mammals OF 0 . APIBSE BY AN ENZYME: f
svsm-a ISOLATED FROM LEMNA MINOR L »

Thesis for the Degree of Ph. D.
MICHmAN STATE umvmsnv
DAVID L GUSTINE
1959

[HUD

This is to certify that the

thesis entitled
Biosynthesis of D-Apiose
By An Enzyme System

Isolated From Lemna minor L.

presented by

David L. Gustine

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Biochemistry

few/é (W

Major professor

Date February 19¢ 1961

0-169

wt”

:5" " -
if“ L l B R A R Y

l Michigm State

I U n‘wet sity

ABSTRACT

BIOSYNTHESIS OF D-APIOSE BY AN ENZYME
SYSTEM ISOLATED FROM LEMNA MINOR L.

 

By

David L. Gustine

An enzyme system was isolated from Lemna minor L.

which converts UDP-D-glucuronic acid-lac (D-glucuronic

 

acid portion uniformly labeled) to a compound containing
D-apiose-luC.

The enzyme was purified 28-fold by the following
steps: (1) homogenization of Lemna plants and centrifu-
gation, (2) ammonium sulfate fractionation, (3) chroma-
tography on Sephadex G100, (u) chromatography on DEAE-
Sephadex and (5) vacuum dialysis. The purified enzyme
showed a requirement for NAD+ with an optimum concentra-
tion of 0.8 x 10-3 M. It had optimum activity over the
pH range of 8.2 to 8.5. The Km for UDP-D-glucuronic
acid was 0.870 x 10’5 M.

Only two products were isolated from the enzyme
reaction: one was UDP-D-xylose-luc and the other was
an unidentified compound (Peak III), which contained
D-apiose-luC. The second compound was not UDP—D-apiose-
INC, D-apiose—luC-phosphate, or D-apiose-luC. When

Peak III was hydrolyzed, the only radioactive component

David L. Gustine

released was D-apiose—luc. The radioactive hydrolysis
product was characterized by the following criteria:
(1) cochromatography with D—apiose on paper in three
different solvent systems, (2) synthesis of D-apiose
phenylosotriazole-luc and crystallization to constant
Specific activity and (3) oxidation with periodate and
isolation and identification of the products.

Peak III was shown to contain phosphate by the fol—
lowing criteria: (I) it behaved as an anion during
electrophoresis and (2) it could be converted to D-apiose-
luC-phosphate. D-apiose-luC-P was shown to contain
phosphate by treatment with alkaline phOSphatase and its
subsequent conversion to D-apiose—luC.

Results of chromatography and solubility experiments
indicated that Peak III possesses non-polar characteris-
tics. Therefore, Peak III may be a phospholipid of
unknown structure. It is also possible that Peak III
is an artifact formed either during the incubation of
the enzyme with substrate or during the isolation pro-

cedure. One such artifact could be D-apiose-cyclic

phosphodiester.

BIOSYNTHESIS OF D-APIOSE BY AN ENZYME

SYSTEM ISOLATED FROM LEMNA MINOR L.

 

By

David L. Gustine

A THESIS

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

DOCTOR OF PHILOSOPHY
Department of Biochemistry

1969

Dedicated to my wife and my children

11

". . . I ask not if God created the
universe but rather, 'How did he do
it?‘ I ask not if God controls his
universe but rather, 'How does he

maintain it?'" (53)

ACKNOWLEDGMENTS

The author wishes to express his sincere apprecia—
tion to Dr. Paul Kindel for his guidance, patience and
understanding during the course of this research. The
stimulating discussions with Barry Rhinehart and David
Hart are very much appreciated. Thanks are also due to
Miss Roswitha Blohm for her technical assistance. The
author would also like to thank Dr. w. A. Wood, Dr.
Robert S. Bandurski and Dr. Eugene LeGoff for serving on
his guidance committee.

The author is especially grateful to his wife,
Diane for her love and encouragement and for her faithful
work as an X-ray technician to help finance his work as
a graduate student. The support of a National Defense

Education Act, Title IV, Fellowship is also appreciated.

iv

TABLE OF CONTENTS

 

Page
DEDICATION o o a o o o o o o o o o ' o 0 ii
ACKNOWLEDGMENTS . . . . . . .~ . .. . . . iv
LIST OF TABLES C I O O O O O O O G O O O Vii
LIST OF FIGURES O O O O O O O O O O O O V111
Chapter
I. INTRODUCTION AND LITERATURE REVIEW . . . . l
The Discovery of Apiose and Its Distribution
in Nature . . . . . . . . 1
Discovery of Apiose and Proof of its
Structure . . . . . . . . . l
Stereochemistry of Apiose . . . .- . . A
Apiose as a Polysaccharide Constituent . . 6
Biosynthetic Studies . . . .~ . . . . 10
Possible Biosynthetic Routes to D-apiose . lO
Isotope Incorporation Studies . . . . ll
Possible Biosynthetic Routes to Glycosides
and Polysaccharides . . . . . . . 18
II. MATERIALS AND METHODS . . . .p . . . . 20
Materials 0 a a ' o c o e o c 20
Substrates and Cofactors . . . . . . 20
Chemicals and Reagents . . . . . . 21
Instruments . . . . . . . . . . . 23
MethOdS O C I O l O O I O C O O O 23
General Methods . . . . . . . . . 2l
Growth of Lemna minor . . . . . . 26
Standard Enzyme Assay . . .u . . 3. . 26
Synthesis of 3H- UDP- -D—glufiose-l C and H-
UDP— —D-glucuronic acid--l . . . . . 27
III. RESULTS 0 O O C I O I O O O O O C 30
Purification of the Enzyme . . . . . . 3O
Purification Procedure . .‘ . . . . . 30
Stability of the Enzyme . . . . . . . A0

V

Chapter

Partial Characterization of the Purified

Enzyme

Determination of the Linearity of the

Enzyme Assay
Determination of the Optimum pH +.
Determination of the Optimum NAD Concen-

tration

Determination of K and V
Effect of MgCl

Components of the Enzymatic Product
Structure and Mechanism of Formation of

Future of the Problem

IV. DISCUSSION
Peak III
BIBLIOGRAPHY

2

onmthe fogfigtion of
Peak III in Different Buffers
Characterization of the Enzymatic Products
The Initial Isolation and Identification
of Enzymatically Synthesized D-apiose—
Enzymatic Synthesis of Peaks I,
Identification of Peak I
Identification of Peak II
Partial Characterization of Peak III

vi

0

C

II and III.

102

103

LIST OF TABLES

Summary of the purification of the enzyme
described in the text . . . . . . . .

Stability of the concentrated enzyme from the
G100 Step at "200 o o o I a o o o 0

Stability of the concentrated enzyme after the
DEAE Step at -200 o I o o 0 O o o 0

Effect of MgCl2 on the formation of Peak III .

Recrystallization of D—apiose phenylosotriazole—
luc to constant specific activity . . . .

Results of periodate oxidation of D-apiose—luC
enzymatically syTBhesized from UDP—D-
glucuronic acid- C . . . . . . .

3H/luC ratios of Peaks I, II and III . . . .

The formation of D—apiose-luC-P from Peak
III With time o o c o o c o o ‘ a 0

vii

Page

39

41

146
63

68

72

75

88

LIST OF FIGURES

Figure Page

1. Postulated pathway for the biosynthesis of
D-apiOse o o o o o o o o o q o a 12

2. Chromatography of a standard enzyme assay
in solvent system A . . . . . . . . 28

3. Chromatography of a dialyzed ammonium sulfate
fraction on Sephadex GlOO. . . . . . . 33

A. Chromatography of 6100 fractions on DEAEg
Sephadex . . . . . . . . . . . . 36

5. Linearity of enzyme activity with increasing
amounts of protein . . . . . . . . . 42

6. Increase in the formation of Peak III with
time 0 O O O O _ O O O O O O O O _ “u

7. Increase in the formation of Peak III with
time at two different substrate concentra-
tions 0 o o p o o o o ‘ o o o o o o ’47

8. The formation of Peak III in the standard 15
minute assay with respect to pH . . . . 50

9. The formation of Peak III in the standard 15
minute assay with respect to pH . . . . 52

10. The formation of Peak III with increasing
concentrations of NAD+ . . . . . . . 5M

11. The increase in the formation of Peak III
with increasing UDP-D—glucuronic acid-luc
concentration . . . .3 . . . . . . 56

12. Increase in the rate of Peak III formation
with increase in UDP-D-glucuronic acid-lac
concentration . . . . . . . . . . 59

13. Lineweaver-Burk plot of the data shown in
Figure 12. O O 0 O O O O O O O O 61

viii

Figure Page

I“. Chromatography in solvent system D of the
hydrolysate from the first enzymatically
synthesized D-apiose—containing compound . 65

15. Chromatography in solvent system C of
hydrolyzed products from an incubation of
UDP-D-glucuronic acid—luC with a 35,000 x g
enzyme preparation . . . . . . . . .. 69

l6. Chromatography of hydrolyzed Peak II in
solvent system C . . . . . . . . . 78

l7. Chromatography of Peak III which had been left
on a silica gel plate for six days . . . 85

l8. Chromatography of D-apiose-luC-P in solvent
system C after treatment with NaBHu . . . 90

ix

CHAPTER I

INTRODUCTION AND LITERATURE REVIEW

The Discovery of Apiose and Its
Distribution in Nature

Discovery of Apiose and
Proof of its Structure

 

 

There have been found to occur in nature fifteen
branched-chain monosaccarides. Two have been identified
in higher plants, and the remainder in microorganisms
(25, 32). The first branched—chain monosaccharide to be
discovered was "a remarkable branched-chain aldopentose"
which was named apiose (I) by its discoverer, Vongerichten
(l, 26). Apiose was originally found as a constituent of
apiin (III), a glycoside isolated from Petrosilinum
crispum L. (parsley). Apiose was not recognized as a con-
stituent of apiin during the early investigations, and
was not discovered until 1901. Apiin was discovered in
1836, when Rump (24) reported that a hot, aqueous extract
from plants yielded a gelatinous material upon cooling.

He called this material apiin. Other investigators (l,
58, 59) studied this material and found that reducing
sugars were released upon hydrolysis. They were also able

to obtain a flocculent material from hydrolysates which

1

was nearly insoluble in cold water. This aglycone was
named apigenin (IV) by Linderborn (l, 59).

Beginning in 1876, Vongerichten started a series of
investigations on apiin which eventually resulted in the
elucidation of the structures of apiin and apiose.
Previous to 1901, the year apiose was discovered, it had
been assumed that each of the two sugars present in apiin
was glucose. In 1901, Vongerichten (26) was able to
hydrolyze apiin in two stages with acid. The first stage
was done with 0.5% H28014 and produced a reducing sugar
plus a new glucoside, 7-apigenin—B-D-glucopyranoside (II).
Hydrolysis of the new glucoside with 15% H230” yielded
D-glucose and apigenin. Using the sugarsolution from the
first stage, Vongerichten prepared a crystalline phenylosa-
zone and p-bromophenylosazone. Upon analysis of these
derivitives, he found that the sugar was a pentose. In
addition, he found that the pentose did not yield furfural
upon acid treatment and concluded that he had discovered a
new kind of pentose. He named the pentose apiose after the
Linnean designation of parsley, Apium petroselium L.

In his next paper, Vongerichten (27) characterized
apiose as a branched-chain aldopentose. To show this, he
first oxidized an impure solution of apiose with bromine
water to form apionic acid. He characterized the compound
by preparing the crystalline phenylhydrazide, the-

crystalline strontium salt, and an amorphous calcium salt.

 

CHO
H—C-OH CH OH
HOH2C-C-OH
CH2OH HO H
OH
I II
D-apiose 7-apigenin-B—D-gluc0pyranoside

 

 

III

Apiin

 

HO l // \\ OH

 

OH 0
IV

apigenin

ll .rll ll it'lltfl‘ulll‘fl

..I|\[ I‘ll) l

Second, he reduced calcium apionate with hydriodic acid
and phosphorous to yield a volatile acid which he con-
cluded was isovaleric acid. From such information,
Vongerichten pr0posed that the structure of apiose was
that shown in (I). The configuration of the single
assymetric carbon atom was not eStablished by Vongerichten.
In 1906, Vongerichten and Mfiller (28) obtained the first
crystalline hydrazone of apiose, d-benzyl-a-phenylhydrazone.
Twenty-four years later, Schmidt (29) repeated the
work of Vongerichten using pure apiose and confirmed the
structure (I). Utilizing several empirical rules for the
determination of the configuration of the a-carbon atom of
a-hydroxycarboxylic acids, Schmidt deduced the correct
stereochemical structure of apionic acid. From this
information, Schmidt concluded that apiose isolated from

apiin has the D-configuration as shown in (I).

Stereochemistry of Apiose

When the aldehyde form of D-apiose (I) converts to
the furanose form as in (V-VIII), two new asymmetric
centers are formed; thus four configurational isomers are
possible. The nomenclature for these isomers has not yet
been agreed upon. Carbon atom l is covered by the usual
0, 6 terms for glycosidic anomers, but for carbon atom 3
there are no universally accepted rules. Cahn (3) has
suggested that the numbering be 1 to A from the glucosidic

position to the other end of the furanose ring, and that

the free hydroxymethyl group be numbered 5 (see VII).

However, most of the authors cited in this thesis number

the free hydroxymethyl group 31 or A1 (in this thesis it

will be numbered 31).

Cahn further suggested that the

"D-apio" or "L—apio" should designate the stereochemistry

of carbon atom 2 of apiose.

When the open chain form con-

verts to the cyclic form, "furanose" should be added to

the stem together with a second D defining the stereo-

chemistry at position 3.

With the addition of the a, B

prefix for carbon atom l, the configuration for all 3

assymetric centers if fixed.

Although these are not

approved rules, this nomenclature will be used in this

thesis.

Furthermore, the term D-apiose will be used

throughout, and the "D" will refer to the configuration of

carbon atom 2 as shown in (I).

