Total synthesis of selaginellin A

Article abstract of DOI:10.1016/j.tetlet.2021.153295

The total synthesis of selaginellin A, isolated from Selaginella tamariscina, was achieved through an eight-step process including a Diels-Alder reaction, formation of a quinone methide moiety and dehydrogenative aromatization. The NMR spectra collected for different samples of synthetic selaginellin A initially showed irregular peak shapes and variable chemical shifts for the signals of the phenol-substituted quinone methide moiety under neutral conditions. After detailed NMR analysis, it was found that the addition of a trace amount of trifluoroacetic acid accelerated the tautomerism of the phenol-substituted quinone methide moiety and improved the reproducibility of the chemical shifts of synthetic selaginellin A.

Full text of DOI:10.1016/j.tetlet.2021.153295

Tetrahedron Letters 79 (2021) 153295
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Total synthesis of selaginellin A
a
b
Kenshu Fujiwara a, , Takaya Itagaki , Tetsuo Tokiwano
⇑
a Department of Life Science, Graduate School of Engineering Science, Akita University, Akita 010-8502, Japan
b Department of Biotechnology, Faculty of Bioresource Sciences, Akita Prefectural University, Akita 010-0195, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The total synthesis of selaginellin A, isolated from Selaginella tamariscina, was achieved through an eight-
step process including a Diels-Alder reaction, formation of a quinone methide moiety and dehydrogena-
tive aromatization. The NMR spectra collected for different samples of synthetic selaginellin A initially
showed irregular peak shapes and variable chemical shifts for the signals of the phenol-substituted qui-
none methide moiety under neutral conditions. After detailed NMR analysis, it was found that the addi-
tion of a trace amount of trifluoroacetic acid accelerated the tautomerism of the phenol-substituted
quinone methide moiety and improved the reproducibility of the chemical shifts of synthetic selaginellin
A.
Received 16 May 2021
Revised 24 June 2021
Accepted 13 July 2021
Available online 27 July 2021
Keywords:
Natural product synthesis
Diels-Alder reaction
Natural polyphenol
Total synthesis
Ó 2021 Elsevier Ltd. All rights reserved.
Dehydrogenative aromatization
Selaginellin A (1) (Fig. 1), isolated from Selaginella tamariscina
by Ma and co-workers as a natural pigment [1], is a member of
the selaginellin family of natural polyphenols [2] that commonly
have a congested 2,3-disubstututed diphenyl skeleton. The struc-
ture of 1, characterized by a contiguously trisubstituted benzene
possessing a 2-(4-hydroxyphenyl)ethynyl group at C14, a 4-
hydroxyphenyl group at C18 and a (4-hydroxyphenyl)(4-oxocyclo-
hexa-2,5-dien-1-ylidene)methyl group at C19, was unambiguously
confirmed by X-ray crystallography. However, it was reported that
the NMR spectra of 1 were quite different from those of selaginel-
lin, a natural congener of 1 [3], and that this may have been caused
by the tautomerism of 1 with the contribution of active hydrogen
[1]. The broadened signals in the NMR spectra of 1 shown in the of
Yang’s isolation [4] also illustrates the challenges in reproducibility
of the structural assignment of 1 by NMR analysis. Since we have
been interested in the unique structures of the selaginellin family
of natural products, we started our research with their total syn-
thesis. Previously, we synthesized a simple congener, selaginellin
S [5,6], by a process including a regioselective Diels-Alder reaction
and dehydrogenative aromatization for the construction of the
central tetrasubstituted benzene [7]. Herein, we describe the
development of an assembly method for the (4-hydroxyphenyl)
(4-oxocyclohexa-2,5-dien-1-ylidene)methyl group to complete
the total synthesis of 1. We also report that the addition of a trace
amount of trifluoroacetic acid (TFA) accelerated the tautomerism
of 1 and improved the reproducibility of the NMR chemical shifts.
Our synthetic plan for 1, shown in Scheme 1, was based on
regioselective construction of the central benzene ring as reported
in our previous total synthesis of selaginellin S [7]. Thus, the Diels-
Alder reaction between enone 5, in which the 4-bromobenzoyl
group was installed to enhance its reactivity, and diene 6 was
scheduled at the initial stage. The Diels-Alder reaction was
expected to produce cyclohexene 4 regioselectively, which would
then be reacted with aryllithium 3 to provide 2 having the full car-
bon skeleton of 1. The successful addition of 3 to 4 relied on the
reduced steric hindrance of the alkynyl and aryl groups at C14
and C18 in the cyclohexene framework of 4. For the completion
of the synthesis of 1 from 2, a process including quinone methide
formation, dehydrogenative aromatization of the cyclohexene moi-
ety, and hydroxylation at C10 was employed. The success of the
final process required finding the proper order of these steps.
