nu. ’ V
v3.1. 557:; a V ‘ -‘

"an NHL. . an - . - ..,~-” 6
{auuvfp‘wr Urn " " " "' f' ', 2.". C "w‘n‘n *1’

‘4. O“.- I F " ’ I I l .1 ' t
I . M. r ’ . , . .u " '. a
,';’!X '13
r V'f'Q.‘

J.‘ 1:~'

7
,_..;4-.‘rava...-.
..‘...... .v..
, ,, -r .

‘. - 1?”. I“!

w 3-" ‘1‘“
‘ . {‘1' ”1:5. ".‘

' 014'!»

n

. .
w .iu-nm
n v 1- .
. .\

a.» u
“n- m.
“v1 “{ch
>4
.

"‘0;
‘JJ' .
war?"
a? I

‘.,h:
-».«

5-4. -.

IDJ
. vu- .
I"

u’
,u’ . ‘ I 1"
z“. @mfib

4? .‘

. u-
. u'v
u .. .
,' my "f V
X’ - w.

P 311; ‘
”‘ 5753

4" ~

u. 44

"11814“:

n n ma
I

.6
up,“
MW." .5:

ifnrj:}1:f.-

, \u ’
"9%)?
W

, ‘
n ' U .
.n .r .,
m... - h
. \ r."

by
.1

 

4..
,' . ‘ ”n l
‘4, A 4... u.

«51:

.' .
nu ‘ vr
u

a ‘
~v~| .
wrw ‘14

r
h.
t'uJ-L
. 4 ‘.'..
"a "'t»: v

y». u
w

s .,v

'u

 

54! mué'i- '

L

r.

" ‘ . -.. I; ’. '13.

4:9. , wx V x. ‘ '... ,..'
.. ‘

.J 3.: ”$9.. 75.2.“2. -.' . ..
,V, -,‘ “311., ‘. . 31!;‘5'EL‘44- w...“
W“ 0,” .~ .. .. . m

I m.- . . J

Am-

‘llv
'Ji.

‘ 1',
3 ' A571. .4.
”.7.-. 4"»

1 .-_

 

lint-ac»:

MICHIGAN STATE UNIVERS I E

2m rum IIIHII lmlln‘rllll

3 1293 00911 7072

ll

 

 

 

 

 

 

 

 

 

g
l

 

 

II

 

 

 

This is to certify that the

dissertation entitled

STUDIES ON THE PURIFICATION AND REGULATION OF
CMP-SIALIC ACID:LACTOSYLCERAMIDE
ALPHA2-3 SIALYLTRANSFERASE

presented by

Lyla J. Melkerson-Watson

has been accepted towards fulfillment
of the requirements for

Ph.D. Biochemistry

degree in

 

 

Major profes r

Date 11-28-90

 

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

‘

I umnr
Michigan State
University J

“fl

fl

 

 

L

PLACE IN RETURN BOX to

remove We check
TO AVOID FINES retum on

or before ode due.

DATE DUE DATE DUE DATE DUE

out from your record.

       

\

  

\

       

\\

MSU Is An Affinnetlve ActiorVEquel Opportunity Institution
CWMMT
.7 , , , _ , /

 

STUDIES ON THE PURIFICATION AND REGULATION OF CIVIP-SIALIC
ACID:LACTOSYLCERAMIDE 012-3 SIALYLTRANSFERASE

By

Lyla Jill Melkerson-Watson

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1991

513;?"527

C" x __
CO

//.
(f

ABSTRACT

STUDIES ON THE REGULATION OF CNIP-SIALIC
AClDzLACTOSYLCERAMIDE «2-3 SIALYLTRANSFERASE

By

Lyla Jill Melkerson-Watson

A procedure is described for assays of CMP—sialic acidzlactosylceramide 012-3
sialyltransferase (SAT-1), using Sep Pak C13 cartridges. Complete separation of the
more polar CMP-sialic acid and sialic acid from the less polar GM3 is simple and rapid
relative to other methods. Chromatographic recovery of 6M3 is high when
phosphatidylcholine is added. The procedure may be applicable for other in vitro
glycosyltransferase assays.

SAT-l has been purified 40,000-fold from rat liver by affinity chromatography
on lactosylceramide-aldehyde Sepharose 4B. Synthesis of the column is described.
Purification was verified by immunoaffinity chromatography on M126C7-Affi Gel-10.
The MlZGC7 monoclonal antibody specifically inhibits and immunoprecipitates SAT-l
activity. The apparent molecular weight by SDS-PAGE is about 60Kd. Studies on
substrate specificity indicate SAT-1 recognition of the glycolipid, GalB 1-4Gchl-O-Cer
and to a lesser extent GalBl-O-Cer or Gchl-O-Cer. SAT-l is a glycoprotein. The
carbohydrate moieties are detected with specific lectins. Deglycosylation of SAT-1
results in a 43Kd band. The two-dimensional electrophoretogram of SAT-1 indicates a
pI range of 5.7 to 6.2 for the 60Kd protein.

Lauryldimethylamine oxide (LDAO) is employed in the purification of SAT-1.
This detergent has advantages over Triton detergents in the solubilization and
stabilization of this sialyltransferase. The ability of LDAO to activate and stabilize

SAT-1 activity may involve the structural similarity between the hydrophobic moieties
and quaternary amino groups of LDAO and phosphatidylcholine.

Co-purification of an endogenous proteolytic activity has been proposed as the
cause for the size heterogeneity of rat liver sialyltransferases. Addition of protease
inhibitors, sulfhydryl-reducing agents and antimicrobial agents to immunoaffinity-
purified SAT-l dramatically affects its activity. All protease inhibitors examined, with
the exception of PMSF, inhibited the purified enzyme. The most inhibitory were the
cysteine (thiol) protease inhibitors. Further, the apparent activation of SAT-l activity
in the presence of B—mercaptoethanol was observed.

Turnover of SAT-1 in butyrate-synchronized KB cells is cell-cycle dependent.
Regulation of 6M3 synthesis in KB cells may involve the phosphorylation of SAT-l.
Analysis of immunoaffinity—purified KB cell SAT-l , during different time points of
expression within the G 1 phase, by SDS-PAGE and its immunodetection on Western
blots with monoclonal specific for phosphotyrosine residues, indicates SAT-1 is a
phosphotyrosine-containing protein. The expression of this tyrosine-phosphorylated
form may regulate SAT-1 activity.

To may parents, for their love and desire to educate their childem
To Greg, for his love and support
To Richard, for a new and youthful perspective of life

iv

ACKNOWLEDGEMENTS

I would like to thank my mentor Dr. Charles C. Sweeley for all of his advice
and guidance. He has an extraordinary zest and love for research to which I aspire.

I would also like to thank the following people: .

The members of my advisory committee, Drs. Thomas Deits, Walter Bsselman,
Estelle McGroarty, and William Smith, for their guidance and assistance during my
dissertation project.