OH OH

CH20H OH

V

a-D-apio—L—
furanose

OH

CHZOH OH

VI

B-D-apio-L—
furanose

VII

d-D-apio-D-
furanose

O
H2OH
OH OH

VIII

B-D-apio-D-
furanose

An investigation to determine the configuration of
carbon atom 3 in glycosidically bound D-apiose has been
reported. Hulyalkar, Jones and Perry (14)
isolated tri-O—methyl-D—apiose from the hydrolysis
products of methylated apiin and found it to be identical
with synthetic 2,3,u-tri-O-methyl-D-apio-D-furanose. In
addition, periodate oxidation studies indicated the hydroxyl
groups at carbon atoms 2 and 3 were in the cis configura-
tion. Optical rotation data suggested that a B-type
linkage exists between the D-apiosyl unit and carbon atom
2 of the D-glucosyl moiety of apiin. All of their data
indicated that the structure of D—apiose in apiin is
B-D-apio-D-furanose (VIII). At the present time, the con-
figuration of carbon atom 3 in glycosidically bound
D—apiose has been determined only for apiin.

Apiose as a Polysaccharide
Constituent

 

D-apiose was first implicated as a polysaccharide
constituent by the work of Bell, Isherwood, and Hardwick
in 195“ (3). They subjected leaves and residual fibers

of Posidonia australis (marine fiber) to acid hydrolysis

 

and isolated what they thought was L-rhamnose. However,
further studies showed that (l) the sugar did not yield
furfuraldehyde on acid treatment, (2) it behaved chroma-
tographically as a pentose, and (3) contained cis

vicinal hydroxyl groups as evidenced by the formation of a

complex with boric acid. These facts led them to compare
the sugar with D—apiose isolated from apiin. Preparation
of the di-O—isopropylidene and 2,5—dichlorophenylosazone

derivitives showed that the Posidonia sugar and D—apiose

 

were identical. Since they isolated D-apiose from
hydrolysates of the residual fibers, their work inferred
that D-apiose could_be present in polysaccharides (see
also 7).

Duff and Knight (6) first reported the presence of
D—apiose in Lemna (duckweed), one of the smallest and
simplest of flowering plants. They also examined Zostera

marina and re-examined Posidonia australis. They found

 

that all three species of plants released D-apiose upon
acid hydrolysis. They reported 6-7% for Lemna, and 4-5% for

Posidonia and Zostera. They stated that "it seems likely

 

that the sugar is a constituent unit of a polysaccharide"
in those plants. The first attempt to prove that D-apiose
was not present as a flavone glycoside, but rather in a
polysaccharide in certain plants was made by Bacon (7).
This he showed by a negative approach. That is, he
demonstrated that only traces of D-apiose were present in

Zostera marina extracts, which contained glycosides. He

 

further showed that hydrolysis of these extracts did not
release any free D-apiose. Since Duff and Knight (6) had

shown that hydrolysates of the whole plant contained much

larger amounts of D—apiose, he concluded that D-apiose
was present in the polysaccharide portion.

The first isolation and demonstration of a poly-
saccharide fraction containing D-apiose was by Williams
and Jones (8). They extracted Zostera marina L. with
(l) methanol, (2) n-butanol saturated with water and (3)
water saturated with n—butanol, respectively. The result-
ing residue was then extracted with cold 2% sodium hydroxide.
This extract was dialyzed and concentrated to a dry resi-
due. This "polysaccharide fraction" contained D-xylose and
D-apiose inealflgratio. Further evidence for the presence
of D-apiose in a polysaccharide was presented by Beck and

Kandler (10). Using Lemna gibba and Lemna minor they were

 

 

able to show that D-apiose occurred only in fractions
insoluble in ethanol and water, and that it was released
by acid hydrolysis. Data on incorporation of 1“C02 into
D-apiose, D-xylose, and D-glucose revealed that D-apiose
is not a part of storage material which is rapidly turning
over, as in D—glucose (starch), but like D-xylose is part
of a cell wall component. Although they did not isolate
polysaccharides containing D—apiose, their results point
to such a compound or group of compounds. That same year

Duff (11) fractionated a holocellulose preparation from

Lemna minor and showed conclusively that D-apiose was.

 

present in a number of the polysaccharide fractions. In

that publication, he also presented results from an

examination of 176 different plants comprising 106
families, wherein he analyzed each plant for D—apiose.
His procedure involved inspection of paper chromatograms
of hydrolysates of whole plants, but did not distinguish
between the presence of D-apiose in glycosides or poly-
saccharides. Of the 175 plants tested, 76 showed no D—
apiose, 31 had trace amounts (about 0.03% of dry weight),
51 had moderate amounts (about 0.3%) and 17 were good
sources (about 3%). Of those that showed no D-apiose one
in ten were selected on a random basis and examined more
closely for D-apiose. All of those showed traces of D-
apiose. That study established the presence of D-apiose
in a wide variety of species in the plant kingdom.

The first reported isolation, purification and par-
tial characterization of a polysaccharide fraction con-
taining D-apiose was by Beck (30) in 1967. He_isolated an

apiogalacturonan from Lemna minor which he concluded was

 

an 0-1,A linked polygalacturonic acid chain, with single
D-apiose molecules attached as side chains to the D-
galacturonic acid residues. He found the composition to
be about 68% D-galacturonic acid and 28% D-apiose.
Recently, Ovodova e£_al. (U8) isolated an apparently
similar polysaccharide fraction from Zostera marina L.
However, no characterization work was done.

D-apiose has now been established as a constituent

of cell wall polysaccharides in Lemna minor L. and Zostera

 

lO

marina L. as well as part of the glycoside apiin in
Petroselinum crispum L. It has also been reported as a

component of a second glycoside in Petroselinum, called

 

petrosilinin (2). D-apiose has been found in four other
flavone glycosides; they are luteolin and chrysoeriol
from Apium graveolens (celery) (35), lanceolarin (an

isoflavonoid) from Dalbergia lanceolaria (54) and an

 

unidentified flavone from Digitalis purpurea (foxglove)
(55). D-apiose has also been found in furcatin, a non-

flavone, from Viburnam furcatum Blume (57). In summary,

 

it has now been established that D-apiose occurs in a
large number of flowering plants.
Biosynthetic Studies

Possible Biosynthetic Routes
to D-apiose

Although D-apiose occurs in a wide variety of plants,
only Petroselinum and Lemna have been used for biosynthetic
studies. D-apiose occurs predominantly in Petroselinum as
the flavone glycoside apiin; whereas, in Lemna, D-apiose
is found predominantly in cell wall polysaccharides. In
each case, D-apiose is bound in a glycosidic linkage. Thus,
the possibility exists that D-apiose could be synthesized
and glycosidically bound in a similar manner in either
plant.

Five possible mechanisms for the formation of the

branched-chain carbon skeleton of D-apiose have been

11

considered (l6, 17, 25): (a) methylation of a tetrose by
methionine, (b) synthesis from acetate units, as in iso-
prenoid metabolism, (0) an intramolecular acyloin conden-
sation, (d) condensation of dihydroxyacetone with acti-
vated glycoaldehyde and (e) an intramolecular rearrange-
ment of the carbon skeleton of a pentose. Only the latter

possibility has proved to be correct.

Isotope Incorporation Studies

Grisebach and D6bereiner were the first to suggest
that D-apiose is synthesized bound to a nucleotide (9).
They postulated the sequence shown in Figure 1. They
based their sequence on (1) the known formation of
UDP—D—xylose from UDP-D-glucuronic acid in which a
transient carbonyl group was pr0posed at carbon atom A
(15, 3A) and (2) a similar rearrangement previously
postulated by Blumson and Baddiley (36) for the biosyn-
thesis of streptose. Grisebach and D6bereiner administered
acetate-l—luc, formate-luC, glucoseeU-luC, and glucose-

,AluC to young parsley shoots for periods of 36 hours

3
and isolated apiin. Radioactivity was incorporated from
all of the precursors into apiin, but D-glucose was by far
the best. Upon hydrolysis, they found that both D-apiose
and D-glucose were labeled. Periodate degradation of
apiin revealed that the 31 carbon atom of D-apiose con-

tained 23% of the radioactivity in D—apiose synthesized

from uniformly labeled D—glucose, and 40% of the

 

l2

 

.omOfiQmI
Q mo mamonpzmmOHn on» mom magnuma popoHSQmom

.H omzwam

l3

emoaeeuaumaa
mo mo

m memm

O

 

Q¢z

+m+mm<z
mo mo

m muo
o

O

smeasxumumm:

mo
mo mo

 

Q¢z

nu

oaom
canonsosamlmlmmb
+m+ma<z L

mo

 

 

 

It

mo mo
m ,m mo mo mo m mo
«00 +m+ aez +m<z o
mo o

3000

 

14

radioactivity in D—apiose synthesized from D-glucose-3,u-luC.

If the carbon skeleton was derived from D-glucose without
it breaking down into smaller molecules, and if their
scheme was correct, the expected values would be 20% and

50%, respectively. Further studies with Petroselinum

 

were done by Mendicino and Picken (12); however, their
results were not in complete agreement with those of

Grisebach and D6bereiner. The extent of incorporation of

acetate-2-luC, L-serine-3-luC, and L-methionine-MCH3 into

D-apiose from isolated apiin was as much or more than

with D-glucose-l-luc. Using Lemna minor they also found

 

acetate-2-luC was a better precurser of D-apiose than
lu

C02. They concluded that the biosynthesis of D—apiose
might involve a 1 carbon transfer reaction. They did not
offer an explanation for the greater incorporation of
acetate-2-luC. They also found that in parsley, D-glucose
was incorporated into D—apiose with the loss of carbon
atom 6. Their experiments included a study on the loca-
tion of the synthesis of apiin. The results indicated
that apiin was synthesized in the roots and translocated
to the leaves. However, they did not test whether D-
apiose itself was synthesized in the roots or in the leaves.

Using Lemna gibba, Beck and Kandler (16) investi-
gated the labeling patterns in D-glucose, D-xylose and

D-apiose derived from D-glucose-U—luC, l-luC, 2-luC and

6-luC. In addition, they administered methionine—luCH3.

Their results showed that all three sugars were labeled

15

only slightly when methionine—MGR3 was used, whereas
they were strongly labeled when the various labeled

glucoses were used. In addition, when methionine-luCH

3
was used, D-apiose was not labeled in preference to
D-xylose. In order to make a decision as to which of the
five possible biosynthetic routes to D-apiose was most
correct, Beck and Kandler isolated D-glucose, D-xylose
and D—apiose from Lemna polysaccharides and determined
the distribution of radioactivity in each carbon atom of
the three sugars. Their results clearly showed that the
carbon skeletons of both D-apiose and D—xylose are
derived from D-glucose. In fact, their data fit the bio-
synthetic pathway in Figure l, and rule out the other four
possibilities mentioned earlier. Almost simultaneous
with the work of Beck and Kandler, Grisebach and
DBbereiner (17) reported similar experiments with 333327
selinum. Part of that work was essentially the same as
described in their earlier paper. Further interesting
data was obtained from tritium experiments. Using D-
g1ucose-6-luC—A-3H and D-glucose-U—luC-3—3H, they found
70% and 60%, respectively of the tritium in D-apiose at
carbon atom 31. They also found that carbon atom 6

of D—glucose was lost during the conversion to D-apiose.
These results also fit the biosynthetic scheme in Figure 1
if upon the formation of the A-keto intermediate, the

tritium at carbon atom A was transferred to NAD+, and

16

subsequently transferred after the rearrangement from
NAD3H to the transient intermediate to give D-apiose. If
carbon atom 3 becomes carbon atom 31, then the tritium

at carbon atom 3 would have remained bound to carbon

atom 31. The results of the 1“C experiments therefore
suggest that carbon atom 3 of D-glucose becomes carbon
atom 31 of D-apiose. The results of Beck and Kandler

and of Grisebach and D6bereiner, were confirmed by Picken
and Mendicino (22). They studied the incorporation of
D-glucose-l-luC, -2-luC, -3,A-1uC, -6-luC, acetate-2-luC,
D,L-serine-3-luC and L-methionine—MCH3 into the D-apiose
moiety of cell wall polysaccharides. They also con-
cluded that D-glucose was converted to D-apiose by the
loss of C-6 and the conversion of the straight-chain
sugar to the branched—chain sugar.

At this point, it still remained to be demonstrated
that D-glucuronic acid could be converted to D-apiose.
Late in 1966, Grisebach and Sandermann (20) reported such
a finding. In their experiments, they administered D-

IN

glucuronic acid-U-luC and D-glucuronic acid-6- C to

Petroselinum plants with and without roots. This resulted

 

in about 0.1% incorporation into the D-apiose portion of
apiin in the case of uniformly labeled D-glucuronic acid.
At the same time, there was a 13% incorporation of D-
glucuronic acid into polysaccharides. They concluded that

the conversion to D-apiose was low because the synthesis

17

of apiin occurs at a slower rate than the synthesis of D-

glucuronic acid containing polysaccharides. Furthermore,

they establishec by periodate oxidation that all five car-
bon atoms of D-apiose were derived from D-glucuronic

14 lHC

acid-U- C. With D-glucuronic acid-6— , they again

proved that carbon atom 6 was lost during the conversion
to D-apiose. Finally, they found essentially no differ~

14 1A

ence in incorporation of C from D-glucuronic acid—U— C

into apiin in Petroselinum with and without roots.

 

In 1967, Roberts, Shah and Loewus (23) published
results which also showed D-apiose was synthesized from
D-glucuronic acid in Lemna gibba as well as in Petroselinum.

They used myo-inositol-2-lu0 with Lemna and myo-inositol-

 

2-3H with Petroselinum. Loewus and others (37, 38) had
14

 

previously demonstrated that myo-inositol-2- C or -2-3H
was oxidized by plants to D-uronic acids and D-pentoses
containing the label exclusively in carbon atom 5. In
essence, they administered D-glucuronic acid-S-luC and
D-glucuronic acid-5-3H to the plants. In the case of
Lemna, both D-apiose (1.7% incorporation in 8 weeks) and
D-xylose derived from the polysaccharide fractions were
labeled. In the case of Petroselinum, only the D—apiose
portion of isolated apiin was labeled. Periodate oxida—
tion of D-apiose-luC from such experiments, showed that

1N

9A% of the C in D—apiose was located in carbon atoms 31

or A., Since the other experiments described above clearly

18

indicate that radioactivity located at carbon atoms 3 or A
of glucose is incorporated into carbon atoms 3 or 31 of
D-apiose, then carbon atom A of D-apiose must have been
labeled by myo-inositol-Z-luC. The results of Roberts
g£_§1. gave positive proof that carbon atom 5 of D-
glucuronic acid becomes carbon atom 5 of D-apiose.

