Asymmetric total synthesis of dragonbloodins A1 and A2

Article abstract of DOI:10.1021/acs.orglett.8b00315

The first asymmetric total synthesis of dragonbloodins A1 and A2, a pair of unprecedented chalcone-flavan heterotrimmers, has been achieved through a series of rationally designed or bioinspired transformations. Key elements of the synthesis include a highly efficient heterotrimerization reaction to assemble the two chalcone units and one flavan unit in one pot and a tandem oxidative dearomatization/cyclization/oxygenation reaction to forge the polycyclic core of dragonbloodins A1 and A2. The present synthesis unambiguously confirms the biogenetic relationship and absolute stereochemistry of dragonbloodins A1 and A2.

Full text of DOI:10.1021/acs.orglett.8b00315

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Asymmetric Total Synthesis of Dragonbloodins A1 and A2
Zhen Guo, Zhiguo Wang, and Yefeng Tang*
School of Pharmaceutical Sciences & Comprehensive AIDS Research Center, Tsinghua University, Beijing 100084, China
S
* Supporting Information
ABSTRACT: The first asymmetric total synthesis of dragonbloodins A1 and A2, a pair of unprecedented chalcone-flavan
heterotrimmers, has been achieved through a series of rationally designed or bioinspired transformations. Key elements of the
synthesis include a highly efficient heterotrimerization reaction to assemble the two chalcone units and one flavan unit in one pot
and a tandem oxidative dearomatization/cyclization/oxygenation reaction to forge the polycyclic core of dragonbloodins A1 and
A2. The present synthesis unambiguously confirms the biogenetic relationship and absolute stereochemistry of dragonbloodins
A1 and A2.
chiral center. Interestingly, it was reported that dragonbloodins
ragon’s blood, also known as Draconis Resina, is a
Drenowned traditional medicine widely used in different A1 and A2 were rather unstable and readily advanced to some
unidentified compounds upon separation from each other.4 By
comparison, the samples consisting of a 1:1 ratio of
dragonbloodins A1 and A2 turned out to be relatively stable,
presumably due to the stabilizing effect of the intermolecular
hydrogen bonding between the two molecules. Notably,
preliminary biological study revealed that dragonbloodins A1
and A2 exerted an inhibitory effect on human neutrophil
superoxide anion generation in a dose-dependent manner. In
addition, they could also inhibit neutrophil elastase release
(IC50 = 6.53 μM), which renders them potential leads for the
development of anti-inflammatory agents.4
Attracted by the novel chemical structures and promising
biological profiles of dragonbloodins A1 and A2, we initiated a
program with the aim of achieving their total synthesis.
However, we were surprised to find that the original isolation
paper was retracted shortly after their disclosure.4 According to
the author’s declaration, the structure of dragonbloodin A2 was
incorrectly assigned due to the misinterpretation of the X-ray
crystal data, although the structure of dragonbloodin A1 should
be correct. While this interlude casted a shadow on our
synthetic program, we quickly realized that a total synthesis of
these targets could provide more conclusive evidence for their
structural assignment. Notably, while our project was ongoing,
an elegant biomimetic total synthesis of dragonbloodins A1 and
A2 was disclosed by the Trauner group.5 In this seminal work,
the structure of dragonbloodin A2 was revised from 2 to 2′ on
the basis of the biogenetic consideration as well as the
comparison of their CD spectra. Compared to dragonbloodin
A1, 2′ bears the same (S)-configuration at the C2″ chiral center
cultures of the world.1 In Chinese medicine, Dragon’s blood is
the major component of the hemostatic preparation “Yun-Nan-
Bai-Yao”, a popular wound-healing agent and a coagulant.
