Chemical synthesis, redox transformation, and identification of sonnerphenolic C, an antioxidant in Acer nikoense

Article abstract of DOI:10.1016/j.bmcl.2017.02.054

Sonnerphenolic C (3), which was predicted in a redox product of epirhododendrin (1) isolated from Acer nikoense, was synthesized for the first time via the epimeric separation of benzylidene acetal intermediates as a key step. From a similar synthetic route, 1 was obtained concisely. As a result of their antioxidative evaluation, only 3 revealed potent activity. The redox transformation of 1 into 3 was achieved in the presence of tyrosinase and vitamin C. Moreover, 3 was identified in the decoction of A. nikoense by HPLC analysis with the effective use of synthesized 3. Thus, a novel naturally occurring antioxidant 3 was developed through the sequential flow including redox prediction, chemical synthesis, evaluation of the activity, and identification as the natural product.

Full text of DOI:10.1016/j.bmcl.2017.02.054

Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Chemical synthesis, redox transformation, and identification
of sonnerphenolic C, an antioxidant in Acer nikoense
a
a,b,
⇑
Takehiro Iwadate , Ken-ichi Nihei
a
Department of Applied Life Science, United Graduate School of Agricultural Science, Tokyo University of Agriculture and Technology, Tokyo 183-8509, Japan
Department of Applied Biological Chemistry, Faculty of Agriculture, Utsunomiya University, Tochigi 321-0943, Japan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Sonnerphenolic C (3), which was predicted in a redox product of epirhododendrin (1) isolated from Acer
nikoense, was synthesized for the first time via the epimeric separation of benzylidene acetal intermedi-
ates as a key step. From a similar synthetic route, 1 was obtained concisely. As a result of their antioxida-
tive evaluation, only 3 revealed potent activity. The redox transformation of 1 into 3 was achieved in the
presence of tyrosinase and vitamin C. Moreover, 3 was identified in the decoction of A. nikoense by HPLC
analysis with the effective use of synthesized 3. Thus, a novel naturally occurring antioxidant 3 was
developed through the sequential flow including redox prediction, chemical synthesis, evaluation of
the activity, and identification as the natural product.
Received 27 December 2016
Revised 16 February 2017
Accepted 22 February 2017
Available online xxxx
Keywords:
Sonnerphenolic C
Epirhododendrin
Antioxidant
Ó 2017 Elsevier Ltd. All rights reserved.
Redox transformation
Tyrosinase (polyphenol oxidase, EC 1.14.18.1), a copper-con-
taining oxidoreductase, is widely distributed in nature. Its active
site recognizes various phenolic compounds as substrates, and
consequently, corresponding o-quinones are generated by this
enzymatic oxidation (Scheme 1).1 The o-quinones are generally
labile and can be easily transformed into biopolymers including
melanin. Melanin formation is related to skin aging, skin cancer,
and Parkinson disease.3 Accordingly, various melanogenesis-con-
trolling agents have been developed from extensive research on
from those of the corresponding monophenols.13,14 For instance,
the antioxidant activity of o-diphenols is preferred to that of their
1
5,16
monophenols.
Thus, this redox transformation can be mean-
ingful for producing unique o-diphenols as natural antioxidants.
Acer nikoense (Aceraceae), called ‘‘Megusurinoki” in Japanese, is
,2
1
7
a folk medicinal plant used for curing eye disease. Epirhododen-
drin (1) is a characteristic monophenol in A. nikoense (Fig. 1),18
while several phenolic compounds have been isolated from this
–5
1
9–23
plant.
Recently, a novel and potent tyrosinase inhibitor, rhodo-
6–10
organic synthesis and/or natural product chemistry.
dendrol glycoside 2, was developed by the structural modification
9
,24
of 1.
This suggests that 1 can easily move closer to the active
site of tyrosinase. When 1 acts as a substrate of tyrosinase, an
o-diphenol sonnerphenolic C (3) is possibly generated with the
aid of appropriate reductive agents.
Scheme 1. Redox transformation of phenolic compounds in the presence of
tyrosinase and reductive agent.