Recently, Sandermann and Grisebach (31) have reported
the isolation of a mixture of UDP-sugars from Petroselinum
which contains UDP-D-apiose. Another publication has

appeared in which Sandermann, Tisue and Grisebach (21)

described a soluble enzyme from Petroselinum, and a soluble

 

and particulate enzyme from EEEEE.Wh1°h converted UDP-D-
glucuronic acid-lac (u.l.) to D-apiose-luC. Their results
indicated that the enzyme product in each case was UDP-D-
apiose-luc.

In summary, the evidence described in this review
supports the postulated pathway shown in Figure 1.
Possible Biosynthetic Routes to
Glycosides and Polysaccharides

Very little research had been done with respect to
the biosynthesis of D-apiose-containing flavone glycosides
or polysaccharides. A possible mechanism is the transfer
of D-apiose to an acceptor molecule from a D-apiose-
nucleotide. There is precedence for this hypothesis.

Barber (33) has shown in mung bean leaves that D-glucose

and L-rhamnose are transferred from UDP-D-glucose and

l9

UDP-L-rhamnose to quercetin, thus forming rutin, a
flavonol. It~has become clear that sugar nucleotides are
involved in the biosynthesis of plant polysaccharides as
well (61-65). There is now some evidence that suggests

a lipid intermediate is required for the transfer of the
sugar moiety from a sugar nucleotide to the growing end of
a polysaccharide in bacteria. Lipid intermediates have
been demonstrated in some bacterial systems (60), and in
the case of cellulose biosynthesis in Acetobacter xylinum,
the lipid intermediate has been identified (39, 40, Al).
Efforts to identify this intermediate in green plants have
so far failed, but its presence has reportedly been
demonstrated (A1).

The subject of this thesis is the demonstration of
an enzyme system which converts UDP-D-glucuronic acid to
D—apiose. The isolation of the enzyme or enzymes, its
partial characterization, the isolation of the enzymatic
product and its partial characterization will be described.
A communication describing part of this research will
appear in the March 12, 1969 issue of the Journal of

Biological Chemistry.

CHAPTER II

MATERIALS AND METHODS

Materials

Substrates and Cofactors

Radioactive UDP-D-glucuronic acid-luC was purchased
from New England Nuclear Corp., Boston, Mass., and had a
specific activity of 173 mcuries/mmole. The glucuronic
acid portion of the molecule was uniformly labeled with
1"AC, and the degree ofxuniformity for each carbon atom was
within i 15%. D—glucose-luC-l-phosphate was also pur—
chased from New England Nuclear and had a specific activity
of 179 mcuries/mmole. The glucose was uniformly labeled
within i 15% for each carbon atom. UTP-3H was a gift from
Dr. R. G. Hansen and was purchased from Schwarz Bio
Research, Inc., Orangeburg, New York. The specific
activity was 2.0 curies/mmole. The position of label was
not specified by the manufacturer.

NAD+ for enzyme assays, and UDP-D-glucose and UDP-D-
glucuronic acid for chromatographic standards were pur—

chased from P-L Biochemicals, Milwaukee, Wisconsin and

Sigma Chemical Co., St. Louis, Missouri.

20

21

Chemicals and Reagents
D-apiose used as carrier and for a chromatographic

standard was isolated from Petroselinum crispum in Dr.

 

P. Kindel's laboratory. The isolation procedure described
by Gupta and Seshadri (18) was used. UDP-D-xylose for a
chromatographic standard was a gift from Dr. R. G. Hansen.

Crystalline calf liver UDP-D-glucose pyrophosphorylase
(EC 2.7.7.9) was kindly supplied by Dr. R. G. Hansen.
Snake venom phosphodiesterase (EC 3.1.“.1.) was a gift
from Dr. J. L. Fairley and was purchased from Worthington
Biochemicals Corp., Freehold, New Jersey. Calf intestine
alkaline phosphatase (EC 3.1.3.1) was a gift from Dr.

N. E. Tolbert and was purchased from Nutritional Biochemical
Corp., Cleveland, Ohio (listed in their catalogue as
"purified," see reference 5 for purification procedure).

Sephadex G100 and DEAE—Sephadex A50, medium, was
purchased from Pharacea Fine Chemicals, Inc., Piscataway,
New Jersey. Before use, the Sephadex gels were swelled
in glass distilled water for the recommended times, and
fine particles removed by repeated decantations with
appropriate buffers.

Dialysis tubing (#8, 0.39 inches in diameter) was
purchased from Union Carbide Corp., Chicago, Illinois.
Before use, it was soaked in glass distilled water as
follows. A length of tubing from 5 to 10 feet long was

placed in a 1 liter beaker, and the beaker was filled with

22

glass distilled water. The tubing was heated at 700 for
A hours. The water was decanted, new water added and the
tubing heated again at 70° for A hours. The cycle was
done a total of 5 times.

The following scintillation fluids were used: (1)
the "dioxane system"; 100 g of naphthalene, 10 g of
2,5-diphenyloxazole (PPO) and 0.25 g of l,A-bis-[2-(5-
phenyloxazolyl)]-benzene (POPOP) in 1 liter of p-dioxane
(56) and (2) the "toluene system"; ”.0 g of 2, 5-bis
[2(5-tgrtfbutylbenzoxazolyl)J-thiophene (BBOT) in 1 liter
of toluene. The source reference of the second scintil-
lation system was the Packard Inst. Co., Downers Grove,
Illinois. The three scintillators were purchased from
Packard Instrument Company.

The culture medium used for growing Lemna minor L.

 

cultures was based on medium V of Norris gt_al. (AA).

The medium was made up as follows: 30.31 g of "major
salts", 10.0 ml of "microelement solution" and 10.0 ml

of "iron solution" were added to 10 liters of distilled
water and stirred for 2H hours. The major salts used for

10 liters of medium consisted of 20.2 g of KNO 2.72 g of

3’
KHZPOu, 4.93 g of MgSOu-7 H20, 2.36 g of 0a(N03)2-u H20,
0.10 g of NaCl and 0.006 g of CuSOu°5 H20. Large quanti-
ties of the combined salts were ground to a fine powder

with a mortar and pestle and stored in a dessicator.

The microelement solution contained 2.86 g of H3BO3, 1.5“

23

g of MnSOu‘HZO, 0.105 g of ZnCl 0.08 g of CuSOu'S H20

2,
in 1 liter of water._

The iron solution (A) was prepared by adding 0.9 g
of KOH in 10 ml of water to 2.12 g of EDTA (free acid)
in 20 ml of water. After the EDTA had dissolved, the
solution was added slowly and with stirring to 2.0 g of
FeSOu'7 H20 in 50 m1 of water. The solution was areated
4-5 hours until the stable ferric complex was obtained
(when solution became clear). The iron solution was

stored in the dark at U° and the microelement solution was

stored at A°.

Instruments

 

The following equipment was used: (a) GME Model
V—10 fraction collector, purchased from Gilson Medical
Electronics, (b) Bfichi Rotovapor, purchased from Rinco
Instruments, Inc., (0) Gilford Model 300 micro sample
spectrophotometer, purchased from Gilford Instruments,
(d) Gilford Model 200 spectrophotometer, (e) Packard Model
7201 Strip Counter, purchased from Packard Instrument Co.,
and (f) Pherograph High Voltage Electrophoresis Apparatus,

Type 6A, purchased from Brinkmann Instruments.
Methods

General Methods

a. Paper and thin layer chromatography and electro-

 

phoresis.--Unless otherwise stated, all paper

2A

chromatography was carried out on Whatman N0. 3MM paper
previously washed with 0.1 M citric acid for 24 hours,
followed by glass distilled water for 48 hours. All

thin layer chromatography was done on Brinkmann pre-
coated silica gel plates, catalogue No. 68 10 00. Carbo-
hydrates were visualized on paper with aniline hydrogen
phthalate (47). Sugar nucleotides were visualized in the
dark with the aid of a Mineralight Model UVS-ll ultra-
violet lamp(l max = 2537 A). The following solvent
systems were used for paper and thin layer chromatography.

A. 95% ethanol-1.0 M ammonium acetate, pH 7.5
(7:3, V/v).

B. Is0propanol-H20 (9:1, v/v).

C. Ethyl acetate-water-acetate acid—formic acid
(18:4:3:l, v/v).

D. n-Butanol—acetic acid-water (Azl:5, upper
phase, v/v).

E. n-Pr0panol—ethyl acetate-water (7:1:2, v/v).

F. Pyridine-ethyl acetate-acetic acid-water
(5:5:l:3, v/v).

G. Ethyl acetate-acetic acid-water (3:1:1, v/v).

H. 95% ethanol-water (8:2, v/v).

I. 0.5 M LiCl in 2.0 M formic acid.

J. 0.3 M LiCl.

K. Chloroform-methanol-water (65:35:A, v/v).

All electrophoresis was done on Whatman No. 1 paper
using a Pherograph High Voltage Electrophoresis Apparatus

at settings oflllO volts and 30 mamps.

25

b. Buffers.-—The following buffers were used:

 

A. 0.1 M sodium phosphate, pH 7.8, containing
10-4 M EDTA, and 0.02 M B-mercaptoethanol.

B. 0.01 M sodium phosphate buffer, pH 7.4,
containing 10' M EDTA-and 0.02 M B-
mercaptoethanol.

C. 0.1 M glycylglycine, pH 8.5, containing
5 x 10-3 M dithiothreitol.

D. 0.1 M ammonium formate, pH 3.7.

E. 0.2 M ammonium acetate, pH 5.4.-

F. 0.2 M sodium phosphate, pH 7.8.

G. 0.2 M sodium phosphate, pH 8.0.

H. 0.1 M Tris-acetate, pH 8.8.

c. Countingfiof radioactive compounds.—-Radioactive
compounds in solution were counted in scintillation vials
containing 1.0 m1 of the aqueous solution to be counted
plus 10 ml of the dioxane scintillation fluid. Radio-
active compounds on paper, which were insoluble in toluene,
were counted on the paper in scintillation vials containing
18 ml of the toluene scintillation fluid. The dimensions
of the papers counted were about 3.7 x 4.0 cm. All samples
in either system were counted for sufficient times to
obtain a total of at least 10,000 counts. The efficiency
of the dioxane system was determined using benzoic acid-lac
(—COOH group labeled) as a standard. The efficiency of

the toluene system was determined using known amounts of

D-apiose—luC and radioactive Peak III on Whatman 3MM paper.

26

d. Determination of protein.-—Protein was determined

by the biuret method (45) and the Lowry method (51).

Growth of Lemna minor L.

A sterile culture of Lemna minor L.1 was obtained

 

according to the procedure of Hillman (42) using plants
collected on campus from the Red Cedar River. Sterile
cultures were maintained in 1 liter erlenmeyer flasks
containing 300 ml of the growth medium described in the
materials section. Larger amounts of non-sterile Lgmna_
were grown on 3 liters of medium in plastic dishpans
covered with clear plexiglas sheets. Pans were started
with 5-50 Lemna fronds. Both sterile and non—sterile
cultures were grown under banks of 60 or 100 watt light
bulbs adjusted to a height which gave a light intensity of
75 foot candles at the surface of the growth medium.
After five weeks of growth in the pans, the plants were

ready for harvesting.

Standard Enzyme Assay_
The standard enzyme assay contained 1-23 ul of enzyme
solution, 0.00102 umoles of UDP-D-glucuronic acid-luC

(42,460 dpm), 0.0375 umoies of NAD+ and 22 umoles of buffer

 

lTentative identification of the species was based on
the key in Gray's Botany Manual (43) and examination of
non-flowering plants by Dr. G. W. Prescott, Department of
Botany and Plant Pathology, Michigan State University.
Positive identification will require examination of
flowering plants.

27

F in a total assay volume of 50 ul. The reaction was
initiated by addition of enzyme and incubated at 25° for
15 minutes. The reaction was terminated by heating at
100° for 30 seconds. Any visible protein precipitate was
centrifuged. The clear reaction mixture was then spotted
on unwashed Whatman 3MM paper, chromatographed for 15 hours
in solvent system A and scanned on the strip counter (see
Figure 2). Peak III was cut out and counted in the toluene
scintillation system. From the counting data the nmoles
of Peak III formed in 15 minutes was calculated. A unit
was defined as the amount of enzyme required to form one
nmole of Peak III in 15 minutes at 25°. The specific
activity was expressed as units/mg of protein.
Synthesis of 3H—UDP-D—glucoserl‘uc
and SH—UDP—D-glucuronic acid-IFC

Doubly labeled UDP—D-glucose was synthesized from
3H-UTP (2.0 ucuries) and D-glucose-luC-l—P (u.l., 0.425
ucuries) using crystalline calf liver UDP-D-glucose
pyrophosphorylose and a modified procedure of Fitzgerald
and Ebner (49). The 3H-UDP-D—glucose-J'U'C was isolated
from the reaction mixture by chromatography in solvent
system I on PEI paper (50). Doubly labeled UDP-D-
glucuronic acid was synthesized from 3H-UDP-D-glucose-luc
by the procedure of Castanera and Hassid (13). The doubly
labeled UDP-D-glucuronic acid was purified by chromato-

graphy in solvent systems A and E.

