Isolation, total synthesis and quantification of caffeoylisocitric acid, a characteristic ingredient of the superfood amaranth

Article abstract of DOI:10.1016/j.tet.2019.06.033

Amaranth is regarded as a new “super-vegetable” in western countries, albeit it is consumed for centuries in Africa and Asia. In addition to common carotenoids, flavonoids and polyphenols, caffeoylisocitric acid has been described as amaranth type-specifi

Full text of DOI:10.1016/j.tet.2019.06.033

Accepted Manuscript
Isolation, total synthesis and quantification of caffeoylisocitric acid, a characteristic
ingredient of the superfood amaranth
Tianyi Zhou, Benedikt Ringbeck, Nils Helge Schebb, Jürgen Scherkenbeck
PII:
S0040-4020(19)30695-7
https://doi.org/10.1016/j.tet.2019.06.033
TET 30425
DOI:
Reference:
To appear in: Tetrahedron
Received Date: 1 March 2019
Revised Date: 17 June 2019
Accepted Date: 19 June 2019
Please cite this article as: Zhou T, Ringbeck B, Schebb NH, Scherkenbeck Jü, Isolation, total synthesis
and quantification of caffeoylisocitric acid, a characteristic ingredient of the superfood amaranth,
Tetrahedron (2019), doi: https://doi.org/10.1016/j.tet.2019.06.033.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Isolation, Total Synthesis and Quantification of Caffeoylisocitric Acid, a
Characteristic Ingredient of the Superfood Amaranth
Tianyi Zhoua, Benedikt Ringbecka, Nils Helge Schebba, Jürgen Scherkenbecka*
a Faculty of Mathematics and Natural Sciences, University of Wuppertal, Gaussstrasse 20, D-
42119 Wuppertal, Germany
* Corresponding author. E-Mail address: scherkenbeck@uni-wuppertal.de (J. Scherkenbeck)
ABSTRACT
Amaranth is regarded as a new “super-vegetable” in western countries, albeit it is consumed for
centuries in Africa and Asia. In addition to common carotenoids, flavonoids and polyphenols,
caffeoylisocitric esters have been described as amaranth type-specific secondary metabolites.
Remarkably, nothing is known on biological effects of this specific polyphenol. Here we detail a
concise, diastereoselective synthesis of caffeoylisocitric acid, deuterium-labelling studies and a
quantitative determination of the caffeoylisocitric acid content of three different amaranth types.
Keywords: Amaranth; polyphenols; caffeoylisocitric acid; isolation; total synthesis; natural product
1

ACCEPTED MANUSCRIPT
1.
Introduction
Polyphenols constitute a structurally highly diverse class of plant secondary metabolites, including
the hydroxybenzoic acids, hydroxycinnamic acids, flavonoids, stilbenes and lignans as the major
families [1]. Polyphenols are an important element of human nutrition ever since. Several
beneficial effects of polyphenols such as anti-inflammatory activity, antibacterial or cancer
protecting activities as well as neuroprotective effects have been described, though most of them
are rapidly metabolized and excreted [2-6]. In particular, polyphenols and glucosides, i.e. their
adducts with glucose in traditional fruits and vegetables have been studied intensively in the past
[7]. However, in the course of globalization, several uncommon food plants have become quite
popular in western countries. Those vegetables constitute a still buried treasure of new, largely
uncharacterized phenolic secondary metabolites and other plant natural products with the potential
for the development of new healthy plant-based foodstuffs. In addition to African nightshade
(Solanum scabrum), cowpea (Vigna unguiculata) and others, amaranth (Amaranthus) is another
“super-vegetable” with increasing importance [8]. Amaranth originates from middle- and south
America where it is used as a pseudo-grain. In Asia and Africa, the leaves are used as a vegetable.
Amaranth leaves provide a broad spectrum of secondary metabolites, such as carotenoids,
flavonoids and polyphenols [9]. While esters of hydroxycinnamic acids with glucose and aldaric
acids are common in all European traditional vegetables, esters with isocitric acid are characteristic
for amaranth plants and the grass Dactylis glomerate [10,11].
2

ACCEPTED MANUSCRIPT
Fig. 1. Structures of Amaranthus cruentus caffeic acid esters.
An overview of the natural diversity of hydroxycinnamic acid derivatives in six different amaranth
species has been published very recently by Neugart [10]. Although caffeoylisocitric acid 1 (Fig. 1)
has been isolated and characterized already in 1987 as a major constituent of Amaranthus cruentus
cotyledons, nothing is known on beneficial or adverse biological effects of this compound [12]. To
date, neither a total synthesis nor a quantitative determination of caffeoylisocitric acid in amaranth
leaves has been described.
2.
Results and discussion
2.1. Synthesis of caffeoylisocitric acid (1)
Acetyl protected caffeic acid chloride 4 and anhydride 5 were prepared in good yields according
to published literature methods [13-15].
3

ACCEPTED MANUSCRIPT
Scheme 1. Synthesis of caffeic acid derivatives 4 and 5. Reagents and conditions: a) Ac2O,
°
pyridine, DMAP, rt, overnight, 91%; b) SOCl2, DMF, DCM, 60 C, 4 hours, quantitative; c)
Et3N, triphosgene, EtOAc, 0 °C, 15 minutes, 81%.
Isocitric tert-butylester 12 was synthesized with some modifications according to a known
procedure [16-18]. In particular, we found the oxidative cleavage of the allylic double bond in
diester 9 with RuCl3/NaIO4 more straightforward compared to an ozonolysis. Acid 10 was
obtained in an excellent yield of 96% as a single stereoisomer. Key-step of this chiral-pool
synthesis is a highly diastereoselective Fráter-Seebach allylation of the ester enolate, generated
with LiHMDS from malic ester 7. The excellent diastereoselectivity is based on the predominant
formation of the (Z)-ester enolate followed by addition of allyl bromide to the Si-face of the
enolate [19,20]. NMR spectroscopic and specific rotation data of the allylation product 8 were
identical with reported values [18].
4

