rt.” ‘m' A. “ham“ h” __.
'x'f

“J LIBRARY

Michigan S tau:
University

tween

 
  
   

This is to certify that the

thesis entitled
Enzyme-Mediated Synthesis of the

Antigenic Portion of the Blood Group 0 Substance with
Specific Carbon-13 Enrichment

presented by

Paul R. Rosevear

has been accepted towards fulfillment
of the requirements for

 

Ph- D- degree in _B.iochemj_s.try

 

Major professor

[Mme May 12, 1980

 

0-7639

  
 

   
 

.D ‘

l‘ b
‘1.

“L\

’“‘€:‘\ k
. V .4 (in
:‘glm'r V »

   

1"

   

 
 

 

w:

25¢ per day per item

RETURNING LIBRARY MATERIALS:
M

Place in book return to remove
charge from circulation records

 

ENZYME-MEDIATED SYNTHESIS OF THE
ANTIGENIC PORTION OF THE BLOOD GROUP 0 SUBSTANCE WITH SPECIFIC
CARBON-13 ENRICHMENT

By

Paul Richard Rosevear

A DISSERTATION

Submitted to

Michigan State University

in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Biochemistry

kin m 95

ABSTRACT

ENZYME-MEDIATED SYNTHESIS OF THE ANTIGENIC PORTION OF THE BLOOD GROUP 0
SUBSTANCE WITH SPECIFIC CARBON-13 ENRICHMENT.

By

Paul Richard Rosevear

The use of partially purified glycosyltranferases and chemically
synthesized sugar nucleotides provides an alternative method for
synthesis of complex carbohydrates. Enzymatic synthesis of the
glycosidic linkage has been shown to proceed with anomeric specificity,
and high yield. This approach was used in the synthesis of mmole
quantities of the anitgenic portion of the blood group 0 substance with
specific carbon-13 enrichment.

Partially purified bovine Nfacetylglucosaminide B (1-4)
galactosyltransferase and porcine 8 galactoside a(1-2)
fucosyltransferase were used with chemically synthesized sugar
nucleotides in the synthesis of the antigenic trisaccharide Fuc
a(1-2)GalB(1-4)GlcNAc-B-hexanolamine.

The B-galactoside a(1-2) fucosyltransferase was partially purified
from porcine submaxillary glands by a procedure similar to T.A. Beyer
and R.L. Hill (unpublished results). The purification entailed;
solubilization with Triton X-100, chromatography on the affinity
adsorbant GDP-Sepharose, Sephadex GSO chromatography and concentration

on GDP-Sepharose. The enzyme was purified 11,000 fold and was used in

the synthesis of mmole quantities of several fucosylated di and
trisaccharides.

The enzyme-mediated synthesis of fucosylated oligosaccharides was
performed by adding equimolar ratios of GDP-fucose and acceptor to the
enzyme in sodium cacodylate buffer, pH 6.0 in the presence of manganese
chloride at 33°. After 12 hours of incubation, yields of the fucosyla-
ted compounds were typically 80%. The remaining GDP-fucose was
hydrolyzed to fucose during the incubation. Purification of the product
di or trisaccharide was rapidly achieved by deproteinization, ion
exchange chromatography and gel filtration. Enzymatic synthesis with
Specifically enriched UDP-[1-13CJ-D-Galactose, GDP-[1-13CJ-L-fu-
cose or 13C-enriched acceptors allowed preparation of singly or
doubly enriched di and trisaccharides.

Specific 13C-enrichment and comparison with 13C-enriched
model compounds allowed unambiguous resonance assignments of the carbons
in the trisaccharide using IJC—C’ ZJC-C and BJC-C coupling
constants. Derivatization of a carbon with another aldopyranosyl ring
resulted in a 5-8 ppm downfield shift of that resonance. Carbons
contiguous to the derivatized carbon also underwent small chemical shift
changes. The B-galactoside a(1-2) fucosyltransferase was unambiguously
shown to transfer fucose from GDP-L-fucose to C2 galactose when
[1-13C]Gals(1-4)GlcNAc-B-hexanolamine was used as the acceptor. The
CZ galactose was easily identified by a 'JCI'CZ' coupling of 46 Hz.

Inter-residue 3JC-C’ 3JC-H and ZJC-C coupling constants in
the singly and doubly enriched fucosylated oligosaccharides allowed
estimation of the most abundant conformer for the w and 0 torsion angles

in the Gals(1-4)GlcNAc and FuCa(1-2)Gal glycosidic linkages.

DEDICATION

The author would like to dedicate this work to his wife
Cathryn Carson Rosevear
for the joy that she has given me, her love, and her continual support
throughout the days of research and preparation of this work, and to his
mother
Dorothy B. Rosevear

for her love, encouragment and help throughout the years.

ii

ACKNOWLEDGMENTS

The author would like to thank his thesis advisor, Dr. Robert
Barker, for the Opportunity to work in his laboratory and financial
support. Special thanks are necessary for Dr. Shelagh Ferguson-Miller,
Dr. John Wang, and Dr. Hernan Nunez for their friendship, advice,
valuable discussions during the course of this research, and
colaborations.

The author also wishes to thank the Department of Biochemistry,
Michigan State University, for it's generous support during the final
stages of this work.

Special thanks to the Purdue University Biochemical Magnetic
Resonance Laboratory, Dr. A.S. Mildvan for the use of the Bruker-ZSO MHz
NMR at the Institute for Cancer Research and the Stable Isotopes

Resource of the Los Alamos National Laboratory.

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES
ABBREVIATIONS
INTRODUCTION
REVIEW OF THE LITERATURE
Current Concepts in Glycoprotein Biosynthesis
The Biochemical Structure and Synthesis of the Blood Group
Substances
Review of the Purification and Properties of Lactose
Synthetase
Review of the Literature on the Fucosyltransfease
Review of 13C-NMR Spectrosc0py and 13C-13C Spin-
Spin Coupling Constants
METHODS
Materials
Analytical Procedures
Thin Layer Chromatography
High Voltage Electrophoresis
High Pressure Liquid Chromatography
Gas Chromatography
Radioactivity Measurements
Carbon-13 NMR Spectroscopy

Quantitation of Hexanolamine Compounds

iv

Page

vii
viii

xi

10

14

17
22

28
28
28
28
29
29
29
29
3O
30

Protein Determination

Synthetic Procedures

GMP-Morpholidate and UMP-Morpholidate

6-Amino-1-Hexanol PhOSphate

N-Trifluoroacetyl 6-Amino-1-Hexanol Phosphate

P'-(6-Amino-1-Hexyl)-P-(5'-Guanosine)-Pyrophosphate

P'-(6-Amino-1-Hexyl)-P-(5'-Uridine)-Pyrophosphate

GDP-Hexanolamine Sepharose

UDP-Hexanolamine Sepharose

UDP-s-D-Galactose

GDP-a-L-Fucose

2-Acetamido-3,4,6-tri-ancetyl-2-deoxy-Glucopyranosyl
Chloride

NfTriflouroacetyl-6-Amino-1-Hexyl-2-Acetamido-2-Deoxy-B-
D-Glucopyranoside

6-Amino-1-Hexyl-2-Acetamido-2-Deoxy-e-D-Glucopyranoside

Acetobromogalactose

6-Amino-1-Hexyl-a,3-D-Galactopyranoside

Ethyl-e-D-Galactopyranoside

5-Deoxy-L-Lyxose

[1-13c, 6-3H]-L-Fucose and [1-13c, 6-3HJ-6-Deoxy-
L-Talose

[1-13CJ-D-Galactose

[2-13CJ-D-Galactose

Purification of Bovine UDP-GalactosezGlucose-B-(1-4)

Galactosyltransferase

page
30
3O
30
31
31
31
32
33
33
33
35
35

36

36
36
37
38
38
4o

41
41
41

page

Preparation of Galactosylated Disaccharides 42

Purification of Galactosylated Disaccharides 42

Assay of Porcine B-Galactoside a(1-2)Fucosyltranferase 42

Purification of a Porcine B-Galactoside a(1-2) 44
Fucosyltransferase

Preparation of Fucosylated Oligosaccharides 47

Purification of Enzymatically Synthesized Oligosaccharides 48
RESULTS 49

Synthesis of Galactosylated Disaccharides

Enzymatic Assay of the B-Galactoside a(1-2) 49
Fucosyltransferase

Purification of the Porcine B-Galactoside a(1-2) 50
Fucosyltransferase

Synthesis and Purification of Fucosylated Oligosaccharides 56
Carbon-13 NMR Analysis of Fucosylated Oligosaccharides 71

Conformational Analysis of the Fucosylated Oligosaccharides 88

DISCUSSION 102
Preparation of B-Galactoside a(1-2)Fucosyltransferase 102
Preparation of Fucosylated Oligosaccharides 105
Carbon-13 NMR of Fucosylated Oligosaccharides 106
Conformation of the Fucosylated Oligosaccharides 108

BIBLIOGRAPHY 128

APPENDIX 135

vi

LIST OF TABLES
Page
1. Purification of the Porcine Submaxillary B-Galactose a(I-2) 52

Fucosyltransferase

2. Carbon-13 Chemical Shift Assignments 74

vii

1.

3.

9.

10.

LIST OF FIGURES

Structures Of the Q-Glycosidically Linked Oligosaccharides.
The Scheme Of Synthesis Of Type Two Blood Group Antigens,
Their Participating Genes and Their Primary Products,
Glycosyltransferases.

Purification 0f the Porcine Submaxillary Gland B-Galactoside
a(1-2) Fucosyltransferase By NaCl Elution From GDP-Sepharose.
Purification and Desalting Of The B-Galactoside

a(1-2)Fucosyltransferase.

Concentration Of The Porcine B-Galactoside a(1-2)Fucosyltrans-

ferase By GDP-Sepharose II Chromatography.

The Kinetics Of Fucosylation Of GalB(1-4)GlcNAc-3-Hexanolamine

Nith Porcine s-Galatoside a(1-2)Fucosyltransferase.

Time Course Of Fucosylation of [1- 13C]Gale(1-4)GlcNAc
Followed By High Pressure Liquid Chromatography.

Kinetics Of Hydrolysis 0f GDP-Fucose By The s-Galactoside
a(1-2) Fucosyltransferase In The Absence Of An Acceptor
Substrate.

Separation of FUCa(1-2)Gal-3-Hexanolamine and Gal-B-Hexanola-
mine By Gel Filtration. .
The 15.08 MHz Proton Decoupled, 13C-NMR Of; A. The Galac
tosyltransferase Acceptor Substrate GlcNAc-B-Hexanolamine, B.

The Galactosylated Product, Gal3(1-4)GlcNAc-3-Hexanolamine.

viii

Page

9
12

58

63

65

67

7O

73

11.

12.

13.

14.

15.

16.

17.

18.

19.

The 15.08 MHz, Proton Decoupled, 13C-NMR 0f

Gal-B-Hexanolamine.

The 90 MHz, Proton Decoupled, l3C-NMR Of FUCo(1-2)

GalB(1-4)GlcNAc-3-Hexanolamine.

The 90 MHz, Proton Decoupled, Natural Abundance

13C-NMR Spectrum of the Antigenic Trisaccharide,

FuCo(1-2)Gals(1-4)GlcNAc-B-Hexanolamine.

An Expanded 90 MHz, Proton Decoupled, 133-NMR Spectrum of

the Anomeric Region Of FUCa(1-2)GaIB(1-4)GTCNAC-B-

Hexanolamine.

The 45 MHz, Proton Decoupled, 13C-NMR Of The Trisaccharide

FUCa(1-2)

[1-13C]Gal3(1-4)GlcNAc-3-Hexanolamine.

An Expanded Region From C2 to C6 Of The 90 MHz, Proton

Decoupled, 13C-NMR or FUCa(1-2) [1-13C]Gale(1-4)

GlcNAc-B-Hexanolamine.

The Conformation About The Glycosidic Bond In FUCo(1-2)Gal

Can Be Defined By Two Torsion (Dihedral Angles, w
(Cl-O-C2'-Cl') and 0 (Hl-Cl-O-CZ').

The 90 MHz, Proton Decoupled, 13C-NMR Spectrum Of

The Anomeric Region Of The Trisaccharide FUCo(1-2)

[1-13C] Gal3(1-4)GlcNAc-B-Hexanolamine.

Carbon-13 NMR Linewidth Measurements Of The 13C-Enriched

Resonances in FuCo(1-2)[1-13C]Gals(1-4)GlcNAc-s-Hexanolamine

and [1-I3CJFUCo(1-2)[1-13C]Gal3(1-4)GlcNAc-e-Hexanolamine.

ix

page
77

80

82

85

87

9O

93

95

98

20.

21.

22.

23.

24.

page
The 180 MHz 'H-NMR Showing the Hl Fuc Resonance In; A. 101
FUCo(1-2)Gal-B-Ethyl, and B. Fuc a(1-2)[2-13C]Gal-B-Ethyl.
Newman Projections Describing the Two Torsional (Dihedral) 111
Angles, uyand 0, About the Glycosidic Bond In Fuc a(1-2)Gal.
The Proposed ipTorsion Angle For The Glycosidic Linkage In 115
GalB(1-4)GlcNAc-3-Hexanolamine.
Possible w Torsional Angles About The Fuca(1-2)Gal Glycosidic 121
Linkage.
Torsional Angles About the FUCo(1-2)Gal Glycosidic Linkage 124
Expected To Give Approximately A 1.5 Hz Vicinal Coupling
Between ClGal and ClFuc.

ABBREVEATIONS

2,5-diphenyloxazole, PPO; 1,4 bis [2(-4-methyl-5-phenyloxazolyl)]-
benzene, POPOP; parts per million, ppm; guanosine monophoshate, GMP;
guanosine diphosphate, GDP; guanosine triphosphate, GTP; uridine mono-
phosphate, UMP; uridine diphosphate, UDP; uridine triphosphate, UTP;
hydrochloric acid, HCl; lithium chloride, LiCl; sodium hydroxide, NaOH;
Nfacetylglucosamine, GlcNAc; L-fucose, Fuc; galactose, Gal; glucose ,
Glc; (6-amino-1-hexyl)-2-acetamido-2-deoxy-3-D-glucopyranoside,
GlcNAc:B-hexanolamine; magnesium perchlorate, Mg(ClO4)2; mercuric
cyanide, Hg(CN)2; tetramethylsilane, TMS; lactose, Gal8(1-4)Glc;
.Neacetyllactosamine, GalB(1-4)GlcNAc; (6-amino-1-hexyl)-4gg
(s-D-galactopyranosyl)-2-acetamido-2-deoxy-D-glucose; GalB(1-4)GlcNAc-B—
hexanolamine; (6-amino-1-hexyl)-3-D-galactopyranoside, Gal-B-hexanola-
mine; ethyl-B-D-galactopyranoside, ethyl-Gal; methyl-a-L-fucopyranoside,
methyl-Fuc; (6-amino-1-hexyl)-4597(fucopyranosyla(1-2) B-D-galactopyran-
osyl)-2-acetamido-2-deoxy-8-D-glucopyranoside, Fuc a(1-2)Gale(1-4)
GlcNAc-B-hexanolamine; (6-amino-1-hexyl)-259io-L-fucopyranosyl)-galacto-
pyranoside, Fuc a(1-2)Gal-B-hexanolamine; ZgQTo-L-fucopyranosyl)-D-gal-
actOpyranose, Fuc o(1-2)Gal; 13C nuclear magnetic resonance spectro-
scopy, 13C-nmr; 'H nuclear magnetic resonance spectroscopy, 'H-nmr;
Sulfopropyl-Sephadex, SP-Sephadex, high pressure liquid chromatography,
HPLC.

xi

INTRODUCTION

Carbohydrates present on glycoproteins and glycolipids have
recently become an active area of interest due to their probable role in
cell-cell adhesion, density dependent growth inhibition, molecular
recognition, hormonal control, blood clotting, immunological protection,
structural support, and surface protection (1,2). Glycoproteins are
ubiquitously distributed in nature, occuring in vertebrates,
invertebrates, plants, bacteria and viruses.

One important class of glyc0proteins and glycolipids are the ABO
blood group substances. Aside from the recognized importance of the ABO
antigens in blood typing and transfusion, they have been implicated in
carcinoma, ulcers, pernicious anemia, hemolytic disease of the newborn,
fetal loss and infertility (1,3).

A versatile chemical synthesis of these kinds of complex,
biologically active, oligosaccharides has not been accomplished.

Lemieux and coworkers, have succeeded in the chemical synthesis of the
Lewis and B blood-group antigenic determinants (4,5). However, their
method was extremely laborious, involving complex blocking groups,
difficult o-glycosidations, purifications, and limited in the ability to
vary the aglycon at the reducing end of the carbohydrate moiety.

A more versatile approach to the synthesis of biologically active
oligosaccharides is needed due to the difficulty of isolation from
natural sources, marked heterogenity when isolated, and limited
availability. The necessity for a general method of oligosaccharide
synthesis becomes increasingly important with the advent of 13C nmr

spectroscopy in the elucidation of structures and conformations of

1

oligosaccharides, antigenic determinants, glycoside-protein complexes,
and micelles containing gangliosides (6,7,8,9,10,11,12). Carbon-13 nmr
spectroscopy, with the use of apprOpriately labeled 13C compounds,
provides an excellent nondestructive tool for probing the conformations
and biological interactions of this important class of compounds.

One approach to this problem is the use of a combination of
chemical and enzymatic methods for the synthesis of complex
oligosaccharides. This approach involves the chemical synthesis of
enzyme substrates, with appropriate labels when necessary, and the use
of these substrates with specific enzymes for the synthesis of
oligosaccharide linkages. Enzymatic synthesis of the glycosidic linkage
proceeds with specificity, high yield, and relative ease as compared to
the chemical methods for synthesis of these linkages (4,5,13). One
example of this type of synthesis is the use of the
Nfacetylglucosaminide 3(1-4) galactosyltransferase in the synthesis of
Nyacetyllactosamine, glycosidic derivatives of Nyacetyllactosamine and
specifically labeled [13CJ—N7acetyllactosamine (13,14,15). This
type of enzymatic approach to synthesis of carbohydrates has been used
as early as 1955 by Bean and Hassid for the synthesis of several
disaccharides (16). However, this approach has not been used for the
synthesis of branched chain complex oligosaccharides found on cell
surfaces that are responsible for antigenic activity. The combination
of chemical and enzymatic syntheses requires that quantities of
substrate and at least partially pure enzyme preparations be available.

The ABO blood group substances provide an excellent class of
complex oligosaccharides to attempt a combination of chemical and

enzymatic syntheses. Each carbohydrate is known to be transferred from

a sugar nucleotide donor to the nascent chain by action of a specific
glycosyltransferase (17).

This thesis describes the solubilization and partial purification
of a e-galactoside a(1-2) fucosyltransferase from porcine submaxillary
glands. This enzyme is responsible for the transfer of L-fucose from
GDP-fucose to oligosaccharides containing a terminal B-galactoside
residue. Partial purification was achieved using the affinity adsorbant
GDP-hexanolamine Sepharose. The use of this enzyme in conjunction with
a bovine Bfifleacetylglucosaminide 3(1-4) galactosyltransferase and
chemically synthesized sugar nucleotides, GDP-fucose and UDP-galactose,
enabled the synthesis of quantities of the terminal trisaccharide
portion of the blood group 0 antigenic substance sufficient for
13C-nmr evaluation. Enzymatic synthesis of fucosylated
disaccharides and trisaccharides containing specific 13C enrichment
was accomplished using chemically synthesized UDP-[1-13CJ-galactose
and GDP-[1-13CJ-fucose. Specific enrichment with 13C to the 90%
level allowed Vicinal carbon-carbon and carbon-hydrogen coupling
constants to be evaluated in terms of the conformation of the

oligosaccharride.

REVIEW OF THE LITERATURE

I. Current Concepts in Glycoprotein Biosynthesis

The discovery of glycogen synthetase by Leloir and Cardini (18)
established the role of sugar nucleotides in complex carbohydrate
synthesis as shown below;

UDP-Glucose + Glycogen": Glycogen n +1 + UDP

Recently attention has turned to the carbohydrate moiety which is
covalently attached to either protein or lipid in glycoproteins and
glycolipids, respectively. This interest has largely been due to the
implication of carbohydrates in many important biological processes
(1,2). Complex carbohydrates, in this context, are the oligosaccharide
structure found in glycoproteins and glycolipids.

The general process for synthesis of complex carbohydrates consists
of a transglycosylation reaction as shown below:

XDP-O-Rl + Rz-OH -‘_—_§ Rz-O-Rl = XDP
where XDP-O-Rl is any number of sugar nucleotides and Rz-OH is the
alcohol acceptor, usually a monosaccharide or oligosaccharide. The
common sugar nucleotides used in the synthesis of complex carbohydrates
are UDP-o-D-glucose, UDP-o-D-galactose, UDP-a-Deflyacetylglucosamine,
UDP-o-Nfacetylgalactosamine, GDP-o-D-mannose, GDP-B-L-fucose and
CMP-a-sialic acid. The common anomeric linkages found for each of these
sugars incorporated into complex carbohydrates are; a and 3 for

galactose, Nyacetylgalactosamine and mannose; B for N—acetylglucosamine,

and a for fucose and sialic acid (1,2,19).

Two types of carbohydrate-protein linkages exist. Carbohydrates
linked through asparagine are classified as N-glycosidically linked and
those linked through serine or threonine as Ofglycosidically linked
carbohydrates. The Neglycosidically linked carbohydrates are found in
plasma proteins, hormones, immunoglobulins and enzymes including bovine
ribonuclease and al-acid glycoprotein (2). The nglycosidically linked
glycoproteins include those from mucous secretions, immunoglobulins,
fetuin, plasma membranes and the Antarctic fish freezing-point
depression glycoprotein (2).

The biosynthesis of N-glycosidically linked carbohydrates has
recently been described by Schachter and Roseman (19). These structures
share a common core containing mannose and Nfacetylglucosamine. The
assembly of this core occurs through a polyiSOprenoid phospholipid
acceptor, dolichol phosphate (20). Initiation of the oligosaccharide
core occurs by the transfer of Nracetylglucosamine from UDP-GlcNAc to
the lipid acceptor (20). The chain is then extended by the sequential
transfer of Nyacetylglucosamine and mannose from UDP-GlcNAc and GDP-Man
to form Man8(1-4)GlcNAc3(1-4)GlcNAc-P-P-Dol. Further extension occurs
with branching by the subsequent addition of mannose and glucose to form
mannosylphosphoryl dolichol (19). The entire oligosaccharide is then
transferred from dolichol phosphate to an asparagine residue on the
nascent glyc0protein. This transfer of the oligosaccharide from
dolicohol to the protein occurs in the rough endoplasmic reticulum (19).
Further elongation occurs in the Golgi apparatus of the cell. Before
the immature glycoprotein reaches the Golgi complex the oligosaccharide
is processed to remove all glucose residues and some mannose residues.

