Utilizing an o-quinone methide in asymmetric transfer hydrogenation: Enantioselective synthesis of brosimine A, brosimine B, and brosimacutin L

Article abstract of DOI:10.1002/anie.201507269

A concise and highly enantioselective synthesis of the flavonoids brosimine A, brosimine B, and brosimacutin L is reported for the first time. The key transformation is a single-step conversion of a flavanone into a flavan by means of an asymmetric transfer hydrogenation/deoxygenation cascade. You got much flava: The single step transformation of a flavanone into a flavan in a highly enantioselective fashion by means of a non-enzymatic kinetic resolution enabled the concise synthesis of all three title compounds.

Full text of DOI:10.1002/anie.201507269

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201507269
German Edition: DOI: 10.1002/ange.201507269
Flavonoids
Utilizing an o-Quinone Methide in Asymmetric Transfer
Hydrogenation: Enantioselective Synthesis of Brosimine A,
Brosimine B, and Brosimacutin L
Anton Keßberg and Peter Metz*
Abstract: A concise and highly enantioselective synthesis of
the flavonoids brosimine A, brosimine B, and brosimacutin L
is reported for the first time. The key transformation is a single-
step conversion of a flavanone into a flavan by means of an
asymmetric transfer hydrogenation/deoxygenation cascade.
B
rosimum acutifolium is a tree from the genus Moraceae,
which is distributed throughout Brazil. The extract of its stem
bark is valued in traditional medicine owing to its anti-
inflammatory and anti-rheumatic properties.[1] More recent
investigations on its crude ground bark extract revealed
promising antibacterial activity against both multidrug-resist-
ant and ATCC strains of Staphylococcus aureus.[2] Intense
studies towards the identification of the bioactive compounds
led to the isolation of a variety of flavans, including
brosimine A (1),[3,4] brosimacutin L (2),[4] and brosimine B
(3).[3,5] Brosimine B (3) was recently also isolated from the
leaves of Morus yunnanensis[6a] and from mulberry leaves.[6b]
Of note, flavan 2 was shown to display significant cytotoxic
activity against vincristine-resistant P388 cancer cells.[4] While
the absolute configuration of flavan 2 was assigned as (S) on
the basis of CD measurements, that of 3 is assumed to be (S)
but this is so far unknown, and neither the relative nor the
absolute configuration of 1 has yet been determined. Herein
we report the first enantioselective synthesis of flavans[7] 1–3
by using a novel domino asymmetric transfer hydrogenation
(ATH)/deoxygenation of a racemic flavanone with kinetic
resolution as the key transformation.
Scheme 1. Retrosynthesis of brosimine A (1), brosimacutin L (2), and
brosimine B (3) leading to racemic naringenin (rac-6).
As outlined in Scheme 1, brosimine A (1) might be
derived from brosimine B (3) through a catalyst-controlled
Shi epoxidation[8] followed by 5-exo-tet cyclization. Brosima-
cutin L (2) can also be connected to 3 by means of
a regioselective cobalt-catalyzed hydration according to
a method reported by Mukaiyama.[9] Flavan 3 should be
attainable from compound 4 by reductive removal of the
phenolic hydroxy substituent[10] and cleavage of the carbonate
groups. We envisioned that enantiopure phenol 4 might be
directly available from a racemic flavanone rac-5 bearing
a suitable 5-O-acyl substituent through the recently disclosed
rhodium-catalyzed ATH of flavanones[11] by utilizing a cata-
lyst developed by Noyori.[12] Similar ketone deoxygenations
lacking the enantioselectivity are well known for the reaction
of 5-O-acetyl[13] or 5-O-methoxycarbonyl[14] flavanones with
sodium borohydride.[15] A range of ketones rac-5 would be
readily obtained from racemic naringenin (rac-6) in analogy
to the previously reported synthesis of 8-prenylflavanones.[16]
The preparation of ATH substrates rac-5a,b from com-
mercially available rac-6 is depicted in Scheme 2. After
chemoselective conversion of rac-6 to give the di-O-Boc
derivative rac-7, the prenyl ether rac-8 was synthesized under
Mitsunobu[17] conditions in excellent yield. Subsequently,
a Eu-catalyzed sigmatropic rearrangement accelerated by
microwave irradiation afforded the 8-prenylflavanone rac-9
in only five minutes. Finally, treatment of rac-9 with either
acetic anhydride or methyl chloroformate gave rise to the
desired 5-O-acyl derivatives rac-5a and rac-5b, respectively.