The synthesis of cyclohexene 4 via the Diels-Alder reaction of
enone 5 and diene 6 is shown in Scheme 2. Enone 5 was prepared
from aldehyde 7 [8] and 4-bromoacetophenone by a previously
reported procedure (64% over 2 steps) [7]. Diene 6 was obtained
from known alcohol 8 [9] by mesylation followed by elimination
(64% from 8). The Diels-Alder reaction between 5 and 6 was per-
formed in the presence of 3,5-di-tert-butyl-4-hydroxytoluene
(BHT) in o-dichlorobenzene at reflux (180 °C) for 1 h. As a result,
cyclohexenes 4a (24%) and 4b (15%) were regioselectively pro-
duced along with a trace amount of regioisomer 9 (<5%).
With the desired cyclohexenes 4a and 4b in hand, we next
examined the installation of the (4-(tert-butyldimethylsilyl)oxy)
phenyl group to C7 after aromatization of the cyclohexene cores
of 4a and 4b. The dehydrogenative aromatization of 4a and 4b with
⇑
Corresponding author.
E-mail address: fjwkn@gipc.akita-u.ac.jp (K. Fujiwara).
https://doi.org/10.1016/j.tetlet.2021.153295
0040-4039/Ó 2021 Elsevier Ltd. All rights reserved.

K. Fujiwara, T. Itagaki and T. Tokiwano
Tetrahedron Letters 79 (2021) 153295
Br
Br
O
O
R
R
CHO
f
Br
a, b
4a (24%)
O
R
R
R
Br
g
O
TBSO
TBSO
R
10
R
7
5a
e
h
4b (15%)
OH
Br
OH
c, d
Fig. 1. Selaginellin A (1) and congeners.
Br
R
R
R
TBSO
TBSO
(i) Quinone Methide Formation
(ii) Aromatization
(i)
(iii)
8
6
1
10
HO
O
R
O
11
(iii) Hydroxylation
7
30
TBSO
HO
Br
22 OH
9
O
10
Br
OTBS
OTBS
(ii)
1
OTBS
OH
7H
TBSO
R=
H
5b
R
H
Scheme 2. Reagents and conditions: (a) 4-bromoacetophenone, LDA, THF, –78 °C,
15 min, then 7, –78 °C, 30 min, 85%; (b) MsCl, Et3N, CH2Cl2, 0 °C, 5 min, 75%; (c)
MsCl, Et3N, CH2Cl2, 0 °C, 1 h; (d) DBU, CH2Cl2, 23 °C, 2.5 h, 64% from 8; (e) BHT, o-
dichlorobenzene, reflux (180 °C), 1 h, 4a: 24%, 4b: 15%; (f) DDQ, benzene, 60 °C,
11 h, 53%; (g) DDQ, benzene, 60 °C, 11 h, 68%; (h) 1-((tert-butyldimethylsilyl)oxy)-
4-iodobenzene, BuLi, THF, –78 °C, 10 min, then 10, –78 °C.
2
OTBS
OTBS
O
OTBS
Li
19
5
3
Br
14
TBSO
18
15
7
O
H
14
15
H
6
Diels-Alder
Reaction
18
Br
OH
R
H
R
4
7
TBSO
R
R
a
b
4a
4b
Br
c
Scheme 1. Synthetic plan for 1.
