Anthocyanins from wild carrot suspension cultures acylated with supplied carboxylic acids

Article abstract of DOI:10.1016/S0008-6215(98)00124-4

The anthocyanins accumulated by carrot cell cultures include some that are acylated. Addition of cinnamic and benzoic acid analogues to the culture medium resulted in the production of fourteen novel monoacylated anthocyanins. Four additional anthocyanins were prepared by partial acid hydrolysis, which selectively removed the xylose side chain from the branched trisaccharide part of the molecule. The carrot cells accommodate a wide range of carboxylic acids used to acylate anthocyanins and provide a system for the preparation of anthocyanins that have a variety of acyl groups attached. The compounds were characterized by ¹H NMR spectroscopy and mass spectrometry, and by UV-vis absorption spectroscopy at both pH 2 and 6. Copyright (C) 1998 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(98)00124-4

Carbohydrate Research 310 (1998) 177±189
Anthocyanins from wild carrot suspension
cultures acylated with supplied carboxylic acids1
Donald K. Dougall a,*, David C. Baker b, Elena G. Gakh b,2
Marc A. Redus a, Neil A. Whittemore b
,
a Department of Botany and the Biotechnology Program, Life Sciences Graduate Degree Program,
The University of Tennessee, Knoxville, TN 37996, USA
b Department of Chemistry, The University of Tennessee, Knoxville, TN 37996, USA
Received 4 March 1998; accepted 22 April 1998
Abstract
The anthocyanins accumulated by carrot cell cultures include some that are acylated. Addition
of cinnamic and benzoic acid analogues to the culture medium resulted in the production of
fourteen novel monoacylated anthocyanins. Four additional anthocyanins were prepared by par-
tial acid hydrolysis, which selectively removed the xylose side chain from the branched tri-
saccharide part of the molecule. The carrot cells accommodate a wide range of carboxylic acids
used to acylate anthocyanins and provide a system for the preparation of anthocyanins that have a
variety of acyl groups attached. The compounds were characterized by 1H NMR spectroscopy and
mass spectrometry, and by UV-vis absorption spectroscopy at both pH 2 and 6. # 1998 Elsevier
Science Ltd. All rights reserved
Keywords: Anthocyanin; Tissue culture; Acylation; Nutraceuticals; Natural colorants
1. Introduction
(¯avylium) salts that have the unfortunate draw-
back in that they lose most of their color at near-
neutral pH. This phenomenon is understood to be
due to hydration of the ¯avylium nucleus at C-2, a
process that interrupts the conjugation in the
chromophore [4±6]. A small number of anthocya-
nins, characteristically those acylated on the sugar
moiety with cinnamic acids, show greatly increased
color retention in near-neutral solutions; that is,
they exhibit a decreased rate of hydration or an
increased rate of dehydration (reversal). The resul-
tant color retention of anthocyanins is highly
desirable and enhances their potential utility in the
food and related industries.
Anthocyanins, which are highly colored sub-
stances found in plants, are possible viable
colorants for use in food, nutraceutical, and phar-
maceutical preparations [2,3]. Chemically antho-
cyanins are glycosides of benzopyrylium
* Corresponding author: Fax: (423) 974-2258; e-mail:
ddougall@utk.edu
1
Part 2 in the series, Studies on the Stability and Con-
formation of Monoacylated Anthocyanins. For Part 1, see ref
[1].
2
Current address: Wyeth-Ayerst Research, Pearl River,
NY 10965, USA.
0008-6215/98/$Ðsee front matter # 1998 Elsevier Science Ltd. All rights reserved
PII S0008-6215(98)00124-4

178
D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
Acylation of the carbohydrate moiety is a pri-
mary method for increasing anthocyanin color
retention. To improve our understanding of the
ways in which these acyl groups alter the color
retention of anthocyanins, it is desirable (1) to
measure the e€ects of these acyl groups and sub-
stituents on the rate constants of the hydration±
dehydration reaction, and (2) determine the physi-
cochemical interrelationships (i.e., the distance,
orientation, and any possible stereoelectronic
e€ects) between the acyl group and the ¯avylium
chromophore. These objectives require the avail-
ability of a series of anthocyanins that di€er sys-
tematically in the acyl group present. As such a
series is not available by isolation from plants in
nature, a semibiosynthetic method using plant cell
cultures was pursued.
When naturally occurring cinnamic acids are fed
to wild carrot suspension cultures, the proportion
of acylated to nonacylated anthocyanins increases.
Work from these laboratories has shown that the
cultures incorporate 3,4-dimethoxycinnamic acid
and 3,4,5-trimethoxycinnamic acid into new
anthocyanins when these acids are provided in the
culture medium [7]. In these prior studies, some
evidence for the presence of anthocyanins acylated
with 4-hydroxybenzoic acid was also obtained, a
®nding that is in line with a previous report by
Glaûgen et al. [8].
We now report our ®ndings concerning the abil-
ity of carrot cell suspension cultures to biosynthe-
size acylated anthocyanins by providing variously
substituted cinnamic and benzoic acids to the cul-
tures. We show that the anthocyanins produced
di€er in their ability to retain color at pH 6 and
provide further evidence that there are interactions
between the acyl groups and the chromophore. A
preliminary account of these ®ndings has been
reported [9].
2. Results and discussion
Isolation, puri®cation, and characterization of the
anthocyanins.ÐAnthocyanins were acylated with
variously substituted benzoic and cinnamic acids to
give compounds 5±24 by addition of the requisite
carboxylic acids to a tissue culture of Daucus car-
ota spp carota (wild carrot) by previously reported
methods [7]. These were then isolated, puri®ed, and
characterized (see Experimental section). The non-
acylated anthocyanins ꢀ-d-glucopyranosyl-(1!6)-
[ꢀ-d-xylopyranosyl-(1!2)]-ꢀ-d-galactopyranosyl-
(1!O3)-cyanidin (1), ꢀ-d-glucopyranosyl-(1!6)-ꢀ-
d-galactopyranosyl-(1!O3)-cyanidin (2), ꢀ-d-xylo-
pyranosyl-(1!2)-ꢀ-d-galactopyranosyl-(1!O3)-