28

.2 mloa x 5.0 was coapohpcoocoo +o<z can one 2 mloa x :o.m mm; momma
onmocmpm on» CH coapmnpzoocoo oponmeSm one .monmocmpm canamprmeOAQO mm
poms ohms omoosamlmlmmb one pace OHCOLSOSHwIQImQD .Rm a Canvas mos mammmo

omeHHQSU hog coapmfi>oo Hmucoaapooxo no pHEHH Uo>nomoo one .Eap ooo.m

on ooo.a so am op m was HHH xmom on opopmeSm mo coamao>coo on» has» om a:
wow ohms mmomm< .osmmco ooafion spas ocoo mommm so Bonn mm: coon sozoa one
.oEzmco mo moan: mo.o use cfiopopa no w: m.m pocfiousoo a: m one .mpadmop
one as ponfiaomoo osmuco oonHpSQ on» mo H: m spa: ocoo mo: coauoDSocH one

.< Eopmmm uco>aom CH momma oEmNco pamUQMpm o no mcdmnwoumEonnoll.m ouswfim

 

 

29

HHH xmom

 

omoosawlalmoa

HH
seed

 

oaoo
unacknosamlmummo

 

H seem

 

 

 

 

CHAPTER III

RESULTS

Purification of the Enzyme

 

Purification Procedure

In the introduction, it was stated that the primary
goal of this research was to isolate and study the enzyme
or enzymes responsible for the biosynthesis of D—apiose
from UDP—D—glucuronic acid. The results of the partial
purification of an enzyme system from Lemna minor L. which
converts UDP-D-glucuronic acid-lac to a compound containing
D-apiose-luC are described in this section. A typical
purification is described below.

Lgmna_plants were collected from dishpans by filter-
ing through one layer of cheesecloth, washed with distilled
water, blotted dry and weighed (42 g). All subsequent
steps were done at 4°. The plants were homogenized in 65
ml of buffer A with a Waring Blendor until whole fronds
were no longer present. This normally required about one
minute. The resulting slurry was squeezed through four
layers of cheesecloth and the filtrate was centrifuged at
35,000 x g for 20 minutes. A 0.85 ml aliquot of the 35,000

x g supernatant solution (83 m1 total volume) was assayed

30

31

for protein by the biuret method. A second aliquot (23 ul)
was assayed for enzyme activity using the standard enzyme
assay.

Since the 35,000 x g fraction is stable only for a
few hours at 4°, it was immediately fractionated with
ammonium sulfate (reagent grade). Following the addition
of 22.8 g of ammonium sulfate with continuous stirring,
the pH was adjusted to 7.7-8.0 with 7.5 N NHuOH and the aid
of pH paper. The resultant suspension was centrifuged at
10,000 x g for 10 minutes, the precipitate discarded and
8.0 g of ammonium sulfate added to the supernatant solution.
The resulting mixture was stirred slowly for 5 minutes,
allowed to stand for 5 minutes and centrifuged at 10,000
x g for 10 minutes. The protein precipitate containing
the enzyme activity was dissolved in 3 ml of buffer A.

This fraction could not be assayed for protein since
ammonium ion interferes with the biuret color formation,
or for enzyme activity since the enzyme is inhibited by
ammonium ion. The ammonium sulfate fraction was stored
at -20° until further purification was done. The enzyme
could be stored at this stage for 1-3 months with only a
slight loss of activity.

The ammonium sulfate fraction was then chromato-
graphed on a Sephadex 0100 column. The column was prepared
according to the procedures described in the "Technical

Data Sheets," Pharmacia Fine Chemicals. Its dimensions

32

were 2.2 x 40 cm and it was equilibrated before use with
buffer B. Using a 1000 pl micropipet attached to a 1 ml
syringe, the sample (volume, 3.5 ml; approximately 180 mg
of protein) was applied to the top of the gel bed and
allowed to enter the t0p of the bed. It was washed in
with three 2 m1 portions of buffer B and the reservoir
attached to the top of the column. Using a hydrostatic
head of 15 cm, the flow rate obtained was about 30 ml/hour.
Fractions of 4.8 ml were collected using a Gilson fraction
collector.

In previous experiments it had already been determined
that the enzyme activity was contained in fractions 11-15
(48-72 ml). This fact had been substantiated by a number
of experiments. In one experiment, a 0100 column was used
to fractionate a dialyzed ammonium sulfate fraction. In
that experiment, an elution profile ofprotein and enzyme
activity was obtained by determining the absorbance of
each fraction at 280 mu and by assaying the fractions
containing protein for enzyme activity. The results are
shown in Figure 3.

The enzyme fractions from the 0100 step were combined
for chromatography on a DEAE-Sephadex A-50 column. Before
this step, the combined fractions were assayed for enzyme
activity and for protein by the Lowry method. The sample,
which contained 52 mg of protein in a volume of 24 ml,

was chromatographed on a 1.0 x.10 cm column equilibrated

33

 

.zpfi>fipoo oEmNno mom nozommo onoz AHE omlmnv
malm mnOHpoonm no muonvaao H: OH was o.m mos nospoonm nooo mo oenao> one
.oEmNno mo mpHn: am one naopond no we mm nonfiopnoo nno HE m.H mos oasao>

oHQEom one

.ooumamfiv eno .pxop on» nfi nonanomoo onsoooonn onp en onqu
no m mm Bone oonononn mos nOHpoone opoMHSm EdflnoEEo one

.ooao xoeennem
no nonpoonm opoMHSm EanoEEo eonzaofio o no mnemnmo»o€onnoul.m onswfim

34

scrum

(UIw SI/PamJQJ III Mead setomU)

omH

HE .oenao>

 

0H.r

om 1..

0:1.

om i.

 

 

nHouonmN

 

 

 

 

 

ro.H

ro.m

 

088V

35

with buffer B. The combined G100 fractions were poured
into a funnel attached to the top of the column and allowed
to flow at a rate of 20 ml/hour until the sample had
emptied out of the funnel. Then, the column was connected
to a reservoir containing buffer B and washed at a flow
rate of 25 ml/hour. Beginning with the sample application,
u.8 ml fractions were collected and the absorbance at 280
mu determined for each. When the absorbance dropped to

the background level of 0.1, the buffer was changed to
buffer B containing 0.1 M NaCl. When the absorbance
dropped again to 0.1, the buffer Was changed to buffer B
containing 0.2 M NaCl. The protein peak obtained from the
0.2 M NaCl elution contained the enzyme activity. The
protein fractions from the wash were combined (fractions
U-6) and then assayed for enzyme activity, and for protein
by the Lowry method. The same was done for the protein
fractions from the 0.1 M NaCl elution (fractions 12—1“).
Each protein fraction from the 0.2 M NaCl elution (frac-
tions 20-23) was assayed for enzyme activity and also for
protein by the Lowry method. The protein elution profile
is shown in Figure 4. Fraction 20 and combined fractions
21 and 22 from the DEAE-Sephadex step were concentrated by
vacuum dialysis. The following concentration procedure
was devised, based on the method of Peterson and Sober (52).
Two pieces of dialysis tubing previously soaked in water

were cut to a length of 30 cm. A triple knot was tied at

 

36

.mE»Nco mo mafia:

omm paw campopm mo we mm oocfiopcoo mHQEmm one .pxmp on» CH oooamomoo

cEsHoo ooam mnp Eonm AHE o.m~|o.m:v malaa mcofipommm mo oopwfimcoo madamm
one .xoomnaomlm<mm co mcoauomnm ooaw mo mnmmmmOmeoscoll.: muswfim

37

HE .mESHo>
OOH mm om . mm

b
d

. I'l.

\ //. ./.

/ moan: mam .//moacs o.H moms: m:

./.\._

 

 

.TIIsz z m.o||LT|.I.|:sz 2 To T 233 VF

 

 

 

Hm 088v

38

one end of each and the other end slipped over the tip of

a funnel stem that was about the same diameter as the
dialysis bag. The loose end of the dialysis bag was pulled
through a hole in a #8 rubber stopper, and the funnel

stem carefully pushed through the hole so that the dialysis
bag did not tear. Each dialysis bag was placed inside a 1
liter suction flask containing 100 ml of buffer B. Each

of the two enzyme solutions (at no) to be concentrated
was.poured into a dialysis bag and any air bubbles worked
out of the bag with a thin glass rod. The flasks were
connected to a vacuum pump with a stopcock in the line
between the flask and the pump, and evacuated until the
buffer in the flask began to bump (after 30 to 60 seconds).
These steps were done at room temperature. Following
evacuation, the stOp cock was closed off and each apparatus
moved into the cold room. The concentration time for each
enzyme solution was about 10 hours. After concentration of
the enzymes to a volume of 0.1—0.3 ml, the dialysis bags
were washed with 0.5 ml of buffer A and the enzyme solu-
tions removed with a pipet. An aliquot of each was

assayed for protein by the Lowry method, then an equal
volume of 95% glyéerol added, and 5 ul aliquots assayed

for enzyme activity. A summary of the results of the

purification are shown in Table l.

39

.mm was Hm mcoapomso

omcHoEoo mafia om coapommmo

.mmlom mQOfipommmn

.mHIHH meadpomsmm

 

 

ms Hma :.mm oa.om :.m oeoapmspemoeoo
:m mam m.mm oa.ma 2H omamo
om 0mm m.m mm.: mm mooao
OOH mmm o as.o mm m x ooo.mm
R as
ano>ooom wwwmm coaummwwwnsm we\mpacs menao> doom

 

.pxmu on» CH omnfimommo mEzNCm map mo soapmofimaszo on» no msoEEJmII.H mqm¢9

“0

Stability of the Enzyme

The following eXperiment was done to determine the
effect of glycerol on the stability of the enzyme. An
enzyme preparation was made using 38 g of Lemna and
carried through the 6100 step of the purification proce-
dure. At that point the enzyme was concentrated by
vacuum dialysis. The concentrated enzyme was divided into
two equal aliquots; one was assayed for protein and enzyme
activity, and the other was mixed with an equal volume of
95% glycerol and assayed for enzyme activity. Each
aliquot was stored at -20° and periodically assayed for
enzyme activity. The results in Table 2 show that the
presence of 50% glycerol increases the stability of the
enzyme.

A similar experiment was done with a concentrated
enzyme from the DEAE step prepared from “3 g of Lemna.
The results in Table 3 show that the purified enzyme was
stable for 3 days and that it had not yet lost 50% of the
activity after 5 weeks.

Partial Characterization of the
Purified Enzyme

Determination of the Linearity
of the Enzyme Assay

In the course of this research, one criterion used
for determining the linearity of enzyme assays at each

step of the purification procedure was the determination

41

TABLE 2.--Stability of the concentrated enzyme from the
G100 step at -20°.*

 

Specific Activity

 

 

Days
Without With
Glycerol Glycerol
o (4.20 11.39
2 3.85 “.28
5 2.20 3.18
12 1.89 2.23
25 1.13 1.65
39 O.NO 1.H8

 

*The concentrated enzyme solutions with and without
glycerol had protein concentrations of u.u and 8.8 mg/ml,
respectively. The amount of protein in each assay was
varied so that the units of enzyme was constant (about
0.07 units).

of the amount of Peak III formed in 15 minutes for varying
concentrations of protein. That criterion plus a second
was used for the purified enzyme, namely, the formation

of Peak III with respect to time. Figure 5 shows that the
amount of Peak III formed in the standard 15 minute assay
using purified enzyme is linear over a protein concentra-
tion range of 0.05 to 0.20 mg/ml of assay mixture. When
the formation of Peak III with time was determined, the
results shown in Figure 6 were obtained. Inspection of

the data showed that the assay was not linear for 15 minutes,

 

M2

.sfiouopa no wE\mpHcs

ma mo: oEmNso onp mo zpfi>fipoo camaooam one .coapospcoocoo :Hoposa onp mo

om\H ma mommo so ca pom: cfioposa mo pesoso one .osspxfie zommo mo HE\wE mm

commosaxo ono occapospcoosoo cfioposm .mommo oEmNco unaccoum one ca pom:

opoz doom m<mo ogp Eosm oEmNco ooponpcoosoo mo mundoeo wcamno> .cfiopona
mo mpssoso wcfimmonosfi npfiz zpfi>dpoo osmNco mo moasooCHQIl.m opswfim

opsuxfie.hommo no HS\wE .QOfiuonpcoocoo :fioponm

 

 

mm.o om.o mH.o OH.o mo.o
I]? 4. x .1 1
O
: mo.o
3
.u. 0
: oa.o
0 AW
: ma.o

arm 91 u: pamao; III need JO satomu

 

uu

.o.m ma no m o>CCo oCo m.» mg no oCoo mo3.< o>sso .mommo UCoUCopm

on an ooCHECopoo mo .mHo>HuooCmoC .CHoponq mo w8\mpHCs m.mH oCo 0.:H mm:

m UCo C mo>CCo pom pom: moEmNCo on mo mpH>Hpom OHMHoon one .Qoum m<ma on»

Eosm oEmNCo ooponCooCoo Co CoHpoCoaonQ pCopommHo o Cqu oCoo mo: o>uCo Comm

.oCCCxHE zommo mo HE\wE mmH.o m CH oCo opsprE momma mo He\we mmH.o mo:

< CH COHponCooCoo CHouosa oCu mmHo>HpoonoC .2 mIOH x ww.m oCo z mIOH x :o.m

omoz m oCo < mo>CCo CH mCoHpospCooCoo oHoo oHCopsousnmanD ”com: mo: mommo
opmoCopm one .oEHu Csz HHH xoom mo COHpoECom on CH omooCoCHll.m oCCme

CHE .oEHB

 

cm on om om OH
\ ..mo.o
. o
.\\\\\\ o\\\\\\\
5 .\ .
A :3 o
< o>C5o
. o
o
o :35
m o>Czo
o
L.om.o

 

pewaog III Head JO setowu

46

TABLE 3.--Stability of the concentrated enzyme after the
DEAE step at -20°.*

 

 

Days Specific Activity
0 13.00-
1 lu.u1
2 13.30
3 15.39
7 10.60
20 8.6a
3“ 7.5“

 

*

The enzyme solution had a protein concentration of
2.52 mg/ml. The amount of protein in each assay was
varied so that the amount of enzyme was constant (0.08
units).

but only for about 5 minutes. The experiment was then
repeated for shorter times. The results in Figure 7 show
that at a substrate concentration of 0.58 x 10"5 M, the
rate was linear for 2.5 minutes, while at 2.00 x 10"5 M,

the rate was linear for at least 7 minutes.