Extensive phytochemical investigations have been carried out
on Dragon’s blood, leading to the discovery of various
compounds that showed diverse biological effects, including
antitumor, antiviral, analgesic, and anti-inflammatory activ-
ities.1,2 Structurally, the major constituents identified from
Dragon’s blood belong to the flavonoid family, existing in either
monomeric or oligomeric forms.3 As a paradigm, two
unprecedented flavan trimers, namely dragonbloodins A1 (1)
and A2 (2), were isolated by Wu and co-workers from the
genus Daemonorops draco in 2016 (Figure 1).4 It was suggested
that dragonbloodins A1 and A2 share a common polycyclic
backbone, only differing in the stereochemistry at the C2″
Received: January 29, 2018
Figure 1. Structures of dragonbloodins A1 and A2.
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b00315
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 1. Proposed Biomimetic Synthetic Strategies for Dragonbloodins A1 and A2
but differs in seven other stereocenters. In this context,
dragonbloodin A2 represents the pseudoenantiomer of
dragonbloodin A1 instead of the originally proposed C2″-
epimier. While such rationalization appears to be reasonable, a
more convincing evidence to verify the identity of dragon-
bloodins A1 and A2 could be obtained only through an
asymmetric synthesis, which allows for the direct comparison of
the specific rotation of synthetic and natural samples.
rise to the corresponding bridged compounds 10a and 10b.
Upon exposure to suitable oxidant, 10a and 10b could undergo
tandem oxidative dearomatization/cyclization/hemiketalization
to afford dragonbloodins A1 and A2, respectively.8
Our synthesis commenced with the preparation of the
chalcone unit (Scheme 2A). Thus, selective acetylation of 2,4-
Scheme 2. Syntheses of the Chalcone and Flavan Units
It is worth noting that although Trauner and co-workers
achieved the asymmetric synthesis of a key intermediate en
route to dragonbloodins A1 and A2, they did not complete the
asymmetric total syntheses. Keeping this in mind, we continued
to push our project forward, which has culminated in the first
asymmetric total synthesis of dragonbloodins A1 and A2. On
the basis of this work, we unambiguously verified the identity of
dragonbloodins A1 and A2, particularly regarding their absolute
configuration.
From a structural point of view, dragonbloodins A1 and A2
consist of two units of chalcone 3 and one unit of flavan 4.
Apparently, the most efficient way to access these targets
should build on a biomimetic heterotrimerization reaction.
Keeping this in mind, we designed two distinct strategies
(Scheme 1). The first one features a [3 + 3 + 4]-mode (path a).
In this scenario, the chalcone 3 could undergo homodimeriza-
tion through a photoinduced [2 + 2]-cycloaddition9 to give the
cyclobutane derivative ( )-5 that would readily advance to
( )-6 through hemiketalization. Upon treatment with suitable
oxidant, ( )-6 could undergo tandem oxidative dearomatiza-
tion/cyclobutane ring-expansion10 to generate the carbon
cation I-2. Finally, I-2 could be trapped with (−)-4 through
sequential nucleophilic attack and ketalization to give
dragonbloodins A1 and A2.
Alternatively, we assumed that the heterotrimerization
reaction could also proceed through a [3 + 4 + 3]-mode
(path b). Thus, the chalcone 3, upon irradiation or treatment
with acidic conditions, could undergo sequential double-bond
isomerization and hemiketalization to give the bicycle 7.6 The
hemiketal 7 readily advances to the benzopyrylium inter-
mediate 8 through dehydration, and the latter could further
undergo nucleophilic addition with flavan (−)-4 to from two
diastereomeric heterodimers 9a and 9b.7 Both 9a and 9b could
react with the second unit of benzopyrylium 8 through
sequential nucleophilic addition and ketalization, thus giving
dihydroxy-3-methylbenzaldehyde 1111 afforded 12 in 67%
yield, which after methylation, provided the fully protected
aldehyde 13. To introduce the third phenolic hydroxyl, we
resorted to an interesting method developed by Ackermann
and co-workers, that is, the aldehyde-assisted Ru(II)-catalyzed
C−H oxygenation reaction.12 Gratifyingly, the transformation
proceeded smoothly, providing the desired product 14 in a
reasonable yield. Of note, the reaction could be improved by
replacing the acetyl group to sulfonyl group (Ts- or Ns-)
( T a b l e S 1 ) . F i n a l l y , t r e a t m e n t o f 1 4 w i t h
(triphenylphosphoranylidene)acetophenone in refluxing tol-
uene provided the chalcone derivative 15a smoothly. In
parallel, 15b and 15c, two other protected chalcone derivatives,
were also prepared on the basis of the above procedure (for
details, see the SI).