ACCEPTED MANUSCRIPT
Scheme 2. Synthesis of caffeoylisocitric acid (1). Reagents and Conditions: a) N,N’-diisopropyl-
O-tert-butylisourea, DCM, rt, 48 hours, 74%; b) LiHMDS, allyl bromide, THF, -78 °C 1 hour to
°
0 C 1 hour, 82%; c) Ac2O, pyridine, Et3N, DMAP, THF, rt, 5 hours, 97%; d) RuCl3, NaIO4,
MeCN:CCl4:H2O=5:5:8, rt, 10 hours, 96%; e) N,N’-diisopropyl-O-tert-butylisourea, DCM,
°
reflux, 48 hours, 94%; f) K2CO3, MeOH, 0 C, 1 hour, 84%; g) acid chloride 4, pyridine, rt, 48
hours, 86%; h) TFA, DCM, rt, 48 hours, quantitative; i) 3 M HCl (aq.), acetone, reflux, 5 hours,
61%.
Tri-tert-butyl protected isocitric derivative 11 was obtained by esterification with a large excess
of N,N’-diisopropyl-O-tert-butylisourea in 94% yield. Removal of the acetoxy protecting group
with K2CO3 in methanol gave alcohol 12 (84%) required for coupling with a suitably protected
caffeic acid derivative.
5

ACCEPTED MANUSCRIPT
Direct coupling of isocitric tri-tert-butylester 12 and acetyl caffeic acid 3 with N,N’-
diisopropylcarbodiimide and DMAP in DCM resulted only in low yields of 17%. Also,
disappointingly low yields (21%) were observed in the coupling with the acid anhydride 5, which
was prepared in 81% yield by reacting acid 3 with triphosgene in the presence of Et3N (Scheme 1).
A very good yield (86%) of ester 13 was indeed obtained with the acid chloride 4 in pyridine. A
two-step sequence, consisting of first cleaving the tert-butyl esters with TFA, followed by removal
of the acetoxy groups with 3 M HCl finally afforded caffeoylisocitric acid (1) in 61% yield over
the two deprotection steps. Based on all available analytical and spectroscopic data, the synthetic
material was identical with the natural substance we isolated from amaranth leaves and literature
reports [12].
2.2.
Synthesis of deuterium-labeled caffeoylisocitric acid (1D)
For subsequent biochemical studies we were also interested in a deuterium-labelled derivative of
caffeoylisocitric acid (1D). Since our synthesis of isocitric acid does not provide any chemical
transformation which would allow an economic introduction of a deuterium label, we focused on
the synthesis of caffeic acid. In principle, cinnamic acids can be easily prepared by a Perkin
reaction of arylaldehydes and acetic anhydride [21]. 3,4-Dihydroxybenzaldehyde (15), the starting
material for caffeic acid, was first protected as deuterated diacetate. Remarkably, undeuterated
acetate groups were found to exchange significantly protons with the deuterium atom of the double
bond, introduced in the Perkin condensation. By consequently using deuterated reagents and bases
a labelling rate of 20:1 was achieved for caffeic acid (16), though with a low chemical yield.
However, ester formation of caffeic acid chloride 17 and isocitric acid derivative 12 resulted in
some loss (5:1) of the deuterium label, probably caused by the acidic work-up [21].
6

ACCEPTED MANUSCRIPT
Scheme 3. Synthesis of deuterium-labelled (E)-caffeoylisocitric acid 1D. Reagents and
°
°
conditions: (a) Ac2O-d6, NaOAc-d3, 140 C, 48 hours, 11%; (b) SOCl2, DMF, DCM, 60 C, 4
hours, quantitative; (c) alcohol 12, pyridine, rt, 48 hours, 14%; (d) TFA, DCM, rt, 48 hours,
quantitative; (e) 3M HCl (aq.), acetone, reflux, 5 hours, 67%.
7

ACCEPTED MANUSCRIPT
2.3.
Isolation of caffeoylisocitric acid and quantitative determination in amaranth leaves
The content of (E)-caffeoylisocitric acid 1 in three amaranth species was determined by LC-UV.
High concentrations of caffeoylisocitric acid were found in A. lividus L. (1.41 ± 0.20 g/100 g dry
weight) and A. tricolor L. (red) (1.11 ± 0.04 g/100 g dry weight). Despite the same cultivation
conditions, A. tricolor L. shows a strong deviation in the caffeoylisocitric acid content (0.0099 ±
0.0002 g/100 g dry weight), while concentrations of other UV active substances were
comparable to A. tricolor L. (red) and A. lividus L. The caffeic acid content in amaranth (A.
cruentus), cultivated in Kenya, was recently determined in a semi-quantitative approach [10]. Of
the total phenolic compounds (0.92-1.42 g/100 g dry weight), caffeic acid derivatives make up
the highest percentage with concentrations of 0.57 and 0.92 g/100 g dry weight [9].
3.
Conclusion
Our studies have shown unequivocally the relevance of caffeoylisocitric acid as one of the major
polyphenols in amaranth leaves. Our synthesis of labelled caffeoylisocitric acid provides the basis
for more detailed investigations, which will address the question of beneficial or adverse biological
effects and mode of action. Further studies will also include additional amaranth types to clarify
the apparently strong variation of caffeoylisocitric acid contents in different amaranth species.
4.
Experimental section
4.1. Abbreviations
Cyh, cyclohexane; DCM, dichloromethane; DIC, N,N’-diisopropylcarbodiimide; DMAP, 4-
dimethylaminopyridine; DMF, dimethylformamide; DMSO, dimethyl sulfoxide; DW, dry weight;
Et2O, diethyl ether; EtOAc, ethyl acetate; n-Hex, n-hexane; LiHMDS, lithium
8