The crucial ATH of the flavanones rac-5 with kinetic
resolution was screened by using the catalyst (S,S)-10[12] and
a mixture of formic acid and triethylamine in ethyl acetate at
[*] M. Sc. A. Keßberg, Prof. Dr. P. Metz
Fachrichtung Chemie und Lebensmittelchemie
Organische Chemie I, Technische Universität Dresden
Bergstrasse 66, 01069 Dresden (Germany)
E-mail: peter.metz@chemie.tu-dresden.de
Homepage: http://www.chm.tu-dresden.de/oc1/
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201507269.
1160
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1160 –1163

Angewandte
Communications
Chemie
the flavanone (R)-5b and the desired flavan 4 proved to be
difficult, the corresponding yields could be readily deter-
mined by 1H NMR integration of the product mixture.
Subsequent treatment with triflic anhydride allowed the
clean separation of the resultant triflate 11 from (R)-5b.
Analysis by HPLC on a chiral stationary phase revealed an
excellent enantiomeric excess of 99% for both flavanone (R)-
5b and flavan 11. A palladium-catalyzed deoxygenation[10] of
11 at C-5 then furnished flavan 12 in almost quantitative yield.
In order to experimentally verify the absolute configuration
of 4 and 5b obtained by ATH, we synthesized the corre-
sponding (R) enantiomers starting from (R)-8-prenylnaringe-
nin,[11] the absolute configuration of which is well known.
Comparison of the specific rotation data confirmed the
anticipated (R) configuration of 5b and therefore the (S)
configuration of 4.[18]
An o-quinone methide[19] is believed to be the crucial
intermediate of the ATH shown above. In analogy to the
mechanistic rationale for the deoxygenation of 5-O-acyl
substituted flavanones[13,14] and related substrates[15] with
sodium borohydride, a plausible pathway from flavanone
rac-5b to flavan 4 is illustrated in Scheme 4. Initially, the
ketone is converted into the benzylic alkoxide 13 in a highly
enantioselective fashion. Subsequent migration of the
methoxycarbonyl group to give the phenoxide 14 is followed
by elimination with formation of the reactive o-quinone
methide 15, which eventually undergoes a conjugate reduc-
tion to afford flavan 4.
With flavan 12 in hand, brosimine B (3) can be synthe-
sized by simple cleavage of the carbonate groups (Scheme 5).
While trifluoroacetic acid failed to remove the Boc moieties
cleanly, lithium aluminum hydride reduction gave rise to the
natural product 3 in good yield and virtually enantiopure
form. Starting from brosimine B (3), brosimacutin L (2) was
synthesized in three steps. Formation of the silyl ether 16 and
Scheme 2. Preparation of flavanones rac-5 from naringenin (rac-6).
a) Boc2O, 10 mol% DMAP, Et3N, THF, RT, 1 h, 94%; b) Ph3P, DIAD, 3-
methyl-2-buten-1-ol, THF, RT, 14 h, 95%; c) 10 mol% [Eu(fod)3],
CHCl3, 1108C, 5 min, microwave, 70%; d) for rac-5a: Ac2O, Et3N, THF,
468C, 20 h, 80%, for rac-5b: ClCO2Me, Et3N, THF, RT, 1 h, 89%.