B
DOI: 10.1021/acs.orglett.8b00315
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
With the requisite chalcone unit in hand, we moved to
achieve the asymmetric synthesis of the flavan unit (Scheme
2B). To this end, the aldehyde 1613 was first converted to the
chalcone 17 through acetylation, followed by Wittig olefination.
Next, reduction of 17 through catalytic hydrogenation gave the
ketone 18. Subsequently, an asymmetric reduction was effected
with the chiral reagent (+)-DIP chloride, which gave the
secondary alcohol 19 with excellent yield (91%) and
enantioselectivity [(R)-configuration, 97% ee].14 Finally,
removal of the acetyl group followed by intramolecular
Mitsunobu reaction afforded the desired flavan unit (−)-4
smoothly, with the C2 stereochemistry being inverted to (S)-
configuration.15
resorcinols.17 Inspired by this work, we attempted to effect
the heterotrimerization of 22 and (−)-4 under identical
conditions (PTSA, toluene, 110 °C) (Scheme 4). Gratifyingly,
Scheme 4. Heterotrimerization of Chalcone and Flavan
Units
Having both chalcone and flavan units secured, we turned to
explore the bioinspired heterotrimerization reaction. Initially,
the [3 + 3 + 4]-mode (path a, Scheme 1) was attempted. On
the basis of our design, the first step is to realize the
homodimerization of the two chalcone units through a [2 + 2]-
photocycloaddition. To our surprise, although there exist some
relevant precedents,9d−g such transformation was much more
challenging than our expectation. Indeed, we evaluated a
number of potential substrates (e.g., 15a−c) under various
conditions involving different light sources (Hg lamp, UV lamp,
and sunlight), solvents (DCM, THF, acetone, CH3CN,
hexanes, and neat), and additives [Cu(OTf)2 and benzophe-
none] (Scheme 3). Unfortunately, we failed to get promising
a small amount of the expected heterotrimers 23a/23b (ca.
15% yield) was identified in the reaction, which existed as an
inseparable diastereoisomers in a ratio of 1.2:1. Additionally,
the heterodimers 24a/24b (1:1) were isolated in 70%
combined yield. Encouraged by this result, we performed a
comprehensive condition optimization to improve the yield of
the heterotrimers (Table S2). It was found that both the
solvent effect and reaction temperature played crucial roles in
the reaction. Eventually, the best result was obtained when the
reaction was performed in THF at room temperature using 3.0
equiv of 6a, which delivered a mixture of 23a/23b in 80%
combined yield, favoring 23b as the major isomer (23a/23b =
1:2). More interestingly, based on the above results, a
operationally simple one-pot protocol was developed to effect
the photoinduced double-bond isomerization/hemiketalization
and subsequent acid-promoted heterotrimerization, which
enabled the direct access of 23a/23b from 15a and (−)-4 in
a single operation with high efficiency (67% yield). Notably, no
heterotrimerization products could be obtained in the reaction
without irradiation.
Scheme 3. Attempts to Effect Homodimerization of the
Chalcone Units through [2 + 2]-Photocycloaddition
results. In most cases, the starting material remained
unchanged. Compared to those precedents,9 we assume that
the poor reactivity in our cases could be attributed to the
notable steric effect associated with substrates, which precludes
two monomers from approaching each other. Of note, this
hypothesis was validated by the following observation: the less
substituted chalcone 15d readily underwent [2 + 2]-photo-
cycloaddition upon irradiation (365 nm) in its solid state,16
which gave rise to the head-to-tail dimerization product 21d in
90% yield. Notably, the structure of 21d was confirmed by the
X-ray crystallographic study of its deprotection product 21d′.