In the presence of reductive agents such as vitamin C, the gen-
1
1
erated o-quinones are readily reduced to stable o-diphenols. The
redox transformation of polyphenols occurs frequently in plant
materials because they are rich sources of tyrosinase, phenolic
compounds, and reductive agents.12 Several o-diphenols have been
identified in nature, with their biological activities clearly differing
⇑
Corresponding author at: Department of Applied Biological Chemistry, Faculty
of Agriculture, Utsunomiya University, Tochigi 321-0943, Japan.
Fig. 1. Structure of 1–3.
E-mail address: nihei98@cc.utsunomiya-u.ac.jp (K. Nihei).
http://dx.doi.org/10.1016/j.bmcl.2017.02.054
0
960-894X/Ó 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Iwadate T., Nihei K.-i. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.054

2
T. Iwadate, K.-i. Nihei / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
2
5
+33.4, c 1.8, MeOH).25 According to previous
Recently, 3 was isolated from Sonneratia ovata (Sonneratiaceae)
that of natural 3 ([a]
D
reports,
2
4,29,32
which showed selective and moderate cell-growth inhibitory effect
against MCF-7, a hormone-responsive breast cancer cell.2 How-
ever, the existence of 3 in A. nikoense and its antioxidant potency
are still unclear. Accordingly, we studied concise syntheses of 1
and 3 via glycosylation and epimeric separation as key steps, and
evaluated their antioxidant activity. In addition, redox conversion
of 1 to 3 and identification of 3 in the extract of A. nikoense were
investigated by the effective use of synthetic compounds.
all congeners such as 1 and 2, and their epimers, pos-
5
sessed the negative sign, suggesting that unanticipated errors occur
in the measurement step of the specific rotation value of natural 3.
Although asymmetric synthesis of 1 has been accomplished via
the enzymatic procedure, the application of the synthetic scheme
of 3 led to a concise preparation of 1 (Scheme 3). Racemic alcohol
2
9
3
5
10 was glycosylated in 88% yield by Koenigs-Knorr method using
7 as the glucose donor. The b-glucoside 11 obtained was converted
Through aldol condensation between aldehyde 426 and acetone
under the basic aqueous condition, enone 5 was obtained in 91%
to diol 12 ([
a]
ꢀ20.2, c 0.23, MeOH) in 30% yield over consecutive
D
steps including transesterification and benzylidene acetalizaion. By
2
7
yield (Scheme 2). In the presence of sodium borohydride and
using silica gel column chromatography, 12 was isolated as a single
cobalt chloride hydrate, reduction of 5 smoothly proceeded to fur-
isomer, which was confirmed by comparison with the NMR data of
28
24,31
Hydrogenolysis of 12 afforded 1 in 74% yield. 1
H
nish alcohol 6 in 84% yield, although this reaction was sluggish in
analogs of 12.
9
,24
13
the synthesis of 2.
A comparison of 2,4-dibenzyloxy derivatives
and C NMR data of 1 were fully consistent with those reported pre-
3
6
indicates that the effects of steric hindrance and/or electronic
donation from the 3,4-dibenzyloxy group are insignificantly influ-
enced on its enone part. Alternatively, selective hydrogenation
using a palladium ethylenediamine complex on activated carbon,
followed by hydride reduction, has been applied in the transforma-
tion of 5 into 6 in 90% yield.9,24 However, the reduction with
sodium borohydride and cobalt chloride hydrate was selected in
order to reduce synthesis steps and to avoid the dissolution prob-
lem of 5 in toluene which is an appropriate solvent for selective
viously. Therefore, 1 was synthesized from 10 in 20% over four
steps.
OPv
PvO
O
OPv
OPv
OH
7, Ag CO ,
O
2
3
MS3A, Et O
2
rt
88%
BnO
BnO
OH
2
4
hydrogenation.