Drs. Jim Fairley and Richard Anderson for their special guidance and assistance
over the years and for just being there...

Dr. Kimihiro Kanemitsu for his friendship and help in teaching me GC analysis
of sialic acid and ozonolysis derivatization of LacCer.

Doug Wiesner and Kiyoshi Ogura for their assistance with the quantitation of
GM3 and its sialidase activity, respectively.

Dr. Elizabeth Cowles for her advice on 2D SDS-PAGE and comradeship.

Ms. Patty Voss for all "the tricks of the trade" on monoclonal antibody
production and for ”babysitting the guys” on occasion.

Dr. Rawle Hollingsworth for the use his/ our HPLC.

Mr. Joe Leykam and Ms. Melanie Corlew for assisting me during the long
hours on the microbore HPLC and protein analysis.

Carol Smith for her excellent secretarial assistance.

My friends and fellow Grads in Dr. John Wang's lab, Liz Cowles, Jamie Laing,
Neera Agrwal, and Kim Hamann, for being ”my lab away from lab. "

All my laboratory mates who have supported and challenged me over the years,
Barbara Myskwiecz-Sullivan, Frank Wilkinson, Kimihiro Kanemitsu, Mitsu Nakamura,
Seigo Usuki, Masahiko Shigematsu, Shu-Chen Liu, Dr. Lyu, Patricia Hoops, Lynne
Royer, Kazz, Kamada, Musti Swamy, Rawle Hollingsworth, Rob Dodds, Amanda

Poxon, Kiyoshi and Misa Ogura, Barb Kuhns, Bev Chamberlain, Zhi-Heng Huang,
Bao-Jen Shyong, and last but not least, Doug Wiesner. My friends. . .thank you all!

vi

TABLE OF CONTENTS

Page
ABSTRACT ........................................................................... ii
DEDICATION ........................................................................... iv
ACKNOWLEDGEMENTS .................................................................... v
TABLE OF CONTENTS .................................................................... vii
LIST OF TABLES ......................................................................... xiv
LIST OF FIGURES .......................................................................... xv
Chapter 1. Literature Review ............................................................. 1
Introduction ............................................................................ 1
Ganglioside Structure ................................................................... 2
Ganglioside Biosynthesis ............................................................... 5
Regulation of Ganglioside Biosynthesis ............................................ 12
Topography of Glycosyltransferase ................................................. 13
Molecular Organization of Glycosyltransferases .................................. 19
Purification of Glycosphingolipid Glycosyltranferases ........................... 20
Regulation of Cell Proliferation by Glycosphingolipids ......................... 21
GM3 in the Regulation of Cell Growth ............................................ 23
Regulation of GM3 Synthesis ........................................................ 26
References .......................................................................... 30
Chapter 2. A Quantitative Method for Separating Reaction
Components of CMP-Sialic Acid:Lactosylceramide
Sialyltransferase Using Sep Pair C13 Cartridges ...................... 40
Abstract ................................................................................. 41
Introduction ............................................................................. 42

Materials and Methods ................................................................ 44

Materials ....................................................................... 44
Rat Liver Golgi Membranes ................................................. 44
CMP-Sialic Acidzlactosylceramide Sialyltransferase
* Assay and Sep Pak C13 Separation Strategy .............................. 45
Quantitative Recovery of SAT—1 Reaction ...................................
Components from Sep Pak C13 Cartridges ............................... 46
Qualitative Recovery of the [14C]GM3 Product ......................... 46
Results ................................................................................... 48
Separation of SAT-1 Reaction Components ............................... 48
Application of SAT—1/Sep Pak C13 Strategy in
Determining SAT-1 Activity ................................................ 54
Discussion ......................... A ..................................................... 5 8
Acknowledgements .................................................................... 60
References .............................................................................. 61

Chapter 3. Purification to Apparent Homogeneity by Immunoaffinity
Chromatography and Partial Charterization of the GM3
Forming Enzyme, CMP-Sialic AcidzLactosylceramide

«2-3 Sialyltransferase(SAT-1) from Rat Liver Golgi ................ 63
Abstract ................................................................................. 64
Introduction ............................................................................. 65
Experimental Procedures ............................................................. 67

Materials ....................................................................... 67

Animals and Cell Lines ...................................................... 68

Assays .......................................................................... 68

Protein Concentration ............................................... 68
SAT-1 Activity Assay ............................................... 68
812-6 Sialyltransferase Activity Assay ........................... 69
Thiamine Pyrophosphatase Activity Assay ....................... 69
Quantitation of Lauryldimethylamine Oxide ..................... 69

viii

Affinity Purification of SAT-l .............................................. 69

Preparation of Golgi-enriched Microsomes ...................... 69
Detergent Extraction ................................................. 7O
Affinity Chromatograph on
CMP—hexanolamine Sep so I ................................... 7O
Rechromatography on CMP—hexanolamine
Sepharose II ........................................................... 71
Preparation of LacCer—aldehyde Sepharose ...................... 71
Chromatography on LacCer-aldehyde Sepharose ............... 74
Strategy for Developing a Monoclonal Antibody to
SAT-1 and Its Application in SAT-l Purification ........................ 74
Production of Monoclonal Antibodies to
SAT-1 ................................................................. 74
Inhibition of SAT-l Activity by M12GC7 ....................... 75
Specificity of M12GC7 ............................................. 76
Immunoaffinity Purification of SAT-1 ............................ 76
Assessment of Purity and Biological Activity ............................ 76
Verification of SAT-l Homogeneity .............................. 76
Localization of SAT-l Acitivty with
Silver-Stained Protein ............................................... 77
SAT-1 Specificity .................................................... 77
Method #1 - Specificity for Various
Glycosphingolipid Substrates .............................. 77
Method #2 - Specificity for Various
Glycoprotein Substrates .................................... 78
Glycan Detection ..................................................... 79
Two—Dimensional Gel Electrophoresis ............................ 79
Results .................................................................................. 81
Detergent Extraction ......................................................... 81
Preliminary Characterization of SAT-1 .................................... 81

Affinity Purification of SAT-l from

Rat Liver Golgi ............................................................... 81

Assessment of Affinity-purified SAT—1

by Gel Electrophoresis ....................................................... 83

Specificity of M12GC7 Anti-SAT-l

Monoclonal Antibody ........................................................ 86

Purification of SAT-l to Apparent Homogeneity

by Immunoaffinity Chromatography ....................................... 87

Specificity of SAT-l ......................................................... 93

SAT-1 Glycan Detection ..................................................... 93

Two-dimensional Protein Pattern of SAT-1 ............................... 96
Discussion .............................................................................. 99
Acknowledgments .................................................................... 103
Footnotes ............................................................................... 104
Abbreviations .......................................................................... 105
References ............................................................................. 107

Chapter 4. Special Considerations in the Purification of CMP-Sialic
Acid:Lactosylceramide 012-3 Sialyltransferase (SAT-l):
Solubilization and Stabilization of SAT-l by

