Short Approach toward the Nonracemic A,B,E Tricyclic Core of Calyciphylline B-Type Alkaloids

Article abstract of DOI:10.1021/acs.orglett.7b01566

A suitably functionalized tricyclic adduct containing the common A,B,E rings found in calyciphylline B-type alkaloids was obtained in nine linear steps. The key transformation features an efficient one-pot sequence of intramolecular Vilsmeier-Haack cycliz

Full text of DOI:10.1021/acs.orglett.7b01566

Letter
pubs.acs.org/OrgLett
Short Approach toward the Nonracemic A,B,E Tricyclic Core of
Calyciphylline B‑Type Alkaloids
Patrick Boissarie and Guillaume Belanger*
́
́ ́ ́ ́
Departement de Chimie, Universite de Sherbrooke, 2500 boulevard de l’Universite, Sherbrooke, Quebec J1K 2R1, Canada
S
* Supporting Information
ABSTRACT: A suitably functionalized tricyclic adduct containing the
common A,B,E rings found in calyciphylline B-type alkaloids was
obtained in nine linear steps. The key transformation features an efficient
one-pot sequence of intramolecular Vilsmeier−Haack cyclization and
azomethine ylide 1,3-dipolar cycloaddition in which three cycles, three
new carbon−carbon bonds, and three stereocenters are formed, one
being fully substituted. This work also demonstrates the first use of a
chiral, nonracemic cyclic enol ether as an internal carbon nucleophile.
Scheme 1. Retrosynthetic Analysis toward the Tricyclic Core
of Calyciphylline B-Type Alkaloids
ith over 300 structurally distinct structures identified to
Wdate, Daphniphyllum form one of the most diverse and
unique family of polycyclic alkaloids. Their complex architec-
tures along with their interesting biological activities have
attracted attention for many years now, resulting in the
development of several elegant and inventive synthetic
approaches1 triggered in great part by the pioneering work of
Heathcock.2 In 2003, Morita and Kobayashi isolated calyciphyl-
line B from the leaves of D. calycinum, presenting a unique
hexacyclic core (Figure 1).3
This key step would allow the formation of three of the six
rings of calyciphylline B-type general structure in a single
transformation. To achieve this sequence in a stereocontrolled
fashion, our strategy relies on the use of a chiral cyclic silyl enol
ether, namely, a 1,3-dioxa-2-silacyclohexene. The latter would act
as the nucleophile in the Vilsmeier−Haack cyclization. The
single stereocenter would force the adoption of the only possible
reactive chair conformation 3b, locking the dipolarophile chain
in the axial position. This specific arrangement would lead to a
single diastereomer at C1, C2, and C8 during the formation of
Figure 1. Representative calyciphylline B-type alkaloids.
Several other natural products possessing the same skeleton
have since been identified. Surprisingly, synthetic efforts toward
members of this family remained essentially nonexistent until
recent work from the Hanessian group,4 ultimately leading to the
synthesis of isodaphlongamine H.5
We envisaged preparing the tricylic central core 1 of
calyciphylline B-type alkaloids using a synthetic strategy
developed by our group,6 namely, a sequence of intramolecular
Vilsmeier−Haack reaction and azomethine ylide cycloaddition
(Scheme 1).
Received: May 23, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b01566
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
the A, B, and E rings. Azomethine ylide 2 would be generated by
desilylation (X = TMS) or deprotonation (X = electron-
withdrawing group, EWG) of the corresponding iminium ion,
itself obtained from the Vilsmeier−Haack cyclization initiated by
activation of formamide 3. The cyclic silyl enol ether should be
obtained from β-hydroxyaldehyde 4. However, special attention
will be needed when setting reaction conditions to prevent any
potential elimination of the hydroxyl group or retro-aldol
reaction. Aldehyde 4 would be derived from an Evans aldol
reaction between oxazolidinone 5 and aldehyde 6.
