BIOSYNFHESIS 0F UBP-APEOSE AND APQIN

Thesis for the Degree of Ph. D.
MICHfiGAN STATE UNNERSHY
RONALD ROSS WATSOfl
1971

This is to certify that the

thesis entitled

BIOSYNTHESIS OF UDP-APIOSE AND APIIN

presented by

Ronald Ross Wats on

has been accepted towards fulfillment
of the requirements for

gal/(W

Major professor

Dateww

0-7639

 

ABSTRACT

BIOSYNTHESIS OF UDP-APIOSE AND APIIN

BY

Ronald Ross Watson

A method was developed for synthesizing the product of the
reaction catalyzed by UDPGA cyclase. The reaction product was syn-
thesized from the substrate of the enzyme, UDP-glucuronic acid,
using purified UDPGA cyclase isolated from Lemna minor. Catalytic
quantities of NAD+ were required for the synthesis of the reaction
product. The reaction product was extensively characterized. It
migrated on paper in four different solvents with an RF of uridine
5'-(d-D-xy10pyranosyl pyrophosphate), UDP-xylose, and had the same
mobility as UDP-xylose when subjected to paper electrophoresis at
pH 5.8. When [BHJUDP-[U-luc]glucuronic acid was used as the sub-
strate, the 3H:1uC ratio in the reaction product was that expected
if D-apiose remained attached to the uridine. The reaction pro-
duct yielded [3H]urid1ne 5'-diphosphate ([3HJUDP) and D-[U-luCJ-
apiose when hydrolyzed at pH 2 and 1000 for 15 minutes. When
boiled at pH 8.0 for 5 minutes, it phosphorylated intramolecularly,
yielding [3H]uridine 5'-monophosphate ([3H]UMP) and a-D—[U-luCJapio-
D-furanosyl cyclic 1:2-phosphate. The reaction product served
as the [U—luc]apiosy1 donor in the enzymatic synthesis of apiin
from 7-(4',3,7-trihydroxyf1avonyl)O-B-D-glucopyranoside. The
other product of this reaction was identified as [BHJUDP. These
results established that the reaction product is uridine

5'-(a-D-apio—D-furanosy1 pyrophosphatel‘UDP—apiose. Through the

O

Ronald Ross Watson

use of purified UDPGA cyclase UDP-apiose could be prepared in
approximately 50% yield with a radiochemical purity of at least
60%. Stability studies showed that UDP-apiose was hydrolyzed to
UMP and a-D-apio-D-furanosyl cyclic-1:2-P by exceedingly mild
alkaline conditions. In a solution which contained 0.1 nmoles of
UDP-[U-lucjapiose approximately one-third of the UDP-[U-140]apiose
was hydrolyzed in 2 hours at pH 8.0 and 25°. Stability studies
showed that UDP-apiose was hydrolyzed by exceedingly mild acid
conditions (pH 4 and 25°). UDP-Apiose was stored for months with-
out degradation at pH 5-6 at .20° in a solution 50$ ethanol by
volume. UDP-xylose was not degraded by mild acid or alkaline conditions.

The purified duckweed UDPGA cyclase contained UDPGA decarboxy- .
lase which formed [3H]UDP-[U-1uc]xylose from [BHjunP-[U-luCngu-
curonic acid and NAD+.

D-Apiose was shown to be transglycosylated from UDP-apiose in
a reaction catalyzed by apiin synthase. This enzyme catalyzed the
transglycosylation of D-apiose from UDP-apiose forming apiin.
Apiin synthase [UDP-apiose: 7-(u',5,7,-trihydroxyflavonyl) O-B-D-
glucopyranoside, D-apiose transferase] was isolated from parsley
and purified #5 fold. During purification it was separated from
the parsley UDPGA cyclase. The substrates of the reaction, UDP-
apiose and 7-(4',5,7-trihydroxyf1avony1) B-glucopyranoside were iden—
tified. Apiin and UDP were identified as the products of the reac-
tion. The pH optimum for apiin synthase is 7.6-8.4. The rate of
formation of apiin increased linearly with increasing amounts of
apiin synthase or increasing incubation. The Km for 7-(4',5,7-
trihydroxyflavonyl) O-B-D-glucopyranoside is 7.0 x 10"5 M.

Ronald Ross Watson

The Km for UDP—apiose is 0.6 x 10‘5 M. Apiin synthase occurs
primarily in the leaves. None of 12 compounds tested which con-
tained metal ions increased apiin synthase activity. Five com-
pounds which chelate various metal ions were incubated with apiin
synthase and none inhibited it. Sulfhydryl reagents inhibited
apiin synthase when incubated with it at 1 mM final concentrations.
A11 (100%) of the apiin synthase activity was lost when p-chloro-
mercuribenzoate, 75% with iodoacetamide and less with oxidized
glutathione and N-ethylmaleimide. Uridine, UMP, UDP, UTP, UDP-
galactose, UDP-glucuronic acid and UDP-xylose inhibited apiin syn-
thase to some extent, but UMP and UDP inhibited it most at the

'6 and l x 10'5 M). Apiin synthase

lower concentrations (1 x 10
was also isolated from Digitalis purpurea (foxglove) as shown by
the formation of a compound with the Rf of apiin from UDP-apiose
and 7-(4',5,7-trihydroxyf1avonyl) O-B—D-glucopyranoside. Enzyme
isolated from foxglove was necessary for the formation of the com-
pound with the Rf of apiin.

A method for isolating and purifying UDPGA cyclase from
Petroselinin crispum (parsley) is described. It was purified 93
fold from parsley leaves. The pH optimum was 8.0-8.3 and the Km
was 0.33 x 10'5 M for UDP-glucuronic acid. Exogenous NAD+ was re-
quired to synthesis any UDP—apiose. Optimum synthesis was obtained
with 1-2 x 10"3 M NAD+. UDPGA decarboxylase activity was found
with purified UDPGA cyclase. The pH for optimum activity of
parsley UDPGA decarboxylase activity was 8.0-8.2. UDP—xylose in-
hibits UDPGA decarboxylase more than UDPGA cyclase and at lower

'6 M). Exogenous NAD+ was not required to

concentrations (1 x 10
synthesize 65$ of the amount of UDP-xylose formed with the optimum

amount of NAD+,

BIOSYNTHESIS OF UDP-APIOSE AND APIIN

BY

Ronald Ross Watson

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Biochemistry
1971

,/ ’73". ‘2’}
" J J"

VITA

Ronald Ross Watson was born in Texas in December of 1942.
He was raised and educated in Moscow, Idaho. He graduated
from Moscow High School in 1961. That same year he began
his Sophomore year at the University of ldaho which was in-
terrupted to preform service for the Church of Jesus Christ
of Latter-day Saints in Chile. After two and one-half years
in Chile he returned obtaining his B.S. in Chemistry from
Brigham Young University on May 27, 1966. That significant
day in May marked his marriage to Anita Ann Hebert and the
start of their journey to the East for further study. After
four and one-half years at Michigan State University they
are continuing their trek eastward to Harvard University
where Ronald will participate with Drs. G. Edsall and A. B.

MacDonald in research.

11

' . . . . . There is in science quite a gap between
belief and certainty. But would one ever have the patience
to wait and establish the certainty if the inner conviction
was not already there [Jaques Monod (75)] ”Truth is the
knowledge of things as they are, and as they were, and as
they are to come.‘ [Joseph Smith (74)]

111

ACKNOWLEDGMENTS

I wish to thank Dr. Paul K. Kindel for his example,
advice and counsel during my research and graduate training.

I also wish to thank him for his invaluable contribution
towards my development as a biochemist.

I appreciate the valuable advice, assistance and time
of my thesis committee members, Drs. S. H. Wittwer, N. E.
Tolbert, R. A. Ronzio and F. M. Rottman.

My special thanks to Bernard Bruinsma for the learning
opportunity his 1970 summer research project became for both
of us. I also appreciate the technical assistance of Mrs.
Beth Laine.

The love, encouragement and support of my parents Drs.

R. D. and E. K. Watson throughout my life, especially during
the past few years is greatfully acknowledged. Without their
encouragement and example I would not have set and met the
goals such as this thesis represents. Their love of learning
and improvement has always been an inspiration. The faith
and support by my wife's mother, Hardis Stott, from afar, was
especially valuable and appreciated.

Finally, my most sincere thanks and love to my wife Anita
for her continued support. My thanks for her understanding as
she did without some of the material benefits of a working
husband as well as being along way from home and family. I
appreciate her wisdom in ordering the important phases of our
family life, her love and her encouragement.

The financial assistance of the National Science Foundation
and the National Aeronautics and Space Administration is

appreciated.
iv

TABLE OF CONTENTS

VITA O O O O O C O O O O O O O O O O O O O O O O O O 0

ACKNOWLEDGMENTS. . . . . . . . . . . . . . . . . . . .
LIST OF TABLES . . . . . . . . . . . . . . . . . . . .
LIST OF FIGURES. . . . . . . . . . . . . . . . . . . .
ABBREVIATIONS. . . . . . . . . . . . . . . . . . . . .

PART 1. CHARACTERIZATION AND BIOSYNTHESIS OF D-APIOSE
AND OF COMPOUNDS CONTAINING D-APIOSE. . . .

Occurrence of D-Apiose in Nature . . .

Characterization of Compounds
Containing D—Apiose. . . . . .

In vivo Biosynthesis of D-Apiose . . .

PART 2. BIOSYNTHESIS AND CHARACTERIZATION OF
UDP-APIOSE O O O O O C O O O O C O O O O O 0

INTRODUCTION 0 O C O O O O O O O O O O O O 0
MATERIALS AND METHODS . . . . . . . . . . .

Materials. . . . . . . . . . . . . . .
General Methods. . . . . . . . . . . .
Paper Chromatography . . . . . . .
Preparation and Assay of UDPGA Cyclase
and Apiin Synthase . .
Preparation of [ 3H]UDP—[U-1&C]-
glucuronic Acid. . . . . . . . . .

RESULTS 0 O O O O C O C O O O I O O O O O 0

Identification of the Product of the
Reaction Catalyzed by UDPGA Cyclase.

DISCUSSION 0 O O O O O O O O O O O O O O O O

Page

ii
iv
ix
xi
xiii

OUNH

26
49

Page

PART 3. PURIFICATION AND CHARACTERIZATION OF APIIN
SYNTHASE FROM 2. crispum. . . . . . . . . . . . 66

INTRODUCTION 0 O O O O O O O O O O O O O O O O O 67

MATERIALS AND METHODS . . . . . . . . . . . . . 68

materials 0 O O O C O O O O O 0 O O O O O 68
General Methods . . . . . . . . 69
Standard Assay for Apiin Synthase . . . . 70
Absorbance and Solubility of '7GA'. . . . 75

RESULTS 0 O O O O O O O O O O O O O O O O O O O 76

IDENTIFICATION OF ONE OF THE REACTION

PRODUCTS FORMED Bx APIIN SINTHASE . . . .76

In vitro Enzymatic Formation of Apiin
(Compound IV). . . . .76
Identification of Compound IV as Apiin, .‘77

DISTRIBUTION OF APIIN SYNTHASE IN THE

LEAVES, STEMS AND ROOTS OF PARSLEY. . . .82

PURIFICATION OF APIIN SYNTHASE. . . . . . . .83

EXtraction-o O O O O O O O O O O O O O O O 83
Ammonium Sulfate. . . . . . . . . . . . .84
Sephadex G-100. . . . . . . . . . . . . .84
DEAE-Sephad-ex O O O O O O O O O O O O O O 8“

PROPERTIES OF APIIN SYNTHASE. ... . . . . . .91

Linearity of the Apiin Synthase Reaction. 91
Factors Affecting the Stability of

Apiin Synthase. . . . . . . . . .91
pH Optimum of Apiin Synthase. . . . . . .98
Affinity of Apiin Synthase for UDP-

apiose and '7GA'. . . . . . . . .98
Effect of Various Ions on the Activity

of Apiin Synthase . . . . . . . .98
Effect of Sulfhydryl Reagents on the

Activity of Apiin Synthase. . . . .105
Inhibition of Apiin Synthase by Five '

Uracil Compounds. . . . .105

Fractionation of Apiin Synthase with
Ammonium Sulfate. . . . . . . . . . . .106

DISCUSSION. 0 I O O O O O O O O O O O O O O O O 1'10

PART 4. PURIFICATION AND CHARACTERIZATION OF UDPGA

CYCLASE FROM 3.

crispug . . . . . . . . . . . .115

INTRODUCTION. . . . . . . . . , , , , , , , , .116

vi

Page

MATERIALS AND METHODS. . . . . . . . . . . . .117

materials 0 O O O O O O O O O O O O O .117
General Methods. . . . . . . . . . . . .118
Definition of Units. . . . . . . . .119
Standard Assays I and II for Parsley

UDPGA Cyclase. . . . . . . . . .119
Standard Assays I and II for Duckweed

UDPGA Cyclase. . . . . . . .121
Standard Assay for Parsley UDPGA

Decarboxylase. . . . . .121
Standard Assay for Duckweed UDPGA

Decarboxylase. . . .122

Identification of the Products Formed
by Parsley UDPGA Cyclase and Parsley
UDPGA Decarboxylase. . . . . . . . . .122
RESULTS 0 O O O O O O O O O O O O O O O O O . O O 123

DISTRIBUTION OF UDPGA OIOLASE IN THE LEAVES,
STEMS AND ROOTS 0F PARSLEY . . . . . . .123

PURIFICATION OF PARSLEY UDPGA CYCLASE. . . .125

EXtraCtion O O C O O O O O O O O O O O .125
Ammonium Sulfate . . . . . . . . . . . .125
Sephadex 6-100 . . . . . . . . . . . . .125
DEAE-Sephadexo o o o o o o o o o o o o 0125

PROPERTIES OF PARSLE! UDPGA CYCLASE. . . . .126
Linearity of the Parsley UDPGA Cyclase

Reaction . . . . .126
Factors Affecting the Stability of
Parsley UDPGA Cyclase. . . . .132

pH Optimum of Parsley UDPGA Cyclase. . .133
Affinity of Parsley UDPGA Cyclase for

UDP-glucuronic acid. . . .133
Effect of Various Ions on the Activity
of Parsley UDPGA Cyclase . . . . .133

Effect of Sulfhydryl Reagents on the
Activity of Parsley UDPGA Cyclase. . .133

Determination of Optimum NAD+ Con-
centration for the Formation of
UDP—apiose . . . .138

Inhibition of Parsley UDPGA Cyclase and
Parsley UDPGA Decarboxylase by

UDP-xylose . . . . .138
The Energy of Activation of Parsley
UDPGA Cyclase. . . .138

Fractionation of Parsley UDPGA Cyclase
and Parsley UDPGA Decarboxylase with
Ammonium Sulfate . . . . . . . . . . .143

vii

Page
PROPERTIES OF PARSLEY UDPGA DECARBOXYLASE. . . .145

pH Optimum of Parsley UDPGA Decarboxylase. 11.5
Determination of Optimum NAD Concentra-
tion for the Formation of UDP-xylose . .145

PROPERTIES OF DUCKWEED UDPGA CYCLASE AND
DUCKHEED UDPGA DECARBOXYLASE . . . . . . .150

Factors Affecting the Stability of

Duckweed UDPGA Cyclase . . . .150
Effect of Various Ions on the Activity
of Duckweed UDPGA Cyclase. . . . . .150

Effect of Sulfhydryl Reagents on the
Activity of Duckweed UDPGA Cyclase . . .153
The Energy of Activation of Duckweed

UDPGA cyclase. o o O O O 153
Inhibition of Duckweed UDPGA Cyclase
and Duckweed UDPGA Decarboxylase by
Five Uracil Compounds. . . . . . . . . .153
DISCUSSION 0 0 O O O 0 O O O O O 0 O O O O O 0 O 157

viii

PART 1.

1.

PART 2.

1.

PART 3.

1.

LIST OF TABLES

Small molecular weight compounds containing
D-apioseo o o o o e o e o o o o e o o o o e o

The ratio of 3H:1“c in the product from the
rgaction caigl zed by UDPGA cyclase with
[H]UDP—[U— Cfiglucuronic acid as the sub-

strate. . . . . . . . . . . . . . .

Products obtained on acid hydrolysis of
doub y labeled reaction product and their
1C ratios 0 O 0 O O O O O O O O O O O O 0

Products obtained on alkaline hydrolysis
8f ggubly labeled reaction and their
H 6 ratios 0 O O O O O O O O O O O O O O 0

Products obtained on incubation of doubly
labeled reaction product and “7GA' with
apiin synthase and their 3H:14C ratios. . . .

The stability of UDP-[U—lnc]apiose at -200
and at pH 5.5.. . . . . . . . . . . . . . .

Hydrolysis of UDP-[U—luc]apiose at pH 6.0
and 25° with time . . . . . . . . . . . . . .

Recri allization of a mixture of "apiin”
dfii C] ]Compound IV and a mixture of
'api n" acetate and [ CJCompound 13V acetate.
Distribution of apiin synthase in the leaves,

stems and roots of parsley. . . . . . . . . .

Purification of apiin synthase from parsley
leaves. . . . . . . . . . . . . . . . . . . .

Inhibition of apiin synthase by various
uracil compounds. . . . . . . . . . . . . . .

Fractionation of apiin synthase with ammonium
sulfate 0 0 O O O 0 O O 0 O O O O O O O O O 0

ix

Page

31
32
35
‘ 37
1+7

48

81
83
86
107

107

PART 4.
1.

Page

Distribution of UDPGA cyclase in the leaves,
stems and roots of parsley. . . . . . . . . . . . 124

Purification of UDPGA cyclase from parsley
leaves. . . . . . . . . . . . . . . . . . . . . . 127

Inhibition of parsley UDPGA decarboxylase
and parsley UDPGA cyclase by UDP-xylose . . . . . 143

Fractionation of parsley UDPGA decarboxylase
and parsley UDPGA cyclase with ammonium sulfate . 144

Inhibition of duckweed UDPGA decarboxylase
and duckweed UDPGA cyclase by five uracil
compounds 0 O O O O O O O O O O O O O O I O O O O 1 56

PART 1.

1.

PART 2.
1.

2.

3.

PART 3.

1.

LIST OF FIGURES

Transglycosylation of D-apiose from UDP-apiose.

The extent of hydrolysis of UDP-[U-luc]apiose
after 90 minutes at 25° as a function of pH . .

The rate of drolysis of UDP-{U-1 “CJapiose
to a-D-f [U- “Cfiapio-D-furanosyl cyclic-1: 2- P
and UMP at pH8 .0 and 1000 with time. . . . . .

The rate OEW h drolysis of UDP— [U- 1ucjapiose
to a-D-[ [U-4 Capio-D-furanosyl cyclic-l: 2- P
and UMP at pH 8. 0 and 250 with time . . . . . .

Two possible mechanisms for the acid hydro—
lysis of glycosidic linkages between
pyraHOSj-des 0 O O O O I O O O o O O 0 O O O O O

Torsional isomerism in cyclohexane. . . . . . .

Chromatography of some of the compounds pre-
sent in the standard assay mixture on
polyethylenimine impregnated paper with LiCl
solutions. . . . . . . . . . . . . . . . . . .

The elution pattern of UDPGA Cyclase and
apiin synthase from Sephadex 6—100 . . . . . .

Chromatography of apiin synthase and UDPGA
cyclase on DEAE-Sephadex . . . . . . . . . . .

Formation of apiin by apiin synthase as a
function of time with low concentrations
(<Km)°f.7GA.oooeeeeooooeeooo

Effect of apiin synthase concentration on
reaction velocity with low concentrations
(<Kfn) Of '7GA. O O O O O O O O O O O O O O O 0

Formation of apiin by apiin synthase as a
function of time with low concentrations
(<Km) Of UDP-apiose. o o e o e o o e o o o o 0

xi

Page

11

39

41

58
61

72

87

89

92

92

9h

10.

11.

12.

Effect of the amount of apiin synthase on
reaction velocity with low concentrations
(<Km) of UDP-apiose. . . . . . . . . . . . . .

Formation of apiin by apiin synthase as a
function of time with exogenous UDP-apiose . .

Effect of the concentration of apiin syn-
thase on reaction velocity with exogenous
UDP-apiose . . . . . . . . . . . . . . . . . .

Effect of pH on the reaction velocity of
apiin synthase . . . . . . . . . . . . . . . .

Effect of “VGA“ concentration.on the velocity
of the apiin synthase reaction . . . . . . . .

Effect of UDP-apiose concentration on the
velocity of the apiin synthase reaction. . . .

Formation of UDP-apiose by parsley UDPGA
cyclase as a function of time. . . . . . . . .

Effect of parsley UDPGA cyclase concentra-
tion on reaction velocity. . . . . . . . . . .

Effect of pH on the reaction rate of parsley
UDPGA cyclase. . . . . . . . . . . . . . . . .

Effect of UDP-glucuronic acid concentration
on the velocity of the parsley UDPGA cyclase
reaction 0 O O O O O O o O O O O O 0 O O O O O 0

Effect of NAD... concentration on parsley
UDPGA cyclase reaction velocity. . . . . . . .

Effect of temperature on parsley UDPGA
cyclase activity . . . . . . . . . . . . . . .

Effect of pH on the reaction rate of parsley
UDPGA decarboxylase. . . . . . . . . . . . . .

Effect of NAD+ concentration on parsley
UDPGA decarboxylase reaction velocity. . . . .

Effect of NAD+ on the stability of duckweed
UDPGA cyclase at 4°. . . . . . . . . . . . . .

Effect of temperature on duckweed UDPGA
cyclase activity . . . . . . . . . . . . . . .

xii

Page

94

96

96

99

101

103

128

130

134

136

139

141

146

148

151

154

ABBREVIATIONS

UDP-glucuronic acid

UDP-apiose

UDP-galacturonic acid

UDP-galactose

UDP-xylose
UDP-pentose
UMP

UDP

UTP

a-D-apio-D-furanosyl
cyclic-1:2-P

70A

NAD +
Tris-RC1

Tris-acetate

EDTA

dpm

uridine 5'-(a-D-glucopyranosylr
uronic acid pyrophosphate)

uridine 5'-(a-D-apio-D-furanosyl
pyrophosphate)

uridine 5'-(a-D-galactopyranosyl-
uronic acid pyrophosphate)

uridine 5'-(a-D-galactopyranosyl
pyrophosphate)

uridine 5'-(a—D-xylopyranosyl)
uridine 5'-(pentosyl pyrophosphate)
uridine 5'-phosphate

uridine 5'-diphosphate

uridine 5'-triphosphate

a-D-apio-D-furanosyl cyclic 1:2-
phosphate

7-(4'.5.7-trihydroxyf1avonyl)
O-B—D—glucopyranoside

nicotinamide adenine dinucleotide

tris (hydroxymethyl) amino
methane-H01 '

tris (hydroxymethyl) amino
methane-acetic acid

ethylene diamine tetraacetic acid

disintegrations per minute

The following trivial names are used for enzymes which have
been assigned systematic names: UDP-glucose pyrophosphorylase,
UTP: a-D-glucose-l-phosphate uridyltransferase (EC 2.7.7.9);
inorganic pyrophosphatase, pyrophosphate phosphohydrolase .
(EC 3.6.1.1) and UDP-glucose dehydrogenase, UDP-glucose: NAD"
oxidoreductase (Ec 1.1.1.22).

xiii

PART 1

CHARACTERIZATION AND BIOSYNTHESIS OF D-APIOSE AND

OF COMPOUNDS CONTAINING D-APIOSE

Occurrence of D-Apiose in Nature. -- At the turn of the
century the first branched-chain sugar was isolated (1).
Its discoverer, Vongerichten, named the unusual sugar
apiose (2,3). In the 55 years following the discovery of
D-apiose few studies of the sugar were reported. For many
years D-apiose was considered a rare sugar and a curiosity.
During the last 15 years considerable work on the chemistry
and biochemistry of D-apiose has appeared. The chemistry
of this branched-chain sugar and a number of its synthetic
derivatives has been investigated and reviewed elsewhere
(4,5). Extensive characterization has shown D-apiose to be
3-C-hydroxymethyl-aldehydo-D-glycero-tetrose (4). The 4
carbon atoms of D-apiose constituting part of the furanose
ring have been numbered 1 to 4 starting with the anomeric
carbon. With D-apiose in the furanose ring the carbon atom
of the free hydroxymethyl group is numbered 3' (Fig. 1).
Research published during the last 10 and particularly
the last 5 years has shown that D-apiose occurs in a large
number of different plants from many families. Duff (6)
investigated the hydrolysates from 175 plants belonging to
many families and found D-apiose in about 60% of the
plants. Duff discerned no taxonomic pattern in the plants
giving a negative test for the sugar. Therefore a random
sample of these plants was taken and re-examined more
carefully. All were found to contain traces of the sugar

encouraging him to suggest that all of the plants which

3

originally did not give a positive test for D-apiose did

in fact contain the sugar. Duff detected D-apiose by
hydrolyzing plant material with sulfuric acid, neutralizing
the hydrolysate and then chromatogramming the hydrolysate

on paper. The sugars present were visualized by spraying

the chromatogram with benzidine-trichloroacetic acid. Other
investigators have found D-apiose in various plants. Van
Beusekom (7), for example, did a study with 27 species of
monocotyledons and found D-apiose in 10 of them. Appreciable
amounts of D-apiose were found only in the hydrolysates of

Lemnaceae, and species of Hydrocharitaceae, Potamogetonaceae

 

 

and Zannichelliaceae which grow in the sea or in brackish

 

water.
No report of the occurrence of D-apiose outside of the
plant kingdom has yet appeared. A bacterial mutant has been

isolated which metabolizes D-apiose. Neal and Kindel (8)

 

isolated a mutant of Aerobacter aerogenes PRL-R3 which used
D-apiose as a sole source of carbon. A new enzyme was
isolated, D-apiose reductase, which catalyzes the NAD+-

dependent reduction of D—apiose to D-apiitol.

Characterization of Compounds Containing D—Apiose. -- In
Posidonia australis (70), Tilia Sp. (9), Zostera marina
(9,10,11), Lemna gibba (12), E. minor (6,12-16), g. nana

(6), g. pacifica and Phyllospadix (ll) D-apiose appears

 

 

to be primarily a component of polysaccharides. In the
case of g. marina (10,11), 2. Sp. (9), L. minor (6,13—16),

g. pacifica and Phyllospadix (11) a polysaccharlde fraction

 

or fractions were isolated and shown to contain D-apiose.
Beck (14,15) reported the isolation of two apiogalacturonans
from E. minor which contained 28% and 25% D-apiose. One

of these polysaccharides was reported to also contain D-
xylose and D-galactose. Hart and Kindel (16) isolated cell
walls of ;. EEEQE that contained 83% of the total D-apiose
present in the intact plant. The cell wall polysaccharides
containing D-apiose that were solubilized all appeared to
be of one general type, namely apiogalacturonans. Some of
the polysaccharides were further purified and characterized.
D—Apiose in these polysaccharides was found in a disac-
charide, O-B-D-apio-(?)-furanosyl-(1+3')-D-apiose. The
disaccharides were attached as side chains to the galac-
turonan (l7).

D-Apiose has been isolated glycosidically linked to
certain small molecular weight compounds. The small
molecular weight types of compounds that have been isolated
are isoflavone, phenolic, and flavone glucosides. The
parsley plant, Petroselinum cris um, contains at least two
D-apiose flavonoids, the principal one is apiin [7-(4',5,
7-trihydroxyflavonyl) 0-B-D-apio-D-furanosyl-(1+2)-B-D-glu-
copyranoside]. Apiin is also the principal flavonoid com-
pound in the stems and leaves of this plant (18).