Br
OH
7
7
R
R
2a
R
13
14
O
R
Br
R
2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) afforded 10 in
53% and 68% yield, respectively. However, ketone 10 barely reacted
with aryllithium 3, prepared from 1-((tert-butyldimethylsilyl)oxy)-
4-iodobenzene and BuLi, to give addition product 11, which might
be attributable to the close proximity of the alkynyl and aryl
groups to the ketone moiety in the same benzene ring of 10. A sim-
ilar observation in the synthesis of selaginellin analogues was
reported by Wang and co-workers [10]. Therefore, ketones 4a
and 4b were first treated with 3 with the expectation of decreased
steric hindrance of the alkynyl and aryl groups in the cyclohexene
framework of 4.
d
OH
R
R
R
7
12
OTBS
2b
R=
Scheme 3. Reagents and conditions: (a) 1-((tert-butyldimethylsilyl)oxy)-4-
iodobenzene, BuLi, THF, –78 °C, 10 min, then 4a, –78 °C, 10 min, 100%; (b) 1-
((tert-butyldimethylsilyl)oxy)-4-iodobenzene, BuLi, THF, –78 °C, 10 min, then 4b, –
78 °C, 10 min, 94%; (c) DDQ, benzene, 60 °C; (d) DDQ, benzene, 60 °C.
The addition reactions of aryllithium 3 to ketones 4a and 4b in
THF at –78 °C produced alcohols 2a (100%) and 2b (94%), respec-
tively (Scheme 3). Thus, the construction of the full carbon skeleton
of 1 was achieved.
Next, the dehydrogenative aromatization reactions of 2a and 2b
with DDQ were examined. However, both reactions failed to afford
aromatized product 12 and only gave benzophenone 13 and biphe-
nyl 14 as byproducts. Since the production of 13 was attributable
to the hydroxy group at C7 in 2a and 2b, which would result in a
retro-aldol type elimination of 13 during dehydrogenation with
DDQ, the aromatization was rescheduled after the formation of a
quinone methide by the elimination of the C7 hydroxy group.
Quinone methide formation was achieved by treatment of 2a
with p-toluenesulfonic acid (TsOH) in benzene at 60 °C to give
15a, which was then reacted with DDQ to afford aromatized pro-
duct 16 (54% over 2 steps) (Scheme 4). Diastereomer 2b was also
converted to 16 by the same process (37% over 2 steps).
2

K. Fujiwara, T. Itagaki and T. Tokiwano
Tetrahedron Letters 79 (2021) 153295
Br
OTBS Br
O
OH
R
R
R
R
R
a
b
Br
O
2a
15a
OTBS
R
R=
R
Br
OTBS Br
O
OH
16
R
R
d
c
R
2b
15b
Scheme 4. Reagents and conditions: (a) TsOHÁH2O, benzene, 60 °C, 2 h; (b) DDQ,
benzene, 60 °C, 11 h, 54% from 2a; (c) TsOHÁH2O, benzene, 60 °C, 2 h; (d) DDQ,
benzene, 60 °C, 11 h, 37% from 2b.
Scheme 5. Reagents and conditions: (a) bis(pinacolato)diboron, KOAc, PdCl2(dppf)ꢀ
CH2Cl2 (10 mol%), DMSO, 80 °C, 22 h; (b) 30% aqÁH2O2, 5 mol L–1 aq. NaOH, THF,
23 °C, 2 h, 71% from 16.
The final hydroxylation at C10 of 16 was performed according
to the procedure employed in our previous synthesis of selaginellin
S [7] (Scheme 5). When bromide 16 was treated with bis(pinaco-
lato)diboron in the presence of PdCl2(dppf)ꢀCH2Cl2 in DMSO at
80 °C [11], substitution of a boronic ester group at C10 along with
concomitant removal of TBS ethers occurred to produce 17. Oxida-
tion of the boronic ester moiety of 17 with NaOH and H2O2 fur-
nished
1 (71% over 2 steps). Thus, the total synthesis of
selaginellin A (1) was achieved over 8 steps from aldehyde 7 in
8.3% yield.
During the NMR analysis of synthetic 1, we initially observed
that the chemical shifts moved unpredictably each time the spec-
trum was measured. For example, when the NMR spectra of inde-
pendently prepared samples of 1 were measured in methanol d4
(Fig. 2), the signals of H3, H5, H8 and H12 (indicated by black
arrows) as well as H2, H6, H9 and H11 (indicated by outlined
arrows) showed broad peaks [(a) and (b)] or sharp peaks [(c)] with
variable chemical shifts. Since the influence of the tautomerism of
1 on the chemical shift was suggested by Ma’s report [1], we
deduced that the broadening of the signals was caused by the slow
tautomerism of 1 under neutral conditions. Therefore, the intro-
duction of a trace amount of TFA to the NMR samples of 1 was
examined to accelerate the tautomerism and to stabilize the peak
shapes and chemical shifts. For convenience, a trace amount of
TFA was introduced using a thin glass capillary tube to a solution
of 1 in an NMR tube [12]. As a result, the NMR signals of 1 showed
consistent chemical shifts with good reproducibility for subse-
quent measurements [(d)] [13]. However, the addition of an excess
amount of TFA was found to induce a high-frequency shift of the
signals of the phenol-substituted quinone methide moiety of 1
Fig. 2. 1H NMR spectra of independent samples of 1 (500 MHz, methanol d4, 22 °C).
