alkyne cross metathesis as the last step to make
7
amphidinolide V. Second, Krische et al. employed a ring-
Scheme 1. Retrosynthetic Disconnections of Amphidinolide
P Using Ene−Yne Metathesis
closing diene−alkene metathesis en route to 6-deoxyery-
8
thronolide B. These syntheses were successful; however, if a
metathesis approach fails at a late stage, the entire synthetic
9
plan may need to be abandoned. A better understanding of
metathesis approaches that work (or those that are less
successful) at late stages would help inform future synthesis
planning using ene−yne metathesis. For a successful ene−yne
metathesis, the alkene must be reactive enough to form the
1
0,11
metal carbene needed in catalysis
and alkyne polymer-
ization must be effectively suppressed.
The two metathesis-oriented approaches are outlined in
Scheme 1. Path a disconnects the macrocycle into seco ester 2,
which in turn could be garnered by a stereoselective cross
(
intermolecular) ene−yne metathesis between alkyne 3 and
dienol 4. Alkyne 3 features a terminal alkyne lacking propargylic
substitution, a less reactive substrate for EYM. The alkene cross
partner 4 bears two different alkenes, which demands
chemoselectivity in the metathesis step. In path b, a ring-
Received: June 1, 2015
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.5b01601
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
closing enyne metathesis (RCM) of 5 would access 1. The
syntheses of 5 and 3 are each derived from common aldehyde
intermediate 6 which possesses the C −C segment of the
Table 1. Optimization Studies for the Ene−Yne Metathesis
Cross-Coupling
3
12
amphidinolide P backbone.
The synthesis commenced with the preparation of the
common aldehyde intermediate (Scheme 2). Epoxyalcohol 7
a
a
entry
equiv of 4
temp/°C
time/h
yield/%
Scheme 2. Synthesis of Common Aldehyde Intermediate 6
b
c
1
2
3
1.5
1.5
0
−20
−20
8
5
80(E) + 10(Z)
53(E) + 13(Z)
78(E) + 22(Z)
d
3
18
a
All reactions were quenched by addition of KO CCH NC in MeOH.
Yield was determined by H NMR integration vs internal standard
mesitylene). 60% isolated yield; 3% alkyne was recovered. 33%
alkyne was recovered. Excess 4 was present in the crude reaction
mixture; no attempts were made to isolate unreacted 4.
2
2
1
b
c
(
d
17
balances and gave dienol products that were colorless.
Significantly, the good mass balance indicated that alkyne
polymerization was successfully controlled at the low reaction
temperature. Low temperatures are not used in metathesis
possibly due to perceptions of low reactivity. However, the
Hoveyda catalyst reacts by an associative mechanism, unique
among most metathesis catalysts, and concentration is more
important. In entries 2−3, alkyne concentration was doubled to
was reported previously in our model studies and is available
4
,12
from commercial starting materials in six steps. Oxidation of
with the Dess-Martin periodinane gave the aldehyde 8
7
18
needed for the Hosomi−Sakurai reaction. In the event, allyl
silane 9 underwent transmetalation with SnCl giving an
4
intermediate allyl stannane which coupled with 8 in good yield,
albeit with low diastereoselectivity. High stereoselectivity was
conveniently obtained through a sequential oxidation and
chelation-controlled reduction with Zn(BH ) . In this way,
0
.12 M allowing lower temperatures to be used. With a longer
reaction time, complete conversion was obtained (entry 3). At
these catalyst loadings, a ruthenium removal method was
critical. We found that treatment with KO CCH NC was
19
4
2
2
2
intermediate 10 was obtained in a >99:1 anti/syn ratio. Next, a
series of steps were needed to protect the secondary alcohol at
C7 and to oxidize the C3 carbon, providing common aldehyde
intermediate 6 in 87% overall yield.
highly effective for the quench and removal of Ru emanating
20
from the metathesis catalyst.