Determination of the Optimumng

The amount of Peak III formed in 15 minutes was
determined in four different buffers and at different pH
values using the standard enzyme assay. The buffers were:

0.1 M sodium phOSphate, pH range 6.5-8.0; 0.1 M tris-HCl,

147

 

 

.CHopoma no we\mpHCC m.oH mos oEmNCo on mo sz>Hpoo

OHQHoon one .mommo oEmNCo mo HE\wS mMH.o mo COHuonCooCoo CHoCOCQ o con

oCo o.m ma pm oCoo oCoz mummmo oCB .Qopm m<mo on Eopm oEmNCo oopoCuCooCoo

on mo COHpoCoquQ o Cqu oCoo oCoz whommo Had .mHo>HpooamoC .2 mIOH x :o.m

oCo z mIOH x mm.o oCoz m oCo < mo>CCo Com mCOHmeuCooCoo o HIoHoo oHCopdoCHw

IQImQD on poooxo .Uoms mo: momma osmmCo UCoUCopm oCB .mCOHpo pCooCoo opompmnsm
pCoCoMMHo 03» no oEHp Csz HHH xoom mo COHpoECom oCu CH omooCoCHll.~ oCCme

CHE .oEHB

 

om m m a m
< o>CSo

..mmo.o

O
00 .r mo.o

U.
0
:mso.o
O

m o>Czo 4 0H.o

 

pewaog III xeaa JO setomu

49

pH range 7.5-9.0; 0.1 M sodium glycylglycinate, pH range
7.0-9.5; and 0.1 M sodium glycinate, pH range 8.0-10.5.

The results of this experiment (Figures 8 and 9)
showed that the enzyme had an Optimum pH of 8.2 in Tris
and phosphate buffers, and 8.5 in glycine and glycylglycine
buffers. Optimum formation of Peak III was obtained with
phosphate buffer. Even though the assays were not done
using conditions which gave linear rates, the optimum pH
values observed are valid. This is true because linear
rates would only increase the sharpness of the pH Optimum
curves and would not change the observed optimum.

Determination of the Optimum
NAD+ Concentration

 

 

The amount of Peak III formed in 15 minutes was
determined over a 100 fold concentration range of NAD+.
The optimum NAD+ concentration, as shown in Figure 10,
was 0.8 x 10"3 M. These assays were also done under
assay conditions which gave non-linear rates. Again the
optimum concentration found is still valid since linear
rates would only change the sharpness of the peak and

would not shift the optimum.

Determination of Km and V ax

Using the standard 15 minute assay, the amount of
Peak III formed was determined for UDP-D—glucuronic acid-140
concentrations ranging from 0.22 x 10'"5 M'to 18.04 x 10"5 M.

Figure 11 shows the results. These data indicated that

 

50

.Commsn opoCmmoCQ CH o.m mm pm CHopopa mo mE\mpHC: o.mH no: COHpoCoquQ

on mo mpH>Hooo oHMHooam one .aopm m<mo on» Sony oEmNCo ooponCooCoo

on» mo COHpoCoaoCQ o Csz oCoo onoz mmommo HHC .zommo oEmNCo mo HE\wE

mmH.o mos momma Como CH COHponCooCoo CHoposq one .HomImHCp z H.o Csz

m o>Cso oCo opoCQmoCQ ECHoom z H.o Csz mos < o>CCo .ma on pooamoh Csz
momma opsCHe mH UCoUCoom on CH HHH xoom mo COHpoECom oCBII.m oCCme

51

.<
o
> a
$4
3
o
0

curve B

 

10.0

9.0

8.0

pH

.0

 

 

Ln C) ux
Hi .4 C3
c> c: c:

utm gt a: pamao; III Head JO sstomu

6.0

 

52

.m oCCMHm CH pom:

mo: non» oszuCo oEom on Csz oCoo oCoz whommo HH< .Am.oH oCo 0.0H .m.m
.o.m ma no m o>Cso CH HE\wE mmH.ov oCCprE momma no Hs\ms omH.o mos zommo
Como CH COHponCooCoo CHoCOCQ one .opoCHosz ECHoom z H.o Csz m o>sso
oCo opoCHomHmHzome ECHoom 2 H.o Cqu mm: C o>Cso .ma op pooamos Csz
momma opsCHE mH oCooCopm on CH HHH xoom no CoHpoECom oCBIl.m onstm

ma
0.0H o.m o.w 0.5 o.m

 

m o>nso

O x o
O
/ .. 3.0
O
/ . 3.0
\ :36
< o>pso

 

uIm gt a: pewao; III neea JO satomu

5“

.HO

.oEzuCo mo mpHCC mHH.o ooCHopCoo mommo Comm .mommo

oCooCopm on CH ooCHECopoo mo CHopOCQ no we\mpHCs m.mH mos COHpoCoQoCQ

on mo muH>Huom oHMHooam one .Qopm mCMQ on Eopm oEano oopoppCooCoo

oCu mo COHooCoooCo o Csz oCoo opoz mmommo HH< .ohsprE momma no He\me

omH.o mos COHponCooCoo CHopoma on» oCo com: mo: momma UCoUCoum oCe .+Q<z
mCOHuonCooCoo wCHmooCoCH CpHB HHH xoom mo COHpoECom oCenI.OH oCCmHm

55

 

k%_
100

1

’40
[NAD+] x 10”, moles/1

 

 

 

0...;

6 Lb
H O
o' o'
uIm SI UT pemJOJ III Heed JO SQIOwU

 

56

.oEmNCo mo mpHC:

mHH.o UoCHopCoo momma Comm .mommo chooCopm on CH ooCHECopoo mo CHouoCQ

mo wE\mpHCC m.mH mos COHpoCoQoCQ on mo mpH>Hpoo OHMHooCm oCB .Qopm m<mo

on EoCm oEano ooponCooCoo on mo COHpoCoQoCQ o Csz oCoo oCoz ozommo HH<

.ohspre momma no HE\mE omH.o mos momma Como CH COHpoCuCooCoo CHopomq on» oCo

pom: mo: momma oEmNCo oCooCopm oCB .CoHponCooCoo o HnoHoo oHConsousIQImQD
wCHmooCoCH Cqu HHH xoom mo COHpoECom on CH omooCo H oCBII.HH oCCme

57

mH

-
1

‘H\oHos .mOH x Hammoog
OH

«PI-n

 

 

1r om

 

l

TOOH

 

%
Woe

UTm SI u: pamJOJ atOI x III Mesa J0 seIow

58

the eXperiment should be repeated using a substrate con-
centration range of 0.4 to 2.0 x 10"5 M. Also, the assays
had to be done at shorter times so that the reaction rates
would be linear. The results of that experiment are shown
in Figure 12. The data were then plotted as l/v vs l/[S]
by the least squares method and are shown in Figure 13.

The equation as determined by the least squares method was:

1 = (0.05741 x 107)x = 0.06604 x 1012.
- a 12
Since l/Vmax - y intercept 0.06604 x 10 ,
then Vmax = 2.405 nmoles/min/mg of protein.
- = 7
Since Km/Vmax - m 0.05741 x 10 ,

then Km = 0.870 x 10’5 M.

Effect of MgCl9 on the Formation
of Peak III in“Different Buffers

 

Standard assays were done in 0.1 M sodium phosphate
buffer, pH 7.8 and in 0.1 M Tris-HCl.buffer, pH 8.25 with
and without 2 mM MgC12. Each assay had a protein concen-
tration of 0.126 mg/ml of assay mixture. As shown in Table

4, 2 mM MgCl had very little effect on the enzyme

2
activity in phOSphate buffer; however, in Tris buffer the
activity was increased to the same level as in phosphate
buffer. An increase in activity was also noted with
glycylglycine and glycine buffers. These results indicate.
that there may be a metal ion requirement, but this require-

ment is either not absolute or it is satisfied by metal

contaminants. Other metal ions were not tested.

 

59

.oEmuCo no mpHCC mmo.o ooCHopCoo 2ommo Comm .o.w ma no
mommo oCooCopm on CH ooCHECopoo mo CHopoua mo we\mpHC: m.HH mos COHpoCoCoCQ
on» no 2pH>Hpoo oHMHooam one .aopm m<mo on Eomm osmuCo oomepCooCoo oCu no

COHmeoaoCQ m Csz oCoo oCoz mmomoo HH< .mopsCHE m mo: oEHp momma on,.mC0Hp

IonCooCoo ConHC no ”mopsCHE m was oEHp mommo on .2 mIOH x :mm.o no mmouCCHE

m mo: oEHp mommo oCu 2 mIOH x :m:.o oCo 2 mIOH x mo:.o mo mCOHuonCooCoo

o HloHoo oHCopsosHmlolmQD p< .o.m mos mo oCu oCo oCCuxHE mommo mo HE\mE

mmH.o mo: COHponCooCoo CHopOCC oCB .COHponCooCoo o HloHoo OHCOCCoCHwIQImQD
CH ommoCoCH Csz COHpoECom HHH xoom mo ouoC on CH om oCoCHll.mH oCCme

60

o.m

1p

H\moHos .mOH x Hammopu

m.H o.H

d

y

+m

:OH

 

uIm/pemaog ZIOI x III'HBGJ J0 setom

 

61

.CHopOCQ mo wE\CHE\ooECom HHH xoom no moHosC mo:.m n

on> m2mIOH x onm.o u Ex .UOCpoE mopoddm pmooH on 2n oopuoHQ onos muCHoo

22.

.NH oCCme CH Csosm mpmo on mo COHQ xCCmICo>oo3oCqul.MH onstm

 

b
d

H\moHos .mOH x Haemosg

 

db

H

1b

9

rm.o

 

OI x 0A

8T

‘

qu/sstom

 

63

TABLE 4.--Effect of MgCl on the formation of Peak III.*

 

 

2

Units/mg
Buffer and pH With MgCl2 WithoutMgCl2
0.1 M Phosphate, 7.8 14.00 13.30
0.1 M Tris, 8.25 14.11 12.24

 

*

‘A preparation of the concentrated enzyme from the
DEAE step was used. The protein concentration in each
assay was 0.150 mg/ml of assay mixture.

Characterization of the Enzymatic
Products

 

The Initial Isolation and
Identification of Enzymatically
Synthesized D—apiose-qu

The first experiments tried in which enzymatically
synthesized D-apiose-luC was tested for directly, were
started in August, 1966. Ten months later, after many
unsuccessful attempts, that goal was finally achieved.
The two experiments described below were the first experi-
ments performed in which D-apiose-luC was enzymatically
formed from UDP—D-glucuronic acid-lac and identified.

a. Crystallization of D—apiosegphenylosotriazole-
:iflg.--In the first experiment, 0.5 g of Lgmg§_plants
were ground with buffer C in a glass homogenizer at 4° and
used without filtering or centrifuging.. The following

incubation mixture was used: 0.5 m1 of crude homogenate,

0.08 nmoles of NAD+, 0.008 nmoles of UDP-D-glucuronic

64

acid-luC (1.0 ucuries) and 90 nmoles of buffer C in a total

volume of 2.0 ml. The mixture was incubated for four
hours at 25°, the reaction stopped by the addition of 4

m1 of absolute ethanol and the solution filtered through
Whatman No. 1 filter paper. The filtrate was concen-
trated, chromatographed on paper in solvent system E for
16 hours and scanned on the strip counter. The majority
of the radioactivity remained at the origin and was eluted
with water. An aliquot (571,500 dpm) was hydrolyzed in
1.0 N H280“ for one hour at 100°. The hydrolysate was

neutralized with BaCO filtered and the filtrate deion-

3,
ized by passage through a column of Dowex 50W—X8, 50-100
mesh, H+ form, and Amberlite IR—45, 20—50 mesh, 0H' form.
The deionized solution was concentrated in vacuo,
chromatographed on paper in solvent system D for 24 hours
and scanned on the strip counter. The results from the
scan are shown in Figure 14. Since solvent system D

does not separate D-apiose from D-glucuronolactone, the
third peak from the origin was eluted with water and the
following experiments were done. An aliquot (10,935

dpm) was mixed with 0.25 nmoles of carrier D-apiose,
chromatographed on a thin layer plate in solvent system

F and scanned on the strip counter. The results from the
scan showed that two radioactive ' components were present.
One had the same R as D-glucuronolactone and the other

f
had the same Rf as D—apiose. In order to identify the one

 

65

.pCoEHCoaxo oHouoHCu on» pom pom: mo: oCo
ozHloCOpooHOComsosHmlo oCo ozHlomoHaoIQ mo oCCprE o mm: CHwHCo on EOCM xooa
UCHCp one .oCzanoo wCHCHopCooIomOHColm ooNHmonCmm 2HHooHpo82NCo umCHm on
Eopm opom2Honozn oCu mo Q Eopmmm pCo>Hoo CH 2CQonouoEoCColl.:H oCCmHm

66

oCouooHOCOCCoCHmIQ oHoo
oCo oooHQotn omOHaxIQ oHCopsoszuo

000

 

CHwHCo

67

radioactive compound as D-apiose-luc, a crystalline
derivitive, D-apiose phenylosotriazole—luC was synthe-
sized. If D-apiose-luC was present, the derivitive could
then be crystallized to constant specific activity. An
aliquot of the material from the third peak (22,500 dpm)
was combined with 0.2012 g (1.34 mmoles) of carrier
D-apiose. The triazole derivitive was synthesized
according to the procedure of Kindel.2 The derivitive
was recrystallized four times from diethyl ether-
petrolium ether, B.R. 30-60° (1:2.5, v/v). The results
in Table 5 showed that the derivitive was crystallized
to a constant specific activity. The data permit two
conclusions; first, D—apiose-luC was formed in the incu-
bation mixture, and second, the radioactive solution
used contained 33% D-apiose-luC (based on the starting
and final dpm/mmole of D—apiose).

b. Periodate oxidation of enzymatically synthe-
sized D-apiose-luC.--The D-apiose—luC used for this
experiment was obtained using an enzyme preparation dif-
ferent from that used in the triazole experiment. In
this case, about 2 g of Lgmna_was ground in a glass
homogenizer in buffer C and centrifuged at 35,000 x g

for 20 minutes. Using that preparation, the following

 

2P. Kindel, manuscript in preparation for submis-
sion to Carbohydrate Res.