D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
179
cyanidin (3), and ꢀ-d-galactopyranosyl-(1!O3)-
cyanidin (4) were prepared to provide reference
compounds to determine the e€ects of various acyl
groups on the properties of the anthocyanins.
Compounds 1 and 3 have been partially character-
ized [1,7,8], and 3 has been reported by Terahara et
al. [10]. Compounds 5, 7, 9, and 11±24 are all 6-O-
acyl-ꢀ-d-glucopyranosyl-(1!6)-[ꢀ-d-xylopyranosyl-
(1!2)]-ꢀ-d-galactopyranosyl-(1!O3)-cyanidins. Of
these, 5 and 9 have been characterized previously
[7,8]. Compounds 6, 8, and 10 are 6-O-acyl-ꢀ-d-
glucopyranosyl-(1!6)-ꢀ-d-galactopyranosyl-(1!O3)-
cyanidins.
independently showed this connectivity for 5 and
the carrot anthocyanin acylated with 4-hydroxy-3-
methoxycinnamic acid (ferulic acid) by 1H-detected
short- (HMQC) and long-range (HMBC) correla-
tion experiments.
Comparison of the NMR data of a mono-
acylated compound with its nonacylated counter-
part, e.g., 5 and 1, respectively, indicates that
acylation has occurred on C-6III of the glucose
moiety as the chemical shift of the H-6IIIA of 5 is
observed at ꢁ 5.339 and that for 1 at ꢁ 3.846. There
is also a much smaller down®eld shift of H-6IIIB
observed with the acylated compounds. For 24,
which is acylated with 3-(4-hydroxyphenyl)-
propionic acid, a similar down®eld shift, albeit one
Characterization.Ð(a) NMR studies. The 1H
NMR data for anthocyanins 1±24 are given in
1
Table 1. By making use of the H NMR chemical
smaller in size, is observed for H-6IIIA
.
shifts and coupling constants for oligosaccharides
in general [11], together with data that have been
reported from 2D NMR spectroscopy of antho-
cyanins [8,12], speci®c protons in this series of
anthocyanins were assigned. For example, the
anomeric proton of each unit of the glycone was
observed in regions typical for that particular sugar
moiety. Ranges were as follows: d-galactose sub-
unit ꢁ 5.18±5.40; d-xylose subunit, ꢁ 4.64±4.86; and
d-glucose subunit ꢁ 4.23±4.64. All these protons
exhibited H-1±H-2 coupling constants between 7.1
and 7.8 Hz, showing that all the sugar linkages are
in the ꢀ-d-con®guration.
The presence of a cinnamic or benzoic acid resi-
due on the glucose moiety has a marked e€ect on
the chemical shift of the cyanidin H-4 proton in the
acylated compounds. The chemical shifts for H-4
in nonacylated 1±4 are in the range of ꢁ 9.029 to
8.981, while those for the acylated 5±24 appear at ꢁ
8.474±8.953. In the anthocyanins acylated with
cinnamic acids, but not in those acylated with
benzoic acids or with 3-(4-hydroxyphenyl)propionic
acid, there is a change in the chemical shift of
H-8 from the region of ꢁ 6.887±6.964 in the non-
acylated 1±4 to ꢁ 6.448±6.606 in the cinnamoylated
compounds 5±19. Glaûgen et al. [8] have noted a
pronounced up®eld shift of H-4 in carrot antho-
cyanins acylated with 4-hydroxy-3,5-dimethoxy-
cinnamic (sinapic), ferulic, 4-hydroxycinnamic
(4-coumaric), and 4-hydroxybenzoic acids relative
to that for the nonacylated compound. These
authors have interpreted this as evidence for intra-
molecular copigmentation (stacking). Although
not noted by them, the data of Glaûgen et al. [8]
also show the up®eld shift of H-8 in anthocyanins
When xylose is absent, H-2I is found up®eld at
less than ꢁ 4.0 (see data for 2, 4, 6, 8, and 10), and
when xylose is present, H-2I is found down®eld of ꢁ
4.169, con®rming the attachment of xylose at the 2-
position of the galactose residue. The connectivity
between the galactose and the glucose moieties was
unequivocally established by the use of NOESY
spectroscopy on 1, 7, 15, and 16. A partial NOESY
spectrum for 7 is shown in Fig. 1. Glaûgen et al. [8]
Fig. 1. Portion of the NOESY spectrum (500 ms mixing time) of 7 at 20 ꢀC. The cross-peaks are labeled, (a) H-1III±H-5I, (b) H-1III
H-6IA, (c) H-5I±H-6IA, and (d) H-1III±H-6IB
±
.

180
D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
Table 1
1H NMR data of anthocyanins 1±24 a
H
4
1 b,c
2
3 d
4
5 b
6 b
7 b
8 b
9
10
11
12
1
Flav
8.981
br s
9.029
br s
8.976
br s
9.027
br s
8.474
d
8.605
d
8.478
d
8.633
br s
8.524
br s
8.702
br s
8.524
br s
8.526
br s
8
b
(0.9)
6.684
d
(2.0)
6.512
dd
(0.9)
6.652
d
(2.0)
6.536
d
(0.9)
6.695
d
(2.0)
6.525
d
6
8
6.715
d
(2.0)
6.964
br d
6.655
br s
6.643
d
(1.8)
6.893
br d
6.694
d
(1.9)
6.887
d
6.634
br s
6.618
br s
6.612
d
(2.0)
6.593
d
6.609
d
(2.0)
6.532
d
6.600
br s
6
d
(1
6
d
6.899
br d
6.548
br d
6.559
br d
6.483
br d
(0.8)
(2.0
0.9)
7.913
d
(2.4)
7.022
d
(8.8)
8.136
dd
(8.8;
2.4)
5.256
d
(7.3)
4.276
dd
(9.2;
7.3)
4.182
dd
(9.2;
3.3)
3.944
d
(3.3)
4.483
dd
(0.9)
(0.9)
(2.0)
(2.0)
(1
20
50
60
8.090
d
(2.4)
7.041
d
(8.8)
8.292
dd
8.103
d
(2.4)
7.018
d
(8.7)
8.271
dd
8.065
d
(2.4)
7.016
d
(8.8)
8.290
dd
8.074
d
(2.3)
7.015
d
(8.8)
8.259
dd
(8.7;
2.4)
5.241
d
(7.7)
3.982
dd
(9.6;
7.7)
3.650
dd
(9.6;
3.4)
3.942
d
(3.4)
ꢁ3.7
7.916
d
(2.3)
7.033
d
(8.7)
8.266
dd
(8.7;
2.3)
5.191
d
(7.3)
3.994
dd
(8.9;
7.3)
3.929
dd
(8.9;
3.3)
3.914
d
(3.3)
4.427
dd
7.915
d
(2.3)
7.018
d
(8.8)
8.186
dd
(8.8;
2.3)
5.286
d
(7.5)
4.296
dd
(9.3;
7.5)
4.184
dd
(9.3;
3.4)
3.951
d
(3.4)
4.500
dd
7.932
d
(2.4)
7.028
d
(8.8)
8.270
dd
(8.8;
2.4)
5.218
d
(7.5)
3.997
dd
(8.9;
7.5)
3.925
dd
(8.9;
3.3)
3.909
d
(3.3)
4.446
dd
7.934
d
(2.3)
7.051
d
(8.7)
8.066
dd
(8.7;
2.3)
5.268
d
(7.4)
4.218
dd
(9.4;
7.4)
4.125
dd
(9.4;
3.3)
3.932
d
(3.3)
4.427
dd
7.945
d
(2,3)
7.030
d
(8.8)
8.241
dd
(8.8;
2.3)
5.216
d
(7.6)
3.966
dd
(8.6;
7.6)
ꢁ3.9
7.901
d
(2.4)
7.046
d
(8.7)
8.110
dd
(8.7;
2.4)
5.260
d
(7.4)
4.223
dd
(9.3;
7.4)
4.127
dd
(9.3;
3.4)
3.930
d
(3.4)
4.420
dd
7.913
d
(2.2)
7.042
d
(8.7)
8.095
dd
(8.7;
2.2)
5.294
d
7
d
(2
7
d
(8
8
d
(8
2
(8.8;
2.4)
5.398
d
(7.6)
4.196
dd
(8.7;
2.4)
5.217
d
(7.7)
3.956
dd
(8.7;
2.4)
5.402
d
(7.6)
4.213
dd
Gal
1I
2I
5
d
(7
4
d
(9
7
(7.3)
ꢁ4.2
(7.6;
9.4)
3.919
dd
(7.7;
9.7)
3.649
dd
(7.7;
9.3)
3.882
dd
3I
ꢁ4.1
4
d
(9
3
(9.4;
3.3)
4.002
d
(3.3)
4.064
dd
(9.7;
3.4)
3.966
d
(3.4)
ꢁ3.9
(9.3;
3.3)
3.953
d
(3.3)
ꢁ3.8
4I
5I
ꢁ3.9
ꢁ4.4
3.934
d
(3.3)
ꢁ4.4
3
d
(3
4
d
(9
2
4
d
(
9
3
d
(
(7.3;
4.9)
3.997
dd
(4.9;
11.2)
3.879
dd
(7.3;
(9 6;
2.0)
4.215
dd
(9.4;
1.5)
4.224
dd
(9.2;
1.9)
4.222
dd
(9.6;
2.0)
4.229
dd
(9.4;
2.0)
4.195
dd
(9.4;
2.0)
4.192
dd
6IA
ꢁ3.9
ꢁ3.9
ꢁ3.8
ꢁ3.8
ꢁ3.7
ꢁ3.7
4.184
dd
ꢁ4.1
ꢁ3.7
(
13.1;
(
12.8;
(
13.3;
(
13.0;
(
13.0;
(
12.8;
(
12.9;
9.6)
3.747
dd
9.4)
3.734
dd
9.2)
3.758
dd
9.6)
3.732
dd
9.4)
3.742
dd
9.5)
3.743
dd
9.4)
3.744
dd
6IB
(
13.1;
(
12.8;
(
13.2;
(
13.0;
(
13.0;
(
13.2;
(
12.9;
11.2)
4.695
d
(7.6)
3.158
dd
(9.0;
7.6)
3.296
dd
(9.0;
8.8)
3.392
ddd
(10.2;
8.8;
5.3)
3.636
dd
2.0)
4.829
d
(7.8)
3.193
dd
(9.0;
7.8)
3.333
dd
(9.0;
9.0)
3.251
ddd
(10.9;
9.0;
5.2)
3.610
dd
1.5)
Ð
1.9)
4.856
d
(7.8)
3.148
dd
(8.9;
7.8)
3.342
dd
(8.9;
8.9)
3.266
ddd
(10.5;
8.9;
4.6)
3.627
dd
2.0)
Ð
2.0)
4.660
d
(7.3)
3.240
dd
(9.1;
7.4)
3.298
dd
(9.7;
9.1)
ꢁ3.40
2.4)
Ð
2.0)
4.694
d
(7.4)
3.244
dd
(9.1;
7.4)
3.303
dd
(9.1;
9.6)
ꢁ3.41
1
4
d
(7
e
Xyl
1II
Ð
Ð
4.684
d
(7.7)
3.135
dd
Ð
Ð
4.667
d
(7.4)
2II
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
e
(8.9;
7.7)
e
e
e
e
3II
Ð
Ð
Ð
Ð
4II
3
e
d
(9
9
5IIA
Ð
Ð
3.606
dd
Ð
Ð
Ð
Ð
Ð
Ð
3.484
dd
Ð
3.521
dd
ꢁ3.5
3
d
(
5
2
d
(
(
11.6;
(
11.6;
(
11.4;
(
11.6;
(
11.2;
(
11.2;
5.3)
3.019
dd
( 11.6;
10.2)
5.3)
3.007
dd
( 11.3;
10.3)
5.2)
3.121
dd
( 11.4;
10.9)
4.6)
3.144
dd
( 11.6;
10.5)
5.3)
2.810
dd
( 11.2;
10.5)
5.3)
2.850
dd
( 11.2;
10.8)
5IIB
Ð
dd
2.812
dd
( 10.4;
10.4)
1