ACCEPTED MANUSCRIPT
bis(trimethylsilyl)amide; MeCN, acetonitrile; MeOH, methanol; RP, reversed-phase; sat.,
saturated; TFA, trifluoroacetic acid; THF, tetrahydrofuran; TMS, tetramethylsilane.
4.2. General experimental procedures
Optical rotations were measured on a Krüss P-8000T polarimeter. IR spectra were recorded on a
Bruker ALPHA FTIR spectrometer. 1H- and 13C-NMR spectra were recorded on Bruker Avance
III 600 and Bruker Avance 400 spectrometers operating at 600 and 400 MHz (1H) respectively
and 150 and 100 MHz (13C) respectively. Accurate mass determinations were achieved with a
Bruker micrOTOF mass spectrometer. The reactions were monitored by TLC carried out on
Macherey Nagel silica gel plates (60F-254) or Merk silica gel 60 RP-18 F254 plates using UV
light and aqueous solution of KMnO4, K2CO3, NaOH and heat as the visualizing agents. HPLC
analysis was conducted on a PerfectSil Target ODS-3 HD 5 ꢀm 100×4.6 mm column using an
Agilent 1100 instrument. Reagents and solvents were purchased from commercial sources and
used without further purification, unless otherwise stated. Allyl bromide was redistilled over
CaCl2. TFA was redistilled over P2O5. DCM, THF were dried with MB-SPS-800 solvent
purification system. Dry Et2O was obtained by passing through an activated aluminium oxide
column. DMF and pyridine were redistilled over CaH2. Reactions were stirred magnetically
under an argon atmosphere unless otherwise stated.
4.3. Plant material
The three amaranth species, A. tricolor, A. tricolor red and A. lividus were cultivated over an 8-
week period during April and May 2017 in a greenhouse (Wermelskirchen, Germany). Leaves
were harvested before flowering and stored till extraction at -80 °C.
9

ACCEPTED MANUSCRIPT
4.4.
Extraction and quantification of caffeoylisocitric acid from amaranth
The extraction was based on an optimized procedure, originally developed by Neugart [9]. 20 mg
shred and freeze-dried amaranth leaves were extracted with 600 µL methanol/water (60/40, v/v)
for 40 minutes under continuous shaking at room-temperature. The mixture was centrifuged
afterwards for 10 min at 4 °C and 20,000 ×g. The extractions were repeated four times, each with
300 µL methanol/water (60/40, v/v). The solvent from the collected supernatants was completely
evaporated under nitrogen. The residue was suspended in 1 mL methanol/water (5/95, v/v). The
solution was analyzed by LC-UV at 295 nm for quantitative analysis and ESI(-)-MS/MS for
structure elucidation. For quantification the reconstituted extract was mixed (1/1, v/v) with a
solution of internal standard (2 µM chlorogenic acid in methanol/water (5/95, v/v)). Accuracy of
the linear calibration curve (R2 = 0.9989) was in a range of 88 % and 104 % over the whole
calibration in a range between 1 µM and 30 µM. Quantification was carried out by external
calibration in a range between 1 µM and 30 µM. LC-UV analysis was performed using Merck
Hitachi liquid chromatography system (Interface D-7000, Diode Array Detector L-7455,
Programmable Autosampler LaChrom, Pump L-7100 LaChrom; Hitachi High Technologies,
Tokyo, Japan) with a column oven at 40 °C (IGLOO-CIL, ERC-GmbH, Riemerling, Germany) and
a 3-line degasser (Jasco DG-2080-53, JASCO Corporation, Tokyo, Japan). Separation was
performed on a RP-18 column of the dimensions 150 x 4.6 mm, a particle size of 5 µm and a pore
size of 80 Å (C18 XDB Zorbax Eclipse, Agilent Technologies, Santa Clara, USA). The analytes
and amaranth extracts (injection volume 10 µL) were separated by a binary gradient at a flow rate
of 0.8 mL/min of water containing 5 % acetic acid as solvent A and pure MeCN as solvent B. The
following linear gradient was used: 0.0 to 15.0 minutes 5 % B, 15.0 to 25.0 minutes linear to 20 %
B, 25.0 to 35.0 minutes linear to 95 % B, 35.0 to 45.0 isocratic 95 % B, 45.0 to 45.1 minutes
10

ACCEPTED MANUSCRIPT
returning to initial conditions of 5 % B, and reconditioning between 45.1 and 60.0 minutes.
Analytes were detected at 295 nm. LC-MS measurements were performed with an Agilent 1200-
LC system (Agilent Technologies, Santa Clara, USA) with an Autosampler (CTC-Pal, CTC
Analytics, Zwingen, Swiss Confederation) and a AB Sciex API 3200 QqQ mass spectrometer (AB
Sciex, Darmstadt, Germany). Separations were performed on an RP-18 column (100 x 2 mm,
particle size 1.8 µm. EC102 Nucleodur C18 Gravity from Macherey-Nagel, Düren, Germany). The
analytes and amaranth extracts (injection volume 10 µL) were separated by a binary gradient at a
flow rate of 0.3 mL/min of water containing 5 % acetic acid as solvent A and pure MeCN as
solvent B. The following gradient was used: 0.0 to 2.0 minutes 5 % B, 2.0 to 8.5 minutes linear to
100 % B, 8.5 to 9.5 minutes isocratic 100 % B, 9.5 to 9.7 minutes returning to initial conditions of
5 % B, and reconditioning between 9.7 and 11.0 minutes. Analytes were detected in ESI (-)
ionization mode with a declustering potential of -15 V and a collision energy of -30 V. Product ion
spectra were measured with an entrance m/z of 353.4 Da and a SCAN range between 100 and 360
Da. The further instrument settings were: Ion Spray voltage: -4500 V, Heater gas temperature: 500
°C, Ion source gas 1: 70, Ion source gas 2: 55, Curtain gas: 20.
4.5.
Synthesis
4.5.1. 3,4-Diacetoxycinnamic acid (3)
Acetic anhydride (2.20 mL, 23.30 mmol, 2.1 eq.) was added dropwise to a solution of trans-
caffeic acid 2 (2.00 g, 11.10 mmol, 1.0 eq) and DMAP (0.03 g, 0.25 mmol, 0.02 eq.) in pyridine
(10.0 mL) over 10 minutes at 0 °C. The reaction mixture was stirred at rt overnight and was then
poured onto cold water. The aqueous layer was extracted with EtOAc. The organic layer was
washed with 1 M HCl (aq.) and water, dried over Na2SO4, filtered and concentrated in vacuo.
11

