Methylation of cyanidin-3-o-glucoside with dimethyl carbonate

Article abstract of DOI:10.3390/molecules26051342

The approach presented in this study is the first for the hemisynthesis of methylated anthocyanins. It was possible to obtain cyanidin-3-O-glucoside derivatives with different degrees of methylation. Cautious identification of 4'-, 5-, and 7-OH monomethylated derivatives was also accomplished. The methylation agent used was the “green chemical” dimethyl carbonate (DMC), which is characterized by low human and ecological toxicity. The influence of the temperature, reaction time, and amount of the required diazabicyclo[5.4.0]undec-7-en (DBU) catalyst on the formation of the products was examined. Compared to conventional synthesis methods for methylated flavonoids using DMC and DBU, the conditions identified in this study result in a reduction of reaction time, and an important side reaction, so-called carboxymethylation, was minimized by using higher amounts of catalyst.

Full text of DOI:10.3390/molecules26051342

molecules
Article
Methylation of Cyanidin-3-O-Glucoside with
Dimethyl Carbonate
Sarah Straßmann, Tillman Brehmer, Maike Passon * and Andreas Schieber
Department of Nutritional and Food Sciences, Molecular Food Technology, University of Bonn,
Friedrich-Hirzebruch-Allee 7, D-53115 Bonn, Germany; sarah.strassmann@magenta.de (S.S.);
brehmer@uni-bonn.de (T.B.); schieber@uni-bonn.de (A.S.)
*
Correspondence: passon@uni-bonn.de; Tel.: +49-228-73-4107
Abstract: The approach presented in this study is the first for the hemisynthesis of methylated
anthocyanins. It was possible to obtain cyanidin-3-O-glucoside derivatives with different degrees
0
of methylation. Cautious identification of 4 -, 5-, and 7-OH monomethylated derivatives was also ac-
complished. The methylation agent used was the “green chemical” dimethyl carbonate (DMC),
which is characterized by low human and ecological toxicity. The influence of the temperature,
reaction time, and amount of the required diazabicyclo[5.4.0]undec-7-en (DBU) catalyst on the forma-
tion of the products was examined. Compared to conventional synthesis methods for methylated
flavonoids using DMC and DBU, the conditions identified in this study result in a reduction of
reaction time, and an important side reaction, so-called carboxymethylation, was minimized by using
higher amounts of catalyst.
ꢀ
ꢁꢂꢀꢃꢄꢅꢆꢇ
ꢀ
ꢁꢂꢃꢄꢅꢆ
Keywords: cyanidin-3-O-glucoside; cyanidin; anthocyanins; LC-IMS; HRMS; methylation; DMC;
phase II metabolites
Citation: Straßmann, S.; Brehmer, T.;
Passon, M.; Schieber, A. Methylation
of Cyanidin-3-O-Glucoside with
Dimethyl Carbonate. Molecules 2021,
2
6, 1342. https://doi.org/10.3390/
1. Introduction
molecules26051342
Anthocyanins are natural plant colorants and are a component of fruit and vegeta-
bles in the human diet [1]. Like other polyphenols, they show antioxidant properties,
Academic Editors: Toyonobu Usuki,
Akira Kawamura and
e.g., against radical oxygen species. As such, they are associated with the prevention of
certain diet-related illnesses, such as cardiovascular disease and cancer [2]. The increased
interest in anthocyanin-rich nutrition is accompanied by the elucidation of the physiologi-
cal relationships between intake and metabolism. Metabolites of anthocyanins are inter alia
compounds glucuronidated, sulfated, and/or methylated at the aglycone [3]. Knowledge of
the structure and quantity of these metabolites after consumption of anthocyanin-rich meals
in the human body provides information about the intake and bioavailability and thus,
Masaru Hashimoto
Received: 5 February 2021
Accepted: 24 February 2021
Published: 3 March 2021
in the long term, about the health significance of anthocyanins in vivo [4]. For such investi-
gations, standard substances are required, and their production gives rise to new research
areas [5].
In living organisms, conjugation of flavonoids having a catechol structure with
methyl groups is carried out by catechol-O-methyl transferase enzymes (COMT) and
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
S-adenosylmethionine (SAM) as a coenzyme [6]. Since cyanidin has this catechol struc-
ture only on the B ring, there is no methylation of the A ring. Consequently, methylated
0
derivatives observed so far in vivo carry the methyl group either at the 3 -(peonidin-3-
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
0
O-glucoside) or at the 4 -hydroxyl group of cyanidin-3-O-glucoside (isopeonidin-3-O-
glucoside) [7]. The metabolites themselves can either be excreted directly or be subject to
further metabolism. The physiological properties of the metabolites may differ from those
of the native starting compounds [8].
Metabolically significant methylated reference substances can be synthesized using
different approaches. A promising approach for glucuronidation, sulfation, and methy-
lation is enzymatic hemisynthesis, but costs and efforts are too high to result in usable
4
.0/).
Molecules 2021, 26, 1342. https://doi.org/10.3390/molecules26051342
https://www.mdpi.com/journal/molecules

