Radical [3 + 2]-annulation of divinylcyclopropanes: Rapid synthesis of complex meloscine analogs

Article abstract of DOI:10.1021/ol403078e

A radical [3 + 2]-divinylcyclopropane annulation cascade has been extended to encompass five D-ring variants of the meloscine/epimeloscine core structure. Representative ABCD tetracyclic intermediates were further elaborated with novel substituted E-rings through subsequent transformations of advanced intermediates that provided opportunities for late-stage variation of the B-ring (lactam) N-substituents which were also developed.

Full text of DOI:10.1021/ol403078e

Letter
pubs.acs.org/OrgLett
Radical [3 + 2]-Annulation of Divinylcyclopropanes: Rapid Synthesis
of Complex Meloscine Analogs
Hanmo Zhang, Kyu Ok Jeon, E. Ben Hay, Steven J. Geib, Dennis P. Curran,* and Matthew G. LaPorte*
Department of Chemistry and Center for Chemical Methodology and Library Development, University of Pittsburgh, Pittsburgh,
Pennsylvania 15260, United States
S
* Supporting Information
ABSTRACT: A radical [3 + 2]-divinylcyclopropane annulation cascade has been extended to encompass five D-ring variants of
the meloscine/epimeloscine core structure. Representative ABCD tetracyclic intermediates were further elaborated with novel
substituted E-rings through subsequent transformations of advanced intermediates that provided opportunities for late-stage
variation of the B-ring (lactam) N-substituents which were also developed.
he fascinating structures and powerful biological activities
of natural products continue to inspire synthetic and
2] radical annulation reaction of a divinylcyclopropane
(DVCP) to assemble the C-ring and the adjacent bonds
(represented in red in Figure 1).14−17 The two examples
reported in that synthesis are shown in Scheme 1.15 Radical
cyclization of sec-amide 4 provided lactam 6 (R = H) in 38%
yield while tert-amide 5 gave the corresponding lactam 7 (R =
Bn) in 55% yield. Both products have the epimeloscine
configuration at the BC-ring fusion.
T
medicinal chemists.1 Recent cheminformatic analyses of natural
products reveal increased structural diversity compared to
typical commercial sample collections largely because the
natural product samples contain more sp3-hybridized carbon
atoms.2a−c Accordingly, the construction of natural products
and derivatives or natural product inspired libraries is an
important goal.2d
It can be challenging to make core analogs of natural
products like the members of the meloscine family shown in
Figure 1.3 To be suitable for analog preparation, syntheses of
Scheme 1. Divinylcyclopropane Cascade Cyclization with
Likely Intermediates
Figure 1. Structures of meloscine 1a, epimeloscine 2s, and scandine 3.
such complex molecules need to be both short and flexible. The
four interlocking B−E-rings with the central hexasubstituted
cyclopentane C-ring poses a significant challenge. Methods to
make complex cyclopentanes are often based on [3 + 2]
cycloadditions to alkenes. These can involve 1,3-zwitterions or
diradicals, or their equivalents.4 Both transition-metal-cata-
lyzed5−11 and radical-mediated reactions of vinylcyclopropanes
(VCPs) are also important.12,13
We have recently reported a short total synthesis of the
pentacyclic alkaloid meloscine 1a that featured a tandem [3 +
Received: October 25, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ol403078e | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
This transformation is believed to occur through addition of
the tributyltin radical to one of the vinyl groups of 4 or 5,
followed by fragmentation of the cyclopropane ring to give
radical A. A subsequent 6-exo-trig-cyclization gives tricyclic
intermediate B, which in turn undergoes an exo-trig cyclization.
Elimination of the starting tin radical then gives the final
product 6 or 7. The better yield afforded by 5 compared to 4
probably reflects amide rotamer differences.15 Intermediate 6 is
then converted to epimeloscine 2a in only three steps followed
by an epimerization to give meloscine 1a.
Scheme 2. Synthesis of Annulation Precursors 16a−c
We hypothesized that this synthesis anchored by the DVCP
annulation would be short and flexible enough for oppor-
tunities to diversify the meloscine core at the B,D- and E-rings.
The plan for making the desired analogs incorporates the four
fragments 8−11 in Figure 2.
Scheme 3. Completion of the D-Ring Analog Synthesis
Figure 2. Fragments in the design of the pentacyclic analogs.
