Total Synthesis of Neomarchantin A: Key Bond Constructions Performed Using Continuous Flow Methods

Article abstract of DOI:10.1021/acs.orglett.7b01127

A synthesis of neomarchantin A has been achieved wherein key bond constructions involving C-O or C-C bond formations were augmented via continuous flow techniques. Of note, the synthesis of neomarchantin A represents the first demonstration of catalytic m

Full text of DOI:10.1021/acs.orglett.7b01127

Letter
pubs.acs.org/OrgLett
Total Synthesis of Neomarchantin A: Key Bond Constructions
Performed Using Continuous Flow Methods
́
Emilie Morin, Michael Raymond, Amaury Dubart, and Shawn K. Collins*
̈
Department of Chemistry and Centre for Green Chemistry and Catalysis, Universite
Montreal, Quebec, Canada H3C 3J7
́ ́
de Montreal, CP 6128 Station Downtown,
́
́
S
* Supporting Information
ABSTRACT: A synthesis of neomarchantin A has been achieved wherein key bond constructions involving C−O or C−C bond
formations were augmented via continuous flow techniques. Of note, the synthesis of neomarchantin A represents the first
demonstration of catalytic macrocyclic olefin metathesis as a key step for the synthesis of a macrocyclic bisbibenzyl natural
product.
acrocyclic bisbibenzyl natural products have been
M
isolated from liverworts, other bryophytes,1 and higher
flowering plants.2 Their biosynthetic origin stems from the
dimerization of two molecules of lunularin,3 which results in
macrocycles having ethano and biaryl ether linkages (Figure 1).
Figure 1. Representative structural features of the macrocyclic
bisbibenzyl family of natural products: lunularin (the monomeric
unit) and examples of macrocycles having ethano, biaryl ether, and
biaryl linkages. Ring sizes are indicated in red.
Members of the bisbibenzyl family can exhibit varying
connectivity patterns and substitution by hydroxy and/or
alkoxy groups. Some macrocyclic bisbibenzyls possess
aromatics that are halogenated or further oxidized.4 An
additional structural feature of the macrocyclic bisbibenzyl
natural products is that a number of the compounds exhibit
atropisomerism.5 Neomarchantin A (Figure 2) is a member of
the marchantin subclass of macrocyclic bisbibenzyls.6 Although
the macrocycle is achiral, the presence of two biaryl ether
subunits in conjunction with their respective connectivities
(two para- and two meta-substituted benzene rings) results in a
rigidified structure.7 Aside from their interesting structural
features, the marchantins have displayed antibacterial and
antimycotic effects as well as inhibition of 5-lipoxygenase and
Figure 2. Retrosynthetic analysis of neomarchantin A. Ring size is
indicated in red.
Received: April 13, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b01127
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
activity as antioxidants.8 Marchantins have shown cytotoxicity
in P-388 mouse leukemia cells and KB cell lines and inhibition
of the growth and migrating ability of human glioma A 172
cells.9 Recently, the marchantins have been revealed to possess
activity against influenza A (both H3N2 and H1N1) and
influenza B viruses.10 The combination of impressive diversity
in biological activity and intriguing macrocyclic structures has
bolstered interest in bisbibenzyl natural products.4 The key
synthetic challenges of macrocyclic bisbibenzyls are generally
regarded as formation of the biaryl ethers and formation of the
eventual macrocycle. The former is generally addressed via
SNAr reactions or Ullmann couplings, which typically employ
harsh conditions (excess reagents, high temperatures). The
macrocyclization processes are often olefinations (Wittig,
McMurry), which can afford excellent yields but require
lengthy syntheses and stoichiometric reagents. Some other
macrocyclization strategies have included Pd-catalyzed cross-
coupling for the construction of biaryl linkages and de novo
synthesis of an arene via a Diels−Alder reaction.11 Herein, a
total synthesis of neomarchantin A is presented that employs
catalysis and continuous flow methods to address the key
synthetic challenges of biaryl ether formation and macrocyclic
ring closure (Figure 2).
synthesis of C−N bonds via the coupling of amines and
arylboronic acids has recently been reported in continuous flow
employing tube-in-tube reactors for efficient delivery of
molecular oxygen.13 However, similar cross-couplings for the
formation of C−O bonds have not been reported. When the
conditions reported by Baxendale and co-workers for C−N
bond formation were employed (25 mol % Cu(OAc)2, CH2Cl2,
O2) in continuous flow using the tube-in-tube reactor for C−O
bond formation, the yield of biaryl ether 3 was increased 3-fold
(34% vs 11% using the Ullmann conditions; Scheme 1).14 For
comparison, the Chan−Evans−Lam cross-coupling of boronic
acid 8 and phenol 5 was performed with the same catalyst
system in a batch reactor, but the oxygen pressure was only 1
atm or ∼14 psi (vs 100 psi in flow), and only an 8% yield of the
desired product was isolated. Consequently, the flow reactor
afforded higher efficiency than the batch reactor. The synthesis
of the second biaryl ether unit 4 was also first investigated using
the Ullmann coupling (Scheme 1). Again the yield was
disappointing, as the reaction afforded 26% biaryl ether 4 (25
mol % CuO, pyridine).15 With the continuous flow/Chan−
Evans−Lam conditions, a similar yield of 33% was obtained.