68

TABLE 5.—-Re0£ystallization of D-apiose phenylosotriazole
-1 C to constant specific activity.

 

 

51:23:13.2:- chazzizl ”3:121“: ifiififif,
tallization

mg % °C fi%§%3
1 114.7 49.2a 91.5-93.5 6170
2 89.3 83.6 95.0-96.5 4660
3 67.6 84.0 95.0-96.0 5860
4 46.7 79.5 95.0-96.0 5320
5 26.0 66.8 94.5-95.5 5520

 

aOverall yield based on the starting mmoles of
D-apiose.

incubation mixture was used: 0.2 m1 of enzyme, 0.125
umoles of UDP-D-glucuronic acid-lac (217,400 dpm), 0.5
nmoles of NAD+ and 25 nmoles of buffer c. The reaction
mixture was incubated for 90 minutes at 25° and the
reaction terminated by heating at 100° for 3 minutes.
After the addition of 2.0 nmoles of carrier D—apiose,

the mixture was hydrolyzed in 1.0 N H250, at 1000 for

1 hour, deionized as before and chromatographed on paper
in solvent system C for 10 hours. After the chromatogram
was scanned on the strip counter, the peak which corre-

sponded to known D—apiose was eluted with water (Figure

15). An aliquot of the D-apiose-luC solution (about

 

69

.pCoEHCono

COHpoono opoBOHCoo one pom pom: mo: owHIomOHmolo oCB .COHpoCoQoCQ

osmmCo w x ooo.mm o Csz o
muoCUOCC noN2H0C©2C mo 0 Eopm

m

HIUHoo oHCOCCoCH Dumas mo COHpoCCoCH Co Eopm
m pCo>Hom CH 2CmomwouoEOCColl.mH oCCmHm

7O

oCopooHOCossoiHmln

O

omOHQ<IQ

O

oHoo
omonxlo oHCopsouslm

OO

 

 

 

 

 

 

 

CHwHCo

71

5,000 dpm) was chromatographed on thin layer plates in
solvent systems C and F. The plates were scanned on the
strip counter; the scan showed that the radioactive com—

pound had the same R as D-apiose. To further character-

f
ize the compound as D-apiose-luC, the following experi-
ment was carried out. The remainder of the D-apiose—luC
solution (21,957 dpm) was combined with 2.83 poles of
carrier D-apiose, oxidized with sodium meta-periodate
and the products separated on a Dowex l X-8 column (200-
400 mesh, acetate form) by the procedure of Kindel.3
The results of the periodate oxidation are shown in
Table 6. The per cent recoveries were based on the
assumptions that exactly 1 mole each of formaldehyde and
glycolic acid and 2 moles of formic acid were formed

from each mole of D-apiose, and that the D-apiose-luC was
uniformly labeled.

Since the formation of glycolic acid upon periodate
oxidation is a distinguishing characteristic of D-apiose,v
further studies were done with the glycolic acid fraction.
The glycolic acid concentration was determined by the
colorimetric procedure of Calkins (46) and the specific
activity calculated. An aliquot of the glycolic acid
fraction (1,634 dpm, 0.815 nmoles) was chromatographed
on Whatman No. 1 paper ascendingly in solvent system G

3P. Kindel, manuscript in preparation for sub-
mission to Anal. Biochem.

72

 

 

TABLE 6.——Results of periodate oxidation of D-apiose-ific
enzymatically synthesized from UDP—D-glucuronic acid- C.
Theoretical Observed Z of
Fraction dpma dpm theoretical
Formaldehyde 4391 3891 88.6
Glycolic acid 8783 8172 93.0
Formic acid 8783 8082 91.9

 

aThe sample contained 21,957 dpm in 4.63 nmoles of-
D-apiose. The column size was 0.8 x 10.5 cm.

for 6 hours. The chromatogram was scanned on the strip
counter, the radioactivity eluted and the specific
activity determined as described for the fraction from
the column.

The theoretical specific activity of glycolic.acid
based on the specific activity of D-apiose-luC used,
was 1,896 dpm/mmole. The specific activities of glycolic
acid isolated from the column and after paper chromato-
graphy were 2,003 dpm/mmole and 1,710 dpm/mmole, respect-
ively. Since the uniformity of labeling for each carbon
atom of D-glucuronic acid is within t 15%, the results
from these two experiments are well within this varia-
tion and further prove that the enzymatically synthesized

lac-compound isolated from the chromatogram was D-apiose-

14C.

73
Enzymatic Synthesis of
Peaks 1, II and III

The results described thus far have been concerned
only with the identification of D-apiose-luC isolated
from hydrolyzed enzyme incubation mixtures. The remain-
ing sections will describe the characterization of the
enzyme products, Peaks I, II and III (see Figure 2).

a. Synthesis of Peaks I,-II and III usingluC
labeled substrate.--Using 78 g of Lemna, a dialyzed
ammonium sulfate fraction was prepared as described
earlier and used for this experiment. After the centri-
fuged protein was dissolved in buffer A, it was dialyzed
against sufficient changes of the same buffer until the
Nessler test for ammonium ion in the dialysate was.
negative. The incubation mixture contained 0.150 ml of
the dialyzed enzyme preparation, 0.0112 nmoles of UDP-
D—glucuronic acid—lac (1.9 ucuries), 0.5 umoles of NAD+,
and 25 nmoles of buffer F in a total volume of 0.5 ml.
The mixture was incubated at 25° for 30 minutes and the
reaction terminated by heating at 100° for 30 seconds.
The precipitated protein was centrifuged, the super—
natant solution spotted on paper and chromatographed in
solvent system A for 12 hours. The chromatogram was
scanned on the strip counter and each of the three radio-
active compounds eluted with water. At this point, the

per cent conversion to each peak was determined; Peak I

was 16.4%, Peak II was 49.4% and Peak III was 21.8%.

 

74

Since Peaks II and III contained only pentoses, the

1”002 was taken into account for

making those calculations. Each of the three peaks was

loss of carbon atom 6 as

rechromatographed on paper in solvent systems E and H.
Unless otherwise stated, these purified compounds were
used for the identification of Peaks I and II, and for
the characterization studies of Peak III.

b. Synthesis of Peaks I, II and III using 3H—UDP-

D—glucuronic acid—luC.--Doubly labeled products were

 

synthesized to determine whether or not Peaks I, II and
III contained uridine.
An incubation was done using 0.000123 umoles of

1”0 (17,500 cpm of 1”C and

3H-UDP-D-glucuronic acid-
31,150 cpm of 3H, or a 3H/luC = 1.78) with 100 01 of the
dialyzed enzyme used in part a, 0.2 nmoles of NAD+ and

10 nmoles of buffer F in a total volume of 200 ml. The
reaction mixture was incubated at 25° for 30 minutes and
the reaction terminated by heating at 100° for 30 seconds.
After removal of precipitated protein by centrifugation
the supernatent solution was chromatographed on paper in
solvent system A for 14 hours and scanned on the strip
counter. Peaks I, II and III were eluted with water,
chromatographed on paper in solvent systems E and H and
the 3H/luC ratios determined after each step for each

compound.

75

Peaks I, II and III, synthesized from 3H-UDP-D-

glucuronic acid—luc, contained 2,657 dpm, 10,573 dpm and

14

7,748 dpm, respectively (calculations based on C only).

These numbers were obtained after chromatography in sol-

vent system E. The 3H/luC ratios are shown in Table 7.

TABLE 7.--3H/luC ratios of Peaks I, II and III.

 

 

Step Peak I Peak II Peak III
Incubation 1.78 --- -—-
Solvent System E 1.36 1.95 0.125
Solvent System H ' 1.48 1.98 0.038

 

The significance of the values are discussed in the

following sections.

Identification of Peak I

 

Peak I was identified as UDP-D-glucuronic acid-lac

by three criteria.

a.j Paper cochromatography.—-Each of two aliquots

 

of radioactive Peak I (3,309 dpm) was mixed with 0.2
umoles of UDP-D-glucuronic acid. One was chromatographed
in solvent system A and the other was chromatographed in
solvent system J on PEI paper., The chromatograms were
then scanned on the strip counter. UDP-D-glucuronic

acid and Peak I had the same R in each solvent system.

f

76

 

b. Hydrolysis and identification of the products.-—
An aliquot of Peak I (6,670 dpm) was hydrolyzed in 0.1
N H280“ for 15 minutes at 100°. The hydrolysate was,
neutralized with 5.0 N NaOH, and 9 volumes of absolute
ethanol was added. After cooling to 4°, the precipi-
tated inorganic salts were centrifuged at top speed in r
a clinical centrifuge for 5 minutes. The supernatent
solution, which contained the free sugars, was concen-

trated in vacuo and chromatographed on paper in solvent

 

system C for 10 hours. The chromatogram was scanned on
the strip counter. There were two radioactive peaks on
the scan; one had the same Rf as D-glucuronic acid and
the other had the same Rf as D-glucuronolactone.

c. Determination of the 3H/luC ratio.-—As shown
in Table 7, the 3H/luC ratio of Peak I was 1.36 after
chromatography in solvent system E and 1.48 after
chromatography in solvent system H., Since the ratio was

1.78 for the 3H-UDP-D-glucuronic.acid'lu

C used in the
doubly labeled substrate experiment, it was concluded
that Peak I contained uridine. Therefore, the results
from the three experiments described above prove that

Peak I was UDP-D-glucuronic acid.

Identification of Peak II

 

Peak II was identified as UDP-D—xylose-luC by three

criteria.

 

77

a. Paper chromatography.-—Each of two aliquots
of radioactive Peak II (3,388 dpm) was mixed with 0.1
nmoles of UDP-D-xylose. One aliquot was chromatographed
in solvent system A and the other was chromatographed in
solvent system J on PEI paper; then both chromatograms
were scanned on the strip counter. The results showed

that Peak II had the same R as UDP-D—xylose in both

f
solvent systems.

b. ,Hydrolysis and identification of the product.--
The possibility existed that UDP-D—apiose-luC could be
present in small amounts in Peak II. Therefore, a large
amount of radioactivity was used for the hydrolysis of
Peak II in order to detect D—apiose-luC present in levels
greater than 0.5%. An aliquot of Peak II (67,750 dpm)
was mixed with 0.25 nmoles of D-apiose and 0.5 umoles
of D-xylose and hydrolyzed in 0.02 N H280“ for 15
minutes at 100°. The hydrolysate was worked up as
described for Peak I, chromatographed on paper in solvent
system C for 10 hours and scanned on the strip counter.
The hydrolysis of Peak II resulted in the release of D-
xylose-luC as shown by the chromatogram scan (Figure 16).
Of the total radioactivity in Peak II, 89% was recovered
as D-xylose—luC, 2% was present in a peak having a lower
R value than D-xylose, and the remainder was unaccounted

f
for. No D-apiose-luC was detected, which means that if

 

 

78

.o Eoumm
m
pCo>Ho
m CH
HH xoo
m
woumHoComg no 2
Cmmhmo
meOC
Conn.
mH oCCw
Hm

79

omOHQCIQ

omonxIQ

 

 

CHwHCo

80

14

D-apiose- C was present, the level was less than 0.5%

of the total radioactivity in Peak II.
c. Determination of the 3H/luC ratio in the doubly
labeled substrate experiment.--As_shown in Table 7, the

final 3H/lLIC ratio of Peak II was 1.98. With the loss

14

of carbon atom 6 as 002, and using the 3H/luC ratio r

of 1.48 for 3H-UDP-D-glucuronic acid-l”

3H/luC for UDP-D-xylose-luc would be 1.78. Again, the

C, the theoretical

presence of 3H label proves that uridine is present in

Peak II. The three experiments described above proved

 

that Peak II was predominantly UDP-D-xylose-luC, and
furthermore, that UDP-D-apiose was not present in Peak II
to an extent greater than 0.5%.

Partial Characterization of

PeakrIII

a. Paper and thin layer chromatography.--Separate

 

aliquots of Peak III (3,609 dpm) were chromatographed in
solvent systems A, E and H on paper and solvent system K
on thin layer plates, then scanned on the strip counter.
In the case of paper chromatography, Peak III chromato-
graphed as a single peak in all three solvent systems.

On thin layer plates, Peak III chromatographed as a single
peak with an Rf of 0.18. Phosphatidyl choline chromato-

graphed on the same thin layer plate with an R of 0.23.

f
Compounds such as UDP-D-glucose or glucose—l-P do not

81

migrate on thin layer plates in that solvent system. Thus,
Peak III appears to have non-polar characteristics.

b. Electrophoresis.——Separate aliquots of Peak
III (3,609 dpm) and 0.1 nmoles each of UDP—D-glucose and
UDP-D—glucuronic acid were used for electrOphoresis as
described in the general methods. Electrophoresis was

conducted using buffers D and E, and the papers were then

_ u' 3| manta-MW

scanned on the strip counter. Peak III migrated as a

single peak at pH 3.7 and 5.4., After 1 hour at pH 3.7,

 

the relative mobilities of UDP-D—glucose, Peak III and
UDP-D—glucuronic acid were 1.0, 1.14 and 1.26, respect-
ively. After 2 hours at pH 5.4, the relative mobilities
were 1.0, 1.22 and 1.37, respectively. These results
showed that Peak III behaved as an anion.

The results from parts a and b indicated that Peak
III was a single, homogeneous compound.

0. Determination of the 3H/luC ratio of Peak III
from the doubly labeled substrate experiment.--One purpose
for doing the doubly labeled substrate experiment was to

determine whether or not Peak III contained uridine. The Rf

of Peak III in solvent systems A and K already suggested
that Peak III was not UDP-D—apiose. Sandermann (31)

reported that UDP-D-apiose had the same R as UDP-D-

f
glucose in solvent system A. The 3H/luC ratio of Peak

III synthesized from 3H-UDP-D-glucuronic acid was 0.04.

82

This demonstrated conclusively, that Peak III did not

contain uridine, and therefore was not UDP-D-apiose—luC.

 

d. Identification of D-apiose-luC from hydrolyzed
Peak III. Two experiments were done: (1) paper cochromato-
graphy and (2) periodate oxidation.