The key intermolecular ene−yne metathesis cross-coupling
was successful and led to completion of the total synthesis
(Scheme 3). First, common aldehyde 6 was converted to the
required β-ketoester 3 by a three-step sequence. Next, the cross
EYM was conducted. Low reaction temperature was found to
limit decomposition of 3 and led to a good yield of 1,3-diene 2
obtained as a 4.5:1 E/Z mixture. Addition of Hov2 in
installments or by slow infusion did not improve the product
yield or increase the stereoselectivity. Completion of the
synthesis followed Williams’ endgame to give (−)-1. Synthetic
amphidinolide P was identical in all respects to that reported,
except specific rotation. Our synthetic 1 was found to be pure
enantiomer as analyzed by chiral HPLC vs Williams’ original
material, (+)-1. In addition, intermediates 2 and 15 were
identical to Williams’ intermediates, and each matched all
spectroscopic properties, including the magnitude of their
specific rotation. These data together show that synthesis of
The dienol coupling partner was prepared using Leighton’s
13
diastereoselective allyl silane addition to acrolein eq 1. To the
best of our knowledge, acrolein has not been previously used in
this allylation. The diastereomer ratio of 4 was assigned based
1
on integration of the diastereomers in the H NMR spectrum
of the crude mixture: anti-4 had a vicinal J value of 8.5 Hz,
1
3b
similar to the reported J values for anti-stereodyads.
Assignment of configuration of the secondary alcohol was
made by comparison of the R- and S-Mosher esters, establishing
the R-configuration. Dienol 4 is required for the intermolecular
pathway. Esterification of 4 gave acetate 12, a starting material
needed for the second, ring-closing metathesis pathway.
2b,c
(−)-1 was achieved and support the earlier contention that
Williams originally synthesized (+) amphidinolide P. From the
common aldehyde 6, the yield is 21% (10% overall starting
from 7).
At this stage, the chemo- and stereoselective EYM was
investigated with a model alkyne. Model alkyne 13 bears an
epoxide in close proximity to the alkyne and lacks propargylic
substitution (eq 2, Table 1). Terminal, unhindered alkynes are
difficult substrates for EYM because they may undergo
competitive alkyne polymerization and can trigger catalyst
In the second approach, attempted ring-closing EYM
demonstrated how sensitive EYM is to the enyne substrate
(Scheme 4). The key enynes were formed by an aldol reaction
between the ester enolate of 12 and common aldehyde 6,
followed by oxidation. Using a different ending sequence, 16A
and 16B were each prepared. The direct EYM of 16B (20 mol
% Hov2 at 110 °C for 5 h) failed to give any ring-closing
product, 15. Decomposition was observed, and none of 16B
could be recovered. To control and suppress alkyne polymer-
ization, a TMS group was retained on the alkyne, to give 16A.
Attempted RCM under similar conditions gave only 30%
recovered 16A and additional byproducts, but no ring closure.
14
decomposition particularly with the Hoveyda catalyst, Hov2.
The alkene partner requires a free allylic alcohol for favorable
15
reactivity with Hov2, but at the same time can potentially
16
decompose the ruthenium carbenes into Ru hydrides. Using
.5 equiv of 4, an excellent yield of diene 14 was obtained
entry 1). Low temperatures produced the highest mass
1
(
B
DOI: 10.1021/acs.orglett.5b01601
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 3. Key Ene−Yne Metathesis and Completion of the Synthesis (Intermolecular Route)
Scheme 4. Preparation of Enyne and Unsuccessful Ring-
Closing Enyne Metathesis
constraint in the flexible backbone might further improve the
metathesis cyclization step.
An improved RCM route to 1 was achieved by converting
the β-keto ester into a silyl enol ether prior to the RCM step
(
Scheme 5). Locking the enol form would introduce an
element of rigidity in the backbone joining the alkenes.