Recently the results obtained by several workers have

led to the elucidation of the structure of apiin. After

methylation of apiin followed by hydrolysis, Hamming and
01118 (19) isolated a tri—0_methy1 derivative of U—upihnn
and 3,4,6-tri-O~methy1-D-g1ucose. From these data, they
concluded that the linkage between the two sugars was
from carbon atom 1 of D-apiose to carbon atom 2 of the
glucose moiety. Halyalker, Jones and Perry (20) in 1965
investigated the stereochemistry at carbon atoms 1 and
3 of the D—apiose moiety. They also methylated apiin
and isolated the resulting tri-O-methyl-D-apiose. The
compound was identical by paper, thin layer and gas-liquid
chromatography, optical rotation, and infrared spectra,
to chemically synthesized, 2,3,3'-tri-O-methyl-D-apio-D-
furanose. This shows that D-apiose-D-furanose is the naturally
occurring form, at least in parsley. Periodate oxidation
data also indicated that the hydroxyl groups at carbon
atoms 2 and 3 were in the gig configuration. From a know—
ledge of the molecular rotation of 7GA1 and apiin and by
application of the molecular rotation theory of Klyne (21)
these workers concluded that the stereochemistry of the
linkage between carbon atom l of D-apiose and carbon atom
2 of D-glucose was 8. Therefore, apiin is 7-(4',5,7-
trihydroxyflavonyl) O-B—D-apio-D-furanosyl-(14>2)-B-D-
glucopyranoside.

Besides apiin there is two and perhaps more fla-

vonoid glycosides in parsley which contain D—apiose.

 

1Abbreviations are listed on page xiii.

.coNHmopompmso no: mo: oasooaoo pseudo one .hsommepmaohno hogan an mpsosooaoo
ono>sam one woman on» no sofipdoamenooa unmodomDSm one mamhaopemn pace an mane eouapopomnmnoss

ozonhmouoonahsonnun u o:o>mamomH "ouchaaouoopahsonaum u esopmams

 

 

sopoonsu aznuspa>

 

maumaomonmfl ofimuonamn
. masonic: .m
can mnoaoo>muw .¢

mnoaooemuw azamd
use Away hoamhmm

 

 

 

monzmuzm,mdampawan

Aev summaho .M.

Ra .fiwumtfisflb m3

.mpmoombo caspmapomv.monm o
.w .mmoHanm monoomsmp
.maooona .2 .MHHHaoamzw

mammomppoz .msoomo on .

.maafinoa maaosus< .manaonom
mHHHom Mascaaho asswaop

.aosaxma wwl.a:moswwaflz MD

.soaoSpsoosoa mesoSpqmmaasw
.maoaooemuw .4 .aommano .m

 

 

 

 

 

 

Honosoamad>nm. Honozoahna>nm_ Edamonom
oso>mamomahxouehsad
.m.nuhxospo=u.: <nsacmzooam adamaooonmq
o:o>dam»x0Aeh:aha
-a.m..:-sxospoz-.m Hoanoomstno m medmownaoossno
oso>mHthonohs
nonpoauu.m..:..m naaoopzq < oeamoanaoo>mpo
eno>mam
saxonehsdaeum.m..: nanowao4 ssgaaaomdomaamoHomoosao.
“cocoaanopmo
ono>mamzw0Lehsanp mHo>Hm5Honoo no: monopmaxov
um.m..mnhxospo=u.e :Hposmoa sandaomOppom
oso>oam
-axonessdtaua.m..e edemmno< enwo<

 

oonoom usuam

Aoaez .ofipmsopmam.v Amsmz Hedswpev
s282391 oncomaw< omofln<un wnasacpaoo enzooaoo

 

omoaomun moanaopaoo monsooaoo unwaoz amasooaos Hanan "H mqm<a

Vongerichten (1,2,86) obtained from the leaves of 13. crispum
what he believed was a crude mixture of two glycosides. This
was Shown by hydrolyzing the mixture and obtaining two
aglycones which were partially characterized as apigenin
(4',5,7-trihydroxyf1avone) and desmotin (3'.5,7-trihydroxy-
u'-methoxyflavone)(4). Later Gupta and Seshadri (45) isolated
crystalline apiin from 2. crispum, moss—curled variety,

and obtained only one glycoside containing D-apiose. The
D-apiose containing compound was well characterized and
found to be apiin. It was not characterized by paper
chromatography. They found that some of the apiin isolated
had a melting point 100 lower than authentic apiin and was
otherwise identical in every other aspect tested. Shortly
thereafter Nordstrom, Swain and Hamblin (18) isolated
flavonoid glycosides containing D-apiose from parsley.

They extensively characterized the two glycosides they
obtained after separating them from each other by paper
chromatography. They identified the aglycone portions by
their RF values, color reactions, u.v. spectra in alcohol
and other tests as apigenin and luteolin (3',4',5,7-
tetrahydroxyflavone). They also extensively characterized
the glycosides before and after hydrolysis as apiin and the apio-
syl glucoside of luteolin. In 1966 Grisebach and Bilhuber
(60) partially characterized the two glycosides they
isolated from g. hortense Hoffm. (parsley). The major
component was identified as apiin and the minor component

was identified from its hydrolytic and Spectral properties

as the apiosyl glucoside of chrysoeriol. The apparent
discrepancy between these last two studies may be that

the flavonoids were isolated from different species of
parsley. Nordstrom and Swain and Hamblin (18) did not
identify the specie of parsley they tested although they

most probably used 2. crispum as they were reinvestigating
the work of Vongerichten (1,2,86) who investigated the
flavonoid glycosides from E. crispum. Therefore the flavone
glycosides containing D-apiose in.§. crispum are apiin
(18,45) and small amounts of apiosyl glucoside of luteolin (18).
The flavone glycosides containing D-apiose in 2.2hortense

are primarily apiin and some apiosyl glucoside of chrysoeriol
(60).

Recently, D-apiose was found in glycosidic combination
with a flavone in Digitalis purpurea (foxglove)(23). This
compound had all the components of apiin i.e. D-apiose,
D-gluoose and apigenin but it was not further characterized.
Apiin has also been isolated from the white flowers of
Chrysanthemum uliginosum, g. leucanthemum, g. maximum,
Bellis_perennis (English daisy), and Anthemis nobilis by
Wagner and Kirmayer (24). The yield of apiin was 1-3%.
Apiin was mixed with authentic apiin isolated from E.
crispum and then identified by its mixed melting point.

The isolated apiin and its acid hydrolysis products were
also identified by paper chromatography. Apiin was
isolated from the flewers of g. frutescens, Matricaria

chamomilla, fl. inodora, Centaurea scabiosa, Q. yanus,

 

9

Serratula coronata, and Eckinops gmelini. The apiin
isolated from these last seven plants was identified only
by paper chromatography in one solvent. Apiin has been
isolated from Vigig hirsuta (vetch) by Nakaoki, Morita,
Motosume, Hiraki and Takeuchi (25). Some workers in
India isolated apiin from an ethanol extract of one of

five species tested, Cuminum cyminum (78).

 

In addition to the parsley flavonoids, and the plants
containing apiin, D-apiose-containing flavonoid glucosides

have been found in Apium graveolens (celery) (22,76,79).

 

Farooq, Gupta, Kiamuddin, Rahman and Seshadri (22) isolated
from celery the two principal glucosides containing D-apiose,
graveolbioside A and B. Both have the structure of apio—
syl-glucosyl-7—aglycone. The aglycone portion of graveol-
bioside A is luteolin and that of graveolbioside B is
chrysoeriol, the 3' methyl ester of luteolin (See Table
I for structures).

Malhotra, Marti and Seshadri (26) reported the
occurrence of an apiosyl glucoside of biochanin-A, lanceolarin,

in the root—bark of the tree Dalbergia lanceolaria. Bio-

 

chaninin-A is an isoflavone.

D-Apiose has been identified glycosidically linked,
to D-glucose which in turn is glycosidically linked to a
non-flavonoid aglycone. Hattori and Imaseki (27) isolated
this compound from Viburnum furcatum Blume. They showed
that the new compound, furcatin, was p-vinylphenyl O-(?)-
D-apio-(?)-furanosyl-(l—)6)-B-D-glucopyranoside. Imaseki

and Yamamoto (28) later isolated a glycosidase from

10

y. furcatum which hydrolyzes furcatin into the aglycone
and O-(?)-D-apio-(?)-furanosyl-(T#6)—B-D-glucose. The
addition of alcohols to the mixture containing the gly-
cosidase and furcatin allowed the glycosidase to transfer
the apiosyl-glucose residue from furcatin to the alcohols,
thereby forming alkyl O-D-apio-furanosy1-(le6)-glucosides,

instead of transferring the residue to water.

In vivo Biosynthesis of D-Apiose.-- D-Apiose biosynthesis
has been investigated only in 2. crispum, g. gibbg G3, and
L. mingg. Enzymes isolated from these species were used
for the biosynthetic studies to be discussed in this thesis.
Enzymes isolated from several other plants have been found
to convert UDP-glucuronic acid into UDP-apiose as measured
by the formation of a-D-apio-D-furanosyl cyclic-1:2-P

after intramolecular phosphorylation (Part 2) and by the
release of D-apiose after acid hydrolysis of the incubation

mixture.2 These plants are Medicago sativa _I_... (alfalfa),

 

Trifolium repens L. (white clover) and Trifolium protense L.

 

(red clover).2

The biosynthetic studies to date support the principal
portions of a hypothesis proposed for the synthesis of
D-apiose and apiin by Grisebach and Dobereiner (30). They
postulated that D-apiose was formed as part of a sugar
nucleotide by a decarboxylation reaction which used a
nucleoside diphosphate glucuronic acid as the substrate.

They further postulated that a rearrangement reaction must

 

2Gustine, D. I... Watson, R. R. and Kindel, P. K., manuscript

in preparation.

11

Fig. l. Transglycosylation of D-apiose from UDP-apiose.
D—Apiose is transglycosylated to 76A, an acceptor molecule,
to form apiin. This is shown by the data presented in

Part 2 and 4. A pathway for the formation of glycosidically
linked D-apiose in polysaccharides from UDP-apiose has been

proposed by Roberts, Shah and Loewus (35). It is shown in
this figure.

12

32.38329...
:03 :oo

 

/

823 I do:
:0 o...

co .
_ :ofw

O

13

also occur in order to form the branched-chain D-apiose
from this intermediate. This latter suggestion is based
upon a hypothesis earlier presented by Baddiley, Blumson,
Di Girolano and Di Girolano (37) for the rearrangement of
an intermediate in the biosynthesis of L-rhamnose to form
streptose, another branch-chain sugar. The NDP-apiose
formed would be involved in the synthesis of polysaccharides
and glycosides such as spiin. Grisebach's hypothesis was
initially tested by carrying out isotope incorporation
experiments with parsley and duckweed. The experiments of
Grisebach as well as others, including those reported in
this thesis, have confirmed this hypothesis.

Grisebach and Dobereiner (30,31) gave radioactive pre-

[14C]Formate was admin-

cursors to young parsley shoots.
istered to the stems and although the percentage of radio-
activity in D-apiose was high, the nonspecific incorporation

of [IHCJformate into D-glucose showed that [luCJformate was

not functioning as a one carbon donor and that randomization

of 1“c had occurred. These above results and additional
experiments with [l-lucjacetate suggested that one and two

carbon units do not participate directly in the biosynthesis

of D-apiose. The specific activities of D-apiose and of

Deglucose from apiin were about equal when either D-[U-luCngucose
or D-[3,4-luC]glucose were precursors (30). Almost half

of the radicactivity from D-[3,4-1uc]glucose was found in

carbon atoms 3 and 3' of D-apiose, thus carbon atom 3'

comes from either carbon atom 3 or 4 of D-glucose. This

14

observation suggested that D-apiose does not occur in a free
(aldehyde) form during the biosynthes's of apiin. If it
did, randomization of 1“C activity between carbon atoms

3' and 4 of D-apiose would be expected to occur. To further
test his hypothesis, Grisebach fed D-[6—1uc,4-3h]glucose

to parsley shoots. About 70% of the tritium was located

at carbon atom 3' of D-apiose. This result could be
explained by the assumption that hydrogen is removed from
carbon atom 4 of UDP—glucuronic acid by coenzyme—enzyme
complex and that the same coenzyme-enzyme is used for the
reduction of the transient carbonium ion Grisebach postulates
is formed during the rearrangement step.

Additional support for his own hypothesis was obtained
by Grisebach and Sandermann in experiments with.D-[U-luC]
glucuronic acid (32). The results showed a very large
incorporation of 140 from D—glucuronic acid into the D-apiose
(92%) and very little incorporation into the D—glucose (5%)
and the apigenin (3%) of apiin. In contrast, with
D-[6-]uc]glucuronic acid as precursor, the incorporation
into D-dpiosc is much lower. This is to be expected since
carbon atom six was postulated to be lost as [1403002.

This would be true even if refixation of [MCJCO2 by photo-
synthesis occurred.

Hendicino and Picken (13,33) independently studied tie
biosynthesis of D-apiose in E. crispum and in.L. miggg.

They compared the relative extent of incorporation of
various precursors into D-apiose. On the basis of their

r11;
L

findings that [Z-IuCJacetate, CH3]methionine and

15

[3-1ucjserine were incorporated into D-apiose, Mendicino
and Picken suggested that the transfer of a one carbon
group might be involved in D-apiose biosynthesis. In
these experiments there was incorporation of 140 into D-
glucose. The potential carbon atom 1 donor compounds were
metabolized and that 146 had randomized in this experiment.
This cast doubt upon the conclusion that one carbon units
participate in the biosynthesis of D-apiose. If they did

140 from one carbon unit

participate the incorporation of
precursors into the branched-chain sugar should have
occurred with little incorporation into other sugars. The
work of Grisebach, reviewed above, showed that a one car-
bon unit is not important in the biosynthesis of D—apiose

as does the work reported in this thesis. Mendicino and
Picken also found that carbon atoms 1 through 5 of D-glucose
were converted into D-apiose.

In studies with whole parsley plants in which various
1“Cu-containing compounds were metabolized, Mendicino and
Picken found that the specific activity of the apiin iso—
lated was ten times higher in the roots than the leaves.

Since the greatest concentration of apiin is in the leaves
they suggested that it was synthesized in the roots and trans-
ported to the leaves. In contrast to these findings and
conclusions, Grisebach (34) and Grisebach and Dobereiner (31)
have reported that little if any difference existed in the
incorporation of D-glucose into D-apiose of apiin in parsley

with or without roots. Mendicino and Picken (13) also

16

found that free D—apiose was not utilized by the parsley
plant. When D-[U—lu01apiose was administered to the plant
negligible incorporation into apiin was observed.

Roberts, Shah and Loewus (35) studied D-apiose
biosynthesis in.§. crispum and L. gigba G3 with isotopic
incorporation experiments utilizing mygsinositol. Elg-
_inositol is an effective precursor of D-glucuronic acid
in higher plants and is rapidly and specifically utilized
in the biosynthesis of uronic acid residues (71,72).
[Z-luC]flyg-inositol is oxidized in plant to D-[5-14Cngu-
curonic acid. Roberts, Shah and Loewus (35) observed that
[2-14C]myg-inositol functions in the biosynthesis of
D-apiose-and D-xylose-containing polysaccharides of L.

‘gibba G3' They also found that leaves of E. crispum
utilized [Z-BHJm o-inositol for the biosynthesis of D-apiose
of apiin. Apiin isolated from 2. crispum after the isotope
incorporation study contained 3H only in the D-apiose moiety.
Periodate oxidation of this D—apiose showed that 94% of the
3H in D—apiose was located in carbon atoms 3' and 4.
Experiments by other workers discussed above showed that
radioactivity located at carbon atoms 3 or 4 of D—glucose
becomes carbon atoms 3 or 3' of D-apiose. The experiments
of Roberts, Shah and Loewus can then be interpreted as
showing that carbon atom 4 of D—apiose obtained from
[2-3H]m‘o-inositol is equivalent to D-CS—thglucuronic
acid. The results they obtained independently provided

Idirect evidence for a pathway of D—apiosc biosynthesis

involving D-glucuronic acid metabolism.

17

When L-[l-lucjarabinose was supplied to g. gibbg 63’
both L—arabinosyl and D—xylosyl units of cell wall poly-
saccharides contained 14C, but no 1“C was found in Deapiose
(35). Some plants contain an epimerase which can convert
L-arabinose to D-xylose (UDP-L-arabinose aUDP-D-xylose).
The work of Roberts, Shah and Loewus snowaithat there was
no mechanism for converting D-xylose or L—arabinose to
D-apiose in p. 33323 G3.

The in X212 biosynthetic studies of D-apiose in
p. 21223 by Beck and Kandler (36) confirm those already
described above. They studied the patterns of 1&3
incorporation into D—glucose, D—xylose and D-apiose obtained
after administering D-[U-lucjglucose, D-[l-lucjglucosc,
D-[Z—luCngucose and D—[6-1uC]g1ucose. They found very

significant incorporation of 1LLC into the three sugars

studied. The 14C of [luCHBJmethionine, however, was in-
corporated only slightly into these sugars. Furthermore,
D—apiose did not contain more 140 than D-xylose as would
have been expected if the addition of a one carbon unit

was an important step in the biosynthesis of D—apioso

but not in D—xylose biosynthesis. D-Xylose has been shown
to be synthesized from UDP—glucuronic acid in plants by

a decarboxylation reaction (34,71,77). lhxflcand Kandler (12)iso-
lated D-apiose, D—xylose and D-glucose from g. mlngg polysac-
charides and determined the distribution of radioactivity

in each carbon atom. The results showed that both D-apiose

and D_xylose were derived from D-glucose. They (12) also

18

I
found that during photosynthesis l*COZ was incorporated into

D-apiose and D-xylose. The incorporation of lLLCOZ into
D-xylose and D-apiose during the first two hours was
very similar and yet different from other sugars synthesized.
He therefore suggested that the biosynthesis of D-apiose
and D-xylose was part of the same biosynthetic pathway.

Very recently in gitrg studies of the biosynthesis
of D-apiose,'UDP-apiose and apiin have been reported.

This literature will be reviewed and cited as necessary

in the rest of the dissertation.

PART 2

BIOSYNTHESIS AND CHARACTERIZATION OF UDP-APIOSE

19

20

INTRODUCTION

Cell-free systems from Lgmng mingg (duckweed) and
Petroselinin crispum (parsley) have been described which
carry out the biosynthesis of D—apiose (3—C-hydroxymethyl-
aldehydo-D-glzcero-tetrose) from UDP-glucuronic acid
(38, 39). Although not fully characterized the product
of the reaction was reported to be UDP-apiose (38, 40).
UDP-apiose was not found when assay mixtures were incubated,
subjected to a brief heat treatment and chromatographed
under slightly alkaline conditions (39). Under these
conditions a single compound containing D-apiose was
formed which proved to be neither UDP-apiose, a D-apiose
phosphate monoester nor free D-apiose. This compound has
been identified as a-D-apio-D-furanosyl cyclic-1:2-P and
has been shown to be formed non-enzymatically from UDP-
apiose1 under exceedingly mild conditions (42). The

tentative identification of this compound has recently

been reported by Sandermann and Grisebach (40).

 

1Although not established, the configuration at carbon atom
3 of the D-apiose moiety in UDP-apiose and in a-D-apio-D-
furanosyl cyclic-1:2-P is assumed to be D as in apiin (20).
The configuration at carbon atom 1 of a-D-apio-D-furanosyl
cyclic—1:2-P must be a in order for this sugar cyclic-1:2-
phosphate to exist. This was shown clearly by using
Framework Molecular Models of the above compounds made

from components manufactured to scale.

21

Apparently a single enzyme catalyzes the conversion of
UDP-glucuronic acid to UDP-apiose and C022. This enzyme has
been purified about 55-fold from L. mingg and has been given
the common name UDPGA cyclasez. Only through the use of
purified UDPGA cyclase has it been possible to prepare UDP-
apiose of sufficiently high radiochemical purity to permit
complete characterization of this compound. A preliminary
account of this work has been presented (42).

The purified UGPGA cyclase contains UDPGA decarboxylase.
This will be indicated by characterization of one of the
two sugar nucleotides formed from UDP-glucuronic acid,
UDP-xylose.

In this Part, two things will be described: 1) a method
for the enzymatic synthesis of UDP-apiose from UDP-glucuronic
acid and 2) an extensive characterization of one of the

products of the UDPGA cyclase reaction and its identification
as UDP-apiose. The other product presumably is C02.

MATERIALS AND METHODS
Materials. —- UDP-glucuronic acid, UDP-xylose, UMP, UDP,
UTP, and NAD+ were obtained from P—L Biochemicals, Inc.
Dquiose was isolated as described elsewhere (8). Framework
Molecular Models were manufactured by Prentice-Hall, Inc.
UDPfU-luc]glucuronic acid (233 mCi/mmole) and the dipotassium
salt of a-D-[U-lucjglucose-l-phosphate (194 mCi/mmole) were

 

2Gustine, D. L., Watson, R. R. and Kindel, P. K., manuscript
in preparation

22

obtained from New England Nuclear Corp. The tetralithium
salt of [5-3HJUTP (12.1 Ci/mmole) was purchased from
Schwarz/Mann. Crystalline calf liver UDP-glucose pyrophos-
phorylase (93 units/mg protein) was a gift from Dr. J. A.
Boezi. One unit of enzyme will form 1 umole of d-D—glucose
l-phosphate (NADPH) per minute at 25° and pH 7.8 (61).
Crystalline yeast inorganic pyrophosphatase, Type II (approx.
250 units/mg protein) was purchased from Sigma Chemical Co.
One unit of enzyme will liberate 1 umole of inorganic ortho-
phosphate per minute at 25° and pH 7.2. Partial purified
calf liver UDPG dehydrogenase was a gift from Dr. R. G. Hansen.
One unit of enzyme will oxidize l;umole of UDP-glucose per
minute at 25° and pH 8.7. Purified Escherichia 921; alkaline
phosphatase, Code BAPC (35 units/mg protein) was purchased
from Worthington Biochemical Corp. One unit of enzyme will
liberate lyumole of p-nitrophenol per minute at 25°. '7GA'
was isolated from crystalline “apiin“ by acid hydrolysis,
collected by suction filtration, washed thoroughly with
distilled water and recrystallized from 95% aqueous ethanol.
Crystalline 'apiin' was isolated from parsley seeds by the
method of Gupta and Seshadri (45). Although crystalline,

it was a mixture, hence the quotation marks (8).

General Methods. -- Electrophoresis of UDP-apiose, UDP-
xylose, D-apiose and a-D-apio-D-furanosyl cyclic-1:2-P

was carried out on unwashed Whatman No. 3MM paper at 4°.
The buffer employed was 0.2 M ammonium acetate, pH 5.8.

Electrophoresis was performed at 30 volts per cm with a

23

Pherograph Original Frankfurt, Type 64 (distributed by
Brinkmann Instruments, Inc., Westbury, N.Y.). Radio-
activity was detected on chromatograms with a Packard
radiochromatogram scanner, Model 7201 (Packard Instrument
Co.). All other radioactivity measurements were made with
a Packard Tri-Carb liquid scintillation counter, Model
3310, employing either (A) a scintillation solution made
as described by Bray (84) or (B) 2,5-bis-[2-(5-tggt-
butylbenzoxazolyl)]-thiophene in reagent grade toluene

(4 g/l.). When using solution B the radioactive compounds
were counted directly on paper by cutting out the appro—
priate portions of the chromatogram and completely immersing
them in this solution. The counting efficiencies with
solutions A and B were 79 and 60$, respectively. Duckweed
was cultured as described by Kindel and watson3.

Paper Chromatography. -- Paper chromatography was by the
descending technique and was carried out with washed

Whatman No. 3MM paper at 4° unless otherwise stated. The
paper was washed with 0.1 M citric acid followed by distilled
water. Unwashed Whatman No. 3MM paper impregnated with 2.5%
neutralized polyethylenimine was used with solvent F (46).
The following solvents were employed: (A) acetone-ethanol-
2-propanol-H20 (3:1:1:2, v/v); (B) l-propanol—ethyl acetate-
H20 (7:1:2, v/v); (C) acetone-ethanol-Z-propanol-0.05 M
borate buffer, pH 8.6, (3:1:1:2, v/v) (47); (D) 75% EQueous

 

3Kindel, P. K. and Watson, R. R., manuscript in preparation

24

ethanol; (E) ethyl acetate-HZO-acetic acid-formic acid
(18:4:3:l, v/v); (F) 0.3 M LiCl; (G) 95$ aqueous ethanol-
1.0 M ammonium acetate, pH 7.5, (7:3, v/v), (H) 1-butanol-
acetic acid-H20 (4:1:5, v/v, upper phase); (1) ethanol-
methyl ethyl ketone-0.5 M morpholinium tetraborate, pH
8.6, in 0.01 M EDTA (70:20:30, v/v)(90).

Preparation and Assay of UDPGA Cyclase and Apiin Synthase. --
UDPGA cyclase was purified from duckweed. The purification
and assay procedures were modifications2 of those described
by Gustine (41). One unit of enzyme has been defined as

that amount which will form 1 umole of UDP-apiose per min-
ute under the conditions of the standard assay. Apiin
synthase was purified from parsley leaves and assayed by
procedures described in Part 3. One unit of enzyme has

been defined as that amount which will form 1 umole of apiin
per minute under the conditions of the assay.

Preparation of r BHWUDP-[U-M'C'Iglucuronic Acid. .. [3RJUDP-
[U-lucjglucose was prepared from [BHJUTP and a-D-[U-14Cngu-
cose l-phosphate by a modification of the procedure of
Fitzgerald and Ebner (51) and was converted to [BHJUDP-
[U-lucjglucuronic acid by the procedure of Castanera and

Hassid (52). The incubation mixture contained 1.4 nmoles

of [3H]UTP (16.9 uCi), 10.3 nmoles of a-D—[U-lucjglucose-
l-phosphate (2.0 uCi), 7 nmoles of UTP, 1.8 nmoles of magnesium

25

acetate, 25 pmo1es of glycine-NaOH buffer, pH 8.7, 5.4 units
of UDP-glucose pyrophosphorylase and 1.2 units of inorganic
pyrophosphatase in a volume of 0.48 ml. The mixture was
incubated 60 minutes at 25° and then 3 nmoles of UTP and
2.7 units of UDPG pyrophorylase were added making the final
volume 0.5 ml. The reaction was incubated an additional 30
minutes at 25° and then heated for 1 minute at 100°. To
the cooled solution were added 0.4 umole of NAD... and 0.05
units of UDP-glucose dehydrogenase. After incubating for

1 hour at 25° the mixture was heated for 1 minute at 100°.
To establish that [3H]UDP-[U-1°C]glucuronic acid had been
formed and to determine its purity, small portions of the
preparation were examined by paper chromatography with
solvents B, c, F and G. Only [3R]UDP-[U-1“c]glucuronlc
acid and a small amount of d-D-[U-lucjglucose were observed.
The latter compound had been present in the commercial
a-D—[U-lucjglucose l-phosphate. It does not contaminate
the final product, [3H]UDP-[U-1°C]apiose, since it is
separated from the UDP-pentoses during paper chromato—
graphy in solvent A. None of the radioactive substrates
remained since all the [BHJUTP had been converted to pro-
duct when it was the limiting substrate and when the
subsequent addition of UTP made a-D—[U-lucjglucose-l-
phosphate limiting, this radioactive sugar phosphate was
all converted to product. [3H]UDP-[U-1u0]glucuronic acid
was converted to [3H]UDP-[U-1°C]apiose as described in

the Results.

26

RESULTS
Identification of the Product of the Reaction Catalzggg
prDPGA Cyclase. -- The assay procedure which has been
used to detect UDPGA cyclase involved incubating the enzyme
with UDP-[U-lucjglucuronic acid and NAD‘+ in sodium phosphate
buffer, pH 8.0, for 5 minutes at 25°, heating the assay
mixture briefly at 100° and then chromatographing the mixture
in solvent G (39). With this procedure Gustine and Kindel
(39) found that all the D-[U-lucjapiose that was synthesized
by duckweed UDPGA cyclase was found in a compound they called
Compound III. They showed that Compound III was neither
UDP-[U-lucjapiose, a D—[U-lucjapiose phosphate monoester
nor free D-[U-1“c]aplose. Later Kindel3 showed that if
the assay mixture was not heated but immediately cooled
to 4°, spotted and chromatographed in solvent G at this
same temperature, little Compound III was obtained. Instead
a D-[U-1°C]apiose-containing compound with an RF of UDP-
xylose was obtained. He also showed that this new compound
was converted to Compound III when heated at 1000 in pH
8.0 buffer. These results showed that the true product of
the reaction was not Compound III but was a compound which
migrated with an RF of UDP-xylose. He also showed that the
reaction product could be converted to Compound III. Similar
results were obtained with enzyme isolated from parsley and
are described in Part 4. Compound III was subsequently
identified as a-D-[U-lncjapio-D-furanosyl cyclic-l:2-P3.