The processed nascent glycoprotein strucure is shown below:

Man o(1-6)
Man B(1-4)GlcNAcB(1-4)GlcNAc-AsN-protein
Man a(1-3

v

Once the nascent glycoprotein reaches the Golgi area final
elongation occurs by the sequential action of specific glycosyl
transferases (19). The structures of the mature glycoproteins are
dictated by the strict substrate specificities of the
glycosyltransferases (21). The action of a number of these
glycosyltransferases are mutally exclusive (21,22). Vacuoles then bud
off from the Golgi apparatus and the mature glycoprotein is transferred
to the exterior of the cell.

Biosynthesis of Q-glycosidically linked oligosaccharides does not
occur by the same mechanism as Neglycosidically linked oligosaccharides
(19,23,24). Initiation does not involve pre-assembly of the
oligosaccharide as a lipid intermediate (24). The Ser(Thr)-GalNAc
linkage is synthesized by a UDP-GalNAc: polypeptide
Nfacetylgalactosaminyltransferase (25). Immediately after the
incorporation of GalNAc there is a branch point in the synthetic pathway
of the oligosaccharide (25). The incorporation of sialic acid a(2-6) to
the GalNAc prevents further carbohydrate incorporation and the
predominant form of ovine submaxillary mucin results. Incorporation of
galactose 8(1-3) to the GalNAc allows further incorporation of
carbohydrates to form the common porcine submaxillary mucins. The
relative proportions of these two enzymes control this important branch

point (26). Additions of carbohydrates to the Gal8(1-3)GalNAc-Ser can

then proceed to form the ABO blood group megalosaccharides as shown in

Figure (1) (27).

.mmuwcmzoommOPmme azocm voopn

om< mcp eo weapuchm mpecuaconcmu mpm_qeou on» .HHH .cowuecoaLOUCW mpecuxconceo convene mzoPPe u<z_mu
eo cowumcoacoucH .cwuze memF—wxesnzw mcwucoa eo Ecom pcmcwaoumca ms» .HH .cowpocoacoucw mpmcuxconcmu
segues» mucm>mca o<zpmw mew op Amumva ovum ow—mwm mo co_umcoacoocH .cwose Acmppwxm5nzm mcw>o

mo Ecom ucecweoumca on» .H .mmuwcmzuommomwFO nmx:_4 x——euwvwmooxponm mg» mo mmczpuscum .H meam_d

H mesmwu

AIS

03d
Am-Hve _ee Amy

AN-HVe
eomfleeev-o-o<z_eefim-fivm_eoflm-fivmo<zo_efim-fivm_eefle-evmo<zo_ofle-flvm Few ee

Am-five o<zpeo A<v

eomfieeev-o-o<z_ewAm-va_eo

eomfloeev-o-o<z_eofie-mve e_oe oe_e_m

HHH

HH

10

II. The Biochemical Structure and Synthesis of the Blood Group

Subtances

The blood group antigens are one important class of cell surface
oligosaccharides. They are gene-dependent structures expressing the
“individuality" of cell surfaces, body fluids, and secretions. Blood
group antigens are classified into the following groups depending on
their specificities: ABO, Lewis, MN, P,I,Y, and Rh antigens. There is
no O-antigen, group 0 erythrocytes contain the H antigen, but the Group
0 designation has been retained for historical reasons. Pure ABO
antigens were isolated in water—soluble form from pepsin, saliva,
gastric juice, gastric mucin, ovarian cyst fluid and urine (28).
Membrane bound ABO blood group antigens have recently been isolated from
erythrocytes using a combination of organic solvents and detergents
(29).

The scheme of synthesis of the blood group antigens, their
participating genes and their primary products, glycosyltransferases, is
schematically represented in Figure 2. Biologically, these substances
are synthesized by glycosyltransferases through the sequential addition
of monosaccharide units, derived from sugar nucleotides, to carbohydrate
acceptors. These glycosyltransferases have strict specificity toward
the structure of the sugar nucleotide and somewhat less specificity
toward the structure of the precursor carbohydrate chain (acceptor).

The serologic specificity of these oligosaccharides is determined
by the structure and linkage of the monosaccharides at the nonreducing

ends of the carbohydrate chains. Two types of nonreducing ends, Type 1

11

.mmmmcmmmcocupxmoux_m .mpozuocq apnerca gems“

new mmcmm mcwuoapu_ucma e_m;u .mcmm_uco qzocm noopn oz» waxy mo mwmmgucxw yo osmgom as» .N me=m_d

12

N mczmwd

 

 

 

 

 

 

compuz< m
.502;
N
. pduam puozu pauzu
5395‘ am.— e e e
m N N
-u-mu<zu_ueAT|i—m—mw chili—m_mcmm peu<zpmw
gown. .mouaoa .iv no:

38% m . 626m v.
oea-ee o<z_eu-eoa
pecan pound

$53.2 x M. 4—. 2335‘ 2
N
-imuezoae 4.1.23.0 inmoezuBeTpfec

.5 Y .25
_ *Nmzmm mm;
ozomk . drew a

oea-aa. . . o:a-aeo

-uuuumzo—ceATiipmpmw

eeeeo ~_ ease

13

and Type 2, are commonly found in the A80 and Lewis systems. Type 1
contains a Gal8(1-3)GlcNAc sequence while Type 2 contains a
Gal3(1-4)GlcNAc sequence, commonly called Nfacetyllactosamine. The Type
2 sequence can be synthesized by the galactosyltransferase isolated from
milk (15). The A,B, and H determinants can be synthesized from either
Type 1 or Type 2 chains. The most important carbohydrate in each chain
is known as the immunodominant sugar and for H specificity this
carbohydrate is L-fucose. A fucosyltransferase, the product of the H
gene and designated B-galactoside a(1-2) fucosyltransferase, is
responsible for the synthesis of the H-active substance through addition
of L-fucose from GDP-fucose, to either Type 1 or Type 2 chain

acceptors.

The Lewis system is characterized by the addition of fucose to the
Nracetylglucosamine of Type 1 chains. This is an a(1-4) linkage and is
catalyzed by B-galactoside o(1-4) fucosyltransferase which is a Le gene
product. The fucose linked a(1-4) to the Nfacetylglucosamine is the
immunodominant structure of the Lea antigen. However, the Leb
antigen contains not only fucose linked o(1-4) to Nracetylglucosamine
but also a fucose o(1-2) linked to galactose to form it's immunodominant
structure. The fucosyltransferase responsible for the a(1-2) linkage is
the H gene product.

The immunodominant sugar in the blood group A substance is
Nracetylgalactose linked o(1-3) to the galactose of the H-active
substance. The Nracetylgalactosaminyltransferase designated [fucosyl
o(1-2)] galactoside o(1-3)N:acetylgalactosaminyltransferase, responsible
for this activity has very strict substrate specificity for

FUCa(1-2)GalB(-) and uses UDP-Nracetylgalactosamine as the sugar

14

nucleotide donor (30,31). This enzyme, the product of the A gene, has
recently been purified from porcine submaxillary glands and
characterized (30,31).

Finally, the immunodominant sugar in blood group B antigenic
substance is galactose. The enzyme that catalyzes the transfer of
galactose from UDP-galactose to FUCo(1-2)Gale(-) is a galactosyltrans-
ferase designated [fucosylo(1-2)] galactoside a(1-3) galactosyltransfer-
ase, and is the B gene product.

111. Review on the Purification and Properties of Lactose Synthetase

Lactose Synthetase (UDP-galactose:glucose-B(1-4)galactosyltransfer-
ase; EC 2.4.1.22) catalyzes the synthesis of a galactose 8(1-4)
glycosidic linkage between Neacetylglucosamine or glucose. In the
presence of a specific protein, o-lactalbumin, the enzyme will only
transfer galactose to glucose to form lactose (32). However, in the
absence of aelactalbumin the enzyme shows a high acceptor specificity
for Nracetylglucosamine (14).

Lactose synthetase activity and the levels of a—lactalbumin have
been measured in the mammary glands from pregnant and lactating animals
and shown to be approximately 20% higher than in non-lactating animals
(33,34). Increased synthesis of lactose synthetase and o—lactalbumin
were observed by treatment of mammary explants from virgin mice with
insulin, hydrocortisone, and prolactin (35). The presence of
progesterone was found to decrease the synthesis of o—lactalbumin but
the synthesis of lactose synthetase remained unaltered (36). The
transferase accumulates in Golgi membranes during the last half of
pregnancy and once lactation commences the rate of a-lactalbumin

synthesis increases tremendously. The greatly increased orlactalbumin

15

synthesis is due to the drop in progesterone levels at parturition.
Lactose after it's synthesis, together with o-lactalbumin, casein, and
other milk proteins is enclosed in membranous vacuoles, derived from the
Golgi, and secreted by exocytosis (37). This mechanism accounts for the
release of proteins and lactose into the milk but also explains the
presence of galactosyltransferase in the milk. The inner surface of the
secretory vacuole corresponds to the intracisternal membrane of the
Golgi apparatus to which the galactosyltransferase is embedded. Once
the vacuole has fused with the cell surface, the transferase could
either dissociate from the membrane into the milk or be attacked by
proteolytic enzymes in the milk which cleave it from the membrane (39).

In tissues other than the mammary gland the galactosyltransferase
functions in the catalysis of plasma glyc0proteins (40). Under
physiological conditions, competition of glucose with glyc0proteins
containing terminal N:acetylglucosamine residues is not possible since
the Michaelis constant for glucose is 2.5 M in the absence of
o-lactalbumin. The galactosyltransferase is secreted from the Golgi
apparatus in vaccuoles and released into the extracellular fluid.

The enzymatic activity of lactose synthetase was first demonstrated
by Watkins and Hassid (41). Brodbeck and Ebner (42) showed that the
enzyme system is composed of two components, A and B, and Brew gt 11.
(14) clearly identified the A component as UDP-galactose:
N:acetylglucosamine 8(1-4) galactosyltransferase. Babad and Hassid (43)
were the first to report that the bovine galactosyltransferase was a
particulate enzyme and to partially purify the soluble form from milk.
The purification yielded a 70-fold enriched enzyme using ammonium

sulfate precipitation, heat denaturation, and hydroxylappatite

16

chromatography (43). The most common assay system for the
galactosyltransferase was also developed by Babad and Hassid (43) and
consisted of using a low molecular weight acceptor, glucose or
N-acetylglucosamine, and radioactive galactose in UDP-galactose. After
the appropriate incubation period, the reaction was diluted with ice
water and passed over a pipet column of Dowex 1 chloride resin. The
unreacted UDP-galactose remains bound to the column and the radioactive
galactose transferred to the acceptor passes through the column directly
into a scintillation vial and is counted.

Pure preparations of galactosyltransferase were first obtained from
bovine milk by chromatography on DEAE-Sepharose, cellulose phosphate and
a-lactalbumin attached covalently to Sepharose (44). o-Lactalbumin was
coupled to Sepharose by the cyanogen bromide activation method of
Porath, Axen, and Ernback (45). The partially purified
galactosyltransferase was applied to the o-lactalbumin Sepharose column
in the presence of manganese and glucose. After all the inert protein
had been eluted, the galactosyltransferase was specifically eluted by
removal of the glucose in the buffer (44). The galactosyltransferase
purified from the a-lactalbumin Sepharose was found to be pure by gel
electroohoresis (44). Later, Barker, et al. (15) introduced the use of
argarose derivatives of uridine diphosphate and Neacetylglucosamine for
the purification of the bovine galactosyltransferase. The
galactosyltransferase could be bound to the UDP-hexanolamine Sepharose
affinity resin in the presence of manganese ions and specifically eluted
by removal of the manganese ions and addition of EDTA. The
Nyacetylglucosamine Sepharose affinity column was also found to bind the

galactosyltransferase in the presence of UMP or UDP. The enzyme could

17

be specifically eluted with the addition of glucose and removal of the
UMP or UDP.

SDS-polyacrylamide gel electrophoresis showed three main bands with
molecular weights corresponding to 54,000, 48,000 and 42,000. All three
polypeptides have been shown to be active by gel filtration studies
(38). Treatment of the three components with trypsin was found to give
a progressive decrease in the high molecular weight form and increasing
amounts of the lower molecular weight forms (39). The
galactosyltransferase is a glycoprotein with a considerable amount of
carbohydrate (44).

Bell et 31. (46) have shown that the galactosyltransferase follows
a random equilibrium mechanism. In the absence of o-lactalbumin, the
enzyme first binds manganese and then can either bind UDP-galactose or
acceptor (46). When o-lactalbumin is added to the assay system it has
been shown to bind to either the E-Mn-UDP-Gal or the E-Mn-acceptor

complex (46).

IV Review of the Literature on the Fucosyltransferase

 

Fucosyltransferases are known which transfer L-fucose, from
GDP-L-fucose, onto various acceptor molecules in an a(1-2), o(1-3),
a(1-4) and a(1-6) linkage (47,48,49,50,51,52,53).

The enzyme initially was reported and the products partially
characterized in human milk and serum (48,52,53,54). Later, the enzyme
was reported in human submaxillary glands, human stomach tissue, human
bond marrow, rat small intestinal mucosa, porcine liver, porcine
submaxillary glands, ovine submaxillary glands, testis, bovine spleen,

human neuroblastoma cells, rat lymphocytes, Hela cells, and monkey

18

kidney cells (47,55,56,57,58,59,60). Enzymatic activity was measured
either directly from the soluble source, after ammonium sulfate
precipitation or after isolation of a Golgi or membrane fraction from
the various tissues. The major problem with many of the studies except
those done on the bovine spleen and porcine liver was that only
radioactive GDP-fucose, containing no measureable mass, was used as the
sugar nucleotide donor. One example of this is the use of 0.2 nmoles of
GDP-L-[14CJ-fucose containing 135,000 cpm as the nucleotide donor

and Nfacetylactosamine as the acceptor in an enzyme assay. After 20
hours only 4% of the radioactivity migrated in an area corresponding to
a trisaccharide following paper chromatography (55). The inherent low
activity of the enzyme and lack of substrate saturating amounts of
GDP—fucose lead to incubation times ranging from one hour to three days
(54,58). Isolation of the product usually entailed descending paper
chromatography, a double chromatographic method, or high voltage
electrophoresis followed by descending paper chromatography (54,57,58).
Thus, in addition to long incubation times, isolation of the product for
scintillation counting required, at least, an additional 24 hours.

The possibility of various positional isomers of fucose in
oligosaccharides led to the development of procedures for determination
of the linkage position. A common procedure used to show fucose o(1-2)
linked to galactose when low molecular weight acceptors were used took
advantage of the Lobry de Bruyn-Alberda van Ekenstein transformation.
Under alkaline conditions, at elevated temperatures, 2' fucosyllactose
yields 2' fucosylgalactose, a small amount of 2' fucosyltalose and
degradation products of glucose. The 2' fucosylgalactose can be

identified by descending paper chromatography and compared against

19

standard 2' fucosylgalactose isolated from milk (47,48). Alternatively,
an a(I-2) fucosidase, which is specific for the 2' fucosylgalactose

linkage, can be isolated from Colstridium perfringens and used to

 

determine if the fucose linkage is o(1-2) (58,61). Two fractions can be

isolated from Trichomonas foetus that contain fucosidase activity (62).

 

One is specific for the o(1-2) linkage to galactose and the second
fraction specific for fucose linked either o(1-3) to Nfacetylglucosamine
(or glucose) or o(1-4) to Nfacetylglucosamine (62,63).

Serological activity is often used for information on the linkage
position of fucose. For example, fucose linked o(1-2) to galactose of
.N-acetyllactosamine, H-active trisaccharide, should inhibit the
agglutination action on 0 cells of anti H sera. This type of inhibition
of hemagglutination has been used to characterize the various fucosyl
linkages in oligosaccharides (64,65). For determination of H-active

products the lectins Ulex europeus and Lotus tetragonalobus were used

 

 

with blood group 0 cells. Chester gt_al. (63) synthesized various
methyl, phenyl, and nitrophenyl glycosides for use as artificial
acceptors for the fucosyltransferase. One acceptor,
B-D-phenylgalactOpyranoside was said to be specific for accepting fucose
in an 0(1-2) linkage to galactose (63).

Most of the procedures described above make use of the following
various artificial acceptors; lactose, Hyacetyllactosamine and
B-D-phenylgalactoside. Studies have also been done with macromolecular
acceptors isolated from various tissue sources. Some of the
macromolecular acceptors used are: asialofetuin, asialo-ol-glycopro-
tein, and porcine submaxillary mucin. Fucosyltransferase activity is

assayed by precipitation of the macromolecular acceptor with

20

trichloracetic acid and either the radioactivity determined directly or
the protein subjected to chromatography (56). Glycolipids have also
been used for acceptors in the fucosyltransferase reaction. The
reaction product glycolipids are usually extracted with organic solvents
and subjected to chromatographic analysis (58).

Jabbal and Schachter (57) published one of the first papers in
which substrate saturating amounts of GDP-fucose were used to assay the
enzyme. The enzyme was isolated from porcine liver by preparation of a
48,000 X g membrane pellet. Both low molecular weight and
macromolecular acceptors were used. These were assayed for radioactive
fucose transfer by high voltage electrophoresis followed by descending
paper chromatography (57). Their enzyme was found to be unstable at 4°C
and -20°C for over 2 to 4 days, respectively (57). The
fucosyltransferase activity had a broad pH optimum between 6.5 and 8.5
and both magnesium and manganese were equally effective in stimulating
enzyme activity. Triton X-100 (0.2%) and GTP (.5mM) had a stimulating
effect. The enzyme was inhibited 65% by SOmM EDTA. It was shown that
the enzyme obeyed Michaelis-Menten kinetics with a Km calculated to be
7.8 x 10'5 M for the nucleotide sugar. Basu §t_al, (58)
using substrate saturating concentrations of GDP-fucose, investigated a
fucosyltransferase in bovine spleen. They isolated a fraction rich in
Golgi, end0plasmic reticulum and mitochondria enriched 24 to 46 fold
over the crude homogenate (58). The membrane bound enzyme transferred
fucose from GDP-fucose to various glycosphingolipids. Treatment of the
enzyme with EDTA resulted in a 95% loss of activity. In contrast to the
enzyme reported by Jabbal and Schachter (57), GTP (0.5 mM) was found to

inhibit the enzyme 30 percent. Maximal activity was found with the

21

cationic detergent G3634A (58). GDP-Fucose was found to obey
Michaelis-Menten kinetics with a Km of 3.6 x 10-4M (58). The
calculated Km for the acceptor lactosylceramide was found to be
6 x 10'4M. Products were characterized by microimmunodiffusion

using Ulex europeus and hemagglutination inhibition techniques (58).

 

The fucosyltransferase reported by Basu gt a1. (58) had a wider
range of specificity for the acceptor substrate than the porcine liver
enzyme reported by Jabbal and Schachter (57). These conclusions were
based on enzymatic assay of crude tissue preparations containing
membrane bound fucosyltransferase. It is important to note that the
substrate specificities using membrane fragments can be assigned only
tentatively since the influence of the membrane fragment is not known.
Likewise, only apparent kinetic values and physical properties can be
established. Further characterization of the fucosyltransferase with
respect to the number of isoenzymes present, substrate specificities,
(various glyc0proteins verses glycolipids), kinetic properties and
physical properties will have to await solubilization, purification and
characterization of the enzyme from various tissue sources.
Purification of these isoenzymes will be extremely difficult due to the
inherant low activity (pmoles.hr"1.mg'1 protein), difficulty in
solubilization, separation of the isoenzymes, and characterization of
their substrate specificities and products.

Several investigators attempted to overcome the solubilization
difficulties by isolation of the enzyme from soluble sources such as
serum (54,66). However, the major problem with soluble sources of the

fucosyltransferase are the inherant low activities. A soluble

22

B-galactoside o(1-2)fucosyltransferase has been reported in all ABO
secretors. All sera examined were also found to contain a B-galactoside
a(1-3) fucosyltransferase of unknown function. Schenkel-Brunner et_gl.
(67) reported an activity for both fucosyltransferases of 7.5 pmoles of
product formed/25ml serum/72 hours. Such a low activity makes
purification and characterization from soluble sources almost
impossible.

Recently, the B-galactoside o(1-2) fucosyltransferase has been
purified to homogenity from porcine submaxillary glands using
solubilization with Triton X-100 and GDP-hexanolamine Sepharose affinity
chromatography. (T.A. Beyer and R.L. Hill, personal communication).

V Review of 13C-NMR Spectroscopy and 13C-13C Spin-Spin

 

Couplinngonstants.

 

Nuclear magnetic resonance spectroscopy was initially devoted
almost entirely to the proton. The proton is ubiquitous in organic
molecules has a high natural abundance of the 'H isotope (99.98%), a
spin of 1/2, and a high magnetic susceptibility which produces strong
resonance signals.

With the advent of fast Fourier transform techniques and pulsed
excitation of nmr signals the observation of 13C-nmr spectra and
measurement of 13C-X spin-spin coupling constants have become
possible.

The common isot0pe of carbon, 12C, has a nuclear spin of 0 and
is nmr inactive. The isotope of carbon which is nmr active, 13C,
exists in nature to the extent of 1.1%. The second inherent problem
with the 13C nucleus is the low magnetic susceptibiltity, only 0.016

that of 'H. However, 13C nmr spectroscopy does offer several

23

advantages over conventional 'H nmr. Carbon-13 nuclei resonate over 200
ppm of the applied field where 'H nuclei only resonate over 12 ppm of
the applied field. 13C NMR spectra, normally obtained with broad

band 'H decoupling to remove spin-spin couplings between 13C and 'H,

are usually much less complex than corresponding 'H spectra allowing the
interpretation of more complex molecules. Proton decoupling also
enhances the carbon signals due to the nuclear Overhauser effect.

Carbohydrates provide excellent examples of the usefullness of
13C nmr spectrosc0py. Chemical shifts and coupling constants are
very sensitive to substitution, configuration and conformation. This
sensitivity often makes assignment of resonances quite difficult but
once assignments are made, the same sensitivity yields information about
structure that would not other wise be obtainable.

Assignments of chemical shifts in 13C nmr spectra of
carbohydrates have been established by selective [13C] enrichment
(68,69), deuterium isotope shifts (70,71), effects of derivatization on
chemical shift (72) and the use of one bond 13C-‘H coupling
constants (68,73).

Since organic compounds containing carbon also have other nuclei
possessing spin in close proximity to the carbon nucleus, these spins
may be coupled so that various combinations of spin states may differ in
energy and fine structure in the resonance bands produced.

The first spin-spin couplings involving the carbon nucleus to be
observed were one bond carbon-hydrogen couplings ('JCH). Due to the
higher sensitivity of the 'H, these couplings were usually extracted

from the 'H spectra. When specifically 13C-enriched compounds are

24

available, however, the 'JCH coupling can be obtained more directly
from the 13C spectra.

Bock_et-al. (73), Bock and Pedersen (74) and Walker gt _1. (68)
have measured 13C-‘H coupling constants ('JCH) in hexoses,
pentoses and their derivatives. Their data show that in pyranosyl
derivatives 'JCl-Hl is found to be approximately 10Hz smaller
( 160 Hz) when the H-1 is axial than when the H-1 is equatorial ( 170
Hz).