Glycosylation of 6 was performed under the Koenigs-Knorr con-
10
11
29
dition using pivalate 7 as the glucose donor. Accordingly, b-glu-
coside 8 was furnished in high yield (93%). In this step, diethyl
ether, rather than dichloromethane, acted as an appropriate sol-
vent, whereas the use of ethereal solvents was unsuitable for
HO
O
O
Ph
O
1
) NaOMe, MeOH,
reflux
O
H -Pd(OH) /C,
2 2
THF, MeOH
1
2) PhCHO, ZnCl
rt, 30% (2 steps) BnO
,
rt
74%
increasing the production rate of a
-glucoside.30 Schmidt glycosyla-
2
3
1
tion and Koenigs-Knorr glycosylation using acetylated sugar
failed to obtain the corresponding b-glucosides and complex mix-
tures were detected by TLC analysis.
12
Scheme 3. Synthesis of 1.
Removal of all pivalyl groups in 8 with sodium methoxide, fol-
lowed by acetalization with benzaldehyde in the presence of Lewis
acid, afforded diol 9 in 38% yield over two steps. Epimeric mixtures
of 8 as well as the corresponding epimeric intermediates without
pivalyl groups could not be separated by various methods includ-
ing crystallization and flash chromatography. In contrast, 9 and
its epimer, which was obtained in 33% over two steps, were easily
isolated by silica gel column chromatography.32 The stereochem-
Antioxidant activities of
,2-diphenyl-2-picrylhydrazyl (DPPH) (Table 1). Monophenol 1
was inactive to free-radical scavenging, while 1 mol of diphenol
1 and 3 were evaluated using
3
7
1
3
8–40
3
eliminated 2.23 mol of DPPH.
This efficacy was comparable
to those of vitamin C and vitamin E. As observed above, the
Table 1
Antioxidant activity of 1, 3, vitamin C, and vitamin E.
istry of the aglycone part of 9 was determined as S by comparison
1
with the previously reported
glucosides.
H
data of similar phenolic
Compounds tested
DPPH consumption
a
3
3
1
0
Finally, catechol 3 was furnished from 9 in excellent yield (99%)
3
2.23 ± 0.08
2.47 ± 0.19
1.86 ± 0.31
by hydrogenolysis using Pearlman’s catalyst.34 The H and C NMR
1
13
Vitamin C
Vitamin E
data of synthesized 3 were fully consistent with those of natural
2
5
3.
Consequently, 3 was synthesized in 27% yield from 4 over
a
The values indicate that one molecule of the tested com-
six steps for the first time. However, the sign of the specific rota-
pound scavenges how many molecules of DPPH, and represent
means ± SE of three different experiments.
27
tion of synthesized 3 ([
a]
D
ꢀ17.5, c 1.0, MeOH) was different from
Acetone, NaOH,
O
NaBH
MeOH, H
0 o
84%
4
, CoCl
2
·6H
2
2
O,
OH
OPv
BnO
BnO
CHO
H
2
O, EtOH
BnO
BnO
O
BnO
BnO
PvO
Br
OPv
+
0 oC - rt
91%
C
OPv
O
7
4
5
6
OPv
OH
O
PvO
O
OPv
OPv
HO
O
O
Ph
O
Ag
2
CO
3
,
2
O
1) NaOMe, MeOH,
reflux
2 2
H -Pd(OH) /C,
MS3A, Et O
BnO
BnO
BnO
EtOH, EtOAc
3
rt
3%
2) PhCHO, ZnCl2,
rt, 38% (2 steps) BnO
rt
99%
9
8
9
Scheme 2. Synthesis of 3.
Please cite this article in press as: Iwadate T., Nihei K.-i. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.054

T. Iwadate, K.-i. Nihei / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
3
In conclusion, we expeditiously synthesized 1 and 3, including
the epimeric separation step of their benzylidene acetal intermedi-
ates. The evaluation of the antioxidant activity indicated that only
o-diphenol 3 is capable of DPPH scavenging. The physiologically
active 3 can be generated from 1 through the redox step with
tyrosinase and vitamin C. Furthermore, the existence of 3 in the
decoction of A. nikoense was clarified using synthetic 3 as an ana-
lytical standard. This sequential flow including structural predic-
tion, organic synthesis, and analysis of the enzymatic product
would be useful to identify the trace bioactive compounds in var-
ious natural resources.