D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
181
11
12
13
14
15
16
17
18
19
20
21
22
23
24
8.524
br s
8.526
br s
8.603
br s
8.583
br s
8.588
br s
8.515
br s
8.538
d
8.554
br s
8.475
d
8.447
br s
8.518
br s
8.953
br s
8.500
br s
8.821
br s
(0.8)
6.603
d
(1.1)
6.622
d
6.609
d
6.600
br s
6.552
d
6.604
br s
6.587
br s
6.613
br s
6.673
d
6.483
d
6.632
d
6.551
d
6.538
d
6.632
d
(2.0)
6.532
d
(1.9)
6.448
d
(2.0)
6.514
d
(2.0)
6.568
br d
(2.0)
6.521
dd
(2.0)
6.741
d
(2.1)
6.879
br d
(1.7)
6.855
br d
(2.0)
6.818
br d
(2.3)
6.842
br d
6.483
br d
6.497
br d
not
seen
6.606
br d
(2.0)
(1.4)
(0.8)
(2.0;
1.1)
7.856
d
(1.2)
7.901
d
7.913
d
7.920
d
7.915
d
7.952
d
7.961
d
7.951
d
7.958
d
7.973
d
8.022
d
8.023
d
8.018
d
8.066
d
(2.4)
7.046
d
(2.2)
7.042
d
(2.4)
7.037
d
(2.4)
7.046
d
(2.3)
7.056
d
(2.3)
7.062
d
(2.4)
7.044
d
(2.4)
7.080
d
(2.3)
7.026
d
(2.3)
7.013
d
(2.3)
7.056
d
(2.4)
7.038
d
(2.4)
7.030
d
(2.3)
7.024
d
(8.7)
8.110
dd
(8.7)
8.095
dd
(8.8)
8.171
dd
(8.7)
8.093
dd
(8.8)
8.114
dd
(8.7)
8.100
dd
(8.8)
8.080
dd
(8.7)
8.116
dd
(8.7)
8.146
dd
(8.7)
8.219
dd
(8.7)
8.230
dd
(8.7)
8.222
dd
(8.7)
8.208
dd
(8.7)
8.257
dd
(8.7;
2.4)
5.260
d
(8.7;
2.2)
5.294
d
(8.6;
2.4)
5.392
d
(8.7;
2.4)
5.304
d
(8.8;
2.3)
5.277
d
(8.7;
2.3)
5.208
d
(8.8;
2.4)
5.254
d
(8.7;
2.4)
5.237
d
(8.7;
2.3)
5.257
d
(8.7;
2.3)
5.388
d
(8.7;
2.3)
5.232
d
(8.7;
2.4)
5.178
d
(8.7;
2.4)
5.190
d
(8.7;
2.3)
5.289
d
(7.4)
4.223
dd
(7.3)
ꢁ4.2
(7.5)
4.240
dd
(7.4)
4.218
dd
(7.4)
ꢁ4.2
(7.4)
4.221
dd
(7.4)
4.241
dd
(7.5)
4.244
dd
(7.3)
4.232
dd
(7.5)
ꢁ4.2
(7.4)
4.169
dd
(7.5)
4.182
dd
(7.4)
4.186
dd
(7.5)
4.203
dd
(9.3;
7.4)
4.127
dd
(9.4;
7.3)
4.128
dd
(9.4;
7.4)
4.100
dd
(9.4;
7.4)
4.102
dd
(9.4;
7.4)
4.108
dd
(9.4;
7.5)
4.091
dd
(9.2;
7.3)
4.152
dd
(9.2;
7.3)
3.983
dd
(9.3;
7.5)
3.964
dd
(9.3;
7.4)
3.974
dd
(9 2;
7.5)
3.912
dd
ꢁ4.1
4.091
dd
4.043
dd
(9.3;
3.4)
3.930
d
(9.2;
3.4)
3.940
d
(9.4;
3.4)
3.935
d
(9.3;
3.3)
3.935
d
(9.4;
3.3)
3.932
d
(9.4;
3.4)
3.943
d
(9.4;
3.4)
3.972
d
(9.2;
3.3)
3.920
d
(9.4;
3.3)
3.917
d
(9.2;
3.4)
3.902
d
(9.4;
3.4)
3.896
d
(9.3;
3.4)
3.902
d
(9.2;
3.3)
3.942
d
3.934
d
(3.4)
4.420
dd
(3.3)
ꢁ4.4
(3.3)
4.490
dd
(3.4)
ꢁ4.4
(3.3)
ꢁ4.4
(3.3)
4.368
dd
(3.3)
4.381
dd
(3.4)
4.361
dd
(3.3)
4.410
dd
(3.3)
ꢁ4.2
(3.4)
4.334
dd
(3.5)
4.282
dd
(3.4)
4.271
dd
(3.3)
4.102
dd
(9.4;
2.0)
4.192
dd
(9.4;
2.2)
4.20
dd
(8.7;
1.1)
ꢁ4.1
(9.5;
2.2)
4.178
dd
(9.5;
2.5)
4.154
dd
(9.7;
1.9)
4.190
dd
(9.3;
2.4)
4.108
dd
(8.6;
2.4)
4.111
dd
(8.8;
2.6)
4.093
dd
(7.9;
3.7)
3.986
dd
ꢁ4.1
ꢁ3.7
ꢁ4.1
ꢁ4.1
4.201
dd
(
12.9;
(
13.1;
(
12.9;
(
12.8;
(
13.0;
(
13.1;
(
12.7;
(
12.6;
(
12.5;
(
11.9;
9.4)
3.744
dd
9.3)
3.767
dd
9.5)
3.769
dd
9.1)
3.802
dd
9.7)
3.729
dd
9.2)
3.783
dd
9.3)
3.799
dd
8.8)
3.798
dd
8.8)
3.812
dd
7.9)
3.857
dd
3.763
dd
3.766
dd
3.752
dd
(
12.9;
(
12.0;
(
12.9;
(
13.1;
(
12.9;
(
12.9;
(
12.8;
(
13.0;
(
12.9;
(
12.7;
(
12.5;
(
12.5;
(
11.9;
2.0)
4.694
d
(7.4)
3.244
dd
1.5)
4.731
d
(7.4)
e
2.3)
4.655
d
(7.4)
3.229
dd
3.1)
4.659
d
(7.4)
e
2.0)
4.656
d
(7.4)
3.221
dd
2.2)
4.686
d
(7.4)
3.235
dd
2.3)
4.702
d
(7.5)
3.219
dd
1.9)
4.755
d
(7.3)
3.267
dd
2.0)
4.765
d
(7.8)
ꢁ3.3
2.4)
4.678
d
(7.5)
3.198
dd
2.6)
4.657
d
(7.5)
3188
dd
2.6)
4.645
d
(7.4)
3.200
dd
3.7)
4.666
d
(7.5)
3183
dd
4.667
d
(7.4)
e
(9.1;
7.4)
3.303
dd
(9.1;
9.6)
ꢁ3.41
(9.1;
7.4)
(9.1;
7.4)
(9.1;
7.4)
3.300
dd
(9.1;
9.1)
ꢁ3.38
(9.2;
7.5)
3.317
dd
(9.2;
8.9)
ꢁ3.39
(9.2;
7.3)
3.323
dd
(9.2;
9.2)
ꢁ3.40
(9.2;
7.5)
ꢁ3.4
(9.2;
7.5)
ꢁ3.4
(9.1;
7.4)
3.308
dd
(9.1;
9.2)
ꢁ3.40
(9.0;
7.5)
3.282
dd
(9.2;
9.0)
ꢁ3.30
e
e
e
e
e
e
ꢁ3.3
ꢁ3.4
3.672
dd
(9.1
9.1)
3.640
dd
(9.1;
9.0)
3.620
dd
(9.2;
9.2)
3.631
dd
(9.4;
9.0)
ꢁ3.6
ꢁ3.55
ꢁ3.55
3.521
dd
ꢁ3.5
3.622
dd
ꢁ3.5
ꢁ3.4
ꢁ3.4
3.506
dd
3.538
dd
3.589
dd
ꢁ3.6
3.594
dd
3.540
dd
3.582
dd
(
11.2;
(
11.6;
(
11.4;
(
11.4;
(
11.4;
(
11.6;
(
11.4;
(
11.4;
5.3)
2.850
dd
5.2)
2.932
dd
4.9)
2.856
dd
5.2)
2.860
dd
4.3)
2.922
dd
5.4)
2.837
dd
5.3)
2.833
dd
5.2)
2.967
dd
2.812
dd
2.811
dd
2.825
dd
2.824
dd
3.049
dd
2.841
dd
( 11.2;
10.8)
( 10.4;
10.4)
( 10.9;
10.9)
( 11.3;
10.9)
( 11.1;
10.4)
( 10.8;
10.4)
( 11.4;
10.2)
( 11.4;
10.4)
( 11.4;
10.5)
( 11.4;
10.6)
( 11.6;
10.5)
( 11.3;
10.6)
( 11.4;
10.6)
( 11.4;
10.4)