The synthesis of 1,3-dioxa-2-silacyclohexene is poorly
documented,7 and its use as a carbon-based nucleophile has
not been investigated to date. To validate such a nucleophile in
the Vilsmeier−Haack cyclization, we synthesized a model
substrate lacking the dipolarophile chain that was replaced with
a simple methyl group. The synthesis of model substrate 13a
started with alkylation of 3-aminopropanol (7) with iodomethyl-
trimethylsilane and then protection as a carbamate using ethyl
chloroformate (Scheme 2).
molecular reduction, we turned our attention toward Red-Al.
Gratifyingly, aldehyde 13a was obtained in high yield.12
At this point, we were ready to install the 1,3-dioxa-2-
silacyclohexene as the chiral nucleophile needed in the first step
of the key reaction sequence, namely, the Vilsmeier−Haack
cyclization. We chose dialkylsilyl bis(trifluoromethanesulfonate)
reagents as they are commercially available and possess a good
Lewis acidity. Indeed, after silylation of the free alcohol 13a, we
expected the resulting silicon monotriflate to act as an internal
Lewis acid to promote enolization of the aldehyde in the desired
geometry. Although di-tert-butylsilyl ditriflate led to no
conversion (Table 1, entry 1), the less bulky diisopropyl
Table 1. Preparation of the 1,3-Dioxa-2-silacyclohexene
entry
R1
R2
base
DIPEA
temp (°C)
yield (%)
Scheme 2. Synthesis of Model Substrate 13
1
Et
Et
Et
Et
Et
Et
Et
Et
Et
all
all
t-Bu
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
rt
0
17
51
25
0
2
DIPEA
DIPEA
DIPEA
none
rt
3
−40
−78
−40
−40
−40
−40
−40
−40
−40
4
5
6
LDA
0
7
Et3N
32
22
27
47
72
8
2,6-lutidine
DIPEA, DMAP
DIPEA
DIPEA
9
10
a
11
a
From anti substrate 16 (cf. 13b, β-Me).
equivalent furnished the expected silyl enol ether 14a, albeit in
poor yield (entry 2). Significant improvement was obtained
when the reaction was run at −40 °C (entry 3). Lower
temperature resulted in a slower reaction without any significant
yield improvement (entry 4). Further investigation revealed that
the base was required (entry 5). However, a strong base appeared
to be too reactive (entry 6) and led to decomposition. DIPEA
was the most suitable base (entry 4) as it led to yields higher than
those of triethylamine or 2,6-lutidine (entries 7 and 8,
respectively). The use of a pyridine catalyst (DMAP, entry 9)
had no significant effect on the outcome of the reaction.
From the perspective of using the synthetic route described in
Scheme 2 to access a substrate bearing the dipolarophile branch,
we foresaw compatibility issues for the eventual ethyl carbamate
cleavage. For this reason, we also tested the allyl carbamate as the
nitrogen protecting group.13,14 The synthesis of allyl carbamate
13b uneventfully followed the sequence described in Scheme 2.