27

I had previously observed that parsley UDPGA cyclase
was necessary to synthesize apiin. Presumably it formed
an unidentified D-apiose donor molecule from UDP-glucuronic
acid. The reaction synthesizing apiin also required the
presence of '7GA' and was catalyzed by parsley apiin syn—
thase. I believed that this D-apiose donor molecule could
be UDP-apiose which had been postulated to function as a
D-apiose donor molecule in this glycosidic reaction (34).

A method for preparing the D—[U-1°C]apiose containing
reaction product enzymatically from UDP-[U—1°C]glucuronic
acid was developed and is described.

An incubation mixture containing 11 nmoles of
UDP-[U-lucjglucuronic acid (5,104,000 dpm) 0.4 umole of
NAD+ 10 nmoles of sodium phosphate buffer, pH 8.0, 20
nmole of EDTA, and 14.5 milliunits of duckweed UDPGA cyclase
(2.4 mg) purified through the DEAE-Sephadex step (41) in a
total volume of 0.4 ml was prepared. The mixture was in-
cubated 15 minutes at 25° and then cooled to 4°. The remaining
steps were carried out at 4°. The incubation mixture was
streaked on washed Whatman No. 3MM paper over 5 cm and the
paper was developed in solvent A for 16 hours. Without
having been dried, the chromatogram was cut and that por-
tion with an RF of UDP-xylose was eluted with 3.2 m1 of
solvent A. This solution had a pH of approximately 6.0
and contained the reaction product. It was concentrated
in 0.5 hour to 1.0% of the starting volume by rotary

evaporation under reduced pressure and at 25°. From

28

radioactivity measurements, the yield Of reaction product
plus UDP-[U-1°C]xylose in this particular preparation was
79%. Paper chromatography of aliquots of the solution
containing the reaction product in solvents B and C at 40
together with an examination of the original chromatogram
showed that 10% of the starting radioactivity was present

in a-D-[U—luc]apio-D-furanosyl cyclic-l:2-P (Compound III)
of which 3% had separated from the UDP-[U-1°C]apiose during
chromatography in solvent A and 7% was formed during elution
of the reaction product. During elution of the reaction
product D-[U-lucjapiose was formed to the extent of 5% from
UDP-[U-1°C]apiose. No UDP-[U-1°C]glucuronic acid was
detected. Where appropriate, percent figures are corrected
to take into account the loss of carbon atom 6 of glucuronic
acid as [1°CJCOZ. No attempt was made to account for the
remainder of the starting radioactivity. Chromatography

in solvent A separated the reaction product from NAD+o

An aliquot of the solution containing the reaction pro-

duct was made 0.05 M with respect to sodium phosphate
buffer, pH 8.0, heated for 4 minutes at 100° and the hy-
drolysis products chromatographed on paper in solvent G.
Measuring the amount of a-D—[U-1°C]apiO-D-furanosyl
cyclic-1:2-P formed showed that 54% of the radioactivity

was in the reaction product and 34% was in UDP-[U-lucjxylose.
The remainder was in the compounds mentioned above. An
‘alternative procedure for determining the amount of reaction

product present was to adjust an aliquot Of the solution to

29

pH 2 with HCl, heat the solution for 15 minutes at 100°,
chromatograph the hydrolysis products on paper in solvent

140 content of D-[U-1°C]apiose and

E and then measure the
D-[U-lucjxylose formed. Both procedures gave comparable
results. The reaction product was stored at -20° in 50%
aqueous ethanol. Reaction product prepared in this way
was used for the characterization experiments. In this
paper reaction product refers to the product of the reaction
catalyzed by UDPGA cyclase, i.e., UDP—apiose. The UDP-[U-
1L’C‘jxylose present in the reaction product is formed in the
reaction catalyzed by UDPGA decarboxylase. Reaction pro-
duct prepared in this manner was used in all of the following
experiments. Rechromatography of the reaction product in
solvent B proceeded with less than 2% degradation of the
reaction product. After chromatography in solvent B at 4°
for 24 hours UDP—[U-1°C]apiose was eluted with 0.001 M sodium
phosphate, pH 6.0 which was 50% ethanol by volume. This
step removed the UMP, a-D-[U-1°C]apiO-D—furanosyl cyclic-
1:2-P and D-[U-1“c]apiose which contaminates the reaction
product and UDP-[U-lncjxylose after chromatography in solvent
A as described above. Elution with sodium phosphate buffer
prevented the.degradation of the reaction product to D-apiose
and D-apio-D-furanosyl cyclic-1:2-P. Degradation often occurred
when the reaction product was eluted with solvent A.

The reaction product and UDP-[U-lucjxylose obtained
migrated on paper at 4° as a single peak in solvents A-D.

In each solvent they had the same RF as UDP-xylose. When

30

subjected to paper electrophoresis at pH 5.8, the mobility
of the reaction product was the same as UDP-xyloSe. An
essentially identical preparation of the reaction product
as that described above was carried out except that
[3H]UDP-[U-140]glucuronic acid was used. The reaction pro-
duct was isolated and its 3H:1°C ratio was that expected
if the D-[U-lucjapiose and D-[U-1°C]xylose each remained
attached to the uridine and one carbon atom of the D-[U-

1Mcjglucuronic acid moiety was lost as [14

CJCOZ.

Doubly labeled reaction product (0.07 nmoles) in 0.25
ml of 0.01 N HCl was heated for 15 minutes at 100°. The
hydrolysis mixture,UDP, UMP, D-xylose and D-apiose were
spotted separately on paper and the paper was developed in
solvent C at 4°. Four radioactive hydrolysis products were
obtained. Starting from the origin, these compounds were
numbered 1 to 4. The compounds were eluted from the chro-
matogram with water at 4°. Compounds 3 and 4 were eluted
together.

A portion of each of the products from acid hydrolysis
was used for determining the isotope content and 3H:1°C
ratio. The remainder was used for identification. The
results are seen in Table II. In solvent C compounds 1

and 2 had the same R as UDP and UMP, respectively. In

F
solvents F and G compounds 1 and 2 migrated as single peaks
and co-chromatographed with UDP and UMP, respectively. On

this basis they were identified as [3RJUDP and [3R]UhP.

31

TABLE I: The ratio of 3H:1l+C in the product from the reaction
catalyzed by UDPGA cyclase with [3H]UDP-[U-lu0]glucuronic

acid as the substrate*.

 

 

Compound Ratio of 3mmc
3 14

[ HJUDP-[U— nglucuronic acid 8.91

Reaction product 10.6 (observed)

10.4 (calculated)u

 

*Thiuincubation mixture contained 1.57 nmole of [3H]UDP-

[U- nglucuronic acid, 0.55 umole of NAD‘, 25.5 nmoles of
sodium phosphate buffer, pH 8.0, 0.055 umole of EDTA, 10
nmoles of B-mercapto-ethanol, 3 milliunits of duckweed UDPGA
cyclase (2.0 mg) purified through the DEAE-Sephadex step in

a total volume 08 0.55 ml. The mixture was incubated for

25 minutes at 25 and then the reaction product was isolated
as described in the Regults. After paper chromatography in
sglvent A, n3 (i-D-[U--l C]apio-D-furanosy1 cyclic -l:2-P or

[ HJUDP-[U-l nglucuronic acid could be detected in the
reaction product. A It contamination of either would have
been detected. Acid hydrolysis of the reaction produ t,

as described in the Results, showed that fix of the 1 C was

in the reaction product while 29$ was in E HJUDP-[U-lucjxylose.
A portion of the reaction product was rechromatographed on
paper in solvent C at 4° and the paper was scannefi. The
reaction product (with contaminatifig [3H]UDP-[U-1 ijylose)
was eluted with water and its 3H: C ratio was determined.
This rechromatography was necessary in order to remove con-
taminating [3H]UMP. This latter compound presumably was
formed by alkaline hydrolysis during the preparation of the
reaction product. or the total tritium present in the solution
containing the reaction product 7.8% was in [3HJUMP. (A por-
tion of the [BHJUDP-[U—l nglucuronic acid was chromatographed
in solviflt B in order to separate it from the contaminating

a-D-[U- nglucose. After the paper was scanned the
EBHJUDP- U- nglucuronic acid was eluted with water and
he 3H:1 0 ratio was determined.

**Calculated assuming 1/6 of the 140 in [BHJUDP-[U-lucjglu-
curonic acid was released as [ CJCOZ.

32
TABLE 11: Products obtained on acid hydrolysis of doubly
labeled reaction product and their 3H:1uC ratios.

 

14

 

Compound Ratio of 3H C
3H:1“C (fi) (5)
Doubly labeled reaction product‘ 10.6 100 100
Products from acid hydrolysis**
UDP at origin 549 21.2 0.4
UDP 1028 64.7 0.6
UMP 109 10.3 1.0
D-apiose and D-xylose 0.005 0.1 95.0
96.3 97.0

 

*The doubly labeled reaction product was the ame as that used
in Table I. The starting percent of 33 and 1 C is based on that
ffiom both the reaction product and the contaminating [BHJ-[U-

1 ijylose.

**The products of hydrolysis were identified as described in the
Results. The [3H]UMP present in the reaction product was taken
into account in calculating the values for UMP.

 

Compounds 3 and 4 were only partially resolved by solvent C.
Both authentic D-apiose and D-xylose had migrated to the same
area as these two compounds. Rechromatography in solvent E
with authentic D-apiose and D-xylose showed that compound 3
was D-[U-lncjxylose and compound 4 was D-[U-lncjapiose. Fre-
quently, some radioactive material remained at the origin or
extended from the origin to compound 1. Co-chromatography in
solvents F and 6 showed that this material was [3HJUDP. The
amount of this material could be reduced, and in some cases
was completely eliminated, by adding carrier UDP to the hydro-

lysate before chromatography in solvent C. Control experiments

33

were run in which the reaction product in water at pH 5-6 was
kept at 4° for the duration of the experiment and then was
chromatographed in solvent C. No free D-[U-luCJapiose or
D-[U-lucjxylose could be detected in the controls and less
than 2% of the 3H migrated with the RF of [3HJUMP. These
data established that the products from the acid hydrolysis
of the reaction product were [3H]UDP and D-[U-lncjapiose.
These data and others presented earlier established that
the reaction product contained [3HJUDP-[U-lucjxylose since
[BHJUDP and D-[U-luc]xylose were obtained after acid hydrolysis.
UDPGA decarboxylase which contaminates the UDPGA cyclase would
account for the formation of [BHJUDP-[U-lucjxylose.

Doubly labeled reaction product (0.077 nmoles) in 0.1
ml of 0.07 M sodium phosphate buffer, pH 8.0, was heated
for 5 minutes at 100°. This mixture, UMP and UDP-xylose
were spotted separately on paper and the paper was developed
in solvent G at 22° for 48 hours. Three radioactive products
were obtained, one of which was allowed to run off the paper.
Chromatography for 48 hours was needed in order to separate
[3HJUMP and [3HJUDP-[U-1“c]xylose. The compounds, including
[BHJUDP-[U-lucjxylose were eluted from the chromatogram with
water at 4°. A portion of each was used for determining the
isotope content and the 311:140 ratio. The remainder was used
for identification. The results are seen in Table III. There
was no detectable breakdown of [3H]UDP-[U-1u0]xylose as
evidenced by 1t8 complete recovery and unchanged 3H:1“c ratio.
The two compounds remaining on the paper had the same RF values

.3“

as the UMP and UDP-xylose standards. In solvent F one compound,
[BHJUMP, migrated as a single radioactive peak and co-chro-
matographed with UMP. In solvent F [BHJUDP-[U-luc]xylose
migrated as a single radioactive peak and co-chromatographed
with UDP-xylose. Rechromatography of the material which ran
off the chromatogram showed that it migrated as a single peak
in solvent B and had the same RF as a-D-[U-lnCJapio-D-
furanosyl cyclic-1:2-P. When this compound was hydrolyzed
with acid by the procedure previously used with Compound III
(39) and the ethanol-soluble material chromatographed on
paper in solvent E, a single radioactive peak co-chromato-
graphing with D-apiose was obtained. The controls were the
same as those described above for the acid hydrolysis of

the reaction product. No [BHJUMP and no d-D-[U-lucjapio-
D-furanosyl cyclic-1:2-P were detected on the chromatogram.
These data established that the products from the alkaline
hydrolysis of the reaction product were [BHJUMP and
a-D-[U-luCJapio-D-furanosyl-cyclic-l:2-P. They further showed
that [BHJUDP-[U-lucjxylose was not degraded.

The reaction product acted as a D-[U-lucjapiose donor.
However, [BHJUDPQ[U-lu0]xylose and a-D-[U-lucjapio-D-furanosy1
cyclic-1:2-P did not act as D-[U-lucjxylose and D-[U-luc]-
apiose donors in the reaction catalyzed by apiin synthase
using '7GA' as the accepting compound. These results are
shown by the data in Table IV. When doubly labeled reaction
product and '7GA' were incubated together with apiin synthase,
two products were obtained. They had the same BF values in

35

TABLE III: Products obtained on alkaline hydrolysis of
doubly labeled reaction product and their 3H:1I+C ratios.

 

 

 

comm 333324? at t?
Doubly labeled reaction product* 10.6 100 100
Products from alkaline hydrolysis**

UMP 418.0 61.6 1.5

a-D-apio-D—furanosyl cyclic-1:2-P 0.01 0.1 57.7
Other compounds

UDP-xylose 10.2 35.4 36.2

97.1 95.4

 

*The doubly labeled reaction product was the ame as that used
in Table I. The starting percent of 3H and 1 C is based on
that from be h the reaction product and the contaminating
[BHJUDP-[U-1 cjxylose.

**The products of hfidrolysis were identified as described in
the Results. The [ H]UMP present in the reaction product
was taken into account in calculating the values for UMP.

 

solvent G as the UDP and apiin which had been spotted separately
on the paper. In solvent F the material with an RF of UDP
migrated as a single peak and co-chromatographed with authentic
UDP. In solvent H the material with an RF of apiin migrated

as a single peak and co-chromatographed with apiin. When the
product with an RF of apiin was hydrolyzed with acid by the
procedure previously used with Compound III (39) and the
ethanol-soluble material chromatographed on paper in solvent

E, a single radioactive peak was obtained which ,co-chromatographed

with D—apiose. These data identified UDP and apiin as the products

36

of the reaction catalyzed by apiin synthase. Two other
compounds were present on the original chromatogram in
addition to [BHJUDP-[U-lucjxylose. These two compounds
had the same RF values as UMP and a—D-apio-D-furanosyl
cyclic-1:2-P. In solvent C after 18 hours at 22°, UMP
and UDP-xylose have approximately the same RF. When the
material migrating with an RF of UMP and UDP-xylose was
rechromatographed on paper in solvent F for 4 hours, part
of the radioactivity co-chromatographed with UMP while the
remainder co-chromatographed with UDP-xylose. In solvent
B the material with an RF of d-D-apio-D-furanosyl cyclic-
1:2-P migrated as a single peak and chromatographed with
the same RF as the authentic compound. When this radio-
active compound was hydrolyzed with acid by the procedure
previously used with Compound III (39) and the ethanol-
soluble material chromatographed on paper in solvent E,

a single radioactive peak with the same RF as D-apiose

was obtained. These data established that the two compounds
were [3H]UMP and d-D-[U-lucjapio-D-furanosyl cyclic-1:2-P.
Alkaline hydrolysis of unreacted doubly—labeled reaction
product accounts for the formation of these two compounds.
The following three control experiments were run: (1) with
heated apiin synthase; (2) with heated reaction product;
(3) with both heated apiin synthase and heated reaction
product. The incubation mixtures for the controls were
identical to those described in Table IV except that apiin

synthase or doubly-labeled reaction product or both were
heated for 2 minutes at 100° and pH 8.0 before being added

3?

TABLE IV: Products obtained on incubation of doubly labeled
reaction product and '7GA' with apiin synthase and their

3 14 *

H: C ratios

 

 

 

Compound rigging!“ (2) 3;;

Reactant

Doubly labeled reaction product** 10.6 ' 100 100

Products

Apiin 0.1 0.4 54.8

UDP 5512.0 50.2 0.1

Other compounds

a-D-apio-D-furanosyl cyclic-1:2-P 0.9 1.4 14.9

UDP-xylose and UMP 16.7 44.1 28.3
96.1 98.1

*The incubation mixture contained 0.0133 nmole of doubly

labeled reaction product, 0.015ilmole of '7GA', 7.5 umoles of
sodium phosphate buffer, pH 8.0, and 0.1 milliunit of apiin
synthase (0.36 mg) in 0.15 ml. The mixture was incubated 5 ~
minutes at 25° and then heated for 2 minutes at 100°. The in-
cubation mixture, UDP, UMP, d—D-[U- “CJ-apio-D-furanosyl cyclic-
1:2-P, and apiin were spotted separately on paper and the paper
was developed in solvent 0 for 18 hours at 22° and scanned. The
various radioactive compounds were eluted at 40 with 75%

aqueous ethanol. A portion of each was used for determining

the isotope content and 3H:14C ratio. The remainder was used
for identification. The compounds were identified as described
in the Results. The [BHJUMP present in the reaction product

was taken into account in calculating the values for UMP.

**The doubly labeled reaction product was the same as that
used in Table I. The starting percent of 3H and 1"‘C is based
03 that froT both the reaction product and the contaminating
[ HJUDP-[U- “Cnylose.

38

to their respective mixtures. The incubation mixtures were
treated and the per cent of each isotope in a compound was
determined as described in Table IV. A value of 100% was
1“C and 3H content of the doubly-labeled
l4

(l4

assigned to both the
C and 3H content of the following
compounds was: apiin C-2.l%, ave. of 1.6, 2.3, 2.5%;
3H.0.6%, ave. of 0.0, 0.4, 1.3%). UDP (Ibo-0.5fi, ave. of

reaction product. The

0.0, 0.0, 1.4; 3H-3.9%. ave. of 2.1, 2.3. 7.2%), a-D-apio-
D-furanosyl cyclic-1:2-P (luC-57f. 3H-2.0%); UMP (IhC-Ofi;
33-531), UDP-xylose (1“0-355; 3H-36x).

Stability of UDP:[U-1“c1apioss. -- The stability of UDP-
[U-lucjapiose was greatest at pH 5.6 (Fig. 1). Some hydrolysis
occurred even at pH 4 and 7. Under the conditions used no
significant hydrolysis of UDP—[U-lucjxylose was detected at
any pH except possibly a small amount at pH 2 and 11. Hydro-
lysis of UDP-[U-lucjapiose to a-D-[U-1aC]apio-D—furanosyl
cyclic-1:2-P and UMP occurred rapidly at pH 8 and 100°, being
virtually complete after 2 minutes (Fig. 2). There was no
detectable hydrolysis of UDP-[U-lucjxylose even during the
longest heating periods. This was established by hydrolyzing
the compounds with an RF of a—D—[U—lucjapio-D-furanosyl
cyclic-1:2-P and UDP-[U—lucjxylose with acid from those
incubation mixtures that were heated for time periods longer
than 2 minutes. From the former only D-[U-luC]apiose was
obtained and from the latter only D-[U-lucjxylose. Hydro-
lysis of UDP-[U-luc]apiose to d-D-[U—luc]apio-D-furanosyl
cyclic-1:2-P and UMP occurred much more slowly at ph 8 and

39

 

 

omodaofioeaubgun A. 0V

mnmuaiogoao Hzmocoaanaioaomosansuunus A0 ............. 0.5.5.... 8

.ooussaa om :a o: coo 0 mg no onaooo

omoaaomoaauaunmop no ouoaaoadan oanopSmooa on nos» 30:» monsoom on» ad voodoooha
moon .oeooo coouaooou no cocoon» son» duo psoaaaoaxo on» we soapoasu on» son

0: coo can ma no one: oao: macanzaoo one .coaoaonn oao: .aopox an cocoaaon

mo: soapsaoo common on» non» paooxo opono donaaoooc muoHpSaOo oaaaoo on» on
HooHpsocd msoapsaom Hoppsoo mooduoohaoo ooonu oxoa oe .soaponsosa oaouon soapaaoo
omoaaomozauaunmas on» a“ usomoha moooAHo oooaaomosausuuo can mnmuauofinoho Hamosonsmun
uoaaomoeauagnnnu one new cocooaaoo ohoz moaaaoo HH< .omouaomoaauagin ho muNHH
aoaaomo Hmoosoasmuaaoaaomoeaupguano o» oopaobsoo not scan: doe: omoaaomoaausunmop
on» go psoo son on» ma mamhfloacmn usoo pom .m soapsaom soauoflaapaaom ad oopssoo
and use use oaoz moons o>apooo«uoa opoaaaohaao on» coo cosqooo .oaaon on com m
poo>aom so“: 0: no coaoHo>oc no: momma one .oasuohoaaou osoo on» no cocoa donoox no
coupons mos» com c: on UoHooo one: oaoausHoo on» mopsnaa om hon omN no soduonso
use aopm< .Hs ma.o no: soapsaom oanaom sooo no oasno> Honda one .o.HH so 0.0H
.o.e .o.m .e.a .o.o .o.m .o.s .e.m .o.~ me access as one: asoaaaaoe access use
.mooz no mussoeo msahao> was ac: no moaoaa m.~ .oposamoga nowohchnao asadom mo
moaaaaam.a .nommm mo moaoaa m.H dosaopsoo soans_:odpsaom houmsn o cocoa no: sumo oa
.Aeae ooe.nv oaoaawmosa-ag-daa uo sacs: emo.o one lane oem.aav sooaaomosa-sguaap
mo oHoas omo.o doodopsoo nods: oohoaoaa onoz o: no osoapsaom n: .:a no moduossm o
no 0mm no mopsaas om houmo omodaomooanauumaa no oaohaoaoh: mo unopxo one .H .wam

mwh:2_<< 00 «On—

 

40

 

 

 

 

e. , OO—

aaznoaom asouv-do n iwaoaaa

41

.oaasoo esnpuonou on» an onomaondhn erepoeuee o: no: ones»

co:ose eHonusoo on» sonn connoppo oped one .m:bap:~oe Honusoo on» non enono
oennnoeoc mo deacon» and oenoaona no: .o.o ma no pun enono confinemec ooosu on
Hooapneen .osonpzaom Honuaoo paeanneaxo on» wannsc eHaaee eanpuonow on» an
eennsooo do: mnonaonunn ado na esnenepee oa .eaasoe eanpnoneu on» an usemena
mumuauonao»o aheosonsnaauonaomo: tsunaua on» non douoennoo one: eeaaaoe

HH< .mamuauonaozo HamosenznuouOHQo oaausuiniu on dopnensoo mo: sens: com:
ooonaomuaaupgima: on» no psoo nea on» on enmhaonch: unoo non .m acnusaom
nonuoaannsnom an oopssoo use poo use one: noeno o>npoooncon eponnaonaao

on» use eessoom .mnson ma non o pse>Hom Spa: 0: no eeaoaenec mo: neaon one
.o: no neaoa doses: :0 ueppoao one: use can an op oeumznoo one: escapadoe

on» HHo cone .Aooussns om haopoaaxonaaev psoannoawo on» no noduonsu on» non
o.m ma one o: no paex son» use omN no oeuonzosuonn no: oHaaem eanpnonou one
.co>oaen mo: ego peed es» Hanna 0: no anex use eanp eponnaonnao on» no oe>oaon
me: some use naaonusesoee wounopm one: encapsaom .ecaoooo om: no can .oNH
.om .oo .om .nn .o non soon so eonspaoan one: nos» scan .omm no seasons N
uenensosaona one: .Ha H.o no easaob Hogan e an <sam no modes: m.m one o.m ma
.sonnso ensconced senses no aonoaa n .nsae oom.m~n oeonnwmoenupu-nn= no monoss
mm.e .neao ooo.o~v ononaomusnuouunas no hence: mm.o wananaasoo some .oa as
msonpsaom an .eanp spa: coca use o.m an no ms: use nuNHHuOHH050 Hamosensn
-a-onasmoenuouuoua on osonasmosnuagunoa no unannoaena no seen one .N .mnn

42

 

4"!
480

 

..r
l
240
secowos AT pH 3.0 AND 100°

 

i
120

 

 

So

100

OJZA'IOUOAH asouv-aon 1N3 083d

43

250 than at pH 8 and 100° (Figs. 1 and 2). Some hydrolysis
occurred even at pH 8 and 4° (Fig. 3). Stability of
UDP-[U-140]apiose at 40 and.pH 6.0 was determined. Each of
three incubation mixtures contained 0.23 nmoles of UDP-
[U-lncjapiose (17,600 dpm), 0.24 nmoles of UDP-[U-lncj-
xylose (18,500 dpm). 4.8 nmoles of sodium phosphate buffer,
pH 6.0,and‘2=.4 nmoles of EDTA in a total volume of 0.05 ml.
After incubation for 5 hours at 4° the mixtures were spotted
and chromatographed on washed Whatman No. 3MM paper in
solvent B for 26 hours. The results obtained were compared
to those obtained from three identical incubation mixtures
which were made up 5 hours later. UDPer-lucjapiose for
this latter set of incubation mixtures was kept at .200 in
solvent A during the 5 hours that the first set of incubation
mixtures was incubated at 4°. The amount of radioactivity
which migrated with an RF of D-[U-140]apiose and a-D-[U-lucj-
apio-D-furanosyl cyclic-1:2-P from both sets of incubation
mixtures was measured. The data showed that no detectable
hydrolysis of UDP-[U-luc]apiose occurred during the 5 hours
at pH 6.0 and 4°. Hydrolysis of 11 or more of the UDP-[U-lucj-
apiose would have been detected.
The stability of UDP-[U-lucjapiose was determined at
pfl 5.5 and in 501 ethanol by volume. This was done by storing
the sugar nucleotide at -20° for 2 months as described in
Table V. It was not hydrolyzed to a-D-[U-luCJapio-D-furanosyl
cyclic-1:2-P or to Dn[U-1“C]apiose during the storage period.
The stability of UDP-[U-1“c]spiose at 25° and pH 6.0 was
determined to check the feasibility of concentrating solutions

.eanaee

oanpaoneu on» an eewhaonehn eeonnemoaalpuamna on» no nm.m pen» oesose oaonasoo
one sonn confidano wood one .onone connnoeec no condone soap and useaanenwo

on» no senpensu on» non o: no paex one: encuusaoe Honusoo 0:» one .epoonfinzo

an eoneaonn mo: .o.o mg no asp onono eennnoeec anon» on Hounpneua .sonuaaoe
Honpsoe e anoaanoaxo on» manned eaaaee eanuionou on» an connsooo do: enemaone»:
has nn osnaneuoe ca .oaaaoe eauuuonow on» an usooenn mumuaionaoho wheosensn
nauonaomo anaguans on» non eeuoennoo one: eeaaaoe Had .mimuauOnHOho Haeosensnua
acnaomoeaua cans on depnepsoo he: sons: coo: eeonaemoeauagumn: on» no psoo non on»
on unenaonens ucoo nem .m soapsaoe sonuoaanpsnoe an oeussoo one use use one:
moeno ennuooonuon ouonnnonano on» use .oossooe .enao: 5H non o unopaoe sun:

as no eeaoao>eo no: nonea e29 .02 no noaen venue: so cannons one: use wan an

on denounce one: econosaoe es» sens .do>osen an: ego peed on» Hanna 0: no

ago: one eanu enonnaonano on» no eeeoaen mo: node use eanu osee on» we eepnoae
one: ononpaaoe HH< .0: on unmaonn one: son» one eonssda oua no ow .om .ma .0
non onN no oeuonzosn one: .Hs ma.o no eesaoe Hanan e an «saw no eeaoas nn.n use
.o.m ma .aonnao oossaaosa senses no sonoaanm.n .naae eo~.mn swonnwmesnuauunaa

no monoa: na.o .nsae emn.nn oeonasmoenusgunaa no sense: ooo.e manansasoo sous
.o: as asonosnom .. .oana can: own one e.m an as as: use numanuonnono nanoseaan
-auonasmoen-=uua-e on suonaomosn-pg-nn= no unannonens no soon see .m .mnn

45

 

    

L
90

 

 

 

#
O
In

100

oaznouoiw asonav-aan 1N3 bill

so

 

MINUTES A'l' pH 8.0 AND 25°

46

of it under reduced pressure with little or no degradation.
The degradation of UDP-[U-luCJapiose under these conditions
yielded both D-[U-lucjapiose and a-D-[U-lncjapio-D-furanosyl
cyclic-1:2-P about equally throughout the course of the
experiment (Table VI). After 1 hour 1.2x of the UDP-[U-luCJ-
apiose was degraded. After 5 hours 3.01 was degraded to
a-D-[U-lucjapio-D-furanosyl cyclic—1:2-P and 3.21 was degraded
to D-[U-lucjapiose. The data show that the degradation of
UDP-[U-lnCJapiose to D-[U-1“c]apiose and to a-D-[U-luCJapio-
anuranosyl cyclic-1:2-P is each linear with time at 25° and
pH 6.0.