Molecules 2021, 26, 1342
2 of 14
quantities [
9
]. Due to the enzyme properties, another drawback of this synthesis route is
the lack of the ability to generate derivatives with more than one methyl group, which can
be relevant for already genuinely methylated compounds.
Thus, chemical synthesis is more promising. This may either mean full synthesis, as
performed successfully by Cruz et al. [10], or hemisynthesis. The advantage of hemisynthe-
sis is that only a few steps are required to achieve the goal. Initially, a suitable methylation
agent is required. For flavonoids such as quercetin or catechin, successful syntheses
with dimethyl sulfate (DMS) or methyl iodide, also known as iodomethane, have been
described [5,11]. In addition, mono-, di- and tetramethylated naringenin derivatives
were synthesized by using anhydrous acetone and N,N-dimethylformamide (DMF), re-
spectively [12]. Since DMS and methyl iodide are toxic and carcinogenic, working with
dimethyl carbonate (DMC) is an attractive alternative.
For DMC, two reactions are of particular importance: carboxymethylation, also known
as methoxycarbonylation, at low temperatures, and methylation at temperatures above
◦
1
20 C [13]. Both reactions are shown in Scheme 1.
Scheme 1. Possible reactions of cyanidin-3-O-glucoside with DMC and DBU.
The crucial difference in carboxymethylation and methylation lies in the mechanism
of base-mediated ester hydrolysis of DMC. Successful O-methylation (in the following
referred to as methylation) depends on a suitable catalyst to prevent carboxymethyla-
tion [14]. Methylation and carboxymethylation are also dependent on the type of al-
cohol [15]. Thus, methylation is preferred on aryl alcohols and carboxymethylation on
aliphatic alcohols. Most syntheses of phenols with DMC take place at temperatures above
◦
its boiling point, i.e., between 120 and 200 C [16
,17]. Considering the sensitivity of an-
thocyanins to high temperatures and long reaction times [18], this is rather critical for
their methylation. By using diazabicyclo[5.4.0]undec-7-ene (DBU) as a catalyst, the temper-
◦
ature of the methylation can be reduced below 100 C, [19] which is nevertheless still too
high for the direct synthesis of the cyanidin aglycone.
The synthesis of methylated anthocyanins for the exploration of metabolites in phys-
iological samples poses a special challenge, as only aglycone-substituted compounds
are relevant. Although the aglycone is an aryl alcohol, the glucose moiety is an aliphatic al-
cohol. These types differ in the way they react with DMC. Thus, the aim of the synthesis
was to determine the reaction conditions for highly selective O-methylation of the agly-
cone without carboxymethylation of the sugar. For cyanidin-3-O-glucoside, these are the
derivatives methylated at the B ring [7]. Furthermore, the yield of mono- and dimethy-
lated derivatives should be maximized as much as possible, as they are interesting from a
metabolic point of view. At the same time, the amount of higher methylated compounds
should be minimized.
2. Results and Discussion
Using the methods described above, three monomethylated, three dimethylated,
three trimethylated, and one tetramethylated derivative were generated (Table 1). In addi-

Molecules 2021, 26, 1342
3 of 14
tion, three carboxymethylated compounds were detected. The path that led to these results
and the identification of the monomethylated derivatives is described below.
Table 1. LC-DAD-ESI-IMS-qTOF results of the synthesis reaction of methylated cyanidin-3-O-glucoside.
+
+
Compound
Number
[M] (m/z)
[M] (m/z)
Mass Error
(mDa)
CCS Value
[M] (Å )
Fragment
Ions (m/z)
RT (min)
Yield (%) a
λmax (nm)
+
2
Observed
Calculated
Reference Substances
cyanidin-3-O-glucoside
21 b
4
4
49.1078
63.1234
449.1078
±0.0
3.42
202.8 ± 0.4
287.0551
518
519
peonidin-3-O-glucoside
2
86.0471,
01.0707
463.1235
−0.1
4.19
207.0 ± 0.3
Methylated Compounds
monomethylated cyanidin-3-O-glucoside
3
8
2
3
2
3
2
3
86.0470,
01.0707
86.0473,
01.0704
86.0471,
01.0706
1
2
3
463.1230
463.1228
463.1231
463.1235
463.1235
463.1235
−0.5
−0.7
−0.4
3.95
4.16
4.36
207.2 ± 0.2
208.1 ± 0.3
207.3 ± 0.3
520
517
517
dimethylated cyanidin-3-O-glucoside
3
1
3
3
00.0624,
15.0861
315.0862
00.0638,
15.0858
6
7
8
477.1386
477.1397
477.1390
477.1391
477.1391
477.1391
−0.5
+0.6
−0.1
4.60
4.99
5.17
213.5 ± 0.4
212.9 ± 0.5
212.3 ± 0.1
516
518
516
3
3
trimethylated cyanidin-3-O-glucoside
9
491.1547
491.1543
491.1548
491.1548
491.1548
491.1548
−0.1
−0.5
±0.0
5.38
5.97
6.08
218.3 ± 0.2
216.5 ± 0.4
218.3 ± 0.3
329.1019
329.1026
329.1019
520
518
517
1
1
0
1
tetramethylated cyanidin-3-O-glucoside
505.1704 6.97
−1.8
carboxymethylated cyanidin-3-O-glucoside
1
4
12
505.1686
223.3 ± 0.2
343.1201
518
13
14
15
507.1132
507.1134
507.1134
507.1133
507.1133
507.1133
−0.1
+0.1
+0.1
4.45
4.56
5.00
215.7 ± 0.3
212.2 ± 0.3
212.8 ± 0.6
287.0544
287.0549
287.0551
515
515
521
a
expressed as peonidin-3-O-glucoside equivalents b remaining after reaction.
2.1. First Observations
Simple hemisyntheses of monomethylated quercetin derivatives with DMS take be-
tween 4 and 24 h, depending on the solvents and bases used [20]. The synthesis of
monomethylated flavonoids with DMC requires 24 to 36 h according to the literature [21].
In the first experiments with 10 mg of anthocyanin extract, 2 mL DMC, and 0.3 mmol DBU
◦
at 90 C, the compounds that were fully methylated could already be identified as the
main products after 22 h. The samples contained methylated cyanidin-3-O-glucoside at
+
a mass to charge ratio (m/z) 505 and 563 ([M] ). The former is the derivative that is fully
methylated at the aglycone, whereas the latter is carboxymethylated in addition to per-
methylation. No monomethylated derivatives were detectable in the samples examined.
For this reason, it was decided to shorten the reaction time to stop the reaction at the
point where compounds with a lower degree of methylation, such as monomethylated
+
+
molecules at m/z 463 ([M] ) and dimethylated molecules at m/z 477 ([M] ) are formed.