The corresponding 1,3-dioxa-2-silacyclohexene 14b was ob-
tained in approximately the same yield as that for product 13a
(Table 1, entries 3 and 10). A side product observed in all these
silylation conditions was the corresponding unsaturated
aldehyde 15, arising from the elimination of the alcohol on
substrate 13 (Scheme 3). A conformational analysis brought
some insight on this side reaction. Indeed, when the syn substrate
13 is subjected to silylation, the O-silylated alcohol is presumed
to intramolecularly activate the aldehyde as a six-membered ring
complex. Two chairlike conformations are then possible: the top
conformation exposes a severe steric interaction between the
carbamate side chain (R) and one of the isopropyl on silicon,
The resulting alcohol was oxidized under Swern conditions to
give aldehyde 8a. Evans aldol reaction with oxazolidinone 9
afforded alcohol 10a with good diastereoselectivity.8 Trans-
amidation toward Weinreb amide 11a was performed smoothly
without affecting the carbamate group.9 Standard reduction
conditions using DIBAL-H failed to afford aldehyde 13a.10 We
believe that, after a rapid deprotonation of the free alcohol, the
resulting O−AliBu2 complex hinders the amide to a point where
reduction with an additional equiv of DIBAL-H can no longer
occur. Switching to a reducing agent containing several hydrides
was appealing: after the initial deprotonation, an aluminate
complex such as 12 could deliver another hydride intra-
molecularly to the amide. The proximity of the hydride source
to the amide group may become very important on more
functionalized substrates bearing the dipolarophile branch to
ensure a chemoselective reduction. Unfortunately, LiAlH4 led to
poor yields, mainly because of over reduction of the resulting
aldehyde.11 After a quick survey of reducing agents bearing only
the two needed hydrides for initial deprotonation and intra-
B
DOI: 10.1021/acs.orglett.7b01566
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
material.16 We chose here to use a cyano group instead of the
TMS to later generate the azomethine ylide by deprotonation
rather than by desilylation as it usually requires milder
conditions.6a,19 The direct reduction of 20 proved to be more
difficult than with model compound 13a,20 so we chose to
protect the free alcohol with a TMS group, which was removed in
the acidic quench of the subsequent reduction reaction, affording
the diol 21 in essentially quantitative yields for these two steps. A
chemoselective oxidation of the primary alcohol was then
performed using catalytic TEMPO and sodium hypochlorite as
co-oxidant.21 During this step, we observed an erosion of the
diastereomeric ratio probably due to epimerization at the C8
position. However, this was inconsequential because the next
step was the enolization to the cyclic silyl enol ether. Formation
of the latter was performed using the conditions previously
developed (see Scheme 2), albeit in lower yield for reasons that
remain unknown. Deallyloxycarbonylation14 of 23 followed by
immediate formylation using Katrizky reagent22 satisfyingly
afforded key polycyclization precursor 24.
Once submitted to amide activation conditions (1.1 equiv of
Tf2O and 1.1 equiv of DTBMP), the Vilsmeier−Haack
cyclization proceeded cleanly and almost instantaneously to
form iminium ion 25 (Scheme 5). When the reaction solution
was transferred to a flask containing a refluxing solution of
DIPEA (5.0 equiv) in dichloromethane, the cycloaddition
reaction also occurred very rapidly.23 To our delight, the excess
of DIPEA employed in the azomethine ylide generation was
sufficient to promote elimination of the cyanide. Two different
Scheme 3. Effect of Relative Configuration on the Formation
of 1,3-Dioxa-2-silacyclohexene 14
thus favoring the bottom conformation. The latter has no proper
alignment to allow enolization of the aldehyde, leading mainly to
elimination product 15. When the same rationale was applied to
a substrate 16 presenting an anti relationship between the
carbamate side chain (R) and the methyl in the six-membered
ring complex, the top conformation should be favored. The latter
is perfectly suited for enolization of the aldehyde, and the desired
1,3-dioxa-2-silacyclohexene 14 should now be favored. This was
confirmed by preparing anti substrate 16,15,16 and as expected, a
remarkable yield increase was observed (Table 1, entry 11).
Having developed the chemoselective Weinreb amide
reduction and the successful preparation of the sensitive 1,3-
dioxa-2-silacyclohexene moiety, we were ready to synthesize the
key cyclization precursor 25 (Scheme 4). A cross-metathesis
performed on compound 1717 with methyl acrylate led to the
unsaturated ester 5.18 The latter was then submitted to
Heathcock aldol conditions with aldehyde 18 to afford the anti
product 19 in 75% yield (97% based on recovered starting
Scheme 5. Possible Reaction Pathways to 30 and 3125
Scheme 4. Synthesis of Key Step Precursor 24
C
DOI: 10.1021/acs.orglett.7b01566
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(2) (a) Heathcock, C. H.; Davidsen, S. K.; Mills, S.; Sanner, M. A. J. Am.