ACCEPTED MANUSCRIPT
The crude product was recrystallized from hot water to obtain acid 3 as a white powder (2.67 g,
1
10.10 mmol, 91%). IR (ATR) νmax 2551, 1753, 1675, 1629, 1500, 1432, 1375, 1184 cm-1; H
NMR (600 MHz, DMSO-d6) δ 12.45 (br. s, 1H), 7.66 (d, J = 1.9 Hz, 1H), 7.63 (dd, J = 8.5, 2.1
Hz, 1H), 7.57 (d, J = 15.8 Hz, 1H), 7.31 (d, J = 8.3 Hz, 1H), 6.53 (d, J = 15.8 Hz, 1H), 2.29 (s,
13
3H), 2.28 (s, 3H); C NMR (150 MHz, DMSO-d6) δ 168.1, 168.0, 167.3, 143.3, 142.3, 142.1,
133.1, 126.7, 124.0, 122.9, 120.3, 20.3, 20.3. EIMS m/z (%) 205.1 [M-OAc]+ (61), 265.1 [M+H]+
(55), 282.1 [M+NH4]+ (75), 287.1 [M+Na]+ (27); HREIMS m/z 287.0523 [M+Na]+ (calcd for
+
C13H12NaO6 , 287.0532); Rf= 0.46 (DCM:MeOH=19:1, v/v).
4.5.2. 3,4-Diacetoxycinnamoyl chloride (4)
Thionyl chloride (1.90 mL, 26.10 mmol, 6.9 eq.) and DMF (2.90 ꢀL) were added to the solution
of acid 3 (1.00 g, 3.78 mmol, 1.0 eq.) in DCM (5.0 mL). After refluxing at 60 °C for 7 hours, the
solvent and excess thionyl chloride were removed in vacuo. The residue was used directly
without any further purification.
4.5.3. 3,3,4,4’-Tetraacetoxycinnamic anhydride (5)
Et3N (0.79 mL, 5.68 mmol, 1.0 eq.) were added dropwise at 0 °C to a suspension of acid 3 (1.50
g, 5.68 mmol, 1.0 eq.) in MgSO4-dried EtOAc (110.0 mL). After all solids have dissolved,
triphosgene (0.29 g, 0.97 mmol, 0.2 eq.) was added in one portion. The reaction mixture was
°
stirred at 0 C for 10 minutes and then warmed up to rt over 15 minutes. The reaction mixture
was filtered. The filtrate was concentrated in vacuo. The residue was recrystallized from toluene
(18.0 mL) to give acid anhydride 5 as white powder (1.17 g, 2.29 mmol, 81%). IR (ATR) νmax
1
1778, 1760, 1630, 1503, 1366 cm-1; H NMR (600 MHz, Acetone-d6) δ 7.93 (d, J = 15.8 Hz,
2H), 7.74-7.71 (m, 4H), 7.38 (d, J = 8.3 Hz, 2H), 6.74 (d, J = 16.2 Hz, 2H), 2.32 (d, J = 1.5 Hz,
12

ACCEPTED MANUSCRIPT
12H); 13C NMR (150 MHz, Acetone-d6) δ 169.3, 169.2, 163.7, 148.1, 146.4, 144.7, 134.2, 128.7,
+
125.9, 125.2, 119.5, 21.2, 21.1; HREIMS m/z 533.1054 [M+Na]+ (calcd for C26H22NaO11
,
533.1060); Rf=0.7 (Cyh:EtOAc=3:1, v/v).
4.5.4. di-tert-Butyl-(R)-malate (7)
N,N’-Diisopropyl-O-tert-butylisourea (30.00 g, 149.90 mmol, 6.7 eq.) was added to the
suspension of (R)-malic acid (3.00 g, 22.40 mmol, 1.0 eq.) in DCM (150.0 mL). After stirring at
rt for 48 hours, the mixture was filtered through a short pad of celite. The filtrate was
concentrated in vacuo and dissolved in EtOAc (100.0 mL). The solution was washed with 1 M
HCl (aq.), sat. NaHCO3 (aq.), brine, dried over Na2SO4, filtered and concentrated in vacuo. The
crude product was purified by column chromatography (Cyh:EtOAc=9:1, v/v) to obtain ester 7
as colorless oil (4.06 g, 16.50 mmol, 74%). [ꢀ]ꢂꢁꢃ +6.2 (c 0.87, CHCl3); IR (ATR) νmax 3500,
2979, 2935, 1727, 1367, 1252, 1144, 1101 cm-1; 1H NMR (400 MHz, CDCl3) δ 4.29 (dd, J = 5.9,
4.5 Hz, 1H), 2.71 (dd, J = 16.3, 4.5 Hz, 1H), 2.63 (dd, J = 16.3, 5.9 Hz, 1H), 1.45 (s, 9H), 1.44
(s, 9H); 13C NMR (100 MHz, CDCl3) δ 172.8, 169.8, 82.6, 81.3, 67.6, 39.9, 28.1, 28.0; EIMS
m/z (%) 269.1 [M+Na]+ (52), 515.3 [2M+Na]+ (40); HREIMS m/z 269.1359 [M+Na]+ (calcd for
+
C12H22NaO5 , 269.1365); Rf= 0.51 (Cyh:EtOAc=4:1, v/v).
4.5.5. di-tert-Butyl-(S)-2-allyl-(R)-3-hydroxy-succinate (8)
1.1 M LiHMDS in THF (27.30 mL, 30.03 mmol, 2.1 eq.) was added dropwise to a solution of
ester 7 (3.52 g, 14.30 mmol, 1.0 eq.) and allyl bromide (1.27 mL, 14.70 mmol, 1.0 eq.) in THF
°
°
°
(50.0 mL) at -78 C. The solution was stirred at -78 C for 1 hour and for another hour at 0 C.
The reaction mixture was then poured onto 1 M HCl (aq.) (191.0 mL). The aqueous layer was
extracted with EtOAc. The combined organic layers were washed with brine, dried over Na2SO4
13