182
D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
Table 1 contd
H
1 b,c
2
3 d
4
5 b
6 b
7 b
8 b
9
10
11
12
13
Glc
1III
4.267
d
(7.7)
3.178
dd
(9.3;
7.7)
ꢁ3.25
4.234
d
(7.5)
Ð
Ð
4.474
d
(7.4)
ꢁ3.44
4.459
d
(7.5)
ꢁ3.44
4.478
d
(7.5)
ꢁ3.44
4.453
d
(7.3)
ꢁ3.45
4.442
d
(7.2)
ꢁ3.41
4.409
d
(7.3)
ꢁ3.41
4.440
d
(7.1)
ꢁ3.40
4.359
d
(7.1)
4
d
(7
e
2III
3III
4III
5III
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
e
ꢁ3.16
ꢁ3.20
e
e
ꢁ3.44
ꢁ3.44
ꢁ3.44
ꢁ3.45
ꢁ3.43
ꢁ3.43
ꢁ3.40
ꢁ3.30
ꢁ3.30
ꢁ3.20
ꢁ3.18
3.746
dd
(9 5;
9.6)
3.697
dd
(9.5;
9.5)
3.747
dd
(9 6;
9.4)
3.669
dd
(9.4;
9.4)
3.673
dd
(9.4;
9.2)
3.618
dd
(9.4;
9.3)
3.661
dd
(9.4;
9.1)
ꢁ3.7
ꢁ
e
e
ꢁ3.44
ꢁ3.44
ꢁ3.44
ꢁ3.45
ꢁ3.42
ꢁ3.42
ꢁ3.40
6IIIA
6IIIB
2A
3.846
dd
3.835
dd
Ð
Ð
Ð
Ð
Ð
Ð
5.339
dd
5.295
dd
5.340
dd
5.302
dd
5.253
dd
5.169
dd
5.252
dd
5.278
dd
5
d
(
3
ꢁ
(
11.7;
(
11.7;
(
12.1;
(
11.9;
(
12.2;
(
12.0;
(
12.0;
(
11.9;
(
12.0;
(
11.9;
1.8)
3.633
dd
2.2)
3.503
dd
2.8)
4.115
dd
2.8)
4.085
dd
3.0)
4.134
dd
2.8)
4.106
dd
2.7)
4.113
dd
2.6)
4.078
dd
2.7)
4.115
dd
2.6)
ꢁ4.2
(
11.7;
(
11.7;
(
12.2;
(
11.9;
(
12.2;
(
12.0;
(
12.0;
(
12.0;
(
12.0;
5.8)
Ð
5.9)
Ð
1.3)
6.195
s
1.2)
6.267
s
1.6)
6.221
s
1.5)
6.303
s
1.6)
6.949
d
2.1)
6.904
d
1.7)
7.046
d
Acyl
6.99f
6.92
6.92
6.81
6.65
7
7
7
6
6
(8.6)
(8.6)
(8.7)
3A
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
6.368
d
(8.6)
6.356
d
(8.6)
6.492
d
(8.7)
4A
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
5A
6.368
d
(8.6)
6.356
d
(8.6)
6.492
d
(8.7)
6A
Ð
Ð
Ð
Ð
6.195
s
6.267
s
6.221
s
6.303
s
6.949
d
(8.6)
6.904
d
(8.6)
7.046
d
(8.7)
7A
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
7.297
d
(15.9)
7.362
d
(15.9)
7.310
d
(15.9)
7.397
d
(16.0)
7.364
d
(16.0)
7.406
d
(15.9)
7.404
d
(16.0)
7.385
d
(16.0)
7
d
(1
8A
6.197
d
(15.9)
3.413
s
3A,5A
6.137
d
(15.9)
3.431
s
3A,5A
6.308
d
6.240
d
6.130
d
(16.0)
6.103
d
(16.0)
6.178
d
(16.0)
3.787
s
4A
6.338
d
(16.0)
3.602
s
3A
6
d
(1
3
s
2A
(15.9)
3.415
s
3A,5A
3.823
4A
(16.0)
3.431
s
3A,5A
3.762
4A
MeO