Chem. Soc. 1986, 108, 5650. (b) Heathcock, C. H.; Kath, J. C.; Ruggeri,
R. B. J. Org. Chem. 1995, 60, 1120. (c) Heathcock, C. H.; Stafford, J. A.;
Clark, D. L. J. Org. Chem. 1992, 57, 2575. (d) Stafford, J. A.; Heathcock,
C. H. J. Org. Chem. 1990, 55, 5433.
cycloadducts were obtained: the expected tricyclic aldehyde 30
and the tetracyclic silylated cyanohydrin 31, in a 1:1 ratio. To
explain the formation of these two products, we suspect a partial
conversion of silyl triflate 28 to cyclic oxocarbenium ion 29. The
latter can be trapped by the cyanide generated during the
elimination, leading to cyanohydrin 31. Hydrolysis of the
remaining silyl triflate 28 in the aqueous reaction quench
would lead to aldehyde 30. Gratifyingly, both 30 and 31 can
independently be desilylated using TBAF to generate the
tricyclic targeted product 32 in good yields. When the key
Vilsmeier−Haack/cycloaddition sequence followed by TBAF
treatment was run as a one-pot procedure, an excellent 69%
overall yield of product 32 was obtained from substrate 24.
Relative configuration was secured by X-ray analysis of crystalline
p-nitrobenzoyl derivative 3324 and was consistent with stereo-
chemical analysis presented in Scheme 1 (see 3 to 1).
In conclusion, we described the preparation and the
unprecedented cyclization of a chiral cyclic enol ether onto an
activated formamide. The following one-pot intramolecular
azomethine ylide generation and 1,3-dipolar cycloaddition
generated a functionalized enantiopure tricyclic core obtained
in as few as nine linear synthetic steps. This core will be used as a
common intermediate toward the synthesis of various members
of the calyciphylline B-type alkaloids family. Methods to
implement the remaining C, D, and F rings are currently under
investigation.
(3) Morita, H.; Kobayashi, J. I. Org. Lett. 2003, 5, 2895.
(4) (a) Chattopadhyay, A. K.; Berger, G.; Hanessian, S. J. Org. Chem.
2016, 81, 5074. (b) Chattopadhyay, A. K.; Menz, H.; Ly, V. L.; Dorich,
S.; Hanessian, S. J. Org. Chem. 2016, 81, 2182.
(5) Chattopadhyay, A. K.; Ly, V. L.; Jakkepally, S.; Berger, G.;
Hanessian, S. Angew. Chem., Int. Ed. 2016, 55, 2577.
(6) (a) Boudreault, J.; Lev
81, 9247. (b) Belanger, G.; Boudreault, J.; Lev
13, 6204. (c) Hauduc, C.; Belanger, G. J. Org. Chem. 2017, 82, 4703.
́
esque, F.; Bel
́
anger, G. J. Org. Chem. 2016,
́
́
esque, F. Org. Lett. 2011,
́
(7) (a) Singh Batra, M.; Brunet, E. Tetrahedron Lett. 1993, 34, 711.
(b) Ryu, I.; Murai, S.; Shinonaga, A.; Horiike, T.; Sonoda, N. J. Org.
Chem. 1978, 43, 780.
(8) (a) Evans, D. A.; Bartroli, J.; Shih, T. L. J. Am. Chem. Soc. 1981, 103,
2127. (b) Evans, D. A.; Vogel, E.; Nelson, J. V. J. Am. Chem. Soc. 1979,
101, 6120.
(9) Evans, D. A.; Gage, J. R.; Leighton, J. L. J. Am. Chem. Soc. 1992, 114,
9434.
(10) Direct reduction of 10a failed to deliver aldehyde 13a.
Transamidation to Weinreb amide 11a then reduction to aldehyde
13a was the best option.
(11) Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815.
(12) For examples of partial reduction of tertiary amides with
bis(alkoxy)aluminum hydrides, see: (a) Brown, H. C.; Tsukamoto, A. J.
Am. Chem. Soc. 1964, 86, 1089. (b) Cha, J. S.; Lee, J. C.; Lee, H. S.; Lee,
S. E.; Kim, J. M.; Kwon, O. O.; Min, S. J. Tetrahedron Lett. 1991, 32,
6903.