The magnitude and sign of two bond C-H couplings (ZJCCH) in
carbohydrates have been studied by Perlin and his coworkers (75, 76) and
Walker gt_al. (68). Perlin and his coworkers (75,76,77) demonstrated
that the orientation and electronegativity of the substituents on the
13C nucleus affected the magnitude of the two bond coupling between
13C and 'H.

The establishment of a Karplus-type (78) relationship between
Vicinal coupling of 'H nuclei and the magnitude of the torsional angles
led to investigations concerning the angular dependence of three-bonded
(Vicinal) carbon-proton couplings. Calculations for propane suggest a
classical Karpus-type relationship for 3JCH with a minimum at a
dihedral angle of 90°, a maximum at 0° and a larger maximum at 180°
(79). Experimental studies on aliphatic 3JCH values and
conformational studies on propionaldehyde strongly suggest this
dependence of coupling constant on dihedral angle (80,81). Schwarcz and
Perlin (76) have established a Karpus type relationship for 3JC-H
coupling in carbohydrates. Coupling constants of 0,2 and 6H2 were found

to correspond approximately to dihedral angles of 100, 60 and 180°,

25

respectively. They also observed that Vicinal 13C-H coupling was
very sensitive in the 60° region of the dihedral angle (76).

Measurement of natural abundance 13C-13C coupling constants
is exceedingly difficult since two atoms of 13C are present in the
appropriate relationship in 1 of every 104 molecules. However, one
bond ('JCC) coupling constants have been observed on natural
abundance compounds using Fourier transform instruments and
concentrations of compound near 70% (82). The vast majority Of‘JC-C
values have been determined by the use of singly or multiply 13C-
enriched compounds. Carbon-13 enrichment is necessary when large
amounts of compounds are not available and when small coupling constants
are to be observed. Karabatsos et a1. (83) have predicted that
13C-13C coupling constants would be 0.16-0.49 times the
corresponding 'H-‘H coupling constants. The small magnitudes of the
coupling constants are compensated for by the simplicity in which they
can be observed. Carbon-carbon coupling constants have recently been
the subject of three excellant reviews (84,85,86).

The value of one bond, 'JC—C’ carbon-carbon coupling constants have
been shown to be a function of hybridization, substitution, ring size
and substituent orientation (87). Substituent orientation effects, when
observed, appear to be small in most cases. Walker ££.El° (68) observed
one bond carbon-carbon coupling constants in specifically enriched
[1-13CJ-aldopyranoses in the range of 46 Hz. The magnitude of the
'JC-C coupling constant did not appear to be sensitive to substituent
orientation at C1 or C2.

Two-bond couplings between carbon nuclei have been known for some

time to depend on the hybridization of the carbons involved, the

26

electronegativity of substituents on the carbons and the bond angles
between the two coupled carbon nuclei (88). Values for ZJC-C

coupling between two sp3 hybridized carbons fall in the range of
0.5-4.4 Hz (85). Walker et a1. (68) observed in carbohydrates that
2J(Cl-0-C5) was large in o-anomers but small and not resolvable in the
B anomers. This was explained in terms of the "dihedral angle rule"

adopted by Perlin (75,76) to explain 20 coupling constants. A

C—H
substituent trgߤ_to the carbon gives a positive contribution and a
substituent in the gagghg_position gives a negative contribution to the
coupling. Carbons in the carbohydrate ring were assumed to be
equivalent to oxygen ring atoms in their contribution to the coupling
constant (68). However, this approach does not explain the behavior of
2J(C1-C3) coupling in carbohydrates.

Three bond carbon-carbon couplings have been investigated in
aliphatic carboxylic acids and alcohols by Marshall §t_gl, (89) and
Barfield (90,91). They obtained a “modified" Karplus type curve with a
minimum and maximum displaced from 90° and 180° when compared to the 'H
Karplus relationship. The dependence on the dihedral angle, 0, was
similar to the curve found for 3J(FF) (92). Dihedral angles of 0, 80
and 180° were found to correspond to coupling constants of 2, 0 and 4H2,
respectively (89). Recently, Barfield (91) demonstrated from
theoretical calculations that nonbonded interactions are substantial for
coupling to 13C and may lead to deviations from the "Karplus type"
relationship of dihedral angle and coupling constant. This deviation
has been explained by the importance of interactions of the "rear lobes"

of the carbon hybrid orbitals in the trans arrangement (91).

27

Vicinal carbon-carbon couplings of the magnitude of 3-4Hz were
observed in [1-13CJ-enriched aldopyranoses between C1 and C6 (68).
This coupling constant corresponds to a trgns relationship between C1
and C6 and may therefore represent a near maximum for 3JC-C
coupling in carbohydrates. Coupling between C1 and C4 (dihedral angle =
60°) in the [1-13CJ-enriched carbohydrates was not observed (68).
The Vicinal relationship between C1 and C4 in the pyranose ring is
gagghe and therefore only a small coupling constant would be predicted.
However, coupling within cyclic systems is complicated by the fact that
there are two pathways through which coupling can occur. In the case of
C1-C4 coupling, the pathway through C2 and C3 is the mirror image of the
pathway through 05 and C5. If the couplings have different signs there
would be a cancelation of couplings decreasing the likelihood observing
a coupling which would have a small value. Although a well defined
relationship between dihedral angle and coupling constant in

carbohydrates for 30 coupling has not been established, values of

C-C
3JC-C observed in the pyranose ring can serve as estimates for the

maximal and minimal values expected for 30 coupling in

C-C
carbohydrates.

METHODS

1. Materials

Bovine milk was obtained from the Michigan State University dairy.
Porcine submaxillary glands were purchased from Bio-Resources Dallas,
Texas. [6-3HJ-L-Fucose, [UL-14c1-L-fucose, GDPEUL-14CJ-L-fu-
cose, UDPEUL-14CJ-D-galactose, [UL-14CJ-D-galactose, D-lyxose,
and K[14C]CN were purchased from New England Nuclear. Potassium
[13C] cyanide (K13CN) was supplied by the Los Alamos Scientific
Laboratory, University of California, Los Alamos, N.M., with 99.64
percent purity and 90.7 atom percent 13C enrichment. Galactose,
Nracetylglucosamine, L-fucose, hexanolamine, UDP-galactose, UMP, GMP,
GDP, Dowex 50-X8, Dowex 1-X8, Bio-Gel P2 (-400 mesh),
Sulfopropyl-Sephadex, Sephadex G50, Sepharose 4B and Chelex-IOO were
purchased from Sigma Chemical Company. Silica Gel G, 250 or 500

microns, thin layer chromatography plates were purchased from Analtech.

11. Analytical Procedures
1. Thin Layer Chromatography

Thin layer chromatography was performed on Silica Gel G plates, 250
or 500 microns, using two solvent systems: (I) ethyl acetate: hexane
(1:1v/v); (II) n-propanol:glacial acetic acid:water (85:22:3 v/v).
Plates were developed with 2N H2504 followed by heating at 110° for
5-15 min. Radioactivity was detected on a Berthold radioscanner.
Ascending chromatography was performed on Whatman #1 paper in solvent
system (111), isopropanol: 1.3 M aqueous sodium acetate buffer, pH 5.0
(7:3 v/v).

28

29

2. High Voltage Paper Electrophoresis

High voltage paper electrophoresis was performed on Whatman #1 paper
in 0.05 M triethylammonium bicarbonate buffer, pH 7.4, at 2100V and 32
d.c. amps for 1-2 hr. Compounds were identified by either molybdate
phosphate Spray (93), ninhydrin spray (94), UV absorbance or
radioactivity.
3. High Pressure Liquid Chromatography

High pressure liquid chromatography was performed on a Whatman
Partisil PXS 10/25 SAX column at 1000 psi using 0.2M potassium phosphate
buffer, pH 3.4. Compounds were detected by monitoring the column
eluate with a Altex dual wavelength UV detector.
4. Gas Chromatography

Gas chromatography was performed on a Varian Aerograph Series 2100
gas chromatograph, with a flame ionization detector. The column was
either an 0V-17(3%) on High Performance Chromosorb G or a Hi-eff 3CP
(10%) on GAS-CHROM 0, 80-100 mesh, (both supplied by Applied Science
Laboratories, INC) with helium as the carrier gas. Silyl derivatives
were prepared by drying 1-10 mg of material in a derivatization vial and
adding 150 pl of pyridine followed by 250 pl N, 0-bis-(trimethylsilyl)-
trifluoroacetamide (BSTFA) with 1% trimethylsilylchloride obtained from
Pierce. The derivative mixture was incubated at 60° for 30 min.
Separations were achieved with a linear temperature program from 100 to
250° at 8° per min.
5. Radioactivity Measurements

Radioactivity measurements were made with a Beckman LS-100 or

LS-7000 scintillation counter. Aqueous samples (0.2 to 1.0 ml) were

30

mixed with 2-5 ml of scintillation fluid (39 PPO, 0.29 dimethyl POPOP, 1
liter Triton X-100 and 2 liters toluene) and counted.
6. Carbon-13 NMR Spectroscopy

Carbon-13 nmr spectra were obtained with a Bruker WP-60 or 180,
15.08 or 45 MHz, Fourier-transform spectrometer equipped with quadrature
detection. Spectra were obtained in the proton-decoupled mode with 4K
spectral points at 36°. High field carbon-13 nmr spectra were obtained
on a Nicolet 360, 90 MHz, Forrier transform spectrometer equipped with
quadrature detection. Spectra were obtained using a two level
decoupling program to prevent excessive sample heating due to the high
decoupling power required. A spectral window of 1800 Hz was used when
the entire spectrum of the compound was required and a spectral window
of 1000 Hz used for doubly enriched compounds. All spectra were
obtained with 16384 spectral points at 14°. Chemical shifts are given
in ppm and referenced to the Cl resonance of B-D-glucopyranose in a
water solution and are accurate to 1 0.01 ppm with respect to internal
TMS (68).
7. Quantitation of Hexanolamine Compounds

Quantitation of hexanolamine derivatized compounds containing a
primary amino group was performed using the fluorescamine assay (95).
Hexanolamine, 0 to 80 pmoles was used as a standard. To a solution
containing the primary amino group, 1.5 ml of 0.2M borate buffer, pH
9.25, is added and the solution mixed well. With constant vortexing,
0.5 ml of a 30% w/v solution of fluorescamine in dry pndioxane is added.
Complete mixing is essential since fluorescamine is decomposed by water.
The fluorescence was determined by excitation at 390 nm and emission at

475 nm on an Aminco-Bowman spectrOphotofluorometer using a Xenon lamp

31

power supply. The fluorescamine solution in dioxane, if kept dry at 4°,
remains good for about 2 months.
8. Protein Determination

Protein concentrations were measured using the fluorescamine assay
of Bohler gt_§l, (95) with bovine serum albumin as a standard.
111. Synthetic Procedures

1. GMP-Morpholidate and UMP-Morpholidate

GMP-Morpholidate and UMP—morpholidate were synthesized by the

method of Moffat (96). The sodium salts of GMP and UMP were converted
to the free acids by passage over a 4 x 20 cm column of Dowex 50
(hydrogen form; 20-50 mesh) resin. The progress of the reaction was
followed by high voltage paper electrophoresis and was determined to be
complete when only the morpholidate derivative was detected by UV and
phosphate analysis. Trituration with ether yielded a fluffy white
powder in approximately 60% yield.
2. 6-Amino-l-Hexanol Phosphate (I)

6-Amino-1-hexanol phosphate was prepared by the method of Barker gt
V91. (68).
3. NrTrifluoroacetyl 6-Amino-1-Hexanol Phosphate (II)

NrTrifluoroacetyl 6-amino-1-hexanol phosphate was synthesized from
6-amino-1-hexanol phosphate (1) by the method of Barker et_gl, (68).
4. P'-(6-Amino-1-Hexyl)-P-(5'-Guanosine)-Pyrophosphate (III)

N:Trifluoroacetyl-6-amino-l-hexanol-GDP (III) was synthesized from
compound (II) and GMP-morpholidate by the procedure of Moffat (96).
Solutions containing 1mmole of GMP-morpholidate or lmmole compound (11)

plus 1 mmole tri-N-octylamine in 100 ml pyridine, that had been

32

distilled over calcium hydride, were evaporated three times with
pyridine. The compounds were dryed jg laggg over Mg(Cl04)2. In a
glove box, under a stream of N2 gas, the GMP-morpholidate was
dissolved in compound (II) and tri-N-octylamine. This was evaporated to
dryness at 40° and another 50 ml dry pyridine added to dissolve the gum.
The flask was then flushed with N2 gas, stoppered, and stored in a
desiccator over P205 under a slight vacuum for 48 hours at room
temperature. The pyridine was then removed and 100 ml H20 added.
This was extracted 3 times with 50 ml aliquots of ether. The aqueous
phase was applied to a 1 x 25 cm Dowex-1 (chloride form; 200-400 mesh)
column and washed with 125 ml H20. A linear gradient of 200 ml 0.01 N
HCl as starting solvent and 200 ml 0.01 N HCl containing 0.6 M LiCl as
the limit solvent was applied. Fractions were collected every 6 ml and
absorbance monitored at 262 nm. Three UV absorbing peaks were found,
pooled, and analyzed by high voltage paper electrophoresis, ascending
paper chromatography in solvent III and by 13C-nmr spectroscopy.
The first peak contained unreacted GMP-morpholidate and GMP. Peak 11
contained unreacted compound (II) and P1, PZ-diquanosine-S‘pyrophos-
phate. Peak III contained the desired product N-trifluoroacetyl
6-amino-1-hexanol-GDP in 73% yield. N-Trifluoroacetyl-6-amino-l-hex-
anol-GDP was identified by 13C nmr spectroscopy. Compound III,
6-amino-1-hexanol-GDP, was obtained by hydrolysis of the trifluoroacetyl
derivative at pH 12 overnight at room temperature. Completion of
hydrolysis was checked by high pressure liquid chromatography.
0. P'-(6-Amino-1-Hexyl)-P-(5'-Uridine)-Pyr0phosphate (IV)

Compound IV, 6-amino-1-hexanol-UDP, was synthesized as described

above for compound III.

33

E. GDP-Hexanolamine Sepharose

The procedure used is modified from that of Cuatrecasas (97).
Sepharose 4B (559;110 ml) was washed with 2 liters of water and
suspended in 55 ml 5M phosphate buffer, pH 12.0, and 55 ml of H20.
This was placed in an ice bath with efficient stirring. Cyanogen
bromide (229), dissolved in 25 ml of acetonitrile, was injected into
the Sepharose solution with a syringe and the reaction allowed to
proceed for 9 min maintaining pH 12 wih 8N NaOH. This was quickly
filtered and washed with 2-3 liters of ice water and the gel added to
0.70 mmoles of compound III in 70 ml of H20. Total time for filtering
was 3 minutes. The initial pH of compound III and the activated gel was
12.5. After 5 minutes the pH was brought to 10.2 with HCl and the
mixture shaken at 4°C for 15 hours. The gel was then filtered and
washed with 3 liters of H20. The UV absorbance in the filtrate was
used to determine that 5 pmoles of GDP-hexanolamine per ml of Sepharose
was bound.

Before use in the purification of the fucosyltransferase, the
affinity resin was washed with 300 ml each of 0.1M boric acid-potassium
hydroxide, pH 9.0, containing 0.5M NaCl; water; 0.1M sodium acetate, pH
5.0, containing 1M NaCl; and finally 2 liters of water.

7. UDP-Hexanolamine Sepharose

UDP-hexanolamine Sepharose was prepared as described for
GDP-hexanolamine Sepharose. This affinity resin was used for the
preparation of the galactosyltransferase.
8. UDP-B-D-Galactose

B-D-Galactopyranose pentaacetate (V) was prepared by the procedure

of Bates gt g1. (98). The product was characterized by gas

34

chromatography on a Hi-eff 3CP column and 13C nmr spectrosc0py in
deuterated chloroform. The 0:8 ratio of the pure pentaacetates was
found to be 10:90, respectively, by 13C nmr. The C1 resonances

for the a and B anomers in CDCl3 were found to be 90.6 and 93.3 ppm,
respectively. a-D-Galactopyranose-l-phosphate (VI) was prepared by
melting compound (V) in crystalline phosphoric acid following the
procedure of McDonald (99). After deacetylation and removal of the
lithium phosphate, according to the above procedure, 13C nmr showed
approximately an 85:15 ratio of the a and 8 sugar phOSphates,
respectively. The C1 resonance for the o-galactose-l-phosphate

appears as a doublet at 95.7 and 95.4 ppm. The galactose-l-phosphate
was passed over a Dowex 50 (hydrogen cycle; 50-100 mesh) column and the
eluate collected in cyclohexylamine. The overall yield for the
synthesis of galactose—l-phosphate, based on phosphate analysis, was
58%. The anomers of galactose-l-phosphate were separated by passing 15
mmoles of galactose-l-phosphate in the cyclohexylammonium salt through a
6 X 48 cm Dowex 1 (bicarbonate form; 200-400 mesh) column, washing with
1.2 liters of H20 and eluting the anomers with a linear gradient of

3.5 liters of H20 as starting solvent and 3.5 liters of 0.35 M
triethylammonium bicarbonate buffer, pH 7.4, as the limit solvent.
Fractions of 10 ml were collected and assayed for phosphate. Two
phosphate containing peaks were eluted from the column. The first peak
was shown to be orD-galactose-l-phosphate and the second peak
B-D-galactose-l-phosphate by 13C nmr spectroscopy. The
a-D-galactose-l-phosphate was pooled, concentrated and converted to the
pyridinium salt by repeated evaporation from pyridine. UDP-galactose

was synthesized by condensing 1 mmole each of o-D-galactose-l-phosphate

35

and tri-n-octylamine with 1 mmole of UMP-morpholidate following the
modified procedure of Moffat (96) used to synthesize compound (III).
The reaction was followed by high voltage electrophoresis and high
pressure liquid chormatography. After 24 hours of reaction at room
temperature the reaction was judged to be complete based on UV
absorbance under the peaks corresponding to UMP-morpholidate and
UDP-galactose plus P1, PZ-diuridine-S' pyrophosphate. The reaction
mixture was concentrated to dryness, extracted three times with 200 ml
ether, and the aqueous layers combined and applied to a 2 X 55 cm
Dowex-1 (chloride form, 200-400 mesh) column at a flow rate of 50 ml per
hour. The column was washed with 0.5 liters of water and eluted with a
linear gradient of 2.25 liters of 0.01M HCl as starting solvent and 2.25
liters of 0.01M HCl with 0.1M LiCl as limit solvent. Fractions of 8 ml
were collected and assayed for UV absorbance at 260 nm. Three peaks
were eluted from the column in the order of increasing salt
concentration and determined by high pressure liquid chromatography to
be UMP-morpholidate, UDP-galactose, and Ple-diuridine-S'pyrophos-
phate, respectively. The fractions containing UDP-galactose were
pooled, concentrated, and desalted by passage over a P2 column (-400
mesh). UDP-galactose was characterized by 13C nmr and the purity
checked by high pressure liquid chromatography and high voltage
electrophoresis. The overall yield of UDP-galactose from galactose-1-
phosphate was 20%.
9. GDP-o—L-Fucose

GDP-o—L-Fucose was synthesized by the method of Nunez gt_§l, (100).

10. 2-Acetamido-3,4,6-Tri-0-Acetyl-2-Deoxy-Glucopyranosyl Chloride VII

36

Compound VII was synthesized by the procedure of Horton (101) and
the purity determined by gas-liquid chromatography, on a high efficiency
3 CP column.

11. NfTrifluoroacetyl-6-Amino-l-Hexyl-Z-Acetamido-2-Deoxy-B-D-Glucopy-
ranoside VIII. Compound VIII was synthesized from compound (VII) by the

procedure of Barker et al. (68).

12. 6-Amino-l-Hexyl-2-Acetamido-2-Deoxy-8-D-Glucopyranoside (IX)
Compound VIII was incubated overnight at pH 12 in NaOH and
de-Nracetylation was judged complete by quantitative fluorescamine.
The product, compound IX, GlcNAc-B-hexanolamine was purified by binding
to a 1 X 10cm Dowex-50 (hydrogen form; 100-200 mesh) column and eluted,
after washing with water, with 1M pyridinium acetate. The pyridinium
acetate was removed by repeated evaporation.
13. Acetobromogalactose (2,3,4,6-Tetraacetylgalactopyranosyl Bromide)
(X)
Galactose pentaacetate, [1-13CJ-galactose pentaacetate and
[2-136] galactose pentaacetate were synthesized by the method of .
Bates et_gl, (98). Acetobromogalactose, 2,3,4,6-tetraacetylgalactOpy-
ranosyl bromide, was synthesized by dissolving 1.7 mmoles of dry
galactose pentaacetate in 4ml glacial acetic acid and adding 4ml of 30%
hydrogen bromide in glacial acetic acid. The reaction was allowed to
proceed for 30 min at room temperature and then poured onto ice in a
separatory funnel. Dichloromethane, 100ml, was added and the aqueous
layer removed and discarded. The dichloromethane layer, containing the
acetobromogalactose, was extracted once with 200ml of cold water, 3

times with cold saturated sodium bicarbonate, 2 times with water and

37

finally dried over M9504 for 30 min. The M9804 was removed by
filtration through a pad of celite and the dichloromethane solution
concentrated to a gum by rotatory evaporation at 40°. The gum was
further dryed over Mg(Cl04)2 under vacuum for 1-2 h.

14. 6-Amino-l-Hexyl-o,3-D-Galactopyranoside (XI)

Compound XI was synthesized from 2,3,4,6-tetraacetylgalactopyranosyl
bromide and Nytrifluoroacetyl-6-amino-1-hexanol. .N: Trifluoroacetyl-6-
amino-l-hexanol was prepared as described for Nytrifluoroacetyl-6—amino-
1-hexanol phosphate (11) and dryed overnight before use over
Mg(Cl04)2 under vacuum. Mercuric cyanide was also dryed overnight
over M9(Cl04)2 under vacuum. Nitromethane:benzene (1:1v/v) used in
the condensation reaction was dryed over 4 A molecular sieves before
use.