The variable fragments are alkenyl iodides 10 and late-stage
electrophiles 11, which provide access to various sized D-rings
and differently functionalized E-rings, respectively. The
invariant fragment 9 is the core of the DVCP annulation.
The aniline fragment 8 is also constant with the selection of a
N-Boc substituent to favor radical cyclization (due to amide
conformation15) and to facilitate simultaneous protecting group
removal of advanced intermediate 17. In this manner, no
additional steps are required for analog preparation and the N-
substituent R on the B-ring of the final pentacycle 12 is
conveniently varied through alkylations. Finally, if the radical
annulation remains stereoselective providing the less-stable BC-
fusion isomer, then the number of final analogs can potentially
be doubled simply by adding end-game epimerizations.
Scheme 2 shows the synthesis of annulation precursors 16a−
c with three different sized D-rings. To make the five-
membered precursor 16a, N-Boc-2,3-dihydro-1H-pyrrole 13a
was converted to vinyl iodide 10a by reaction with ICl/NaOMe
followed by exposure to TFA.18 Palladium-catalyzed coupling
of 10a with boronic ester 8 provided aniline 14a in 69% yield.
Exposure of this aniline to Boc-anhydride gave bis-carbamate
15a in 94% yield. Finally, acylation with acid chloride 9
provided the annulation precursor 16a. The six- and seven-
membered precursors 16b and 16c were assembled by similar
reaction sequences in comparable yields.
Rapid syntheses of the six target pentacycles (1/2a−c) from
16a−c are summarized in Scheme 3. Again illustrating with the
five-membered D-ring series, slow addition of Bu3SnH to 16a
in refluxing toluene provided 17a in 53% yield. The trans BC-
ring fusion was confirmed by removal of the two Boc groups
with TFA to give 18a, which is an intermediate in the prior
meloscine synthesis.15 Conversion of crude 18a to allyl amine
19a followed by RCM provided epimeloscine 2a which was
then epimerized to meloscine 1a using t-BuOK.
a
b
crude product used directly in the next step. Two-step yield.
Radical annulations of higher homologues 16b,c also gave
single isomers 17b,c in 52% and 35% isolated yields.19 Removal
of the Boc groups, allylation (59% and 76% over two steps),
and RCM (73% and 76%) provided the ring-expanded
epimeloscines 2b,c. Finally, epimerization of these two samples
provided ring-expanded meloscines 1b,c in 82% and 84% yield.
Assignment of the trans configurations of the products in the
ring-expanded series followed from similarities in the NMR
spectra of 18a−c. (Spectra of 17a−c have broadened
resonances due to the Boc groups and were not instructive.)
The trans BC fusion of 18b was ultimately confirmed by X-ray
crystallography (Figure 3).
B
dx.doi.org/10.1021/ol403078e | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Information shows selected calculated properties for all of the
final products described herein). The treatment of lactams 1a,
2a, and 25 with potassium hexamethyldisilazane followed by
the corresponding alkyl halides provided the N-alkylated
derivatives 28−31 in moderate to good yields (Scheme 5).
Scheme 5. Synthesis of Substituted B-Ring Analogs
Figure 3. ORTEP structure of amine 18b.
Overall, the short synthetic sequence to the pentacycles 1/
2a−c (4−5 steps from N-Boc DVCP intermediates 16a−c) is
reliable with comparable yields across the three series. The
radical cyclizations were all selective for the epi-configuration,
and epimerization of epimeloscine 2a and its fused piperidine
and homopiperidine D-ring congeners 2b−c gave the
corresponding meloscines 1a−c.
We then explored the incorporation of additional diversity
elements by synthesis of lactam and sultam E-ring analogs
(Scheme 4). The pyrrolidine intermediates 18a and 20 were
Scheme 4. Synthesis of E-Ring Analogs
The molecular weights for the analogs range upward from
292.4 to 432.5 g M−1, though most remain below 400 g M−1.