Pentaene 17 was enolized to afford silyl enol ether 18, assigned
1
as the E-isomer on the basis of H NMR studies. Ring-closing
metathesis of 18 gave complete conversion to RCM products,
isolated as a mixture of 42% intermediate 19 along with 37%
1
5. The latter product provides evidence of some silyl enol
ether deprotection occurring under the thermal conditions. The
RCM products were separated for characterization, but for
completion of the synthesis, the mixture was convergently
deprotected with HF−pyridine to give 1 in 67% yield (two
steps). The material obtained in this way also matched the
spectroscopic properties with those of known amphidinolide.
From the common aldehyde 6, the yield was 50% (24% overall
starting from 7).
It was concluded that the terminal alkene in 16 is not
sufficiently reactive to initiate ring-closing enyne metathesis.
These findings agree with previous observations that showed
that ester- or silyl-protecting groups of secondary allylic
2
1
alcohols abrogated intermolecular metathesis reactivity.
Without a sufficiently reactive alkene,
3
,10,11
In conclusion, two different metathesis routes were used to
complete the total synthesis of amphidinolide P. In the
intermolecular ene−yne metathesis, low temperature con-
ditions were critical for a successful intermolecular EYM. In
the second pathway, direct ring-closing enyne metathesis was
not successful. To suppress the suspected alkyne polymer-
ization pathway and improve the initiation profile, the alkyne
was converted to a 2-substituted-1,3-butadiene. Ring-closing
alkene metathesis of this substrate produced the macrolide ring.
The RCM also benefitted from rigidifying the backbone. By
way of comparison, the intramolecular (RCM) pathway was
one step longer than the cross EYM approach, but the RCM
proceeded with the highest chemical yield and highest E-
selectivity. These studies highlight the relative strengths of
inter- and intramolecular ene−yne metathesis at advanced
stages of total synthesis.
the required metal
carbene fails to form, giving no reaction or resulting in
competitive alkyne oligomerization. If the alkyne reacts first, an
oligomerization pathway results.
A successful ring-closing sequence was realized by first
converting the terminal alkyne into a 1,3-diene prior to the
RCM step. Butadiene 17 was synthesized in quantitative yield
by ethylene−alkyne metathesis of 16B (Scheme 5). Ring-
Scheme 5. Ring-Closing Metathesis Approach to
Amphidinolide P (Intramolecular Approach)
ASSOCIATED CONTENT
Supporting Information
■
*
S
Experimental procedures and NMR spectra. The Supporting
Information is available free of charge on the ACS Publications
website at DOI: 10.1021/acs.orglett.5b01601.
AUTHOR INFORMATION
Corresponding Author
■
*
E-mail: diver@buffalo.edu.
Notes
The authors declare no competing financial interest.
closing metathesis of the TBS ether 17 gave a good yield of the
E-isomer (57% yield 15, E-only). Comparison of these data
with the unsuccessful direct RCM of enynes 16 suggested that
the 1,3-butadiene serves as the initiating alkene. This approach
alters the ring-closing reaction from an enyne metathesis to that
ACKNOWLEDGMENTS
■
We thank Professor David R. Williams (Chemistry, Indiana
University) for helpful discussions and for supplying a sample
of authentic (+)-amphidinolide P. We also thank Dr. Richard
3c
of a 1,3-diene-alkene metathesis. We considered that ring
C
DOI: 10.1021/acs.orglett.5b01601
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Pederson (Materia) for a sample of the Hoveyda−Grubbs
catalyst. This work was supported by the National Science
Foundation through Grants CHE-1012839 and CHE-1300702.
(13) (a) Kubota, K.; Leighton, J. L. Angew. Chem., Int. Ed. 2003, 42,
46−948. (b) Coleman, R. S.; Gurrala, S. R.; Mitra, S.; Raao, A. J. Org.
9
Chem. 2005, 70, 8932−8941.
(14) Diver, S. T.; Kulkarni, A. A.; Clark, D. A.; Peppers, B. P. J. Am.