D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
183
11
12
13
14
15
16
17
18
19
20
21
22
23
24
4.440
d
(7.1)
ꢁ3.40
4.359
d
(7.1)
4.430
d
(7.3)
4.422
d
(7.6)
4.426
d
(7.3)
4.431
d
(7.3)
4.434
d
(7.4)
3.354
dd
4.458
d
(7.5)
ꢁ3.37
4.452
d
(7.2)
ꢁ3.42
4.452
d
(7.4)
ꢁ3.4
4.429
d
(7.3)
ꢁ3.40
4.403
d
(7.4)
4.644
d
(7.4)
3.312
dd
4.278
d
(7.5)
3.206
dd
e
e
e
e
e
e
(9.4;
7.4)
3.417
dd
(9.8;
9.4)
3.630
dd
(9.8;
8.8)
ꢁ3.38
ddd
(8.8;
2.7;
(9.2;
7.4)
ꢁ3.40
(9.3;
7.5)
ꢁ3.40
e
e
ꢁ3.40
3.640
dd
(9 1;
9.0)
3.621
dd
(9.4;
9.2)
e
3.631
dd
(9.4;
9.0)
3.422
dd
(8.8;
9.4)
3.680
dd
(9.8;
8.8)
3.490
ddd
(9.8;
2.5;
ꢁ3.42
ꢁ3.4
ꢁ3.6
ꢁ3.4
ꢁ3.40
ꢁ3.4
ꢁ3.5
3.661
dd
(9.4;
9.1)
ꢁ3.7
ꢁ3.6
ꢁ3.4
ꢁ3.4
3.661
dd
(9.4;
9.2)
3.644
dd
3.589
dd
ꢁ3.30
ꢁ3.30
(9.6;
9.4)
3.495
ddd
(9.6;
3.1;
(9.8;
9.6)
3.505
ddd
(9.8;
3.3;
e
e
ꢁ3.40
ꢁ3.4
ꢁ3.4
ꢁ3.4
ꢁ3.42
3.484
ddd
(9.5;
3.0;
2.1)
5.210
dd
2.6)
5.154
dd
2.2)
5.094
dd
3.0)
5.105
dd
2.5)
5.122
dd
5.252
dd
5.278
dd
5.230
dd
5.208
dd
5.176
dd
5.221
dd
5.294
dd
5.297
dd
4.589
dd
(
12.0;
(
11.9;
(
12.8;
(
11.9;
(
10.8;
(
11.8;
(
12.0;
(
12.0;
(
12.0;
(
11.8;
(
12.0;
(
11.9;
(
12.0;
(
12.3;
2.7)
4.115
dd
2.6)
ꢁ4.2
3.1)
ꢁ4.2
2.7)
ꢁ4.1
2.6)
ꢁ4.1
2.6)
4.151
dd
2.5)
4.189
dd
2.5)
4.187
dd
2.8)
4.086
dd
2.3)
ꢁ4.2
2.2)
4.250
dd
2.3)
4.282
dd
2.5)
4.383
dd
2.0)
4.083
dd
(
12.0;
(
11.8;
(
12.0;
(
12.0;
(
12.0;
(
12.0;
(
12.2;
(
12.0;
(
12.3;
1.7)
7.046
d
1.9)
7.154
d
2.1)
7.378
d
2.6)
7.465
d
1.5)
6.921
d
3.1)
7.759
d
2.4)
7.887dd
dd
(6.6;
2.0)
3.3)
8.126
d
4.6)
6.861d
d
6.99f
6.92
6.92
6.81
6.65
7.23f
7.13
7.13
6.71
6.59
7.135
d
(7.1)
7.207
dd
dd
(8 6;
5.3)
6.776
dd
7.109
s
(8.7)
(8.5)
(8.2)
(8.8)
(8.9)
(8.8)
(8.2)
(8.6)
6.492
d
(8.7)
7.046
d
(7.8)
6.971
d
(8.5)
7.300
d
(8.2)
7.870
d
(8.8)
6.208
d
(8.9)
Ð
6.405
d
(8.8)
7.097dd
dd
(6.6;
2.0)
7.468
d
(8.2)
6.506d
d
(8.6)
(8.6;
8.7)
Ð
7.264
dd
(7.4;
7.2)
7.046
d
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
6.492
d
(8.7)
6.776
dd
dd
(8.6;
8.7)
7.207
dd
6.971
d
(8.5)
7.300
d
(8.2)
7.870
d
(8.8)
6.208
d
(8.9)
6.405
d
(8.8)
7.097dd
dd
(6.6;
2.0)
7.468
d
(8.2)
6.506d
d
(8.6)
(7.8)
7.046
d
(8.7)
7.135
d
(7.1)
7.154
d
(8.5)
7.378
d
(8.2)
7.465
d
(8.8)
6.921
d
(8.9)
7.109
s
7.759
d
d
(8.8)
7.887dd
dd
(6.6;
2.0)
8.126
d
(8.2)
6.681d
d
(8.6)
dd
(8.6;
5.3)
7.439
d
7.404
d
(16.0)
7.385
d
(16.0)
7.535
d
(16.2)
7.450
d
(16.1)
7.445
d
(16.1)
7.532
d
(16.1)
7.538
d
(16.1)
7.382
d
(16.0)
Ð
Ð
Ð
Ð
Ð
Ð
Ð
ꢁ2.74
ꢁ2.56
(16.1)
6.178
d
(16.0)
3.787
s
4A
6.338
d
(16.0)
3.602
s
3A
6.521
d
(16.2)
3.593
s
2A
6.353
d
(16.2)
6.308
d
(16.1)
6.371
d
(16.1)
6.524
d
(16.1)
6.568
d
(16.1)
6.046
d
(16.0)
Ð
3.623
s
3A,5A
3.605
s
4A
a
Apparent, ®rst-order couplings are indicated (br s, broad singlet; br d, broad doublet; d, doublet; dd, doublet of doublets; ddd doublet of
doublets) in parentheses in Hz.
b
Chemical shift data have been reported. See ref [1].
For a comparison spectrum, see ref [8].
For a comparison spectrum, see ref [10].
Obscured by the MeOH-d4 signal.
Protons of the acyl moiety could not be assigned.
c
d
e
f