Molecules 2021, 26, 1342
4 of 14
The optimal reaction time, temperature, and quantity of the catalyst for the formation
of noncarboxymethylated monomethylated cyanidin-3-O-glucoside derivatives can be
determined with the help of a randomized, central composite design.
2.2. Experimental Design
In the synthesis presented here, it is presumed that the nonnucleophilic base, DBU,
deprotonates cyanidin-3-O-glucoside. In addition, DBU reacts with DMC via the preferred
carboxymethylation mechanism to generate a more activated methylating agent. This re-
acts in the next step with the deprotonated cyanidin-3-O-glucoside to form the desired
methylated compound while the catalyst is regenerated. The resulting methoxycarbonate
anion decays rapidly to CO , and, with a proton, to methanol. It becomes clear that the
2
DBU added is only a catalyst. The mechanism was previously described [13,19], but still
needs to be elucidated in more detail. Since anthocyanins have a pH-dependent struc-
ture, this certainly plays a major role in the reaction. It is expected that the flavylium
cation form, which is present in acidic environments, is more stable and less reactive than
the deprotonated form.
◦
Figure 1 shows the relationship between the yield of synthesized compounds at 90 C
for six hours and the volume of catalyst.
◦
Figure 1. Yield of the synthesis products of cyanidin-3-O-glucoside with 2 mL DMC at 90 C for 6 h as a function of the
volume of catalyst DBU.
The tri- and tetramethylated compounds are formed only at higher amounts of DBU,
while the amount of mono- and dimethylated derivatives decreases at this point. It should
be mentioned that small amounts of peonidin-3-O-glucoside were sometimes detected in
the blackberry extract used as the starting material. A consideration of the undesired side
reaction of carboxymethylation shows that these reactions can be controlled by the con-
+
centration of DBU. Without catalyst, the product at m/z 507 ([M] ) is not formed, but with
0.6 mmol DBU, its yield is over 6%. This small amount of DBU is immediately consumed as
a base and cannot be used to generate the methylation agent. In this way, the anthocyanin
reacts directly with DMC, so carboxymethylation is preferred. This consumption of DBU is
also described in the literature [19]. With increasing catalyst quantities, a decrease in car-
boxymethylated cyanidin-3-O-glucoside can be observed. With more than 1.2 mmol DBU,
the concentration of carboxymethylated cyanidin-3-O-glucoside is no longer significant.

Molecules 2021, 26, 1342
5 of 14
The result is plausible since DBU reacts as a base and creates the methylating agent
from DMC, and thus, only the methyl group is transferred to the cyanidin-3-O-glucoside.
This pathway also produces CO as the driving force of the reaction.
2
These assumptions could be confirmed in tests with different concentrations of the
starting compounds but the same amount of catalyst. Increasing the amount of methylating
agent only promotes the side reaction of carboxymethylation. In one of the reactions
investigated, a yield greater than 8% was achieved. From the synthesis of p-cresol to 4-
methyl anisole described in the literature, it can be concluded that a substrate concentration
that is too high slows down the reaction kinetics. The reason for this effect is the possibility
of the substrate interfering with the formation of the phenolate ion, which is necessary for
the transfer of the methyl group, by forming hydrogen bonds between them [22]. A similar
connection is also likely for anthocyanins. With 1 mL of DMC but the same anthocyanin-
to-catalyst ratio, the yield decreases compared to the optimized conditions with 2 mL.
Thus, it may be concluded that in case of an excess of methylating agent and a too low
concentration of the catalyst, the direct contact of DMC with the anthocyanin leads to
carboxymethylation of the sugar residue.
The experimental design also confirmed the assumption that carboxymethylation
◦
is favored below 90 C. The experiment resulted in more than one optimal method for
maximum yield of monomethylated compounds, while at the same time minimizing the
other products. However, since the shortest possible reaction time at high temperatures
and high pH values is desirable for anthocyanins, the following conditions were considered
◦
optimal: 95 C for 90 min with 2.95 mmol of DBU. Under these conditions, a total yield
of 8% monomethylated cyanidin-3-O-glucoside and 4% carboxymethylated cyanidin-3-O-
glucoside was achieved. After the reaction, approximately 21% of cyanidin-3-O-glucoside
remained unreacted. Approximately 40% of the compounds present after the reaction
could be characterized (Table 1). Among the remaining 60%, the degradation products of
cyanidin-3-O-glucoside, phloroglucine aldehyde, protocatechuic acid, and their methy-
lated derivatives, were detected in the synthesis mixture. Whether these compounds
are degradation products of methylated cyanidin-3-O-glucoside derivatives, or whether
the methylation of these compounds took place after the degradation of cyanidin-3-O-
glucoside, was not investigated.
2.3. Increase in Solvation of Anthocyanins
During the experiments, it was observed that although the solution took on a red color,
some freeze-dried anthocyanin particles were still visible as a suspension in the DMC.
This observation is also mentioned in the literature on the methylation of quercetin, where
the yield after 72 h is only 30% [21]. To promote the conversion of anthocyanins, in addition
to controlling the reaction by adjusting temperature, time, and catalyst concentration,
measures to increase the solubility were considered. With this aim, experiments with the
application of ultrasound and different solvents were used.
2.3.1. Ultrasound
The overall influence of ultrasound on the synthesis was small. Compared to the
synthesis under the optimized conditions, neither an improvement nor a deterioration
was observed. It has been reported that the yield of the O-alkylation of 5-hydroxy-4-oxo-
4
H-chromene-2-carboxylic acid could be increased from 28% to 97% by ultrasound [23].
To achieve this, the authors performed 90-min continuous ultrasound treatment with
an ultrasonic sonotrode at a frequency of 20 kHz while the system continued to be ex-
ternally heated. Such an experimental setup could not be realized for the synthesis of
◦
methylated cyanidin-3-O-glucoside derivatives at the optimal temperature at 95 C because
the sonotrode (Hielscher Ultrasonics GmbH, UP200St) can be used only up to tempera-
◦
tures of approximately 70 C. Furthermore, ultrasonic treatment during chemical synthesis
might result in significant degradation of the products. The mechanisms associated with
ultrasound are radical reactions [24]. Those types of reactions, against which anthocyanins

