Anthocyanins from wild carrot suspension cultures acylated with supplied carboxylic acids

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

,

^aDepartment of Botany and the Biotechnology Program, Life Sciences Graduate Degree Program,

The University of Tennessee, Knoxville, TN 37996, USA

^bDepartment 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

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.

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 eects 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

eects) between the acyl group and the ¯avylium

chromophore. These objectives require the avail-

ability of a series of anthocyanins that dier 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

dier 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 eects 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-6^IIIof the glucose

moiety as the chemical shift of the H-6^IIIAof 5 is

observed at ꢁ 5.339 and that for 1 at ꢁ 3.846. There

is also a much smaller down®eld shift of H-6^IIIB

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-6^IIIA

.

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 eect 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-2^Iis found up®eld at

less than ꢁ 4.0 (see data for 2, 4, 6, 8, and 10), and

when xylose is present, H-2^Iis 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-1^III±H-5^I, (b) H-1^III

H-6^IA, (c) H-5^I±H-6^IA, and (d) H-1^III±H-6^IB

±

.

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)

(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

1^I

2^I

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

3^I

ꢁ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

4^I

5^I

ꢁ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

6^IA

ꢁ3.9

ꢁ3.8

ꢁ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

6^IB

(

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

1^II

Ð

4.684

d

(7.7)

3.135

dd

Ð

4.667

d

(7.4)

2^II

Ð

e

(8.9;

7.7)

e

3^II

Ð

4^II

3

e

d

(9

9

5^IIA

Ð

3.606

dd

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)

5^IIB

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

4.201

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

ꢁ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

3.521

dd

ꢁ3.5

3.622

dd

ꢁ3.5

ꢁ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

1^III

4.267

d

(7.7)

3.178

dd

(9.3;

7.7)

ꢁ3.25

4.234

d

(7.5)

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

2^III

3^III

4^III

5^III

Ð

e

ꢁ3.16

ꢁ3.20

e

ꢁ3.44

ꢁ3.45

ꢁ3.43

ꢁ3.40

ꢁ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

ꢁ3.44

ꢁ3.45

ꢁ3.42

ꢁ3.40

6^IIIA

6^IIIB

2^A

3.846

dd

3.835

dd

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

6.99^f

6.92

6.81

6.65

7

6

(8.6)

(8.7)

3^A

Ð

6.368

d

(8.6)

6.356

d

(8.6)

6.492

d

(8.7)

4^A

Ð

5^A

6.368

d

(8.6)

6.356

d

(8.6)

6.492

d

(8.7)

6^A

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)

7^A

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

8^A

6.197

d

(15.9)

3.413

s

3^A,5^A

6.137

d

(15.9)

3.431

s

3^A,5^A

6.308

d

6.240

d

6.130

d

(16.0)

6.103

d

(16.0)

6.178

d

(16.0)

3.787

s

4^A

6.338

d

(16.0)

3.602

s

3^A

6

d

(1

3

s

2^A

(15.9)

3.415

s

3^A,5^A

3.823

4^A

(16.0)

3.431

s

3^A,5^A

3.762

4^A

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

(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

ꢁ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

3.661

dd

(9.4;

9.2)

3.644

dd

3.589

dd

ꢁ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

ꢁ3.40

ꢁ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

6.99^f

6.92

6.81

6.65

7.23^f

7.13

6.71

6.59

7.135

d

(7.1)

7.207

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

6.492

d

(8.7)

6.776

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

(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

ꢁ2.74

ꢁ2.56

(16.1)

6.178

d

(16.0)

3.787

s

4^A

6.338

d

(16.0)

3.602

s

3^A

6.521

d

(16.2)

3.593

s

2^A

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

3^A,5^A

3.605

s

4^A

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-d₄signal.

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-2^Aand H-6^A(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 dierent skeletal arrangements of

sugars between the chromophores and acyl groups

may account for the lack of eect 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 l_vis:maxof 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

l_max(nm)

MeOH/HCl

l_max(nm)

pH 2.0

l_max(nm)

pH 6.0

Ratio of

Abs_maxpH 6:Abs_maxpH 2

1

2

3

4

5

6

7

8

743^c

529, 282

526, 282

529, 281

513

509

514

510

527

530

529

524

523

525

524

523

524

525

526

528

525

528

522

523

520

537

530

538

533

547

550

544

549

546

547

545

546

547

546

549

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

611^b,c

581^b,c

449^b

527, 281

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

817^b

963^c

831^b

889^c

757^b

903^b

873^b,c

891^b

907^b

941^c

918^c

916^c

937^b,c

863^b

881^b

915^c

891^c

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

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 l_vis:maxat 534 nm or longer wave-

lengths. This bathochromic shift was generally

greater when the spectra were measured in buer

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 l_vis:maxat 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

dierent 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 buer 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 dier-

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

dierent 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 dierence 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 t₁and

multiplied by the sine²window function in both

dimensions before Fourier transformation. The

mixing time was 500 ms, and the transmitter oset

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 suciently 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-

centrated_ꢀ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 C₁₈column

(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 C₁₈reversed-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 C₁₈ꢂBondapak column (Waters

Associates, Inc., Milford, MA) by isocratic elution

with a mixture of the solvents described above for

the preparative C₁₈column. 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 C₁₈media 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

C₁₈reversed 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 C₁₈reversed 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 from_ꢀ5, 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 C₁₈reversed

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 C₁₈ꢂBondapak 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 Abs_maxpH 6.0:Abs_max

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 Jerey 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.

Article Doi

Anthocyanins from wild carrot suspension cultures acylated with supplied carboxylic acids

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

Source and publish data:

Authors:

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

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

Products guided by the article

R&D Labs maybe for 164203-04-7

Relevant to this article

Hot Product