Dry 2,3,4,6-tetraacetylgalactOpyranosyl bromide, 1.6 mmoles, was
mixed with 1.8 mmoles of Ngtrifluoroacetyl-6-amino-1-hexanol, 1.7 mmoles
H9(CN)2 in 50 ml nitromethane:benzene (1:1v/v) and 29 4 A molecular
sieves. The solution was heated in an oil bath, with efficient
stirring, at 56°. After 12h the reaction was complete, as determined by
13C nmr spectroscopy. The reaction mixture was filtered,
concentrated to a gum by rotatory evaporation at 40°, diluted in
dichloromethane and extracted with water to remove traces of H9(CN)2.
The dichloromethane layer was again concentrated to a gum and 20 ml of
triethylamine:methanol:water (1:11:4 v/v) added and the solution stirred
overnight for complete de-ancetylation of the galactosides. After
removal of the triethylamine:methanol:water by rotatory evaporation, the
gum was dissolved in 20 ml water and the pH maintained for 4h at 11.0

for de-N:acetylation. Carbon-13 nmr revealed 4 major components;

38

6-amino-1-hexyl B-D-galactopyranoside, 6-amino-1-hexyl
o-D-glactopyranoside, B-D-galactopyranose and a-D-galactopyranose in the
ratio of 36:36:21:8, respectively. The a and B anomers of the
6-amino-1-hexyl galactopyranosides were separated by chromatography on a
3 x 70 cm Dowex 1-X2 (hydroxide form; 200-400 mesh) column by elution
with water. The o-anomer was the first to elute followed by the
B-anomer. Unreacted galactose was converted to acids which remain bound
to the column.
15. Ethyl-B-D-Galactopyranoside (XII)

Ethyl-B-D—galactopyranoside was prepared by the method of Chiang et
91. (102). Compound X, 2,3,4,6,-tetracetylgalactopyranosyl bromide, was
prepared as previously described. Acetobromogalactose, 2.8 mmoles, in
24 ml of dichloromethane is placed in a foil covered reaction flask.
With efficient stiring, the following additions are made: 0.23 ml (4
mmoles) of ethanol which had been dryed over 3 A molecular sieves, 0.559
silver carbonate that had been dryed at least 12h, 0.049 iodine and 19
4 A molecular sieves. The condensation was shown to be complete within
2h by tlc in solvent system (I). Ethyl-B-D galactose tetraacetate had a
Rf of 0.43 and acetobromogalactose had a Rf of 0.54 in solvent
system (I). The reaction mixture was filtered, concentrated to a gum
and 20 ml triethylamine:methanol:water (1:11:4w/v) added for
de-Qracetylation overnight. The triethylamine:methanol:water was
removed by rotatory evaporation and the compound purified over a 3 X 20
cm Dowex 1-X2 (hydroxide form: 200-400 mesh) column by elution with
water. Yield, based on weight was 92%. Carbon-13 nmr revealed the
presence of only 1 anomer, B. at 104.2 ppm downfield from TMS.

16. 5-Deoxy-L-Lyxose (XIII)

39

Compound XIII was prepared by the oxidative decarboxylation of
L-fuconic acid.

L-Fuconic acid was prepared by bromine oxidation of L-fucose, by
the method of Isbell (103). Passage of the reaction mixture through a
4 X 30 cm Dowex 50-X8 (hydrogen form; 20-50 mesh) column and rotatory
evaporation at 45° yielded the lactone as a syrup. The lactone was
shown to be pure by glc of it's trimethylsilyl derivative and by
13C-nmr spectroscopy. Yields of L-fuconolactone based on weight of
the dryed syrup were 98%. Hydrolysis of the lactone to the sodium salt
was accomplished by maintaining a pH of 9.4 with 5M sodium hydroxide for
1 hour.

The oxidative decarboxylation of the sodium salt of L-fuconic acid
was carried out using the Ruff degradation (104). The sodium salt (60
mmoles) of [6-3HJ-L-fuconate at pH 9.5 was dissolved in 200 ml H20
and 1.25 9 barium acetate and 0.62 9 ferric sulfate added. The mixture
was boiled for 3 minutes and allowed to cool to 65°. Hydrogen peroxide
(30%), 8 ml, was added and the mixture stirred with maintenance of the
temperature between 65°-75°. When the temperature cooled to 65°,
another 8 ml of 30% hydrogen peroxide was added and the temperature
again maintained between 65-75°. This was repeated once more with an
additional 8 ml of 30% hydrogen peroxide. After the reaction mixture
had cooled to room temperature, 5 g decolorizing charcoal was added and
the reaction mixture filtered through Celite. The solution was
immediately placed on a 4 X 50 cm Dowex 50-X8 (hydrogen form; 20-50
mesh) column connected to a 2 X 50 cm Duolite A6 (hydroxide form; 20-50
mesh) column at 4°. The only species eluting from the columns was

[5-3H]-5-deoxy-L-lyxose. The 5-deoxy sugar was shown to be pure by

4o

glc and C-13 nmr. The resonances for Cl through C5 in
[5-3HJ-5-deoxy-B-L-lyxose are 102.3, 78.2, 79.9 and 15.3,
respectively. The yield of [5-3H]-5-deoxy-L-lyxose by radioactivity
and weight as a dry syrup was routinely 40%.
17. [1-13c, 6-3H]-L-Fucose and [1-13c, 6-3H]-6-Deoxy-L-
Talose (XIV)

Potassium [13C] cyanide, 13.2 mmoles, was dissolved in 20 ml
H20 at 18° and the pH adjusted to 7.2 with 1M acetic acid.
[5-3HJ-5-Deoxy-L-lyxose, 12 mmoles, was dissolved in 6 ml H20 and
added to the cyanide solution. The pH of the reaction was maintained at
7.2 for 30 minutes. After 20 minutes, glc showed two main
TMS-derivatives appearing from the 0V-17 3% column at 172 and 175°,
corresponding to fucono and 6-deoxy-L-talononitriles, respectively. A
small amount of unreacted 5-deoxy-L-lyxose was also present, emerging
at 142°. The solution was adjusted to pH 4.2 and airated for 15 minutes
to remove any excess cyanide. The nitriles were reduced catalytically
with palladium-barium sulfate (5%, 60 mg/mmole nitrile) at 50 psi of
hydrogen for 1.5 hours.

1-Amino-1-deoxyalditols and aldonic acids, produced as side
products from the hydrogenation reaction were removed with Dowex 50-X8
(hydrogen form; 20-50 mesh) and Dowex 1-X8 (acetate form; 100-200 mesh),
respectively, as previously described (105). Separation of the epimeric
[1-13C, 6-3HJ-L-fucose and [113C, 6-3H]-L-talose was
achieved using a 4 X 90 cm Dowex 50-X8 (barium form; 200-400 mesh)
column and elution with water. [1-13C, 6-3H]-L-Fucose was found
to elute near the void volume of the column and

[1-13C, 6-3HJ-L-talose eluted approximately 150 ml later. The

41

overall yield for the synthesis of [1-13C, 6-3H]-L-fucose and
[1-13C, 6-3H]-6-deoxy-L-talose was 70% based on starting
[5-3HJ-5-deoxy-L-lyxose.
18. [1-13C]-D-Galactose (XX)

Compound XX was prepared by the general method of Serianni gt _1.
(105) from D-lyoxse.
19. [2-13CJ-D-galactose (XXI)

Compound XXI was prepared by the sequential addition of cyanide to
threose by the method of Serianni gt_gl, (105).
20. Purification of Bovine UDP-Galactose:Glucose-B(1-4)
Galactosyltransferase

The galactosyltransferase was purified by the method of Barker gt
al. (68). The purified enzyme, 0.05 mg/ml, was stored at 4° in 25 mM
sodium cacodylate, pH 7.4, 25 mM MnClz and 1 mM B-mercaptoethanol.
Under these conditions no loss in enzymatic activity was observed in 2
years. Dilute solutions of the galactosyltransferase were concentrated
in the presence of 0.01% Triton X-100. The nonionic detergent allowed
concentrations greater than 1 mg protein/ml and stabilized the enzyme at
these concentrations for up to a week at 4°. The rate of synthesis of
N-acetylactosamine, lactose or (6-amino-1hexyl)-4-9r(B-D-9alactopyran-
osyl)- 2-acetamido-2-deoxy-B-gluc0pyranoside was followed by measuring
the rate of transfer of 14C from UDP [UL-14C] galactose into '
disaccharide as described previously (46). All assays were initiated
with galactosyltransferase and incubated for 10 min at 37°. The enyzme
concentration was adjusted so no more than 15% of the radioactivity was

transferred. Standardization of enzyme solutions were preformed with

1.5mM UDP [UL-14CJ-galactose in the standard assay mixture.

42

21. Preparation of Galactosylated Disaccharides

Syntheses of the galactosylated compounds were carried out on a 140
pmole scale at 33° by a procedure similar to Nunez and Barker (13). The
reaction mixture contained 16 ml of N-acetylglucosaminide 8(1-4)
galactosyltransferase (43 units/ml of enzyme solution) in 25 mM sodium
cacodylate, pH 7.4, 25 mM MnClz and 2 mM B-mercaptoethanol, 60pl 0.01
mCi/ml UDP [UL-14C] galactose, 140 pmoles UDP-galactose, disodium
salt, 150 pmoles acceptor, and 4 mg BSA in a total volume of 18 ml.
After 6 h another 6 ml of N—acetylglucosaminide 8(1-4) galactosyltrans-
ferase was added and the incubation continued for another 12 hr.
22. Purification of Galactosylated Disaccharides

Reaction mixtures were passed over a 1 X 4 cm column of Dowex 1-X8
(chloride form; 100-200 mesh) and concentrated. Protein was removed by
repeated concentrations in a 10ml Amicon cell using an Amicon PM 10
membrane or by precipitation using the method of Somogyi (107).
Alternatively, glycosides of hexanolamine could be purified by binding
to a 1 X 5 cm column of Dowex 50-X8 (100-200 mesh) and elution with 1M
pyridinium acetate. The pyridinium acetate could be removed by repeated
evaporation lg 19999 at 40°. Separation of the galactosylated
disaccharide from the acceptor was achieved on a 4 X 90 cm column of
Bio-Gel P2 (-400 mesh) equilibrated in 0.1M triethylammonium
bicarbonate, pH 7.4. Triethylammonium bicarbonate could easily be
removed from the final product by repeated evaporation jg_yaggg at 40°.
23. Assay of a Porcine B-Galactoside o(1-2) Fucosyltransferase

Fucosyltransferase activity was determined by incubating 100pl
mixtures containing 0.5 pmoles lactose, N-acetyllactosamine, or Gal

B(1-4)GlcNAc-B-hexanolamine, 5 pmoles sodium cacodylate, pH 6.0, 2

43

pmoles MnClz, 0.2 mg bovine serum albumin, 2 nmoles GDP-fucose
containing approximately 5,000 cpm GDP [UL-14CJ-fucose and enzyme
solution at 33° for 15-60 min. Appropriate controls were included to
correct for the amount of nonSpecific hydrolysis of GDP-fucose.
Standardizations of enzyme solutions were performed, as above, with 0.08
pmoles GDP-fucose containing approximately 15,000 cpm of GDP
[UL-14CJ-fucose. Activity units are defined as pmoles of fucose
transferred from GDP-fucose to acceptor per minute. Reactions were
stopped by placing the tubes on ice and adding 0.5 ml water. Two
methods were then used in the work-up of the reaction mixtures. The
first method was used only with crude enzyme preparations and involved
the use of the acceptor GalB(1-4)GlcNAc-B-hexanolamine. The reaction
mixture was applied to a Pasteur pipet column of Dowex 50-X8 (hydrogen
form; 100-200 mesh) and washed under gravity with 9.5 ml of water. The
product was then eluted into a scintillation vial with 4 ml of 1 M
pyridinium acetate. The pyridinium acetate was removed on a hot sand
bath, scintillation fluid added and the vial counted for radioactivity.
Reaction blanks were approximately 60 cpm per assay. Standard
deviations of replicate analyses were 5 to 15 cpm. Alternatively, high
voltage electrophoresis can be used to isolate and quantitate the
product.

The second method was routinely used for assaying enzymatic
activity in the purification of the fucosyltransferase. The reaction
mixture is applied directly to a Pasteur pipet column of Dowex 1-X8
(chloride form; 100-200 mesh) and washed with 1.5 ml water. The eluate

is collected directly into a scintillation vial, 14 ml scintillation

44

fluid added and the vial is counted for radioactivity. Standard
deviations of replicate analysis were 2 to 8.
24. Purification of a Porcine B-Galactoside a(1-2) Fucosyltransferase.

The porcine B-galactoside 0(1-2) fucosyltransferase was partially
purified with modifications from a procedure devel0ped by T.A. Beyer and
R.L. Hill (personal communication).

The buffers employed are identified by letter as follows: Buffer A,
25 mM sodium cacodylate, pH 6.0; Buffer B, 25 nM sodium cacodylate, pH
6.0, 0.3M NaCl; Buffer C, 25 mM sodium cacodylate, pH 6.0, 0.15 M NaCl;
Buffer D, 25 "M sodium cacodylate, pH 6.0, 0.5 M NaCl; Buffer E, 25 mM
sodium cacodylate, pH 6.5; Buffer F, 25 mM sodium cacodylate, pH 6.0, 2
M NaCl.

The procedure used for large scale solubilization of the
fucosyltransferase was modified from the procedure used to solubilize a
porcine CMP-N-acetylneuraminate:B-D-galactoside 0(2-3)
sialyltransferase (106). All procedures were done at 4°.

Stgp_1: Removal of Mucin from the Porcine Submaxillary Glands

Porcine submaxillary glands, 2 kg, obtained frozen from
Bio-Resources, were quickly thawed under cold running water and placed
on ice. The glands were trimed of fat and connective tissue and ground
in a meat grinder at 4°. Portions of the ground glands, 5009, were
added to 2 liters of water and homogenized in a Waring blender on high
for three 30 sec bursts with one min rest inbetween bursts. The
homogenate was centrifuged at 7000 X g for 45 min. The fat and mucin
layer at the top of the centrifuge tube was removed by aspiration.
Pellets from 1kg of material were resuspended in 2 liters of Buffer A by

homogenization in a Waring blender with three 30 sec bursts on medium

45

and 30 sec intervals inbetween bursts. The slurry was brought to 20 mM
MnClz by dropivise addition of 60 ml 1M MnClz and stirred for 15
min. This slurry was centrifuged for 30 min as above and the
supernatant removed by aspiration. The pellet was resuspended and the
procedure repeated again. The washed pellets from 1 kg of tissue were
resuspended in 1.5 liters of Buffer A by homogenization in a Waring
blender with three 10 sec bursts on low and 15 sec intervals inbetween
bursts. This procedure was repeated with the second 1kg batch.
Stgp_2_ Triton X-100 Extraction

The resuspended pellets from 1 kg of starting material were brought
to 1% Triton X-100 and 20 mM MnClz, over a period of 15 min by the
addition of 100 ml 20% Triton X-100 and 40 ml 1M MnClz, and stirred
for 45 min. After stirring for 45 min, the solution was centrifuged for
30 min as above. This was repeated with the second k9 batch (II). The
supernatants (I and II) were removed and combined. Pellets (I and II)
were resuspended by homogenization on low in l.5 liters of Buffer A and
brought to 1% Triton X-100 and 20 mM MnClz, over a period of 15 min,
and stirred for 45 min. This was again centrifuged, as above, for 30
min. The supernatant (III) was removed and the pellets (III) were
discarded.
§tgp_§ Sulfopropyl-Sephadex Chromatography

The combined supernatants I and II were immediately treated
batchwise with 150 g SP-Sephadex, previously equilibrated with Buffer A,
and stirred for 30 min. Supernatant III was treated similarly with
509 SP-Sephadex. The gel, containing the bound fucosyltransferase,
was poured into a 5 X 20 cm column, washed with 500 ml of Buffer A and

eluted rapidly with Buffer B. Fractions, 20 ml, were collected and

46

tubes containing the fucosyltransferase pooled and node 50% (v/v) with
ice cold glycerol. The gel, from supernatant (III) was similarly poured
into a column and eluted with Buffer B.
Stgp_4, Sodium Chloride Elution of the Enzyme From GDP-Sepharose

The SP-Sephadex pooled enzyme from 4kg of tissue, stored in 50%
glycerol, was diluted 50% with Buffer C and applied to a 5 X 15 cm
column of GDP-Sepharose, equilibrated in Buffer C, at a flow rate of 100
ml/hr. After application of the enzyme solution, the column was washed
with 250 ml of Buffer C followed by 250 ml Buffer D. A 500 ml linear
gradient of 0.5 M NaCl in Buffer A as the starting solvent and 2.0 M
NaCl in Buffer A as the limit solvent was applied to the column followed
by 200 ml of Buffer F. Fractions of 20 ml were collected and assayed
for fucosyltransferase activity. The enzyme was found to elute with 2 M
NaCl. Fractions, 400 ml, containing the fucosyltransferase were pooled
and stored at 4°. The enzyme was stable at 4° for at least 3 months.
§tep_§ Sephadex G-50 Chromatography

Pooled fucosyltransferase from the GDP-Sepharose column was
desalted in batches of 200 ml, by application to a 5 X 110 cm column of
Sephadex G-50 (fine ) equilibrated in Buffer A. Fractions of 20 ml were
collected at a flow rate of 200 ml/hr. Fucosyltransferase activity was
pooled and quickly applied to a second GDP-Sepharose column.
Step_§_ Second GDP-Sepharose Column

The desalted fucosyltransferase, from Step 5, was applied at a flow
rate of 50 ml/hr to a l ml GDP-Sepharose column equilibrated in Buffer A
in a plastic syringe. The column was then washed with 20 ml Buffer A
and the fucosyltransferase eluted with Buffer F. Fractions of 1 ml were

collected and assayed for fucosyltranferase activity. The enzyme was

47

found to elute in a volume of 5-8 ml and could be stored with no loss of
activity for at least 2 months at 4°. Longer periods of storage
required dilution of the enzyme with 50% glycerol and storage at -20°.
25. Preparation of Fucosylated Oligosaccharides

All enzymatic syntheses were carried out at 33° and the extent of
reaction followed by radioactivity or high pressure liquid
chromatography. Incorporation of radioactivity from
GDP[UL-14C]-fucose into oligosaccharide was measured by removing a
20pl aliquot from the reaction and diluting it with 0.5 ml water. The
sample was transferred to a Pasteur pipet column containing Dowex 1-X8
(chloride form; 100-200 mesh) and collected in a scintillation vial.
The column was rinsed with 1.5 ml water and the eluate, collected in a
scintillation vial, counted for radioactivity incorporated into neutral
oligosaccharides. Nonspecific hydrolysis of GDP-fucose in the reaction
mixture could be measured when glycosidic derivatives of hexanolamine
were used as acceptors. A 20 pl aliquot was removed from the reaction
mixture and diluted with 0.5 ml water. This sample was applied to a
Pasteur pipet column containing equal proportions of Dowex 1-X8
(chloride cycle; 100-200 mesh) and Dowex 50-X8 (hydrogen cycle; 100-200
mesh). The column was washed with 1.5 ml of water and the eluate,
collected in a scintillation vial, counted for radioactivity. Only
radioactive fucose which had been hydrolyzed from GDP-fucose would elute
from the column.

High pressure liquid chromatography on a Partisil PXS 10/25 SAX
column could be used to estimate the proportions of GMP, GDP and
GDP-fucose in the reaction mixture. Retention times for GMP, GDP-fucose

and GDP, were 7, 12 and 24 min, respectively.

48

Synthesis of compounds were routinely performed in 50 umole
batches. The reaction mixture contained 4 ml of B-galactoside 6(1—2)
fucosyltransferase (0.62 units/ml enzyme solution), 30 pl 0.01 mCi/ml
GDP [UL-14C] fucose, 0.04 mmoles sodium cacodylate, pH 6.0, 0.15
mmoles MnClz, 7.5 mg BSA, 50-75 pmoles of acceptor oligosaccharide and
50 pmoles GDP-fucose in a total volume of 7.4 ml.

Reactions were incubated for 6 h and an additional 2 ml of B-galactoside
0(1-2) fucosyltransferase added and incubation continued for 12 h.
26. Purification of Enzymatically Synthesized Oligosaccharides

Reaction mixtures were first placed in a 10 ml Amicon cell
containing an Amicon PM10 membrane. The mixture was concentrated to 0.5
ml and diluted to 10 ml with water and reconcentrated. This was
repeated 2 more times. The protein free filtrate was then applied to a
small column containing 2 ml Dowex 1-X8 (chloride form; 100-200 mesh)
and 0.5 ml Chelex-IOO (sodium form ) and eluted with water.

The mixture was concentrated and applied to a 4 X 90 cm Bio-Gel P2 (-400
mesh) column equilibrated in 0.1M triethylammonium bicarbonate, pH 7.4
and pumped at a flow rate of 5 ml/h. The fucosylated product was
separated from the acceptor and concentrated. Occasionally, some salt
eluted with the fucosylated product and was removed by rechromatography
on a 1 X 40 cm Bio-Gel P2 (-400 mesh) column equilibrated in water at a
flow rate of 5 ml/h. Nonreducing compounds were treated with 0.5 ml
Chelex-IOO (sodium form) and reducing compounds treated with 0.5 ml
Chelex-lOO (hydrogen form) before nmr analysis. Compounds containing
glycosidically linked hexanolamine were completely eluted from the
Chelex-lOO (sodium form) with 1 M pyridinium acetate which was removed

by rotatory evaporation.

RESULTS

I. Synthesis of Galactgsylated Disaccharides

The Ngacetylglucosaminide B (1-4) galactosyltransferase can be
stored for 2 years with small losses in enzymatic activity in 25 mM
sodium cacodylate, pH 7.4, 25 mM MnClZ. 2 mM B-mercaptoethanol and 50
nM N—acetylglucosamine. The enzyme is stabilized by N-acetylglucosamine
(68).

Before synthesis of Gal8(1-4)GlcNAc-B-hexanolamine the stablizer,
‘Nracetylglucosamine, was removed by dialysis. The dialyzed enzyme was
found to lose 30% of it's enzymatic activity during 5 days at 4°.

Yields for the synthesis of disaccharides varied depending on
acceptor. When N—acetylglucosamine was used as an acceptor the yield
was routinely 85%. Yields using GlcNAc-B-hexanolamine were always lower
and varied between 60 and 75%.

II. Enzymatic Assay of the geGalactoside 6(1-2) Fucosyltransferase

During initial steps, the purification of this enzyme is
complicated by the presence of large amounts of endogenous acceptors and
hydrolysis of the sugar nucleotide, GDP-fucose. A small amount of
enzymatic activity could be measured in the crude homogenate using the
substrate Galp(1-4)GlcNAc-e-hexanolamine. This acceptor, containing a
primary amino group, allowed the fucosylated product, Fuc 0(1-2)
Gal8(1-4)GlcNAc-p-hexanolamine, to be bound to a cation exchange resin
and eluted with pyridinium acetate after unreacted GDP-fucose,
hydrolysis products, and fucosylated endogenous glyc0proteins were

washed through the column.

49

50

The routine assay of the B-galactoside o(1-2) fucosyltransferase,
following removal of the mucin, utilized an anion exchange resin to bind
unreacted GDP-fucose. The fucosylated product and free fucose are
eluted into a scintillation vial for counting. Controls were used to
account for free fucose released by the hydrolysis of GDP-fucose.

Using GDP-fucose, (0.01 mM) the concentration of enzyme was
adjusted so that (10% of the GDP-fucose was utilized during the
reaction. Saturating amounts of GDP-fucose were not used in routine
assays during the enzyme purification because the substrate is difficult
to prepare. The Michaelis constant (Km) for GDP-fucose determined
with GDP-Sepharose purified enzyme was 0.07 mM at pH 6.0. The maximal
velocity with the same preparation is 0.10 umoles/min/mg protein.
Standardizations of fucosyltransferase activities were carried out using
0.8 mM GDP-fucose.

Low molecular weight acceptors were used in the assay of
fucosyltransferase activity. An approximate Km of 0.5 mM for the
disaccharides, lactose and Nfacetyllactosamine was determined. The
monosaccharide, Gal-B-hexanolamine, was determined to have an
approximate Km 0f 6 mM.