As an alternative approach to prepare biaryl ether
intermediate 4, an SNAr route arising from aryl fluoride 11
was investigated. The coupling of phenol 6 and aryl fluoride 11
occurred in 86% yield when performed under K2CO3/DMSO/
130 °C reaction conditions. Various SNAr reactions have been
reported to provide high yields and short reaction times under
continuous flow conditions,16 while the synthesis of biaryl
ethers specifically has been shown to proceed when performed
in supercritical CO2.17 Given the possibility to augment the
process via continuous flow, the SNAr synthesis of biaryl ether 4
was repeated using a heated column of K2CO3 at 130 °C,
affording the desired ether 4 in 71% yield but with a
significantly shorter reaction time of 12 min.18 The synthesis
of the ethano bridge between the two biaryl ethers was
performed using Wittig chemistry (Scheme 2). Functional
The synthesis of neomarchantin A commenced with the
synthesis of two key biaryl ether intermediates following
established Ullmann coupling protocols, which proved to be
challenging (Scheme 1). The preparation of the first biaryl
Scheme 1. Comparison of Ullmann, Chan−Evans−Lam
Coupling, and SNAr Routes to Key Biaryl Ether
Intermediates
Scheme 2. Synthesis of Macrocyclization Precursor 2
ether 3 from phenol 5 was catalyzed by CuI (25 mol %) using
Cs2CO3 as a base in refluxing pyridine, affording a low yield of
ether 3 (11%). An extensive optimization was performed, and
higher temperatures, alternative ligands, palladium-based
catalysis, or the use of an iodo coupling partner did not
improve the yield.12
group manipulation was necessary to prepare a phosphonium
salt from ester 12. A series of reduction (NaBH4), halogenation
(CBr4),19 and phosphonium salt generation (PPh3) steps were
then performed in a linear sequence to afford the salt 12. Biaryl
ether aldehyde 3 and phosphonium salt 14 could then undergo
Wittig olefination (Cs2CO3 in refluxing CH2Cl2) to afford a
96% yield of the desired stilbene. Subsequent reduction (H2,
As an alternative route to biaryl ethers, the use of Cu-
catalyzed Chan−Evans−Lam couplings was envisioned. The
B
DOI: 10.1021/acs.orglett.7b01127
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Pd/C) completed one of the ethano bridges of neomarchantin
A. A sequence involving reduction (LiAlH4), oxidation (PCC),
and olefination (MePPh3Br, Cs2CO3) was used to prepare the
bis-styrenyl intermediate 2.
Table 2. Synthesis of the Macrocyclic Core of
Neomarchantin A in Continuous Flow via Olefin Metathesis
In view of its widespread use in macrocyclization of natural
products, it is surprising that macrocyclic olefin metathesis has
yet to be successfully described as a viable tactic for the
synthesis of macrocyclic bisbibenzyls. A single report by
Harrowven and co-workers describes attempts to prepare the
macrocyclic core of riccardin C using macrocyclic ring-closing
metathesis,20 but only a cyclic dimer product was isolated in
32% yield. The macrocyclization of diene 2 to form the
macrocyclic core 16 of neomarchantin A was first investigated
in batch, and both the catalyst and diene 2 were added in one
portion without the aid of any slow addition techniques (Table
1). When catalyst G2 was investigated for macrocyclization in
a
entry
1
time (min)
yield of 16 (%)
10
10
10
30
60
49 (51)
37 (45)
15 (85)
42 (45)
33 (55)
b
2
c
3
4
5
a
Yields following chromatography. Recovered 2 in parentheses. Ring
b
c
Table 1. Synthesis of the Macrocyclic Core of
Neomarchantin A in Batch via Olefin Metathesis
size in red. At 5 mM. with G2 (5 mol %).