For the first experiment, an aliquot of Peak III

(about 15,000 dpm) was mixed with 1.0 umoles of D-apiose

 

and hydrolyzed with 0.1 N H280“ at 100° for 60 minutes.
The free radioactive sugars were isolated from the hydro-

lysate using the same procedure described for the hydroly-

 

‘l‘

sis of Peak I, chromatographed in solvent system B for

17 hours and scanned on the strip counter. A small strip
from the chromatogram was sprayed with aniline hydrogen
phthalate. The D-apiose portion of the other part of the
chromatogram was eluted with water, rechromatographed in
solvent system D for 23 hours and scanned on the strip
counter. Again, a small strip of the chromatogram was
sprayed with aniline hydrogen phthalate. The D—apiose
portion of the other part of the chromatogram was eluted
Awith water, rechromatographed in solvent system C for

12 hours and scanned on the strip counter. The whole
chromatogram was then sprayed. On each of the three
chromatograms, carrier D—apiose had the same Rf as the
radioactive peak on the scans, indicating that the radio-

active compound was indeed D-apiose-luC.

83

In the second experiment, Peak III (72,180 dpm) was
combined with 0.5 umoles of carrier D-apiose and hydrolyzed
with 0.1 N H280“ at 1000 for 90 minutes. A control
hydrolysis was done under the same conditions using
D-apiose-luC (5,140 dpm). Both hydrolysates were worked
up as before and chromatographed on paper in solvent
system C for 12 hours. The D—apiose-luC was eluted from
the chromatograms with water and the recovery of D-apiose~
lMC determined for each hydrolysis. The recoveries were
71% from the hydrolysis of Peak III and 66% from the con-
trol hydrolysis. Therefore, essentially all of the
radioactivity in Peak III must be in D-apiose-luC, since
(1), no other radioactive products were detected and (2),
the recovery of D-apiose-luC is the same from the Peak
III hydrolysis and the control hydrolysis.

An aliquot of the D-apiose—luC from Peak III
(49,175 dpm) was oxidized by Dr. P. Kindel“ with sodium
meta-periodate as described earlier. The results obtained
were similar to those shown in Table 6. Those results,
along with the others in this section identify D-apiose-luC
as the only radioactive hydrolysis product of Peak III.

e. The formation and characterization of D-
apiose-luCphosphate from Peak III.--While doing chromato—
graphy experiments with Peak III it was discovered that

when Peak III was eluted from pre-coated silica gel plates

 

“Ibid., p. 64.

84

and chromatographed on paper in solvent system A, a new
compound was formed (Figure 17). The following two
experiments showed that the new compound was D-apiose-luC-P.
The new compound was isolated from a silica gel plate but:
not chromatographed on paper in solvent system A. A
reaction mixture containing the new compound (5,234 dpm),
15 ug of snake venom phosphodiesterase, 4.5 nmoles of
MgAc2 and 5 umoles of buffer H in a total volume of 80
ul was incubated at 25° for 30 minutes. A control incu-
bation, without enzyme, was also done. The incubation
mixtures were immediately chromatographed on thin layer
plates in solvent system A for 4 hours. In addition,
the new compound (5,234 dpm) was also chromatographed,
and all 3 plates were scanned on the strip counter. This
treatment did not change the new compound.

In the second experiment, the new compound (5,510
dpm) which had been chromatographed in solvent system A,
was incubated with 18 ug of alkaline phosphatase, 1.5
umoles of MgAc2 and 5.0 umoles of buffer H in a total
volume of 100 pl at 25° for 30 minutes. A control incu-
bation was done with 3,320 dpm of unknown, but without
enzyme. Each was immediately chromatographed on paper
in solvent system C for 12 hours and scanned on the strip
counter. Treatment with alkaline phosphatase converted
the new compound to a second compound. When that second

compound was eluted with water and rechromatographed on

 

 

85

.2 Eopwmm pCo>Hom Co ooQoHo>oo mos sapwopoEOCCo oCB .mmoo me Com opoHQ How
ooHHHm o Co umoH Coon om; CoHC3 HHH xoom mo 2CQonouoEOCCoII.>H oCCme

86

HHH xwom

anonoHdaao

 

CHwHCo

87

paper in solvent system C, it had the same R as D-apiose.

f
The results of this experiment therefore showed that

Peak III was converted to D-apiose—luC-P on the silica

gel plates and that Peak III therefore contained phosphate.
When Peak III was Spotted on cellulose thin layer plates

it was not converted to D-apiose-luC-P. When Peak III

was added to aqueous suspensions of silicic acid, or
silica gel of the type used for making thin layer plates,
it was also not changed..

The following experiment was done to determine the
rate and extent of conversion of Peak III to the D-apiose-
lL‘C-P. Aliquots of Peak III (3,609 dpm) were spotted on
separate portions of a Brinkmann pre-coated silica gel
plate and left, unchromatographed, at room temperature for
various times. After 30 minutes, 24 hours, 48 hours,

72 hours and 6 days, the radioactive material was eluted
with water and chromatographed in solvent system A for
12 hours. After scanning the chromatograms, D-apiose-P
was eluted with water and an aliquot counted in the
dioxane scintillation system, while Peak III was cut out
and the paper counted in the toluene scintillation
system. The results are presented in Table 8.

In an effort to determine which carbon atom the
phosphate of D-apiose-luC-P was attached to, the follow-

ing experiment was conceived. If the phosphate was

attached to carbon atom 1, then treatment of

1

88

TABLE 8.--The formation of D-apiose-luC-P from Peak III

 

 

with time.
% of starting % of starting
Time radioactivity radioactivity in.

in Peak III D-apiose-l4C-P
30 min 91.5 not determined E
24 hours 68.9 26.7 3
48 hoursa 46.6 33.4 E
72 hours 33.6 50,4 1
6 daysb not determined 71.7 ;

 

 

aUsing 36,090 dpm of Peak III.
b
tion.

Using 15,087 dpm of a different Peak III prepara—

D—apiose-luC-P with NaBHu, followed by treatment with
alkaline phosphatase would result in the isolation of
D-apiose-luC. On the other hand, if the phosphate was
attached to carbon atoms 2, 3, or 31, such an experiment
would result in the isolation of D-apitol-luc.
D-apiose-luC-P was obtained from Peak III which
was left on silica gel plates for 6 days, as described
above, and chromatographed in solvent systems A.and E.

First, D-apiose-lu

C was isolated from 3,012 dpm of~
D-apiose-luC-P, using alkaline phosphatase. Then, the
following three apiose compounds were treated with NaBHu:

(l) D-apiose-luc isolated from D-apiose-luC-P, (2)

89

D-apiose-luC—P (5,045 dpm) and (3) D-apiose-luC (6,508
dpm;. isolated from a hydrolysis of Peak III and purified
by paper chromatography in solvent systems B, C and D).
Each was incubated at 55° for 2 hours in a conical centri-
fuge tube containing 0.25 umoles of carrier D-apiose,

10.0 umoles of NaBHu and 50 umoles of NaHCO After

3.
cooling in ice, the reaction mixtures were neutralized
with 30 nmoles of H280“, 9 volumes of absolute ethanol
added and the precipitated salts pelleted by centrifu-
gation in a clinical centrifuge. The ethanol supernatent
solutions were concentrated in_xagug, chromatographed on
paper in solvent system C for 14 hours and scanned on the
strip counter (Figure 18). The NaBHu treated D-apiose-
l“CwP was eluted with water, treated with alkaline
phosphatase as described earlier, chromatographed_on
paper in solvent system C and scanned on the strip
counter.

The control reduction done with D-apiose-luC (from
the hydrolysis of Peak III) showed that the yield of
D-apitol-luC from D-apiose-luC was about 80%. The other
control reduction of D-apiose—luC (isolated from D-
apiose—luC-P) resulted in a 60% conversion to D-apitol-
l“C. Alkaline phosphatase treatment of reduced.
D-apiose-luC-P resulted in the release of about 50%
D-apitol-luC and 50% D-apiose—luC. When the experiment

was repeated, similar results were obtained, but a yield

 

 

90

.omo Q I
H o 0 mo COHposooC

OthHopoo 2
o UoCHoo
no mo
3 UCoUCopm HOpHQo
no one .emmmz aeH
3 pCoEpooC
9

Congo 0 So
pmmm pCo>
How CH m
to

:H

O

91

 

omOHQ<IQ

O

HopHaauo

0

2:1
moz seas ooeoooev

mno I
2H omOHQCIQ

CHwHCo

" "'o- "H

 

92

of about 80% D-apitol—luc was obtained from reduced
D-apiose—luC-P. It would seem that reduced D~apiose-
ll‘lC-P should have been at least partially separated from

the non-reduced compound after chromatography in solvent

system C. Figure 18 shows that D—apiose-luC-P after

NaBHu treatment chromatographs as a single peak. This F
is probably because the compounds did not migrate far

enough to separate. On the basis of these results, it

therefore seems likely that phosphate is attached to

 

carbon atom 2, 3 or 31. However, it is also possible

‘T

that phosphate in Peak III is present as a cyclic phos-
phate, and when Peak III is converted to D-apiose-luC-P,
some D-apiose-3uC-l-P is formed. This will be further
considered in the discussion section.

f. Treatment of Peak III withgphosphodiesterase

and alkaline phosphatase.--The following experiment was.

 

done to determine if Peak III contained a phosphodiester
linkage. A different Peak III preparation was used. An
incubation mixture was set up containing 75 ul of a

dialyzed ammonium sulfate preparation from 29 g of Lemna,

0.00056 umoles of UDP-D-glucuronic acid-1Ll

C (212,300 dpm),
0.2 umoles of NAD+ and 5.0 umoles of buffer F in a total
volume of 200 01. The mixture was incubated for 30
minutes at 25°, chromatographed on paper in solvent

system A and Peak III eluted with water. An aliquot of

Peak III (5,100 dpm) was combined with 10 ug of snake

93

venom phosphodiesterase, 4.5 umoles of MgAc2 and 5.0
umoles of buffer H in a volume of 100 pl. A minus
enzyme control was also done. Both were incubated at
25° for 30 minutes, chromatographed on paper in
solvent system A for 12 hours and scanned on the strip
counter.

Another aliquot of Peak III (5,100 dpm), was mixed
with 1.8 ug of alkaline phosphatase, 1.5 umoles of

MgAc and 5.0 umoles of buffer H in a volume of 100 ul.

2
A minus enzyme control was also done. Both were incu-
bated at 25° for 30 minutes, chromatographed on paper

in solvent system A for 12 hours and scanned on the strip
counter.

The above experiments were negative in the sense
that Peak III was unchanged. Since alkaline phosphatase
is a non-specific phosphomonoesterase, and phosphodiester-
ase requires a free 31-OH group of the ribose portion
of a nucleotide base for activity, the following conclu—
sions can be drawn. First of all, the phosphate present
in Peak III was not in a phosphomonoester bond, i.e.,
as in sugar phosphates. Second, it can be concluded
therefore, that the phosphate was present inva phospho—
diester bond. This is true even though snake venom

phosphodiesterase did not act on Peak III, since Peak

III was shown not to be a sugar nucleotide.

I

94

g. Solubility prOperties of Peak III.--Three
results from the proceeding experiments indicated that
Peak III could be a phospholipid. They were: (1) Peak
III migrated in a lipid solvent system, (2) Peak III
contained phosphate, and (3) Peak III was not a sugar
nucleotide. Therefore the two experiments described F
below were done to determine the solubility character-
istics of Peak III.

The first experiment involved the partitioning of

 

Peak III from water into non—polar solvents. Separate
aliquots of Peak III (3,609 dpm) in 1.0 m1 of water
were mixed with 1.0 m1 of the following solvents: (1)
chloroform, (2) chloroform-methanol (2:1; v,v), (3)
chloroform-methanol (1:1; v,v) (4) diethyl ether, and
(5) n—butanol. After thorough mixing, aliquots of the
aqueous and non—polar phases were counted in the

dioxane scintillation system. In every case except one,
less than 5% of the starting radioactivity was found in
the non-polar phase. The exception was with chloroform—
methanol (2:1), where 20% was found.

The second experiment tested the actual solu—
bility of Peak III in some solvents. Separate aliquots
of Peak III (3,609 dpm in 25 ul) were transferred to
test tubes and evaporated to dryness at room temperature
with a stream of air. Then 1.0 ml volumes of the

following solvents were added: (1) water, (2) chloroform,

95

(3) chloroform-methanol (1:1; v,v) and (4) absolute
ethanol. After swirling the solvents, aliquots were
counted in the dioxane scintillation system. In the
case of water, all of the radioactivity went into
solution, and in the case of chloroform-methanol (1:1),
93% went into solution. With chloroform, essentially
none went into solution. Using absolute ethanol, about
70% went in solution.

The results of these tests confirmed thatheak
III is somewhat non-polar., Therefore, Peak III may be
a lipid, but not in the "classical" sense in terms of

solubility.

 

CHAPTER IV

DISCUSSION

Components of the Enzymatic Product

 

Experiments and results have been presented which
extablish that an enzyme or enzymes are present in Lemna
mingr L. that are capable of converting the D-glucuronic-
acid-1M0 moiety of UDP—D-glucuronic acid—lac to
D—apiose—luC. The D—apiose compound (Peak III) isolated
from the enzyme catalyzed reaction has not been identi—
fied. It contains at least two components, D-apiose-luC
and phosphate. These may be the only components or there
may be one or more other components.

D—apiose-luC was shown to be a component of the
product as follows. Peak III was hydrolyzed and the
radioactive product identified as D-apiose-luC by
three criteria: (a) cochromatography on paper in
three different solvent systems, (b) crystallization of
D-apiose phenyosotriazole-luc to constant specific
activity (assuming that the D-apiose-luC used in that
experiment was hydrolyzed from Peak III, even though

Peak III was not isolated), and (c) periodate oxidation

and identification of the products.