Acknowledgements
We are grateful to Mr. Yutaka Kashiwakura for his technical
assistance with the preparation of the A. nikoense extract. This
study was supported in part by JSPS KAKENHI Grant Number
15K01797. T.I. thanks a Grant of Utsunomiya University Explora-
tory Research for Young Scientists for partial support of this
research.
Fig. 2. HPLC analysis of the reaction medium with 1, tyrosinase, and vitamin C. The
sampling time was chosen as 10 s (left) and 24 h (right). The HPLC operating
conditions were as follows: column: Inertsil ODS-2 (4.6 mm ꢁ 250 mm), solvent:
2
2% MeCN-H
2
O, flow rate: 0.8 mL/min, detection: UV at 254 nm, injected amount:
2
5
lL.
A. Supplementary data
physiological activities of monophenol 1 and o-diphenol 3 differ,
and the latter can be classified as a novel naturally occurring
antioxidant.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2017.02.
0
54.
Next, redox transformation from 1 to 3 was traced by HPLC
analysis using synthetic 1 and 3 as the standard substances
References
1
1,41,42
(Fig. 2).
In the presence of tyrosinase and vitamin C, 1 was
immediately converted to 3. Vitamin C acted not only to reduce
the generated quinone to the corresponding o-diphenol but also
as a cofactor to produce oxy-tyrosinase, which was catalyzed by
the oxidation of monophenol.43 After 24 h, the peak area of 3
increased comparable to that of 1.44 These results indicate that 3
still cannot be identified in the extract of A. nikoense because of
its trace amount. Thus, HPLC analysis of the extract was performed
with the aid of synthetic 3.
1. Ramsden CA, Riley PA. Bioorg Med Chem. 2014;22:2388.
2.
3.
4.
Sugumaran M. Int J Mol Sci. 2016;17:1576.
Kanlayavattanakul M, Lourith N, Chaikul P. J Ethnopharmacol. 2016;193:607.
Park DJ, Sekhon SS, Yoon J, Kim Y, Min J. Mol Cell Biochem. 2016;413:119.
5. Segura-Aguilar J, Paris I, Muñoz P, Ferrari E, Zecca L, Zucca FA. J Neurochem.
014;129:898.
2
6
7
.
.
Verdoni M, Roudaut H, De Pomyers H, et al. Bioorg Med Chem. 2016;15:5685.
Oode C, Shimada W, Yokota M, Yamada Y, Nihei K. Carbohydr Res. 2016;436:45.
8. Manse Y, Ninomiya K, Nishi R, et al. Bioorg Med Chem. 2016;24:6650.
9. Iwadate T, Kashiwakura Y, Masuoka N, Yamada Y, Nihei K. Bioorg Med Chem Lett.
2014;24:122.
The boiling-water extract45 of the leaves of A. nikoense was
injected in the HPLC system equipped with an ODS column
1
0. Oode C, Shimada W, Izutsu Y, Yokota M, Iwadate T, Nihei K. Eur J Med Chem.
014;87:862.
11. Nihei K, Kubo I. Bioorg Med Chem Lett. 2003;13:2409.
2
(Fig. 3). As a result of co-injection with synthetic 3, a peak at
1
1
1
2. Mayer AM. Phytochemistry. 2006;67:2318.
3. Zhou Z, Xiao J, Fan H, et al. Food Chem. 2017;214:644.
4. Wang Y, Xiang L, Wang C, Tang C, He X. PLoS One. 2013;8:e71144.
1
9.6 min was identified to be natural 3. In addition, the existence
of 3 was confirmed by HRMS analysis of this peak fractionated
by HPLC separation. A. nikoense, especially its decoction, has been
15. Okawa M, Kinjo J, Nohara T, Ono M. Biol Pharm Bull. 2001;24:1202.
16. Ponomarenko J, Trouillas P, Martin N, Dizhbite T, Krasilnikova J, Telysheva G.
Phytochemistry. 2014;103:178.
18
used as a healthy food in Japan. This is a first report to clarify
antioxidant 3 as an ingredient of the food product of A. nikoense,
although it is in minute amount.
1
7. Inoue T. Yakugaku Zasshi. 1993;113:181.
18. Inoue T, Ishidate Y, Fujita M, Kubo M, Fukushima M, Nagai M. Yakugaku Zasshi.
978;98:41.