Molecules 2021, 26, 1342
6 of 14
may have a stabilizing effect due to their chemical properties, may then lead to countless
isomeric polymeric compounds that are difficult to detect [18].
2
.3.2. Solvents
Another approach to obtain more initially dissolved cyanidin-3-O-glucoside molecules
was the addition of different solvents. The synthesis followed the procedure mentioned
under Section 3.2, whereby the sample substance was dissolved in 200 L solvent such
µ
as methanol, DMF, dimethyl sulfoxide (DMSO), water, or pyridine before the addition of
DMC and DBU and stirred vigorously. None of the selected solvents were able to increase
the yield of monomethylated anthocyanins compared to a reaction without the addition
of solvents. In fact, the yield actually decreased. The yield of the higher methylated
cyanidin-3-O-glucoside derivatives did not exceed 2%. Although the solutions were per-
ceived as more homogeneous before and after the reaction, the yield could not be improved.
Although to our knowledge there are no reports of a reaction with the abovementioned
solvents except water, where hydrolysis of DMC may occur, such reactions cannot be
excluded. It is also possible that a solvent layer forms around the cyanidin-3-O-glucoside,
and the DMC is unable to react with it. The loss of cyanidin-3-O-glucoside increased for all
solvents: only between 2 and 4% of the initial amount could be detected without producing
more methylated compounds.
A plausible explanation for this observation might be an improved heat transfer to
the cyanidin-3-O-glucoside due to the increased solubility, and thus increased surface
and contact possibility. When water is used as the solvent, it is possible that a water
molecule reacts with the anthocyanin to form the colorless chromenol, which would no
longer be detectable at 520 nm [8]. It should also be taken into account that nucleophilicity
is strongly dependent on the solvent. Although DMSO and DMF should not have any
influence on the nucleophilicity, as they are not able to form hydrogen bridge bonds,
water can have an influence even in small quantities, as it can build hydrate shells around
the cyanidin-3-O-glucoside hydroxyl groups.
The method presented may be further optimized using protective groups, as employed
by Bouktaib et al., [5] or the use of phase transfer catalysts, as proposed by Tundo et al. [13].
2
.4. Identification
Identification of the synthesized compounds was first carried out with ultra-violet
UV)/Vis detection at 520 nm, since anthocyanins have an absorption maximum in this re-
(
gion, and then by mass spectrometric analysis of the molecular and fragment ions using
ultra-high-performance liquid chromatography coupled to electrospray ionization multi-
n
stage mass spectrometry (UHPLC-ESI-LIT-MS ). Only those compounds that showed a
+
2
loss of glucose ([M
(
−
162] ) in the MS spectrum followed by the loss of a methyl group
+
3
[M − 162 − 15] ) in the MS spectrum were identified as methylated anthocyanins. To de-
termine the accurate mass and collision cross section (CCS) values, LC-IMS-qTOF-MS
was used. Table 1 gives a summary of the identification characteristics of cyanidin-3-O-
glucoside and its derivatives at their respective degrees of methylation.
2.4.1. Retention Times (RT)
Regarding the elution order in Figure 2, it is evident that methylated cyanidin-3-O-
glucoside derivatives are, as expected, less polar. The higher the degree of methylation,
the later the compounds are eluted. It can also be observed that the retention time of
peak
2 is in accordance with that of the reference compound peonidin-3-O-glucoside
0
(
3 -OH methylated). Another methylated cyanidin-3-O-glucoside which may be used as
0
reference is described in the literature as isopeonidin-3-O-glucoside (4 -OH methylated).
It is formed as a metabolite of human and porcine enzymes [
1
,
7
]. Considering only those
0
reports of authors who used acetonitrile (ACN) and water as eluents, the 3 -OH methylated
0
compounds eluted earlier than the 4 -OH. Regarding the elution order of methylated
0
quercetin or catechin derivatives, the literature indicates also that 3 -OH is eluted before

Molecules 2021, 26, 1342
7 of 14
0
4
-OH, and 5-OH earlier than 7-OH [25
–
27]. Although the structure is different from
the anthocyanins, it may be an orientation in elucidation.
Figure 2. UHPLC-MS chromatograms of the reaction of cyanidin-3-O-glucoside with 2 mL DMC
+
+
and 3.01 mmol DBU. (
monomethylated cyanidin-3-O-glucoside, peaks 1–3). (
20 nm.
A
), EIC of the m/z 449 ([M] , cyanidin-3-O-glucoside) and the m/z 463 ([M] ,
B
), UHPLC-DAD chromatogram extracted at
5
2
.4.2. UV Spectra
Considering shifts in UV spectra after conjugation can provide hints about the posi-
tion [28 29]. Unfortunately, the absorption maxima of the methylated derivatives are not
,
significantly lower than that of cyanidin-3-O-glucoside. The absorption maxima of the
monomethylated compounds and of cyanidin-3-O-glucoside are approximately 519 nm.
Considering the ratio of
λ
4
to λmax, differences become more noticeable. It can be seen
40 nm
that with 40% peak 2 and 3 have a different ratio than peak 1 with 33%, and the reference
substances cyanidin-3-O-glucoside, peonidin-3-O-glucoside, and 7-O-methylcyanidin-3-O-
galactoside have ratios of approximately 31%. The observations described in the literature,
0
where methylation of cyanidin at the 3 -OH position almost doubles the ratio, could not
be confirmed. However, the solvent used was methanol and not ACN [30]. Through the
sulfation of cyanidin (unpublished results), the slightest change in this ratio was found at
the B ring. If the UV shifts in the two reactions are similar, it can be assumed that peak
1 is
derivatized at the B ring and peaks 2 and 3 at the A ring.
2
.4.3. Mass Spectrometry
3
+
As the MS spectra of the monomethylated derivatives (m/z 463; [M] ) represent
+
the fragmentation behavior of the aglycone (m/z 301; [M
determine their substitution pattern. The fragment at m/z 273 is characteristic of methy-
lated cyanidin and marks the loss of CO of the aglycone ([M
same way, the fragment at m/z 259 results from the additional loss of the methyl group
−
162] ), these were used to
+
−
162
−
28] ). In the
+
(
m/z 287; [M − 162 − 15 − 28] ). All these fragments can also be found in their homolytic
cleavage product form [31]. Figure 3 shows fragments that can be identified as characteris-
tic of the A ring (m/z 150, 149, 139) and the B ring (m/z 151, 121, 109) [11 32]. A simple
,
consideration leads to the conclusion that depending on the substitution, characteristic
fragments a mass increment of 14 u are detectable in the spectra of the monomethylated
cyanidin-3-O-glucoside molecules. In the case of methylation of cyanidin-3-O-glucoside,
however, this consideration cannot be applied. Since the methyl group is such a small