184
D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
acylated with cinnamic acids, but not in an example
acylated with 4-hydroxybenzoic acid, which is in
accord with our ®ndings. These observed up®eld
shifts of H-4 and H-8 show that the cinnamoyl
groups are interacting with these two protons in
the chromophore, while the benzoyl groups are
interacting only with H-4. We have previously
shown that H-2A and H-6A (the ortho protons) of
the cinnamic acid acyl groups also have signi®cant
up®eld shifts, showing their interaction with the
chromophore [1]. However, the speci®c nature of
the interactions of the acyl groups with the chromo-
phore are not clear at this time.
Changes in the chemical shifts for H-4 and H-8
of smaller magnitude, but in the same direction as
seen with the monoacylated anthocyanins, are
reported by Idaka et al. [13] and Yoshida et al.
[14], but not by other investigators [15±17] for
anthocyanins acylated with cinnamic acids. In
these latter cases di€erent skeletal arrangements of
sugars between the chromophores and acyl groups
may account for the lack of e€ect of any of the acyl
groups on either H-4 or H-8.
(b) Mass spectral studies. The molecular
weights for compounds 1±24 were supported by
the molecular ions (m/z) observed in both fast-
atom-bombardment mass spectrometry (FABMS)
and electrospray-ionization mass spectrometry
(ESIMS) (see Table 2). While FABMS gave essen-
tially only the molecular ions, ESIMS gave, in
addition to the molecular ion, a series of peaks at
higher m/z values. Except for 14, these peaks were
at M++22 (i.e., M+ H++ Na+), at M++ 40
(i.e., M+ H++ Na++ H2O), at M++ 54 (i.e.,
M+ H++ Na++ CH3OH), and at M++ 70
(i.e., M+ H++ K++ CH3OH). The adducts at
M++ 54 and at M++ 70 did not appear when
H2O was the solvent. Compound 14 gave peaks at
m/z 873 (M+), 905 (M++CH3OH), and 927
(M+ H++ Na++ CH3OH). The mass spectra
of 16 and 22 both showed the expected distribution
of ions resulting from the natural abundances of
35Cl and 37Cl.
(c) UV±vis spectral data. Table 2 contains data
on UV-vis absorbance by the anthocyanins. In
MeOH±HCl the lvis:max of the nonacylated 1±4
Table 2
Concentrations of acids added to carrot cell cultures to produce acylated anthocyanins and the spectral characteristics of the
anthocyanins obtained
Compd
Added acids
(mM) a
m/z
lmax (nm)
MeOH/HCl
lmax (nm)
pH 2.0
lmax (nm)
pH 6.0
Ratio of
Absmax pH 6:Absmax pH 2
1
2
3
4
5
6
7
8
743c
529, 282
526, 282
529, 281
513
509
514
510
527
527
530
529
524
523
525
525
524
523
524
525
526
528
525
528
522
523
523
520
537
530
538
533
547
547
547
550
544
549
546
547
545
546
547
547
546
549
545
545
542
543
542
544
0.09
0.13
0.09
0.13
0.49
0.45
0.42
0.54
0.41
0.57
0.35
0.42
0.41
0.38
0.35
0.38
0.28
0.20
0.37
0.48
0.31
0.15
0.14
0.18
611b,c
581b,c
449b
527, 281
10
10
30
538, 297, 319 d
538, 334, 285
542, 287, 314 d
540, 286, 316 d
536, 316, 285
535, 316, 285
537, 286, 310 d
538, 283, 321 d
537, 326, 281
537, 282
817b
963c
831b
889c
757b
903b
903b
903b
873b,c
891b
907b
941c
918c
916c
937b,c
863b
881b
915c
891c
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
20
6
25
35
15
14
6
538, 282
539, 284
539, 278
541, 285
8
20
60
60
535, 381, 280
539, 278
534, 260±280
536, 282
536, 283
4
100
531, 282
a
Concentration of acid in Me2SO added to cultures at days 4 and 8 at the rate of 0.01 v/v.
Determined by positive-ion ESIMS.
Determined by positive-ion FABMS.
b
c
d
Peak observed as a shoulder.

D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
185
were at 526±529 nm. The acylated compounds
except 24 had lvis:max at 534 nm or longer wave-
lengths. This bathochromic shift was generally
greater when the spectra were measured in bu€er
at pH 2.0. Compounds 5±13 in MeOH±HCl all
had either a shoulder or a peak at 310±335 nm in
addition to the visible and other UV peaks. These
compounds are acylated with cinnamic acids
bearing an electron-donating substituent. In con-
trast, the anthocyanins acylated with cinnamic
acids not having an electron-donating sub-
stituent, with substituted benzoic acids, and com-
pound 24 do not show a peak or shoulder in this
region.
Because anthocyanins undergo a pH-dependent
hydration reaction to give colorless hemiacetals,
the ratio of absorbance at pH 6:pH 2 is a semi-
quantitive measure of color retention by an antho-
cyanin at near-neutral pH. For the anthocyanins
examined here, the lvis:max at pH 6 is increased by
15±25 nm compared to those values a pH 2. The
ratio of absorbance at the visible peak at pH 6:pH
2 is smallest with 1±4 and tends to be greatest with
acyl groups bearing electron-donating substituents.
In general, color retention at pH 6 is greater in the
acylated anthocyanins compared to the non-
acylated ones, is less when the acyl group is a ben-
zoic acid than when it is a cinnamic acid, and tends
to be lower in the anthocyanins those whose acyl
groups bear electron-withdrawing groups. Com-
pound 19 in MeOH±HCl has a spectrum that is
di€erent from those of 5±18. The compound has a
peak at 381 nm whose absorbance is approximately
30% that of the other peaks, which may be due to
the protonated 4-(N,N-dimethylamino)cinnamoyl
group in 19. Support for this contention is that 4-
(N-dimethylamino)cinnamic acid has a small peak
(approximately 8% of the 265 nm absorbance) near
360 nm in MeOH±HCl and in aqueous bu€er at
pH 2.
The range of carboxylic acids that these carrot
cell cultures used to acylate anthocyanins is not
surprising. In related examples, Tabata et al.
showed that suspension cultures from four di€er-
ent species could each glucosylate three ¯avonoids,
three anthraquinones, four coumarins and four
hydroxylated benzoic acids [18], and Naoshima
and Akakabe [19] and Akakabe et al. [20] showed
enantiospeci®c reduction of four ketoesters, ®ve
aromatic ketones, and three heterocyclic ketones
by plant cell cultures. The earlier literature on bio-
conversions by plant cell cultures has been
reviewed by Pras [21].
Partial hydrolysis in HCl, but not in tri-
¯uoroacetic acid, of most acylated and non-acy-
lated anthocyanins of this series preferentially
removed the xylose residue. This occurred in NMR
samples in CD3OD acidi®ed with DCl and stored
in a freezer, on freeze drying solutions containing
HCl, and in 5 N HCl at room temperature. The
anthocyanin 24 was particularly sensitive to acid
hydrolysis of the acyl function. When a sample of
puri®ed 24 was stored in 0.1% tri¯uoroacetic acid
at 20 ꢀC for a month, 1 was formed in signi®cant
amounts. Storage in 0.1% formic acid either at 4 ꢀC
or 20 ꢀC did not lead to detectable deacylation of
the compound during 2 weeks.
In all but two cases, the feeding of a substituted
cinnamic acid to the cultures resulted in the pre-
sence of only one novel, cinnamoylated anthocyanin
in the tissues. However, when either 4-(tri-
¯uoromethyl)cinnamic acid or 4-chlorocinnamic
acid was provided, two novel anthocyanins were
obtained. In each of the latter examples, there was
found one anthocyanin acylated with the cinnamic
acid (17 and 16, respectively) and another acylated
with the corresponding benzoic acid (23 and 22,
respectively). A speci®c attempt to detect the pro-
duction of anthocyanins acylated with two other
benzoic acids in a manner similar to the pre-
vious examples was made by feeding 4-meth-
oxycinnamic acid and 3,4,5-trimethoxycinnamic
acids. Using an HPLC system known to separate
the benzoylated anthocyanins from their cin-
namoylated counterparts, no benzoylated antho-
cyanins were detected in the tissue extracts for
these latter two examples. Thus, the process in
these carrot cell cultures that is responsible for the
conversion of cinnamic acids to benzoic acids
appears to have a greater speci®city than does the
mechanism for the addition of the benzoic acids to
the anthocyanins.
Further observations on the production of novel
anthocyanins.ÐIn addition to the anthocyanins
acylated with the carboxylic acids shown for com-
pounds 5±24, novel anthocyanins in limited
amounts (data not shown) were detected on HPLC
analysis of extracts of tissues fed 3,4-methylene-
dioxycinnamic acid, 4-methoxybenzoic acid, 4-
hydroxy-3,5-dimethoxybenzoic acid, and 4-hydro-
xyphenylacetic acid. The anthocyanin product
acylated with 4-hydroxyphenylacetic acid was not
pure after chromatography on the three adsorbents
routinely used in this study.