Chemical shifts are based on the residual solvent peak (3.31 ppm). Black arrows
indicate the signals of H3, H5, H8 and H12. Outlined arrows indicate the signals of
H2, H6, H9 and H11. (a) Sample 1. (b) Sample 2. (c) Sample 3. (d) Sample 4 with a
trace amount of TFA. (e) Sample 5 with a slightly larger amount of TFA than that of
(d).
[(e)], which might be attributable to the equilibrium shift caused
by the increasing protonated species of 1. Therefore, the addition
of TFA should be limited to the minimum amount required for
stimulating tautomerism.
After establishing the conditions for NMR analysis, the structure
of synthetic 1 was independently determined to be the same as
that reported for the natural selaginellin A (see supporting infor-
mation). The NMR chemical shifts for synthetic 1 [Fig. 2(d) Sample
4 with a trace amount of TFA] were also compared with those for
natural selaginellin A [1] (Table 1). The NMR data for natural sela-
ginellin A showed a symmetry-equivalence of the phenol-substi-
tuted quinone methide moiety. This suggests that the NMR data
3

K. Fujiwara, T. Itagaki and T. Tokiwano
Tetrahedron Letters 79 (2021) 153295
Table 1
Comparison between the NMR chemical shifts of 1 in this work and those in the literature.1.
1H NMR [methanol d4] d
13C NMR [methanol d4] d
Position
Lit.1
1 (synthetic)a
D
dLit.–synth.
Lit.1
Suggested correction
1 (synthetic)b
DdLit.–synth.
1
1
1
2
3
4
5
6
7
8,12
9,11
10
13
14
15
16
17
18
19
20,24
21,23
22
25
26
27
28,32
29,31
30
165.6
122.3
138.9
131.4
138.9
122.3
133.0
138.9
122.3
165.6
131.4
126.4
131.4
130.6
131.0
145.0
142.3
131.0
115.7
158.0
115.7
87.2
? missed
175.2
122.0
139.2
131.9
139.2
122.0
167.4
139.2
122.0
175.2
131.9
126.5
131.5
130.9
131.1
145.1
142.1
131.0
115.8
158.0
132.9
87.1
─
0.3
6.47
7.08
6.58
7.18
–0.11
–0.10
–0.3
–0.5
–0.3
0.3
–1.8
–0.3
0.3
7.08
6.47
7.18
6.58
–0.10
–0.11
? 165.6
7.08
6.47
7.18
6.58
–0.10
–0.11
? missed
─
–0.5
–0.1
–0.1
–0.3
–0.1
–0.1
0.2
0.0
–0.1
0.0
0.1
0.1
–0.2
0.0
0.0
7.55
7.45
7.26
7.62
7.54
7.34
–0.07
–0.09
–0.08
6.72
6.48
6.78
6.54
–0.06
–0.06
? 133.0
95.2
95.4
6.91
6.56
6.96
6.62
–0.05
–0.06
134.1
116.4
159.4
114.6
134.1
116.4
159.4
114.5
0.0
0.1
33
a
Chemical shifts are based on the residual solvent peak at 3.31 ppm.
Chemical shifts are based on the solvent peak at 49.0 ppm.
b
resulted from a relatively rapid tautomerism of 1. The differences
(d
) in the 1H NMR chemical shifts between 1 and the natural
D
compound were relatively large, with values in the range of
À0.11 to À0.06 ppm. However, the negative values of all differ-
ences suggest the presence of systematic deviations, which may
result from the choice of the reference signal in the measurement
of the natural compound. Therefore, the net deviation width is esti-
mated to be 0.06 ppm or 0.03 ppm, which would be small enough
to identify the structure. All chemical shifts in the 13C NMR spec-
trum (methanol d4) of natural selaginellin A, except C1, C7, C10,
and C25, agreed well with those for synthetic 1 within the d
D
range of –0.5 to 0.3 ppm. It was also determined that the chemical
shifts of C7 and C25 in the natural data should be corrected to
165.6 ppm and 133.0 ppm, respectively, and that the signals of
C1 and C10 of natural selaginellin A were missing. The misassign-
ment of C1, C7 and C10 was likely due to the weakness of the sig-
nals. In our 13C NMR measurement of synthetic 1, these signals
were also observed as weak peaks (Fig. 3), and, therefore, the
chemical shifts of the signals were confirmed by HMBC, for which
correlations of H3/5/8/12 with C7 and C1/10 were clearly observed.