The macromolecular acceptor asialo-al-glycoprotein, containing an
oligosaccharide core with a terminal Gal8(1-4)GlcNAc-B-sequence is an
excellent acceptor for the fucosyltransferase.

III. Purification of the Porcine g-Galactoside o(1-2)

Fucosyltransferase

The 10,000 fold purification of the 8 galactoside 6(1-2)
fucosyltransferase, by a modification of the procedure developed by T.A.

Beyer and R.L. Hill (personal communication), from 4 kg of porcine

51

submaxillary glands is summarized in Table 1. Due to the small amount
of enzymatic activity present it was necessary to process a large
quantity of starting material. The procedure utilized is designed to
obtain the maximal amount of enzyme sufficiently pure for the synthesis
of fucosylated compounds. The partial purification of the B-galactoside
0(1-2) fucosyltransferase can be accomplished in less than 10 days.

Fucosyltransferase has been shown to be membrane bound except in
secretory individuals where the enzyme is also present in a soluble form
in the secretions of several glands (17).

Two of the initial steps in preparing enyzme in large quantities
entailed removal of the mucin and solubilization of the enzyme from the
membrane. Much of the mucin could be removed by homogenization of the
crude submaxillary glands with water and centrifugation. Isolation of a
membrane fraction can be accomplished by high speed ultracentrifugation
or aggregation of the membrane fraction with manganese chloride.

Induced aggregation of the membrane fraction permits a large volume of
homogenate to be centrifuged at low speed in a Lourdes Beta centrifuge
(106,108). The membrane fraction is obtained as a pellet from which the
supernatant containing soluble proteins can be removed by aspiration.
The membranes are washed a second time and reaggregated by the addition
of manganese chloride.

Several non-ionic detergents are capable of solubilizing the
fucosyltransferase from the membrane. In the presence of 0.94 sodium
chloride; Triton X-100, Brij 99 and octyl-B-D-glucopyranoside are
capable of solubilizing practically all of the enzyme from the membrane.
However, the ionic strength necessary to solubilize all of the enzyme

from the membrane prohibited adsorption to SP-Sephadex. All of the

52

qmc_m H

vcxdmnnmfiso: om «3m wowndzm mcuaoxdddmwk mimm.mnfiomdam pflwiwv manomkafiwmzmmmxmmm

wmmc.am mam mzoz: nos nym camumxmfidoz +103 6 we om mccamxnd.o1k mdmsam.

 

 

mflmu <odcam flowed vwodmaz flowed mnfis<aak mcmnamdn >nfis<nu< <dmda mwmu 40am.
ucxamanmfiao:Twcxdmflnmudo:

ad an czsflmm czdfim\ac a

:oaommamfim News www.mmo Hm.m o.oooom Hoo H H

adzcm acns:

zmacsmzm mwem mm.~me o.m o.oooHN mo N N

mxmoados

Hasno: mxfismON mHoo um.umo o.m o.ooome poo H.» N

mvimmuzoamx HNNN m.wmw N.m o.ooHN em m Hm

mow mausowomm H wmm em m.o o.HwN um HHo Hmmo

mmuzmamx mmo ewe HN m.o o.e- No N.» mNHN

mow mmuzmsomm HH mo e.m N.N o.mm~ NN N.H HHoHN

 

m cam :33” nowwmmuozam no H :aod om uwoacnfl «013mg can 33:. ea mmflcwmadzu

donaomm ma wwo.

noznmzaxmflflozm om moviwcnomm one

53

fucosyltransferase could not be extracted with Triton X-100 in low ionic
strength buffer. Complete solubilization of the fucosyltransferase from
the membrane is achieved by repeated extraction with the detergent in
low ionic strength buffer. The enzyme in this extract is completely
adsorpted on SP-Sephadex. Batch adsorption of the enzyme onto
SP-Sephadex was used since the enzyme is inactivated in Triton X-100 and
unstable on cation exchange resins (T.A. Beyer and R.L. Hill, personal
communication). The fucosyltransferase is easily eluted from the
SP-Sephadex resin with 0.3 M sodium chloride. Dilution of the enyzme at
this stage in purification with 50% glycerol allows the enzyme to be
stored for at least 6 months at -20° with little loss of activity.

The major purification is achieved on the affinity resin
GDP-Sepharose. This resin was synthesized for use in preparation of the
fucosyltransferase based on the success of similar resins in the
purification of N-acetylglucosaminide 3(1-4) galactosyltransferase,
CMPeflfacetylneuraminate: aeflyacetylgalactosaminide a(2-6)
sialyltransferase and [fucosyl o(1-2)] galactoside 6(1-3)
Nyacetylgalactosaminyltransferase (30,31,68,106). The SP-Sephadex
pooled fucosyltransferase was diluted with buffer for faster loading on
the GDP-Sepharose column. No activity could be detected in the elute
from the affinity resin. A linear sodium chloride gradient from 0.5 to
2.0 M eluted many proteins adsorbed through ionic interactions.
Fucosyltranferase activity was eluted with 2.0 M sodium chloride, as
shown in Figure 3, giving a 1600 fold purified enzyme preparation. The

GDP-Sepharose I purified enzyme can be stored for at least 3 months at

4° without loss of enzymatic activity. The enzyme can be stored up to 6

 

54

.52 z o.N 66 In Jpn—.383 FEES z: N .o 3:638 PEI 2: mm 6oz z

o.N .o.m In .mum_xuooeo Ezwuom 25 mN use acm>_om mcwpcmum mm _umz z m.o .o.m In .mpm_xuooeo
5.68 z: a .6 23%; :2: E 8m e .0 £er 2 m5 .oo .3 662.2388 .533. 2..

mm .m .82 z 35 .3 1e 6:588 5.68. 2:. mm 5.5 .< :3 meet... 2: oe eoeio 2e
ass—co H mmocezqmmieoo so mH x m m on nmw_aqm we: mEaNcm .mmocmcqmmuaow seed covp=_m Fonz

xm mmoemwmcmchAmooam ANiHve wuwmouompmuiu ucmpw ago—F_xmea:m mcwucoe one eo :o_umu_mwc:e

.m mcamwd

Hutu/6111) Ulalmd

160

 

 

 

 

 

 