Solvents A, B, C, D, and I effectively prevented the'”
appearance of d-D-[U-lucjapio-D-furanosyl cyclic—1:2-P and
D-[U-lucjapiose from UDP-[U-14C]apiose during chromatography
at 4°. Solvents F and G permitted the degradation of some
of the sugar nucleotide during chromatography at 40 to
a-D-[U-lucjapio-D-furanosyl cyclic-1:2-P.

47

TABLE V: The stability of UDP-[U-luCJapiose at .20° and at
pH 5.5*

 

X‘UDP- U-lucjapiose % UDP- U-lucjapiose Total UDP-
[U-1 C apiose

 

9"“ °33§3€3§%i§3o3'n' E331351§§1§3. D- degraded
uranosyl cyclic-
1:2-P
0 -+0.00 -+0.00 -+0.00
10 +0.04 +0.23 +0.2?
14 -o.21 +0.34 +0.13
21 -0.47 '+O.34 -0.13
29 .0.11 -0.20 -o.31
41 +0.39 +0.13 +0.52
60 -0.23 +0.12 .0.11

 

*The storage mixture contained 1.97 nmoles of UDP—[U-luC apiose

(147,400 dpm) in each ml of a mixture of equal volumes o 0.04
mM Tris-acetate. pH 5.5 and absolute ethanol. The mixture was
stored at -200. At the time intervals indicated, 2-100 ul
portions were removed, spotted on washed paper at 40 and chro—
matographed in solvent B for 30 hours at the same temperature.
The radioactivity migrating with an a of D-[U-19CJapiose and
a-D—[U-luglapio-D-furanosyl cyclic-l: -P was measured and the
average v as for the 2-100 ul portions was recorded. The
values for these two compounds in the zero-time samples were
set equal to zero. Some fluctuation in the values of these
two compounds during the course of the experiment was expected
as the amount of them in the zero-time samples was relatively
high (1.8!) compared to the amount formed upon storage.

48

TABLE VI: Hydrolysis of UDP-[U-thJapiose at pH 6.0 and 25°
with time'.

 

 

Time UDP- U-luCJapiose a-D-[U-luc apio- D-[U-lucj-
(minutes) rema ning D-furanosy cyclic- apiose
m 1‘“ m m
0 100 0.0 0.0
30 99.5 0.1 0.4
60 97.9 0.5 0.7
90 ' 97.3 1.7 V 1.0
120 97.2 1.5 1.3
210 96.3 2.2 1.5
300 93.8 3.0 3.2

 

(26,600 dpm), 0.38 nmole of UDP C] ]xylose (28, 90 dpm) and

5 nmoles of sodium phosphate bu er, pH 6. 0, and 2. 5 nmole of
EDTA in a final volume of 0.1 ml, were incubated at 25° for 0,
30, 60, 90, 120, 210, or 300 minutes and then were brought to

4°. All solutions were started at the same time and each was
removed at the appropriate time and kept at 4° until the last

one was removed. The solutions were spotted on washed paper

at 4°. The paper was developed at 4° with solvent B for 30
hours, scanned, and the appropriate radioactive areas were cut
out and counted in scintillation solut on B. Hydrolysis was
measured by thi a pearance of a-D—[U-l C]apio-D-furanosyl cyclic—
1: Z-P or D-[ U- “C apiosg or both. Per cent hydrolysis is the
per cent of[ the UDP-[ U- “C apiose used which was co verted to
a-D~[U- “0] Japio-D-furanosy cyclic-1:2-P and D- U-1 Cjapiose

as determined by radioactility measurements. A 1 samples were
correc ed for the a-D- U-1 C ]apio-D-furanosyl cyclic-1:2-P and
Dn[U-1 Ca apiose presen in the zero-time sam le
senteda ve show that no hydrolysis of UDP- 0-140
curred in the zero-time sample during the course o
experiment.

*Solutions at 4°, each containin§0125 nmole of UDP- U-luc Cjapiose
v-1
f

Data pre-
gapiose oc-
the

49

DISCUSSION

One of the richest sources of D-apiose is in the cell
wall polysaccharides of duckweed. Some of these polysaccharides
have recently been isolated and characterized (15-17). Hart
and Kindel (17) found D-apiose glycosidically linked to
polygalacturonic acid (apiogalacturonans). As the biosynthe-
sis of polysaccharides appears to involve transglycosylation
reactions from sugar nucleotides (68) similar reactions could
be expected to occur in the biosynthesis of apiogalacturonans
of duckweed.

The search for a sugar nucleotide containing D-apiose
produced several preliminary and inconclusive communications.
Sandermann and Grisebach (69) isolated sugar nucleotides from
lg. crispgm and obtained 7250 nmoles of ultraviolet absorbing
material. These were hydrolyzed with acid liberating the
sugars of the.sugar nucleotides. One of the sugars (0.15
nmole) migrated with the RF of D-apiose. These data, they
said, showed the presence and existence of a sugar nucleotide
containing Dnapiose. Sandermann, Tisue and Grisebach (38)
later reported the isolation.of a.cell-free extract from
duckweed which formed several radioactive compounds from
UDP-[U-luCngucuronic acid. These compounds migrated with
the RF of UDP-xylose. The cell-free extract was incubated
with UDP-[U-lucjglucuronic acid (1,100,000 dpm) and NAD1'for
4 hours and 20 minutes at 25°. Then the UDP-pentoses (14,500
dpm) formed during incubation were isolated by paper chro-
matography and electrophoresis. They were hydrolyzed

50

with acid and the liberated pentoses were isolated. One
of these pentoses (260 dpm) migrated with the RF of Dbapiose.
As reported by Grisebach and co-workers, the amount of UDP-
apiose formed (38) or isolated (69) was quite small and the
possibility that they were a contamination or an artifact can
not be easily discounted. In both reports very limited data
were used to show the existence of the new compound.
Independently Gustine and Kindel (39) reported the=isola_
tion."ofa cell-free extract from duckweed which converted 22%
of the substrate, UDP-[U-lucjglucuronic acid, into a D-apiose
containing compound. No UDP—apiose was isolated under the
conditions they carried out their experiments. As mentioned
in the Results of Part 2 of this thesis, Kindel afterwards
found that much less of this D-apiose containing compound
(later identified as a-D-[U-lucjapio-D-furanosyl cyclic-1:2-P)
was formed in incubation mixtures which were not boiled at
pH 8 and which were chromatographed at 4°. Re postulated
that a-D-[U-lucjapio-D-furanosyl cyclic-1:2-P was the degra-
dation product of UDP-[U-lucjapiose rather than the primary
enzymatic product of the reaction with UDP-[U-lucjglucuronic
acid catalyzed by the extract from duckweed. Early work
by Paladini and Leloir (62) supported this hypothesis. They
studied the base catalyzed intramolecular phosphorylation
of UDP-glucose and found that UDP-glucose would phosphorylate
intramolecularly above pH 8. If UDP-apiose was formed by the
extract and if it was more sensitive than UDP-glucose to

lower pH then it might phosphorylate intramolecularly at pH 8.

51

I had previously made an observation that suggested that
UDP-apiose was being formed by parsley UDPGA cyclase from
UDP-glucuronic acid. I observed that parsley UDPGA cyclase
was necessary to synthesize apiin through its formation of
an unidentified D-apiose donor molecule. This reaction also
required the presence of '7GA' and was catalyzed by parsley
apiin synthase.

The above observations suggested that UDP—apiose was
formed from UDP-glucuronic acid by enzymes isolated from
duckweed and parsley. This was unequivocally established by
the isolation of all the sugar nucleotides formed from
[3H]UDP-[U-140]glucurcnic acid by purified duckweed UDPGA
cyclase and duckweed UDPGA decarboxylase and their subse—
quent characterization. A very mild isolation and partial
purification procedure was devised which separated the sugar
nucleotides from the other substances present in the incu-
bation mixture (NADT, UMP, a-D-[U-luCJapio-D-furanosyl
cyclic-1:2-P and duckweed enzymes).

The presence of base, sugar and phosphate in the reaction .
products (sugar nucleotides) was demonstrated in the following
five ways, which showed that they were sugar nucleotides.
First, the reaction products were found to contain the

14

expected amounts of 3H and C based upon the amounts of 3H

and 14
1("C in the substrate was lost as [MC]C02. These data

C present in the substrate and assuming that 1/6 of the

show that the assumption that the uridine portion remained

attached to the sugar portion. Second, both of these sugar

52

nucleotides had the same RF as authentic UDP-xylose in a
variety of paper chromatographic systems. Third, mild
acid hydrolysis (pH 2) yielded [3HJUDP, a little [3h]0hp,
D-[U-1u0]xylose and D-[U-luCJapiose. Paladini and Leloir
(62) had observed that mild acid hydrolysis (pH 2) was a
good indication of the presence of a sugar nucleotide. The
glycosyl phosphate linkage of a sugar nucleotide is less
stable than that of the corresponding glycosyl phosphate.
For example, D-glucose can be quantitatively split from
UDP-glucose by heating it at 1000 for 15 minutes in 0.01 N
hydrochloric acid (62), whereas complete cleavage of
D—glucopyranosyl l-phosphate to D-glucose and inorganic
phosphate requires similar treatment in N acid (63). The
A00 of hydrolysis of the D-glucopyranosyl l-phosphate bond
of UDP-glucose has been calculated to be approximately
-7600 calories at pH 7.4 (64), whereas that of D-gluco-
pyranosyl l-phcsphate is -4800 calories at pH 8.5 (65).
Fourth, the product of intramolecular phosphorylation of
one of the two reaction products resulted in the formation
of [3HJUMP and a-D—[U-lucjapio-D-furanosyl cyclic-1:2-P.
The reaction product containing D-[U-lucjxylose like a
similar pyranoid sugar nucleotide, UDP-glucose (62), did
not phosphorylate intramolecularly at pH 8 and 100°. This
reaction product was recovered and its 3H:luc ratio was
unchanged. Fifth, apiin synthase was employed to further
characterize one of the reaction product as [BHJUDP-[U-lucj-
apiose by showing that it would function as a D-apiose donor

53

molecule in a glycosidic reaction. A reaction mixture was
prepared with [3H]UDP-[U-1“c]xylose, [3HJUDP-[U-1“c]epiose,
'70A' and apiin synthase. [BHJUDP-[U-luc]apiose did react
to form [lucjapiin. This was shown by the formation of
[BHJUDP and [1”C]apiin and by the disappearance of [BHJUDP-
[U-luc]apiose. The experiment also shows that the products
of the reaction catalyzed by apiin synthase are UDP and apiin.
[BHJUDP-[U-lucjxylose did not react in the above incubation
mixture. This was shown by the recovery of the added
[3H]UDP-[U-1“c]xylose and its unchanged 3Hzmc ratio. A,
reaction mixture was prepared with [3H]UDP-[U-luC]xylose
[33]UMP, a-DnEU-luc]apio-D-furanosyl cyclic-1:2-P, '7GA' and
apiin.synthase. After incubation the above radioactive com-
pounds were recovered unohanged. Radioactive compounds were
not observed with the RF values of the expected products of
the above transglycosylation reaction i.e. [BH]UDP and [14C]-
apiin. These results indicate that UDP-[U-luc]xylose did
not react in the reaction catalyzed by apiin synthase. The
above results which show that apiin and UDP are formed from
'70A' and UDP-apiose by apiin synthase agree with those of
various glycosylation reactions with sugar nucleotides (65,68).
UDP has not been identified before as a product of transglyco-
sylation reactions where the substrates were a nucleoside
diphosphate sugar and a flavonoid or similar phenolic com-
pound (68.54.55.57).

The final proof of the proposed structure of UDP-apiose
[uridine-5' (a-D-apio-D-furanosyl-pyrophosphate)] will require

its chemical synthesis.

The stability of UDP-xylose and UDP—apiose was investigated
in both acidic and alkaline solutions. UDP—xylcse was stable
under the conditions tested except at pH 2 and 100°. The hydro-
lysis properties of both UDP-xylose and UDP-apiose at pH 2 and
1000 are those that are characteristic of sugar nucleotides (65).
UDP-apiose was unstable under most conditions tested. Optimum
conditions for storing UDP-apiose without degradation are -20°,
pH 5 to 6 and in a solution which is 50$ ethanol by volume.

The other organic solvents used to store UDP-apiose at .200
were 503 acetone by volume and solvent A. UDP-apiose stored

at -200, in solvent A and at pH 5 to 6 was stable for months.
At 25° and below pH 5 acid hydrolysis of UDP-apiose occurred.
At 250 and at pH values above 6 alkaline hydrolysis (phos-
phorylatesintramolecularly) occurred. Lowering the temperature
retarded hydrolysis at these pH values. At pH 8.0, for example,
alkaline hydrolysis at 1000 was nearly complete after 2 min-
utes. At 250 it was 405 complete after 90 minutes and at 40

it was 5.51 complete after 90 minutes. At all 3 temperatures
and pH 8 some hydrolysis occurred.

The above reaction is an example of reactions which occur
with some sugar nucleotides in the presence of alkali. The
result is a nucleoside monophosphate and a cyclic 1:2-phosphate
of the sugar. This type of reaction occurs if the hydroxyl
group on carbon atom 2 of the sugar is located sufficiently
close to the phosphorus atom on carbon atom l of the sugar

residue to permit the formation of a 5-membered ring (67).

55

Khorana, Tener, wright and Moffatt (67) called this type
of a reaction intramolecular phosphorylation and represented

the general reaction as given below.

1’ fi
('3- 0—l|’—0-l"— on’ £13
in - on on
H o
“r“ :: ‘6
+R'OP-OH
C- \ow \
ow

Intramolecular phosphorylation has been found to

occur in bopsgluouronic acid ianHuOH (50,66).
Basic hydrolysis (intramolecular phosphorylation) at pH 8

of [BHJUDP-[U-lucjxylose and [3HJUDP-[U-lucjapiose revealed
that only [BHJUDP-[U—lhcjxylose was stable at pH 8. [BHJUDP.
[U-lucjapiose was'degraded to [3HJUMP and a-D-[U-lacjapio-
D-furanosyl cyclic-l:2-P. Both UDP-apiose and UDP-xylose
would be expected to form a cyclic sugar phosphate because

of the stereochemistry at carbon atom l and 2 of the sugars.
However, this was not expected under the mild conditions that
resulted in the basic hydrolysis of UDP-apiose (4° and pH 8)
since Leloir and Paladini (62) did not observe intramolecular
phosphorylation with UDP-glucose except under more severe

56

conditions, i.e. higher pH and temperature. UDP-glucose, a
compound with a pyranoid ring and with the same stereo-
chemistry as UDP-xylose, is quite stable at pH 8. After 18
hours at 180 and pH 8, UDP-glucose is undegraded (62). How-
ever, it did begin to cyclize at a more alkaline pH and 100°.
The tendency of UDP-apiose to phosphorylate intramole-
cularly at pH 8 has recently been mentioned by Sandermann
and Grisebach (40). They suggested that the ready formation
of d-D-apio-D—furanosyl cyclic-1:2-P could be explained by
its great stability. This stability is due to the presence
of two condensed 5-membered rings which would be in contrast
to the cyclic 1:2-phosphate of anylose and of D-glucose.
These latter compounds have a 5-membered ring condensed to a
6-membered ring. The facile degradation of 5-phospho-ribo-
furanosyl-pyrophosphate was cited as an analogous reaction
resulting in the formation of two condensed 5-membered rings.
Besides the suggestion of Sandermann and Grisebach (40)
another reason for the unusual sensitivity of UDP-apiose to
basic conditions can be suggested. It would also explain
the unusual sensitivity of UDP-apiose to acidic conditions
which the suggestion of Sandermann and Grisebach (40) does
not explain. (They did not report investigating the stability
of UDP-apiose in acidic conditions nor its isolation and
characterization.) My suggestion (which will be explained
in.more detail later) is that a pyranose sugar of a sugar
nucleotide requires more energy to change into the best con-

formation for intramolecular phosphorylation of the sugar

57

nucleotide or acid hydrolysis of the glycosidic linkage than
does a furanose sugar of a sugar nucleotide. The amount of
energy necessary to convert the sugar into the ideal confor-
mation for the greatest reactivity of the sugar nucleotide
affects the sugar nucleotide's lability in acidic and basic
solutions.

My suggestion can be readily understood from the infor-
mation supplied by studies on the hydrolysis of the glycosidic
linkages of oligosaccharides. The glycosidic linkages of
oligosaccharides are readily hydrolyzed by acids, but are
relatively stable in alkaline conditions. Acid hydrolysis
of oligosaccharides occurs, as with other glycosides, by
fission of the bond between carbon atom l of the glycosyl
residue and the glycosidic oxygen atom (91). The mechanism
of the acid-catalyzed hydrolysis of pyranosides has been
extensively investigated. Little information is available
on the hydrolysis of furanosides by acid. Pazur and Gordon
(92) and Aspinall and Telfer (93) reported that glycosidic
linkages involving furanose residues were extremely acid-labile
in polysaccharides. Shafizadeh (82) says that there are two
possible mechanisms consistent with the observations stated
above and other observations which would explain the acid-
catalyzed hydrolysis of pyranosides. "They are shown in
formulas [3] to [10] (Fig.l$). One mechanism involves rapid
reversible protonation of the glycosidic oxygen atom to
yield the protonated oligosaccharide [3] which undergoes a

slow unimolecular decomposition to a stable monosaccharide

58

Fig.l+. Two possible mechanisms for the acid hydrolysis
of glycosidic linkages between pyranoses (88).

CH20H

 

 

H30

 

/( /( H "—9
"H H . (H \9
on H C. - OH H ‘ OH H -- H+ROH
no OR' so go - H R0 '
l I l " g
H H

011

 

60

and a cyclic carbonium ion [4]. Nucleophilic addition of
water to the electron-deficient carbon center yields a
protonated reducing sugar [6], and subsequent loss of‘a
proton yields the expected hydrolytic products [7]: A
possible, but not widely accepted alternative mechanism,
Shafizadeh (82) says,"would involve protonation of the ring
oxygen atom of the glycosyl moiety to yield a protonated
oligosaccharide [8] followed by the cleavage of the pyranose
ring to give an acyclic carbonium ion [9]. Nucleophilic
addition of water on the acyclic ion would yield an unstable
intermediate [10], which would predictably eliminate a mono—
saccharide residue to yield an aldehydo-monosaccharide.
Rearrangement of the latter into the pyranose structure
would yield the final hydrolytic product [7].I Indirect evi-
dence from experiments on the rates of hydrolysis of D-xylose
derivatives supports the first mechanism as being operative
in the hydrolysis of these compounds (94). The relative
rates of hydrolysis of pyranosides in terms of the first
mechanism have been explained by conformational effects and
intramolecular steric interactions (95).

Pazur (88) says, ”the rate-determining step is the
formation of the carbonium ion, which is considered to exist
in a ”half-chair" conformation.” The most unstable confor-
mation of cyclohexane, a compound which is very similar in
conformation to the various pyranoses discussed, is the
"half-chair" conformation (96). About 5.5 kcal/mole is

required to change from the stable, "chair“ conformation to

61

Fig. 5. Torsional isomerism in cyclohexane (96).

62

 

 

63
the "twist" conformation. This means that at 250 there are
10,000 "chair'I conformations for each "twist" conformation,
(96). About 0.5 kcal/mole is required to change from the
"twist" conformation to the “boat" conformation (Fig.55).
However, to change from “chair“ conformation to the "half-
chair' conformation requires 11 kcal/mole (96). This is
because formation of the "half-chair" conformation involves
a small rotation about the carbon atoms 2 to 3 and carbon
atom 3 to 4 bonds. The rate of acid hydrolysis is dependent
primarily on the extent of the interaction of the equatorial
substituents on carbon atom 2 relative to carbon atom 3 and
on carbon atom 4 relative to carbon atom 5. These groups
eclipse each other in reaching the transition state for the
reaction--the "half-chair“ conformation.(88).

In cyclopentane, a furanose, the change from "envelope"
conformation to "half-chair" conformation and back again
involves almost no change in potential energy (89). In sub-
stituted cyclopentanes one form or the other may have greater
stability sometimes by as much as 2 kcal/mole (89). Even so
this difference is much smaller than that observed in cyclo-
hexane when it changed from the "chair” to "half-chair“
conformation. Actually the shape of cyclopentane is not
fixed. The individual carbon atoms move up and down at right
angles to the average plane of the ring in such a manner as to
cause the irregularity or puckering to move around the ring (89).

The above observations explain the acid lability of UDP-

apiose resulting in its' hydrolysis under mild conditions

64

(pH 4 and 25°). The comparative stability of UDP—glucuronic
acid, UDP-xylose and UDP-glucose to mild acid hydrolysis is
easily understood. Apparently in each case the amount of
energy required to reach the rate limiting or critical
reaction conformation or intermediate determines the reacti-
vity of the sugar nucleotide. UDP-glucuronic acid, UDP-xylose
and UDP—glucose, require a large amount of energy to change
the conformation of the sugars from the stable, "chair” con-
formation to the critical, unstable, reaction conformation.
Some of the D-apiose of UDP-apiose either exists in the
critical, reactive conformation (”half-chair“) or requires
very little energy to bring D-apiose into this conformation.
An examination of Framework Molecular models representing
UDP-apiose and UDP-xylose reveals that UDP-apiose can readily
phosphorylate intramolecularly. In sugar nucleotides free
rotation about phosphorus to oxygen to phosphorus bonds is
restricted by steric hinderance. Thus, the models reveal
that UDP-xylose must change from the more stable ”chair” con-
formation to the less stable ”half-chair" conformation (96,89)
to bring the hydroxyl group on carbon atom 2 to its closest
position to the a phosphorus atom of the sugar nucleotide so
that intramolecular phosphorylation is more likely to occur.
The hydroxyl group of carbon atom 2 of the stable, "half-chair"
form of D-apiose in UDP-apiose is much closer to the a phos-
phorus atom than it is in the stable, "chair" form of
D-xylose in UDP-xylose. In fact the "half-chair“ conformation

and the "envelope” conformation of furanoses change back and

65

forth and back again via intermediate asymmetric arrangements
which involve no substantial change of potential energy (89).
The change from one form of D-apiose to the other involves
little movement of the hydroxyl group on carbon atom 2 away
from the B phosphorus atom. These facts support my suggestion
that the furanoid ring of D-apiose is important in causing
instability at alkaline pH values as well as acidic pH

values.

PART 3

PURIFICATION AND CHARACTERIZATION OF APIIN SYNTHASE FROM

2. crispum

66

67

INTRODUCTION

A wide variety of phenolic compounds exist in higher plants
including the flavonoids and related compounds. Flavonoids
normally occur in higher plants not as free phenolic compounds
but rather in an amazing variety of glycosides containing mono-
or oligosaccharides (53,68).

Barber (54) first described the glycosylation of a flavonoid,
quercetin (3',4',3,5,7-pentahydroxyf1avone), in which the glycosyl
donor was UDP—glucose. The product of the reaction was 3-(3',4',
3,5,7-pentahydroxyflavonyl) O-B—D—glucopyranoside. The reaction
was catalyzed by a cell-free extract from Phaseolus aureus (mung
bean). Further incubation of the enzyme preparation in the pre-
sence of TDP-L-rhamnose catalyzed the transfer of L—rhamnose to
form 3—(3',4',3,5,7-pentahydroxyflavonyl) O-B-L-rhamnopyranosyl-
(1+6)-B-D-glucopyranoside. Barber (55) later reported the gly-
cosylation of quercetin by UDP-L-rhamnose to quercetrin, 3-(3',
4',3,5,7-pentahydroxyf1avonyl) O-B—L-rhamnopyranoside. This
reaction was catalyzed by a cell-free extract from Leucaena glauca.
Miles and Hagen (48) reported the formation of the 3-monoglucoside
of kaempferol, quercetin and myricetin from UDP—glucose by a

cell-free extract of Impatiens balsamina. Marsh (49) reported

 

the formation of 3-(3',4',3,5,7-pentahydroxyflavonyl) O-B-D-
glucopyranosyluronic acid from UDP-glucuronic acid and quercetin
by an extract of Phaseolus vulgaris (French bean).

In 1970 there were two preliminary reports on the enzymatic
glycosylation of another flavonoid. Watson and Kindel (56) re-
ported the purification of an enzyme from P. crispum which
catalyzed the transfer of D-apiose (3-C-hydroxymethy1—
aldehyd -D-g1ycero-tetrose) from UDP apiose to

68

7-(4'5,7-trihydroxyflavonyl) 0-8-D-glucopyranoside to form
apiin (Fig. 1, Part 1). They gave the common name apiin
synthase to this enzyme. Ortmann, Sandermann and Grisebach
(57) also reported the isolation from 2. hortense of a cell-
free extract forming the same compound.

I report here the partial purification and characterization
of apiin synthase. The substrates and products of the reac—
tion catalyzed by apiin synthase are identified. Kindel1
had previously obtained evidence for the existence of this
enzyme in cell-free extracts of P. crispum and established
that one of the products of the reaction was apiin. UDP-
apiose:7-(4'-5,7—trihydroxyf1avony1) B-O—D-glucopyranoside,

2

D-apiose transferase is the systematic name Watson and Kindel

suggest for apiin synthase.

MATERIALS AND METHODS

Materials. -- NAD+ was obtained from Sigma Chemical Co.
UDP—[U-luc]glucuronic acid was obtained from New England
Nuclear Corp. Sephadex G-25, 0-100 and DEAR—Sephadex A50
were obtained from Pharmacia Fine Chemicals, Inc. and
polyethylenimine was obtained from Dow Chemical Co. '7GA"
was isolated from crystalline ”apiin" and recrystallized
from 95% aqueous ethanol (8). Crystalline ”apiin" was
isolated from parsley seeds by the method of Gupta and
Seshadri (45). Although crystalline, it was a mixture,
hence the quotation marks for both it and '7GA“ (8).

 

1Kindel, P. K., unpublished data.

2Watson, R. R. and Kindel, P. K., manuscript in preparation

69

UDPGA cyclase was isolated from L. minor (41). d-D-[U-luc]-

l and UDP-[U-luC]apiose were

Apio-D-furanosyl cyclic-1:2-P
formed enzymatically from UDP—[U-luC]glucuronic acid

(Part 2). D-Apiose was isolated from "apiin" (8).

All other materials used were of the highest quality avail-
able from commercial sources. The parsley used to obtain
apiin synthase and UDPGA cyclase was 2. crispum, moss-curled
variety. They were mature (8-12 months of age) plants grown
in a greehnouse from seeds obtained from Ferry—Morse Seed Co.,
Fulton, Ky. Bovine serum albumin was purchased from Research
Products Division, Miles Laboratories, Inc. UMP, UDP, UTP,
UDP-glucuronic acid, UDP-galactose and UDP-xylose were from
Sigma Chemical Co. Diaflo cells and UM-lo Diaflo membranes
were purchased from the Amicon Corporation.

General Methods. -- Paper chromatography was by the descending
technique and was carried out with Whatman No. 3MM paper.