ACCEPTED MANUSCRIPT
and concentrated in vacuo. The crude product was purified by column chromatography
(Cyh:EtOAc=19:1, v/v) to afford alcohol 8 as colorless oil (3.34 g, 11.70 mmol, 82%). [ꢀ]ꢁꢂꢃ
-
1
10.1 (c 0.99, DCM)18; IR (ATR) νmax 3518, 2979, 1726, 1367, 1250, 1149 cm-1; H NMR (400
MHz, CDCl3) δ 5.81 (dddd, J = 17.1, 10.2, 7.6, 6.3 Hz, 1H), 5.18-5.05 (m, 2H), 4.10 (dd, J = 7.6,
3.1 Hz, 1H), 3.18 (d, J = 7.6 Hz, 1H), 2.79 (ddd, J = 7.9, 7.1, 3.1 Hz, 1H), 2.60-2.51 (m, 1H),
2.43-2.33 (m, 1H), 1.49 (s, 9H), 1.43 (s, 9H). 13C NMR (100 MHz, CDCl3) δ 172.7, 171.4,
135.3, 117.4, 82.6, 81.6, 70.6, 48.9, 32.7, 28.1, 28.0. EIMS m/z (%) 309.2 [M+Na]+ (39), 575.4
+
[2M+Na]+ (60); HREIMS m/z 309.1683 [M+Na]+ (calcd for C15H26NaO5 , 309.1678); Rf= 0.48
(Cyh:EtOAc=9:1, v/v).
4.5.6. di-tert-Butyl-(R)-2-acetoxy-(S)-3-allyl-succinate (9)
Et3N (1.75 mL, 12.57 mmol, 1.2 eq.), DMAP (0.15 g, 1.30 mmol, 0.1 eq.) and acetic anhydride
(1.19 mL, 12.57 mmol, 1.2 eq.) were added to the solution of alcohol 8 (3.00 g, 10.50 mmol, 1.0
eq.) in THF (90.0 mL) at 0 °C. The reaction mixture was stirred at rt for 5 hours and then poured
onto 1 M HCl (aq.) (105.0 mL). The organic layer was washed with sat. NaHCO3 (aq.) and brine,
dried over Na2SO4 and concentrated in vacuo. The crude product was purified by column
chromatography (Cyh:EtOAc=4:1, v/v) to afford ester 9 as colorless oil (3.35 g, 10.20 mmol,
97%). [ꢀ]ꢂꢁꢃ +6.9 (c 0.80, Et2O); IR (ATR) νmax 2979, 2934, 1734, 1643, 1369, 1222, 1149,
1
1045, 845 cm-1. H NMR (400 MHz, CDCl3) δ 5.76 (ddt, J = 17.0, 9.9, 6.9 Hz, 1H), 5.10-5.04
(m, 3H), 2.96 (m, 1H), 2.52-2.40 (m, 1H), 2.31-2.19 (m, 1H), 2.10 (s, 3H), 1.46 (s, 9H), 1.43 (s,
9H); 13C NMR (100 MHz, CDCl3) δ 170.0, 169.9, 167.5, 134.5, 117.6, 82.5, 81.2, 72.3, 47.0,
32.2, 28.0, 28.0, 20.5; EIMS m/z (%) 351.2 [M+Na]+ (7), 679.4 [2M+Na]+ (10); HREIMS m/z
+
351.1779 [M+Na]+ (calcd for C17H28NaO6 , 351.1784); Rf= 0.65 (n-Hex:Et2O=4:1, v/v).
14

ACCEPTED MANUSCRIPT
4.5.7. tert-Butyl-(R)-2-acetoxy-(S)-3-tert-butoxycarbonyl-pentanedioic acid (10)
RuCl3·2H2O (11.50 mg, 0.05 mmol, 0.04 eq.) and NaIO4 (1.12 g, 5.23 mmol, 4.1 eq.) were
added to the solution of alkene 9 (0.42 g, 1.27 mmol, 1.0 eq.) in CCl4: MeCN: H2O (9.0 mL,
5:5:8, v/v/v). The reaction mixture was stirred at rt for 10 hours and was then diluted with DCM
(10.0 mL). The organic layer was extracted with DCM (10.0 mL). The extract was washed with
water, dried over Na2SO4 and concentrated in vacuo. The crude product was purified by column
chromatography (Cyh:EtOAc=4:1, 0.5% AcOH, v/v) to afford acid 10 as a white solid (0.42 g,
1.22 mmol, 96%). [ꢀ]ꢂꢁꢃ +35.1 (c 1.07, DCM); IR (ATR) νmax 2980, 1732, 1369, 1220, 1148 cm-
1; 1H NMR (600 MHz, CDCl3) δ 5.20 (d, J = 3.4 Hz, 1H), 3.38 (dt, J = 9.8, 4.5, 3.5 Hz, 1H), 2.80
(dd, J = 17.1, 9.9 Hz, 1H), 2.54 (dd, J = 17.2, 4.6 Hz, 1H), 2.12 (s, 3H), 1.47 (s, 9H), 1.45 (s,
9H); 13C NMR (150 MHz, CDCl3) δ 176.7, 169.9, 168.9, 166.5, 83.2, 82.4, 71.9, 43.5, 32.1,
28.0, 27.9, 20.5; EIMS m/z (%) 347.2 [M+H]+ (10), 364.2 [M+NH4]+ (87), 369.2 [M+Na]+ (33).
+
HREIMS m/z 369.1520 [M+Na]+ (calcd for C16H26NaO8 , 369.1525); Rf= 0.30 (Cyh:EtOAc=4:1,
1% AcOH, v/v).
4.5.8. di-tert-Butyl-(R)-2-acetoxy-(S)-3-tert-butoxycarbonyl-pentanedioic ester (11)
N,N’-diisopropyl-O-tert-butylisourea (1.68 g, 8.40 mmol, 10.0 eq) was added to a solution of
acid 10 (0.29 g, 0.84 mmol, 1.0 eq.) in DCM (18.0 mL). The reaction mixture was refluxed for
48 hours and then filtered through a short pad of celite. The filtrate was concentrated in vacuo.
The residue was purified by column chromatography (Cyh:EtOAC=4:1, v/v) to obtain ester 11 as
colorless syrup (0.32 g, 0.78 mmol, 94%). [ꢀ]ꢂꢁꢃ +15.6 (c 0.32, Et2O); IR (ATR) νmax 2979, 1730,
1
1368, 1222, 1144 cm-1; H NMR (400 MHz, CDCl3) δ 5.15 (d, J = 3.5 Hz, 1H), 3.36 (ddd, J =
9.6, 5.1, 3.3 Hz, 1H), 2.67 (dd, J = 16.7, 9.6 Hz, 1H), 2.38 (dd, J = 16.8, 5.2 Hz, 1H), 2.11 (s,
15