1
19. Nagai M, Kubo M, Fujita M, Inoue T, Matsuo M. Chem Pharm Bull. 1978;26:2805.
20. Morita H, Deguchi J, Motegi Y, et al. Bioorg Med Chem Lett. 2010;20:1070.
21. Okuda T, Yoshida T, Hatano T, Ikeda Y, Shingu T, Inoue T. Chem Pharm Bull.
1
986;34:4075.
2. Morikawa T, Tao J, Toguchida I, Matsuda H, Yoshikawa M.
003;66:86.
2
J Nat Prod.
2
2
2
2
2
3. Akihisa T, Taguchi Y, Yasukawa K, et al. Chem Pharm Bull. 2006;54:735.
4. Iwadate T, Nihei K. Bioorg Med Chem. 2015;23:6650.
5. Nguyen T, Pham H, Pham N, et al. Bioorg Med Chem Lett. 2015;25:2366.
6. Acuña AU, Amat-Guerri F, Morcillo P, Liras M, Rodríguez B. Org Lett.
2009;11:3020.
2
2
2
3
7. Shi J, Liu Y, Wang M, Lin L, Lin X, Feng X. Tetrahedron. 2011;67:1781.
8. Aramini A, Brinchi L, Germani R, Savelli G. Eur J Org Chem. 2000;1793.
9. Mori K, Qian Z. Bull Soc Chim Fr. 1993;130:382.
0. Demchenko AV. General aspects of the glycosidic bond formation. In:
Demchenko AV, ed. Handbook of Chemical Glycosylation: Advances in
Stereoselectivity and Therapeutic Relevance. Weinheim: WILEY-VCH Verlag
GmbH & Co. KGaA; 2008:1–27.
3
3
3
1. Schmidt RR, Michel J. Angew Chem Int Ed Engl. 1980;19:731.
2. Huang LJ, Hou SJ, Li JB, Yu DQ. J Asian Nat Prod Res. 2007;9:223.
2
5
3. Compound 9: Colorless solid; [
a
]
D
ꢀ19.9 (c 1.42, CHCl
3
); IR (film) mmax 3432,
Fig. 3. HPLC analysis of the decoction of A. nikoense leaves. The HPLC operating
conditions were as follows: column: Wakosil-II 5C 18AR Prep (4.6 mm ꢁ 250 mm),
ꢀ1 1
1606, 1509, 1071, 750 cm
;
H NMR (400 MHz, CDCl
Bn), 6.85 (d, J = 8.2 Hz, 1H, H-5 ), 6.78 (d, J = 1.9 Hz, 1H, H-2 ), 6.69 (dd, J = 8.2,
3
) d 7.47–7.24 (m, 15H, Ph,
0
0
0
solvent: 5% MeCN-H
2
O (0–5 min) and 5–30% MeCN-H
2
O (5–30 min), flow rate:
1.9 Hz, 1H, H-6 ), 5.51 (s, 1H, PhCH), 5.13 (s, 2H, Bn), 5.11 (s, 2H, Bn), 4.78 (bs, 2H,
0
0
0
0
1
.0 mL/min, detection: UV at 254 nm, injected amount: 0.25 mL.