r T T
,J
F. ’fl”¢”’ 8
at”...
~~~~-c>-....____
1.
O o)
.. O .0
o .— 2.5
Z
.. 8 c m
o 5
'4': .2,
o u.
r- 8 Us;
a) -—-
- a
<); «u-
. g? i 1 1 a O
m N —
g o. O. o.
O' O O O

o---o( |u1/siiun) Ail/010v esmeisuoulfisoong

 

56

months by dilution with 50% glycerol and storage at -20°.

The fucosyltransferase is not stable in low ionic strength buffers
and desalting before application to another affinity column or a cation
exchange column must be accomplished quickly. Dialysis of large volumes
is difficult to accomplish technically and results in loss of enzymatic
activity. Desalting was accomplished in 200 ml batches on a 5 X100 cm
Sephadex G50 column at a flow rate of 150 ml/hr. Fractions of 20ml were
collected. The Sephadex G50 chromatography step completely desalted the
fucosyltransferase and resulted in a 3 fold purification, as shown in
Figure 4. The enzyme was found to be extremely unstable in low ionic
strength buffers, loosing 90% of it's activity in a 24 h period.
Concentration of the Sephadex G50 pooled enzyme was achieved by
adsorption to a 1 ml column of GDP-Sepharose, in a plastic syringe, at a
flow rate of 50 ml/h. The column was washed with buffer and the enzyme
eluted with 2 M sodium chloride as shown in Figure 5. Fractions of 1.0
ml were collected and those containing the fucosyltransferase pooled.
The pooled enzyme is stable at 4° for at least 2 months. Storage for
longer periods of time was best accomplished in 50% glycerol and storage
at -20°.

IV Synthesis and Purification of Fucosylated Oligosaccharides

Fucosylation of galactosylated derivatives was performed at 33°
because of the greater stability at this temperature of the
GDP-Sepharose II enzyme.

Reactions were initially performed with a 2-fold excess of acceptor
over the GDP-fucose concentration as described under "Methods". Using

the acceptor, Gal3(1-4)GlcNAc-B-hexanolamine, it is possible to fellow

57

.umuowp_ou
mew: _E oN mo mcowuoecd .o.o :a .mumpxuoomo Ezwuom ze mN :? umumcnw_w:cm ass—co
Amcwev omw xwumgawm so ooH x m a co umppemmu mm; mmeememcecppxmooae ANiavo muwmouompmmim

one .mmmcmemcmcp_xmoo:u ANiHvo mu_mopumpmwiu mg» yo mcqummmo use =o_uouwmwcze

.e mc:m_d

H (3w) eouolonpuoo

 

o

g g 8 8 9 8 o
.__.(|u1/6u1) ugatmd

9. 9.. 8

 

I U

   

r

 

a...
Q)
‘8:
_ 3
2
C Q”
.2 i'.’
3 .8»
U L:
P ________....—o’" L:
VT"—

 

 

 

J I L 1 O
8 92 Q 8 o
o. - -o (mi/spun) thhgiov esoieisuonpisoong

 

59

8.5 2 e 2e 82 z e .o :02 2 ON 6.0
In .mumpzuoomo Ezwuom zE mN .m mo.m In .mpm_»uoumo E:_uom z; mN cup: .< ”AA¢ mzoccm ecu um
665mm; use ass—co mmocmcammieow PE H a op noncomum we: mex~cm uoppmmmu use .asamcmopmsocsu

NH mmocmgqmmiaoo xm mmecmemcmcu_amoo:d ANiHve mnemouumpmw a mcwucoa one yo cowumcucmucou

.m mczm_m

60

H( “Ll/5111) Ulalmd

 

 

 

 

 

in LO Q" N
93. O. O. O. O.
O O O O O o
T T j—'__T T m
. o- 8
if:
.0
i E
:3
+0 Z -
h- ¢ c a)
_____ .0—43 E E;
""""" o 1::
fr»— ————— ——— E
(De-- L‘-
- a
. <1- 5
‘1
O

1 1 1 1 L
(I) (.0 N

91- 9- O. o. 3. O. O

O O O O O O

o.

“cum/spun) lip/111w esmelsuonpisoong

 

61

the kinetics of the reaction and estimate GDP-fucose hydrolysis, as
described under “Methods". Under these conditions, 94% utilization of
GDP-fucose could be achieved with 72% incorporated into product and 22%
hydrolysis to fucose, as shown in Figure 6.

Large scale reactions with 13C-enriched acceptors were
performed using a 1:1.2 ratio of GDP-fucose to acceptor. The reaction
profile, for the synthesis of Fuc6(1-2)[1-13C]Gal8(1-4)GlcNAc
followed by high pressure liquid chromatography, as described in
"Methods", is shown in Figure 7. No unreacted GDP-fucose is detected in
the reaction mixture after 20 h, Figure 7C, the only products observable
are GMP and GDP. The yield of Fuc6(1—2)[1-13C]Gal8(1—4)GlcNAc was
83% based on starting GDP-fucose.

Hydrolysis of GDP-fucose to fucose was approximately 20% in all
reactions, independent of acceptor concentration. The factors
influencing hydrolysis of GDP-fucose were further investigated by
determining the kinetics of hydrolysis in the absence of acceptor and in
the absence of enzyme, as shown in figure 8. The hydrolysis of
GDP-fucose by the enyzme preparation in the absence of acceptor
indicates that the fUcosyltransferase might act as a phosphohydrolase.

Purification of the large scale fucosylation reactions is easily
achieved by deproteinization, deionization and gel filtration. Several
methods for removal of proteins were investigated. Deproteinization
with barium hydroxide and zinc sulfate by the method of Somogyi (107)
gave a white precipitate of barium sulfate and protein which was
difficult to filter and gave in large volumes of filtrate containing the
product. Alternatively, products containing a primary amino group could

be adsorbed to a small Dowex 50 (hydrogen) column and eluted with 1 M

.ooNHHocox;

omooaeioow Ho “cacao ecu mucomocooe :Hmoc H xozoo oco om xoZoa goon m:H=_oucoo csaHoo on»
.2 soc» mcwszo xHH>HHoooHooa .ooNHHoLoH: omooowioow mo “cacao asp oco o:HEoHo=oxo;1mio<onw
,o HeivaHow op ooecoemcocp omooom Ho pesoEo ogp mucomocooe csoHoo H xozoo
ozp Eoce mcwpoHo HHH>HpoooHUom .HH_>_poooHooc co» ooucaoo oco emoogpoz= Loon: ooQHLomoo mo
oops—o moo: mcszHoo :pom .:_moe HooHLonov H xozoo oco Hammocoxgv om xozoa zuoo mchwopcoo
esoHoo HoQHQ Lampoon o op omHHooo poooHHo ocooom ocu oco HooHLoHcov H xozoo Ho can—cu

pooHo cooumoo m on ooHHooo mo; HoaoHHo oco .couoz cup; :oHuoHHo an oouo:_ELop oco om>oeog
mom; muooo_Ho oz» .moeHu poocoemHo H< .cowpoozocw mo moEHH m=HHgo> Ho mcoHuoooc mcwxommo

Ho oocouwcos mo: pooooco Ho :oHpoELow mo own; one .omocoemcocpmeoood HNiHvo oonouooHow

m ocwocoo on» ssz o:HEoHocoxozim1o<zoHwHoiHVmHow eo :oHpoHAmoozd mo mowuocHx on» .m ocomwd

63

o---o 001 x esoong Io uouooig enow

8 8 a 9 a o

 

 

I V W I j

8 IO I5 20 25
Time (h)

6

 

 

 

o o o o 0
Q) (9 v (\l

IOO-

H00] x Ionpon ,to UOTIODH e|ow

Figure 6

64

.HHo>HHuoomoc .:HE «N oco NH.N

mew; new use omoooHiooo .ezu com moswp :oHHcopom .HHo>HHooomoL .coHuonoo:_ Ho ;ON oco H.o
op ocoomoccoo .0 use .m .< mosogm mmHHHoco oer owes“ one .:E=Hoo x<m meoH mxo HHmecoo o
ooee oooooHeH Haow eea .e.m Ia . eoeeee ooaeemeea sewmmaoea z N.o goes He oe op eo_o=_wo
.ocouxHE coHponoocH on“ soc» Ham.H wo Ho>oeoc Ho noxommo mom: muoooHH< .acoocmouoeoccu

oHooHH mcommoco gmHI Hm ooonHod o<onuHoivaHowmumHng we :oHHoHamooom Ho mmcoou oEHH .N mesmHN

0.02

 

0.0l5

QOIO}

Absorbonce (254nm)

g

 

 

 

 

 

 

 

JUN L

onlojzolao'o'lo 20.30 0 IO 20 so
Time(min)

 

 

 

 

Figure 7

 

66

.cE:Hoo HooHLoHcov H xozon o mch: ooeomoos omHo mo; Heiiiov Hommo

:oHuonoocH mg» :H oEHNco wo oocomno on» :H mmeHoco»; omooodiooo .mHmaHoco»; omoooeioow
mo ocomooe o we now: mo: oco HHH>HuoooHooL co+ oopcsoo mo: ouoaHo one .oazHoo HooHLoHcov
H xozoo o ow :oHHooHHooo oco .emooguoze :H ooowcomoo HHmooH>oco .ocopxwe :oHHoooocH

ogu soc» muoooHHo mcH>oeoL an oocomome mew: omocommcoLHmeooae ogu Ho .HOIIJQV copooooo

Ho wocomao asp cw .mHmHHoLox; omoozeieoo wo mopom .mpocumnam eopoooo< c< Ho oucomo<

one cH omocoemcocpmeoozd HNiHVo oonouooHowim och Hm mmooomioaw mo mHmHHocoH: Ho moHHo:_¥

.w meomwd

 

67

 

m mczmwm

2: 22:. cozooom

8
8
-9
oo

   

8

0
v-

0
(D

8

 

 

c3
(0
0le P32901994 Wild-dos) IO Uouomg epw

68

pyridinium acetate. Removal of the pyridinium acetate by rotatory
evaporation at 30° resulted in slightly acidic conditions that led to
some fucose hydrolysis. However, purification of the disaccharide, Gal
8(1-4)GlcNAc-B-hexanolamine from the galactosyltransferase reaction,
with Dowex 50 (hydrogen) chromatography did not result in any detectable
hydrolysis of the product.

The best method for removal of proteins from the fucosylation
reaction was concentration by Amicon filtration. After washing the
concentration cell several times, the product was recovered
quantitatively in the filtrate free of proteins. Deionization and
removal of GMP, GDP and GDP-fucose was achieved by passing the filtrate
through a 1 X 4 cm column containing Dowex 1 (chloride) and Chelex 100
(sodium). The solution was then concentrated to 2 ml and applied to a 4
X 90 cm gel filtration column containing Bio-Gel P2 (-400 mesh)
equilibrated in 0.1 M triethylammonium bicarbonate pH 7.5. A typical
separation profile, for the fucosylated and unreacted acceptor is shown,
in Figure 9, for FuCo(1-2)Gal~8-hexanolamine. A two fbld excess of
acceptor was used in the reaction incubation. The fucose containing
disaccharide eluted first containing [UL-14CJ-L-fucose
glycosidically linked to the galactose residue. The primary amino group
of the hexanolamine, present in both FuCo(1-2)Gal-8-hexanolamine and
Gal-B-hexanolamine, was quantitatively measured using a fluorescent
assay for primary amines (95). Removal of the triethylammonium
bicarbonate buffer by rotatory evaporation yielded the fucosylated
product ready for nmr analysis. Fucosylated compounds synthesized with
the e galactoside a(1-2)fucosyltransferase were; Fuc 0(1-2)

Gal3(1-4)GlcNAc-B-hexanolamine, FUCa(1—2)Gal-3-hexanolamine, FUCO(1-2)

69

.omocoemcocpHHmooze HNiHVo oonouooHomim 6:» Ho omooom mooHinzuioow Eoce omccommcocu

mo; m:_EoHocoxo;1m1Hoo HNiHvo one .oUHLogooomHo ogp :H Hommoco omooowinimueH142H

one .emoocuoze cH ooowcomoo mo .xommo HcoomocooHe o m:_m: cocomoms mo:

zoom—mo ozHEoHocoxo; ox» cw ooocm o:_Eo AcoeHco och .m.N Io .opoconcooHo EzwcoEEonzumHLH
z H.o cH oouoanHwooo csoHou Acmos oooiv No Hooiowm o :o coacomcoo mo; coHuocpHHm

How .cowpocuHHN How an o:_soHocoxozim1How oco o:Ham—ocoxozimiHooHNiHvoood we :oHHocooom

.m ocsde

70

o__ _o (mu 91.7) UO!SS!UJ3 GOUSOSSJOmd

 

 

 

 

 

 

 

 

 

o 0 o
Q Q 8 8 st 8 0
IT I T I 1 r (:>
to
- a

______ -<r

oc": ————
§“°. ~~~~~~~~
q‘ ———————

- _ _ O
. O
N
1 1 1 L g

 

CO In SF 1'0

Humane/9-0: x 114103,.) 014110001909

Fraction Number (1.0 ml)

Figure 9

71

[1-13C]Gal8(1-4)GlcNAc-B-hexanolamine, [1-13CJFUCa(1-2)
[1-13C]Gal8(1-4)GlcNAc-8-hexanolamine, FUCa(1-2)[113C]Gal8
(1-4)GlcNAc, [1-13CJFUCa(1-2)[1-13CJGalB(1-4)GlcNAc, [1-13c]
FUCa(1-2)[1-13C]Gal8(1-4)Glc, [1-13CJFUCa(1-2)[1-13C] Gal-3-
hexanolamine, [1-13CJFUCa(1-2)[1-13C]Gal-B-ethyl, FUCa(1-2)
[1-13CJGal-3-ethyl, [1-13CJFUCa(1-2)Gal-B-ethyl, and
FUCa(1-2)[2-13C]GalB-ethyl.

V Carbon-13 NMR Analysis of Fucosylated Oligosaccharides

 

The enzymatic synthesis of Gal3(l-4)GlcNAc-B-hexanolamine was
carried out using the bovine fl-acetylglucosaminide 8(1-4)
galactosyltransferase. The 15MHz, proton decoupled, 13C-nmr of this
acceptor substrate is shown in Figure 10A. The Cl of GlcNAc, 8
glycosidically linked to hexanolamine, resonates at 102.4 ppm. Figure
108, shows the 15 MHz, proton decoupled, 13C nmr of the disaccharide
GalB(1-4)GlcNAc-B-hexanolamine synthesized by the enzymatic transfer of
galactose to C4' of GlcNAc-B-hexanolamine. The Cl of Gal 8
glycosidically linked to C4' of Glch is found to resonant at 104.3 ppm.
Derivatization of the C4' of GlcNAc results in a downfield shift from
71.6 to 79.9 ppm as discussed by Nunez and Barker (13). Glycosidation
at C4'results in small chemical shifts in C3' (75.5 to 73.9 ppm) and CS'
(77.4 to 76.5 ppm). Unequivocal assignment of carbon resonances in the
dissaccharide were made using [1-13C]Gal3(1-4)GlcNAc-B-hexanolamine
and are given in Table II (13).

Synthesis of quantities of fucosylated compounds sufficient for
13C-nmr analysis was attempted first with the acceptor Gal-

B-hexanolamine. The 15.08 MHz, proton decoupled, 13C-nmr spectrum

72

.:o_puwmc

mmmgmwmcmgu—xmoozw mcu cw Louqmuom cm m? muwgmguummwu ms» .Eaa «.moH an mmucmcommg u<zu~a._u

age .Aeaa m.m~ a“ m.HNV a<zapw.eo co chgm u_avc=zou m =_ mp_:mma .au an _au ;p_z a=_Ea_o=axm;-m
-o<zu_w we :owpmNPpm>_Lmo .m:m_wmpazum5mguwu mo upmwmczou Egg m.¢oH um macmcommg cu venom mw paw Po
mzp .m:FEw_o:mxmzuuno<zoFwAequmpmu .pozvoga umumFAmouum_mw ugh .m .mcwampocmxmzumuu<zupw mumgumnsm
gouamou< mmocmmmcmLHFxmouompmu ugh .< ”co mzzémH .uw_a=oomo copoga .~:z mo.mH ugh .oH mg=m_u

73

OH atsmwa

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

flung

 

 

 

2%.
q. i... sealeeqei
.888 .3 .6 .0
fie.Jszr};:4fixC(J\fj1;;wsxaczgsz)flL<;1r¢Iéesyarxx;ziz1c3:s£¢12¢¢g314111fi1¢1rtsscas
mono no

 

 

 

 

 

m.m~ m.mn H.m~ m.mn H.m~ m.m~ cam
mg: Wow mg: «.2 9.3 fig: «.2 ads mg: P3
H.wm m.- n.m~ H.om m.mn m.m~ o.- N.MN m.Hm u<zupw
m.H~ ¢.HN upw
S
~.HN m.H~ o.H~ m.Hm m.vm o.- can
o.- 9: Q: 9: 0.: 0.: Y: «.3 w.om Foo
m.om H.m~ m.¢~ o.o~ w.mN m.mm m.mn m.mm H.~n o<zopw
m.n~ m.¢~ ope
8
m“ w.mo N.o~ m.mo m.mo ¢.m~ m.mm use
Wt 93 ex: imm Q: m.- mums .93 o4: Eu
m.~m m.mm o.~m H.mm w.om m.om N.~m H.wm m.mm u<zupo
m.m~ N.Mm opw
No
¢.HoH m.HoH m.ooH N.HoH w.~m w.mm can
m.~oH ~.voH m.moH m.¢oH n.HoH m.¢oH o.¢oH o.wm w.mm Foe
m.om H.Nm N.om m.Hm ~.NoH «.mofi ~.NoH «.mm N.Nm o<zo~o
TB 0.8 08
—u

a a Eula 5.5 5.2.... E; 5.; on“. Eu 95

a<za_w a<za_¢_aw _awa=a _aw a<za_o a<zo_w_au a<za_o .xspmz _»;pm am
Zeus“. 0 F82: u u

ampcas=m_mm< “cwgm Paaweazu mH-=oacau

mm wpnnh

.Amfiv va—me Dcm chzz EOLL. cwxmw muwwzm —mo.:=w£u .U

.Aon ._o um me_m3 soc» cmxmu xumwos mcwoanmc on» we pewsm _mu_em;o mcp mmwwwomam m .n

.Amov omm um cowuapom Luna: 5 cw gmsocm mmocmgaaoua—mum
as» mo Pu Lee Egg ¢.~o we mspm> wwwzm _ouweo;u m on mocmgmmmg saw: Eng cw cm>pm men mpmwsm Pmo_sm;o .m

 

 

5
7

 

 

 

¢.~m w.Hm o.- ~.H~ m.HN mewsmpocmxm:
mo Po
m.om
m.m~ m.m~ m.m~ m.m~ m.m~ m.m~ _»»aoa-z co
_acume
oNH aka ~.mua m.m~H N.©~H m.m~H qumumuz do
meoagmu
w.oH ¢.~H m.mH m.oH N.NH ~.~H as;
“.mm m.mc m.mm m.mm o.Ho m.~o m.mo m.~o m.~o Paw
o.Ho o.Ho m.Hc m.Ho m.Ho m.Ho «.mm N.~o o.~o u<zo_w
m.~o m.~o ope
mu
m.wo ~.mo ¢.wo a.~o m.- m.~o uzm
m.m~ m.m~ o.- N.©N m.o~ m.o~ o.on m.m~ o.~m _mo
5.05 w.o~ H.m~ m.HN m.o~ m.om ¢.N~ m.- o.m~ u<zopo
¢.~N o.m~ u_w
mo
mic again ecu Eco ecu Egm uzm Paw m a
u<zu_w u<zu_o_mu _mwo=d Paw u<zupw nw<zupu_mw nw<zu_u _»;umz —»;pm a

meozm

quuam

76

.w:Pampocmxmcumn_momumaufig
op comwgmaeou xn Eng ¢.- ou umcmwmmm mm; Paw to N congmu .Pmo _o m? sag m.¢oH no mucmcommg

a_accczou use .ac_5a_ocaxmz-m-_mw co mzz-omH .ua_a=oaao copoca .NIZ mo.mH ash .HH «czar;

77

 

 

 

 

 

HH acsmwa

SEQ

3;

 

 

 

 

 

 

 

 

.0

78

of Gal-s-hexanolamine is shown in Figure 11. The downfield resonance at
104.3 ppm was assigned to Cl of galactose B-glycosidically linked to
hexanolamine. Carbon 2 of galactose was assigned to 72.4 ppm by compar-
ison to hexanolamine e-D-[1-13C] galactopyranoside. The 90 MHz,

proton decoupled, 13C-nmr of the fucosylated product, FuCa(1-2)Gal-
B-hexanolamine is shown in Figure 12. The two resonances found at 103.5
and 101.3 ppm correspond to Cl of Gal and Fuc, respectively. The Cl'
Gal is shifted upfield as a result of derivatization at C2 with fucose.
Fucose was shown to be a-glycosidically linked to galactose by
comparison with standard methyl-a-L-[1-13C] fuc0pyranoside and by
enzymatic removal of the fucose by an a-L-fucosidase from bovine kidney.
Derivatization at C2 with fucose resulted in a characteristic downfield
shift (from 72.4 to 78.4 ppm) of the C2 galactose resonance. Carbon 3
was found to be shifted upfield as a result of derivatization at CZ with
fucose (from 74.6 to 75.8 ppm). Carbon-1 was also shifted upfield
(104.3 to 103.5 ppm). Complete assignment of the chemical shifts are
given in Table II.

The antigenic portion of the blood group 0 substance was
synthesized by enzymatic fucosylation of the disaccharide Gal5(1-4)
GlcNAc-B-hexanolamine. The 90 MHz, proton decoupled, natural abundance
13C-nmr spectrum of the antigenic trisaccharide, FUCa(1-2)Gals(1-4)
GlcNAc-s-hexanolamine is shown in Figure (13). Resonances marked, X,
are due to the disaccharide acceptor Gal3(1-4)GlcNAc-B-hexanolamine, an
impurity in the sample. The spectrum can be divided into three regions:
a glycosidic region from 105 to 95 ppm containing Cl of the
carbohydrate; a second region from 85 to 55 ppm containing C2 through C6

of the carbohydrates and C1 of the hexanolamine; and a third region from

79

.auwgagew co m? x mucocommg
mg» .uzm guw: cowumNPum>wgmu mo upsmmg a mo Agaa v.w~ op «.mn eogmv cpm_w=zou umuwwcm m? me.~u
use .x_m>_uumammg .ozm ucm —mw we _u on ucoammggoo sag m.HoH vac m.moH an ucaoe mmucmcommg o3» mch

.m:_5aPo=axaz-m-a<zu_wA¢-Hvu_auA~-Hva use we mzz-umH .ua_a=ooao couoea .sz om ash .NH me=m_u

NH menace

Ema ON 0m 01 cm 00 on om

b-P—PP~PPP~Pr_Frpb—rprP-Prbpbhpbbp—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«U

 

 

om OOH

pbpp—kb-pPPPp

 

 

81

.HH mpamh cw cm>wm mgm mmucmcommg mg» we

“cchm_mmm mumeEoo .u<zopo $0 aaogm Pxpmomum.mcu cw :oagmu _a;pme mg» use mmooay we no .mcwsmpocmxm;
mzp yo no cmaogcp Nu mcwcwmpcoo sag mH o» om sexy :oPomL ugwgu a van mchEwFocmxm; mg» yo

—u ecu mmumgua;oagmu any we no cmzoczp No mcwcwmucou sag mm on mm ace» zowmmg ucoumm m ”mmumgu»:oncmu

mgu mo Pu mcwc_mp:oo Eng mm op moH soc» cowmmg owvwmoua—m o "meowmmc mmggu opcw cmu_>_u

ma :mo Eaguomam ugh .m—QEMm mg» cw xuwgaaew cm .mc_Em_ocmxm:umuo<zuFoaaufivmPaw Louqmuum mvwgmzoummwu

an» op mzu mew .x .umxgms mmucmcommm .m:VEo_ocmxmzumuo<zoPwfieiavm_woflmufivaozm .mv_gm;oummwgp

owcmmwuc< ash we Esguumam azzuomH mocmuczn< ngsuoz .umpqaouwo :ouoga .sz om mgp .m~ mg:m_u

3?.

tan. om

pp—

 

 

 

 

»

P

P

P

Or

om

—rpp»—

 

-

r

P

 

P

mH weaned

om

FFLP

 

 

-

Oh
pep

 

 

 

 

 

 

 

 

P

 

p

on
_

P

P

om

PTTPIPPP

OO~
. _

 

P

 

 

P

 

P)

 

83

50 to 15 ppm containing C2 through C6 of the hexanolamine, C6 of fucose
and the methyl carbon in the gracetyl group of GlcNAc. Complete
assignment of the resonances are given in Table II. Unequivocal
resonance assignments were made using methyl-a-L-[1-13C]
fucopyranoside, ethyl-B-D-[1-13C] galactopyranoside and Fuc a(1-2)
[1-13CJGal8(1-4)GlcNAc-B-hexanolamine. An expanded 90 MHz, proton
decoupled, 13C-nmr spectrum (Figure 14) of the anomeric region of
Fuc a(1-2)Gal3(1-4)GlcNAc-3-hexanolamine reveals chemical shift
differences in the anomeric carbons as a result of derivatization with
Fuc at C2. Resonances are labeled in accordance with IUPAC nomenclature
with Fuc, Gal and GlcNAc being represented by C, C' and C",
respectively. Carbons Cld and Cl'd correspond to the contaminating
disaccharide GalB(1-4)GlcNAc-B-hexanolamine. Derivatization of C2' with
Fuc results in a 2.6 Hz upfield chemical shift of Cl' (from 104.3 to
101.1 ppm). The Cl" of GlcNAc is found to shift downfield by 0.11 ppm
as a result of fucosylation. Carbon-1 of the a-L-fuc050pyranosyl residue
resonates at 100.9 ppm.

Th 45 MHz, proton decoupled 13C-nmr of the trisaccharide Fuc
a(1-2)[1-13C]Gals(1-4)GlcNAc-8-hexanolamine is shown in Figure
15. The carbon 1 of Gal is enriched with 13C to the 90% level
allowing unequivocal assignment of the Gal resonances in the
trisaccharide. The anomeric region shows two carbon-13 enriched
resonances. The enriched resonance at 104.3 ppm is the Cl' of Gal in
the contaminating disaccharide [1-13CJGalB(1-4)GlcNAc-
B-hexanolamine. Natural abundance carbons of the disaccharide are not
observable. The carbon-13 enriched resonance at 101.7 ppm is the Cl' of

Gal in the trisaccharide.

34

.mcwsmpocmxmcumuu<zupw
Aeiflvm_mw mu_cm;uommwu mg» op ucoqmmcgoo u._u ucm UFQ mcongmo .a_m>wuumammg .u<zo_w
new —mw .uam “cummgamg ._u can .0 .o cm_mam_ mmucmcommm .m:_smPocmxmzumuo<zo—ufieuavmpmwANnHvauzu

we :opmmm owLmEoc< och we sapwomnm mzzuumH .vm—aaouwo :opoga .sz om cmucqum c< .¢H mg=m_u

85

ea aczmwu

21¢ mm OOH HOH NOH moy Tog mofl mo~ Rog

r——-P_.bF—rppb—P-bPP——PPPpF——PP_—P—PPPP_)LP_Pbl

33%;}

wpnv

 

 

 

 

 

vnu

.nv

 

 

 

.nv

86

.me mm

on Poo mumflufiu seem ANI mev m:_—a:oo ucoaumco mmLMF ms“ Eocm umcmwmmm a—quo>_:cmc: m_ Paw Nu

och .w_nm>gmmao uoc mew muwgmgoummwu use we mcongmo oucmucznm chsgmz .m:Pam_o=mxm;-muu<zupoflvuavm
_mw mumfiuflg mu_gmsuomm_u m:_pm:w5mucoo mg» cw Paw we ._u mgu my sag m.¢o~ um monocommg

um50wgcm ms» .umH saw: pm>mp wom mg» op umsowgcm m? Paw Po mgp .m:Pam—ocmxmzuuuu<zopreuHvuPaw
Humfi-flu A~-Hvaa=a au_ea;ooamvep age co mzz-umH .ua_a=oaao couoca .sz we ask .mH mgamca

mH mesa?“

 

 

 

 

 

 

 

 

 

 

 

 

.C .1
_ E

87

 

 

 

88

An expanded region from c2 to ca of Fuc [1-13CJGalB(1-4)
GlcNAc-B-hexanolamine is shown in Figure 16. Derivatization of Gal at
C2 with fucose results in a downfield shift of C2' (from 72.3 to 77.9
ppm). The large one-bond coupling (46 Hz) from [1-13C] Gal to C2
makes assignment of the C2 resonance unequivocal. The C4" of GlcNAc
derivatized with Gal is clearly observed at 70.6 ppm. Carbon-3 of Gal
is easily assigned at 75.0 ppm by a 2a C1-C3 Coupling of 3.5
Hz, shown in Figure 16. Carbon 2 of GlcNAc is found to resonate upfield
at 56.8 ppm as a result of derivatization with an fl-acetyl group.

Carbon 6 of Gal was assigned to 76.9 ppm as a result of a broadening due
to a small 3a C1-C6 interaction. The C6" of GlcNAc was

assigned to 61.5 ppm. The 6-deoxy carbon of fucose is found to resonate
upfield at 16.3 ppm.

The reducing trisaccharide FuCa(1-2)[1-13C]Gal3(1-4)GlcNAc was
enzymatically synthesized by the sequential addition of [1-130] Gal
and Fuc, from UDP-[1-13CJ-Gal and GDP-Fuc, by the
galactosyltransferase and fucosyltransferase, respectively. Chemical
shift assignments are reported in Table II and were determined as
described for FuCa(1-2)[1-13C]Galp(1-4)GlcNAc-B-hexanolamine. The
ratio of a:B GlcNAc in the trisaccharide was found to be 2:1. The
chemical shifts of the [1-130] enriched galactose in FUCa(1-2)

[1-13c] Gals(1-4)GlcNAc were found to differ (102.32 to 102.28)
depending on the anomeric configuration at Cl of GlcNAc, a or 8. Fuc
6(1-2)[1-13C]Gal8(1-4)Glc was also found to have an azp ratio of 2:1
and differing chemical shifts of the [1-13C] Gal depending on the
anomeric configuration of Glc.

VI Conformational Analysis of the Fucosylated Oligosaccharides

89

.m:_Ee_o=aXdz-m-e<ze_eA¢-HVm_aw numH-Hu
AN-HVad=d eo mzz-umH .ed_a=oaeo capoea .sz om age eo mo oe we seem edema“ edueeaxm =< .eH de=m_e

9O

 

NU

55

00

 

cu

\s

cu

¢H ee=m_d

no

 

mu

 

 

NU.

 

xuzpu
flu

 

 

nu

 

.vu

ms

nu

On

 

NU

swu

 

 

mu
mu

\\

91

Specifically enriched carbon-13 carbohydrates can be used to
evaluate the solution conformation of oligosaccharides by measurement of
interresidue spin-spin coupling (3d) between the 90% enriched 13C
nucleus of one residue and an appropriatively positioned 13C or 'H
nucleus of the adjacent residue. Conformation about the glycosidic bond
can be defined by two torison (dihedral) angles, w (Cl-0-C2'-C1') and 0
(C2'-0-Cl-Hl) as shown in Figure 17. The dihedral angles can be
evaluated by measurement of the inter-residue coupling between
carbohydrates.

The trisaccharide Fuc [1-13C] Gal8(1-4)GlcNAc-B-hexanolamine
was initially examined for interresidue spin-spin coupling between
[1-13C]Cl Gal and Cl Fuc. A partial spectrum of the trisaccharide
FUCa(1-2)[1-13C]Galp(1-4)GlcNAc-B-hexanolamine is shown in Figure

18. Inter-residue 30 coupling between [1-13C]Cl Gal and Cl Fuc

C-C
is small or absent since a coupling constant could not be observed at Cl
Fuc, as shown in Figure 18. The linewidth of ClFuc is broad in
comparison with Cl of GlcNAc or any single resonance in the spectrum.
Repeated attempts to resolve the broadening failed. The broadening is