Molecules 2021, 26, 1342
8 of 14
fragment, some of the masses may also belong to other fragments. For example, the
fragment at m/z 137 is described as a characteristic 0 B -fragment [32]. Nevertheless,
,2
+
1
,2
+
this mass may also belong to a methylated fragment of the m/z of [ B] (m/z 123 + 14
)
or to the 1 A -fragment [25]. Other examples are the fragment at m/z 185 (cyanidin:
,3
+
+
[
M − H O − CO − CO − CO] ), which may also belong to a methylated fragment of cyani-
2
+
din ([M + CH − H O − CO − CO − C H O] (171 + 14)), the fragment at m/z 161 (cyani-
3
2
2
2
+
1,4
+
din: [M
−
CO
−
CO
−
C H O
−
CO] or methylated cyanidin: [ B + CH − H O] ),
2
2
3
2
−
1
,2
+
0,2
+
and the fragment at m/z 123 (cyanidin: [ B] or methylated cyanidin: [ B + CH
CO]
3
(
109 + 14)) [25,32].
Figure 3. Postulated fragmentation scheme for cyanidin derivatives methylated on the A or B ring, based on the proposals
of Barnes & Schug [32] and González-Manzano et al. [25].
2
.4.4. Ion Mobility Spectrometry
Over the drift time of the substances, CCS values can be calculated. These depend
on the instrument settings, the m/z, and the structure and configuration of the molecule.
Based on the retention time, it might be concluded that peak
2
is peonidin-3-O-glucoside.
2
However, at 208.1
of the reference with 207.0
±
0.3 Å , the CCS value of peak
2
deviates too much from the CCS value
2
±
0.3 Å . Thus, the possibility that the cyanidin derivative
0
methylated at the 3 -OH position is among the three detected can be cautiously excluded.
2.4.5. Views on Reactivity and Order of Methylation
As the ratios of the three monomethylated peaks to each other are 43:39:18, only 3 methy-
lated derivatives were formed in significant quantities. Thus, the question remains regard-
ing which OH group of cyanidin-3-O-glucoside is least reactive. In the case of quercetin,
the 5-hydroxyl is by far the least acidic [25], whereas in cyanidin-3-O-glucoside, this po-
a a
sition is the one with the highest pk after 7-OH [33]. The pk values for the OH groups
in catechin are almost equal at approximately 9 [25]. Therefore, no regioselectivity in the

Molecules 2021, 26, 1342
9 of 14
methylation reaction of catechin is expected [25]. However, the lowest amount of catechin
monomethylated is observed in position 5, with slightly more monomethylated in 7. Ad-
0
0
ditionally, it is described that catechin is preferentially methylated at 3 and 4 -OH [27],
which may be due to a better accessibility.
In the past a methylation order that may be dependent on the acidity of the single OH
groups of quercetin was observed, where the most acidic group is methylated first [34].
This might be used to identify the methylation position in anthocyanins, as their OH groups
also show quite different acidities. According to this approach, the three most abundant
0
resulting derivatives of cyanidin-3-O-glucoside are 7, 5, and 4 -OH. However, it should be
noted that the conditions under which the work was carried out were completely different.
2.4.6. Hydrolysis
To obtain more precise information about the aglycones, a sample was hydrolyzed
2
with hydrochloric acid, as the hydrolysis allows the aglycone to be measured as MS
3
n
rather than as MS . The hydrolyzed sample was measured with both LC-ESI-LIT-MS and
LC-IMS-qTOF-MS and compared with a peonidin standard and a hydrolyzed sample of
Brazilian pepper extract. As Brazilian pepper contains 7-O-methylcyanidin galactoside, [28
]
a comparison of the glycosides was not sufficient for identification. The authors succeeded
in distinguishing peonidin and 7-O-methylcyanidin based on characteristic fragments for
+
the A ring and the B ring. Figure 4 shows an extracted ion chromatogram (EIC) of the [M] ,
m/z 301 of a hydrolyzed cyanidin-3-O-glucoside sample after methylation compared to a
hydrolyzed sample of Brazilian pepper and a peonidin reference.
+
Figure 4. Extracted ion chromatogram of the m/z 301 ([M] , methylated cyanidin) of the methylation of cyanidin-3-O-
glucoside after hydrolysis (B) compared to a hydrolyzed sample of Brazilian pepper (A) and a peonidin reference (C).
Table 2 provides information about the aglycones found. In comparison with the
reference substances, it can be seen from peak 18 that peonidin and 7-O-methylcyanidin
have the same retention time. The observation that peonidin and 7-O-methylcyanidin have
similar chromatographic behavior was already described before [30]. The two substances
have almost the same Rf value in TLC. Fortunately, IMS is more accurate as an identification
technique in this instance. The CCS values allow us to exclude that the detected peak 18
2
is peonidin because their values differ from each other by 160.3
±
0.5 Å for peak 18 and
2
2
1
59.1
±
0.1 Å for peonidin. However, the value of 7-O-methylcyanidin at 160.3
is identical to that of peak 18. Assuming that glucose does not change the elution order,
it can be cautiously concluded that peak 3 is 7-O-methylcyanidin-3-O-glucoside.
±
0.5 Å