Although the proposed [2 + 2]-photocycloadditions failed to
give satisfactory results, we found that 15a, upon photo-
irradiation, could undergo sequential double-bond isomer-
ization and hemiketalization to give the bicycle 22 in 90% yield.
This discovery paved the way to explore the alternative
trimerization reaction through the [3 + 4 + 3]-mode (path b,
Scheme 1). Recently, Lee and co-workers reported an acid-
catalyzed heterodimerization of 2-hydroxychalcones and
With 23a/23b in hand, we moved to complete the total
synthesis of dragonbloodins A1 and A2. For this end, the acetyl
group of 23a/23b was removed with the action of NaOMe/
MeOH. In agreement with the observation reported by
Trauner and co-workers,5 the resulting products 10a/10b
were sensitive to air, and partially converted to the hydroperoxy
25a/25b through auto-oxidation during the course of
chromatographic purification. However, we found that the
isolated yield of 25a/25b varied in different experiments. After
several tries, we developed a more reliable and reproducible
protocol to effect the tandem oxidative dearomatization/
intramolecular cyclization/oxygenation reaction. Thus, direct
treatment of the newly generated 10a/10b with silver oxide
under air furnished 25a/25b in 63% yield.8c Finally, reduction
of 25a/25b with dimethyl sulfide led to the formation of
dragonbloodins A1/A2. At this stage, the ratio of dragon-
bloodins A1 and A2 was 1:2. Interestingly, we found that this
sample was inclined to precipitate in EtOAc, forming a
precipitate consisting of a 1:1 ratio of dragonbloodins A1 and
C
DOI: 10.1021/acs.orglett.8b00315
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
A2. As a result, the mother liquid showed an increased
proportion of dragonbloodin A2. After repeating such
operation twice, we obtained the pure dragonbloodins A2 for
structural characterization. Gratifyingly, the spectroscopic data
(1H and 13C NMR) of synthetic dragonbloodins A1 and A2
were identical with those reported for the natural samples.
Moreover, the optical rotation of synthetic dragonbloodin A2
([α]D −54.72, c = 0.27, CHCl3) was also in good agreement
with the reported one ([α]D −58.08, c = 0.06, CHCl3).4,18
Thus, the absolute configuration of dragonbloodin A2 could be
unambiguously assigned as shown in the structure 2′ (Scheme
5). Naturally, the identity of dragonbloodin A1 (1) is confirmed
as well.
request@ccdc.cam.ac.uk, or by contacting The Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge CB2
1EZ, UK; fax: +44 1223 336033.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: yefengtang@tsinghua.edu.cn.
ORCID
Yefeng Tang: 0000-0002-6223-0608
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Scheme 5. Total Synthesis of Dragonbloodins A1 and A2
We gratefully acknowledge financial support from the National
Natural Science Foundation of China (21572112, 21772109)
and Beijing Natural Science Foundation (2172026).
REFERENCES
■
(1) Gupta, D.; Bleakley, B.; Gupta, R. K. J. Ethnopharmacol. 2008,
115, 361.
(2) For selective examples, see (a) Zhu, Y.; Zhang, P.; Yu, H.; Li, J.;
Wang, M.-W.; Zhao, W. J. Nat. Prod. 2007, 70, 1570. (b) Shen, C.-C.;
Tsai, S.-Y.; Wei, S.-L.; Wang, S.-T.; Shieh, B.-J.; Chen, C.-C. Nat. Prod.