Boc=tert-butoxycarbonyl, DMAP=4-(dimethylamino)pyridine,
DIAD=diisopropyl azodicarboxylate, fod=6,6,7,7,8,8,8-heptafluoro-
2,2-dimethyl-3,5-octanedionate.
room temperature. Next to the substrates rac-5a,b, we also
tested the corresponding racemic tri-O-Boc derivative of 8-
prenylnaringenin[18] as well as rac-5,7,4’-O,O,O-triacetyl-8-
prenylnaringenin. The latter compound had already been
subjected to an analogous ATH reaction with the catalyst
(R,R)-10, from which no benzylic alcohol reduction product
could be isolated.[11] From this set of substrates, the methyl-
carbonate rac-5b turned out to be optimal and led to the most
efficient flavan production (Scheme 3). While separation of
subsequent
regioselective
hydration
by
applying
Mukaiyama[9] conditions provided the tertiary alcohol 17. In
the course of the latter reaction, the enantiomeric excess
dropped slightly to 98%. Finally, cleavage of the silyl ether
furnished brosimacutin L (2). The specific rotation[20] of the
synthetic products 3 and 2 confirmed the (S) configuration
assigned to 2 from CD measurements and assumed for 3.[3–6]
We then utilized Shiꢀs asymmetric epoxidation protocol
and a subsequent 5-exo-tet cyclization to generate flavan
1 (Scheme 6). In accordance with the reported epoxidation of
various prenylated arenes featuring an o-silyloxy function,[21]
Shiꢀs diester 18 proved to be superior to the conventional Shi
acetonide catalyst in terms of diastereomeric ratio, and the
stereoisomer (2S,2’’R)-19 was obtained next to only small
amounts of its C-2’’ epimer. For related prenyl-substituted
substrates, approach of the dioxirane derived from 18 was
shown to occur from the Re face.[21a-c] Eventually, treatment of
epoxide (2S,2’’R)-19 with tetrabutylammonium fluoride[21]
furnished (2S,2’’S)-1 in almost quantitative yield and good
diastereomeric purity (d.r. = 94:6). Unfortunately, separation
of the two epimers by column chromatography was possible
for neither epoxide 19 nor flavan 1. Moreover, the NMR
spectra of (2S,2’’S)-1 and its C-2’’ epimer are very similar.
Therefore, a comparison with the published data of the
Scheme 3. ATH of rac-5b and subsequent conversion to give flavan 12.
a) 1 mol% (S,S)-10, Et3N/HCO2H, EtOAc, RT, 15 min, 44% (R)-5b,
42% 4; b) Tf2O, Et3N, CH2Cl2, À408C, 40 min, 91% 11 (99% ee),
100% (R)-5b (99% ee); c) Pd(OAc)2, dppf, HCO2H, DMF, 2 h, 608C,
98%. Ts =p-toluenesulfonyl, Cp*=1,2,3,4,5-pentamethylcyclopenta-
dienyl, Tf =trifluoromethanesulfonyl, dppf=1,1’-bis(diphenylphosphi-
no)ferrocene.
Angew. Chem. Int. Ed. 2016, 55, 1160 –1163
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1161

Angewandte
Communications
Chemie
Scheme 4. Possible mechanism for the deoxygenation of flavanone 5b.
hydride reduction of (R)-5b to give the flavan ent-4, we
followed the reaction sequence established for (2S,2’’S)-1 and
used the Shi catalyst 18 again. Additionally, we prepared their
enantiomers (2R,2’’R)-1 and (2S,2’’R)-1 by utilizing the Shi
catalyst ent-18.[18] For all four stereoisomers of 1, the optical
rotation was measured. Since the diastereomeric ratio and
enantiomeric excess for these compounds were known from
HPLC on a chiral stationary phase, the specific rotation of the
pure isomers could be calculated from the the data obtained.
Comparison of these values with the specific rotation of the
natural product shows that brosimine A (1) possesses
a (2S,2’’R) configuration.[22] This assignment is in line with
the fact that all three flavans (1–3) were isolated from
Brosimum acutifolium. Hence, brosimine A (1) is presumably
biosynthetically derived from brosimine B (3) and has a (2S)
configuration.
Scheme 5. Synthesis of brosimine B (3) and brosimacutin L (2).
a) LiAlH4, THF, 608C, 50 min, 90% (99% ee); b) TPSCl, imidazole,
DMF, RT, 24 h, 96%; c) 20 mol% Co(acac)2, PhSiH3, O2, THF, RT,
24 h, 56% (98% ee); d) TBAF, THF, 08C, 10 min, 75% (98% ee).