estimated to reflect a coupling of no more than 2H2. To increase the
sensitivity in the measurement of the small 3a CC coupling between
[1-13C] Cl Gal and Cl Fuc the trisaccharide was synthesized with 90%
enriched 13C at both ClGal and ClFuc. Again, no observable

splitting pattern could be detected, however, inter-residue interactions
could be detected by increased linebroadening in the two 13C

enriched resonances. Linebroadening was measured by comparison of the
13C enriched linewidths in [1-13C] FUCa(1-2)[1-13C]Gals-
(1-4)GlcNAc-p-hexanolamine to Fuc«(1-2)[1-13C] Gal3(1-4)GlcNAc-B-

92

58-39:: a 2; A.S-.Séév x. .835 22.853

comweoh 03h am umcwemo mm :mu —mwAN-Hvao=g cw ucom u_u_moux_c an» uaon< =o_meLom=ou mg»

.NH deamee

93

 

Figure 17

94

.a_m>_pomnmme ...Fu use .—u ._o cmpmam_ wee
o<zu_u use _mo .03; we H mconeeu .mewEmpocwxwzrmnu<zopuflwrfivm_eomumauflgamrflvdu:m mu_gmcoummwgh
age eo =o_mem deemeo=< ace eo Eaeeaeam mzz-omH .Udpazoamo couoea .sz om use .mH weaned

I, I

Cl Cl

 

Cl

 

 

 

163 ' 162 161 :60
PPM ,

Figure 18

96

hexanolamine and [1-13C]GalB(1-4)GlcNAc-s-hexanolamine using dioxane
as an internal standard. Linewidths of the 13C enriched resonances
in [1-13C]Fuc6(1-2)[1-13C]Gal3(1-4)GlcNAc-p-hexanolamine were
approximately 3 times greater than the linewidth of the 13C enriched
resonance in [1-13CJGalB(1-4)GlcNAc-p-hexanolamine as shown in
Figure 19. Although a coupling constant
(30 Cl-0-C2'-Cl') cannot be measured for the
torsion angle, a small unresolved coupling of 1 to 1.5 Hz must be

occuring to account for the observable increased linebroadening.
Several doubly enriched 13C di and trisaccharides, [1-13CJFUCa
(1-2) [1-13c] Gal-B—ethyl, [1-13CJFUCa(1-2) [1-13c]
Gal8(1-4)GlcNAc. [1-13c] Fucd(1-2)[1-13c]Ga1s(1-4)Glc and
[1-13C] Fuc a(1-2) [1-13C] Gal-B-hexanolamine were prepared for
evaluation of the w torsion angle by 3d 13C-13C coupling
constants. In no case was appreciable inter-residue coupling observed,
but increased linebroadening of approximately 3 Hz was observed for all
doubly enriched oligosaccharides when compared to the singly enriched
analog.

Evaluation of the 0 torsion angle was attempted using proton nmr by
inter-residue 3J13C-‘H coupling. Inter-residue coupling involving
sp3 carbons contained in the sequence 13C-0-C-'H have been shown
to obey a "Karplus type" of relationship (125). The model compound
FUCa(1-2)[2-13CJGal-B-ethyl was used to evaluate the 0 torsion angle
about the Fucd(1-2)Gal glycosidic bond.

Assignment of resonances in the proton nmr of compounds as complex

as disaccharides can be extremely difficult since coupling is not always

97

.mcexo_o .0 van mm:_a

um_ocestrmru<zo_uA¢-HVm~mu~umfirfiuaNufivduzdmumfiufiu cw mwucmcomme umzowecmuomH ugh .m ”were
-mpocmxmsrmru<zopofierfivm_wwmumflrfiufimrfivau=m cw mocecomme umcuwecwromfi ms» .< .m:_sm_o=mxmz
-m-d<za_ofie-fivm_mwmumH-HuA~-Hveazdmumfi-fiu ecu a:_Ee_oeaxaz-m-a<ze_eAe-Hve_aemumH-HuAN-Hva

use cw mmocecomom umguPecmrumH mcp eo mucmemgammmz cpuwzmcpA my: narcongmu .mH we:m_d

98

n1n.— I.

 

mH deemeu

at. 06'.

a,

 

 

 

 

 

 

.a‘mVn

T

99

first order. One exception are the anomeric protons on the
carbohydrates which resonate 1 to 2 ppm downfield from H2 through H6.
Anomeric protons are usually observable as doublets as a result of a

3JH1-H2 coupling. The HlFuc resonance in FUCa(1-2)Gal-B-ethyl was
found to resonate as a doublet at 5.04 ppm downfield of
tetramethylsilane. A 3d H1-H2 Coupling of 3.4 Hz was

measured for FuCa(1-2)Gal-3-ethyl, shown in Figure 20A. The 0 torsion
angle was estimated using 3a H1-13C2 Coupling in FUCa(1-2)
[2-13C]Gal-e-ethyl. Figure 218 shows the H1 resonance of Fuc in
Fucd(1-2) [2-13C]Gal-3-ethyl. The multiplet was interpreted as an

ABX Splitting pattern with a 3d H1-H2 of 3.4 Hz and a

3JH1-13C2' of approximately 3.2 Hz. Computer simulation

would be needed to further refine the coupling constants. The 3d
H1-13C2 coupling of approximately 3.2 Hz can be used to evaluate the

0 torsion angle (the position of C2'Gal with respect to the Fuc ring) in

the Fucd(1-2)Gal glycosidic linkage.

100

..xgem-e-PewmomH-~HAN-Hva and

.m eea ._»gem-m-_eeAN-Hv-e 6:1 .< ”CH deeeeomdm use .1 age meezosm mzz eoeoea NI: omH one .ON meamee

101

cm ee=m_e

 

DISCUSSSION

1. Preparation of B-Galactoside 6(172) Fucosyltransferase

Enzyme mediated synthesis of branched chain complex
oligosaccharides proceeds with high yield and anomeric specificity in
contrast to chemical methods for the synthesis of glycosidic linkages.

The chemical preparation of B-linked glycosides are usually carried
out using the classical Koenigs-Knorr reaction (109) or a modification
of this method. Such methods involve reaction of a blocked halogenated
sugar with the agylcon (bearing a hydroxyl group) in the presence of
silver or mercury salts to produce a B-linked glycosides.

Recently, much effort has been devoted to the stereoselective
preparation of a-linked disaccharides in good yields (110). Much of
this interest was due to the finding that many of the immunodominant
sugars in glycoproteins and glycolipids are d-linked in these struc-
tures. The major breakthrough in this area was the so-called a-halide
ion catalyzed glycosidation reaction developed by Lemieux and co-workers
(4,5,111). Although this method does allow synthesis of a-fucosides,
some limitations exist. Yields are low, complex blocking strategies are
required, and extensive purification of the product is necessary before
and after removal of the blocking groups. The number of readily
obtainable derivatives is limited, most importantly those for blocking
the reducing moiety of the di or trisaccharide. A second method for
a-glycosidations has recently been described which involves the
synthesis of sugar imidates and their alcoholysis in the presence of

p-toluenesulfonic acid to obtain a-linked disaccharides (112). This

102

103

method also suffers from some of the same limitations as mentioned
above.

The advent of carbon-13 nmr Spectroscopy in the elucidation of
structures and conformations of oligosaccharides, antigenic determinants
and glycoside-protein complexes and the recent development of a general
method for synthesis of carbon-13 enriched carbohydrates prompted
investigation into a versatile synthetic method for synthesis of complex
biologically active oligosaccharides. Enzymatic synthesis has been
proven to be a useful method to obtain specificity and high yield. The
enzyme mediated synthesis of the blood group 0 antigenic substance was
performed using a Nfacetylglucosaminide 3(1-4) galactosyltransferase and
a B-galactoside 0(1-2) fucosyltransferase. The Nfacetylglucosaminide
8(1-4) galactosyltransferase has been purified to homogenity by Barker
§t_gl, (15) and has recently been used to synthesize carbon-13 enriched
derivatives of Gale(1—4)GlcNAc by Nunez and Barker (13). The product
disaccharide, Gal8(1-4)GlcNAc, is the building block on which the
antigenic sugars of the A80 blood group system are transferred by
specific glycosyltransferases.

Synthesis of the terminal antigenic portion of the blood group 0
substance required the purification of a B-galactoside 0(1-2)
fucosyltransferase capable of transferring fucose from GDP—Fuc in an
6(1-2) glycosidic linkage to a terminal galactose residue.

Several tissue sources were examined for fucosyltransferase
activity. Porcine submaxillary glands were found to have good enzymatic
activity, were easily obtainable and therefore were chosen as the tissue

from which to purify the enzyme.

104

The two major criteria of the purification procedure were; to
obtain a partially purified enzyme preparation that easily allowed
synthesis of 0.1 mmole quantities of fucosylated compounds and to obtain
the enzyme preparation in the highest yield since concentrations of the
enzyme in the tissue were extremely low.

The s-galactoside a(1-2) fucosyltransferase had been previously re-
ported in several tissues, shown to be membrane bound and purified 70
fold (57,58). Recently, the enzyme has been purified to homogenity from
porcine submaxillary glands (T.A. Beyer and R.L. Hill, personal
communication). A purification procedure was developed that involved
removal of mucin from the glands, manganese chloride aggregation of the
membranes, solubilization of the enzyme with 1% Triton X-100, adsorption
to SP-Sephadex, purification on a GDP-Sepharose affinity adsorbant,
desalting on Sephadex G50 and finally concentration on GDP-Sepharose.
This approach allowed partial purification of the enzyme on a large
scale. Aggregation of the membranes with manganese chloride was chosen
to alleviate the necessity of high speed centrifugation of large
volumes. Extraction with 1% Triton X-100 in 0.5 M sodium chloride
quantitatively solubilized the enzyme from the membrane. However, the
ionic strength prohibited adsorbing the enzyme to SP-Sephadex. The
enzyme could be adsorbed batchwise to GDP-Sepharose and eluted with 2.0
M sodium chloride but the purification achieved was small and the high
ionic strength needed to elute the enzyme presented problems in further
purification. Therefore, the enzyme was extracted twice from the
precipitated membranes with 1% Triton X-100 in low ionic strength
buffer, adsorbed batch-wise to SP-Sephadex and eluted with 0.3 M sodium

chloride. The enzyme, at this stage could be stored in 50% glycerol

105

at -20° for long periods of time. The enzyme was eluted from the
SP-Sephadex resin in the absence of Triton X-100 since the enzyme was
found to rapidly lose enzymatic activity in the presence of this
detergent (T. Beyer and R.L. Hill, personal communication).

Large scale purification of the enzyme was achieved on the affinity
adsorbant GDP-Sepharose. After adsorption of the enzyme to the column,
the column was washed with a 0.5 M to 2.0 M sodium chloride gradient.
The enzyme was eluted with 2.0 M sodium chloride giving a 1600 fold
purification. The next step was to concentrate the enzyme so that it
could readily be used in large scale synthesis of fucosylated
oligosaccharides. To achieve this the fucosyltransferase was desalted
on a Sephadex 050 column, adsorbed on a 1 ml GDP-Sepharose column and
eluted with a samll volume of 2.0 M sodium chloride. The B-galactoside
a(1-2) fucosyltransferase at this stage had been purified 11,000 fold,
had a Specific activity of 0.88 units/mg protein, and was used in all
syntheses of fucosylated compounds reported here.

Increased purification could be achieved by specific elution of the
second GDP-Sepharose column with 5 mM GMP and adsorption onto a Gal
B(1-4)GlcNAc-Sepharose affinity column. The enzyme was eluted from the
disaccharide affinity adsorbant by removal of the GMP. Enzyme prepared
in this manner was over a 100,000 fold purified but the overall yield
was only 5%.

A Similar procedure was utilized by T.A. Beyer and R.L. Hill
(personal communication) in the initial steps in the purification to
homogenity of the B-galactoside 0(1-2) fucosyltransferase.

11. Preparation of Fucosylated Oligosaccharides

106

The use of partially purified enzyme preparations of the
B-galactoside C(l-Z) fucosyltransferase for the synthesis of fucosylated
oligosaccharides proceeds with anomeric specificity, high yields, and
ease of purification, as shown in "Results".

The preparation of fucosylated di and trisaccharides was monitored
by radioactivity and HPLC. Yields of the fucosylated products were
typically 70-85% based on GDP-fucose. Hydrolysis of GDP-fucose was
always 10-20%. The complete enzymatic hydrolysis of GDP-fucose in the
absence of acceptor indicates that the B-galactoside a(1-2)
fucosyltransferase may also act as a phosphohydrolase by transfering
fucose from GDP-fucose to water. Recently, this observation was
confirmed by T. Beyer and R. Hill (personal communication) who have now
succeeded in purifying this enzyme to homogenity.

Purification of the fucosylated products was rapidly achieved by
ultrafiltration, passage over a small Dowex 1 (chloride) and Chelex-100
(sodium) column and gel filtration. Yields from the purification
procedure were routinely greater than 90%.

III. Carbon-13 NMR of Fucosylated Oligosaccharides

 

The assignment of resonances in the carbon-13 nmr spectra of
FUCa(1-2)Gal-B-hexanolamine, Fucd(1-2)GalB(1-4)GlcNAc-p-hexanolamine,
FUCa(1-2)[1-13C]Gal3(1-4) GlcNAc-B-hexanolamine,
FuCa(1-2)[1-13C]Gal8(1-4)GlcNAc and FUCa(1-2)Gal have been
accomplished using selective 13C-enrichment and 13C-enriched
model compounds (Table II).

Roberts and his co-workers (114) have shown that methylation of a
hydroxyl group results in an 8-11 ppm downfield chemical shift in the a

carbon atom and B-carbon chemical shifts of less than 1 ppm. Chemical

107

shift changes in resonances more remote are usually less than 0.3 ppm
(114). The substituent effects were interpreted in terms of a steric
hindrance or a proximity effect as was observed in inositols (113).
Chemical shift differences can arise from neighbor-anisotropy effects,
currents induced in regions of a molecule that are relatively remote
from the carbon atom of interest, electrogenativity effects, and steric
hinderance or proximity effects (12). The structure of several
disaccharides and trisaccharides have been determined by 13C-nmr

using substituent effects and confirmed by biological degradation
experiments (13,87,115,116).

In the synthesis of the disaccharide GalB(1-4)GlcNAc-B-hexanolamine
it was observed that glycosidation of galactose to C4' of GlcNAc
resulted in a downfield shift of C4' (8.3 ppm). Substituent effects 8
to C4' resulted in upfield chemical shifts of 1.6 ppm for C3' and 0.9
ppm for C5' as predicted by Dorman and Roberts (114). Other nuclei were
Shifted upfield also (Cl', 0.3ppm; C2', 0.7 ppm and C6', 0.9 ppm).

Fucosylation of the disaccharide, gave the antigenic trisaccharide
FUCa(1—2)GalB(I-4)GlcNAc-B-hexanolamine. Derivatization of C2' Gal with
fucose was unequivocally shown when [1-13c]GalB(1-4)GlcNAc-
B-hexanolamine was used as an acceptor. The large '0 C1-C2
coupling of 46 Hz permitted easy assignment of this resonance. The C2'
Gal was shifted downfield, deshielded, by 5.6 ppm (from 72.3 to 77.9
ppm). Substituent effects to the B-carbons, Cl' and C3' of Gal, were
also readily observed. The C3' Gal could be unambiguously assigned by a
3d c1-c3 coupling of 3.5 Hz. The C3' Gal was shifted

downfield as a result of derivatization at C2. The Cl' Gal was shifted

108

downfield as a result of derivatization at C2. The Cl' Gal was shifted
upfield by 2.6 ppm (from 104.3 to 101.7 ppm).

Derivatization of C2' with fucose also resulted in a 0.11 ppm shift
in C" GlcNAc, as shown in Figure 15. This small chemical shift could
result from steric effects or a small conformational change in the
disaccharide as a result of derivatization with fucose.

Analysis of the reducing trisaccharide, FUCa(1-2)[1-13c]Gal
3(1-4)GlcNAc, was complicated by the presence of both a and B anomers of
GlcNAc. Complete assignment of the spectrum was again aided by
selective 13c-enrichment and comparison to the disaccharide
[1-13c] GalB(1-4)GlcNAc. The Cl' Gal was found to differ (102.32 to
102.28 ppm) depending on the anomeric configuration at Cl of GlcNAc, a
or B. The pr0porti0n of 0:8 anomers in the reducing trisaccharide was
found to be 2:1.

Enzyme mediated synthesis of the antigenic portion of the blood
group 0 substance allowed quantities of natural abundance and
13C-enriched compounds to be obtained for nmr evaluation. Carbon-13
enrichment, in many cases, allowed unambiguous assignment of resonances
in these compounds. Correct chemical shift assignments are imparative
before the complex carbohydrate can be used for structural analysis or
dynamic studies involving molecular interactions. At least one study on
the interaction of a lectin with various glycosides has already been
reported based on incorrect chemical shift assignments (11).

IV Conformation of the Fucosylated Oligosaccharides

 

The use of selectively enriched 13C di and trisaccharides allow

BJC-C and 3JC-H coupling constants about the glycosidic bond to be

evaluated. The abundance of glycosidic structures in nature and the

109

increasing importance of these structures in glycoproteins and
glycolipids make a detailed study of Q-glycosidic linkages of great
importance.

The conformation about a glycosidic bond with an aglycon more
complex than a methyl group requires the specification of two torsional
linkages, m and 0, as Shown in Figures 17 and 21 for the FUCa(1-2)Gal
glycosidic linkage. Several studies have already attempted to define
the conformation about the glycosidic linkage using primarily 3d
C-O-C-H coupling. and 30 0.0.0.0 Coupling (13, 117, 188, 199).

Lemieux and Koto (119) developed the concept of exo-anomeric effect
for the particular stability of certain conformations of the 0 torsion
angle in glycosidic linkages. The exo-anomeric effect is proposed to be
due to the extra stability provided when the lone pair electrons on the
ring oxygen are antjfperiplanar to the glycosidic C-O bond (119). Two
conformations about the 0 torsion angle allow for an anti-periplanar
orientation of the p-orbital of 01 with the 05-Cl bond. One
conformation has the aglycon in a syn-clinal orientation to Hl and 05
and is expected to be more energetically favorable than the orientation
in which the aglycon is syngclinal to 05 and C2. Using a series of
compounds 13C-enriched at the aglyconic position and a modified
13C to 'H Karplus relationship a 0 torsional angle was found in
agreement with the exo-anomeric effect (119). X-Ray crystallographic
data, where available, confirmed the torsional angles obtained by nmr.

The p torsion angle in peracetylated disaccharides labeled with
13c at the Cl position having a(1-3), 3(1-3), a(1-4) and 3(1—4)

glycosidic linkages were evaluated using 3013C 0 Cn Hn

110

cu pumammg cpwz

wch uzon< .s

.x_m>wpomamme .m:_e use one ou pomamme new: Pmo No mo cowuwmoa on» wee ace; Poo msu
one _u yo cowp_moq mcu waveommu mmrmcm a use i one .Pmoflmifivoozm cH ecom owuwmoox—u

vc< s .mmpm:< A_meum;wuv chovmeop oz» 05H mcenweommo m:o_uumwoea cmszmz .HN mesm_m

111

‘3' H2

 

 

Figure 21

112

coupling (117). Coupling constants in the range of 3.7 to 5.5 Hz were

observed in all cases (117). The 5.5 Hz coupling constant in

B-cellulose octaacetate was related to the 15.7° angle determined by

X-ray crystallography and the conclusion drawn that the urconformation

was the same in solution as in the crystalline state (117). The value

of the J3 coupling constant was noted to be remarkabely constant

regardless of the (1-3) or (1-4) linkage in these disaccharides (117).
Nunez and Barker (13) were the first to use 3d C-C

inter—residue coupling to evaluate the w torsion angle in the glycosidic

linkage, Gal B(1-4)GlcNAc-B-hexanolamine. Carbon-13 enrichment to the

90% level at Cl allowed evaluation of 3a coupling from CT to C3' and

Cl to C5'. A coupling constant could not be observed to either C3' or

C5' of the GlcNAc residue (13). However, C5' GlcNAc was shown to be

broadened by > 1 Hz with respect to other singlets in the spectrum (13).

This was interpreted as indicating a 3a C1-C5.5 1.5 Hz and a

3J C1-C3' = 0 Hz (13). Since a series of model compounds

with defined m torsion angles for an observed coupling constant are not

available the value of the coupling constant was compared to intra ring

3J(>{;coupling constants and dihedral angles in monosaccharides.

Walker §£.El- (68) observed a 3d C1-C6 intra ring coupling

of 3.7 Hz in hexopyranosides in which C1 and C6 are trans. A 30

C1-C4 coupling, corresponding to a dihedral angle in which C1

and C4 are gauche, was not observed in any of the aldopyranoses (68). A

3d C1-C6 of 0 to 2.6 Hz (110 to 140°) for pentofuranoses,

pentofuranosides and 1,4 aldonolactones has also been observed by Nunez,

(personal communication). Using this data to establish limits for

113

31>4: coupling in carbohydrates a conformer which favored C1 Gal at
an angle slightly greater than 120° with respect to C5' GlcNAc and
slightly less than 120° with respect to C3' GlcNAc was postulated (13).
This conformation, shown in Figure 22, was also proposed by X-ray
crystallography (121) and theoretical calculations (122) for lactose and
cellobiose which have the same 3(1-4) glycosidic linkage as
Gal3(1-4)GlcNAc-B-hexanolamine (13).

To obtain an approximation of the solution value for the 0 torsion
angle, Nunez and Barker (13), relied on 2d C1-0-C4'
coupling constants through the glycosidic linkage. Geminal,
ZJC-O-C , coupling through the ring oxygen was found to be influenced
by the configuration of the C, 0 and H substituents on the carbon
involved in the coupling (123). Both a ring oxygen and a ring carbon
are considered to contribute equally to the coupling. An oxygen or
carbon gauche to the coupled carbon is considered to make a negative
contribution to the coupling whereas an oxygen or carbon anti_to the
coupled carbon makes a positive contribution to the coupling (68).
Coupling, 20 1301-04'» from [1-13c] Gal to C4' GlcNAc
was not observed in the disaccharide Gal3(1-4)GlcNAc-3-hexanolamine
(13). However, a linebroadening of approximately 1 Hz was observed
indicating an interaction between 13Cl Gal and C4' GlcNAc (13). Two
possible 0 torision angles would result in a small
20 C1-C4' coupling constant. The rotamer that was favored
had C4' bisecting 05 and HlGal again in agreement with the exo-anomeric
effect (119) and X-ray data on lactose and cellobiose (121).

Fucosylation of [1-1301 Gal3(1-4)GlcNAc-B-hexanolamine allowed
the conformations about the angles 0 and 0 in the Gal3(1-4)GlcNAc

114

.3:
mcEmPocmxmzrmio<zupu3iimEu 5 mafia: 338868 9: Lo“. 035‘ c328 .5 2580.5 9: .NN 85m:

115

CE C5

 

CI
H4

Figure 22

116

glycosidic linkage and the m torsion angle in the Fuc6(1-2)Gal
glycosidic linkage to be evaluated. The 90 MHz, proton decoupled,
13C-nmr completely resolved all resonances except C5" GlcNAc and

C5' Gal as shown in Figure 16. The C3" GlcNAc was not found to be
broadened by or coupled to the 13ClGal. Broadening of the

C5"GlcNAc could not be determined since it was not completely resolved
from C5'Gal. However, the lack of broadening of the C3"GlcNAc is in
agreement with the conformation of the m torsion angle postulated in
Gal8(1-4)GlcNAc-B-hexanolamine since an appreciable change in coupling
to C5" GlcNAc would simultaneous cause an observable change in coupling
to C3" GlcNAc. Thus, it appears likely that fucosylation of the
disaccharide does not appreciably affect the m torsion angle in the
Gals(1-4)GlcNAc glycosidic linkage. An alternative conformation could
exist in which Cl'Gal is gauche to both C3" and C5" GlcNAc Since
linebroadening from Cl'Gal to C5"GlcNAc can not be resolved. This
conformation is expected to be particularly unstable since HlGal would
sterically interact with H3" and H5" of GlcNAc. Also, hydrogen
bonding between 0H3" of GlcNAc and 05'Gal, as predicted in the
disaccharides Gale(1-4)GlcNAc and lactose, would be loss in this
alternative conformation (13,127).