The paper used with solvents A and E was treated with 2.5%
neutralized polyethylenimine as described by Verachtert,

Bass, Wilder and Hansen (46). The paper used with solvents
B—D and F-H was washed with 0.1 M citrate and then with
distilled water. The following solvents were employed: (A)
0.3 M LiCl, (B) 95% aqueous ethanol-1.0 M ammonium acetate,

pH 7.5 (7:3, v/v), (C) 1-butanol-acetic acid-H20 (4:1:5,

v/v, upper phase), (D) l-propanol-ethyl acetate-water

(7:1:2, v/v), (E) 0.5 M LiCl, (F) pyridine-ethy1 acetate-

acetic acid-water (5:5:l:3, v/v), (G) ethyl acetate-acetic

 

1Kindel, P. K. and Watson, R. R., manuscript in preparation

7O

acid-formic acid-water (8:4:l:3, v/v) and (H) 2—propanol-
water (9:1, v/v).

Radioactivity was detected on chromatograms with a
radiochromatogram scanner, Model 7201 (Packard Instrument
Co.). All other radioactivity measurements were made with
a Tri—carb liquid scintillation counter, Model 3310, employing
either: (A) a scintillation solution made as described by 1
Bray (84) or (B) 2,5-bis-[2-(5-tg§t-buty1benzoxazo1yl)]thio-
phone in reagent grade toluene (4 gm/l). The portion of the
chromatogram containing a radioactive compound was cut out and
completely immersed in solution B in a scintillation vial and
counted. The counting efficiencies with solution A and B
were 79 and 60%, respectively.

Protein was determined by the biuret method (58) before
column chromatography of the enzyme. After passage through
Sephadex G-100 it was determined by the procedure of Lowry,
Rosebrough, Farr and Randall (59). After passage through
the DEAR-Sephadex column protein was estimated by 280/260.
Then the fractions containing apiin synthase were concentrated
by ultrafiltration, dialyzed against buffer (2-500 ml volumes
of 0.01 M sodium phosphate, pH 7.4, 1 x 10"+ M EDTA and l x

10'2

M B-mercaptoethanol for 1 hour) and the protein was
measured by the procedure of Lowry, Rosebrough, Farr and
Randall (59).

Standard Assay for Apiin Synthase. -— Apiin synthase was mea-
sured by quantitatively isolating [14C]apiin formed from
I'7GA" and UDP-[U-1uC]apiose. The following standard assay

was used. In a 12 m1 conical centrifuge tube, 2.5 nmoles

71
of sodium phosphate, pH 8.0, containing 1.2 nmoles of EDTA,
80 nmoles NAD+, 2.2 nmoles of UDP-[U—14C]g1ucuronic acid
(110,000 dpm, glucuronic acid portion uniformly labeled
within i15%) and 0.2 thiunits of duckweed UDPGA cyclase
in a final volume of 25 ul. After incubation at 25° for
20 minutes, the amount of UDP-[U-1“c]apiose, UDP-[U-1“c]-
xylose and d-D-[U—luC]apio-D-furanosy1 cyclic-1:2-P present
in the standard assays was determined by chromatographing
and further treating a standard assay as described on page
28.Iof Part2 . More than 95% of the UDP—[U-luC]glucuronic
acid had been converted to UDP-[U-1uC]apiose and UDP-[U-luC]-
xylose in the standard assay. The amount of UDP-[U-luC]apiose
and UDP-[U—luC]xylose formed was about the same since the
purity of various preparations of UDPGA cyclase used was about
the same with respect to UDPGA decarboxylase. As shown in
Part 2 page 28 34% of the UDP-[U-luC]glucuronic acid had
been converted to UDP-[U-lnc]xylose and 63% was UDP-[U-luc]-
apiose. Sometimes the amount of UDP-[U—luc]apiose formed
was 3-5% more and the amount of UDP-[U—14C]xylose formed was
the same amount less when the UDPGA cyclase contained less
contaminating UDPGA decarboxylase. To conserve apiin syn-
thase, UDP-[U-luc]glucuronic acid and duckweed UDPGA cyclase

14C]g1ucuronic

the standard assay contained only enough UDP-[U-
acid to yield a concentration of UDP-[U-luc]apiose which was
approximately five times the Km that apiin synthase has for UDP-
[U-luC]apiose. After incubation of UDPGA cyclase with
UDP-[U-luC]glucuronic acid, 5.5 nmoles of "7GA" in water

and apiin synthase were added to make a final volume of 50 ul.

72

Fig. 1. Chromatography of some of the compounds present in
the standard assay mixture on polyethylenimine impregnated
paper with LiCl solutions. Chromatography of apiin,
d-D-[U-luC]apio-D-furanosyl-1:2-P, UDP-xylose, D-[U-luc]-
apiose and UDP-glucuronic acid was performed on paper pre-
pared as described for solvent A and E in the Materials

and Methods. The solutions of LiCl were allowed to migrate
45 cm which in each case took about 4 hours. After chroma—
tography at 220 the distance each compound had migrated was

determined-

O—O D-apiose A—A d-D-apio-D-furanosy1
cyclic-1:2-P

D—D UDP-xylose
A-g UDP-glucuronic acid

O—O Apiin

73

 

 

 

 

 

P _

 

 

 

"CLARITY OF Li Cl

 

 

 

_ _
O O O O
4 3 2 ....

2530 20m“. mmwbmirrzmo

74

Only 5.5 nmoles of '7GA" would dissolve or stay in solution

at 250 without raising the temperature to dissolve it (see
below). Therefore 20 ul of a solution of '7GA' that had been
boiled (0.28 nmoles/pl) were added to give the maximum quantity
of '7GA' which was soluble in 50 ul at 25°. The solution con-
taining 0.28 nmoles of “7GA” per ul was prepared by briefly
boiling '7GA' in water and then cooling to 25°. Apiin syn-
thase was added to the standard assay mixture, it was incubated
for 5 minutes at 250 and then heated at 1000 for 2 minutes.
Protein was removed by centrifugation and the supernatant was
applied to polyethylenimine impregnated paper for chromato-
graphy in solvent E. The supernatant from each assay was
streaked over a 2 cm portion of the paper. The protein pre-
cipitate from each assay was suspended in 100 ul of 70%

ethanol and removed again by centrifugation. The supernatant
from the ethanol wash was applied on the same 2 cm wide por-
tion of the paper. The chromatography paper was 12.5 cm long.
After 4 hours of chromatography the origin, which contained
[1nC]apiin, was cut out and the radioactivity measured as
described in the General Methods. All the other compounds pre-
sent in the assay mixture containing 1“C had been removed from
the origin (Fig. 1).

Apiin synthase catalyzed the formation of apiin from exo-
genously supplied UDP-[U-luc]apiose. A modification of the
standard assay is presented. The modified standard assay was
used only to obtain the data given in Figs.8 and 9. In a
12 ml conical centrifuge tube, 2.5 umoles of sodium phos-
phate buffer, pH 8.0, containing 1.2 nmoles of EDTA,

 

 

75

5.5 nmoles of "7GA" and a mixture of 0.032 nmoles UDP-[U—luC]-
apiose (17,300 dpm) and 0.014 nmoles of UDP—[U-luc]xylose
(7,500 dpm) were added. The reaction was initiated by the
addition of apiin synthase, the mixture was incubated for

3 minutes at 250 and then heated at 1000 for 2 minutes.

The assay mixture was further treated as described above

for the standard assay mixture.

A unit of apiin synthase is defined as the amount re-
quired to form one umole of apiin per minute at 250 from
UDP-apiose and "7GA" under the conditions of the standard
assay.

Absorbance and Solubility of "7GA". The absorbance of a solution
of recrystallized '7GA' in water at 25° was determined at
335¢nu and at 268 mu. The molar absorbancy index, am, at

1 cm'l. The molar absorbency

index at 268 mu was 17,400 liters mole"1 cm'l.

4

335 mu was 19,000 liters mole‘

At 25°, 1.1 x 10' moles of '7GA“ was dissolved in 1

liter of water. However, after boiling briefly 2.8 x 10"4
moles of '7GA' remained in solution in 1 liter of water after

it was cooled to 25°.

76
RESULTS

IDENTIFICATION OF ONE OF THE REACTION PRODUCTS
FORMED BY APIIN SYNTHASE

In vitro Enzymatic Formation of Apiin (Compound IV). -- In-
cubation of NAD+, parsley UDPGA cyclase, UDP—[U—luC]g1ucuronic
acid, '7GA” and apiin synthase resulted in the formation of
two radioactive D-apiose-containing compounds. In Part 2

one of these compounds was identified as the product of in-
tramolecular phosphorylation of UDP-[U-luC]apiose. The other
is identified below. An incubation mixture consisted of
0.276 nmoles of UDP-[U-luC]glucuronic acid (105,600 dpm), 160
nmoles of NAD+, 4.8 umoles of sodium phosphate, pH 8.0, con-
taining 2.4 nmoles of EDTA, 11.0 nmole of "7GA" and 0.145 mg
of protein containing both parsley UDPGA cyclase and apiin
synthase activity from the ammonium sulfate material which
had been passed through a Sephadex G-25 column (40 cm in
height and 2.2 cm in diameter) in a final volume of 100 ul.
After prior treatment of the enzymes by passage through
Sephadex G—25 or 0-100 exogenous "7GA" was required for the
biosynthesis of the compound with an RF of apiin (Compound
IV below). After incubation at 25° for 30 minutes,

it was heated at 1000 for 2 minutes. T‘101101‘1'11'13 removal
of the denatured protein by centrifugation, the supernatant
liquid was applied to washed Whatman No. 3MM paper and
chromatographed in solvent B. Scanning the chromatogram for

radioactivity revealed that four distinct radioactive compounds

77

were present. Numbering them from the origin, they were

designated Compounds I, II, III, and IV and had R values

F
of 0.12, 0.28, 0.66, and 0.73, respectively. Compound I

was chromatographically identified in solvent systems A and

B as UDP-[U-luC]glucuronic acid. Compounds II and III were
found to have the same RF values in solvent systems A, B

and C as UDP-[U-1u0]xylose and d-D-[U-luC]apio-D-furanosyl
cyclic-1:2—P. Hydrolysis of Compounds I, II and III in 0.1 N
H230“ revealed the presence of only one radioactive component

1I‘J'Ctjglucuronic acid in Compound I,

in each compound: D-[U-
D-[U-luC]xylose in Compound II and D-[U-1uC]apiose in Compound
III. Only Compound IV is discussed in detail since its syn-
thesis alone required "7GA". UDP-[U-laC]xylose and UDP-[U-
1I+C]apiose have been more extensively characterized in Part 2.
a-D-[U-luC]Apio-D-furanosyl cyclic-1:2-P is a degradation
product of UDP-[U-luc]apiose (Part 2).

Identification of Compound IV as Apiin. —- Compound IV migrated
as a single, radioactive area on paper chromatography in sol-
vent systems A to C. In each of these solvent systems, it

had the same RF as apiin whose RF is distinctly different

from those of D-[U-14C]apiose, UDP—[0-1u0]glucuronic acid,
UDP-[U—luC]xylose and d-D—[U-luC]apio-D-furanosyl cyclic-1:2-P.
A small amount of an impurity was observed in the crystalline
“apiin" when chromatograms were developed in solvent C (8).

It is known that besides apiin, g. crispum contains small
amounts of at least one other D-apiose—containing flavonoid.

Nordstrom, Swain and Hamblin rigorously identified one such

flavonoid as 7—(3',4'.5.7-tetrahydroxyflavonyl)

78

O-?-D-apio-?-furanosyl-(?)-B-D-g1ucopyranoside (see also

Part 1). Acid hydrolysis of "apiin” yielded the crystalline
"70A" which was used as a substrate in the incubation mixture
described above for the formation of Compound IV and which
also contained a small amount of impurity (about 5%), based
upon the weight of each compound after elution. The impurity
in “apiin" was observed only on chromatograms developed in

solvent 0 where the R for the impurity was 0.83. The

apiin

R was 1.0 for the major component of "apiin" which was

apiin
assumed to be pure apiin. This major component migrated 28
cm when chromatographed in solvent C for 15 hours. The Rapiin

for the impurity in '7GA" was 0.94 and the R for the

apiin
major component of '7GA' was 1.14. The major component in

'7GA' was assumed to be pure 7GA. Grisebach and Bilhuber (60)
have isolated crystalline "apiin" from g. hortense and found

it contained a small amount of 7-(3'-methoxy, 4',5,7-trihydroxy-
flavonyl) O-B-D-apio-D-furanosyl-(1»2)—B-D-glucopyranoside on

the basis of its hydrolysis and spectral properties. Grisebach
and Bilhuber (60) identified the principal product (isolated

from g. hortense) and Gupta and Seshadri (45) identified the

only product (isolated from P. crispum) which they found after
hydrolysis of crystalline apiin as 7—(4',5,7-trihydroxyflavonyl)
O-B-D—glucopyranoside. Compound IV migrated with an RF of the
major component of the apiin and not the minor component of'apiin:’
Both of the components of “7GA” were eluted from chromatograms
developed in solvent 0. Each component of "70A" was incubated as

described above for the formation of Compound IV. The major com-

ponent of "70A", when incubated with UDP-[U-luC]apiose and

79

apiin synthase resulted in the formation of only Compounds
II, III and IV. Co-chromatography of this Compound IV with
|'apiin" showed that it had the same RF as the major compon-
ent of "apiin" on chromatograms developed in solvent C.
None of Compound IV migrated with an RF of the minor com-
ponent of "apiin". The minor component of ”7GA" when incubated
as described above for the formation of Compound IV, yielded
a very small amount of a substance which had the same RF as
the minor component of l‘apiin" in solvent C. On the basis
of these results and the observation that all of Compound IV
migrated with an RF of the major component of ”apiin" in
solvent C, Compound IV is considered homogeneous. None of
Compound IV (1% or less) migrated with any other RF than the
RF of the major component of “apiin" in solvent C.

Compound IV was hydrolyzed in 0.5 N H230“ for 1 hour
at 1000 in the presence of 0.15 umole of authentic D—apiose.
This mixture and authentic D—[U-lucjapiose similarly treated
were each spotted separately onto Whatman No. 3MM paper.
Chromatography in solvent B showed that Compound IV was no
longer present and only a single radioactive compound with
the same RF as the authentic D-[U-luC]apiose was observed.
This compound contained eighty-six and one-half percent
(86.5%) of the starting radioactivity which was all the
radioactivity' found on the chromatogram.
This is the average of 3 experiments whose
values were 87.9, 87.2 and 84.3 percent. The recovery on
the chromatogram of authentic D-[U-1uc]apiose identically

treated was 87.1 percent. This shows that the amount of

80

I
D-[U-14C]apiose recovered was the same as the amount ofradioaCtive

material obtained from Compound IV. This suggests that all

of the radioactive material from Compound IV was a single

l“0]apiose in

compound with the characteristics of D-[U—
solvent B. The radioactive material with an RF of D-apiose
obtained from Compound IV was eluted. Aliquots of this
radioactive material were co-chromatographed with carrier
D-apiose (0.211mole per aliquot) in solvent A, B, C, D, F, G
and H. This radioactive material in each solvent system
migrated with the same RF as the carrier D-apiose which was
visualized with aniline hydrogen phthalate (73). Apiin has
a distinctly different RF from D-apiose in solvent systems
A, B, C and D. The radioactive product of acid hydrolyzed
Compound IV has the same chromatographic properties as the
compound Gustine and Kindel (39) characterized by periodate
oxidation to be D-apiose.

Compound IV was further identified as apiin by recrystal-
lization of a mixture of Compound IV and crystalline "apiin"
to constant specific activity. Compound IV (41312 dpm) and
2.25 mmoles of "apiin” were dissolved in 144 m1 of 95%
ethanol by boiling. The hot solution was filtered through
Whatman No. 1 filter paper and the crystals were washed with
7.5 ml of boiling 95% ethanol. After this solution stood
at 220 for 21 hours and 3 hours at 4°, the crystals were
collected and air dried. This process was then repeated by
redissolving the dried crystals in 95% ethanol (Table I).

To further identify Compound IV as apiin the acetate

derivative of a mixture of Compound IV and ”apiin" was

81

.oHoEE\Eac swam mm:

opwumom gsfiaaog was opopoom >H ossoasoomo: H mo zpfl>apom candooam oopoonxo 02p one

oaoaa\emo comma mm: asfidnmg one >H ossonaoo Udag mo mpa>wpom camfiooam cocooaxm onwl.

.nopm nofipmuaaawpmacoop w50a>opa on» scum oopo>ooon unmosmm
.2.

 

 

 

 

 

 

 

o.osmm s.~: m.a: s
o.om~m m.ow m.om m
o.HHmm H.mm H.Hm N
o.mamm H.ao m.mo H
oaoss\eao R m
an. omopoow : as V
mpmpmom >H onsomsoo_oaa_ tomam>ooop one» toono>oooh cones:
hpa>apom onHoopm noon >H undonaoomosau endowed :«HQ4. :oHpmNaaampmhho
u.mnam m.mw N.nw m
s.mm:m o.~o H.mm :
m.nmmm d.mm m.nm m
o.mm~m p.35 5.55 N
m.:oom m.mm “.mw H
oHoEE\aao R a
A.sfianm.\>H onsogaoomosagv toopo>ooop *ooho>ooon popes:
*.suw>apoc oscdoo m >H onsonsoomosag .qwan<. noapauafiflmpmsgo
.opwpoom >H onsonaoomoeau can mumpoow asadnmg mo manpxfis
w dam >H onsonsoomodag and asaanm. no ohspxaa a mo scapunaHHMpmhuoom .H mqm<a

82

recrystallized to constant specific activity. Compound IV
and "apiin" were previously recrystallized together five
times from 95% ethanol. The acetate derivatives of these
compounds were prepared by the procedures of Gupta and
Seshadri (45). Then the mixture of "apiin" acetate and Com—
pound IV acetate was recrystallized to constant specific
activity. This was accomplished by dissolving 5405 dpm of
acetylated Compound IV containing 0.66 mmoles of "apiin”
acetate in 99 ml acetone (analytical grade). This solution.
was stirred with heating until the precipitate dissolved.
Further heating resulted in solvent evaporation. Just as a
precipitate began to form the solution was cooled 5 minutes
at 22°, the precipitate collected and air dried.

A summary of the recrystallization data involving Com-
pound IV and "apiin" acetate are presented in Table I. The
data show that the ratio (dpm/mmole) of Compound IV to "apiin”
does not change significantly. The ratio of the acetate de-
rivative of Compound IV and the acetate derivative of "apiin"
also does not change significantly during recrystallization.
DISTRIBUTION OF APIIN SYNTHASE IN THE LEAVES, STEMS AND ROOTS

OF PARSLEY

The distribution of apiin synthase activity was measured
in the leaves, roots and stems of several mature parsley
plants which were examined for activity following the
extraction procedure described below for leaves. A
summary of the analyses for apiin synthase activity is

Presented in Table II. The leaves are the best source of

83

apiin synthase activity as they yield more total activity

as measured in the assay used (Table II), although some
activity is observed in the stems and less in the roots.

The enzyme isolated from stems had 2.5 times higher specific
activity than the enzyme isolated from the leaves and roots.

TABLE II. Distribution of apiin synthase in the leaves,
stems and roots of parsley.

 

 

 

 

Total protein . * Total activity
Tissue 100 gm tissue Speleic Activity 100 gm tissue
mg milliunits/mg protein milliunits
Leaf 1050 0.020 21.07
Stem 147 0.051 7.45
Root 398 0.014 5.60

 

*Apiin synthase was measured in the standard assay with 0.2
nmoles UDP-[U-lucjglucuronic acid (110,000 dpm) as substrate.

PURIFICATION OF APIIN SYNTHASE

A typical purification of apiin synthase is detailed
below and summarized in Table III. All procedures were
carried out at o-5° unless otherwise stated.

Extraction. -- Parsley leaves from mature parsley plants were

 

diced and 300 gms of diced leaves were homogenized in a
Waring Blendor for 2 minutes with 660 m1 of 0.1 M sodium

-4 ‘2 M B—mercapto-

phosphate, pH 8.0, l x 10 M EDTA and 1 x 10
‘ethanol. The resulting slurry was squeezed through four
.1ayers of cheesecloth and centrifuged for 20 minutes at

114,500 x g. Apiin synthase was measured in the standard assay.

84

Ammonium sulfate. -- Ammonium sulfate (105.1 gm) was added

to the supernatant material from the extraction step (641

ml) to bring the saturation to 30%. The solution was stirred
5 minutes and centrifuged at 14,500 x g for 20 minutes.
Ammonium sulfate (75.2 gm) was added to this supernatant to
bring the saturation to 48%, stirred 5 minutes and centrifuged
at 10,000 x g for 10 minutes. The precipitate was taken up

in about 25 m1 of buffer to give a final volume of 50 ml.
Apiin synthase was measured in the standard assay mixture
after dialysis of 1 m1 of the dissolved precipitate against
250 m1 of buffer for 1 hour. Then it was dialyzed 2 more times
against 250 m1 of buffer for one hour until the ammonium.
sulfate was removed from the dissolved precipitate.

Sephadex 0-100. -- A column of defined Sephadex G-100, 65 cm
in height x 5 cm in diameter, was prepared and equilibrated

with 0.01 M sodium phosphate, pH 7.4 containing 1 x 10‘“ M

EDTA and 1 x lo“2

M B-mercaptoethanol. The dissolved pre-
cipitate from the previous step was applied to the column

and eluted with the above buffer. The column was eluted at
1.0 ml/minute and fractions of 12.2 mls were collected. The
void volume began with fraction number 29. Most of the apiin
synthase eluted into 9 tubes, numbers 42 - 50, which were
combined and used below. (Fig. 2). Apiin synthase was
assayed in the standard assay.

DEAR-Sephadex. -- A column 10.0 cm in height and 2.2 cm in
diameter was prepared from defined DEAE-Sephadex. The DEAE-

Sephadex had been pretreated in a beaker by‘washing as

85

follows: (1) with 0.1 M sodium phosphate, pH 7.4 until the
pH remained at 7.4, (2) 4 times with 200 ml of 0.01 M

sodium phosphate, pH 7.4, containing 1 x 10'“ M EDTA and

1 x lo“2

M B-mercaptoethanol. After the column was equili-
brated with 500 m1 of the above 0.01 M buffer, the material
from the Sephadex G-100 step was applied. After application
of the eluant from the Sephadex 0-100 step containing apiin
synthase the column was washed with 60 ml of 0.1 M NaCl in
buffer. Apiin synthase was then eluted by increasing
linearly the NaCl concentration. The linear gradient was

set up with 200 m1 of 0.3 M NaCl in buffer being gradually
added to 200 ml of 0.1 M NaCl in buffer in the mixing flask.
Fraction of 9.2 ml were collected. Almost all of the parsley
apiin synthase eluted into 3 tubes, numbers 13-15, which
were combined and concentrated to about 1.0 ml of ultrafil-
tration in a Diaflo cell (Fig. 3.). After dialysis for 2
hours against buffer, 1.0 ml of glycerol was added. Enzyme
prepared in this manner constitutes the enzyme characterized
below. Apiin synthase was purified 45 fold by this procedure
and was almost free of UDPGA cyclase. Table III shows that
more than 99% of the UDPGA cyclase activity in the original

cell-free extract had been removed. Apiin synthase was

assayed with the standard assay in all four steps.

86

confiscate mm mdmzflmfio one soaumppafimmaufi: an soflpmcpsmosoo seams cossmmms mm: aufi>dpo<
.n.

.pxee we» as

*t.

.ncmmzn pmzammm couaamao mm: pa amped coached HHmEm m :o dopammme mm: mpfipauo<et

m>HpmHmp one cause unmadmmnsm one

.H 0»

.ampm sofipomspxc on» ma H 0» Hence pom some mm: omega

team cases no unfl>fipom cacaoomm on» o» mmmaomo <0ma: mo spfi>daom camaomgm on» mo cause one:

 

 

 

 

ee.o e s.me m ms.m assessesaem-m<ma
He.e mm m.mH em ee.a oeflue xeeesaem
sa.o mm m.e mama no.0 saececasm assesses
e.H eoa o.H mean mo.o soaeeesaxm
Mmcpopm c we
a each as madssawHas
ommnuszm madam n
ampoaa hua>apom
semeaesa «emu: efleas seseeeaaassm fleece caeaeecm deem

 

.mo>moH acamnmn

scum cmmSpchm added mo soaumOHmHLsm .HHH mqm<9

87

Fig. 2. The elution pattern of UDPGA cyclase and apiin
synthase from Sephadex 0-100. The ammonium sulfate frac-
tion containing apiin synthase ( I" """"" I--"""""I )

and parsley UDPGA cyclase ( 0 O o ) was
applied to the Sephadex 0-100 column as described for the
Sephadex G-100 step. The enzymes were eluted with buffer
at 1.0 ml/minute and 12.2 ml fractions were collected. The
amount of protein in each fraction was measured

( o o o ).

 

 

 

 

88

1W I NISLOIJ 9W

 

 

 

 

 

 

 

 

 

O O
N '- _ In
#T T
Q “3 g
1.__fi__.,
I .- _____
.I" ...... v
....'o’ ...... . ........
0’ i. ...“ .....
II "y
9.. ‘J‘...
i o ------ a ..
.... the, “um-u...
c 1.. '9“
i/ 'u ’o/ ' :-
w 5°
\O\O \O .V.
t I
a... o /
o/.—f
\.
\
g~
‘ . __
M _.._ .\ _
d a e’ “\
ALIAIADV DHIDJJS J
is s L
"2
'- O

1W I SWAIN!

JO SllNfll‘l'llW

25

FRACTION NUMBER

89

Fig. 3. Chromatography of apiin synthase and UDPGA cyclase
on DEAR-Sephadex. The Sephadex 0-100 fractions containing
apiin synthase ( I ----- --l--------l ) and some parsley UDPGA
cyclase ( o—o—o ) were applied to the DEAE-Sephadex
column as described for the DEAE-Sephadex step. Protein
was measured ( o O O ) by the procedure of
Lowry, Rosebrough, Farr and Randall (59) after concentra-
tion by ultrafiltration and dialysis. Fractions were col-
1ected starting with the application of the linear gradient
described in the text. Fraction number 1 is the first one
collected after application of the gradient. No UDPGA
cyclase or apiin synthase was observed in the solutiomscol-
1ected before application of the gradient. Protein in frac-

 

 

tions 1-5 and 18—20 was measured by absorbance at 280/260.

90

1W I NIBIOUJ 9W
‘0

+-—~ 4?

umuav amazes I
s 0.
4‘40 7
o/
t’ /.'/° ..
\o / \e’”

\o a: “y”. \

40.2

 

 

15 20

IO

FRACTION NUMBER

 

 

 

‘ O
l
, O
- .l I
O In
- 0.
0 0

1w I iWAZN! :IO SllNfllTllW

91

PROPERTIES OF APIIN SYNTHASE

Linearity of the Apiin Synthase Reaction. -— The data in Fig.
4 (standard assay), Fig. 6 (standard assay) and Fig. 8 (modi-
fied standard assay) show that the transfer of D—apiose from
UDP-apiose to "7GA' was linear with time for 5 minutes. The
data in Fig. 5 (standard assay), Fig. 7 (standard assay) and
Fig. 9 (modified standard assay) show that the reaction
catalyzed by apiin synthase is linear with increasing concen-
trations of apiin synthase over the range of apiin synthase
measured. The rate of apiin formation was directly propor-
tional to the amount of apiin synthase present. Crude pre-
parations of apiin synthase such as described for the extrac-
tion and ammonium sulfate step were almost linear with respect
to enzyme concentration. However, some apparent inhibition

of the reaction was noted particularlyat higher concentra—
tions of enzyme which was removed by passage through Sephadex
G-25 or 0.100. This treatment removes small molecular weight
compounds which probably include some of the large quantity

of the product, apiin, present in the cell-free extract described
in the extraction step and the ammonium sulfate step.