Lauryldimethylamine Oxide (LDAO) .................................. 111
Summary ............................................................................... 1 12
Methods ................................................................................ 1 13

Materials ...................................................................... 113

Rat Liver Golgi .............................................................. 113

CMP-Sialic Acid:Iactosylceramide a2-3 Sialyltransferase

(SAT-1) Activity Assay ..................................................... 113

Two-Dimensional Gel Electrophoresis .................................... 113
Results and Discussion ............................................................... 114
Footnotes ............................................................................... 125
Acknowledgements ................................................................... 126

References ............................................................................. 127

Chapter 5. Special Considerations in the Purification of the GM3
Ganglioside Forming Enzyme, CMP-Sialic
AcidzLactosylceramide 012-3 Sialyltransferase

(SAT-l): Effects of Protease Inhibitors on SAT-l .................. 129
Summary ............................................................................... 130
Introduction .......................................................................... 31
Materials and Methods ............................................................... 132

Materials ...................................................................... 132

Preparation of Golgi-enriched Microsomes ............ 132

Purification of CMP-Sialic AcidzLactosylceramide

022-3 Sialyltransferase ....................................................... 132

CMP—Sialic AcidzLactosylceramide a2-3

Sialyltransferase (SAT- 1) Activity Assay ................................ 132
Results and Discussion ............................................................... 134
Conclusions ............................................................................ 142
Footnotes ............................................................................... 144
Acknowledgements ................................................................... 145

References............................ ................................................. 146

Chapter 6. Regulation of CMP-Sialic AcidzLactosylceramide

022-3 Sialyltransferase (SAT-l)
by Covalent Modification ................................................. 149
Abstract ................................................................................ 150
Introduction ............................................................................ 151
Materials and Methods ............................................................... 153
Materials ...................................................................... 15 3
Cells ........................................................................... 153
Methods ....................................................................... 154
Cell Synchronization ............................................... 154
Butyrate Block of KB Cells ............................... 154

xi

Double—Thymidine Block of KB Cells .................. 154

Determination of DNA Synthesis ................................. 155

Preparation of Cell Membranes ................................... 155
Purification of CMP-Sialic Acid:

Lactosylceramide «2—3 Sialyltransferase ......................... 156

Assays ................................................................ 15 6

Protein Concentration ..................................... 156

SAT-1 Activity Assay ..................................... 156

GM3 Sialidase Activity Assay ............................ 157

Ganglioside Analysis ...................................... 157

Extraction of Gangliosides ....................... 157

Ganglioside Quantitation ......................... 158

Results ................................................................................. 159

Determination of SAT-1 Expression in KB Cells ....................... 159

Turnover of GM Ganglioside, SAT-1 and GM3
Sialidase in KB ells Following Release
Butyrate Treatment ................................................. 159

Determination of DNA Synthesis ................................. 159

Morphological Changes Induced by Butyrate and
Thymidine Blocks in Monolayer Cultures

of KB Cells .......................................................... 162
Regulation of SAT-1 Activity in KB Cells by
Phosphorylation .............................................................. 162
SAT-1 Activity in Synchronized KB Cells ...................... 162
Immunoaffinity Purificaiton of KB Cell SAT-1 ................ 165
Immunodetection of KB Cell SAT-l With
Anti-Phosphotyrosine Monoclonal ................................ 165
Discussion ............................................................................. 173
Footnotes ............................................................................... 177
Acknowledgements ................................................................... 178

Abbreviations .......................................................................... 179

References ............................................................................. 180
Chapter 7. Discussion ...................................................................... 183
History and Perspectives ............................................................. 183
Considerations on the Regulation of SAT-1 ...................................... 186
Protein Sequencing of SAT-1 ....................................................... 194
Analysis of CNBr Digest ................................................... 197

Analysis of Tryptic Digests ................................................ 204

Cloning of cDNA Encoding SAT-1 ................................................ 204
Intracellular Loeation of SAT-1 .................................................... 209
Closing Statement .................................................................... 211
References ............................................................................. 212
Appendix A ................................................................................. 215

Report of Laboratory Examination - Gross Necropsy of

Rabbit Employed in the Production of Polyclonal

Antibodies to SAT-1 from the Animal Health Diagnostic

laboratory ............................................................................. 216

Chapterl

Table 1.
Table 2.

Chapter 2

Table 1.

Table 2.

Chapter 3.
Table 1.
Table 2.
Table 3.
Table 4.

Chapter 4
Table 1.

Chapter 5

Table 1.
Table 2.

Table 3.

Chapter 7

Table 1.

LIST OF TABLES

Summary of Ganglioside Nomenclature .................................... 9
Glycolipid Glycosyltransferase Nomenclature ............................ 16

Recovery of [3H]-GM3 from Sep Pak Chg
Cartridges. The Effect of Phosphatidylc oline

on Recovery. .................................................................. 50
Recovery of Other Gangliosides from SAT-1

Reaction Mixtures from Sep Pak C13 Cartridges. ....................... 57
Affinity Purificatrion of SAT-1 ............................................. 82
Immunoaffinity Purification of SAT-1 ..................................... 92
SAT-1 Specificity ............................................................. 94

Glycan Differentiation of SAT-1 Carbohydrate
Sidechains ...................................................................... 95

Stabili of LDAO Solubilized SAT-1 in

15%(v v) LDAO at -200C for 6 Months ................................. 120
Size Hertogeneity of Sialyltransferase .................................... 135

Effects of Protease Inhibitors on
Immunoaffinity-Purified SAT-1 Activity ................................. 137

Effects of Protease Inhibitors on SAT-1
Activity in Golgi-Enriched Microsomes .................................. 139

Phosphatidyl Inositol as a Potential
Membrane Anchor for SAT-l .............................................. 193

xiv

Chapter 1
Figure 1.
Figure 2.
Figure 3.
Figure 4.
Figure 5.

Figure 6.
Figure 7.

Chapter 2
Figure 1.

Figure 2.

Chapter 3

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5 .
Chapter 4
Figure 1.

Figure 2.

LIST OF FIGURES

The Chemical Structure of GM3 Ganglioside .............................. 4
Biosynthesis of Glycolipid Oligosaccharide ................................. 7
Synthesis of GM3 Ganglioside by CMP-Sialic ........................... ll
Biosynthetic Pathway of Glycosphingolipids ............................. 15

Topography and General Structure of
Glycosyltransferases within the Golgi Apparatus. ....................... 18

Molecular Species Formed Metabolically from
GMiiGanglioside Involved in Cell Growth Regulation
Via odulation of Signal Transduction Mechanism. .................... 25

Hypothetical Model of SAT-1 Regulation ................................. 29

Elution Profile of Radiolabeled SAT-1
Reaction Components on Sep Pak C13 Cartridges. ...................... 52

Enzymatic Synthesis of [14C]-GM3 from

CMP-[l4C]-NeuAc and LacCer as a Function of Time
and Protein ..................................................................... 56

Chemical Synthesis of LacCer-aldehyde

Sepharose ...................................................................... 73
Gel Electrophoresis of Affinity-Purified

SAT-1. ......................................................................... 85
Inhibition of SAT-1 activity by MlZGC7

monoclonal antibody. ........................................................ 89
Assessment of Immunoaffinity-Purified SAT-1 ’

Homogeneity by SDS-PAGE Gel Electrophoresis. ...................... 91
Analysis of SAT-1 by 2D SDS-PAGE ..................................... 98

Two-Dimensional SDS-PAGE of Detergent-
Solubilized Rat Liver Golgi Proteins ...................................... 116

Effects of Various Surfactants on SAT-l
Activity in Rat Liver Golgi. ................................................ 118

XV

Figure 3a and 3b. Reverse-phase C8 HPLC of SAT-1 vs.