Intramolecular interactions between contiguous residues have been
shown to cause wide variations in the 0 angle around the glycosidic
linkage (121). Since fucosylation does not appear to affect the
torsion angle in the Galp(I-4)GlcNAc glycosidic linkage, appreciable
interaction of the fucopyranosyl ring with the Gal and GlcNAc rings in

the trisaccharide may not occur.

117

The 0 torsion angle in the Gal3(1-4)GlcNAc glycosidic linkage in
the trisaccharide was evaluated using 2d 13C1'-C4"
coupling. An angular dependence of 13Cl substituents on the
magnitude of the geminal coupling has been shown (68,76). However, this
relationship between the conformation and the magnitude of geminal
coupling is not well understood (68). The C4" GlcNAc in the
trisaccharide Fuc6(1-2)[1313c]GalB(1-4)GlcNAc- s-hexanolamine was
observed to be broadened with respect to C2", C4‘, C4 or any single
resonance in the spectrum, shown in Figure 16. Although the
line-broadening (1.0-1.5 Hz) can not be quantitatively interpreted with
respect to the 0 torsional angle of the Gale(1-4)GlcNAc glycosidic
linkage, studies with [1-13c] and [2-13C] enriched
aldopyranosides indicated that the aglycon carbon equivalent to C4"
GlcNAc is trans to C2. Ethyl—B-D-[2-13c] galactopyranoside was
found to have a 3J 13C2-C1' coupling of 3.5 Hz indicating
a trans relationship between C2 and Cl'. Two bond,
23 13C1-C1" coupling in these simple glycosides resulted
in increased linebroadening ( 1 Hz) of the Cl' aglycon. The 1 to 1.5 Hz
linebroadening in the C4" GlcNAc resonance, equivalent to the Cl'
aglycon in the [1-13cj and [2-13C] enriched aldopyranosides
indicated that C4" GlcNAc is trans to C2' Gal and between a synfclinal
and elipsed orientation to Hl' and 05' Gal. Derivatization at C2 of Gal
with Fuc may affect the contribution of C2 to the coupling constant
since C4' trans to C2 with CH as a substituent may differ signficantly
from the same geometry with QfFuc at 02. However, a rotamer with C4"
GlcNAc synfclinal to Hl' and 05 Gal not only accounts for the observed

linebroadening but is in agreement with the exoanomeric effect, model

118

studies, theoretical calculations and X-ray crystallography (13, 121,
122, 124). This conformation also permits hydrogen bonding between
0H3" GlcNAc and 05' Gal as predicted for lactose by X-ray
crystallography and proposed in the disaccharide Gal3(1-4)GlcNAC (13,
127).

Comparison of the intra-residue galactose couplings in the
trisaccharide; la C1-C2 = 46 Hz, 20 C1-C3 = 3.5
Hz and a broadening of C6‘ as a result of 3a C1-C6 coupling,
with galactose indicate that the galactopyranosyl ring is not distorted
in the trisaccharide.

The p torsion angle in the FuCa(1-2)Gal glycosidic linkage in the
trisaccharide FUCa(1-2)[1-13C]Gals(1-4)GlcNAc-B-hexanolamine was
evaluated by 3JC-C coupling. Carbon-carbon coupling from 13Cl
Gal to Cl Fuc was used in hope of obtaining a coupling constant which
could be related to the m dihedral angle. The spectrum of the anomeric
region of Fuc6(1-2)[1-13c]GalB(1-4)GlcNAc-B-hexanolamine, shown in
Figure 18, reveals no coupling from 13cl Gal to Cl Fuc although a
broadening of the Cl Fuc resoanance of approximately 3 Hz is observed.
Due to the length of time required to obtain natural abundance data on
100 pmoles of material it was thought that better resolution could be
obtained with doubly 13C-enriched compounds. The sugar nucleotide
GDP[1-13C] Fuc was synthesized and the p-galactoside 0(1-2)
fucosyltransferase used to prepare [1-13CJFUCa(1-2)[1-13C]Gal8
(1-4)GlcNAc-B-hexanolamine. Again, no coupling was observable between
Cl'Gal and ClFuc. Comparison of linewidths of the 13C-enriched
resonances in [1-13CJFUCa(1-2)[1-13C]Galp(1-4)GlcNAc-3-hexanola-

119

mine with FUCa(1-2)[1-13C]GalB(1-4)GlcNAc-e-hexanolamine and
[1-13c]GalB(1-4)GlcNAc-B—hexanolamine unambiguously Shows a
linebroadening of approximately 3 Hz , shown in Figure 19, in the doubly
enriched trisaccharide. A detailed analysis and comparison of these
lines indicate that a 3a C1-Cl' coupling of approximately

1.5 Hz can account for the observed linebroadening. Resolution of the
coupling constant could not be obtained between the temperatures of 45
to -40°. Linewidths at -40° in 60% methanol water for the 13c

enriched resonances in [1-13c]FuCa(1-2)[1-13CJGalB(1-4)GlcNAc-
B-hexanolamine were 16 Hz with an internal dioxane linewidth of 2 Hz.
One possible explanation of the 16 Hz linewidths at -40° would be the
existance of two stable populations of the Fucd(1-2)Gal glycosidic
linkage. Further experiments are necessary to explain the observed
linewidths at -40°. Several other doubly 13C-enriched compounds,
[1-13CJFuca(1-2)[1-13CJGal-B(1-4)GlcNAc,
[1-13cjfuca(1-2)[1-13c]Ca1e(1-4)Cic and
[1-13C]FUCa(1-2)[1-13CJGal-8-ethyl were analyzed and compared

against their respective singly 13C-enriched compound. In all cases

a 3d C1-C1' coupling could not be resolved but an

interaction of 1.5 Hz could be estimated from the 3 Hz linebroadening.
Using this data and assuming a Karplus type relationship for 3d C-C
coupling the most stable rotamers about the FUCa(1-2)Gal glycosidic
linkage depicting theip torsion angle are shown in Figure 23. Rotamers
with Cl Fuc eclipsed or near eclipsed to either Cl or C3 Gal are omitted
since these are highly unfavorable due to nonbonded interactions.

Rotamer B with Cl trans to Cl' can also be omitted since an observable

120

.mmeES PFEmoofu Fewfimridui 2: “:22 8:35 2:328 3 02.58;

.mN oe=m_a

'Cl

03

 

Cl

Cl’

121

H2
Cl
ca’ Cl’
8 .
Hz’
Cl
C3’ Cl’

Figure 23

122

coupling constant of approximately 3.5 Hz would be expected for this
arrangement. A trans relationship between C2 and the Cl' methylene
carbon in the ethyl group of ethyl-B—D-[2-13C] galactopyranoside
gives an obervable 3d C2-C1' coupling of 3.5 Hz.

Similarly, 3J C1-C6’ through the ring oxygen, gives a 3.7

Hz coupling constant for a trans arrangement. Unfortunately, an

appropriate model for 30 coupling, through a glycosidic linkage,

C-C
for an angle of 120° is not available. Values for the 3J(>4: coupling
in rotamers C and 0 would be expected to lie in the range of 0 to > 1.0
Hz for gauche situated carbons since 3J C1-C4 in

aldopyranoses, C1 gauche to C4, is not observable (68). A conformation
with Cl Fuc less than 120° to Cl Gal would not be expected to result in
a 3 Hz linebroadening since C3'GlcNAc is less than 120° to Cl'Gal in the
disaccharide Gal3(1-4)GlcNAc B-hexanolamine and no increased
linebroadening was observed (13). The observed 30 Cl-Cl'

interaction of a 3 Hz linebroadening, approximately a 1.5 Hz coupling,
favors a m torsion angle with Cl at an angle Slightly greater than
eclipsed with H2' Gal (Figure 23 Rotomer A). NMR can not distinguish
between a rotamer with Cl Fuc at an angle slightly greater than 120°
(Figure 24 Rotomer A) from C1' Gal and a conformer with Cl turned
through 180° as Shown in Figure 24; Rotomer 8. Model building studies
do not completely resolve this ambiguity. Steric interactions appear
greater in rotamer B where the fucopyranosyl ring interferes with the
Nfacetylglucosamine ring and the Cl Fuc is nearly eclipsed with Cl Gal.
Results obtained from model building of the O antigenic trisaccharide
and conclusions from theoretical calculations (122) X-ray

crystallography (121) and nmr (13,117,118) on the 0 angle in other

123

.03; F0 uc< Few Pu cmmZumm mew—asoo pmcwor> N: m.H <
»_wueaexocqa< m>wu OP umuuwaxm mmmxcwe uwuwmouapo Pmoamravduzu mg» uson< mm_mc< _mcome0P .em meamwm

124

CI H2

 

ca’ Cl’
CI
B ‘

Figure 24

125

disaccharides favor the rotomer (A) in which Cl Fuc is at an angle
slightly greater than 120° with respect to Cl Gal.

The 0 torsion angle about the Fuc6(1-2)Gal glycosidic linkage
was evaluated using the disaccharide FUCa(1-2)Gal-B-ethyl. Ethyl
B-D-[2-13CJ-galactopyranoside was chemically synthesized and fucose
added to it enzymatically with the use of GDP-fucose and the
B-galactoside 0(1-2) fucosyltransferase. The carbon-13 enriched and
natural abundance FUCa(1-2)Gal-B-ethyl were used as a model for the
blood group 0 trisaccharide. Evaluation of the 0 torsion angle could be
obtained by 3d 13C2'-HI coupling. There is ample
experimental and theoretical evidence that a "Karplus-type" of
relationship exists between 3a C-H coupling and the torsional
angle (56). Homer gt al. (125) have established a "Karplus-type" of
relationship involving inter-residue Sp3 carbons contained in the
specific sequence, 13C-0-C-'H. The region of 0 to 60° was difficult
to define precisely due to the lack of appropriate model compounds
(125). The 180 MHz 'H nmr of FUCa(1-2)Gal-B-ethyl was obtained and the
Hl Fuc examined (shown in Figure 20A). The anomeric proton of fucose
resonates as a doublet at 5.04 ppm aS a result of a 3.4 Hz
3a Hl-H2 coupling. No long range coupling was observed.

The Hl Fuc resonance was confirmed by comparison with the 'H nmr of
[1-13C]FUCa(1-2)Gal-B-ethyl. The Hl of Fuc in
FuCa(1-2)[2-13c]Gal-p-ethyl is a multiplet (shown in Figure 20B)
resulting from 3d H1-H2 and 3d H1_13C2u

coupling. The multiplet is comprised of a 3d H1-H2

Splitting of 3.4 Hz and a 3d H1-13C2' splitting of approximately

126

3.2 Hz. The inter-residue coupling of 3.2 Hz corresponds to a torsional
angle in which C2'Gal is sygfclinal to 05 and Hl. NMR can not
distinguish between this conformer and one turned through a 180°.
However, a conformer turned through 180°, having C2' syfleclinal to C2
and 05 is sterically unfavorable as discussed previously (124).
Interestingly, the former angle corresponds closely to the 0 torsion
angle in gentabiose (glucosee(1-6)glucose) octaacetate

(3.1 13C6-H1 of 3.8 Hz) (118). Two bond, 013C1_C4, ,

in Galp(1-4)GlcNAc also indicated a 0 torsion angle with C4' near
syn-clinal with HI and 05 as was observed in the crystalline structure
of stereo-Similar lactose (13). Lemieux and co-workers (129) estimated
0 for a and 3 alkyl-D-glycopyranosides and found 3a C-H values

of approximately 3.8 and 4.2 Hz for the a and 8 anomers, respectively.
The exo-anomeric effect appears to be the predominant force in
stabilizing the conformation about the 0 torsion angle in the absence of
extenuating circumstances. Small variations observed in the 0 angle may
be the result of steric interactions, intramolecular hydrogen bonding
and a verses 8 glycosidic linkage effects.

Enzyme mediated synthesis of complex oligosaccharides, using
partially purified glycosyltransferases and chemical synthesized sugar
nucleotides, has been shown to proceed with high yield, anomeric purity
and ease of product purification. The galactosyltransferase and the
fucosyltransferase have been shown to utilize acceptor substances
containing a number of different aglycons. Partial purification of
these enzymes using affinity adsorbants allows quantities of enzyme to

be easily prepared in less than two weeks.

127

Sugar nucleotides may be synthesized from specifically
l3C-enriched carbohydrates and used with the glycosyltransferases to
synthesize specifically 13c-enriched complex oligosaccharides. High
resolution 13C-nmr allows product characterization and unambiguous
assignment of many resonances with the use of specific
13c-enrichment. Accurate resonance assignment is the first priority
before using the synthesized oligosaccharide in biological studies and
for future assignment of more complex biological oligosaccharides.
Specifically 13C-enriched compounds allow the measurement of
inter-residue 3d C-C' 20 C-C and 3J C-H coupling
constants which can then be used to evaluate the solution conformation
of the molecule. The specifically 13C-enriched fucosylated
compounds can now be used in biological studies to evaluate
glycoside-protein interactions, in particular antigen-antibody
interactions. Glycoproteins and glycolipids could also be used as
acceptors for the glycosyltransferases allowing specific
13C-enrichment of the carbohydrate moiety in the glycoprotein or
glycolipid. Examination of these specifically 13c-enriched
glycoproteins and glycolipids may provide much information as to the

role of carbohydrates in these molecules.

1.

2.
3.
4.

9.
10.

11.
12.

13.
14.

15.

16.
17.

18.
19.

BIBLIOGRAPHY

Nalborg, Jr., E.F. Glycoproteins and Glycolipids in Disease

Processes A.C.S. Symposium Series 99, (l978).

Spiro, R.G. Advan. Protein Chem. 91, 349 (l973).

Marcu, D.M. New Eng. J. Med. 999, 994 (1969).

Lemieux, R.U. and Driguey, H., J. Am. Chem. Soc. 91, 4009 (1975).
Lemieux, R.U. and Driguey, H., J. Am. Chem. Soc. 91, 4063 (l975).

Bundle, D.R., Smith, I.C.P. and Jerning, H.J., J. Biol. Chem. 999,
2275 (l974).

Harris, P.L. and Thornton, E.R., J. Am. Chem. Soc. 999, 6738
l978 .

Czarniecki, M.F. and Thornton, E.R., J. Am. Chem. Soc. 99, 8279
l977 .

Mazueek, M. and Gorin, P.A.J., Carb. Res. 19, Cl (l979).

Brewer, C.F. and Keiser, H., Proc Nat. Acad. Sci. U.S.A. lg, 3421
(l975).

Villagranca, J.R., Arch. Biochem. Biophys. 199, 465 (l974).

Colson, P., Jennings, H.J. and Smith I.C.P., J. Am. Chem. Soc. 99,
8081 (1974).

Nunez, H.A. and Barker, J. Biochem. (in press) l980.

Brew, K., Vanaman, T.C., Hill, R.L., Proc. Nat. Acad. Sci. U.S.A.
59, 491 (l968).

Barker, R., Olsen, K.N., Shaper, J.H., and Hill, R.L., J. Biol.
Chem. 251. 7135 (l972).

Bean, R.C. and Hassid, H.Z., J. Am. Chem. Soc. 11, 5737 (l955).

Kabat, E.A. Blood Group Stubstances; Their Chemistry and
Immunochemistry. New York: Academic Press (T956).

 

 

Lelori, L.F. and Cardini, C.E., J. Am. Chem. Soc. 19, 6340 (1957).

Schachter, H. and Roseman S: Mammalian Glycosyltransferases: Their
Role in the Synthesis and Funtion of Glycoproteins. In
Biochemistry of Proteoglycans and Glycoproteins, (H.J. Lennarz,
Ed), (in press).

128

20.

21.
22.

23.

24.

25.

26.

27.

28.

29.
30.
31.
32.

33.

34.
35.

36.
37.

38.
39.

129

Haechter, C.J. and Lennarz, H.J., Ann. Rev. Biochem. 99, 95
(l976).

 

 

Beyer, T.A., Rearick, J.I., Paulson, J.L. Prieels, J.P. Sadler,
J.E. and Hill R.L., J. Biol. Chem; 254, 12531 (l979).

Paulson, J.C. Prieels, J.P. Glasgow:_[.R., and Hill, R.L., J.
Biol. Chem. 999, 5617 (l978).

Young, J.D., Tsuchiya, 0., Sandlin, D.E. and Holroyde, M., Fed.
Proc. 99, 1684 (l978).

Schachter, H., Narasimhan, S. and Wilson, J.R. Biosynthesis and
Catabolism of Glycoproteins. In Glycoproteins and Glycolipids in

 

Disease Processes. ACS Symposium Series (Washington l978).

 

Hill, H.D. Jr., Schwyzer, M., Steinman, H.M. and Hill R.L., J.
Biol. Chem. 999, 3799 (l977).

Schachter, H., McGuire, E.J. and Roseman, 5., J. Biol. Chem. 999,
5321 (l971).

Feizi, T., Kabat, E.A. Vicari, G., Anderson, B. and Marsh, w.L.,
J. Immunol. 999, 1578 (1971).

Brah, B., Schiff, F. and Neinman, H. Klin. Hachenschr._99, 1592
(1932).

Hakomor, S, Chem. Phys. Lipids, 9, 96 (l970).
Schwyzer, M. and Hill, R.L., J. Biol. Chem. 999, 2338 (l977).
Schwyzer, M. and Hill, R.L., J. Biol. Chem. 999, 2346 (1977).

McGuire, E.J., Jordian, G.N., Carlson, D.M., and Roseman, S., J.
Biol. Chem. 999, 4112 (1965).

Turkington, R.U., Brew, K. Vanaman, T.C. and Hill R.L., J. Biol.
Chem. 999 (1978) 3382. ,

Palmiter, R.D., Biochem. J. 999, 409 (l969).

Vanderhaar, B., Owens, 1.5., and Topper, Y.J., J. Biol. Chem. 999,
467 (l973).

Turkington, R.w. and Hill R.L., Science 999, 1458 (I969).

Sacke, R.G. and Heald, C.w. In Lactation A Comprehensive Treatise,
Academic Press, New York, 2, 147 (l974).

 

Powell, J.T. and Brew, K., Eur. J. Biochem. 99, 217 (l974).

Mageei E.E. Mawal, R., and Ebner, K.E., J. Biol. Chem. 999, 7565
(l973 .

40.
41.
42.

43.
44.

45.
46.

47.

48.

49.

50.
51.

52.

53.

54.

55.
56.
57.
58.

59.
60.

61.
62.

130

Heath, E.C., Ann. Rev. Biochem. 99, 29 (1971).
Watkins, W.M. and Hassid, W.Z., J. Biol. Chem. 999, 1432 (l962).
Broadbeck, N. and Ebner, K.E., J. Biol. Chem. 999, 762 (1966).

Babad, H. and Hassid, W.Z., J. B101. Chem. 999, 2672 (1966).
Trayer, I.P. and Hill, R.L., J. Biol. Chem. 999, 6666 (l971).

Porath, J., Axen, R. and Ernback, 5., Nature 999, I496 (l967).

Bell, J.E., Beyer, T.A. and Hill, R.L., J. Biol. Chem. 999, 3003
(1976).

Chester, M.A. and Watkins, W.M., Biochem. Biophys. Res. Com. 99,
835 (1969).

Munro; J.R. and Schachter, H., Arch. Biochem. Biophys. 999, 534
l969 .

Baenziger, J. Kornfeld, S. and Kochwa, S., J. Biol. Chem. 999,
1897 (l974).

Baenziger, J. and Kornfeld, S. J. Biol. Chem. 999, 7270 (1974).

Kornfeld, R., Keller, J. Baenziger, J. and Korneled, S., J. Biol.
Chem. 999, 3259 (l971).

Jarkovsky, A., Marcus, D.M., Grullman, A.P., Biochem. 9, 1123
(1970).

Grollman, A.P., Hall, C.W. and Ginsburg, V., J. Biol. Chem. 999,
975 (l965).

Schenkel-Brunner, H., Chester, M.A. and Watkins, W.M., Eur. J.
Biochem. 99, 269 (l9720.

Pacuszka, T. and Koscielak, J., FEBS Letters 99, 348 (l974).
Bella, A. and Kim, Y.S., Arch. Biochem. Biophys. 999, 753 (1971).
Jabbal, I. and Schachter, H., J. Biol. Chem. 999, 5154 (l971).

Basu, S., Basu, M. and Chien, J., J. Biol. Chem. 999, 2956
(1975).

Hoflack, R. and Verbert, A., Eur. J. Biochem. 99, 1 (l978).

Basu, M. Moskal, J., Gardner, 0. and Basu, 5., Biochem. Biophys.
Res. Commun. 99, 1380 (l975).

Aminoff, D. and Furnkawa, K., J. Biol. Chem. 999, 1659 (l970).

Stealey, J.R., Watkin, W.M., Biochem. J. 999, 160 (l971).

63.

64.
65.
66.

67.

68.

69.

70.

71.

72.
73.

74.

75.
76.
77.

78.
79.
80.
81.

82.

83.

131

Chester, M.A., Yates, A.D., and Watkins, W.M., Eur. J. Biochem.
99, 583 (l976).

Annison, E.F. and Morgan, W.T., Biochem. J. 99, 247 (l952).
Kosielak, J., Biochem. Biophys. Acta. 99, 313 (l963).

Shen, L., Grollman, E.F. and Ginsburg, V., Biochem. 99, 224
(l968).

Bosmann, H.B., Hagopian, A. and Eylar, E.H., Arch. Biochem.
Biophys. _1_2_8. 240 (1968).

Walker, T.E., London, R.E., Whaley, T.W., Barker, R. and
Matwiyoff, N.A., J. Am. Chem. Soc. 99, 5807 (l976).

Gagnaire, D.Y., Nardin, R., Taravel, F.R. and Vignon, M.R.,
Nouveau J. Chim. 9, 423 (l977).

Bundle, 0.R., Jennings, H.J., and Smith, I.C.P., Can. J. Chem. 99,
3812 (l973).

Pfeffer, P.E., Valentine, K.M. and Parrish, F.W., J. Am. Chem.
Soc. 999, 1265 (1979).

Gorin, P.A.J. and Mazurek, M., Carbhydr. Res. 99, 171 (I976).

Bock, K., Lundt, I. and Pedersen, C. Tetrahedron lett. 99, 1037
(l973).

Bock, K. and Pedersen, C., Acta Chem. Scand., Ser. B, 99, 258
(1975).

Perlin, A.S. and Casu, B., Tetrahedron Lett. 99, 2921 (1969)
Schwarcz, J.A. and Perlin, A.S., Can. J. Chem. 99, 3667 (I972).

Schwarcz, J.A., Cyr, N., Perlin, A.S., Can. J. Chem. 99, 1872
(l975).

Karplus, M., J. Chem. Phys. 99, 11 (1959).
Karplus, M., J. Am. Chem. Soc. 99, 2870 (l963).
Wasylishen, R. and Schaefer, T., Can J. Chem. 99, 2710 (l972).

Karabatsos, G.J., Orzech, Jr., C.E. and Hsi, H., J. Am. Chem. Soc.
99, 1817 (l966).

Weigert, F.J. and Roberts, J.D., J. Am. Chem. Soc. 99, 5962
(1967).

Karabatsos, G.J., Graham, J.D., and Vane, F.M., J. Am. Chem. Soc.
99, 37 (1962).

84.

85.
86.
87.

88.

89.

90.

91.
92.

93.
94.
95.

96.

97.
98.

99.
100.

101.
102.

103.
104.

132

Marshall, J.L., Miller, D.E., Conn, S.A., Seiwell, R., and Ihrig,
A.M., Acc. Chem. Res. 9, 333 (1974).

Hansen, P.E., Org. Mag. Res. 99, 215 (1978).
Wray, V., Prog. in NMR Spectroscopy 99, 177 (1979).

Stothers, J.B., Carbon-13 NMR §pectroscopy, Academic Press, New
York (1972).

Weigert, F.S. and Roberts, J.D., J. Am. Chem. Soc. 99, 6021
(1972).

Marshall, J.L. and Miller, D.E., J. Am. Chem. Soc. 99, 8305
(1973).

Barfield, M., Burfitt, I., and Doddrell, 0., J. Am. Chem. Soc.
.99, 2631 (1975).

Barfield, M., .1. Am. Chem. Soc. 909, 1 (1980).

Hirao, K., Nakatsuji, H. and Kato, H., J. Am. Chem. Soc. 99, 31
(1973).

Bandursky, R.A. and Axelrod, B., J. Biol. Chem. 999, 405 (1953).
Levy, A.L. and Chung, 0., Anal. Chem. 99, 896 (1953).

Bohlen, P., Stein, 5., Dairman, W., and Undenfriend, 3., Arch.
Biochem. and Biophys. 999, 213 (I973).

Moffatt, J.G., In Methods in Enzymology.9, Academic Press, New
York, 136 (1966).

 

Cuatrecasas, P., J. Biol. Chem. 999, 3059 (1970).

Bates, Frederick and Associates, In Polarimetry, Saccharimetry
and the Sugars, U.S. Government Priniting Office, 488 (1942).

 

MacDonald, D., J. Org. Chem. 99, 1107 (1962).

Nunez, H.A., O'Connor, J., Rosevear, F.R. and Barker, R.,
submitted to Can. J. Chem. (1980).

Horton, 0., Org. Syn. 99, 1 (1966).

Chiang, C.K., McAndrew, M. and Barker, R., Carb. Res. 99, 93
(1979).

Isbell, H.S., Methods of Carbohydr. Chem. 9, 13 (1963).
Fletcher, Jr., H.G., Methods Carbohydr. Chem. 9, 77 (1962).

105.

106.

107.
108.

109.
110.

111.

112.
113.

114.

115.

116.

117.

118.

119.
120.
121.

122.
123.
124.

133

Serianni, A.S., Nunez, H.A., and Barker, R., Carbohydr. Res. 99,
71 (1979)

Sadler, J.E., Rearick, J.I., Paulson, J.L. and Hill, R.L., J.
Biol. Chem. 999, 4434 (1979).

Somogyi, M., J. Biol. Chem. 999, 69 (I945).

Kamath, S.A. and Rubin, E., Biochem. Biophys. Res. Commun. 99, 52
(1972).

Koenigs, W. and Knorr, E. Ber. 99, 957 (1901).

Wulf, G. and Rohle, G., Angew, Chem. Int. Ed. Engl. 99, 173
(1974).

Lemieux, R.U., Hendriks, K.B., Stick, R.V., and James, K., J. Am.
Chem. Soc. 99, 4056 (1975).

Sinay, P., Pure and Appl. Chem. 99, 1437 (1978).

Dorman, D.E., Angyal, S.J. and Roberts, J.D., J. Am. Chem. Soc.
99, 1351 (1970).

Dorman, D.E., and Roberts, J.D., J. Am. Chem. Soc. 99, 1355
(1970).

Usui, T., Yamaoka, N., Matsuda, K. and Tuzimura, K., J. Chem. Soc.
Perkin I, 2425 (1973).

Gagnagre, 0. and Vincendon, M., J. Chem. Soc. Chem. Comm. 99, 509
1977 .

Excoffier, G., Gagnaire, D.Y., and Taravel, F.R., Carbhydr. Res.
99, 229 (1977).

Gagnaire, D.Y. Nardin, R., Taravel, F.R., and Vignon, M.R.,
Nouyeau, J., Nouveau, J. De Chim 9, 5 (1977).

Lemieux, R.U., and Koto, S., Tetrahedron 99, 1933 (1974).
Booth, H., Progr Nucl. Magn. Resonance Spectrosc. 9, 149 (1969).

Marchessault, R.H. and Sundararajan, P.R. Pure and Applied Chem.
99, 399 (1975).

Rees, D.A. and Smith P.J.C., J. Chem. Soc. Perkin II, 836 (1975).
Bock, K. and Pedersen, C., Acta Chem. Scand. 999, 354 (1977).
Lemieux, R.U., Koto, S. and Voisin, 0., In Anomeric Effect Origjn

and Consgguences (W.A. Szarek and D. Horton Ed). ACS Symposium
Series, 99, 17 (1979).

 

134

125. Hamer, C.K., Balza, F., Cyr, N. and Perlin, A.S., Can. J. Chem.
99, 3109 (1978).

126. DeBruyn, Anteunis, M., Garegg, P.J. and Norbert, T., Acta.
Chemica. Scand. 999, 820 (1976).

127. Hirotsu, K. and Shimada, A., Bull. Chem. Soc. Jap. 99, 1872
(1974).

135

APPENDIX

List of Publications

Rosevear, P., VanAken, T., Baxter, J., and Ferguson-Miller, S: Alkyl
Glycoside Detergents: A Simpler Synthesis and Their Effects on the

Kinetic and Physical Properties of Cytochrome g Oxidase (in press

Biochemistry).

Nunez, H.A., O'Connor, J.V., Rosevear, P.R., and Barker, R.: The
Chemical Sythesis and Characterization of a- and 8-L-Fucopyranosyl

Phosphates and GDP-Fucose (submitted to Can. J. Biochem.)

Rosevear, P.R., Murphy, K., and Barker, R.: Evaluation of the Mechanism
of the Alkaline Degradation of Glucose Using Isotopically Enriched
SubstrateszThe Retention of Hydrogen from C2 of Glucose (submitted to J.
Am. Chem. Soc.)

Rosevear, P.R. and Barker, R.: Synthesis and 13c NMR Analysis of

1-13C Enriched L-Fucose and L-Rhamnose (in preparation for Carbhydr.

Res.)