Factors Affecting the Stability of Apiin Synthase. -- Stability
of the enzyme to storage largely depended on the protein con-
centration, temperature and the addition of glycerol. When
the enzyme was mixed with an equal volume of glycerol and
stored at .200 good stability was achieved. Enzyme from the
.Sephadex G-100 step stored at u° (4 mg/ml) for 12 hours lost
almost all activity (>95%). Addition of glycerol to make

‘the solution 50% glycerol by volume preserved enzymatic activity

92

Fig. 4. Formation of apiin by apiin synthase as a function

of time with low concentrations (< Km) of "70A”. The forma-
tion of apiin by apiin synthase was followed using the standard
assay. The assay mixture contained 0.6 nmoles of ”70A" in
place of the 5.5 nmoles in the standard assay. The reaction
was initiated by 0.032 milliunits (1.44 ug of protein) of

apiin synthase purified through DEAR-Sephadex step. The stan-
dard assay mixture was incubated for the indicated times (1 to
5 minutes) at 25° before termination.

Fig. 5. Effect of apiin synthase concentration on reaction
velocity with low concentrations (< Km) of "70A“. The for-
mation of apiin by apiin synthase was followed using the
standard assay. The assay mixture contained 0.6 nmoles of
"70A" in place of the 5.5 nmoles in the standard assay. The
reaction was initiated by the indicated amounts of apiin
synthase purified through the DEAF-Sephadex step (0.0024 milli-
units/1.0 pg of protein). The standard assay mixture was
incubated for 3 minutes at 25° before termination.

93

23.23:. 62

«aha-2.2

 

 

 

 

no.0

 

 

o 3

ClinlOi N lldV $31OWN

94

Fig. 6. Formation of apiin by apiin synthase as a function
of time with low concentrations (< Km) of UDP-apiose.' The
formation of apiin by apiin synthase was followed using the
standard assay except that it contained 0.2 nmoles of UDP—[U-
ll‘nglucuronic acid (100,000 dpm). The reaction initiated
by the addition of 0.057 milliunits of apiin synthase (12.5
ug of protein) from the DEAE-Sephadex step. The assay mix-
tures were incubated for the indicated times at 25° and
further treated as described in the Materials and Methods.

Fig. 7. Effect of the amount of apiin synthase on reaction
velocity with low concentrations (< Km) of UDP-apiose. The
formation of apiin by apiin synthase was followed using the
assay described in the legend of Fig. 6. However, the reac-
tions were initiated by the addition of the indicated amounts
of apiin synthase from the DEAE-Sephadex step (0.0046 milli-
units/ug of protein) and were incubated for 3 minutes at 25°
and further treated as described in the Materials and Methods.

 

95

2.2.03;

13M»)

 

0:

9m
4

 

muh32_¢<

 

— 0.0

L
N
0.
O
aawaoa NlldV so snow N

 

 

. A ‘91.,-

i u .f ,.,.

91.:

) 4 .4 A?»
a lbw v

96

Fig. 8. Formation of apiin by apiin synthase as a function

of time with exogenous UDP—apiose. Apiin synthase was assayed
by the modified standard assay described in the Materials and
Methods. The reaction was initiated by 0.046 milliunits (6.2
ug of protein) of apiin synthase from the DEAE-Sephadex step.
The reaction contained 0.032 nmoles of UDP-[U-lucjapiose
(17,300 dpm) and 0.014 nmoles of UDP-[U-IQCnylose (7,500 dpm).
The reaction was incubated for 3 minutes at 25° and then heated
to 100° for 4 minutes.

Fig. 9. Effect of the concentration of apiin synthase on
reaction velocity with exogenous UDP-apiose. Apiin synthase
was assayed by the modified standard assay described in the
Materials and Methods. The reaction was initiated by the in-
dicated amounts of apiin synthase from the DEAE-Sephadex

step (0.0148 milliunits/pg of protein) and contained 0.032
nmoles of UDP-[U-lucjapiose (17,300 dpm) and 0.014 nmoles.

of UDP-[U-lucjxylose (7,500 dpm). The reaction was incubated
for 3 minutes at 25° and then heated to 1000 for 4 minutes.

97

 

2.2.05.
..m

o:

 

 

mmhayzi

N

 

 

 

 

 

,
3.153
c

M:

000.

000.

SJ'IOWN

GJWUOi NlldV

98

at -200. Apiin synthase after DEAF-Sephadex step (17.3 mg/ml)
increased its apparent activity 30% with addition of an equal
volume of glycerol. After 1 week at .20° it still had 112%

of the activity observed before the addition of glycerol.

Two weeks after adding glycerol only 80% of the original
activity remained. The standard assay was used in the above assays.
pH Optimum of Apiin Synthase. —_ Apiin synthase has optimum
activity between pH 7.6 - 8.4 (Fig. 10). In this range the
enzyme has almost the same activity with the two buffers tested.
As indicated in Fig. 10 apiin synthase activity on either side
of this range was near the optimum value and only gradually
decreased in activity with a change in pH.

Affinity_of Apiin Synthase for "70A" and UDP-apiose. -- The

Km for "70A" was calculated from the data presented in Fig.

11. It is 7.0 x 10-5 M. The Km for UDP-apiose was calculated
from the data presented in Fig. 12. It is 0.6 x 10-5 M.

Effect of Various Ions on the Activity of Apiin Synthase. --
Apiin formation by apiin synthase (from the DEAE-Sephadex step)
was measured in the standard assay. The various ions tested
were present in the standard assay at 1 mM final concentra-
tions. The compounds tested were NHucl, FeClB, CaClz, MgClZ,
NaCl, KCl, NaMoOB, NaBOB, NiClZ, 00012, CuCl2 and ZnClz. None of
these compounds increased apiin synthase activity. Only the
latter four metal ions decreased the enzymatic activity with

a decrease of 29% for NiClz, 40% for C001 88% for CuC12

2’
and 95% for ZnClz. The presence of EDTA, d,a'-dipyridy1,
8-hydroxyquinoline and KF at 1 mM final concentrations in

the standard assay mixture had no effect on enzyme activity

99

Fig. 10. Effect of pH on the reaction velocity of apiin syn-
thase. Incubation mixtures were prepared in a volume of 40 u]
containing 2.25 nmoles of UDP-[U-IQCngucuronic acid (88,000
dpm), 50 nmoles of NAD+, 0.8 moles of sodium phOSphate, pH
8.0 and 0.2 milliunits of UDPGA cyclase. After incubation

at 25° for 60 minutes, 11 nmoles of "70A“ and either 5.6 nmoles
of sodium phosphate ( A. A ) or 5.6 umoles of Tris-HCl

( AdnuunnA ) buffer were added. The pH of these final
solutions was measured in larger assay mixtures which did not
contain the UDPGA cyclase, apiin synthase and UDP-[U-lucjglu-
curonic acid. The reaction was initiated by the addition of
0.012 milliunits of apiin synthase (15 ug of protein) from
the DEAF—Sephadex step making a final volume of 100 ul. The
assay mixtures were further treated as described in Materials
and Methods.

 

 

 

 

 

RELATIVE ACTIVITY
u d
O o 8
I T
’
V— \
D
U
I ”H—
1
ch 0......
<1 .....

 

OOI

 

101

Fig. 11. Effect of "7GA" concentration on the velocity of
the apiin synthase reaction. The formation of apiin was
measured in the standard assay with various concentrations
of "7GA" ranging from 0.02 x 10-5 M to 9.07 x 10-5 M. Each
assay contained 1.24 nmoles of UDP-[U-luC3glucuronic acid
(90,000 dpm) which was incubated with 0.12 milliunits of
duckweed UDPGA cyclase (0.2 mg of protein) 2.5 umoles of
sodium phosphate, pH 8.0, containing 1.2 nmoles of EDTA

and 100 nmoles NA0+ . After 15 minutes at 25° the 30 ul
reaction mixture contained 0.84 nmoles of UDP-[U-lucjapiose
(63,000 dpm), 0.30 nmoles of UDP-[U-luC]xylose (22,500 dpm),
0.1 nmoles of UMP and 0.1 nmoles of d-D-[U-lucjapio-D—
furanosyl cyclic-1:2-P (7,500 dpm) in place of the UDP-
[U-1ucjglucuronic acid previously added to the mixtures.
The various amounts of "70A“ were added to the assay mix-
tures and 0.032 milliunits of apiin synthase from the DEAE-
Sephadex step (1.4 ug of protein) was used to initiate the
reaction. The assay mixtures were further treated as des—
cribed in the Materials and Methods. The insert in a
Lineweaver-Burk plot of the same data. The Km was determined
by the least squares method using the data shown in the
insert.

102

 

 

 

oo—

 

 

\ 1 86

I mad

 

 

ALIDO'IEA

(alnNIW/ssnoww)

103

Fig. 12. Effect of UDP-apiose concentration on the velocity
of the apiin synthase reaction. The formation of apiin was
measured in the standard assay with various concentrations of
UDP-apiose ranging from 0.48 x 10-5 M to 17.8 x 10"6 M.
Assay mixtures were prepared as described in the standard
assay and UDP-[U—luC]glucuronic acid was added in varying
amounts from 0.15 nmoles (66,000 dpm) to 1.58 nmoles (110,000
dpm). After incubation with duckweed UDPGA cyclase the
amount of UDP-[U-lucjapiose formed was determined. ”76A"
(5.5 nmoles) was added and the reaction was initiated by

the addition of 0.023 milliunits of apiin synthase (5.0 ug
of protein) from the DEAE-Sephadex step. The reactions

were incubated at 25° for 2 minutes. The assay mixtures
were further treated as described in the Materials and
Methods. The insert is a Lineweaver-Burk plot of the same
data. The Km was determined by the least squares method
using the data shown in the insert.

104

72:. 3.2: .33
a.
s a.
so 0

J.»
'\

 

a $3. 3.9.2.33 \fi. g . \

15.0

 

 

 

 

 

 

 

\
(alnNIWIsaloww ) AIIDO13 A

 

105

except in the case of EDTA which slightly stimulated enzymatic
activity (10%). The sodium phosphate buffer containing EDTA
in the standard assay mixture was replaced by 2.5 umole of
Tris-HC1 buffer, pH 8.0 in the above 17 assays.

Effect of Sulfhydryl Reagents on the Activity of Apiin Syn-
thase. ~- Apiin synthase after DEAE-Sephadex step was incubated
with 1 mM final concentrations of p-chloromercuribenzoate,
N-ethylmaleimide, iodoacetamide and oxidized glutathione in
the standard assay mixture. All (100%) of the apiin synthase
activity was lost with p-chloromercuribenzoate, 75% with iodo-
acetamide, 25% with oxidized glutathione and 18% N-ethyl-
maleimide.

Inhibition of Apiin Synthase Activity by Various Uracil Com-
pgunds. -~ Enzyme from the DEAR-Sephadex step was assayed by
the standard assay in the presence of varying concentrations
of uridine, UMP, UDP, UTP, UDP-galactose, UDP-glucuronic acid
and UDP-xylose (Table IV). All inhibited apiin synthase to
some extent but UMP and UDP inhibited it most at the lower
concentrations. UDP-glucuronic acid, UDP-xylose and UTP at

4

l x 10' M concentrations only decreased the apiin synthase

activity about 20%. UDP inhibited this much when present at

l/100th this concentration. UDP-glucuronic acid and UDP-xylose

a

at 1 x 10' M concentrations only decreased the apiin synthase

.4
activity about 10% from that obtained with 1 x 10 ’ M and
6

l x 10' M concentrations of these compounds. Based upon
these results it was concluded that UDP-xylose formed with
the UDP-apiose in the standard assay (1.5 x 10"5 M) or added
with exogenous UDP-apiose in the modified standard assay

-6

(0.6 x 10 M) probably does not affect significantly the

106

reaction of apiin synthase at the low concentrations present
in the assays. The formation of UDP occurs as the apiin syn-
thase reaction progresses resulting in the appearance in the
standard assay of not more than a concentration of 0.5 x 10'6
M UDP at the end of a 5 minute assay. A small amount of UMP
would also be formed from the intramolecular phosphorylation
of UDP-apiose in the standard assay resulting in not more than

2 x 10'"6

M UMP at the end of a 5 minute assay for apiin syn- ”
thase. From the data presented in Table IV inhibition by the
UDP formed in the assay should be less than 6% of the Optimum
rate. Inhibition of apiin synthase by the small amount of

UMP formed as a breakdown product during the formation of
UDP-apiose in the standard assay would be less than 10% of

the optimum rate.

Fractionation of Apiin Synthase with Ammonium Sulfate. ~-
During the purification of apiin synthase, it was observed

that most of the UDPGA cyclase isolated by the procedure
described in the extraction step was removed during the puri-
fication of apiin synthase. The same extraction step was used
to prepare UDPGA cyclase and most, but not all of the apiin
synthase present, is removed during the subsequent purifi-
cation step (Part 4). Since some UDPGA cyclase is not

removed, apiin synthase was fractionated with ammonium sulfate
to attempt to isolate apiin synthase without contaminating
UDPGA cyclase. The fractions in which apiin synthase was found
are reported below. The same fractions contained UDPGA cyclase

and UDPGA decarboxylase activity and are reported in Part 4.

Apiin synthase was isolated from parsley leaves as described

107

.OOH ho mSHm> 0:»

sopr we: :oHss onsuxHa homes pudendum on» op needs me: muszomsoo HHoss: e>onw
es» no use: sen: cosHmuno zaH>Hpom ommsushm :Haaw 0:» on copwasoo mH mpH>Huo<m

 

 

 

an. am am am cacaccacwumns

N: «w Hm Ha mmOHhxumnD

am am am we sacs casosacsfimumap

mm mm me He mQHUHhD

5 mm mm cm as:

mH mm mm mm ma:

0:. Hm em NOH me:
2 mIOH x H z :IOH x H z mIOH x H OH x H
.mussoneoo HHoMp: umuomHmm mo msodeLpsoosoo

wstoHHom on» suHs cosHsuno tapH>Huod omdsushm nHHQ< unsonaoo

 

.mussomaoo HHods: msoHsdb an mascuszm sHHnd no soHpHnansH .>H mqm<e

108

.uHom oHsoAsosHmmmenpuumna

 

 

 

no moHoas mo.H Ssz humus pudendum on» :« consumes we: huH>Hpod ones» m sHHndt
mg: Goa 2h 3.3-5.9: oc.~T.S.~
m3. coca om E.$A.oi aflmxlmmé
9+: «86 NS 3.3-m.~$ 3.11.84
:4. @36 HnH An.~:uo.m3 om...” TmoH
NS oN~.o om Ao.mmum.:nv no...” Tom;
Nam Bud S 3.3-981 om; T34
Ac: pd camMHsm
:Homosn mo ma azasosad mo enamHsn ssHsoaau
meassaHHHa Havoc unassHHHHa we scapdnspsm RV no hudndHoa
apabdpod Hduoa hpnpapod edudoonm sHououn Hence 90pm

 

.opduHsu asHsoasd :uHs omezunhm sHHnd mo sodesoH»0dhm u> mqm<a

109

above in the extraction step. The cell-free extract was

made 1.25 M in ammonium sulfate, stirred 5 minutes and cen-
trifuged at 14,500 x g for 20 minutes. The supernatant was
fractionated by the addition of 3.75 M ammonium sulfate in

-4

0.1 M sodium phosphate buffer, pH 8.0, l x 10 M EDTA and

1 x 10"2

M B—mercaptoethanol. It was stirred 10 minutes
and centrifuged at 10,000 x g for 10 minutes. The precipitate
was collected and redissolved in 0.01 M sodium phosphate,

‘” M EDTA and l x 10’2

pH 7.4, 1 x 10 M B-mercaptoethanol.
More ammonium sulfate was added to the supernatant and the
process was repeated until 6 precipitates were collected.
Each precipitate was dialyzed for 4 hours against buffer.
After 4 hours dialysis all of the ammonium sulfate had been
removed. The amount of ammonium entering the-dialysis solu-
tion from the dissolved precipitate during dialysis was
measured with the Nesseler's reagent (80). A summary of the
analyses for apiin synthase activity in the various fractions
is presented in Table V. Most of the protein fractions con-
tained apiin synthase. They also contained UDPGA cyclase and
UDPGA decarboxylase (Part 4). However, the last two fractions
contained 35% of the total apiin synthase activity isolated
in the extraction step and did not contain any UDPGA cyclase
activity. These fractions contained only 4% of the UDPGA

decarboxylase present in the extraction step (Part 4).

110
DISCUSSION

An important consideration in the purification of apiin
synthase was the maintenance of the enzyme in dilute concen-
trations for minimal time periods. In the example given in
Table 1, little activity was lost on passage through a
Sephadex G-100 column. Subsequent passage through a Sephadex
G-200 column further purified apiin synthase by removing more
UDPGA cyclase, however, some (64%) of the activity was lost.
The Sephadex G-lOO step has consistently given an appreciable
purification with very little loss of activity. Most (98%)
of the UDPGA cyclase isolated with apiin synthase in the cell-
free extract was removed during purification of apiin synthase.
Ammonium sulfate fractionation separated 35% of the apiin syn—
thase activity isolated in the extraction step from the UDPGA
cyclase activity (Table V).

The site of apiin synthesis has been a matter of some
confusion and therefore was investigated. In 1965 Medicino
and Picken (13) found in whole parsley plants that 16% of
the apiin isolated per gm of tissue was from the roots and
84% was from the leaves. Their isotopic tracer experiments
with [1“03Nauc03, [3-1l‘cjserine and [Z-lucjacetate in parsley
roots, leaves and whole plants indicated that apiin was
exclusively synthesized in the roots. They concluded that
apiin synthesis occurred exclusively in the roots and that
apiin is transported from the roots to the leaves. Roberts,
Shah and Loewus (35) in 1967 incubated [Z-BHlmyg-inositol
in parsley leaves which had been detached from the plant.

They found incorporation of 3H into D-apiose of apiin. I

111

isolated apiin synthase from the roots, stems and leaves
of parsley. Only 16.5% of the apiin synthase isolated from
equal quantities by weight of these three tissues was obtained
from the roots. I also isolated UDPGA cyclase from equal
quantities of roots, stems and leaves of parsley. Less than
1% of this enzyme was isolated from the roots of the total
UDPGA cyclase isolated from the roots, stems and leaves
(Part 4). The intracellular function of this enzyme is to
form UDP-apiose, a substrate for apiin synthase. These results
show that the enzymes catalyzing the final steps of apiin
synthesis, the formation of UDP-apiose and the transfer of
D-apiose to 76A, occur primarily in the leaves and stems.

Data presented in this Part and Part 2 show that
7-(4',5',7-trihydroxyflavonyl) B-O-D-glucopyranoside and
UDP-apiose are substrates of the reaction catalyzed by apiin
synthase.

The minor component of '7GA' in P. crispum according
to Nordstrom, Swain and Hamblin (18) is the glucoside of
luteolin. The minor component of '7GA" was incubated for
an hour with apiin synthase and UDP-[U-lucjapiose. A very
small transfer of D-[U-1nC]apiose (i<6% of the amount trans-
ferred in a similar experiment with the major component
of "7GA') occurred, resulting in the formation of a com-
pound with the RF of the minor component of "apiin”. Ortmann,
Sandermann and Grisebach (57) tested the minor component of
'7GA” isolated from 3. hortense in an incubation mixture
with UDP-[U-lucjapiose and a cell-free extract isolated
from 2. hortense. They obtained 9% of the transfer of

112

D-[U-luCJapiose obtained when 7GA was the substrate. The
minor component of ”7GA" in P. hortense is reported to be
the glucoside of chrysoeriol (57).

In the glycosylation reactions of flavonoids reported
by Barber (54) a requirement for ATP in the reactions was
noted. He suggested the reaction required ATP in order to
prevent the degradation of the sugar nucleotides involved
in the transglycosylation reactions by enzymes present in his
crude extract from mung bean. MgCl2 was also included
in his incubation mixtures although it did not stimulate his
reactions and its removal did increase the formation of pro-
duct slightly. The apiin synthase reaction was not stimulated

by the addition of MgCl or UTP.

2
The products of the reaction catalyzed by apiin synthase

are apiin and UDP. The formation of apiin was shown by paper

chromatography, recrystallization to a constant specific

activity and other data presented in this study. [BHJUDP

was isolated and shown to be a product of the reaction when

[3H]UDP-[U-14C]apiose was a substrate in the reaction catalyzed

by apiin synthase (Part 2). The intracellular function of

apiin synthase appears to be the transfer of D-apiose to

form apiin. The low Km for 7GA and UDP-apiose suggest

that they are the substrates for apiin synthase. Ortmann,

Sandermann and Grisebach recently reported that they would

present data to show that 7GA had been formed from 4',5,7-

trihydroxyflavone (apigenin) and UDP-glucose with a cell-

free extract isolated from parsley (3. hortense) (57). The

other substrate of apiin synthase, UDP-apiose, is formed in

113
parsley from UDP-glucuronic acid (Part 4).

Consistently 2 to 3 times as many total units of apiin
synthase as total units of UDPGA cyclase (Part 4) were iso-
lated from parsley in the extraction step.

It may be of physiological significance that UDP inhibits
(13%) apiin synthase at concentrations of l x 10"6 M. For
apiin formation to continue at high rate, UDP must be removed
by resynthesis to UTP and UDP-sugars or in some other way at
a sufficient rate to prevent accumulation of UDP. The UDP—
sugars tested do not inhibit apiin synthase appreciably. It
may also be of physiological significance that UDP inhibits
both duckweed UDPGA cyclase (41%) and UDPGA decarboxylase
(53%) even at concentrations of l x 10'”6 M. This suggests
that, for transglycosylation reactions using these sugar
nucleotides as substrates to proceed at high rates, the UDP
formed must be removed by resynthesis to UTP and UDP-sugars.
This must occur at a sufficient rate to prevent accumulation
of UDP and its inhibition of the formation of the sugar nucleo-
tides like UDP-xylose and UDP-apiose (Part 4). These results
agree with conclusions summarized by Horecker (83) who sug-
gests that control of polysaccharide biosynthesis occurs in
some organisms either through the regulation of the produc-
tion of precursor nucleotides or through effects on the
activity of transferase enzymes.

Paris, Paris and Fries (23) reported the isolation of
a I'glucoapiosylapigenin" from Digitalis purpurea (foxglove).
I observed a compound which migrated with an RF of apiin

in solvent A was formed by a cell-free extract isolated

114

from Q. purpurea by the procedure described in the extraction
step. The protein was isolated as described in the ammonium
sulfate step when ammonium sulfate was added to bring the
saturation of ammonium sulfate from 0% to 65%. After the
precipitate was dissolved in buffer some was passed through

a Sephadex 0-100 column. Protein which was retarded to the
same extent as parsley apiin synthase was isolated and mea-
sured. Activity forming the compound with an RF like apiin
was observed in the ce11-free extract, the dialyzed ammonium
sulfate protein solution and the protein solution after pa3~
sate through the Sephadex 0-100 column. To the latter protein
solution '7GA" was added for synthesis of the compound with
the RF of apiin. Although this compound was not further
characterized Q. purpurea enzyme, which had not been boiled,
was required for its synthesis. Boiling the enzyme prevented
the reaction from occurring. This experiment as well as others
described in this part of the thesis suggest that apiin is
formed from UDP—apiose and 7GA. Enzyme catalyzing this reac-

tion has been isolated from 3 sources (3. hortense (57), 2.

 

crispum and 2. purpurea) which indicates that this reaction
represents an important pathway in plants for the biosynthesis
of apiin.

In conclusion the data presented in this thesis shows
that apiin and UDP are synthesized from 7GA and UDP-apiose

by an enzyme partially purified from.§. crispum.

 

(1.1.11

PART 4

ISOLATION AND CHARACTERIZATION OF UDPGA
CYCLASE FROM 13. crismm

115

116

INTRODUCTION

In Part 2, the enzymatic formation of UDP-apiose and
UDP-xylose from UDP-glucuronic acid by duckweed UDPGA cyclase
and UDPGA decarboxylase was reported. Both UDP-apiose and
UDP-xylose were characterized. The formation of UDP-xylose
from UDP-glucuronic acid by enzymes isolated from both plants
and animals has been reported several times (77). The enzyme
catalyzing the formation of UDP—xylose from UDP-glucuronic
acid has the systematic name of UDP-glucuronate carboxy-lyase
(EC 4.1.1.35) and has the common name UDPGA decarboxylase (77).
The enzyme catalyzing the formation of UDP-apiose from UDP-
glucuronic acid has been given the name UDPGA cyclasel. The
systematic name suggested for this enzyme is UDP-glucuronate
carboxy-lyase (cleaving, cyclizing)1.

Sandermann, Tisue and Grisebach (38) have reported the
preparation of a cell-free extract from parsley which catalyzed
the conversion of UDP-glucuronic acid to a D-apiose-containing
compound. Although not fully characterized this compound was
reported to be UDP-apiose. Gustine and Kindel (39) isolated
cell-free extracts from both parsley and duckweed which formed
UDP-xylose and a compound that was later identified as a-D—
apio-D-furanosyl cyclic-1:2-P2. The latter compound has been

shown to be the product of an intramolecular phosphorylation

of UDP-apiose (Part 2). Therefore UDP-apiose was formed by

 

1Gustine, D. L., Watson, R. R. and Kindel, P. K., manuscript

in preparation

2Kindel, P. K. and Watson, R. R., manuscript in preparation

117

these cell-free extracts and non-enzymatically was converted

to the cyclic phosphate. Duckweed UDPGA cyclase was partially
purified and partially characterized by Gustine (41). A pre-
liminary communication dealing with the characterization of
duckweed UDPGA cyclase will appear shortly (29). Two preliminary
reports have been presented which dealt primarily with the
characterization of UDP—apiose and apiin synthase and also con-
tained some of the information on parsley UDPGA cyclase (42,56).
These three reports were taken in part or completely from

Part 2, 3 and 4 of this thesis. The partial purification and
characterization of parsley UDPGA cyclase is described in

this part of the thesis. Some characterization data on

parsley UDPGA decarboxylase, duckweed UDPGA cyclase and duck—
weed UDPGA decarboxylase are also presented. Evidence is
presented that the UDPGA cyclase and UDPGA decarboxylase acti—
vities from parsley are associated with two different and

separable proteins.

MATERIALS AND METHODS

Materials. -- NAD+ was obtained from Sigma Chemical Co.
14

 

UDP-[U- C]glucuronic acid was obtained from New England
Nuclear Corp. Sephadex G-100, 6.25 and DEAR-Sephadex A50
were obtained from Pharmacia Fine Chemicals, Inc. D-Apiose
was isolated from "apiin” (8). Bovine serum albumin was pur-
chased from Research Products Division, Miles Laboratories,

2 and UDP-

Inc. d-D-[U-luC]Apio-D-furanosyl cyclic-1:2-P
[U-14C]apiose (Part 2) were formed from UDP—[U-lnc]glucuronic

acid.

118

Parsley seed (3. crispum, variety moss-curled) was pur-
chased from Ferry—Morse Seed Co., Fulton, Ky. Parsley leaves,
roots and stems were obtained from mature plants grown in a
greenhouse. UMP, UDP, UTP, UDP-galactose and UDP-xylose
were purchased from P-L Biochemicals, Inc. Duckweed UDPGA
cyclase and duckweed UDPGA decarboxylase were prepared from

_1_-. minor by amodification of the isolation procedure described by

Gustine (41). All other materials were from commercial sources.
The Diaflo cell and UM-lO Diaflo membranes were purchased from
the Amicon Corp.

General Methods. -- Paper chromatography was by the descending
technique and was carried out with unwashed Whatman No. 3MM
paper at 22° unless otherwise stated. The paper used with
solvent D was pretreated with 2.5% neutralized polyethylenimine
(46). The following solvents were employed: (A) 95% aqueous
ethano1-l.0 M ammonium acetate, pH 7.5 (7:3 v/v), (B) ethyl
acetate-HZO-acetic acid-formic acid (18:4:3zl, v/v), (C) 1-
butanol-acetic acid-H20 (4:1:5, v/v, upper phase) and (D) 0.3

M LiCl. Radioactivity was detected on chromatograms with a
Packard radiochromatogram scanner, Model 7201 (Packard Instru-
ment Co.). All other radioactivity measurements were made

with a Packard Tri-carb liquid scintillation counter, Model
3310, employing either (A) a scintillation solution made as
described by Bray (84) or (B) 2,5-bis[2-(5-tggt-butylbenzoxazolyl)—
thiophene in reagent grade toluene (4 gm/l). When using solu-
tion B the portion of the chromatogram containing a radioactive
compound was cut out and completely immersed in solution B in

a scintillation vial and counted. The counting efficiencies

119

with solution A and B were 79 and 60%, respectively. Protein
was determined by the biuret method (58) through the ammonium

sulfate step. Thereafter it was determined by the Lowry
Rosebrough, Farr and Randall procedure (59). Crystalline'

bovine serum albumin was used as a standard.