Chapter 6
Figure 1.

Figure 2.

Figure 3.

Figure 4.
Figure 5.

Figure 6.

Chapter 7
Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Reference Buffer Containing 5 % (v/v)
Iauryldimethylamine Oxide (LDAO). .................................... 124

KB Monolayer Culture Metabolism of GM3
Ganglioside ................................................................... 161

Morphological Changes Induced by Butyrate

and Thymidine Blocked Monolayer Cultures of KB

Cells. .......................................................................... 164
SAT-1 Activity Associated with KB cells ................................ 167
Immunopercipitation of SAT-l from KB Cells .......................... 169
Immunodetection of Purified KB Cell SAT—1

by an Antiphosphotyrosine Monoclonal Antibody ...................... 171
Proposed Model of Cell Growth Regulation ............................. 176

Immunodetection of SAT-1 with

Antiphosphotyrosine Monoclonal Antibody .............................. 189
Specificity of the Immunodetection of SAT-1

with Antiphosphotyrosine Monoclonal Antibody. ...................... 191
Reverse-Phase C8-HPLC of Tryptic Digestion of

SAT-1 Tryptic Peptides ..................................................... 196
Reverse-Phase C8-HPLC Following CNBr

Digestion of SAT-1. ........................................................ 199

Reverse-Phase C8-HPLC of SAT-1 and Buffer
containing LDAO. ........................................................... 201

Reverse-Phase C8-HPLC of SAT-1 Following 2D
SDS-PAGE and CNBr Digestion. ......................................... 203

Reverse-Phase C8-HPLC of Tryptic Peptides
from Deglycosylated SAT-1 ................................................ 206

Primary Screen of a cDNA Expression Library
with Rabbit anti-SAT-l Polyclonal Antibodies. ......................... 208

xvi

CHAPTER 1
LITERATURE REVIEW

Introduction

The regulation of cellular growth and differentiation is of fundamental
importance to most biological investigations, especially those mechanisms involved in
oncogenic transformation. Research during the past decade has implicated a functional
role for glycosphingolipids in these phenomena (1-5). Glycosphingolipids comprise a
minor proportion of the complex carbohydrate-containing moieties of the plasma
membrane forming the glycocalyx; yet, they are major contributors to the
communication network modulating cellular activities. Glycosphingolipids serve as
membrane receptors for bacteria (6-8) viruses (9,10), toxins (1 1-14), acetylcholine
(15), interferon (16), and fibronectin (17-19)). Glycosphingolipids also act as
mediators of cellular interaction (19,20), and Na+ transport (21), and interact with the
cell substratum (17,22) and as modulators of membrane receptor-mediated signal
transduction mechanisms (25-31), indicating a role for glycosphingolipids in the
regulation of cell proliferation. Further, shedding of gangliosides by tumors leads to
enhanced tumor formation, possibly through suppression of the immune response
(32,33; for an early review, 34). With increasing diversity in the functional role of
glycosphingolipids, specifically gangliosides, research in the glycoconjugate field has
three foci: (a) gangliosides as receptors for ligands and mediators in cell—cell
interactions; (b) gangliosides as modulators of immune regulation; and (c) gangliosides
as modulators of signal transduction mechanisms and cell growth.

The diversity of these functions reflects the structural diversity among the
glycosphingolipids. About 300 glycosphingolipid structures have been elucidated (35).
However, the functional studies cited above are but a sampling of the potential

modulatory roles played by these glycolipids. The functions performed by the

glycosphingolipids may be as diverse as the structures themselves and the cells which
express them. The physiological role of these glycoconjugates is not completely
understood. Thus, there is great impetus for further investigation of the functional role
played by glycosphingolipids in the modulation of cellular activities and the

mechanisms which control their biosynthesis.

Ganglioside Structure

The different classes of glycolipids contain different oligosaccharide structures
and are tissue- and species-specific (36). Gangliosides are sialic acid-containing
glycosphingolipids. The sialic acid1 confers a net negative charge on the molecule and
contributes to its biological activity (37). Gangliosides are amphipathic molecules

composed of the hydrophilic carbohydrate head group and the hydrophobic lipid tail
2

amide-linked to a long-chain sphingoid base3. The ceramide participates in the

(Figure 1). Their lypophilic residue, ceramide , is composed of a fatty acid which is
formation of the outer leaflet of the plasma membrane bilayer. The carbohydrate
sidechain is positioned out into the extracellular matrix. This strategic orientation
defines the cell surface and contributes to the multi-functional role of gangliosides in
cellular intercellular communication and cell growth.

Glycosphingolipids are structurally unique relative to the glycoproteins (for
review, 38) and proteoglycans (for review, 39), two other major families of cell-
surface glycoconjugates. Gangliosides, and, in fact, glycolipids in general are unique
glycoconjugates in that they contain only one oligosaccharide chain, whereas

glycoproteins and proteoglycans contain several such chains as is the case with

 

lSialic acid is the generic name for neuramicinic acid.
2Ceramide is the trivial name for N-fatty acid acyl sphingosine.

3Sphingosine is the name given to D-erythro—l ,3-dihydroxy-2-(alkylacetamide)-4,5-
trans-octadecene or its analogs.

Figure 1. The Chemical Structure of GM3 Ganglioside.

MG _ Z<¢mo m0 _ ¢<IUO<WO@_ I-O

_.||||l||_||||IlJ

G_0< 0..._<_W
mu.mrflc

a _.l.||I_

_
ckuxoTIou co cknxoTnm

:ooo zxmo o:
o
o :o
:0
con :IO :0
:0 N5 :onxoo

O

:3 in 20 in no we o o<z=ozv
359.62% «.55

proteoglycans, glycolipids can also be sulfated. The oligosaccharide is O-linked to the
ceramide through Gch (globo, ganglio, sialoganglio, lacto, and neolacto series) and
sometimes GalB (series) (Gala series and sulfatides) (See Figure 2). These O—linked
saccharides differ from O- and N-linked oligosaccharides of glycoproteins in that the
carbohydrate Sidechains are typically through a GalNAca to serine or threonine
residues (O—linked) and asparagine (N-linked) in glycoproteins (38). Thus, the lactose
(GalBl-4Gchl-1Cer) linkage is unique to the glycolipid structure. It is important
structural differences, such as this that confer the specificity of each particular
glycoconjugate to a wide variety of factors or allow them to serve as substrate for a
specific glycosyltransferase activity during product-precursor synthesis within the

Golgi.