Molecules 2021, 26, 1342
10 of 14
Table 2. LC-DAD-ESI-IMS-qTOF results of the synthesis reaction of methylated cyanidin-3-O-glucoside after acidic hydrolysis.
+
+
Compound
Number
[M] (m/z)
[M] (m/z)
Mass Error
(mDa)
CCS Value
[M] (Å )
λmax
(nm)
RT (min)
Fragment Ions MS (m/z)
+
2
Observed
Calculated
Reference Substances
cyanidin
2
13.0545, 231.06493, 241.04919,
2
87.0548
287.0550
−0.2
4.86
peonidin
5.98
154.7 ± 0.2
527
2
69.04454
2
01.05474, 229.04958, 258.05195,
86.0471
3
3
01.0711
01.0702
301.0707
301.0707
+0.4
159.1 ± 0.1
527
524
2
7
-O-methylcyanidin
5.99
160.3 ± 0.5
−0.5
286.04672, 258.05171
Methylated Compounds
monomethylated cyanidin
1
1
6
7
301.0706
301.0704
301.0707
301.0707
−0.1
5.48
5.71
160.6 ± 0.5
258.05262, 269.04454, 286.04732
258.05262, 286.04732
n.d.
528
−0.3
−0.6
159.4 ± 0.2
160.3 ± 0.5
2
01.05407, 213.05477, 229.04907,
58.05165, 286.04683
1
8
301.0701
301.0707
5.99
524
2
dimethylated cyanidin
19
20
21
315.0860
315.0860
315.0858
315.0863
315.0863
315.0863
−0.3
−0.3
−0.5
6.67
6.74
6.97
164.4 ± 0.3
164.7 ± 0.6
164.9 ± 0.3
272.06712, 300.06231
57.04412, 201.05023, 272.06728,
527
528
523
2
3
01.06719,
257.04467, 271.06016, 300.06297
trimethylated cyanidin a
3
14.07906, 286.08402, 271.06265,
57.08148
14.07861, 286.08251, 271.06018,
57.08116
13.07039, 314.0769, 285.07596,
68.07533
22
23
24
329.1011
329.1017
329.1014
329.1020
329.1020
329.1020
−0.9
7.78
7.94
8.09
170.0 ± 0.4
526
524
524
2
3
−0.3
−0.6
170.0 ± 0.2
170.9 ± 0.4
2
3
2
tetramethylated cyanidin
3
27.08619, 312.06264, 299.09129,
85.07529
25
343.1177
343.1176
+0.1 8.80
176.8 ± 0.2
522
2
a
one probably resulted from a carboxymethylated and trimethylated cyanidin-3-O-glucoside.
Regarding the absorption maxima at approximately 520 nm, the tendency to undergo
a hypsochromic shift as compared to cyanidin is more pronounced for the trimethylated
derivatives and the tetramethylated derivative than for mono- and dimethylated cyani-
din derivatives.
After hydrolysis, three monomethylated compounds can be found. Peak 16 is the small-
est, and the ratio of the three is 1:45:54. An explanation for only two large peaks remaining
after hydrolysis might be the coelution of two substances. On the other hand, one of the
three compounds may have been decomposed by hydrolysis. This would mean that one of
the methyl group positions might have a destabilizing effect on the cyanidin-3-O-glucoside
molecule. It has been described that methylation in the B ring decreases stability in neutral
media [35]. However, opposite observations under acidic conditions and high tempera-
tures, showing a stabilizing effect of the methyl group of peonidin are also described [36].
Concerning the other OH groups, no report could be found.

Molecules 2021, 26, 1342
11 of 14
3. Materials and Methods
3.1. Chemicals
LC-MS-grade water, acetonitrile, and formic acid were purchased from ChemSolute (Ren-
ningen, Germany). Methanol was obtained from VWR (Darmstadt, Germany). Dimethyl car-
bonate (DMC), dimethyl sulfoxide (DMSO), and 1,8-diazabicyclo[5.4.0]undec7.ene (DBU)
were obtained from Carl Roth (Karlsruhe, Germany). N,N-dimethyl formamide (DMF) was
purchased from Fisher Scientific (Waltham, Massachusetts). Peonidin and peonidin-3-O-
glucoside chloride were obtained from Phytoplan (Heidelberg, Germany).
An anthocyanin extract containing primarily cyanidin-3-O-glucoside (Scheme 2) was
obtained as described before [9].
Scheme 2. Structure of cyanidin-3-O-glucoside and its methylated derivatives.
3.2. Hemisynthesis
Ten milligrams of the above-described anthocyanin extract containing approximately
20 µmol cyanidin-3-O-glucoside was reacted with 2 mL dimethyl carbonate, which served
both as the reactant and the solvent. The reaction was catalyzed by 440
at 95 C for 90 min. It was conducted in a 5 mL reaction vial with cap and septum in a heat-
µ
L (2.95 mmol) DBU
◦
ing unit under constant stirring. This approach was based on a method by Bernini et al. [21]
and was optimized by a randomized, central composite design. The investigation was
supported by the software Design Expert, which is suitable for response surface methods.
◦
◦
A range of 60 C to 100 C was considered for the temperature, and a range of 30 to 480 min
was used for the reaction time. The volumes of catalyst were chosen between 130
µ
L and
◦
4
40
µL. The central point of the experimental design was thus 80 C, 255 min, and 285
µL
(
1.91 mmol) DBU. This central point was repeated five times.
Sample Preparation
To the samples obtained through synthesis, 1 mL of methanol was added to remove
excessive DMC azeotropes under nitrogen flow. The residue was dissolved in a mixture
of acetonitrile, water, and formic acid (18.5/78.5/3; v/v/v) and adjusted to a volume
of 5 mL. After further dilution (1:2 or 1:10) and microfiltration (regenerated cellulose,
0
.2
µ
m), the samples were analyzed using liquid chromatography coupled with ion trap
n
(
LC-LIT-MS ) and ion mobility quadrupole time of flight (LC-IMS-QTOF-MS) multistage
mass spectrometry.
For further identification purposes, the glucose moiety of the resulting reaction prod-
◦
ucts was removed through hydrolysis with hydrochloric acid for 90 min at 90 C. This proce-
dure allowed the comparison with an anthocyanin-rich extract of Brazilian pepper (Schinus
terebinthifolius), which was treated in the same way.
3.3. LC-MS Analysis
UHPLC analysis of the reaction products was performed on an Acquity UPLC I-Class
system (Waters, Milford, MA, USA). The apparatus consisted of a sample manager cooled
◦
at 10 C, a binary pump, a column oven, and a diode array detector scanning from 250 to
6
◦
50 nm. The column oven temperature was set at 40 C. An Acquity HSS-T3 RP18 column
(
150 mm
×
2.1 mm; 1.8 µm particle size) was combined with a precolumn (Acquity UPLC