186
D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
The conversion of cinnamic acids to benzoic
and by ESIMS in the positive-ion mode in 50%
MeOH±H2O, except when noted otherwise.
ESIMS was performed with a Quattro-II mass
spectrometer (Micro-Mass, Inc., Manchester, UK).
UV-vis spectra were measured in 1% v/v concd
HCl in 95% MeOH, and vis spectra were measured
in 0.1 M acetate + 0.1 M phosphate adjusted to
pH 2.0 and to pH 6.0 (HCl) within an hour of
preparing the solution using a Hitachi Model U-
2000 UV-vis spectrophotometer (Hitachi Instru-
acids has been examined in only a few systems.
Loscher and Heide [22] showed in extracts of
Lithospermum erythrorhizon cell cultures that 4-
coumaroyl-CoA and 4-hydroxybenzoyl-CoA were
intermediates in the conversion, while Schnitzler et
al. [23] concluded that 4-hydroxybenzaldehyde was
an intermediate in the conversion by extracts from
carrot cell cultures. Funk and Brodelius [24] con-
cluded that 4-methoxycinnamic acids and 4-meth-
oxybenzoic acids were intermediates in the
formation of benzoic acids in cell cultures of
Vanilla planifolia. Benzoic acid appears to be an
intermediate in the formation of salicylic acid from
trans- cinnamic acid in tobacco [25] and rice [26].
Thus, although the conversion of a cinnamic acid
into a benzoic acid is formally analogous to ꢀ oxi-
dation in fatty acids, there appear to be several
di€erent pathways by which the C2 (ethylene) unit
is removed.
In our studies cultures treated with unsub-
stituted cinnamic acid yielded the 4-hydroxylated
cinnamoyl analogue (4-coumaroyl) 9 as the major
product along with the new, relatively hydro-
phobic, nonhydroxylated anthocyanin 14 in a ratio
of approximately 10:1.
The formation of the anthocyanin acylated with
p-coumaric acid is not surprising because cinnamic
acid has been identi®ed as the precursor of p-cou-
maric acid in the hydroxycinnamic acid pathway
that provides intermediates in the synthesis of
anthocyanins and other compounds [27]. These
acyl groups are donated to anthocyanins via 1-O-
1
ments Inc., San Jose, CA). H NMR spectra were
measured using a Bruker AMX 400 spectrometer
as previously described [1], except that CF3CO2D
was substituted for DCl to avoid the observed
partial hydrolysis of xylose from the compounds.
This change made no observable di€erence to the
spectrum of 17, which demonstrates that spectra
measured in either acid can be directly compared.
For some compounds, the assignment of peaks to
speci®c protons was done from 2D spectra, and in
other cases by analogy to compounds where the
assignments had been determined from 2D spectra.
In the NOESY experiments, a total of 256 FIDs of
2K datapoints were collected with a total of 64
scans per FID with a spectral width of 3205 Hz.
The data was zero-®lled to 1K points in t1 and
multiplied by the sine2 window function in both
dimensions before Fourier transformation. The
mixing time was 500 ms, and the transmitter o€set
was placed on the H2O resonance, which was irra-
diated with a lower power pulse during a relaxation
delay of 2 s to suppress the water signal. All pro-
cessing was performed with FELIX 95 software
(Biosym/Molecular Simulations) on a Silicon Gra-
phics Indigo2 workstation.
Biological material.ÐThe wild carrot (Daucus
carota spp carota) cell clone used in these studies
was WC63-1-9-1-13-1 [30], and a subline was
selected from it for increased anthocyanin yield.
Culture preparation.ÐFor each acid tested, a
concentration that gave little or no growth inhibi-
tion and the best level of total and novel antho-
cyanin in the tissue was selected experimentally,
and these data are given in Table 2. For puri®ca-
tion of the novel anthocyanins, 1.5±2 L of culture
in 100-mL portions was grown for 12 days, with
the chosen concentration of the individual car-
boxylic acids in Me2SO added (1% of culture
volume) at days 4 and 8. The tissue was ®ltered on
Miracloth (Calbiochem±Novabiochem Intl., San
Diego, CA) washed with water, frozen in liquid
nitrogen, and stored at 80 ꢀC.
(4-hydroxycinnamoyl)-ꢀ-d-glucopyranoses
[28].
Previous study of the apparent speci®city of
extracts of these carrot cell cultures that formed 1-
O-sinapoyl-ꢀ-d-glucopyranose showed no activity
with trans-cinnamic acid [29]. This ®nding suggests
that the methods used by these workers may not
have been suciently sensitive to detect the cinna-
moyl ester.
3. Experimental
General procedures.ÐExcept as described below,
the culture and analytical methods used are descri-
bed in Baker et al. [7]. All the acids used were
obtained from Aldrich Chemical Company, Mil-
waukee, WI. Mass spectra were measured by
FABMS in the positive-ion mode in thioglycerol [7]

D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
187
Isolation of anthocyanins.ÐThe frozen tissue
was extracted overnight in a refrigerator with
10 mL/g fresh weight of methanol containing 5%
v/v formic acid. The ®ltered extract was con-
centratedꢀ approximately ®ve-fold in vacuo at less
than 30 C to remove methanol, extracted twice
with 0.5 volumes of CH2Cl2, and then twice with
0.5 volumes of EtOAc. The aqueous residue was
chromatographed on Amberlite XAD-7 resin
(Sigma Chemical Co., St. Louis, MO) (2.5Â20 cm
or 4Â20 cm) using a gradient from 0.1% formic
acid in water to 0.1% formic acid in methanol. The
fractions containing the novel anthocyanin were
selected by HPLC using either a ꢂBondapak C18
column (3.9Â300 mm, Waters Associates, Inc.,
Milford, MA), or a Microsorb-MV C18 column
(4.6Â250 mm, Rainin Inst. Co., Woburn, MA),
and either isocratic mixtures of or gradients of (a)
5% formic acid in water, and (b) 5% formic acid
plus 55% methanol plus 10% acetonitrile in water.
The combined fractions were concentrated under
reduced pressure and chromatographed on a C18
reversed-phase column (1.5Â44 cm, 28±40 ꢂm, E.
Merck, Gibbstown, NJ) with a gradient from 5%
formic acid in water to 5% formic acid plus 55%
methanol plus 10% acetonitrile in water. The frac-
tions containing the novel anthocyanin were com-
bined and concentrated under reduced pressure. In
some cases two novel anthocyanins were separated
on the C18 reversed-phase column and were treated
separately. The new anthocyanins were then chro-
matographed on a lipophilic Sephadex LH-20 column
(2.5Â20 cm, Pharmacia Biotech, Inc., Piscataway,
NJ) with 0.1% formic acid in water as the solvent.
The fractions containing the desired anthocyanin
were combined, concentrated in vacuo, dissolved in
0.1% formic acid and stored frozen. Compound 4
was prepared as described above from the skin of
red apples bought in the local market [31].
graphy was mixed with 3. These two compounds
were then separated by semipreparative HPLC on
a 7.9Â300 mm C18 ꢂBondapak column (Waters
Associates, Inc., Milford, MA) by isocratic elution
with a mixture of the solvents described above for
the preparative C18 column. They were also well
separated on the PVP column, eluting ®rst with
0.1% formic acid in water and then a gradient to
methanol containing 0.1% formic acid.
Compound 1 was prepared from either 5 or 7
that had been chromatographed on Amberlite
XAD-7 and C18 media as described above. The
anthocyanin (30 mg) in water at room temperature
was treated with stirring, under nitrogen, with an
equal volume of 1.8 N KOH for 7 min, followed by
5 volumes of N formic acid. The deacylated pro-
duct was puri®ed by chromatography as above on
C18 reversed phase and on Sephadex LH-20 col-
umns. Compound 2 was prepared from 10 by the
same method. When a mixture of 9 and others,
predominantly 3, was deacylated and chroma-
tographed as above, 1 did not separate from 3 but
was easily separated on the PVP column as above.
Compound 8 was prepared from an NMR sample
of 7 in CD3OD and DCl that had been observed to
lose the xylose residue. The sample was chromato-
graphed on C18 reversed phase and on Sephadex
LH-20 columns to give pure 8. Compound 2 was
similarly obtained from an NMR sample of 1. Xylose
was removed from 5 on freeze-drying a solution of
the compound in 0.1% HCl. Xylose was removed
preparatively fromꢀ5, 7, and 9 by partial hydrolysis
in 5 N HCl at 25 C for 60 to 72 h, and the pro-
ducts were chromatographed on C18 reversed
phase and on Sephadex LH-20 columns to give pure
6, 8, and 10. Compounds 9 and 10 were also easily
separated on a PVP column as described above.
In the cases of 5 and 20, the anthocyanin after
chromatography on Sephadex LH-20 was rechro-
matographed on a PVP column (Polyclar AT
>0.2 mm, 2.5Â20 cm), eluted with 0.1% formic
acid in water, and the fractions containing the
novel anthocyanin were selected and treated as
above. Compound 20, at this stage, was still con-
taminated with a component absorbing at 280 nm
that was removed by semipreparative chromato-
graphy on a 7.9Â300 mm C18 ꢂBondapak column
eluted isocratically with a mixture of 0.1% formic
acid and 80% methanol. The anthocyanin 21 when
puri®ed through the Sephadex LH-20 chromato-
4. Conclusions
We have demonstrated that the systems opera-
tive in tissue cultures of Daucus carota spp carota
(wild carrot) for acylating anthocyanins can tolerate
a wide variety of cinnamic and benzoic acids.
Incorporation of these acids into monoacylated
anthocyanins 5±24 occurs solely at the C-6 of the
glucose moiety. In all but two instances studied, the
only anthocyanin isolated was the one acylated with
the carboxylic acid provided. However, when the
culture was fed with either 4-(tri¯uoromethyl)-
cinnamic acid or 4-chlorocinnamic acid, the