Res. 2007, 21, 377. (c) Hao, Q.; Saito, Y.; Matsuo, Y.; Li, H.-Z.;
Takashi, T. Nat. Prod. Res. 2015, 29, 1419. (d) Kuo, P.-C.; Hung, H.-
Y.; Hwang, T.-L.; Du, W.-K.; Ku, H.-C.; Lee, E.-J.; Tai, S.-H.; Chen, F.-
A.; Wu, T.-S. J. Nat. Prod. 2017, 80, 783.
(3) (a) Brockmann, H.; Haase, R. Ber. Dtsch. Chem. Ges. B 1936, 69,
1950. (b) Nasini, G.; Arnone, A.; Merlini, L. Heterocycles 1989, 29,
1119. (c) Arnone, A.; Nasini, G. J. Chem. Soc., Perkin Trans. 1 1990,
2637. (d) Arnone, A.; Pava, O. V.; Nasini, G. J. Nat. Prod. 1997, 60,
971. (e) Nakashima, K.; Abe, N.; Kamiya, F.; Ito, T.; Oyama, M.;
Iinuma, M. Helv. Chim. Acta 2009, 92, 1999.
(4) Du, W.-K.; Hung, H.-Y.; Kuo, P.-C.; Hwang, T.-L.; Shiu, L.-C.;
Shiu, K.-B.; Lee, E.-J.; Tai, S.-H.; Wu, T.-S. Org. Lett. 2016, 18, 3042.
Retracted on June 7, 2016.
(5) Schmid, M.; Trauner, D. Angew. Chem., Int. Ed. 2017, 56, 12332.
(6) (a) Hsiao, C.; Liao, H.; Sugiono, E.; Atodiresei, I.; Rueping, M.
Chem. - Eur. J. 2013, 19, 9775. (b) Li, B.; Williams, B. D.; Smith, A. B.,
III. Org. Lett. 2015, 17, 3.
(7) (a) Yang, Z.; He, Y.; Toste, F. D. J. Am. Chem. Soc. 2016, 138,
9775. (b) Chand Ganguly, N.; Roy, S.; Mondal, P. RSC Adv. 2014, 4,
42078. (c) Selenski, C.; Pettus, T. R. Tetrahedron 2006, 62, 5298.
(8) (a) Callinan, A.; Chen, Y.; Morrow, G. W.; Swenton, J. S.
Tetrahedron Lett. 1990, 31, 4551. (b) Andrez, J.-C.; Giroux, M.-A.;
Lucien, J.; Canesi, S. Org. Lett. 2010, 12, 4368. (c) Yang, H. Z.; Feng,
J.; Tang, Y. F. Chem. Commun. 2013, 49, 6442. (d) Yang, H. Z.; Liu,
X.; Li, Q. G.; Li, L. B.; Zhang, J. R.; Tang, Y. F. Org. Biomol. Chem.
2016, 14, 198.
In summary, we have achieved the first asymmetric synthesis
of dragonbloodins A1 and A2, a pair of unprecedented
chalcone-flavan heterotrimmers. The key elements of our
synthesis include (1) an aldehyde-assisted Ru(II)-catalyzed C−
H oxygenation to access the chalcone unit, (2) a bioinspired
heterotrimerization to unite the two chalcone and one flavan
units together, and (3) a tandem oxidative dearomatization/
cyclization/oxygenation reaction to forge the polycyclic core of
dragonbloodins A1 and A2. The present synthesis provides
unequivocal evidence for the structural determination of
dragonbloodins A1 and A2. Given that the structural revision
of naturally occurring dragonbloodins A1 and A2 has remained
yet to be made by the isolation team, our study may provide
valuable information to those synthetic chemists who are now
working on these targets.
(9) For leading reviews, see: (a) Xu, Y.; Conner, M. L.; Brown, M. K.
Angew. Chem., Int. Ed. 2015, 54, 11918. (b) Karkas, M. D., Jr.; Porco, J.