G2 resulted in lower yields. Changes in the residence time (30
or 60 min) also did not improve the yield. Interestingly, when
the macrocyclization was repeated with a 10 min residence time
but without the applied vacuum, the yield of 16 neither
improved nor decreased. Attempts to perform the reaction at
higher temperatures or with higher catalyst loadings did not
promote conversion, suggesting that the catalyst degrades fully
in the 10 min reaction window. The completion of the
synthesis was achieved via hydrogenolysis of the olefin of
macrocycle 16 and subsequent deprotection of the methyl
ethers to afford neomarchantin A (1), whose spectral analyses
matched those in the literature (Scheme 3).6
a
entry
catalyst
solvent
yield of 16 (%)
1
2
3
4
G2
PhMe
35 (65)
30 (25)
43 (0)
G2
CH2Cl2
PhMe
GH2
GH2
Scheme 3. Completing the Synthesis of Neomarchantin A
CH2Cl2
30 (60)
a
Yields following chromatography. Recovered starting material 2 in
parentheses. Ring size in red.
either PhMe or CH2Cl2 at reflux, macrocycle 16 was isolated as
the E isomer in 30−35% yield with varying quantities of
recovered starting material 2 (25−65%). The more active GH2
catalyst resulted in a slightly improved 43% yield of the 20-
membered ring 16. Changing the solvent to CH2Cl2 afforded a
slight drop in the yield of macrocycle 16 to 30%. Next, attempts
to improve the macrocyclization via continuous flow were
investigated, where the efficient energy transfer could help to
promote the difficult ring-closing event in macrocyclization.
Macrocyclization via olefin metathesis in continuous flow has
been reported for the synthesis of musc-like macrolactones.
Both Fogg21 and Skowerski22 had previously described
macrocyclic olefin metathesis reactions in continuous flow
that were problematic if the corresponding reaction setups were
not designed to efficiently remove ethylene from the reaction
mixture.
The cyclization of diene 2 in continuous flow was carried out
using a tube-in-tube reactor23 connected to a vacuum pump to
help remove ethylene. Following elution from the reactor coil,
the solution was collected in a round-bottom flask containing
ethyl vinyl ether to quench any residual catalyst. In view of the
propensity of flow processes to improve reaction times, our
initial attempt was performed with a 10 min residence time and
afforded the desired macrocycle 16 in 49% yield with 51%
residual unreacted 2 (Table 2). Increasing the concentration
did not improve conversion and attempts at ring-closing using
In summary, a synthesis of neomarchantin A has been
described that involves the first report of a catalytic macrocyclic
olefin metathesis reaction as a key step for the synthesis of a
macrocyclic bisbibenzyl natural product. In addition, the
chemistries used for the key bond constructions were
augmented through continuous flow techniques. The biaryl
ether formation via C−O bond formation was conducted at
lower temperatures with shorter reaction times and afforded
either comparable or higher yields. An SNAr approach was also
shown to be applicable to prepare a biaryl ether intermediate in
short reaction times versus batch processes. The macro-
cyclization event via C−C bond formation proceeded in similar
yields in batch and in continuous flow, while the latter provided
a dramatic improvement in reaction time (10 min vs 17 h) and
should improve the scalability of the olefin metathesis process.
The continuous flow strategies used for the aforementioned
bond constructions should find additional application in the
synthesis of other members of the bisbibenzyl family of natural
products as well as in the synthesis of other biologically active
compounds containing biaryl ether and macrocycle structural
features.
C
DOI: 10.1021/acs.orglett.7b01127
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
K. Tetrahedron Lett. 2016, 57, 654−657. For other examples of C−N
coupling in flow, see: (c) Yang, J. C.; Niu, D.; Karsten, B. P.; Lima, F.;
Buchwald, S. L. Angew. Chem., Int. Ed. 2016, 55, 2531−2535.
(d) Naber, J. R.; Buchwald, S. L. Angew. Chem., Int. Ed. 2010, 49,
9469−9474.
(14) Optimization of the Cham−Evans−Lam in continuous flow
included changing the ligand (TMEDA, pyridine), varying the quantity
of Et3N, using myristic acid as a solubilizing agent for the copper
catalyst, and changing the residence time (1 → 4 h).
(15) A 91% yield of 4 could be achieved but required excess CuO
(1.4 equiv) and coupling partner 9 (6 equiv).
(16) (a) Comer, E.; Organ, M. G. Chem. - Eur. J. 2005, 11, 7223−
7227. (b) Organ, M. G.; Hanson, P. R.; Rolfe, A.; Samarakoon, T. B.;
Ullah, F. J. Flow Chem. 2011, 1, 32−39. (c) Marafie, J. A.; Moseley, J.
D. Org. Biomol. Chem. 2010, 8, 2219−2227.
(17) Lee, J.-K.; Fuchter, M. J.; Williamson, R. M.; Leeke, G. A.; Bush,
E. J.; McConvey, I. F.; Saubern, S.; Ryan, J. H.; Holmes, A. B. Chem.
Commun. 2008, 4780−4782.
(18) A similar SNAr approach to 3 was unsuccessful.
(19) Dai, C.; Narayanam, J. M. R.; Stephenson, C. R. J. Nat. Chem.
2011, 3, 140−145.