TPS=tert-butyldiphenylsilyl, acac=acetylacetonate, TBAF=tetrabutyl-
ammonium fluoride.
In summary, we have accomplished the first enantiose-
lective synthesis of the three flavans 1, 2, and 3, starting from
commercially available racemic naringenin (rac-6). Central to
our strategy is an ATH/deoxygenation cascade that enables
the single-step conversion of flavanones into flavans in
a highly enantioselective fashion. Brosimine B (3) and
brosimacutin L (2) were obtained with 99% ee and 98% ee,
respectively, in a short and efficient reaction sequence, and
their absolute configuration was unambiguously established.
Likewise, all four stereoisomers of brosimine A (1) were
readily synthesized with > 99% ee, and our results show that
stereoisomer (2S,2’’R)-1 is the natural product.
Keywords: asymmetric catalysis · domino reactions ·
flavonoids · kinetic resolution · natural products
How to cite: Angew. Chem. Int. Ed. 2016, 55, 1160–1163
Angew. Chem. 2016, 128, 1173–1176
[1] E. Elisabetsky, Z. C. Castilhos, Int. J. Crude Drug Res. 1990, 28,
309 – 320.
[2] A. F. Correia, J. F. O. Segovia, M. C. A. GonÅalves, V. L. de O-
liveira, D. Silveira, J. C. T. Carvalho, L. I. B. Kanzaki, Eur. Rev.
Med. Pharmacol. Sci. 2008, 12, 369 – 380.
[3] S. L. Torres, M. S. P. Arruda, A. C. Arruda, A. H. Müller, S. C.
Silva, Phytochemistry 2000, 53, 1047 – 1050.
Scheme 6. Synthesis of the brosimine A stereoisomers (2S,2’’S)-1 and
(2R,2’’S)-1. a) 60 mol% 18, Bu4NHSO4, DMM, MeCN, phosphate
buffer pH 6, K2CO3, oxone, 08C to RT, 11 h, 66%; b) TBAF, THF, 08C
to RT, 30 min, 99% (d.r.=94:6; >99% ee). DMM=dimethoxyme-
thane.
[4] J. Takashima, K. Komiyama, H. Ishiyama, J. Kobayashi, A.
Ohsaki, Planta Med. 2005, 71, 654 – 658.
[5] a) A. F. Teixeira, A. F. de Carvalho Alcântara, D. Piló-Veloso,
Magn. Reson. Chem. 2000, 38, 301 – 304; b) A. F. de Carval-
ho Alcântara, A. F. Teixeira, I. F. da Silva, W. B. de Almeida, D.
Piló-Veloso, Quim. Nova 2004, 27, 371 – 377.
[6] a) X. Hu, J.-W. Wu, M. Wang, M.-H. Yu, Q.-S. Zhao, H.-Y. Wang,
A.-J. Hou, J. Nat. Prod. 2012, 75, 82 – 87; b) Y. Yang, X. Yang, B.
natural product 1 did not reveal conclusive information about
its relative configuration.
With the aim of determining the relative and absolute
configuration of the natural product 1, we also synthesized the
stereoisomer (2R,2’’S)-1. Commencing with a sodium boro-
1162
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1160 –1163

Angewandte
Communications
Chemie
Xu, G. Zeng, J. Tan, X. He, C. Hu, Y. Zhou, Fitoterapia 2014, 98,
[17] O. Mitsunobu, Synthesis 1981, 1 – 28.
222 – 227.
[18] See the Supporting Information for details.
[7] Review: O. Mazimba, N. Keroletswe, IRJPAC 2015, 8, 112 – 146.
[8] O. A. Wong, Y. Shi, Chem. Rev. 2008, 108, 3958 – 3987.
[9] S. Isayama, T. Mukaiyama, Chem. Lett. 1989, 1071 – 1074.
[10] T. A. Engler, K. O. Lynch, Jr., J. P. Reddy, G. S. Gregory, Bioorg.
Med. Chem. Lett. 1993, 3, 1229 – 1232.