Ganglioside Biosynthesis

Generally, gangliosides are primarily localized on the outer leaflet of the
mammalian cell; however, gangliosides have also been observed to be associated with
internal membranes (40-43). Subcellular distribution studies with rat liver showed that
76.3% of the total gangliosides (by sialic acid) is localized in the plasma membrane,
while lesser amounts reside on internal membrane compartments such as the
endoplasmic reticulum (7.4%), Golgi apparatus (1.0%), mitochondria (10.8%) and
nuclear membrane (4.8%) (43). The remaining soluble ganglioside were found in
supernatant fractions. Individual gangliosides patterns were unique to the membrane
examined. Further, Matyas and Morre (43) observed a high concentration of the
ganglioside biosynthetic enzymes within the Golgi and minor amounts associated with
the endoplasmic reticulum (ER). This was one of the first studies to show that
gangliosides are synthesized within the Golgi and then transported, not only to the

plasma membrane, but to the ER and other subcompartmental membranes.

Figure 2. Biosynthesis of Glycolipid Oligosaccharide Linkages Specifying the
Different Carbohydrate Series.

a” HE“ 80¢.H

 

 

 

~87§§A28T§$2$§8
)
humane oeonqouz humane oHaozee
HEB
$T§§A3§Afl$b§é38 Hoozmgozsmozuogg
35/ L >
mnoHMBOBOdq nunmum omoqo
than 2358.2; geeagafio 325-3525
“moaiaofiviagigz A, onsets [Sign £58.39 H8485
l. we Hgoqwnoaoaa
humane 4.20 .. ..
“onASmovASmoimaé AIM“! 83.33.35 $2 :33 ~28
WMHQHBANNQDMW H687 a Hglmg
events.” All. cases seas

mg
Hg/ karma:
mag

32% 58 mag aagas 8 flan—258m

Gangliosides4 (nomenclature used is summarized in Table l) are synthesized by
the stepwise transfer of monosaccharides from sugar nucleotides to growing glycolipid
acceptors (44), as shown schematically in Figure 3 for the synthesis of GM3
ganglioside by CMP-sialic acidzlactosylceramide Sialyltransferase (SAT-1). This series
of glycosylations occurs largely within the Golgi apparatus where specific
glycosyltransferases are localized (45). Little is known about the compartmentalization
of the glycosyltransferases though it has been
speculated that their distribution within the Golgi is related to the sequential order of
addition of carbohydrates in the ganglioside biosynthetic pathway (for review, 46).

Recently, Ghidoni and coworkers (47-49) have begun detailed investigations on
the subcellular biosynthesis and transport of gangliosides in rat liver and the enzymes
which catalyze these glycosylations. They have found that the distribution of
glycosyltransferases for the monosialo (GM3->GM2->GM1->GDla catalyzed by
GalNAcT-l, GalT-3 and SAT-4, respectively) and disialo (GD3->GD2->GD1b-
>GT1b->Gle catalyzed by GD3-GalNACT, GDz-GalT, Gle-SAT, and (3le-
SAT, respectively) pathways of ganglioside biosynthesis are differentially localized in
rat liver Golgi apparatus by the order in which they act and that the
glycosyltransferases of monosialo and disialo pathways co-distribute (48). Further,
there exists a precursor-product relationship involved in ganglioside metabolism (49)
and once formed within the Golgi, a ganglioside is in part made available for transport
to the plasma membrane (or other internal membrane), a rapid turnover event, and in
part serves as a precursor to subsequent chain elongation forming more complex
ganglioside species (47).

The mechanisms through which gangliosides are transported to target

membranes within the cell remain to be elucidated. An early proposal for transport of

 

4Gangllirgside nomenclature is that of Svennerholm (144) as recommended by IUPAC-
IBU ( 4 ).

Table 1

Summary of Ganglioside Nomenclature

 

 

Ganglioside Structure
precursors
GlcCer Gch l- 1 Cer
LacCer Gall! 1 -4Gchl-1Cer
GM3 NeuAca2-3GalB l-4Gchl-1Cer
a-series
GM2 GalNAcB l -4(NeuAca2-3)GalB1-4GlcCer
GMl Gal ll 1 -3GalNAcB 1 -4(NeuAca2-3)Galli l -4GlcCer
GD la (NeuAca2-3)GalBl-BGalNAcB1—4(NeuAca2—3)GalB1-4GlcCer
GT1 a (NeuAca2-8NeuAca2-3)Gal B 1 ~36alNAcB l -4(NeuAca2-3)GalB l -4GlcCer
b—series
GD3 NeuAca2-8NeuAca2-3GalB l —4GlcCer
GD2 GalNAcB l-4(NeuAca2-8NeuAca2-3)GalB1~4GlcCer
GD 1b GalBl—4GalNAcB1-4(NeuAca2-8NeuAca2-3)Galli1-4GlcCer
Gle (NeuAca2-3)GalBl-4GalNAcB1-4(NeuAca2-8NeuAca2-3)GalB1-4GlcCer
GQ 1b (NeuAca2-8NeuAca2-3)GalBl-4GalNAcB1—4(NeuAcaZ-8NeuAca2-3)LacCer
c-series
GT3 NeuAca2-8NeuAca2-8NeuAca2-3Galli1-4GlcCer
GT2 GalNAcB l -4(NeuAcoz2-8NeuAca2-8NeuAca2-3)GalB1-4GlcCer
GT1c GalB 1-3GalNAcB l -4(NeuAca2-8Neu Aca2-8NeuAca2-3)GalB1-4GlcCer
GQ lc NeuAca2-36alB 1-3GalNAcB l -4(NeuAca2-8NeuAca2-8NeuAca2-3)LacCer

 

Figure 3. Synthesis of GM3 Ganglioside by CMP—Sialic AcidzLactosylceramide

a2-3 Sialyltransferase

10

02C“ OH CH2 OH

LACTOSYLCERAMIDE
(Gal [31 -4 Glc [31 - 1 Cer)

CH CH CMP-NeuNAc
2 k 6M3 SYNTHASE(ST-1)

OH NH
0H 3 2
o_p._o \N
OH I I [K
OH CH2 N/Lo

H,e-<u:HN coou w Y CMP
o
0
0H 0H -

601-12 oH CHZOH N R
CHo oHH ° ° \ (CH2),20H3
OH
OH

O

 

H c-ngHN 90°“ 6M3 GANGLlOSlDE

° (NeuNAc a 2-3 Gal ,61-4 Glc [31-1 Cer)

R:--(CH2)n or -CH-(CH2)n.
I .
OH

n=15-25

11

gangliosides to the cell surface was via membrane flow (50,51). However, subsequent
studies with monensin, an inhibitor of membrane flow (52), had no effect on either
neuroblastoma or glioma cells (53). Therefore, this early model has given way to other
eloquent hypotheses. One mechanism suggests regulation of transport through the
modulation of temperature (53) or via lipid transfer proteins (54-57). Possibly a
mechanism(s) exists, as has been described for glycoprotein movements between the
endoplasmic reticulum and Golgi (for review, 58) and within the Golgi cistemae,
namely, vesicular transport (e. g., Golgi-coated vesicles) (59). One new reversible drug
treatment which may prove useful in addressing glycolipid biosynthesis and transport
mechanisms is brefeldin A (BFA), an antifungal metabolite, which blocks intracellular
transport and processing of secretory proteins through disassembly of the Golgi
apparatus (60).