Molecules 2021, 26, 1342
12 of 14
HSS T3 VanGuard, 100 Å, 2.1 mm
×
5 mm, 1.8
µ
m), both from Waters (Milford, MA, USA).
The separation was performed with water (A) and acetonitrile (B) as eluents, both acid-
ified with 3% (v + v) formic acid. The flow rate was set at 0.4 mL/min. Analyses of
the methylated cyanidin-3-O-glucoside derivatives were carried out with linear gradient
conditions from 7% B to 25% B for 5 min, then to 50% B for 5 min. For the methylated
cyanidin derivatives, the linear gradient ranged from 7% B to 25% B for 5 min, then to 70%
B for 9 min. The injection volume was 5 µL. For MS analysis, the UHPLC was coupled
with an LTQ-XL ion trap mass spectrometer (Thermo Scientific, Inc., Waltham, MA, USA).
This was equipped with an electrospray interface operating in positive ion mode. Ion mass
spectra were recorded in the range of m/z 245–1000. The source voltage was kept at 4 kV at
a current of 100
µ
A, and the tube lens was adjusted to 55 V. The capillary temperature was
◦
set at 325 C with a spray voltage of 4 V. The sheath, auxiliary, and sweep gas was nitrogen
at a flow of 70, 10, and 1 arb, respectively. Three consecutive scans were conducted: a full
mass scan, an MS/MS scan of the most abundant ion from the first scan using a normalized
collision energy (CE) of 65%, and an MS of the most abundant ion in MS with a CE
of 65%. In addition, multiple reaction monitoring (MRM) measurements were performed
3
2
+
on the fragment of the monomethylated aglycone (m/z at 301, [M] ) with different colli-
sion energies ranging from 10 to 200%. The Xcalibur (2.2SP1.48, Thermo Scientific, Inc.,
Waltham, MA, USA) was used to evaluate the data. For ion mobility spectrometry mea-
surements, the UHPLC was connected to a Vion IMS QTOF mass spectrometer (Waters,
MA, USA). The ionization mode was positive. The capillary voltage was 0.5 kV, the source
◦
temperature was 120 C, the cone voltage was 40 V, the desolvation gas temperature
◦
was 550 C, and the desolvation gas flow was 1200 L/h. Automatic lock correction was
conducted every 5 min with leucine-enkephalin as the lock mass at a concentration of
1
00 pg/µL. The drift gas was nitrogen, and the MS mode was high definition with a low
collision energy of 6 eV and a high collision energy ramp of 20–40 eV. Data were acquired
and processed using UNIFI v1.9.2.045 (Waters, Milford, MA, USA).
4. Conclusions
With this work, a foundation for the green synthesis of methylated anthocyanin metabo-
lites is provided. It was possible to maximize the synthesis of the desired monomethylated
0
derivatives and minimize undesired side products. Cautious identification of 4 -, 5-, and
7
-OH derivatives was possible. Precise identification after isolation would be the next step
to use the synthesized components as reference substances. Nuclear magnetic resonance
NMR) spectroscopy is the most promising approach here, as other spectroscopic methods
(
reach their limits. In addition, the methylation of other anthocyanins should be included to
investigate whether the aglycone or the sugar moiety have an influence on the synthesis.
Author Contributions: Conceptualization, S.S., M.P. and A.S.; methodology, S.S. and T.B.; validation,
S.S.; investigation, S.S. and T.B.; interpretation, S.S., T.B., M.P. and A.S., resources, A.S.; writing—
original draft preparation, S.S.; writing—review and editing, T.B., M.P. and A.S.; supervision, M.P.
and A.S.; project administration, M.P.; All authors have read and agreed to the published version of
the manuscript.
Funding: This research received no external funding.
Data Availability Statement: Data is contained within the article.
Conflicts of Interest: The authors declare no conflict of interest.
Sample Availability: Samples of the compounds are not available from the authors.
References
1
.
Fernandes, I.; Marques, F.; de Freitas, V.; Mateus, N. Antioxidant and antiproliferative properties of methylated metabolites
of anthocyanins. Food Chem. 2013, 141, 2923–2933. [CrossRef]
2
3
.
.
Mazza, G.J. Anthocyanins and heart health. Ann. Ist. Super. Sanita 2007, 43, 369–374.
Wu, X.; Pittman, H.E.; McKay, S.; Prior, R.L. Aglycones and sugar moieties alter anthocyanin absorption and metabolism after
berry consumption in weanling pigs. J. Nutr. 2005, 135, 2417–2424. [CrossRef] [PubMed]