[11] a) M.-K. Lemke, P. Schwab, P. Fischer, S. Tischer, M. Witt, L.
Noehringer, V. Rogachev, A. Jäger, O. Kataeva, R. Frçhlich, P.
Metz, Angew. Chem. Int. Ed. 2013, 52, 11651 – 11655; Angew.
Chem. 2013, 125, 11865 – 11869; b) P. Metz, P. Schwab, P. Fischer,
M.-K. Lemke, S. Tischer, DE 11 2007 001 531 B4, 2012.
[12] K. Murata, T. Ikariya, R. Noyori, J. Org. Chem. 1999, 64, 2186 –
2187.
[13] a) J. J. Shie, C.-A. Chen, C.-C. Lin, A. F. Ku, T.-J. R. Cheng, J.-M.
Fang, C.-H. Wong, Org. Biomol. Chem. 2010, 8, 4451 – 4462;
b) K. Oyama, T. Kondo, J. Org. Chem. 2004, 69, 5240 – 5246;
c) J. G. Sweeny, G. A. Iacobucci, Tetrahedron 1977, 33, 2927 –
2932.
[19] W.-J. Bai, J. G. David, Z.-G. Feng, M. G. Weaver, K.-L. Wu,
T. R. R. Pettus, Acc. Chem. Res. 2014, 47, 3655 – 3664.
[20] Synthetic 3: [a]D25 = À63.4 (c = 0.51 in CHCl3) and [a]2D5 = À78.3
(c = 0.59 in MeOH); natural 3: [a]2D5 = À4.88 (c = 0.5 in CHCl3),[3]
[a]2D5 = À30.90 (c = 0.608 in CHCl3) and [a]2D5 = À37.32 (c = 0.608
in MeOH);[5a] synthetic 2: [a]2D5 = À52.4 (c = 0.35 in MeOH);
natural 2:[5a] [a]D22 = À25.3 (c = 0.47 in MeOH).
[21] a) K. M. Aumann, N. L. Hungerford, M. J. Coster, Tetrahedron
Lett. 2011, 52, 6988 – 6990; b) H. Jiang, T. Sugiyama, A.
Hamajima, Y. Hamada, Adv. Synth. Catal. 2011, 353, 155 – 162;
c) H. Jiang, Y. Hamada, Org. Biomol. Chem. 2009, 7, 4173 –
4176; d) J. Magolan, M. J. Coster, J. Org. Chem. 2009, 74, 5083 –
5086.
[22] Synthetic (2S,2’’S)-1: calculated [a]2D5 =+ 5.6 (c = 0.75 in CHCl3)
from the (2’’S)-series of 1 and calculated [a]2D5 =+ 4.3 (c = 0.54 in
CHCl3) from the (2’’R)-series of 1; synthetic (2R,2’’S)-1:
calculated [a]2D5 =+ 83.9 (c = 0.56 in CHCl3) from the (2’’S)-
series of 1 and calculated [a]2D5 =+ 88.4 (c = 0.45 in CHCl3) from
the (2’’R)-series of 1; natural 1:[3] [a]D25 = À7.06 (c = 0.35 in
CHCl3).
[14] D. Mitchell, C. Doecke, L. A. Hay, T. M. Koenig, D. D. Wirth,
Tetrahedron Lett. 1995, 36, 5335 – 5338.
[15] See also: a) B. J. McLoughlin, J. Chem. Soc. D 1969, 540 – 541;
b) K. H. Bell, Aust. J. Chem. 1969, 22, 601 – 605; c) M. A.
Stealey, R. L. Shone, M. Miyano, Synth. Commun. 1990, 20,
1869 – 1876; d) R. W. Van De Water, D. J. Magdziak, J. N. Chau,
T. R. R. Pettus, J. Am. Chem. Soc. 2000, 122, 6502 – 6503.
[16] S. Gester, P. Metz, O. Zierau, G. Vollmer, Tetrahedron 2001, 57,
1015 – 1018.
Received: August 4, 2015
Revised: October 28, 2015
Published online: December 4, 2015
Angew. Chem. Int. Ed. 2016, 55, 1160 –1163
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1163