Another interesting observation involving glycolipid biosynthesis in rat liver
Golgi has been the isolation of two asialogangliosides (LacCer- > Gg3- >Gg4- > GM“,—
>GD1C) resulting from the presence of the galactosyl- and N-
acetylgalactosylaminyltransferase activities which are typically quiescent in rat liver
Golgi (61). Assays in vitro of these enzymes gave specific activities and Km values
comparable to those observed for the a and b series gangliosides. The presence of
these glycosyltransferase activities and the absence of their glycosylation products in
normal rat liver leads to interesting speculation as to the mode of regulation of these

and other glycosyltransferases forming the glycosphingolipids.

Regulation of Ganglioside Biosynthesis

The regulation of all glycosphingolipid biosynthesis may reside within the
expression of five glycosyltransferases, specifically, LacCer synthase (62-64), GM3,
GD3 and GT3 synthases (64) and GM2 synthase (63,64). The regulatory model as

12

proposed by Sandhoff and coworkers (62) for ganglioside biosynthesis is illustrated in
Figure 4. As set forth in their model, regulation of the entire ganglioside biosynthetic
pathway is dependent upon GalT galactosyltransferase and by SAT-l, SAT-2 and SAT-
3 sialyltransferases specificity for the unique carbohydrate-lipid bond, Gchl-lCer.
Subsequent glycosylations adding GalNAc, Gal and NeuNAc are parallel reactions
catalyzed by glycosyltransferases (Table 2) recognizing the terminal carbohydrate
structural ”backbone.” This regulation is likely to be at the transcriptional or post-
translational levels, but other factors such as feedback inhibition (65), membrane
fluidity (66), availability of sugar nucleotide as well as its translocation into the lumen
of the Golgi (67,68), protein matrix effects (69), influx of divalent cations (70) and
temperature (71) may also play an important function. However, the molecular details
of ganglioside biosynthesis and functions are still speculative due to the lack of
information on the mechanisms governing the regulation of the glycosyltransferases
involved. This will evolve as these enzymes are purified and their cDNA sequences

are determined.

Topography of Glycosyltransferases

Glycosyltransferases are type-2 membrane (single a-helix) enzymes with a
lumenal orientation within the Golgi (Figure 5) (for reviews, 68,72). Evidence
supporting this type of topography include: complex carbohydrate substrates are
known to have a lumenal orientation; Golgi vesicles must be disrupted or permeabilized
with detergent for activity; glycosyltransferases are glycoproteins; glycosyltransferases
are secretory proteins, being derived from membrane-bound forms through proteolysis.
The lumenal orientation of glycosyltransferases requires a mechanism for the transport
of sugar nucleotides into the Golgi (for review, 68). The proposed mechanism involves

entry of each sugar nucleotide through a specific antiporter (Figure 5). The sugar is

13

Figure 4. Biosynthetic Pathway of Glycosphingolipids.

l4

22559“ 2855605 epic—.930

sooAl. Ho Al «no All fio

I:<m

o. 8
06 a hOAII :5

«ItaO _.I._.o<z_o0

a

Alamo Al no Alcoo

«Itao 7.55200
9 «Jim

aboAl sooAl. :oAl 2o Alaso

mr._.<w I:<m

«IF—ea I:o<z.oO

a

motor. 0

not: a

not: a «23.350

«23.036 0.2: AlgoOomI— All-.5020 All .50

wrtao

15

Table 2. Glycolipid Glycosyltransferase Nomenclature

 

 

Glycosyltransferase Designation Ganglioside
Famed
CMP-NeuAchacCer a2—3 Sialyltransferase SAT-l GM3
UDP—GalNAchM3 B1-4N-acetylgalactosaminy1 GalNAcT-l 6M2
transferase
UDP-GalzGMz 131-3 galactosyltransferase GaJT-3 GM}
CMP-NeuAc:GM1 a2-3 Sialyltransferase SAT-4 GDla
CMP-NeuAchDl a 012-8 Sialyltransferase SAT-5 GTla
CMP-NeuAchMg a2-8 Sialyltransferase SAT-2 GD3
UDP-GalNAchD3 81-4 N-acetylgalactosaminyl GDz-GalNAcT-l GD2
transferase
UDP-Gal:GD2 81-3 galactosyltransferase GDz-GalT Gle
CMP-NeuAc:GD1b a2-3 Sialyltransferase Gle-SAT-4 Gle
CMP-NeuAchle a2-8 Sialyltransferase Gle-SAT-S Gle
CMP-NeuAc:GD3 a2-8 Sialyltransferase GD3-SAT-3 GT3
UDP-GalNAchT3 Bl-4 N-acetylgalactosaminyl GT3-GalNAcT-1 GT2
transferase
UDP-Gal:GT2 31-3 galactosyltransferase GTz-GalT GTlc
CMP-NeuAchTlcaZ-B Sialyltransferase GTlc-SAT-4 GQlc

 

16

Figure 5. Topography and General Structure of Glycosyltransferases within the
Golgi Apparatus.

17

wDF<m<mm<
51.00

2852505 AZ
oijoofio

 

$5432me
o=2w<lioozm

 

2min; 61.00

 

 

 

18

transferred to lumenal glycoconjugate acceptors via specific glycosyltransferases and,
with the exception of CMP-NeuAc, the nucleotide diphosphate is converted to the

monophosphate by nucleoside diphosphatase and shuttled to the cytosol.

Molecular Organization of Glycosyltransferases

The present state of knowledge regarding the molecular nature of
glycosyltransferases comes predominantly from studies on glycoprotein
glycosyltransferases (73-83; for review see 72). The cDNA's have been obtained for
these enzymes. Most mRNA encoding glycosyltransferases are about 4.5 kilobases
(54). They share a common structural homology and about 80% amino acid sequence
homology between species (72,84) permitting antibody and oligonucleotide probes
generated from one species to be used to screen cDNA libraries for clones from another
species. This methodology has permitted the recovery of a 3.4 lcilobase clone for
human Gal a2-6 Sialyltransferase using information from rat liver Gal a2-6
Sialyltransferase (84). Even though there are structural similarities, no apparent
sequence homology within the glycosyltransferase families has been found (72). The
region expected to be the most homologous is the sugar nucleotide binding site. A
common hexapeptide among cloned transferases has been proposed to be the region
coding for the UDP-hexose binding site (82) of bovine Gal a1-3 galactosyltransferase
and bovine GlcNAc 814 galactosyltransferase. A similar site on the C-terminal end of
bovine GlcNAc al-4 galactosyltransferase and a human galactosyltransferase was also
found (74,75). It has been suggested (72) that more sequence homologies will be
found once more glycosyltransferase genes are identified.