OH), 4.39 (d, J = 7.8 Hz, 1H, H-1 ), 4.29 (dd, J = 10.5, 5.0 Hz, 1H, H-6 ), 3.80–3.71
Please cite this article in press as: Iwadate T., Nihei K.-i. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.054

4
T. Iwadate, K.-i. Nihei / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
0
0
00
00
(
m, 3H, H-2, H-3 , H-6 ), 3.54 (t, J = 9.3 Hz, 1H, H-4 ), 3.41 (dd, J = 8.6, 7.8 Hz, 1H,
40. Sodium acetate buffer (1 mL, 100 mM, pH 5.5), ethanol (1.87 mL), and ethanolic
solution of DPPH (0.03 mL, 3 mM) were put into a cuvette. The sample DMSO
00
00
H-2 ), 3.40 (td, J = 9.3, 5.0 Hz, H-5 ), 2.62–2.57 (m, 2H, H-4), 1.91–1.82 (m, 1H, H-
), 1.73–1.64 (m, 1H, H-3), 1.24 (d, J = 6.2 Hz, 3H, H-1); ESIHRMS m/z 613.2804
M+H] (calcd for C37
The epimer of 9: Colorless solid; [
590, 1515, 1091, 748 cm
Bn), 6.84 (d, J = 8.2 Hz, 1H, H-5 ), 6.79 (d, J = 2.0 Hz, 1H, H-2 ), 6.70 (dd, J = 8.2,
.0 Hz, 1H, H-6 ), 5.51 (s, 1H, PhCH), 5.13 (s, 2H, Bn), 5.11 (s, 2H, Bn), 4.40 (d,
J = 7.7 Hz, 1H, H-1 ), 4.29 (dd, J = 10.5, 5.0 Hz, 1H, H-6 ), 3.86–3.74 (m, 3H, H-2,
H-3 , H-6 ), 3.55 (t, J = 9.3 Hz, 1H, H-4 ), 3.47 (dd, J = 8.4, 7.7 Hz, 1H, H-2 ), 3.40
td, J = 9.3, 5.0 Hz, H-5 ), 2.85 (bs, 1H, OH), 2.66–2.51 (m, 2H, H-4), 1.89–1.70 (m,
H, H-3), 1.73–1.65 (m, 1H, H-3), 1.64 (bs, 1H, OH), 1.18 (d, J = 6.2 Hz, 3H, H-1);
ESIHRMS m/z 613.2807 [M+H] (calcd for C37
4. Pearlman WM. Tetrahedron Lett. 1967;8:1663.
3
[
solution was added and the resultant solution was incubated at 30 °C for
+
3
H
41
O
8
, 613.2801).
20 min. The absorbance at 517 nm (DPPH,
e
= 8.32 ꢁ 10 ) was recorded. As a
2
5
a
]
D
ꢀ34.6 (c 1.22, CHCl
3
); IR (film)
m
max 3380,
negative control, DMSO (0.03 mL) was added to the cuvette. From the decrease
in absorbance, the scavenging activity was calculated and expressed as
scavenged DPPH molecules per molecule.
ꢀ
1 1
1
;
H NMR (400 MHz, CDCl
3
) d 7.49–7.26 (m, 15H, Ph,
0
0
0
2
41. Kubo I, Nihei K, Shimizu K. Bioorg Med Chem. 2004;12:5343.
42. A mixture of DMSO solution of 1 (0.2 mL, 10 mM), sodium phosphate buffer
(1.2 mL, 100 mM, pH 6.8), and deionized water (4 mL) was incubated at 30 °C
for 30 min. An aqueous solution of vitamin C (0.4 mL, 50 mM) and tyrosinase
(0.1 mg/mL) in sodium phosphate buffer (0.2 mL, 50 mM, pH 6.8) was added to
the mixture. The resultant solution was vigorously stirred under an aerial
condition and directly injected by HPLC.
0
0
00
00
00
00
00
0
0
(
1
+
41 8
H O , 613.2801).
3
3
3
3
3
3
5. Ramachandra MS, Subbaraju GV. J Asian Nat Prod Res. 2006;8:683.
6. Pan H, Lundgren LN. Phytochemistry. 1994;36:79.
7. Blois MS. Nature. 1958;4617:1199.
8. Nihei K, Nihei A, Kubo I. J Agric Food Chem. 2004;52:5011.
9. Kubo I, Fujita K, Nihei K, Masuoka N. Bioorg Med Chem. 2003;11:573.
43. Ros JR, Rodríguez-López JN, García-Cánovas F. Biochem J. 1993;295:309.
44. Land EJ, Ramsden CA, Riely PA, Stratdord MRL. Arkivoc. 2008;2:258.
45. The dried leaves (25 g) of A. nikoense, which were cut into pieces, were soaked
in boiling water (500 mL) for 10 min. After filtration, the resultant solution was
analyzed by HPLC.
Please cite this article in press as: Iwadate T., Nihei K.-i. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.054