Definition of Units. -- A unit of parsley UDPGA cyclase or
duckweed UDPGA cyclase is defined as the amount of enzyme
required to catalyze the formation of l umole of UDP-apiose
per minute at 25° from UDP-glucuronic acid under the condi-
tions of the standard assay. A unit of parsley UDPGA
decarboxylase or duckweed UDPGA decarboxylase is defined as
the amount of enzyme required to catalyze the formation of

1 umole of UDP-xylose per minute at 250 from UDP-glucuronic
acid under the conditions of the standard assay.

Standard Assays I and II for Parsley UDPGA Cyclase. -- Assay
I for parsley UDPGA cyclase is based on the ability of UDP-
apiose to undergo an intramolecular phosphorylation. After
UDP-[U-luc]apiose was formed it was hydrolyzed at a basic pH
to UMP and a-D-[U-luC]apio-D-furanosyl cyclic-1:2-P. The
latter compound was then isolated and measured. The standard
assay follows. To a 12 ml conical centrifuge tube were added
2.5 nmoles of sodium phosphate buffer pH 8.0, containing

2.5 nmoles of EDTA; 2.2 nmoles of UDP-[U-luC]glucuronic acid
(55,000 dpm, D-glucuronic acid portion uniformily labeled
within 315%); 45 nmoles NAD" and parsley UDPGA cyclase to
make a final volume of 50 ul. The addition of UDPGA cyclase
started the reaction. The mixture was incubated at 250 for

5 minutes and heated at 1000 for 3 minutes. The protein was

120

removed by centrifugation and the supernatant was applied to
Whatman No. 3MM paper. The protein precipitate was suspended

in 100 1 of 70% ethanol and removed again by centrifugation.
The supernatant from the ethanol wash was applied on the

same 2 cm wide portion of Whatman No. 3MM paper and chromato-
graphy was carried out with solvent A. After chromatography

the radioactive compounds, including a-D-[U-lucjapio-D—furanosyl
cyclic-1:2-P, were located and measured as described in the
General Methods.

Assay II for parsley UDPGA cyclase is based upon the acid
lability of UDP-apiose at pH 2.0. After UDPGA cyclase catalyzed
the formation of UDP-[U-lucjapiose, the sugar nucleotide was
hydrolyzed at pH 2.0. The resultant D-[U-lucjapiose was iso-
lated and measured. An assay mixture was prepared as described
in Assay I. After the addition of the enzyme to start the
reaction, it was incubated at 25° for 5 minutes and the assay
mixture was placed in an ice bath (see Fig. 8). Water, 0.1
umole D—apiose, 0.1 nmole D-xylose and 9 ul of 1 N H230“ were
added to make a final volume of 125 ul. The mixture was heated
at 1000 for 20 minutes and applied to Whatman No. 3MM paper.
Even though no protein precipitate was observed, the conical
centrifuge tube was washed with 100 ul of 70% ethanol which
was also applied to the same portion of the Whatman No. 3MM
paper. After chromatography with solvent B the radioactive
portion containing D—[U-luc]apiose was located and measured
as described in the General Methods.

Assay I and II were compared by preparing identical assay
mixtures and then measuring the parsley UDPGA cyclase activity

121

by both assays. Assay II measured an.average of 94.5%‘

of the parsley UDPGA cyclase activity measured by Assay I.

The 94.5% is the average of six experiments whose values were
90.0, 90.1, 91.4, 94.4, 99.9 and 100.8% of the activity mea-
sured by Assay I. The 100.0% value of Assay I is the average
of six experiments whose values were 94.4, 94.4, 95.7, 102.5,
105.6 and 107.0%. Some of the discrepancy between Assay I
and Assay II is due to the small amount of UDP-[U-luc]apiose
which phosphorylates intramolecularly to form a-D-[U-lucjapio-
D-furanosyl cyclic-1:2-P during the incubation period. This
compound is much more stable to acid hydrolysis under these
conditions than is UDP-[U-luCJapiose. More than half of the
a-D-[U-luc]apio-D-furanosyl cyclic-1:2-P is not hydrolyzed

to free D—[U—luc]apiose during the course of the incubation
at 1000 and pH 2. The amount of d-D-[U-lucjapio—D-furanosyl
cyclic-1:2—P which is not hydrolyzed to D-[U-lucjapiose migrates

with a different R than D-[U—luCJapiose in solvent B. The

F
stable product of partial hydrolysis of d-D-[U-luc]apio-D-
furanosyl cyclic-1:2-P, D-[U-luC]apiose-2-phosphate, also

2

migrates with a different R than D-[U-lucjapiose.in solvent B.

F
For this reason they are not measured and the amount of UDP-

[U-lucjapiose measured is less than the amount formed.

Standard Assays I and II for Duckweed UDPGA Cyclase. -- Duckweed
UDPGA cyclase was measured as described for parsley UDPGA
cyclase in Assays I and II.

Standard Assay for ParsleyUDPGA Decarboxylase. -- Parsley
UDPGA decarboxylase was measured with Assay II for parsley

UDPGA cyclase. The radioactive compound migrating with the

122

RF of D-[U-lucjxylose was isolated and measured as described
in the General Methods. The same assay was used for measuring
radioactive D-[U-luC]apiose and D-[U-lucjxylose.

Standard Assayrfor DuCkweed UDPGA Decarboxylase. -- Duckweed
UDPGA decarboxylase was measured as described for parsley
UDPGA decarboxylase.

Identification of the Products Formed by Parsley UDPGA Cyclagg
and ParslgyUDPGA Decarboxylase. -- The identification of the
two compounds formed by parsley UDPGA cyclase and parsley
UDPGA decarboxylase is based upon the comparison of the char-
acteristics of these compounds with the characteristics of
UDP-apiose and UDP-xylose formed by duckweed UDPGA cyclase
and duckweed UDPGA decarboxylase. The characterization of
these later compounds has been described elsewhere (Part 2

and footnote 2).

The products of parsley UDPGA decarboxylase and duckweed
UDPGA decarboxylase were both prepared from UDP-[U-luc]glu-
curonic acid. The products both yielded D-[U-1#C]xylose
after treatment at 1000 for 15 minutes in 0.01 N hydrochloric
acid. Degradation under these conditions which results in the
liberation of a sugar is characteristic of a sugar nucleotide
(62,68). The products of parsley and duckweed UDPGA decar-
boxylase migrated with the same RF as UDP-xylose when chroma-
tographed in solvents A, C and D (Part 2). Neither compound
functions in the reaction catalyzed by apiin synthase (Part 2).
Neither of these compounds phosphorylates intramolecularly
after 5 minutes at pH 8.0 and 100°. 0n the basis of these

data the product of the reaction catalyzed by parsley UDPGA

123

decarboxylase was concluded to be UDP-xylose.

The products of parsley UDPGA cyclase and duckweed UDPGA
cyclase were both prepared from UDP-[U-lacjglucuronic acid.
The products both yielded D-[U-14C]apiose after treatment at
1000 for 15 minutes in 0.01 N hydrochloric acid. After heating
at 1000 and pH 8.0 the products of parsley and duckweed UDPGA
cyclase migrated with the same RF as authentic d-D-[U-lucjapio-
D-furanosyl cyclic-1:2-P when chromatographed in solvents A,
C and D. Acid hydrolysis of the compound with anRF of d-D-
[U-luc]apio-D-furanosyl cyclic-1:2-P yielded only D-[U-lucj-
apiose. Both products function in the reaction catalyzed by
apiin synthase which results in the formation of a compound
with an RF of apiin. Acid hydrolysis of the compound with
an RF of apiin yielded only D-[U-1“C]apiose (Part 3). Based
on these data UDP-apiose was identified as the product of the
reaction catalyzed by parsley UDPGA cyclase.

RESULTS

DISTRIBUTION OF UDPGA CYCLASE IN THE LEAVES,
STEMS AND ROOTS OF PARSLEY

The relative amounts of UDPGA cyclase in roots, stems
and leaves of mature parsley plants were determined by extrac-
tion of protein from these tissues. The extractions were
conducted as described in the extraction step of the purification
of UDPGA cyclase from parsley leaves. After centrifugation
for 20 minutes at 14,500 x g, the supernatant and the precipi-
tate were collected and assayed. A summary of the analyses of

the supernatant for UDPGA cyclase is presented in Table I. The

.oacaanoan as fleas ooo.oaav

UHom oHsoasosHmmunHuDuudop meHoe: «.0 :sz H mmww< :H coasmmoe we: ommHoho «ame:

.Ammv asomdoa poasan spa

poses co: cap pH oosHm :Hopoan no: no: copprnaooan do: :OHns

HsHampse on» no pmoz .oeasmdee no: apH>Huod ommHoho «can: use seamen sH voosonSm
tea was oespHaHoeaa one .moussHe om pom w x oom.:H pm sofipmwsmdapsoo eon“ mumpaaaoeamt

 

no.0 maoo.o
m 2.0 88.0
1

om.a moao.o

am.a amoo.o

sHopoaQ we
mpHasHHHHe mpHesdHHHe

 

on: topmpHaHooaa
mom peopdsaonsm

poem
53H eopm
OMOH mama
we

 

canons am ooa
acts spaaapod Hoaoa aaaaaaoo oaaaooam

saunas aw ooa
eoam :Hopoan Hopes oemmas

 

.monamQ mo muooa use meopm .mopon on» :«

ondaoao «can: so soapsndapaan .H mamas

125

leaves are the best source of UDPGA cyclase activity although
some UDPGA cyclase activity is observed in the stems and less
in the roots. No UDPGA cyclase was observed in the resuspended

precipitate from the stems or leaves.

PURIFICATION OF PARSLEY UDPGA CYCLASE

A typical purification of parsley UDPGA cyclase is detailed
below and summarized in Table II. All procedures were carried
out at 0-50 unless otherwise stated.

Extraction. -- The extraction step was carried out exactly as
described in the purification of apiin synthase in Part 3.
Ammonium Sulfate. -- The ammonium sulfate step was carried out
exactly as described in the purification of apiin synthase in
Part 3.

Sephadex G-100. -- The Sephadex G-100 step was carried out
exactly as described in the purificationof apiin synthase in
Part 3. Most of the parsley UDPGA cyclase was eluted into 6
tubes, numbers 38-43. Their contents were combined and used
in the DEAE-Sephadex step.

DEAE-Sephadex. -- A column 10.0 cm in height and 2.2 cm in
diameter was prepared from defined DEAR-Sephadex. The DEAE-
Sephadex was treated as follows: (1) with 0.1 M sodium phos-
phate, pH 7.4 until the pH remained at 7.4 (2) 4 times with
400 ml of 0.01 M sodium phosphate, pH 7.4, containing

1 x 10'“ '2

M EDTA and l x 10 M a-mercaptoethanol. Just before
use the column was equilibrated with 500 ml of the above buffer.
Half of the material from the Sephadex G-lOO step was applied.

After application the column was washed with 60 m1 of 0.1 M

126

NaCl in buffer. UDPGA cyclase was then eluted by increasing
the NaCl concentration linearly. The linear gradient was set
up with 200 m1 of 0.3 M NaCl in buffer in one beaker and 200
ml of 0.1 M NaCl in buffer in the mixing beaker. Fractions

of 9.2 ml were collected. Almost all of the parsley UDPGA
cyclase eluted into 8 tubes, numbers 23-30, which were combined
and concentrated by ultrafiltration to about 1.0 ml. After

2 hours dialysis against buffer 1.0 ml of glycerol was added.
Enzyme prepared in this manner constitutes the enzyme charac—
terized below. UDPGA cyclase was purified 93 fold and was
almost free from apiin synthase. More than 98% of the apiin
synthase present in the cell-free extract (extraction step)
was removed from the UDPGA cyclase after purification through
the DEAE-Sephadex step. Much of the UDPGA decarboxylase
activity present tithe cell-free extract was not removed from
UDPGA cyclase by purification through the DEAE-Sephadex step
(Fig. 4, 10).

Linearity of the Parsley UDPGA Cyclase Reaction. -- The data
in Fig. 1 show that the reaction catalyzed by parsley UDPGA
cyclase is linear with two different concentrations of
UDP-[U-lnc]glucuronic acid. It is linear for more than four
minutes at the lower concentration and for a longer time at
the higher concentration. The data in Fig. 2 show that the
parsley UDPGA cyclase reaction velocity increases linearly
with increasing concentrations of UDPGA cyclase. When parsley
UDPGA cyclase from the ammonium sulfate step was assayed
(after dialysis for 3 hours which removed the ammonium sulfate)

the formation of UDP-apiose was not completely proportional to

127

as waoneez .mamzHcad use scaumapHHhmApHs an :oHumppsooeoo scams easemmoe mm: mpH>Hpo<

.denumeo mHso mo omopmsH dope xmemsaomum<mn esp 2H come soon we:
deem OOHuo addendum one ea oanomo «can: one do HHm ma mm se>Hw was none woemnaomum<ma esp

.sHopoaa

tit...
mo opmpHaHomaa o no moscamoaam esp

an emceeea>e mm oomHa xoou condo sHopoaa no sowpmaepmeoe meow momoaasn memes pom pom: noun
opmMHsm enanoeem on» no sowpaoa HHmEm on» no mHmmHsHU waHasn .aopm OOHIG woumnaem 0:» seems

apa>apom on» use» nosoH hHHmsmz was dope oumMHsm esHsoeem on» ad hpa>auom omeoho «can:

***

.H ammm< sH condense mm: amen woumsaomnm<mn esp :H mpH>apom omeoao «can:

.opmaumDSm we done we: Asap ooo.msv pace 0HeoazoeHmmo Ha
popaaomoo mm doasmmoe mm: amen OOHno seemsaom on» dad
emmmHoao «can: .oaos oasoasosHMnmoD mo pdsoed esp ma m use zpflooHo> do>aomno on» ma > open:

A ex\m.+HV> u > seaweeds on» seed hpaOOHo> HsHpHsH mm wommoaaxo ma madam meow HHm pom mpH>apo<
.He pea mua>Hpom do useoed 30H 0:» mo bassoon come we: “end ooo.HHv uHow OHsOLSOSkumQD

moHoe: N.o paooxm H admm< pom confluence we deem :oHuomapxo on» sH coasmmoe mm: emmHomo «amazes

Dunmaa moHoee mm.o paooxo H amwm< now
done opMMHSm esfisoeew on» :H th>apom

.aepm soapOmApxm

0:» pm H on Hence pom some one: ommnpShm swans no use omcHoho «omen mo apH>Hpod loHoonr

 

 

 

0H0.0 mm ma ~.0 0m.H sassxoodsaom-m<ma
mma.0 as w.na 0am m~.0 00auo wooasaom
0.0 mm m.H mmma 00.0 assosdcasn eaasoaaa
0.H 00H a mnam No.0 assoaaodtaxm
a sacs we m5\naasaAHHas
ommwoao some: saoaoaa assessor
sonosasam saam< sacs» soapdoaaaaam Hence oaeaooam aopm

 

.mo>mmH monama eoam ommHoho <wmap mo seameHMHaem .HH mqm<e

128

Fig. 1. Formation of UDP-apiose by parsley UDPGA cyclase
as a function of time. The formation of UDP-apiose was
followed using Assay I with either 1.71 nmoles (55,000
dpm) ( O O ) or 0.528 nmoles (55,000 dpm)

( o O ) of UDP—[U—lucjglucuronic acid. The
reactions were initiated by the addition of 0.08 milli-
units of parsley UDPGA cyclase (40.5 ug of protein) from
the DEAE-Sephadex step. The reactions were incubated for
the indicated time (1 to 5 minutes) and further treated
as described for Assay I.

 

 

 

129

 

 

 

l l
V. cw
O ' o‘

OJWUOi 350 NV '1! G n

ssnoww

MINUTES

13)

Fig. 2. Effect of parsley UDPGA cyclase concentration

on reaction velocity. The formation of UDP-[U—lchapiose
was followed with Assay I. The assays were initiated by
the indicated amounts of parsley UDPGA cyclase (2.1 milli-
units/mg of protein) from the DEAE-Sephadex step.

131

 

 

 

 

 

..‘Q
o'

0.075 *-

ilnNIW/CI inOi 350M v-dan Si'IOWN

ROTEIN

HG

132

the amount of UDPGA cyclase present. However, after passage
through a Sephadex G-25 or G-lOO column there was a direct
proportionality between the amount of UDPGA cyclase present

in the assay and the amount of UDP—apiose formed in a
specified time period.

Factors Affecting the Stability of Parsley UDPGA Cyclase. --
Stability of parsley UDPGA cyclase to storage and dialysis
largely depended on protein concentration, temperature and
addition of glycerol; the higher the protein concentration

and the lower the temperature without freezing the extract,
the better the stability. When UDPGA cyclase from the Sephadex
0-100 step was stored at 40 (4-5 mg of protein/ml) for 12
hours it lost almost all activity. When UDPGA cyclase was
concentrated by ultrafiltration and made 50% glycerol by
volume enzyme activity was preserved during storage at -200
which did not freeze the solution. This was especially evi-
dent in more dilute solutions (0.1 to 10.0 mg of protein/ml).
Storage of this material at -200 without glycerol resulted

in its freezing. After 24 hours at 0200 no enzymatic activity
was observed. UDPGA cyclase from the DEAE-Sephadex step (13.4
mg of protein/ml) was made 50% glycerol by volume and stored
at -20°. This treatment preserved 98% of the starting activity
after 11 days as measured by Assay I and 91% after 19 days.
However, when UDPGA cyclase from the DEAE-Sephadex step (10 mg
of protein/m1 and 30% glycerol by volume) was kept at 40 and
25°, enzymatic activity was rapidly lost. Storage of this
material at 250 preserved about 15% more enzyme activity than

at 40 throughout a 24 hour period as measured in Assay I. After

133

7.5 hours at 250 and 4°, 70 and 80%, respectively, of the
activity had been lost.

pH Optimum of Parsley UDPGA Cyclase. -- Parsley UDPGA cyclase
has optimum activity between pH 8.0 to 8.3 (Fig. 3). In this
range the enzyme had the same activity with either of the
buffers used. As indicated in Fig. 3 UDPGA cyclase activity
decreased rapidly at pH values below pH 8-0 t0 8.3.

Affinity of Parsley UDPGA_§yclase for UDP-glucuronic Acid. ~-
The Km of parsley UDPGA cyclase for UDP-glucuronic acid was
calculated from the data presented in the insert in Fig. 4

by the least squares method. It is 3.3 x 10'6 M.

Effect of Various Ions on the Activity of Parsley UDPGA
Cyclase. -- UDPGA cyclase activity was measured in the presence
of various ions. The final concentrations of these ions in
Assay I was 1 mM. The compounds tested were FeClZ, FeCl3,

CaCl MnCl

MgCl NaCl, KCl, NaNoOB, NaBOB, C0012, NH4C1,

29
and CuCl

2' 2’
2’ 2 2'

enzyme activity and the latter four decreased the enzymatic

ZnCl NiCl None of the above compounds stimulated
activity by 30, 65, 70 and 85%, respectively. EDTA, a,a'-
dipyridyl and KP at 1 mM final concentrations in Assay I had
no effect on parsley UDPGA cyclase activity. The addition of
o-phenanthroline and 8-hydroxyquinoline at the same concentra-
tion resulted in a slight (6%) decrease in enzyme activity.
The effect of the above 19 compounds on the activity of UDPGA
cyclase was determined in Assay I with the sodium phosphate
buffer replaced by 2.5;umoles of Tris-HCI, pH 8.0.

Effect of Sulfhyggyl_Reagents on the Activity of Parsley UDPGA

Cyclase. -- UDPGA cyclase was incubated with 1 mM final

134

Fig. 3. Effect of pH on the reaction rate of parsley UDPGA
cyclase. UDPGA cyclase was measured in Assay II with 2.4
nmoles of sodium phosphate buffer ( O O ) pH 6.0
to 8.4 or 2.4 nmoles of Tris-RC1 buffer ( I ------------- I )
pH 6.0 to 9.0. The reaction in each assay was initiated

by the addition of 0.0075 milliunits of parsley UDPGA
cyclase (10 pg of protein) from the DEAR-Sephadex step.

 

135

 

 

'e

 

100 '-

, i

I
O
in

All/“13V JAIIV'HII

 

136

Fig. 4. Effect of UDP-glucuronic acid concentration on
the velocity of the parsley UDPGA cyclase reaction. The
assay conditions were those described in Assay I except
that the UDP-glucuronic acid concentration was varied as
indicated. Each assay was initiated by the addition of
0.056 milliunits of parsley UDPGA cyclase (40.5 ug of
protein) from the DEAR-Sephadex step. The insert is a
Lineweaver-Burk plot of the same data.

137

 

:23 ._.M.:u< u_zo~3u=._onaomn._
e

o.

to _
_

mafia 2an: 2: e... an a“?

$23

    
   

l

2 muo— x and u .00.

6a

 

 

 

 

 

 

5.0

00.0

m 0.0

(JIHNIW/SB'IOWN) AIIDO'HA

138

concentrations of pechloromercuribenzoate, N-ethylmaleimide,
iodoacetamide and oxidized glutathione in Assay I. More

than 90% of the activity was lost with p-chloromercuribenzoate
and N-ethylmaleimide (99 and 94%), 15% with iodoacetamide and
8% with oxidized glutathione.

Determination of Optimum NADl’Concentration for the Formation
of UDP-Apiose. -- Present with the parsley UDPGA cyclase from
the DEAR-Sephadex step was some of the parsley UDPGA decarboxy-
lase present in the cell-free extract of the extraction step.
The data in Fig. 5 show that parsley UDPGA cyclase has an
absolute requirement for exogenously supplied NAD'I. No UDP-
apiose is formed in the absence of exogenous NAD+. The NAD+
concentration for optimum UDP-apiose formation was 1.0 to 2.0
x 10-3 M.

Inhibition of Parslgy UDPGA Decarboxylase and Parsley UDPGA
Cyclase by UDP-Xylose. -- The activity of parsley UDPGA
cyclase and parsley UDPGA decarboxylase were measured with
Assay II in the presence of various concentrations of UDP-
xylose (Table III). UDPGA decarboxylase activity was in—
hibited more at high concentrations of UDP-xylose than UDPGA
cyclase activity. UDPGA cyclase activity was slightly sti-
mulated at lower concentrations of UDP—xylose.

The Energy_of Activation of Parsley UDPGA Cyclase. -- The

energy of activation (Ea) was determined from the equation (85):

‘E = slope x 2.303 (R)
The value for the slope was obtained from Fig. 6. R equals
1.987 cal"1 (KO) (g-mole)"1. Ea is 12.5 Kcal.

139

Fig. 5. Effect of NAD+ concentration on parsley UDPGA
cyclase reaction velocity. The formation of UDP-[U-lucj-
apiose was followed with Assay I. The reaction in each
assay was initiated by the addition of 0.0085 milliunits

of parsley UDPGA cyclase (12.1g of protein) from the DEAE-
Sephadex step to each assay mixture containing the indicated
amounts of NAD+ (0.0 to 500 nmoles).

140

 

 

 

100 _

All/\IIDV

3AI1V'I an

 

IO

5

[NAoj . (mM.)

141

Fig. 6. Effect of temperature on parsley UDPGA cyclase
activity. Parsley UDPGA cyclase from the DEAE-Sephadex
step contained 0.015 milliunits per 20 ug of protein.
Each assay was initiated by this amount of UDPGA cyclase.
The assays were incubated for 5 minutes at the various
temperatures indicated. The reactions were heated at
1000 for 3 minutes and further treated as described for
Assay I. The initial velocity of the reaction, Va, is
defined as the milliunits of UDPGA cyclase activity mea-
sured in Assay I at the various incubation temperatures
tested in this experiment.

142

JIHNIW/OJWHO:I asouv-dan $31OWN

e
8
1

0.003

 

’0’
12 5 ch

4 .

 

 

 

 

100 -

ALIAUDV SAIIV'HII

50)-
o

 

25

TEMPERATURE °c

143

TABLE III. Inhibition of parsley UDPGA decarboxylase and
parsley UDPGA cyclase by UDP-xylose

 

Molar concentration of UDP-xylose in Assay

 

l x 10"6 l x 10-5 1 x 10-8 l x 10“3

UDPGA
decarboxylase 100 93 76 29
activity

 

UDPGA
cyclase 100 113 100 54
activity*

 

*Activity compared to that obtained when no UDP-xylose was
added to the assay mixture and is equal to 100. UDPGA cyclase
was assayed with Assay II and UDPGA decarboxylase was assayed
with the standard assay in the presence of various concentra-
tions of UDP-xylose.

 

Fractionation of ParsleyUDPGA Decarboxylase and Parsley UDPGA
Cyclase Activity with Ammonium Sulfate. -- To determine the
concentrations of ammonium sulfate needed to precipitate each
of the two enzymes, leaf tissue was prepared and treated as
described in the extraction step. It was then fractionated
with ammonium sulfate, as described in Part 3. A summary of
the analyses for UDPGA decarboxylase and UDPGA cyclase acti-
vity in the various fractions is presented in Table IV. As
it shows, 3 of the fractions contained both UDPGA decarboxy-
lase and UDPGA cyclase. The ratios of the two activities

did not vary significantly in each of these fractions. Two
fractions, 1.25-ail.50 and 1.95-9 2.10, contained 5% of the
total UDPGA decarboxylase activity. They contained no UDPGA

.mpcapwnsm we oHom 0Heoasousmo Hipunmap mo moHoes mo.H cosHspsoo
.HH momm< :H vehemdce we: ommHOho «can: use zones deem

madame zoom

swam exp :H essences mm: ommHaxocaccme <omaza

 

144

(

 

 

 

 

 

00.0 0000.0 00.0 000.0 me Am.Hous.mev oo.m +IOH.N
00.0 0000.0 00.0 000.0 0m an.aaua.eav 0H.~ cum0.H
0H.m 0m00.0 mm.0 mmH0.0 «no AH.0aum.mav 00.H 4:00.H
0m.s mom0.0 no.0a maa.0 ame Am.man0.0mv 0m.a «.me.a
mo.H . nmmo.o oo.H omo.o om Ao.mmum.emv mo.H atom.H
00.0 0000.0 mm.0 000.0 0H An.amnm.0m0 00.H atmm.a
Aoa pm opMMHSm
mpHssHHHHe sHopoaa do we mpHssHHHHe cHouoaa do we esHsoeed mo opmMHSm eeHsoeem
Hence muHssHHHHe Hence muassHHHHe we :oHpsHom my no mpHamHoe
eemmHozo «can: sommHaxocamooc «man: sHopoaa Hence deem
.opmMHSm enamoeem
Spa: omcHomo «can: monama use ommHaxonamome «can: monama mo soHpcSoHpowam ">H mqm<e

145

cyclase activity. Kindel* observed that 13% of the UDPGA
decarboxylase activity could be separated from the majority of
the UDPGA cyclase activity with ammonium sulfate in a cell-

free extract from duckweed.

PROPERTIES OF PARSLEY UDPGA DECARBOXYLASE
pH Optimum of Parslgy_UDPGA Decarboxylase. -- Parsley UDPGA
decarboxylase has optimum activity at about pH 8.0 to 8.2
(Fig. 7). As indicated in Fig. 7, UDPGA decarboxylase acti-
vity decreased rapidly at pH values above or below pH 8.0
to 8.2.
Determination of Optimum NAD+ Concentration for the Formation
of UDP-xylose. -- Fig. 8 shows that parsley UDPGA decarboxy-
lase has 65% of optimum activity without exogenous NAD+.
However, exogenous NAD+ does stimulate the UDPGA decarboxylase
showing that exogenous NADI'is required for Optimum activity
of this enzyme in yitgg. The NADl'concentration for optimum
UDP-xylose formation is about 2.0 x 10-3 M. More than opti-
mum concentrations of NAD +suppressed enzymatic activity for
both UDPGA cyclase (Fig. 5) and UDPGA decarboxylase, although
the activity of UDPGA cyclase was suppressed much more than

was the activity of UDPGA decarboxylase.