96

97

The results from the cochromatography experiments
suggested that the hydrolysis product was D-apiose-luC
and ruled out most other sugars. The triazole derivi-
tive was synthesized to conclusively identify the
hydrolysis product. Since the solution used for that
experiment contained both D-glucurono—lactone-luc and
D-apiose-luC, the results obtained were very convincing.
The results showed that 33% of the radioactive mixture
was D-apiose-luC. Later, when D-apiose-luC was obtained
which was chromatographically pure, the periodate oxida—
tion experiment was done. Of the known branched-chain
sugars, only D-hammamelose and D-apiose could yield
glycolic acid. However, Rhinehart has shown that D-
hammamelose-does not produce glycolic acid (66). These
three criteria proved conclusively that the sugar moiety
of Peak III was D-apiose-luC.

The presence of phosphate in Peak III is not as
firmly established. Its presence was based solely on
the formation of D-apiose-luC-P on silica gel plates.
Further studies have demonstrated that Peak III is
converted to D-apiose-luC-P in the presence of 50%
acetic acid or 0.01 N H01.5 For example, hydrolysis of
Peak III in 0.01 N HCl at 1000 for 15 minutes yielded

D—apiose-luC-P (43%) and D-apiose-luC (11%). Such an

 

 

5P. Kindel, personal communication.

98

experiment confirmed the presence of phosphate on Peak
III. It also showed that Peak III is as labile or more
labile than sugar nucleotides. Under the same conditions
UDP-D-glucose is hydrolyzed to D-glucose and UMP. The
D-apiose-luC-P formed by hydrolysis of Peak III in 0.01

14C_P

N HCl had the same characteristics as the D-apiose-
obtained from silica gel plates. The only criterion for
the presence of phosphate in D—apiose-luC-P was the fact
that the compound was converted to D-apiose-luC by

alkaline phosphatase. Although, in addition, the chromato-

 

graphic pr0perties of D—apiose—luC-P are very similar to
those of D-glucose-l—P.
Structure and Mechanism of Formation
of Peak III

There are two general structures possible for
Peak III; one is D-apiose-cyclic phosphodiester and the
second is R-O-POBH—D-apiose. No evidence has been pre-
sented which would indicate whether Peak III contains
one or two phosphate groups.

It seems likely that the first structure (cyclic
phosphate) could be an artifact formed during incubation
experiments. D—apiose-cyclic phosphodiester, or some
other artifactual structure, could have been formed during
the termination_of an enzyme incubation or during the
isolation of the products. Two experiments were done

which tend to weaken that argument. In one experiment,

99

an incubationmixture was spotted directly on paper with-
out heating; the same results were found after chromato-
graphy as in the standard assay. In the second experiment,
9 volumes of absolute ethanol was added and the ethanol
soluble fraction chromatographed. Again, Peak III was
observed as the only enzyme product containing D-apiose—
1“c. The possibility still remains that Peak III was
formed on the paper after the reaction mixture was,
spotted. Finally, a phosphatase, and possibly some other
enzyme may be present in the protein preparations which
splits the P-P bond of UDP-D-apiose with subsequent forma—
tion of D-apiose cyclic phosphodiester. That possibility
seems doubtful, since the enzyme would have to be carried
through several purification steps.

Some evidence was found which is consistant with
the second structure, R-O-PO3H—D-apiose. The behavior
of Peak III on a silica gel plate using solvent system K
was of interest. In that solvent system lipid compounds
migrate readily, but non—lipid compounds, such as sugar
phosphates and sugar nucleotides, do not migrate. The
fact that Peak III does migrate, indicates that Peak III
may also contain a lipid component. Experiments done to
test the solubility properties opreak III, suggested
that Peak III was not a "classical" lipid. Peak III
did, however, show some solubility in non-polar solvents.

One possible structure for Peak III could be a structure

 

100

very similar to phosphatidyl inositol, where the nature
of the diglyceride is such that the R—O-P bond is more
labile than the P-O-D-apiose bond.

Experiments were also done to determine the manner
of attachment of the phosphate in Peak III. The results
from the electrophoresis experiments indicate that Peak
III is anionic at pH 3.7 or 5.4. The relative mobility
of Peak III was very close to that of UDP-D-glucose and
UDP-D—glucuronic acid. In addition, sugar phosphates
are reported to have about the same mobilities as UDP-D-
glucose (19). Therefore, it appears that the phosphate
present in Peak III, rather than some other component,
is responsible for its anionic property.

This leads to the question of the physiological
function of Peak III. If Peak III were indeed a phos-
pholipid, one function comes immediately to mind, namely
as an intermediate in cell wall biosynthesis. In the
literature review that kind of an intermediate was
mentioned. This possible function seems very attractive,
since D-apiose has been shown to be a constituent of cell
wall polysaccharides. However, this suggestion is purely
speculation.

The research efforts described in this thesis do
not prove, nor do they disprove, that the mechanism for
D—apiose biosynthesis presented in the literature review

is correct (see Figure l). The fact that UDP-D-glucuronic

 

101

acid was a substrate for the enzyme does strengthen the
postulated mechanism. No other nucleotide bases attached
to the D—glucuronic acid were tried, and therefore it

is not possible to state whether or not UDP-D-glucuronic
acid is the preferred substrate.

It is really not possible to state a mechanism for
the biosynthesis of Peak III, since its structure is
unknown; however, some generalizations may be made. For
the possibility mentioned above, namely a phospholipid,
the following events could have taken place. First, UDP-
D-glucuronic acid was converted to UDP—D-apiose. The
UDP-D-apiose did not accumulate but was immediately
utilized in a reaction in which the D-apiose of UDP-D—
apiose was transferred by a transferase to an endogenous
diglyceride acceptor to form a phosphatidyl D-apiose.
This diglyceride may or may not have contained a phos-
phate group. This hypothesis requires that the rate of
the transfer reaction would have to be faster than the
rate of D—apiose formation so that the UDP-D-apiose pool
would remain small. In addition, the endogenous
acceptor molecule would have to be bound to one or both
of the two postualted enzymes and carried along through
the purification procedure. This hypothesis requires
further modifications, since the results from the NaBHu
experiments with D-apiose-luC-P indicate that it is very

unlikely that the phosphate was attached to carbon atom l.

 

102

This can be explained if one assumed that carbon atom 1

of D-apiose is attached to the PO“ group in R-O-PO H-D-

3
apiose, and that on a silica gel plate or in the presence
of acid, "R" is produced with the formation of an
intermediate D-apiose 1,2-cyclic phosphodiester, which

then breaks down to form D-apiose-2-P.

Future of the Problem

In order to identify Peak III it will be necessary
to prepare umole quantities of the compound. Those experi-
ments have already been started. This in turn will
require either growing larger amounts of Lemna or isolat-
ing the enzyme from a different source, such as 23352:
selinum. Unfortunately, Lgmga collected from local rivers
in kg quantities have been shown to be too low in enzyme
activity to be of value as a source.

In order to study the details of the mechanism of
D-apiose biosynthesis and Peak III synthesis, it will be
necessary to have enzyme of higher purity than has been
obtained so far. That will require larger starting
quantities of the enzyme for its purification. These
anticipated studies should provide many fascinating

research findings in the future.

 

10.

ll.

12.

13.

14.

15.
16.

BIBLIOGRAPHY

Hudson, C. S. Advances in Carbohydrate Chem.
1. 57 (1949).

Hemming, R. and Ollis, W. D. Chem. Ind. 85 (1953). :-

Bell, D. J., Isherwood, F. A. and Hardwick, N. E.
J. Chem. Soc. 3702 (1954) (with a note by R. S. Cahn).

 

Jacobson, L. Plant Physiol. 88, 411 (1951).

Schmidt, G. and Thannhauser, S. J. J. Biol. Chem.
149. 369 (1943).

Duff, R. B. and Knight, A. H., Biochem. J. 88, 33P
(1953).

Bacon, J. S. D. Biochem. J. 82, 103P (1963).

 

Williams, D. T. and Jones, J. K. N. Can. J. Chem.
8;, 69 (1964).

Grisebach, H. and D6bereiner, U. Biochem. Biophys.
Res. Commun. $1, 737 (1964).

Beck, E. and Kandler, 0. Z. Naturforsch. 20b,
62 (1965).

Duff, R. B. Biochem. J. 28, 768 (1965).

Mendicino, J. and Picken, J. M. J. Biol. Chem. 240,
2797 (1965)-

Castanera, E. G. and Hassid, W. Z. Arch. Biochem.
Biophys. 110, 462 (1965).

Hulyalkar, R. K., Jones, J. K. N. and Perry, M. B.
Can. J.,Chem. 83, 2085 (1965).

Ankel, H. and Feingold, D. S. Biochem. 8, 2468 (1965).

Beclgé E. and Kandler, 0. Z. Pflanzenphysiol. 82, 71
19 ).

103

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.
27.
28.
29.
30.
31.

32.
33.
34.

35.

104

Grisebach, H. and D6bereiner, U. Z. Naturforsch.
21b, 429 (1966).

Gupta, S. R. and Seshadri, T. R., Proc. Indian
Acad. Sci. 35A, 242 (1952).

 

Feingold, D. 8., Neufeld, E. F. and Hassid, W. Z.
"Enzymes of Carbohydrate Synthesis" in Methods in
Plant Analysis, Vol. VII, p. 474 (1964).

Grisebach, H. and Sandermann, H., Jr. Biochim. Z.
112. 322 (1966).

Sandermann, H., Jr., Tisue, G. T. and Grisebach, H.
Biochim. BiOphys. Acta. 165, 550 (1968).

Picken, J. M. and Mendicino, J. J. Biol. Chem. 242,
1629 (1967).

Roberts, R. M., Shah, R. H. and Loewus, F. Plant
Physiol. 8g, 659 (1967).

Rump, Buchner's Repertorium fUr die Pharmacie, [2], 8,
l (1836). (Army Medical Library, Washington, D. C.).
Quoted in reference 1.

Grisebach, H. Biosynthetic Patterns in Microorganisms
and Higher Plants. New York: John Wiley and Sons,
Inc., 1967, 110 p.

Vongerichten. E. Ann. Chem. 318, 121 (1901).

 

Vongerichten, E. Ann. Chem. 321, 71 (1902).

 

Vongerichten, E. and Mfiller, F. Ber. 82, 235 (1906).

Schmidt, 0. T. Ann. Chem. 88;, 115 (1930).

 

Beck, E. Z. Pflanzenphysiol. 21, 444 (1967).

 

Sandermann, H., Jr. and Grisebach, H. Biochim.
Biophys. Acta, 156, 435 (1968).

 

Grisebach, H. Helv. Chim. Acta. 8;, 928 (1968).
Barber, G. A. Biochem. l, 463 (1962).

Neufeld, E. F., Feingold, D. S., and Hassid, W. Z.
Arch. Biochem. Biophys. 8;, 96 (1959).

 

Farooq, M. 0., Gupta, S. R., Kiamuddin,.M., Rahman, W.
and Seshadri, T. R. J. Sci. Ind. Res. 12B, 400 (1953).

36.

37.
38.

39.
40.

41.

42.

43.

44.

45.

46.
47.
48.

49.

50.

51.

52.

53.

105

Blumson, N. L. and Baddiley, J. Biochem. J. 81,
114 (1961).

Loewus, F. Arch. Biochem. Biophys. 102, 96 (1963).

Anderson, L. and Wolter, K. E. Ann. Rev. Plant
Physiol. 21) 209 (1966).

Colvin, J. R. Nature. 882, 1135 (1959).

Kahn, A. W. and Colvin, J. R. J. Polymer Sci.
5;, 1 (1961). r

Colvin, J. R. In The Formation of Wood in Forest
Trees. Edited by M. H. Zimmermann, Academic Press,
New York (1964).

Hillman, W. S. Botan. Rev. 21, 221 (1961).

 

Gray's Manual of Botany, 8th ed. Edited by M. L._ .
Fernald. American Book Co., New York, 1632 p.
(1950).

Norris, L., Norris, R. E., and Calvin, M. J. Exp.
Bot. 88, 64 (1954).

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

Calkins, V. P. Ind. Eng, Chem. 82, 762 (1943).

 

Partridge, S. M. Nature. 288, 443 (1949).

Ovodova, R. G., Vaskovsky, V. E. and Ovodov, Yu. S.
Carbohydrate Res. 6, 328 (1968).

 

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

Verachtert, H., Bass, S. T., Wilder, J. and Hansen,
R. G. Anal. Biochem. 88, 497 (1965).

 

Lowry, O. H., Rosebrough, N. J., Farr, A. L. and
Randall, R. J. J. Biol. Chem. 193, 265 (1951).

Peterson, E. A. and Sober, H. A. in Methods in
Enzymology. Vol. V, p. 25. Edited by Colowick, S. P.
and Kaplan, N. 0. Academic Press, New York (1962).

 

Cassel, J. F. Christianity Today. 2, 1161 (1965).

54.

55.

56.

57.

58.

59.
60.
61.

62.

63.
64.

65.

66.

106

Malhotra, A., Murti, V. V. S. and Seshadri, T. R.
Tetrahedron Letters, No. 36, 3191 (1965).

 

Paris, R. R., Paris, M. and Fries, J. Ann. Pharm.
Franc. 28, 245 (1966).

 

Werbin, W., Chaikoff, I. L. and Imada, M. R. Proc.
Soc. Exp. Biol. Med. 102, 8 (1959).

 

Hattori, S. and Imaseki, H. J. Am. Chem. Soc. 82,
4424 (1959).

 

Von Planta, A. and Wallace, W. Ann. Chem. _8,
262 (1850).

 

Linderborn, A. Chem. Centr. 88, Band 1, 928 (1897).
Horecker, B. 1. Ann, Rev. Microbiol. 22, 253 (1966).

Leloir, L. F. and Goldemberg, S. H. J. Biol. Chem.

Recondo, E. and Leloir, L. F. Biochem. Biophys. Res.
Commun. 8, 85 (1961).

 

Glaser, L. J. Biol. Chem. 222, 627 (1958).

 

Elbein, A. D., Barber, G. A., and Hassid, W. Z.
J. Am. Chem. Soc. 86, 309 (1964).

 

Lin, Tsau—Yen and Hassid, W. Z. J. Biol. Chem.
241, 5284 (1966).

 

Rhinehart, B. L. Masters Thesis, Michigan State
University (1967).

MIC‘IWITIWWIWIii4141715!in73411111ms
3 1293 03062 0151