Thus, synthetic selaginellin A (1) provided NMR data that comple-
mented the literature NMR data for natural selaginellin A.
In conclusion, the total synthesis of selaginellin A, isolated from
Selaginella tamariscina, was achieved through an eight-step process
including a Diels-Alder reaction, formation of a quinone methide
moiety and dehydrogenative aromatization. The NMR spectra of
synthetic selaginellin A initially showed irregular peak shapes
and variable chemical shifts of the signals of the phenol-substi-
tuted quinone methide moiety for all samples under neutral condi-
tions. After detailed NMR analysis, it was found that the addition of
a trace amount of trifluoroacetic acid accelerated the tautomerism
of the phenol-substituted quinone methide moiety and improved
the reproducibility of the chemical shifts of synthetic selaginellin
A.
Fig. 3. HMBC correlations between H3/5/8/12 and C7 and between H3/5/8/12 and
C1/10 of 1 (500 MHz, methanol d4, 22 °C, sample 4 with a trace amount of TFA).
4

K. Fujiwara, T. Itagaki and T. Tokiwano
Tetrahedron Letters 79 (2021) 153295
Declaration of Competing Interest
References
[1] X.L. Cheng, S.C. Ma, J.D. Yu, S.Y. Yang, X.Y. Xiao, J.Y. Hu, Y. Lu, P.C. Shaw, P.P.H.
But, R.C. Lin, Chem. Pharm. Bull. 56 (2008) 982.
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
[2] For a review, see: Li, W.; Tang, G. H.; Yin, S. Nat. Prod. Rep. 2021, 38, 822.
[3] L.P. Zhang, Y.M. Liang, X.C. Wei, D.L. Cheng, J. Org. Chem. 72 (2007) 3921.
[4] C. Yang, Y. Shao, K. Li, W. Xia, Beilstein J. Org. Chem. 2012 (1884) 8.
[5] Y. Cao, Y.P. Wu, J.A. Duan, Acta Pharm. Sinica 50 (2015) 199.
[6] B. Zhu, T.B. Wang, L.J. Hou, H.X. Lv, A.M. Liu, P. Zeng, A.H. Li, Chem. Nat. Compd.
52 (2016) 624.
[7] K. Fujiwara, T. Itagaki, Y. Kondo, U. Akiba, T. Tokiwano, Tetrahedron Lett. 61
(2020) 152031.
[8] T. Washio, R. Yamaguchi, T. Abe, H. Nambu, M. Anada, S. Hashimoto,
Tetrahedron 63 (2007) 12037.
Acknowledgments
We thank Prof. M. Jikei, Prof. K. Matsumoto and Prof. M. Oga-
sawara (Department of Materials Science, Graduate School of Engi-
neering Science, Akita University) for the measurements of NMR
spectra.
[9] P.P. Bora, H.A. Sema, B. Wahlang, G. Bez, J. Chem. Pharm. Res. 3 (2011) 524.
[10] S. Wang, L. Liu, X. Guo, G. Li, X. Wang, H. Dong, Y. Li, W. Zhao, RSC Adv. 9 (2019)
13878.
[11] T. Ishiyama, M. Murata, N. Miyaura, J. Org. Chem. 60 (1995) 7508.
[12] Typically, 0.98
ll of TFA was added using a thin glass capillary tube to a
solution of 1 (9.0-10 mg) in methanol-d4 (1.0 mL).
Appendix A. Supplementary data
[13] The similar elimination of the NMR anomalies of selaginellin by the
acceleration of the interconversion in acidic solution (at pH 3 in DMSO-d6)
was reported by Cao and co-workers. See: Cao, Y.; Wu, Y. P.; Chen, D. J. Chin. J.
Magn. Reson. 2019, 36, 155.
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.tetlet.2021.153295.
5