Chem. Soc. 2007, 129, 5832−5833.
REFERENCES
(15) Clark, J. R.; Diver, S. T. Org. Lett. 2011, 13, 2896−2899.
■
(
16) Dinger, M. B.; Mol, J. C. Organometallics 2003, 22, 1089−1095.
(
1) Isolation: (a) Ishibashi, M.; Takahashi, M.; Kobayashi, J. i. J. Org.
(17) For instance, heating at 60 °C gave dienol products that were
Chem. 1995, 60, 6062−6066. Reviews: (b) Kobayashi, J. i.; Tsuda, M.
Natural Product Reports 2004, 21, 77−93. (c) Kobayashi, J. i.; Kubota,
T. J. Nat. Prod. 2007, 70, 451−460. (d) Kobayashi, J. i. J. Antibiot.
008, 61, 271−284. (e) Fu
2) (a) Williams, D. R.; Myers, B. J.; Mi, L. Org. Lett. 2000, 2, 945−
48. (b) Williams, D. R.; Myers, B. J.; Mi, L. Org. Lett. 2013, 15, 2070.
c) Trost, B. M.; Papillon, J. P. N. J. Am. Chem. Soc. 2004, 126,
3618−13619. (d) Trost, B. M.; Papillon, J. P. N.; Nussbaumer, T. J.
Am. Chem. Soc. 2005, 127, 17921−17937. (e) Hwang, M.-h.; Han, S.-
J.; Lee, D.-H. Org. Lett. 2013, 15, 3318−3321. (f) Williams, D. R.;
Myers, B. J.; Mi, L.; Binder, R. J. J. Org. Chem. 2013, 78, 4762−4778.
g) Chakraborty, T. K.; Das, S. Tetrahedron Lett. 2001, 42, 3387−
390.
3) Recent reviews: (a) Li, J.; Lee, D. In Handbook of Metathesis, 2nd
dark colored, presumably due to greater catalyst decomposition at
higher temperatures.
(18) (a) Kuhn, K. M.; Bourg, J.-B.; Chung, C. K.; Virgil, S. C.;
2
(
9
(
̈
rstner, A. Isr. J. Chem. 2011, 51, 329−345.
Grubbs, R. H. J. Am. Chem. Soc. 2009, 131, 5313−5320. (b) Vorfalt,
T.; Wannowius, K.-J.; Plenio, H. Angew. Chem., Int. Ed. 2010, 49,
5
(
533−5536.
19) Decreased loading of Hov2 resulted in lower product yields.
1
(20) Galan, B. R.; Kalbarczyk, K. P.; Szczepankiewicz, S.; Keister, J.
B.; Diver, S. T. Org. Lett. 2007, 9, 1203−1206.
(21) (a) Michaelis, S.; Blechert, S. Org. Lett. 2005, 7, 5513.
Accelerating effect of allylic alcohols: (b) Hoye, T. R.; Zhao, H. Org.
Lett. 1999, 1, 1123. (c) Imahori, T.; Ojima, H.; Yoshimura, Y.;
Takahata, H. Chem.Eur. J. 2008, 14, 10762−10771.
(
3
(
ed.; Grubbs, R. H., O’Leary, D. J., Eds.; Wiley-VCH: Weinheim, 2015;
Vol. 2, pp 381−444. (b) Diver, S. T., Clark, J. R. In Comprehensive
NOTE ADDED AFTER ASAP PUBLICATION
Scheme 4 was corrected on June 29, 2015.
■
̈
Organic Synthesis, 2nd ed.; Furstner, A., Ed.; Pergamon: London, 2014;
Vol. 5, pp 1302−1356. (c) Diver, S. T. Science of Synthesis 2009, 46.3,
9
(
(
7−146.
4) Jecs, E.; Diver, S. T. Tetrahedron Lett. 2014, 55, 4933−4937.