Calculated log P values span about 2 orders of magnitude from
2.2 to 4.2. These values are squarely within accepted ranges for
biological screening.22
In summary, the recent synthesis of meloscine and
epimeloscine15 was expanded to include new ring sizes, ring
types, and ring substituents. The radical [3 + 2]-annulation of
various DVCP precursors provided the A−D tetracycles with
various sized D-rings. This key reaction was reliably stereo-
selective and tolerant of several different N-substituents (H, Bn,
Boc) that allowed for additional B-ring functionalizations of
epimeloscine and meloscine. New lactam- and sultam-
containing E-rings were made by straightforward conversions
of common advanced intermediates. Products with the
epimeloscine configuration were conveniently epimerized to
the analogous meloscines. The work shows that the DVCP
approach to the melsocines meets the brevity and flexibility
requirements for making diverse analogs of meloscine.
acylated with 21 under standard conditions to give amides 23
and 24a in 53% and 94% yields. Attempts to sulfonylate sec-
amide 18a with 22 were not successful, but the N-benzyl amide
20 gave the desired sulfonamide 24b in 30% yield. Subsequent
RCM reactions of these three precursors with the Hoveyda−
Grubbs catalyst generated the E-ring lactams 25 and 26a in
65% and 84% yield, while the E-ring sultam20 26b was isolated
in 74% yield. The N-benzyl epimeloscine analogs 26a,b were
epimerized to the new melsocine analogs 27a and 27b in 86%
and 84% yield.
Although the primary focus of this work involved synthetic
methodology scope, we were mindful that these alkaloid-like
products are prime targets for biological screens as molecular
probes or in other settings.21 In this respect, meloscine is an
attractive complex core because its molecular weight (292.4 g
M−1) and calculated log P (clog P, 2.2) are both relatively
low.22 Thus, the structure of meloscine can be varied
significantly without high molecular weight or high lipophilicity
limitations. The described methodology to both of the
meloscine epimers provided further diversification opportuni-
ties via small alkyl B-ring lactam substitutions that would
augment the overall lipophilicities (Table S1 in the Supporting
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedures and spectral data for all new
compounds. This material is free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: curran@pitt.edu.
*E-mail: mgl12@pitt.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the NIH/NIGMS CMLD
program (P50GM067082). The authors would like to thank
Prof. Peter Wipf (University of Pittsburgh) for constructive
discussions.
C
dx.doi.org/10.1021/ol403078e | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(20) Jimen
2226.
́
ez-Hopkins, M.; Hanson, P. R. Org. Lett. 2008, 10, 2223−
REFERENCES
■
(1) (a) Huryn, D. M.; Wipf, P. Natural Product Chemistry and
Cancer Drug Discovery. In Cancer Drug Discovery, 2nd ed.;Neidle, S.,
Ed.; Academic Press, 2013; Part II, pp 91−120. (b) Ganesan, A. Curr.
Opin. Chem. Biol. 2008, 12, 306−317. (c) Newman, D. J.; Cragg, G. M.
J. Nat. Prod. 2012, 75, 311−335.
(21) Samples of all the final products have been deposited in the
UPitt CMLD Center library. For synthetic and biological probe aims
of Chemical Methodology and Library Development Centers
(CMLDs), see: http://www.nigms.nih.gov/Research/
FeaturedPrograms/CMLD/.
̌
(2) (a) Clemons, P. A.; Wilson, J. A.; Dancik, V.; Muller, S.;
(22) Lipinski, C. A.; Lombardo, F.; Dominya, B. W.; Feeney, P. J.
Carrinski, H. A.; Wagner, B. K.; Koehler, A. N.; Schreiber, S. L. Proc.
Natl. Acad. Sci. U.S.A. 2011, 108, 6817−6822. (b) Clemons, P. A.;
Bodycombe, N. E.; Carrinski, H. A.; Wilson, J. A.; Shamji, A. F.;
Wagner, B. K.; Koehler, A. N.; Schreiber, S. L. Proc. Natl. Acad. Sci.
U.S.A. 2010, 107, 18787−18792. (c) Lovering, F.; Bikker, J.; Humblet,
C. J. Med. Chem. 2009, 52, 6752−6756. (d) Camp, D.; Davis, R. A.;
Campitelli, M.; Ebdon, J.; Quinn, R. J. J. Nat. Prod. 2012, 75, 72−81.
(3) (a) Isolation: Bernauer, K.; Englert, G.; Vetter, W.; Weiss, E.
Adv. Drug Delivery Rev. 1997, 23, 3−25.