Elucidation of the structural similarities among these cloned
glycosyltransferases, as determined from deduced amino acid sequences (for review,
72), indicate that they all possess a short 6-41 amino acid cytosolic NH2 terminus, a

16-20 amino acid transmembrane domain and a lumenal domain containing a stem

19

region and the catalytic site (Figure 5). It is the stem region which is the least
homologous for the cloned glycosyltransferases (74-83; for review, 72), and which is
the proposed site of proteolytic degradation of these enzymes (72,73,85). A cathepsin
D-like activity has been associated with this endogenous proteolytic processing,
releasing a soluble catalytically-active enzyme from the mature membrane-associated
form (86). The mechanism and the regulation of this proteolytic activity are yet to be
resolved. Further, the stem region seems to play a role in the targeting of the enzyme
within the Golgi as has been identified for Gal a2-6 Sialyltransferase by the use of
altered sequences and fluorescent antibody detection (85). Other glycosyltransferases
have also been localized to the Golgi network (85,87-90). Further, these Golgi

membrane-associated glycoprotein transferases are glycosylated (76).

Purification of Glycosphingolipid Glycosyltransferases

Little information regarding glycosphingolipid (GSL) glycosyltransferases is
known because most current studies in the GSL field are still focussing on the
purification of these enzymes from the Golgi. Presumably, it is the nature of the lipid
substrate, the membrane environment surrounding the enzymes, instability of the
purified GSL glycosyltransferase, and endogenous proteolytic processing which have
lead to difficulties in advancing the state of knowledge of glycosyltransferases in the
GSL field. Thus, there exists a need to purify the various GSL glycosyltransferases
and isolate the genes encoding these enzymes to determine their role and that of their
products (glycolipids) in the regulation of cell growth and differentiation, thereby
defining the role of gangliosides in the transformation process.

Some progress in this regard has been made. Employing standard protein
purification strategies, a few GSL glycosyltransferases have been purified (82,91-97)

and of these only one has been successfully cloned and sequenced (94).

20

Studies on the regulation of purified and cloned Galch3 81-4
galactosyltransferase (GalT-4) and FuchMl a1-3 fucosyltransferase (FucT-3) from
human colon carcinoma Colo 205 cells and P-1698 mouse lymphoma cells (94) indicate
a role for sphingosine in the regulation of GSL glycosyltransferases. Sphingosine
inhibits GalT-4 enzyme activity at low concentration while at high concentration FucT-
3 is inhibited. Apparently, sphingosine interacts with a specific octameric amino acid
sequence within the catalytic domain of the glycosyltransferase.

Recently, through the efforts of Lowe and coworkers (98-103) the gene
sequences of three more glycosyltransferases are known. These are a1-3GalT, a1-
2FucT, and a1-3/1-4 FucT, the galactosyl- and fucosyltransferases responsible for the
synthesis of the blood group B, H and Le antigens, respectively. By employing an
expression cloning by gene transfer approach, they have cloned, sequenced and
expressed genomic and cDNA sequences of the blood group H antigen (100-102) and
the Le determination (103) into mouse LM fibroblasts to define the gene structure and
study the complex level of regulation required for the human blood group H a1-2
fucosyltransferase gene and Le antigen fucosyltransferase gene. Multiple functional
transcripts have been generated with potential to encode either the transmembrane or
catalytic domains. These colinear transcripts differ at the 5' ends suggesting a multiple
polypeptides can be generated from a single fucosyltransferase gene (100-103).
Presumably other glycosyltransferase genes, perhaps existing as a gene family, are

regulated by switching on and off their expression.

Regulation of Cell Proliferation by Glycosphingolipids.

The cell surface has been postulated to play an important role in the regulation
of cell growth and differentiation (4,30,104,105). Gangliosides, ubiquitous sialic acid-
containing glycosphingolipids of mammalian plasma membranes (106), have been

(implicated in these phenomena (4). Changes in ganglioside composition and

21

metabolism have been associated with oncogenic transformation (1); during cellular
transformation, there is often an accumulation of precursor (less complex) gangliosides,
compared with nontransformed cells, due to incomplete synthesis that results from a
block or impairment of the ganglioside biosynthetic pathway. Alternatively, there may
be synthesis of novel gangliosides (enhanced synthesis of the neoglycolipids i.e., the Le
antigens) due to activation of specific quiescent embryonic or fetal glycosyltransferases
(2,4,107).

Several lines of evidence suggest a specific role for glycosphingolipid
composition and metabolism in oncogenic transformation. Exogenous addition of
gangliosides can either stimulate or inhibit cell growth depending on the cell type.
Specifically, Gle, a polysialylganglioside, promotes neurite outgrowth (for review,
109) by stimulating cell surface protein phosphorylation via an ecto-protein kinase
activity (110). Addition of GDla, a disialylganglioside, activates Ca2+lcalmodulin-
dependent kinase activity (111,112) have shown that exogenously added GM1
monosialylganglioside inhibits CAMP-dependent protein kinase (cAK) of histone IIA
and autophosphorylation of cAK, but stimulates cyclic nucleotide phosphodiesterase
(PDE). Other studies have shown that a decrease of GM3 or GM1 ganglioside in
several lines of fibroblasts has been associated with oncogenic transformation of these
cells. Addition of exogenous GM3 or GMl restores normal contact-inhibited cell
growth (3) by altering the binding affinity of the cells to growth factors such as PDGF,
EGF and FGF (24,25). GM3 functions to inhibit cell proliferation by inhibiting the
autophosphorylation of EGF-R tyrosine kinase (24,25). Further, chemical compounds
such as retinoic acid (113-115), tumor-promoting phorbol esters (113,116-120) and
sodium butyrate (121-126) have been shown to modulate specific glycosyltransferases
and cell growth and differentiation. Specifically, retinoic acid, phorbol-lZ-myristate-
13-acetate (PMA) and sodium butyrate stimulate CMP—sialic acidzlactosylceramide
sialyltransferase (SAT-1) activity 5- to 10-fold (113,120,124-126).