5) (a) Kim, C. H.; An, H. J.; Shin, W. K.; Yu, W.; Woo, S. K.; Jung,
S. K.; Lee, E. Angew. Chem., Int. Ed. 2006, 45, 8019−8021. (b) Ko, H.
M.; Lee, C. W.; Kwon, H. K.; Chung, H. S.; Choi, S. Y.; Chung, Y. K.;
Lee, E. Angew. Chem., Int. Ed. 2009, 48, 2364−2366.
(
6) Select examples of ring-closing EYM used in total synthesis
small rings): (a) Kinoshita, A.; Mori, M. J. Org. Chem. 1996, 61,
356−8357. (b) Aggarwal, V. K.; Astle, C. J.; Rogers-Evans, M. Org.
(
8
Lett. 2004, 6, 1469−1471. (c) Brenneman, J. B.; Martin, S. F. Org. Lett.
2
004, 6, 1329−1331. (d) Evans, M. A.; Morken, J. P. Org. Lett. 2005,
7
, 3371−3373. (e) Kummer, D. A.; Brenneman, J. B.; Martin, S. F.
Org. Lett. 2005, 7, 4621−4623. (f) Movassaghi, M.; Piizzi, G.; Siegel,
D. S.; Piersanti, G. Angew. Chem., Int. Ed. 2006, 45, 5859−5863.
(
g) Paquette, L. A.; Lai, K. W. Org. Lett. 2008, 10, 2111−2113. Study
on ring closure mode: (h) Hansen, E. C.; Lee, D. J. Am. Chem. Soc.
2
004, 126, 15074−15080. Macrocyclization ring-closing EYM in total
synthesis: (i) Layton, M. E.; Morales, C. A.; Shair, M. D. J. Am. Chem.
Soc. 2002, 124, 773−775.
(
7) (a) Fu
007, 46, 5545−5548. (b) Fu
Takahashi, Y.; Kubota, T.; Kobayashi, J. i. Chem.Eur. J. 2009, 15,
011−4029.
8) Gao, X.; Woo, S. K.; Krische, M. J. J. Am. Chem. Soc. 2013, 135,
223−4226.
9) For instance, despite attempts to get an enyne metathesis to work
̈
rstner, A.; Larionov, O.; Flu
̈
gge, S. Angew. Chem., Int. Ed.
2
̈
rstner, A.; Flu
̈
gge, S.; Larionov, O.;
4
(
4
(
successfully in their amphidinolide N synthesis, ultimately the authors
had to redesign their route; see: Nicolaou, K. C.; Brenzovich, W. E.;
Bulger, P. G.; Francis, T. M. Org. Biomol. Chem. 2006, 4, 2119.
(
10) Grubbs classifies these as types I−IV, with types I and II alkenes
reactive enough to initiate and give alkene metathesis. See:
a) Chatterjee, A. K.; Choi, T.-L.; Sanders, D. P.; Grubbs, R. H. J.
(
Am. Chem. Soc. 2003, 125, 11360−11370. (b) O’Leary, D. J.; O’Neil,
G. W. In Handbook of Metathesis, 2nd ed., Vol. 2; Grubbs, R. H.,
O’Leary, D. J., Eds.; Wiley-VCH: Weinheim, 2015.
(
11) (a) Galan, B. R.; Giessert, A. J.; Keister, J. B.; Diver, S. T. J. Am.
Chem. Soc. 2005, 127, 5762−5763. DFT studies: (b) Lippstreu, J. J.;
Straub, B. F. J. Am. Chem. Soc. 2005, 127, 7444−7457. (c) Nunez-
Zarur, F.; Solans-Monfort, X.; Rodriguez-Santiago, L.; Pleixats, R.;
Sodupe, M. Chem.Eur. J. 2011, 17, 7506−7520.
(
12) Prepared in six steps from methyl crotonate and TMS acetylene
in 21% overall yield.
D
DOI: 10.1021/acs.orglett.5b01601
Org. Lett. XXXX, XXX, XXX−XXX