̈
̈
A.; Stephenson, C. R. J. Chem. Rev. 2016, 116, 9683. (c) Poplata, S.;
ASSOCIATED CONTENT
* Supporting Information
■
Troster, A.; Zou, Y. Q.; Bach, T. Chem. Rev. 2016, 116, 9748. For
̈
S
relevant examples, see (d) Toda, F.; Tanaka, K.; Sekikawa, A. J. Chem.
Soc., Chem. Commun. 1987, 0, 279. (e) Seguin, E.; Elomri, A.; Magiatis,
P.; Skaltsounis, A. L.; Chao, L. R.; Tillequin, F. Nat. Prod. Lett. 2002,
16, 187. (f) Karpuk, E.; Schollmeyer, D.; Meier, H. Eur. J. Org. Chem.
2007, 2007, 1983. (g) Pagire, S. K.; Hossain, A.; Traub, L.; Kerres, S.;
Reiser, O. Chem. Commun. 2017, 53, 12072.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b00315.
1
Experimental details, characterization; H and 13CNMR
spectra for all newly synthesized compounds (PDF)
(10) (a) Avasthi, K.; Salomon, R. G. J. Org. Chem. 1986, 51, 2556.
(b) Bach, T.; Spiegel, A. Eur. J. Org. Chem. 2002, 2002, 645.
(c) Mandelt, K.; Meyer-Wilmes, I.; Fitjer, L. Tetrahedron 2004, 60,
11587.
Accession Codes
CCDC 1819926 contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
via www.ccdc.cam.ac.uk/data_request/cif, or by emailing data_
(11) Tasior, M.; Gryko, D.; Pielacin
Chem. - Asian J. 2010, 5, 130.
́
D
DOI: 10.1021/acs.orglett.8b00315
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(12) Yang, F.; Rauch, K.; Kettelhoit, K.; Ackermann, L. Angew. Chem.,
Int. Ed. 2014, 53, 11285.
(13) Li, C. C.; Xie, Z. X.; Zhang, Y. D.; Chen, J. H.; Yang, Z. J. Org.
Chem. 2003, 68, 8500.
(14) (a) Brown, H.; Chandrasekharan, J.; Ramachandran, P. V. J. Am.
Chem. Soc. 1988, 110, 1539. (b) Saravanan, M.; Satyanarayana, B.;
Reddy, P. P. Synth. Commun. 2013, 43, 2050.
(15) (a) Bachki, A.; Foubelo, F.; Yus, M. Tetrahedron Lett. 1998, 39,
7759. (b) Palmer, A. M.; Munch, G.; Grobbel, B.; Kromer, W.
̈
Tetrahedron Lett. 2009, 50, 3917. (c) Orr, R.; Campeau, L.-C.;
Chobanian, H.; Pio, B.; Plummer, C.; Nolting, A.; Ruck, R. Synthesis
2017, 49, 657.
(16) For a review of the photochemical reactions of organic crystals,
see: (a) Ramamurthy, V.; Venkatesan, K. Chem. Rev. 1987, 87, 433.
For selected examples, see: (b) Zhang, P. T.; Wang, Y. G.; Bao, R. Y.;
Luo, T. P.; Yang, Z.; Tang, Y. F. Org. Lett. 2012, 14, 162. (c) Tan, H.
B.; Zheng, C.; Liu, Z.; Wang, D. Z. Org. Lett. 2011, 13, 2192.
(d) Shang, H.; Liu, J. H.; Bao, R. Y.; Cao, Y.; Zhao, K.; Xiao, C. Q.;
Zhou, B.; Hu, L. H.; Tang, Y. F. Angew. Chem., Int. Ed. 2014, 53,
14494.
(17) Xia, L.; Cai, H.; Lee, Y. R. Org. Biomol. Chem. 2014, 12, 4386.
(18) The NMR spectroscopic data and optical rotation of natural
samples were provided in the Supporting Information of the retracted
paper (see ref 4).
E
DOI: 10.1021/acs.orglett.8b00315
Org. Lett. XXXX, XXX, XXX−XXX