ACCEPTED MANUSCRIPT
3H), 1.49 (s, 9H), 1.48 (s, 9H), 1.47 (s, 9H); 13C NMR (100 MHz, CDCl3) δ 170.5, 169.9, 169.1,
166.7, 82.8, 81.9, 80.9, 72.1, 43.7, 33.7, 28.0, 28.0, 28.0, 20.5; EIMS m/z (%) 420.3 [M+NH4]+
+
(24); HREIMS m/z 425.2145 [M+Na]+ (calcd for C20H34NaO8 , 425.2151); Rf= 0.74
(Cyh:EtOAc=4:1, v/v).
4.5.9. di-tert-Butyl-(S)-3-tert-butoxycarbonyl-(R)-2-hydroxy-pentanedioate (12)
K2CO3 (0.14 g, 1.04 mmol, 1.5 eq.) was added to the solution of ester 11 (0.28 g, 0.70 mmol, 1.0
°
°
eq.) in MeOH (6.0 mL) at 0 C. The reaction mixture was stirred at 0 C for 1 hour and then
poured onto sat. NH4Cl solution (10.0 mL). The aqueous layer was extracted with EtOAc. The
combined organic layers were washed with brine, dried over Na2SO4, filtered and concentrated
in vacuo. The crude product was purified by column chromatography (Cyh:EtOAc=9:1, v/v) to
obtain alcohol 12 as colorless oil (0.21 g, 0.58 mmol, 84%). [ꢀ]ꢂꢁꢃ +7.3 (c 0.67, DCM)18; IR
1
(ATR) νmax 3500, 2979, 1726, 1367, 1250, 1144 cm-1; H NMR (400 MHz, CDCl3) δ 4.17 (s,
1H), 3.27 (ddd, J=8.4, 6.2, 2.8 Hz, 1H), 3.14 (br. s, 1H), 2.72 (dd, J=16.7, 8.4 Hz, 1H), 2.48 (dd,
J=16.8, 6.2 Hz, 1H), 1.49 (s, 9H), 1.45 (s, 9H), 1.43 (s, 9H); 13C NMR (100 MHz, CDCl3) δ
172.3, 171.0, 170.2, 83.1, 81.7, 80.8, 70.8, 45.8, 33.9, 28.1, 28.0; EIMS m/z (%) 383.1 [M+Na]+
+
(27), 743.4 [2M+Na]+ (84). HREIMS m/z 383.2040 [M+Na]+ (calcd for C18H32NaO7 ,
383.2046); Rf=0.77 (Cyh:EtOAc=3:1, v/v).
4.5.10. 2-O-(3,4-Diacetoxycaffeoyl)-tri-tert-butyl isocitrate (13)
Pyridine (1.3 mL) and acid chloride 4 (0.18 g, 0.62 mmol, 1.5 eq.) were added to a solution of
alcohol 12 (0.15 g, 0.42 mmol, 1.0 eq.) and DMAP (7.60 mg, 0.06 mmol) in DCM (4.2 mL).
After stirring at rt overnight, the reaction mixture was diluted with DCM, washed with 1 M HCl
(aq.) and brine, dried over Na2SO4 and concentrated in vacuo. The crude product was purified by
16

ACCEPTED MANUSCRIPT
column chromatography (Cyh:EtOAc=4:1, v/v) to obtain ester 13 as colorless oil (0.22 g, 0.36
mmol, 86%). [ꢀ]ꢂꢁꢃ -6.3 (c 0.42, DCM); IR (ATR) νmax 2978, 2928, 1723, 1639, 1504, 1367,
1
1201, 1141, 1109 cm-1; H NMR (600 MHz, CDCl3) δ 7.66 (d, J = 16.0 Hz, 1H), 7.39 (dd, J =
8.4, 2.0 Hz, 1H), 7.35 (d, J = 2.0 Hz, 1H), 7.22 (d, J = 8.4 Hz, 1H), 6.41 (d, J = 16.0 Hz, 1H),
5.27 (d, J = 3.4 Hz, 1H), 3.43 (ddd, J = 9.6, 5.0, 3.6 Hz, 1H), 2.73 (d, J = 16.7, 9.7 Hz, 1H), 2.44
13
(d, J = 16.7, 5.0 Hz, 1H), 2.30 (s, 3H), 2.29 (s, 3H), 1.48 (s, 9H), 1.47 (s, 9H), 1.45 (s, 9H); C
NMR (150 MHz, CDCl3) δ 170.7, 169.4, 168.1, 168.0, 166.8, 165.5, 144.2, 143.9, 142.6, 133.2,
126.6, 124.1, 123.0, 118.4, 83.1, 82.1, 81.1, 72.4, 44.0, 34.0, 28.2, 28.2, 20.8, 20.7; EIMS m/z
+
(%) 624.3 [M+NH4]+ (100); HREIMS m/z 629.2568 [M+Na]+ (calcd for C31H42NaO12
,
629.2574); Rf= 0.20 (Cyh:EtOAc=4:1, v/v).
4.5.11. (E)-Caffeoylisocitric acid (1)
Fresh distilled TFA (0.62 mL, 8.10 mmol, 23.8 eq.) was added slowly to the solution of ester 13
(0.21 g, 0.34 mmol, 1.0 eq.) in DCM (5.0 mL). The reaction mixture was stirred at rt for 48
hours. Then it was concentrated in vacuo to dryness, redissolved in acetone, reconcentrated and
dried to afford a white powder (0.15 g, quantitative). This white powder (0.075 g, 0.17 mmol, 1.0
eq.) was dissolved in acetone (11.0 mL). 3 M HCl (aq.) (3.40 mL, 60.0 eq.) was then added. The
reaction mixture was stirred at 75 °C for 5 h. After cooling down the reaction mixture was diluted
with EtOAc. The organic layer was washed with brine, dried over Na2SO4 and concentrated in
vacuo. The residue was purified chromatographically by a RP-C18 column (water: MeCN=19:1,
1% AcOH, v/v) to afford acid 1 as brown oil (0.037 g, 0.10 mmol, 61%). [ꢀ]ꢂꢁꢃ 11.4 (c 0.46,
DCM); IR (ATR) νmax 2980, 1694, 1596, 1146 cm-1. 1H NMR (600 MHz, MeOD-d4) δ 7.61 (d, J
= 15.8 Hz, 1H), 7.07 (d, J = 1.8 Hz, 1H), 6.97 (dd, J = 8.2, 1.8 Hz, 1H), 6.79 (d, J = 8.2 Hz, 1H),
6.32 (d, J = 15.8 Hz, 1H), 5.47 (d, J = 3.36 Hz, 1H), 3.57 (s, 1 H), 2.82 (dd, J = 17.0, 9.0 Hz,
17