Molecules 2021, 26, 1342
13 of 14
4.
5.
6.
Kamiloglu, S.; Capanoglu, E.; Grootaert, C.; van Camp, J. Anthocyanin Absorption and Metabolism by Human Intestinal Caco-2
Cells—A Review. Int. J. Mol. Sci. 2015, 16, 21555–21574. [CrossRef] [PubMed]
Bouktaib, M.; Lebrun, S.; Atmani, A.; Rolando, C. Hemisynthesis of all the O-monomethylated analogues of quercetin including
the major metabolites, through selective protection of phenolic functions. Tetrahedron 2002, 58, 10001–10009. [CrossRef]
Blount, J.W.; Redan, B.W.; Ferruzzi, M.G.; Reuhs, B.L.; Cooper, B.R.; Harwood, J.S.; Shulaev, V.; Pasinetti, G.; Dixon, R.A. Synthesis
and quantitative analysis of plasma-targeted metabolites of catechin and epicatechin. J. Agric. Food Chem. 2015, 63, 2233–2240.
[CrossRef] [PubMed]
7
8
9
1
1
.
.
.
Wu, X.; Pittman, H.E.; Prior, R.L. Pelargonidin is absorbed and metabolized differently than cyanidin after marionberry
consumption in pigs. J. Nutr. 2004, 134, 2603–2610. [CrossRef]
Cruz, L.; Basílio, N.; Mateus, N.; Pina, F.; de Freitas, V. Characterization of kinetic and thermodynamic parameters of cyanidin-3-
glucoside methyl and glucuronyl metabolite conjugates. J. Phys. Chem. B 2015, 119, 2010–2018. [CrossRef]
Schmitt, S.; Tratzka, S.; Schieber, A.; Passon, M. Hemisynthesis of Anthocyanin Phase II Metabolites by Porcine Liver Enzymes.
J. Agric. Food Chem. 2019, 67, 6177–6189. [CrossRef]
0. Cruz, L.; Mateus, N.; de Freitas, V. First chemical synthesis report of an anthocyanin metabolite with in vivo occurrence:
0
Cyanidin-4 -O-methyl-3-glucoside. Tetrahedron Lett. 2013, 54, 2865–2869. [CrossRef]
1. Cren-Olivé, C.; Lebrun, S.; Rolando, C. An efficient synthesis of the four mono methylated isomers of (+)-catechin including
the major metabolites and of some dimethylated and trimethylated analogues through selective protection of the catechol ring.
J. Chem. Soc. Perkin Trans. 2002, 6, 821–830. [CrossRef]
˙
1
2. Kozłowska, J.; Potaniec, B.; Zarowska, B.; Anioł, M. Synthesis and Biological Activity of Novel O-Alkyl Derivatives of Naringenin
and Their Oximes. Molecules 2017, 22, 1485. [CrossRef]
1
1
3. Tundo, P.; Musolino, M.; Aricò, F. The reactions of dimethyl carbonate and its derivatives. Green Chem. 2018, 20, 28–85. [CrossRef]
4. Tundo, P.; Memoli, S.; Hérault, D.; Hill, K. Synthesis of methylethers by reaction of alcohols with dimethylcarbonate. Green Chem.
2
004, 6, 609–612. [CrossRef]
1
1
5. Lee, Y.; Shimizu, I. Convenient O-Methylation of Phenols with Dimethyl Carbonate. Synlett 1998, 1998, 1063–1064. [CrossRef]
6. Ouk, S.; Thiebaud, S.; Borredon, E.; Legars, P.; Lecomte, L. O-Methylation of phenolic compounds with dimethyl carbonate under
solid/liquid phase transfer system. Tetrahedron Lett. 2002, 43, 2661–2663. [CrossRef]
1
1
1
2
2
2
7. Ouk, S.; Thiébaud, S.; Borredon, E.; Le Gars, P. High performance method for O-methylation of phenol with dimethyl carbonate.
Appl. Catal. A 2003, 241, 227–233. [CrossRef]
8. Rein, M. Copigmentation Reactions and Color Stability of Berry Anthocyanins; University of Helsinki: Helsinki, Finland, 2005;
ISBN 952-10-2293-0.
9. Shieh, W.C.; Dell, S.; Repic, O. 1,8-Diazabicyclo5.4.0undec-7-ene (DBU) and microwave-accelerated green chemistry in methylation
of phenols, indoles, and benzimidazoles with dimethyl carbonate. Org. Lett. 2001, 3, 4279–4281. [CrossRef] [PubMed]
0. Tatsuzaki, J.; Ohwada, T.; Otani, Y.; Inagi, R.; Ishikawa, T. A simple and effective preparation of quercetin pentamethyl ether
from quercetin. Beilstein J. Org. Chem. 2018, 14, 3112–3121. [CrossRef]
1. Bernini, R.; Crisante, F.; Ginnasi, M.C. A convenient and safe O-methylation of flavonoids with dimethyl carbonate (DMC).
Molecules 2011, 16, 1418–1425. [CrossRef] [PubMed]
2. Ouk, S.; Thiébaud, S.; Borredon, E.; Le Gars, P. Dimethyl carbonate and phenols to alkyl aryl ethers via clean synthesis. Green Chem.
2
002, 4, 431–435. [CrossRef]
2
3. Mason, T.J. Ultrasound in synthetic organic chemistry. Chem. Soc. Rev. 1997, 26, 443. [CrossRef]
2
4. Riesz, P.; Berdahl, D.; Christman, C.L. Free radical generation by ultrasound in aqueous and nonaqueous solutions.
Environ. Health Perspect. 1985, 64, 233–252. [CrossRef] [PubMed]
2
2
5. González-Manzano, S.; González-Paramás, A.; Santos-Buelga, C.; Dueñas, M. Preparation and characterization of catechin
sulfates, glucuronides, and methylethers with metabolic interest. J. Agric. Food Chem. 2009, 57, 1231–1238. [CrossRef] [PubMed]
6. Cren-Olivé, C.; Déprez, S.; Lebrun, S.; Coddeville, B.; Rolando, C. Characterization of methylation site of monomethylflavan-3-ols
by liquid chromatography/electrospray ionization tandem mass spectrometry. Rapid Commun. Mass Spectrom. 2000, 14, 2312–2319.
[CrossRef]
0 0
7. Donovan, J.L.; Luthria, D.L.; Stremple, P.; Waterhouse, A.L. Analysis of (+)-catechin, ( )-epicatechin and their 3 - and 4 -O-
2
2
2
3
3
3
−
methylated analogs. J. Chromatogr. B Biomed. Sci. Appl. 1999, 726, 277–283. [CrossRef]
8. Feuereisen, M.M.; Hoppe, J.; Zimmermann, B.F.; Weber, F.; Schulze-Kaysers, N.; Schieber, A. Characterization of phenolic
compounds in Brazilian pepper (Schinus terebinthifolius Raddi) exocarp. J. Agric. Food Chem. 2014, 62, 6219–6226. [CrossRef]
9. Day, A.J.; Bao, Y.; Morgan, M.R.; Williamson, G. Conjugation position of quercetin glucuronides and effect on biological activity.
Free Radic. Biol. Med. 2000, 29, 1234–1243. [CrossRef]
0. Toki, K.; Saito, N.; Irie, Y.; Tatsuzawa, F.; Shigihara, A.; Honda, T. 7-O-Methylated anthocyanidin glycosides from Catharan-
thus roseus. Phytochemistry 2008, 69, 1215–1219. [CrossRef]
1. Hvattum, E.; Ekeberg, D. Study of the collision-induced radical cleavage of flavonoid glycosides using negative electrospray
ionization tandem quadrupole mass spectrometry. J. Mass Spectrom. 2003, 38, 43–49. [CrossRef] [PubMed]
2. Barnes, J.S.; Schug, K.A. Structural characterization of cyanidin-3,5-diglucoside and pelargonidin-3,5-diglucoside anthocyanins:
Multi-dimensional fragmentation pathways using high performance liquid chromatography-electrospray ionization-ion trap-time
of flight mass spectrometry. Int. J. Mass Spectrom. 2011, 308, 71–80. [CrossRef]

Molecules 2021, 26, 1342
14 of 14
3
3
3
3
3. Dangles, O.; Fenger, J.A. The Chemical Reactivity of Anthocyanins and Its Consequences in Food Science and Nutrition. Molecules
018, 23, 1970. [CrossRef] [PubMed]
4. Rao, K.V.; Owoyale, J.A. Partial methylation of quercetin: Direct synthesis of tamarixetin, ombuin and ayanin. J. Heterocycl. Chem.
976, 13, 1293–1295. [CrossRef]
2
1
5. Fleschhut, J.; Kratzer, F.; Rechkemmer, G.; Kulling, S.E. Stability and biotransformation of various dietary anthocyanins in vitro
.
Eur. J. Nutr. 2006, 45, 7–18. [CrossRef] [PubMed]
6. Tomaz, I.; Šikuten, I.; Preiner, D.; Andabaka, Ž.; Huzani c´ , N.; Leskovi c´ , M.; Karoglan Konti c´ , J.; Ašperger, D. Stability of
polyphenolic extracts from red grape skins after thermal treatments. Chem. Pap. 2019, 73, 195–203. [CrossRef]