*xindel, P. K., unpublished data

146

Fig. 7. Effect of pH on the reaction rate of parsley UDPGA
decarboxylase. UDPGA decarboxylase was measured in the
standard assay with 2.4 umoles of sodium phosphate buffer

( O O ) pH 7.0 to 9.0 or 2.4 umoles of Tris-HCl
buffer ( o - ------- - ------- O ) pH 7.0 to 9.0. Each assay was
initiated by the addition of 0.0075 milliunits of parsley
UDPGA cyclase (10 pg of protein) from the DEAR-Sephadex
step which contained 0.007 milliunits of parsley UDPGA
decarboxylase. ‘

 

147

 

 

AllAllDV aAIlvna

 

148

Fig. 8. Effect of NAD+ concentration on parsley UDPGA
decarboxylase reaction velocity. The formation of UDP-
[U-14C]xylose was followed with the standard assay except
that each assay mixture contained one of the concentra-
tions of NAD*'indicated in Fig. 5 (0.0 to 500 nmoles).

The reaction was initiated by the addition to each assay
mixture of 0.009 milliunits of parsley UDPGA decarboxylase
(12 ug of protein) from the DEAR—Sephadex step which con-
tained 0.01 milliunits of UDPGA cyclase.

149

 

"\I/J

 

 

 

100 -

AllAllDV

50"

SAHV'IJU

1
”in

5.0

Email (mm)

2.0

1+
1.0

 

150

PROPERTIES OF DUCKWEED UDPGA CYCLASE
AND DUCKWEED UDPGA DECARBOXYLASE

Factors Affecting the Stability of Duckweed UDPGA Cyclase. --
Gustine (41) observed using a modified Assay I which was
incubated for 15 minutes that duckweed UDPGA cyclase lost
activity very rapidly (24 hours) when stored in dilute solu-
tion at -200 or at 4°. The addition of glycerol to make a
final concentration of 50% by volume stabilized the enzyme
especially in dilute solutions at -2o°. After 39 days of
storage with these conditions Gustine (41) found that there
was 3.7 times as much activity remaining in the solution
which was 50% glycerol by volume (34% of the starting acti-
vity) compared to the solution without glycerol.

I investigated the stability of duckweed UDPGA cyclase
from the DEAE-Sephadex step (10 mg protein/ml and 30% gly-
cerol) that was kept at 40 and 25°. Storage at 25° preserved
about 10% more enzyme activity than at 40 throughout a period
of 24 hours. Storage at 4° with 10 mM NAD+ both slowed the
rapid loss of enzyme activity during the first 3 hours and
slowed the loss of enzyme activity throughout the remainder
of the 24 hour incubation period. After 24 hours only 35%
of the original activity remained (Fig. 9) whereas 50% re-
mained in the presence of 10 mM NAD‘I
Effect of Variops Ions on the Activity of Duckweed UDPGA
fllclase. -- Duckweed UDPGA cyclase activity was measured in
the presence of various ions. The final concentration of
these ions in Assay I was 1 mM. The compounds tested were
FeCl MnCl NiCl NaBO

CaClz, MgCl NaCl, KCl, NaMoO

3' 2' 2' 2' 3’ 3’

151

Fig. 9. Effect of NAD*'on the stability of duckweed UDPGA
cyclase at 4°. Duckweed UDPGA cyclase from the DEAR-Sephadex
step (10 mg of protein/ml) was stored at 4° in a solution
25% glycerol by volume. One portion was 10 mM with respect
to NAD+ (O ........... O ...............o) and the other portion con_
tained no NAD“ (O . -— 0). After incubation at
40 for the time intervals indicated (2,4,8,19 and 24 hours)
0.059 milliunits of UDPGA cyclase (60;;g of protein) was
removed and measured with Assay I. NAD+'was added to the
assay mixture before enzyme was added from each portion so
that the final concentration of NADt in all assay mixtures
was the same (2 x 10"3 M). The assays were begun by addi—
tion of enzyme and treated as described in Materials and
Methods.

 

 

 

 

152

mdDOI ..

 

 

 

 

«a a. 0
I1 .7
JON
f/
0.... e
........... /e .
...o ............. Lee
0 .................... C/Iiow
too.

3A|1V1 38

AllAllDV

153

NHnCl, FeCl ZnCl CoClz, and CuClZ. None of the 14 com-

2’ 2’
pounds test increased duckweed UDPGA cyclase activity except
MgCl2 (21%). Gustine (41) tested duckweed UDPGA cyclase with
only MgCl2 and observed a 15% increase in enzymatic activity.
Only the latter five compounds decreased the enzymatic acti-
vity by 9. 17, 76, 85, and 100% respectively. EDTA, d,d';
dipyridyl and KF at 1 mM final concentrations in Assay I had
no effect on duckweed UDPGA cyclase activity. The addition
of o-phenanthroline and 8-hydroxyquinoline at 1 mM levels
resulted in a slight (10%) decrease in enzymatic activity.
The effect of the above 19 compounds on the activity of UDPGA
cyclase was determined in Assay I with the sodium phosphate
buffer replaced by 2.5 vmoles of Tris—HCl, pH 8.0.

Effect of Sulfhydpyl Reagents on the Activity of Duckweed
UDPGA Cyclase. -- UDPGA cyclase was incubated with 1 mM con-
centrations of pbchloromercuribenzoate, N-ethylmaleimide,
iodoacetamide and oxidized glutathione in Assay I. More

than 95% of the activity was lost with p—chloromercuriben-
zoate and N-ethylmaleimide, 60% with iodoacetamide and 6%
with oxidized glutathione.

The Energy of Activation of Duckweed UDPGA Cyclase. -- The
energy of activation for duckweed UDPGA cyclase obtained

from the equation above and from the data in Fig. 10 is

11.4 Kcal.

Inhibition of Duckweed UDPGA Cyclase and Duckweed UDPGAADecarboxy_
lase Activities by Five Uracil Compounds. -~ The,two enzymes
from the DEAE-Sephadex step were measured by Assay II in the

presence of five uracil compounds (Table V). All five

154

Fig. 10. Effect of temperature on duckweed UDPGA cyclase
activity. Duckweed UDPGA cyclase from the DEAE—Sephadex
step contained 0.016 milliunits per 5 pg of protein. Each
assay was initiated by this amount of UDPGA cyclase. The
assays were incubated for 5 minutes at the various temper-
atures indicated. The reactions were heated at 1000 for

3 minutes and further treated as described for Assay I.
The initial velocity of the reaction, v0, is defined as
the milliunits of UDPGA cyclase activity measured in Assay
I at the various incubation temperatures tested in this
experiment.

155

310 NIWIOSWUOH asoudv-aan SB‘IOWN

 

 

50

35
TEMPERATURE °c

20

 

 

 

 

AllAllDV" aAIivuu

0 V
a §
5? , :I—r
O/--
r .. J
\
O
O
I 2
l
O
O In

156

.oopmop no: n 92 .ooH mo o=Hm> on» seeHm mm: sonz ommHozo <omaa pom HH momm< on
so ommmeonamooc <omop new names pudendum on» o» peeve mm: messoaeoo
HHoma: o>oom one no meow first oosHoch apH>Hpom as» on coadaeoo mH th>Hpo<¢

 

1(1)

 

 

an 0s 00 sHH, oaoHkuaae

so «0 mm mm oaoeodchuaae

we :0 mm mm as:

me a: 00 an an:

92 mm :0 as as:
ma0H x H anoH x H mu0H w H 0-0H w H
muesoaeoo HHodas cocooHom mo meodeausoosoo amHoe

wstoHHoe one ceH: oosadaao aaHaHeoo soeoHoao «can: oooseosa essoaaoo

NH an H0 00 oeoHHwiaae

an 0: mm 00 esopooHownmne

00 we so 00 as:

0 0 mm as an:

92 a an no as:
muoH w H eu0H w H mu0H a H 0-0H w H

 

mossoaeoo HHodas douooHom no msodeapsoosoo adHoe weHonHom
one 59H: doeHmpno sth>Hpom odehxocamooe <cmas coosxoea pesoaeoo

 

.mdesoaeoo HHoda: o>Hm an
oanomo <6mnm coozxoep use odeawonhdooc «ame: coozxosd mo eoHpHansH u> mqm<e

157

compounds inhibited both activities. UTP and UDP inhibited
both activities much more at lower concentrations than did
the other three compounds. UDP-xylose inhibited UDPGA
decarboxylase at high concentrations greatly, whereas UDP-
galactose did not. Inhibition of UDPGA cyclase activity

by these two sugar nucleotides even at high concentrations,
was much less than the inhibition of UDPGA decarboxylase

by UDP-xylose.

DISCUSSION

An important consideration in the purification of
UDPGA cyclase was to maintain the enzyme in dilute concen-
trations for a minimal amount of time. Maintainence of the
enzyme in dilute solutions (0.1—10.0 mg/ml) at 40 and 250
resulted in the rapid loss of activity.

Before UDPGA cyclase was isolated, Grisebach (84) sug-
gested that only one enzyme would be necessary to form
UDP-apiose and UDP-xylose from UDP-glucuronic acid because
of the similarity of the intermediates he postulated in
their formation. Subsequently Sandermann and Grisebach (40)
and Sandermann, Tisue and Girsebach (38) described the iso-
lation of cell-free extracts from parsley and duckweed
forming UDP-apiose and UDP-xylose. The activities were not
purified and no attempt was made to separate the two activities.

Partially purified duckweed UDPGA cyclase and parsley
UDPGA cyclase were used in most of the experiments described
above. Both enzymatic activities also contained UDPGA

decarboxylase activity. The presence of two separate enzymes

158

one forming UDP-apiose and another forming UDP—xylose, was
indicated by several experiments. In one of these experi-
ments parsley protein was isolated and then fractionated by
increasing the concentration of ammonium sulfate. After
each increase in the concentration of ammonium sulfate the
protein which precipitated was isolated and assayed. All
of the parsley UDPGA cyclase was found in three of the six
fractions isolated as was much of the parsley UDPGA decarboxy—
lase. Some parsley UDPGA decarboxylase (5%) was found in two
of the other fractions which did not contain any parsley UDPGA
cyclase. Some separation of one of the enzyme activities
from the other has been obtained by fractionating with am-
monium sulfate protein isolated from duckweed. About 13% of
the UDPGA decarboxylase was removed from the majority of the
UDPGA cyclase activity*. Further evidence suggesting that
two enzymes exist in parsley was found by measuring the NAD+
requirement of the enzymes. Parsley UDPGA cyclase requires
NAD+ to form UDP-apiose. Parsley UDPGA decarboxylase
activity does not need exogenously added NAD+ to form UDP-
xylose. NAD'*(exogenously added) stimulates the formation
of UDP—xylose and the concentration required for optimal
UDP-xylose formation is similar to the optimal concentration
needed for UDP-apiose formation.

The inhibitory effect of UDP-xylose is quite similar
on both duckweed and parsley UDPGA decarboxylase. The in-
hibitory effect of UDP-xylose is quite similar on both duck-
weed and parsley UDPGA cyclase. The formation of UDP-xylose
is inhibited more than the formation of UDP-apiose by various

159

concentrations of UDP-xylose. In fact, the UDP-xylose for-
mation is inhibited as much as the UDP-apiose formation
with 10 to 100 times lower concentrations of UDP-xylose.

At concentrations of UDP-xylose where parsley UDPGA cyclase
and parsley UDPGA decarboxylase activity are partly inhibited
(30-88%), the parsley UDPGA decarboxylase activity is in-
hibited by up to twice as much. A similar observation was
made with these enzymatic activities isolated from duckweed
and purified through the DEAE-Sephadex step. All of the
uracil compounds tested, UTP, UDP, UMP, UDP-xylose and UDP-
galactose, inhibited duckweed UDPGA decarboxylase more than
UDPGA cyclase.

The data do not show conclusively that the UDPGA decar-
boxylase activity precipitating with UDPGA cyclase in duck-
weed and parsley extracts represents either co-precipitation
of UDPGA decarboxylase wflxlother protein or one enzyme which
forms both compounds. These data strongly indicate that
separate enzymes are involved in the formation of UDP-apiose
and UDP-xylose.

The inhibition by uracil compounds is greater on duck-
weed UDPGA decarboxylase that on duckweed UDPGA cyclase.
This should allow the formation and subsequent isolation of
UDP-apiose without contaminating UDP-xylose providing that
the uracil compounds are added in the proper amounts. Sub-
sequent removal of several of these uracil compounds could
be accomplished using the isolation procedure for UDP-apiose
described in Part 2., This is important as the preparation
of UDP-apiose without contaminating UDP-xylose has not yet

160

been possible (Part 2) .

It may be of physiological significance that UDP, a
product of transglycosylation reactions such as those pre—
dicted for cell-wall polysaccharide formation (65,68), in-
hibits duckweed UDPGA cyclase 41% and inhibits duckweed
UDPGA decarboxylase 57% at low concentrations (1 x 10'6 M).
This suggests that for transglycosylation reactions to pro-
ceed UDP must be removed. This could occur by conversion
of UDP to UTP and UDP-sugars at a sufficient rate to prevent
accumulation of UDP and its inhibition of UDP-xylose and
UTP-apiose formation. Besides duckweed UDPGA decarboxylase
and duckweed UDPGA cyclase, parsley apiin synthase (Part 3)
was inhibited significantly at low concentrations of UDP.

UMP also had a similar effect on all three enzymes (Part 3).

The intracellular function of UDPGA cyclase and UDPGA
decarboxylase is to catalyze the biosynthesis of UDP-apiose
and UDP-xylose, respectively. The low Km for UDP-glucuronic
acid suggests that it is the substrate for parsley UDPGA
cyclase. It would be of interest to determine the activity
of parsley and duckweed UDPGA cyclase toward UDP-galacturonic
acid as a possible substrate. Both UDP-glucuronic acid and
UDP-galacturonic acid, a possible substrate, are probably
present in parsley and duckweed (81). Nucleoside diphosphate
galacturonic acid is very likely formed in duckweed, as large
amounts of galacturonic acid have been found in the cell
wall polysaccharides of duckweed (l6). UDP-galacturonic acid
has been formed and shown to be a precursor for polygalacturonans

in Phaseolus aureus (43,44, 81).

161

Sandermann, Tisue and Grisebach (38) reported the iso—
lation of particulate UDPGA cyclase from duckweed. They did
not isolate any particulate UDPGA cyclase from parsley leaves.
I also did not observe particulate UDPGA cyclase from parsley
leaves and stems.

The requirement of parsley UDPGA cyclase for exogenously
added NAD+ has not been previously investigated. No synthe-
sis of UDP-apiose with parsley UDPGA cyclase purified through
the DEAE-Sephadex step was observed without exogenously
added NAD+. NAD+ was not required to obtain 65% of the
optimum parsley UDPGA decarboxylase activity observed in
the presence of exogenously added NAD+. Sandermann and
Grisebach (40) reported that duckweed UDPGA cyclase, which
had been purified only by passage through a Sephadex G-25
column, functioned without exogenously added NAD+ at 3% of
the rate obtained in the presence of NAD+. Sandermann and
Grisebach (40) observed that duckweed UDPGA decarboxylase
functioned at 17% of the rate obtained in the presence of
exogenously added NAD+. They also found that NADH inhibited
duckweed UDPGA cyclase and that NADP+ was much less effec-
tive than NAD+ as a cofactor in the UDPGA cyclase reaction.

The only significant differences observed between parsley
and duckweed UDPGA cyclase were the energy of activation
and the temperatures at which heat denaturation of the
enzymes began to occur. The latter is the temperature for
optimal activity of each enzyme. Parsley UDPGA cyclase
begins to be denatured by heat at 37.420 and.duckweed UDPGA

cyclase begins to be denatured by heat at 24.270. The

162

temperature for optimum activity of each enzyme is quite
distinctive. It is about 27° for duckweed UDPGA cyclase

and about 42° for parsley UDPGA cyclase. This shows that
the parsley UDPGA cyclase was more stable to heat in the
purified preparation tested. The only difference between
the two experiments was the source of the protein and the
protein concentration in the assay. These data also suggest
that UDPGA cyclase should be assayed at 250 where heat dena-
turation should not affect either enzyme rather than at 300
which the enzyme commission suggests as a standard tempera-
ture for most enzyme assays (87). Because of the low sta-
bility of duckweed UDPGA cyclase and of UDP-apiose (Part 2)
at temperatures above 250 it is suggested that all assays
for any UDPGA cyclase be carried out at 25°. The energy of
activation of duckweed UDPGA cyclase is lower than that of
parsley UDPGA cyclase. This means that less energy is neces-
sary to activate the substrate by formation of a substrate-
enzyme complex (53) in the reaction catalyzed by duckweed
UDPGA cyclase than the reaction catalyzed by parsley UDPGA
cyclase. Based on the above observations the two enzymes are
apparently not completely identical physically even though
they both catalyze reactions forming UDP-apiose from UDP-

glucuronic acid.

6.

7.
8.

9.
10.

11.

12.
13.

14.

15.
16.

17.

REFERENCES

Vongerichten, E., Ann. Chem., 318, 121 (1901).
Vongerichten, E.,Ann. Chem., 321, 71 (1902).
Vongerichten, E. and Muller, F., Chem. Ber., 32.

235 (1906).

Hudson, C. 5., Advances i3 Carbohydrate Chem., 4, 57
(1949).

Hart, D. A., Ph.D. Thesis, Michigan State University,
East Lansing, Michigan (1969).

Duff, R. B., Biochem. g., 23, 768 (1965).

Van Beusekom, C. F., Phytochem., 6, 573 (1967).

Neal, D. L., Kindel, P. K., g. Bacteriol., 191, 910
(1970).

Bacon, J. s. 0., Biochem. g" 82, 103p (1963).
Williams, D. T., and Jones, J. K. N., 992. g. 9233., fig,
69 (1964).

Ovodova, R. G., Vashovsky, V. E. and Ovodov, Yu. S.,
Carbohyd. 393., _6_, 328 (1968).

Beck, E., and Kandler, 0., g. Naturforsh., gee, 62 (1965).
Mendicino, J., and Picken, J. M., J. Bigl. Eggg., 239,
2797 (1965).

Beck, E., Beg. Qt. Bgt.'§§§., ZZ, 396 (1966).

Beck, E., g. Pflanzenpgysiol. 29., 52, 444 (1967).

 

Hart, D. A., and Kindel, P. K., Biochem. g., 116, 569

(1970).
Hart, D. A., and Kindel, P. K., Biochemistry, 2, 2190

(1970).

163

164

18. Nordstrom, C. G., Swain, T., and Hamblin, A. J., £429.
Indust. (London), 1253, 85.

19. Hemming, R., and Ollie, W. D., £422. and I2Q,, 1253, 85.

20. Halyalker, R., Jones, J. K. N., and Perry, M., gag. g.
9222-. E3. 2085 (1965).

21. Klyne, W., Biochem. 4,, 41, xli (1950).

22. Farooq, M. 0., Gupta, S. R., Kiamuddin, M., Rahman, W.,
and Seshadri, T. R., g, §gi. Industr. 523.,‘1ge, 400 (1953).

23. Paris, R. R., Paris, M., and Fries, J., 423. Baggy. Egggg.,
_2_4_, 245 (1966). A

24. Wagner, H., and Kirmayer, W., Naturwiss., 44, 307 (1957).

25. Nakaoki, T., Morita, N., Motosune, H., Hiraki, A., and
Takeuchi, T., g, Baggy. §gg. gapan, 15, 172 (1955).

26. Malhotra, A., Marti, V. V. 3., and Seshadri, T. R.,
Tetrahedron Letters, 36, 3191 (1965).

27. Hattori, 5., and Imaseki, H., g. 422;. 9422. §gg.. 8;,
4424 (1959).

28. Imaseki, H., and Yamamoto, T., 4394. Biochem. Biophys.,
is, 467 (1961).

29. Kindel, P. K., Gustine, D. L. and Watson, R. R., Egg.
£323., in press.

30. Grisebach, H., and Dobereiner, U., Biochem. Biophys. Egg.
Commun., ll, 73? (1964).

31. Grisebach, H., and Dobereiner, U., Q. Naturforsch., 212,
429 (1966).

32. Grisebach, H., and Sandermann, H., Jr., Biochem. §., 339
322 (1966).

33. Picken, J. M., and Mendicino, J., g. 8121. 9422., ggg.
1629 (1967).

34.

35.

36.

37.

380

39.

40.

41.

42.

43.

45.

46.

47.

165

Grisebach, H., Biogynthetic Patterns lg Microorganisms

 

and Higher Plants, John Wiley and Sons, Inc., New York,
1967, p 97 84-101.

Roberts, R. M., Shah, R. H., and Loewus, F., Plant Physiol.,
3;. 659 (1967).

Beck, E., and Kandler, 0., g. Pflanzenphy§121.,‘55, 71
(1966).

Baddiley, J., Blumson, N. L., Di Girolano, A., and

 

Di Girolano, M., Biochim. Biophys. 5932, 59, 391 (1961).
Sandermann, H., Jr., Tisue, G. T., and Grisebach, H.,
Biochim. Biophy .‘Agtg, 165, 550 (1968).

Gustine, D. L. and Kindel, P. K., g. Bigl. £423., £44,
1382 (1969).

Sandermann, H., Jr., and Grisebach, H., Biochim. Biophys.
Agfié. EQQ. 173 (1970).

Gustine, D. L., Ph.D. Thesis, Michigan State University,
East Lansing, Michigan (1969).

Kindel, P. K., Gustine, D. L., and Watson, R. R., 31924
Physiol., 4Q, Suppl., 27 (1970).

Villemez, C. L., Tsau, Yen, L. and Hassid, W. Z., 2523.
flat. gggd.,§gi., 54, 1626 (1965).

Feingold, D. S., Neufeld, E. F. and Hassid, W. Z., 4324.
Biochem. Biophys., 28, 401 (1958).

Gupta, S. R., and Seshadri, T. R., 2529. Indian Acad.
Sgi. Section.§’35, 242 (1952).

Verachtert, H., Bass, 3. T., Wilder, J., and Hansen, R. 0.,
Anal. Biochem., 11, 497 (1964).

Mukerjee, H., and Ram, J. S., g. Chromatography, 14, 551
(1964).

48.

49.
50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61,

62.
63.

166

Miles, 0. D.,and Hagen, Jr., C. W., Plant Physiol.,
Marsh, C. A., Biochem. Biophys. Acta, 44, 359 (1960).

 

Storey, E. D. E.,and Dutton, G. J., Biochem. 8., 52,
279 (1955).

Fitzgerald, D. K., and Ebner, K. E., Anal. Biochem.,
15, 150 (1966).

Castanera, E. G., and Hassid, W. Z., Arch. Biochem.
Biophys., 119, 462 (1965).

Dixon, M., and Webb, E. 0., Enzymes, Academic Press
Inc., New York, 1958, pp. 5, 150-170.

Barber, G. A., Biochem., 1, 463 (1962).

Barber, G. A., Phytochem., Z, 35 (1968).

 

Watson, R. R., and Kindel, P. K., £1881 Physiol., 48,
Suppl., 28 (1970).

Ortmann, R., Sandermann, H., Jr., and Grisebach, H.,
FEBS Letters, Z, 164 (1970).

Beisenherz, G., Boltze, H. J., Bucher, T., Czok, R.,
Garbade, K., H., Meyer-Orendt, E.,and Pflieder, G.,
8. Naturforsch. 88, 555 (1953).

Lowry, O. H., Rosebrough, N. J., Farr, A. L., and
Randall, R. J., 8. Biol. Chem., 143, 265 (1951).

 

Grisebach, H., and Bilhuber, W., 8. Naturforsch., 888,

746 (1967).

Albrecht, G., Seifert, L. L., Bass, S. T., and Hansen,

R. 0., g. 2121- 9828., g41, 2968 (1966).

Paladini, A. C.,and Lelori, L. F., Biochem. 8., 51, 426 (1952)
Leloir, L. F., and Cardini, C. E., Methods Enzymol., 3,

840 (1957).

64.

650

66.

67.

68.

69.

70.

71.
72.

73.
74.

75.
76.
77.

78.

167

Leloir, L. F., Cardini, C. E. and Cabib, E., QQEB-
Biochem., 2, 97 (1960).

Neufeld, E. E. and Hassid, W. Z., Agygg.lggggg.,gggg.,
‘18, 312 (1963).

Strominger, J. L., Maxwell, E. S., Axelrod, J. and
Kalckar, H. M., 1. Biol. Chem., 284. 79 (1957).

Khorana, H. G., Tener, G. M., Wright, R. S. and Moffatt,
J. 0., 1. AE- Chem. Soc., 12, 530 (1957).

 

Pigman, W. and Horton, D., The Carbohydrates Chemistzy
and Biochemistgy, 111, Academic Press Inc., New York

and London, 1970, 302-36.

Sandermann, H., Jr. and Grisebach, H., Biochim. Biophys.
AEEE:.12§: 435 (1968).

Bell, D. J., Isherwood, F. A. and Hardwick, N. E., 8.
Chem. Soc., 1254, 3702.

Loewus, F., 1328. Biochem. Biopgys., 122, 96 (1964).
Anderson, L. and Wolter, K. E., Ag§.l§§y. £1881 Physiol.,
11, 209 (1966).

Partidge, s. M., Biochem. 8., 42, 238 (1948).

McKay, D. 0., The Doctrine 888 Covenants 21 the Church
gg‘ggggg Christ 81 Latter-day Saints, The Church of
Jesus Christ Latter-day Saints, Salt Lake City, Utah,
1968, p. 159.

Monod, J., Science, 154, 477 (1966).

Rahman, A. U., Z. Naturforsch, 128, 201 (1958).

Ankel, H., Ankel, E., Feingold, D. S., Schutzbach, J. 5.,
Biochim. Biophys. A223: 138, 172 (1967).

Sachindra, I. and Chakraborti, K., Igggg. Base Research

Inst., Calcutta, 21, 61 (1956-58).

168

79. Faroog, M. 0., Varshneg, I. P., and Rahman, W.,
Naturwissenschaften, 45, 265 (1958).

80. Hodgman, C. D., Handbook 2: Chemistry and Physics, edition

 

43, The Chemical Rubber Publishing Co., Cleveland, Ohio,
1961, p. 1653.
81. Sandermann, E. J., 82222. 8. Biochem., 2, 404 (1968).
82. Shafizadeh, F., 22122. Carbohyd. 2222., 13, 9 (1958).
83. Horecker, B. L., Ann. Rev. Microbiol., 22, 253 (1966).

 

84. Bray, G. A., Anal. Biochem., 1, 279 (1960).

85. Maron, S. H., and Prutten, C. F., Principles 28 Physical

 

Chemistry, Ed. 3, The MacMillan Co. New York, 1962, p. 622.

86. Vongerichten, E., Chem. Ber., 33, 2334 (1900).

87. International Union of Biochemistry,
Enzyme Nomenclature, American Elsevier Inc., New York,
1965.

88. Pigman, W. and Horton, D., The Carbppydrates ggemistry
and Biochemistpy, 111, Academic Press, New York, 1970,
pp. 96-98.

89.7 Eliel, E. L., Stereochemistry 2; Carbon Compounds,

 

McGraw-Hill Book Company, Inc., 1962, pp. 204-255.

90. Carminatti, H., Passeron, S., Dankert, M., and Recondo,
E., 8. Chromatpg., 18, 342 (1965).

91. Bunton, C. A., Lewis, T. A., Llewellyn, D. R., and Vernon,
C. A., 8. 9222. 822., 1255, 4419.

92. Pazur, J. H., and Gordon, A. L., 8. Amer. Chem. 822., 25
3458 (1953).

93. Aspinall, G. 0., and Telfer, R. G. J., 8. Chem. Soc.,

1255, 1106.

 

94.

95.

96.

169

Whistler, R. L., and Howell, R. M., 8. Org. Chem., 82,
3290 (1964).

 

Feather, M. S., and Harris, J. F., 8. Org. Chem., 32,
153 (1965).

Mislow, K., Introduction 22 Stereochemistry, W. A. Ben-

 

Jamin, Inc., New York, 1966, pp. 76-77.