(23) Physicochemical properties were calculated using: Maestro
8.5.207-QikProp 2.1; Schrodinger, LLC: New York, NY, USA; all new
̈
analogs are tabulated in the Supporting Information.
Helv. Chim. Acta 1969, 52, 1886−1905. (b) Short review: Szabo,
́
L. F.
ARKIVOC 2007, 7, 280−290.
(4) (a) Wender, P. A.; Verma, V. A.; Paxton, T. J.; Pillow, T. H. Acc.
Chem. Res. 2008, 41, 40−49. (b) Rubin, M.; Rubina, M.; Gevorgyan, V.
Chem. Rev. 2007, 107, 3117−3179.
(5) Trost, B. M.; Morris, P. J.; Sprague, S. J. J. Am. Chem. Soc. 2012,
134, 17823−17831.
(6) Wang, H. N.; Yang, W.; Liu, H.; Wang, W.; Li, H. Org. Biomol.
Chem. 2012, 10, 5032−5035.
(7) Dieskau, A. P.; Holzwarth, M. S.; Plietker, B. J. Am. Chem. Soc.
2012, 134, 5048−5051.
(8) Jiao, L.; Lin, M.; Yu, Z. X. J. Am. Chem. Soc. 2011, 133, 447−461.
(9) Goldberg, A. F. G.; Stoltz, B. M. Org. Lett. 2011, 13, 4474−4476.
(10) Li, Q.; Jiang, G. J.; Jiao, L.; Yu, Z. X. Org. Lett. 2010, 12, 1332−
1335.
(11) (a) Jiao, L.; Lin, M.; Yu, Z.-X. Chem. Commun. 2010, 46, 1059−
1061. (b) Jiao, L.; Ye, S. Y.; Yu, Z. X. J. Am. Chem. Soc. 2008, 130,
7178−7179.
(12) (a) Feldman, K. S.; Berven, H. M.; Romanelli, A. L.; Parvez, M.
J. Org. Chem. 1993, 58, 6851−6856. (b) Feldman, K. S.; Romanelli, A.
L.; Ruckle, R. E.; Jean, G. J. Org. Chem. 1992, 57, 100−110.
(c) Feldman, K. S.; Burns, C. J. J. Org. Chem. 1991, 56, 4601−4602.
(d) Feldman, K. S.; Romanelli, A. L.; Ruckle, R. E.; Miller, R. F. J. Am.
Chem. Soc. 1988, 110, 3300−3302.
(13) Miura, K.; Fugami, K.; Oshima, K.; Utimoto, K. Tetrahedron
Lett. 1988, 29, 5135−5138.
(14) Preliminary studies: Guthrie, D. B.; Geib, S. J.; Curran, D. P. J.
Am. Chem. Soc. 2009, 131, 15492−15500.
(15) Total synthesis: Zhang, H.; Curran, D. P. J. Am. Chem. Soc.
2011, 133, 10376−10378.
(16) First total synthesis: Overman, L. E.; Robertson, G. M.;
Robichaud, A. J. J. Am. Chem. Soc. 1991, 113, 2598−2610.
(17) Other recent syntheses and approaches: (a) Selig, P.;
Herdtweck, E.; Bach, T. Chem.Eur. J. 2009, 15, 3509−3525.
(b) Hayashi, Y.; Inagaki, F.; Mukai, C. Org. Lett. 2011, 13, 1778−1780.
(c) Feldman, K. S.; Antoline, J. F. Org. Lett. 2012, 14, 934−937.
(d) Coldham, I.; Burrell, A. J. M.; Guerrand, H. D. S.; Watson, L.;
Martin, N. G.; Oram, N. Beil. J. Org. Chem. 2012, 8, 107−111.
(18) (a) The route is patterned after a synthesis of 9b in: Kiewel, K.;
Luo, Z. S.; Sulikowski, G. A. Org. Lett. 2005, 7, 5163−5165. (b) A
higher yielding synthesis of 9a from 4-Boc-amino-1-butyne is
described in the Supporting Information.
(19) The related N-benzyl-containing DVCP intermediate C with the
pendant N-sulfonyl-dihydropyrrolidine also cyclizes in 58% yield.
Although the N-sulfonyl group on D is difficult to remove, the result
further shows the generality of the DVCP radical cascade.
D
dx.doi.org/10.1021/ol403078e | Org. Lett. XXXX, XXX, XXX−XXX