ACCEPTED MANUSCRIPT
1H), 2.60 (dd, J = 17.0, 5.07 Hz, 1H); 13C NMR (150 MHz, MeOD-d4) δ 174.9, 173.8, 168.0,
149.8, 148.1, 146.8, 127.7, 123.3, 116.5, 115.3, 114.1, 73.1, 44.4, 33.4; EIMS m/z (%) 353.1 [M-
H]- (33), 191.0 [M-caffeoyl]- (21), 173.0 [M-caffeoyl-H2O]- (39), 111.0 [M-caffeoyl-2H2O-CO2]-
-
(19); HREIMS m/z 353.0513 [M-H]- (calcd for C15H13O10 , 353.0509); Rf= 0.27 (RP,
water:MeCN=19:1, 1% AcOH, v/v).
4.5.12. 3,4-Diacetoxy-d6-cinnamic acid-d1 (16)
3,4-Dihydroxybenzaldehyde (15) (0.35 g, 2.51 mmol, 1.0 eq.) and sodium acetate-d6 (0.34 g,
3.97 mmol, 1.6 eq.) were refluxed in acetic anhydride-d6 (2.5 mL, 25.0 mmol, 10.0 eq.) for 48
hours. The reaction mixture was then diluted with EtOAc and treated with a solution of Na2CO3
(0.10 g) in deuterium oxide (10.0 mL). The aqueous layer was acidified with a few drops of 12
M HCl (aq.) and extracted with EtOAc. The organic layer was washed with brine, dried over
Na2SO4, filtered and concentrated in vacuum. The residue was purified by column
chromatography (Cyh: EtOAc=4:1, 1% AcOH, v/v) to obtain 16 as white powder (0.073 g, 0.27
mmol, 11%). IR (ATR) νmax 2544, 1759, 1668, 1629, 1501 cm-1; 1H NMR (600 MHz, Acetone-
d6) δ 7.65 (s, 1H), 7.60 (m, 2H), 7.31 (d, J= 8.2 Hz, 1H), 6.53 (d, J= 16.0Hz, 0.05H); 13C NMR
(150 MHz, Acetone-d6) δ 169.3, 169.2, 168.2, 145.6, 144.6, 144.3, 134.8, 127.9, 125.7, 124.5;
EIMS m/z (%) 272.1 [M+H]+ (24), 289.1 [M+NH4]+ (49), 294.1 [M+Na]+ (40); HREIMS m/z
+
294.0965 [M+Na]+ (calcd for C13H5D7NaO6 , 294.0971); Rf= 0.26 (DCM:MeOH=24:1, v/v).
4.5.13. 3,4-Diacetoxy-d6-cinnamoyl-d1 chloride (17)
Thionyl chloride (0.13 mL, 1.74 mmol, 6.9 eq.) and DMF (1.0 ꢀL) were added to the solution of
acid 16 (0.069 g, 0.25 mmol, 1.0 eq.) in DCM (1.0 mL). After refluxing at 60 °C for 7 hours, the
18

ACCEPTED MANUSCRIPT
solvent and excess thionyl chloride were removed in vacuo. The residue was used directly
without any further purification.
4.5.14. 2-O-(3,4-Diacetoxy-d6-cinnamoyl-d1)-tri-tert-butyl isocitrate (18)
Pyridine (0.5 mL) and acid chloride 17 (0.073 g, 0.25 mmol, 1.5 eq.) were added to a solution of
alcohol 12 (0.061 g, 0.17 mmol, 1.0 eq.) and DMAP (0.003 g, 0.03 mmol) in DCM (2.1 mL).
After stirring at rt overnight, the reaction mixture was diluted with DCM, washed with 1 M HCl
(aq.) and brine, dried over Na2SO4 and concentrated in vacuo. The crude product was purified by
column chromatography (Cyh:EtOAc=4:1, v/v) to obtain ester 18 as colorless oil (0.015 mg,
0.02 mmol, 14%). IR (ATR) νmax 2979, 2932, 1723, 1504, 1368, 1210, 1143 cm-1; 1H NMR (600
MHz, CDCl3) δ 7.66 (s, 1H), 7.39 (dd, J= 8.4, 2.0 Hz, 1H), 7.35 (d, J= 2.0 Hz, 1H), 7.23 (d, J=
8.4 Hz, 1H), 6.41 (d, J= 16.0 Hz, 0.18H), 5.27 (d, J= 3.5 Hz, 1H), 3.44 (ddd, J= 9.8, 4.9, 3.4 Hz,
1H), 2.73 (dd, J= 16.7, 9.7 Hz, 1H), 2.44 (dd, J= 16.8, 5.0 Hz, 1H), 1.48 (s, 9H), 1.48 (s, 9H),
1.45 (s, 9H); 13C NMR (150 MHz, CDCl3) δ 170.5, 169.2, 168.0, 167.9, 166.7, 165.3, 144.0,
143.7, 142.5, 133.1, 126.5, 123.9, 122.9, 118.2, 82.9, 81.9, 81.0, 72.3, 43.9, 33.8, 28.1, 28.0;
EIMS m/z (%) 631.4 [M+NH4]+ (100); HRMS (ESI) m/z 636.3008 [M+Na]+ (calcd for
C31H35D7NaO12+, 636.3013); Rf= 0.20 (Cyh:EtOAc=4:1, v/v).
4.5.15. (E)-Caffeoylisocitric acid-d1 (1D)
Fresh distilled TFA (0.02 mL, 0.26 mmol, 6.5 eq.) was added slowly to the solution of ester 18
(0.022 g, 0.04 mmol, 1.0 eq.) in DCM (1.0 mL). The reaction mixture was stirred at rt for 48
hours. Then it was concentrated in vacuo to dryness, redissolved in acetone, reconcentrated and
dried to afford a white powder. This white powder (0.016 g, 0.04 mmol, 1.0 eq.) was dissolved in
acetone (1.5 mL). 3 M HCl (aq.) (0.5 mL, 1.5 mmol, 37.5 eq.) was then added. The reaction
19