188
D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
anthocyanin acylated with the corresponding ben-
zoic acid (compounds 23 and 22, respectively) was
also isolated.
[4] R. Brouillard, Chemical Structure of Anthocyanins,
in P. Markakis (Ed.), Anthocyanins as Food
Colors, Academic, New York, 1982, chapter 1, pp
1±40.
We have used the ratio of Absmax pH 6.0:Absmax
at pH 2.0 for compounds 5±24 to show that acyla-
tion enhanced the color retention of these com-
pounds at near-neutral pH. Cinnamoylated
anthocyanins in general exhibited greater color
stability compared to their benzoylated counter-
parts. Moreover, if the acyl group bore an electron-
donating substituent, the color retention of the
anthocyanin was further increased. Additional
investigations into their ability to retain color will
be reported in the future.
1H NMR studies of the monoacylated antho-
cyanins 5±24 showed that there is an interaction
between the acyl group and the chromophore
where the pH-dependent hydration reaction occurs
that results in the loss of color. The precise nature
of this interaction is being probed as a part of
continued studies.
[5] G. Mazza and E. Miniati, Anthocyanins in Fruits,
Vegetables, and Grains, CRC Press, Boca Raton,
FL, 1993, pp 6±10.
[6] O. Dangles, N. Saito, and R. Brouillard, J. Am.
Chem. Soc., 115 (1993) 3125±3132.
[7] D.C. Baker, D.K. Dougall, W.E. Glaûgen, S.C.
Johnson, J.W. Metzger, A. Rose, and H.U. Seitz,
Plant Cell, Tissue and Organ Culture, 39 (1994) 79±
91.
[8] W.E. Glaûgen, V. Wray, D. Strack, J.W. Metzger,
and H.U. Seitz, Phytochemistry, 31 (1992) 1593±
1602.
[9] D.K. Dougall, D.C. Baker, E. Gakh, and M.
Redus, Food Technol., 51 (1997) 69±71.
[10] N. Terahara, N. Saito, K. Toki, Y. Sakata, and T.
Honda, Phytochemistry, 31 (1992) 1446±1448.
[11] H. van Halbeek, Curr. Opin. Struct. Biol., 4 (1994)
697±709, and references cited therein.
[12] D. Strack and V. Wray, Anthocyanins, in P.M. Dey
and J.B. Harborne (Eds.), Methods in Plant Bio-
chemistry, Vol. 1, Academic Press, New York,
1989, chapter 9, pp 325±356.
Acknowledgements
[13] E. Idaka, K. Suzuki, H. Yamakita, T. Ogawa,
T. Kondo, and T. Goto, Chem. Lett., (1987) 145±
148.
[14] K. Yoshida, T. Kondo, and T. Goto, Tetrahedron,
48 (1992) 4313±4326.
[15] N. Saito, K. Toki, K. Uesato, A. Shigihara, and T.
Honda, Phytochemistry, 37 (1994) 245±248.
[16] T.S. Lu, N. Saito, M. Yokoi, A. Shigihara, and T.
Honda, Phytochemistry, 31 (1992) 289±295.
[17] K. Toki, N. Saito, K. Kawano, T.S. Lu, A. Shigi-
hara, and T. Honda, Phytochemistry, 36 (1994)
609±612.
[18] M. Tabata, Y. Umetani, M. Ooya, and S. Tanaka,
Phytochemistry, 27 (1988) 809±813.
[19] Y. Naoshima and Y. Akakabe, Phytochemistry, 30
(1991) 3595±3597.
[20] Y. Akakabe, M. Takahashi, M. Kamezawa, K.
Kikuchi, H. Tachibana, T. Ohtani, and Y. Naosh-
ima, J. Chem. Soc., Perkin Trans. 1, (1995) 1295±
1298.
We thank Je€rey Morgan for his excellent tech-
nical assistance during this work. We are grateful
to Richard Cox and Engin Serpersu for their help
in the construction of Fig. 1. Financial support of
this work by USDA±NRICGP Grant 94-37500-
0512 is greatly appreciated. Mass spectra were
obtained by Dr. A. Tuinman, University of Ten-
nessee, Knoxville, Chemistry Mass Spectrometry
Center, which is funded by the Science Alliance, a
State of Tennessee Center of Excellence. The
National Science Foundation contributed to the
acquisition of the Quattro-II mass spectrometer
(Grant no. BIR 94 08252). The NMR instrument
was purchased on a grant from the Department of
Energy (DEF6 05893460892).
[21] N.J. Pras, J. Biotechnol., 26 (1992) 29±62.
[22] R. Loscher and L. Heide, Plant Physiol., 106
(1994) 271±279.
[23] J.-P. Schnitzler, J. Madlung, A. Rose, and H.U.
Seitz, Planta, 188 (1992) 594±600.
[24] C. Funk and P.E. Brodelius, Plant Physiol., 94
(1990) 102±108.
[25] N. Yalpani, J. Leon, M.A. Lawton, and Raskin,
Plant Physiol., 103 (1993) 315±321.
[26] P. Silverman, M. Seskar, D. Kanter, P. Schweizer,
References
[1] E.G. Gakh, D.K. Dougall and D.C. Baker, Phy-
tochem. Anal., 9 (1998) 28±34.
[2] F.J. Francis, CRC Crit. Rev. Food Sci. Nutr., 28
(1989) 273±314.
[3] P. Markakis, Anthocyanins as Food Additives, in
P. Markakis (Ed.), Anthocyanins as Food Colors,
Academic, New York, 1982, chapter 9, pp 245±
253.

D.K. Dougall et al./Carbohydrate Research 310 (1998) 177±189
189
J.-P. Metrauz, and I. Raskin, Plant Physiol., 108
(1995) 633±639.
[29] F.T. Halaweish and D.K. Dougall, Plant Sci., 71
(1990) 179±184.
[27] W. Heller and G. Forkmann, Biosynthesis, in J.B.
Harborne (Ed.), Flavonoids, Advances in Research
Since 1980, Chapman and Hall, New York, 1988,
chapter 11, pp 399±425.
[28] W.E. Glaûgen and H.U. Seitz, Planta, 186 (1992)
582±585.
[30] D.K. Dougall and D.L. Vogelien, Plant Cell, Tis-
sue and Organ Culture, 23 (1990) 79±91.
[31] G. Mazza and E. Miniati, Anthocyanins in Fruits,
Vegetables, and Grains, CRC Press, Boca Raton,
FL, 1993, pp 29±31.