(20) Kostiuk, S. L.; Woodcock, T.; Dudin, L. F.; Howes, P. D.;
Harrowven, D. C. Chem. - Eur. J. 2011, 17, 10906−10915.
(21) Monfette, S.; Eyholzer, M.; Roberge, D. M.; Fogg, D. E. Chem. -
Eur. J. 2010, 16, 11720−11725.
(22) Skowerski, K.; Czarnocki, S. J.; Knapkiewicz, P. ChemSusChem
2014, 7, 536−542.
(23) For a review of applications of the tube-in-tube reactor, see:
Brzozowski, M.; O’Brien, M.; Ley, S. V.; Polyzos, A. Acc. Chem. Res.
2015, 48, 349−362.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.7b01127.
■
S
Experimental procedures and characterization data for all
new compounds (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: shawn.collins@umontreal.ca.
ORCID
Shawn K. Collins: 0000-0001-9206-5538
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
The authors acknowledge the Natural Sciences and Engineering
Research Council of Canada, Universite
■
́
de Montreal, the
Centre for Green Chemistry and Catalysis, and the NSERC
CREATE Program in Continuous Flow Science for generous
funding. The Canadian Foundation for Innovation is acknowl-
edged for generous funding of the flow chemistry infrastructure.
Ms. V. Kairouz is thanked for assistance with continuous flow
́
infrastructure. E.M. and M.R. thank NSERC and the FRQNT
for graduate scholarships. Materia Inc. is thanked for generous
donations of catalysts G2 and GH2.
REFERENCES
■
(1) Asakawa, Y. Phytochemistry 2004, 65, 623−669.
(2) Kosenkova, Y. S.; Polovinka, M. P.; Komarova, N. I.; Korchagina,
D. V.; Kurochkina, N. Y.; Cheremushkina, V. A.; Salakhutdinov, N. F.
Chem. Nat. Compd. 2007, 43, 712−713.
(3) Asakawa, Y. Prog. Chem. Org. Nat. Prod. 1995, 65, 335−351.
(4) Harrowven, D. C.; Kostiuk, S. L. Nat. Prod. Rep. 2012, 29, 223−
242.
(5) Toyota, M.; Yoshida, T.; Kan, Y.; Takaoka, S.; Asakawa, Y.
Tetrahedron Lett. 1996, 37, 4745−4748.
(6) Keseru, G. M.; Nogradi, M. Nat. Prod. Rep. 1995, 12, 69−75.
̋
(7) Bocskei, Z.; Keseru, G. M. J. Chem. Soc., Perkin Trans. 2 1994, 2,
595−597.
̈
̋
(8) (a) Schwartner, C.; Michel, C.; Stettmaier, K.; Wagner, H.; Bors,
W. Free Radical Biol. Med. 1996, 20, 237−244. (b) Hsiao, G.; Teng, C.-
M.; Wu, C.-L.; Ko, F.-N. Arch. Biochem. Biophys. 1996, 334, 18−26.
(9) (a) Asakawa, Y.; Toyota, M.; Taira, Z.; Takemoto, T.; Kido, M. J.
Org. Chem. 1983, 48, 2164−2167. (b) Shen, J.; Li, G.; Liu, Q.; He, Q.;
Gu, J.; Shi, Y.; Lou, H. Cancer Biol. Ther. 2010, 9, 33−39.
(10) Iwai, Y.; Murakami, K.; Gomi, Y.; Hashimoto, T.; Asakawa, Y.;
Okuno, Y.; Ishikawa, T.; Hatakeyama, D.; Echigo, N.; Kuzuhara, T.
PLoS One 2011, 6, e19825.
(11) Zhao, P.; Beaudry, C. M. Org. Lett. 2013, 15, 402−405.
(12) Optimization included exploring the effects of temperature (80
→ 130 °C), electrophile (Br vs I), solvent (pyridine, DMF, PhMe,
dioxane, DMSO), base (K2CO3, Cs2CO3), copper source (CuO,
Cu2O, CuCl, CuBr2, CuI, Cu(acac)2), and ligand (tetramethylphenan-
throline, tetramethylheptadione, glycine, proline, 8-hydroxyquinoline).
At best, a 29% yield of 3 could be achieved when 1.4 equiv of CuI and
6 equiv of the coupling partner 7 were used. Ullmann couplings of 3
and 7 were not attempted in flow because of their heterogeneous
nature.
(13) For examples of Chan−Lam couplings in flow, see: (a) Mallia,
C. J.; Burton, P. M.; Smith, Al. M. R.; Walter, G. C.; Baxendale, I. R.
Beilstein J. Org. Chem. 2016, 12, 1598−1607. (b) Bao, J.; Tranmer, G.
D
DOI: 10.1021/acs.orglett.7b01127
Org. Lett. XXXX, XXX, XXX−XXX