ACCEPTED MANUSCRIPT
°
mixture was stirred at 75 C for 5 h. After cooling down the reaction mixture was diluted with
EtOAc. The organic layer was washed with brine, dried over Na2SO4 and concentrated in vacuo.
The residue was purified chromatographically by a RP-C18 column (water: MeCN=19:1, 1%
1
AcOH, v/v) to obtain acid 1D as brown oil (0.0084 g, 0.02 mmol, 67%). H NMR (600 MHz,
MeOD-d4) δ 7.63 (s, 1H), 7.09 (d, J=1.9 Hz, 1H), 7.00 (dd, J=8.2, 2.0 Hz, 1H), 6.81 (d, J=8.1
Hz, 1H), 6.35 (d, J=15.8 Hz, 0.2H), 5.49-5.48 (m, 1H), 3.59-3.58 (m, 1H), 2.85 (dd, J=17.0, 9.0
13
Hz, 1H), 2.62 (dd, J=17.0, 4.9 Hz, 1H); C NMR (150 MHz, MeOD-d4) δ 175.0, 168.0, 149.8,
148.0, 146.8, 127.7, 123.3, 116.5, 115.3, 114.1, 73.7, 44.4, 33.4; EIMS m/z (%) 354.3 [M-H]-
(3); HRMS (ESI) m/z 378.0542 [M+Na]+ (calcd for C15H13DNaO10+, 378.0547); Rf= 0.27 (RP,
water:MeCN=19:1, 1% AcOH, v/v).
Conflicts of interest
There are no conflicts of interest to declare.
Acknowledgements
The authors thank Ms. I. Polanz, Ms. S. Bettinger and Mr. A. Siebert for measuring the high-
resolution MS and NMR spectra, Prof. R. Maul for providing us with amaranth seeds and G.
Scherkenbeck for breeding the amaranth plants.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
References
[1] S. Quideau, D. Deffieux, C. Douat-Casassus, L. Pouységu, Angew. Chem. 123 (2011) 610-
646.
20

ACCEPTED MANUSCRIPT
[2] M. Daglia, Current Opinion in Biotechnology. 23 (2012) 174-181.
[3] W.E. Hardman, Nutrition Res. Pract. 8 (2014) 233-240.
[4] R.K. Lall, D.N. Syed, V.M. Adhami, M.I. Khan, H. Mukhtar, Internat. J. Mol. Sci. 16 (2015)
3350-3376.
[5] C. Ramassamy, Eur. J. Pharmacol. 545 (2006) 51-64.
[6] G.G. Duthie, S.J. Duthie, J.A.M. Kyle, Nutrition Res. Rev. 13 (2000) 79-106.
[7] A. Scalbert, G. Williamson, J. Nutrition. 130 (2000) 2073S-2085S.
[8] R. Cernansky, Nature. 522 (2015) 146-148.
[9] S. Neugart, S. Baldermann, B. Ngwene, J. Wesonga, M. Schreiner, Food Res. Int. 100 (2017)
411-422.
[10] D. Schröter, S. Baldermann, M. Schreiner, K. Witzel, R. Maul, S. Rohn, S. Neugart, Food
Chem. 267 (2018) 376-386.
[11] I. Parveen, A. Winters, M.D. Threadgill, B. Hauck, P. Morris, Phytochemistry. 69 (2008)
2799-2806.
[12] D. Strack, P. Leicht, M. Bokern, V. Wray, L. Grotjahn, Phytochemistry. 26 (1987) 2919-
2922.
[13] M. Barontini, R. Bernini, I. Carastro, P. Gentili, A. Romani, New J. Chem. 38 (2014) 809-
816.
21

ACCEPTED MANUSCRIPT
[14] T. Hill, L.R. Odell, J.K. Edwards, M.E. Graham, A.B. McGeachie, J. Rusak, A. Quan, R.
Abagyan, J.L. Scott, P.J. Robinson, A. McCluskey, J. Med. Chem. 48 (2005) 7781-7788.
[15] W.E. Zeller, Synth.Commun. 43 (2013) 1345-1350.
[16] M.F.A. Amer, K. Takahashi, J. Ishihara, S. Hatakayama, Heterocycles. 72 (2007) 181-185.
[17] F. Calo, J. Richardson, A.G.M. Barrett, J. Org. Chem. 73 (2008) 9692-9697.
[18] S. Tsegay, H. Hügel, M.A. Rizzacasa, Aust. J. Chem. 62 (2009) 676-682.
[19] D. Seebach, D. Wasmuth, Helv. Chim. Acta. 63 (1980) 197-200.
[20] M. Sefkow, A. Koch, E. Kleinpeter, Helv. Chim. Acta. 85 (2002) 4216-4229.
[21] G.A. Reed, D.R. Dimmel, E.W. Malcolm, J. Org. Chem. 58 (1993) 6364-6367.
22