(13) Orthogonal deprotection of the allyl carbamate could be effected
using Pd(0) and morpholine as nucleophile (see ref 14). These
conditions should not affect the cyclic silyl enol ether or the unsaturated
ester.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.7b01566.
(14) (a) Lee, H.; Suzuki, M.; Cui, J.; Kozmin, S. J. Org. Chem. 2010, 75,
1756. (b) Bihelovic, F.; Ferjancic, Z. Angew. Chem., Int. Ed. 2016, 55,
2569.
(15) The anti aldol product 16 was prepared from 8b and 9 using
Heathcock’s modification of Evans aldol reaction conditions (see ref
16). Procedures and characterization data are found in the Supporting
Information.
(16) Walker, M. A.; Heathcock, C. H. J. Org. Chem. 1991, 56, 5747.
(17) Kaliappan, K. P.; Ravikumar, V. J. Org. Chem. 2007, 72, 6116.
(18) Chatterjee, A. K.; Choi, T.-L.; Sanders, D. P.; Grubbs, R. H. J. Am.
Chem. Soc. 2003, 125, 11360.
Experimental procedures, characterization data, and
1
copies of the H and 13C NMR spectra for all new
compounds (PDF)
Single-crystal X-ray data for 33 (CIF)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: guillaume.belanger@USherbrooke.ca
ORCID
́ ́
(19) Levesque, F.; Belanger, G. Org. Lett. 2008, 10, 4939.
(20) Upon treatment of alcohol 19 with NaBH4, we suspect a
conjugate addition of the resulting sodium alkoxide to the unsaturated
ester.
Guillaume Belanger: 0000-0002-7454-8760
́
Notes
(21) Lucio Anelli, P.; Biffi, C.; Montanari, F.; Quici, S. J. Org. Chem.
1987, 52, 2559.
The authors declare no competing financial interest.
(22) (a) Katritzky, A. R.; Chang, H.-X.; Yang, B. Synthesis 1995, 1995,
503. (b) Pasqua, A. E.; Matheson, M.; Sewell, A. L.; Marquez, R. Org.
Process Res. Dev. 2011, 15, 467.
(23) Longer reaction times for that step provided the same products 30
and 31 albeit in a lower combined yield (50%).
ACKNOWLEDGMENTS
■
The authors thank Dr. Daniel Fortin (U. Sherbrooke) for X-ray
diffraction analysis. This research was supported by the Natural
Science and Engineering Research Council (NSERC) of Canada
and the Universite
́
de Sherbrooke.
(24) An ORTEP representation of 33 is found in the Supporting
Information. Crystallographic data for this compound have been
deposited with the Cambridge Crystallographic Data Centre (CCDC
no. 1548926). Coordinates can be obtained at http://www.ccdc.cam.ac.
uk/deposit.
REFERENCES
■
(1) (a) Yamada, R.; Adachi, Y.; Yokoshima, S.; Fukuyama, T. Angew.
Chem., Int. Ed. 2016, 55, 6067. (b) Diaba, F.; Martínez-Laporta, A.;
(25) For representation of X-ray structure, see: CYLview, 1.0b; Legault,
́
Coussanes, G.; Fernandez, I.; Bonjoch, J. Tetrahedron 2015, 71, 3642.
C. Y., Universite
́
de Sherbrooke, 2009 (http://www.cylview.org).
(c) Wang, W.; Li, G. P.; Wang, S. F.; Shi, Z. F.; Cao, X. P. Chem. - Asian J.
2015, 10, 377. (d) Chattopadhyay, A. K.; Hanessian, S. Chem. Rev. 2017,
117, 4104. (e) Shvartsbart, A.; Smith, A. B. J. Am. Chem. Soc. 2014, 136,
870. (f) Coldham, I.; Burrell, A. J. M.; Guerrand, H. D. S.; Oram, N. Org.
Lett. 2011, 13, 1267. (g) Coldham, I.; Watson, L.; Adams, H.; Martin, N.
G. J. Org. Chem. 2011, 76, 2360.
D
DOI: 10.1021/acs.orglett.7b01566
Org. Lett. XXXX, XXX, XXX−XXX