22

GM3 in the Regulation of Cell Growth

An 'allosteric regulation" model for the interaction of GM3 with the EGF-
receptor has been proposed by Schlessinger in 1983 (80) and supported by Hakomori in
1984 (3), in which the growth factor receptor has multiple loci for growth factor
binding, protein kinase activity and gangliosides or other components serve to mediate
the effect. Specifically, GM3 ganglioside inhibits EGF-stimulated phosphorylation of
the EGF receptor without affecting the binding. EGF binding to its receptor is a
necessary pre-replicative factor controlling commitment of cells for DNA replication.
Thus, enrichment of specific gangliosides, like GM3, in the plasma membrane
indirectly or directly affects the receptor functions and possibly receptor-receptor
interaction.

GM3 ganglioside has been demonstrated to interact directly with the EGF-R.
Immunoprecipitation of the EGF-R with a specific monoclonal antibody co-precipitated
GM3 along with the receptor (26,27). GM3 was demonstrated to inhibit the
autophosphorylation of EGF/EGF-R complex by interfering with the EGF-dependent
dimerization of the receptor (27). Further, these researchers could enhance the effect
on GM3 inhibition of EGF-dependent tyrosine kinase activity in the presence of lyso-
PC (27).

The degradation of cell-surface GM3 ganglioside to LacCer by endogenous
Sialidase activity has been demonstrated to stimulate cell growth (128,129), presumably
through the release of the inhibition of the EGF-dependent dimerization of the EGF-R;
autophosphorylation continues and cell growth is stimulated (130). The fate of the
LacCer may be further degraded by glycosidases or recycling of LacCer back to the
endoplasmic reticulum and Golgi for resynthesis of GM3 (129-131). The effects of

23

Figure 6. Molecular Species Formed Metabolically from GM 3 Ganglioside
Involved in Cell Growth Regulation Via Modulation of Signal Transduction
Mechanism.

24

. mz _ m8: _ Emizhmfoi i m: _ moo: _ Em

:0 .209. :80:
T n a.
{023.23 :0 35.
z
.xo .5

    

2.2 a c
:0
_
{xorzuio :foTi

.xoo:

.x......§\/.\.a/...\ “9.2me

we _ :gwofimgofi

. a a :0
u
:0 .20 o 0:

”0-20 5.20

so 8: g 55;: .33 z...
zxmo... :0
.,...._Q ,.
a :- \leox “W .2031xo.\/\F1/Q0 GNU: rem—:0
xu .xu.\/\§U.\/o oo o 0.23.20 20:6
0-: 20.5

58:26 n5

:0
x xo
.6 .20
n 2. o
:o :6. o o oxxofo
¢/.\z xoax
o

xozxmdax

25

other naturally occurring GM3 catabolites (see Figure 6 for their structures), lyso GM3
(26,27) and de-N-acetyl-GM3 (28), have been investigated. In A431 cells, lyso-GM3
was found to inhibit cell growth by dramatically inhibiting PK-C and EGF-R signal
transduction (26,27). The effect of exogenous addition of de-N-acetyl-GM3 was
opposite to the inhibition of cell growth by GM3 and lyso-GM3. De-N-acetyl-GM3
stimulated EGF-dependent EGF-R tyrosine kinase activity (28).

Complete enzymatic degradation of GM3 by glycosidases and ceramidase yields
sphingosine. The effect of sphingosine on protein kinase C (PK-C) and EGF tyrosine
kinase is being investigated, because of a suggested role for sphingosine as a modulator
of transmembrane signaling (132-135; for review, 136). Observations including those
by Igarashi and Hakomori (137), who have recently shown N ,N-dimethyl enzymatic
derivatization of erythro-sphingosine to give a strong inhibitor of PK-C, whereas the D-
erythro unsubstituted form is not. They have indicated that this effect is different from
that observed for the EGF tyrosine kinase. In a related study, the inhibition of PK-C
by sphingosine was found to be pH dependent; inhibition of PK-C required a positive
charge (138).

Regulation of GM3 Synthesis

In consideration of the down-regulation by exogenous GM3 on the
autophosphorylation of tyrosine on EGF-R, a similar role for endogenous GM3 in
regulation of cell growth has been proposed (131). GM3 is a naturally occurring
constituent of most mammalian cell membranes. The amount of cell surface GM3 and
the level of CMP sialic acidzlactosylceramide sialyltransferase activity is cell-cycle
dependent and is maximal at late M/early G] of the cell cycle (124,126). One
proposed mechanism for increased sialyltransferase activity may involve
phosphorylation/dephosphorylation (139). PMA, which can bind to and stimulate PK-
C activity (140,141), stimulates SAT-1 activity in cultured cells (113, 120,140).

26

Further, Burczak et al. (139) have shown that SAT-l is activated following treatment
with a CAMP-dependent protein kinase (cAK). In studies on the regulation of GM3-
GalNAcT-l (GM2 synthase) by Dawson et al. (142), showed the activity of this N-
acetylgalactosaminyltransferase was stimulated by cAK-activators. Further,
phosphoproteins and an ATP transport mechanism have been identified within the
Golgi (143). Accordingly, the regulatory mechanism for the synthesis (see chapter 6)
and degradation (128,129,131) of GM3 becomes paramount to the role of GM3
ganglioside in the modulation of cell growth and oncogenic transformation. As such,
the purification and investigation of the mechanism of regulation of SAT-l activity was
the focus of this dissertation research project. Based on the aforementioned data,
recognizing GM3 as a key modulator of cell growth, a working hypothetical model for
the regulation of SAT-l is presented (Figure 7).

27

Figure 7. Hypothetical Model of SAT-l Regulation. This schematic diagram
summaries reported data on the regulation of cell growth by GM3 ganglioside and its
catabolites and the potential of phosphorylation as a regulatory mechanisms for this key

enzyme of ganglioside biosynthesis.

28

 

Ill“

5023 oEmmujoE
jwo amazfizool

 

 

Pkmoawz<mkh

 

w

$335 ”so V5.)

 

« ”so“

..).. wzo_.r<mm._.z_ m<qnnjmo

3% E m no ..mooz. 38.51.59:

 

 

 

or

29

9'95”.“

10.

11.
12.
13.

14.
15.
16.
17.

18.

REFERENCES

Hakomori, S. (1981) Ann Rev. Biochem. 50, 733.

Hakomori, S. and Kannagi, R (1983) J. Natl. Cancer Inst. 71, 231-251.
Hakomori, S. (1984) Annu. Rev. Immunol. 2, 103-126.

Hakomori, S. (1985) Cancer Res. 45, 24-5-2414

Ledeen, R.W. (1989) in Neurobiology of Glycoconjugates (Margolis, R.U. and
Margolis, R.K., eds.) Plenum Press, NY, p.p.43-83.

Kirivan, H.C., Ginsburg, V. and Roberts, DD. (1988) Arch. Biochem.
Biophys. 260, 493-496.

Karlsson, K.-A. (1989), Annu. Rev. Biochem. 58, 309-350.

Deal, CD. and Krivan, H.C. (1990) J. Biol. Chem. 265, 12774-12777.
Haywood, A.M. (1974) J. Mol. Biol. 83, 427-436.

Markwell, M.K., Svennerholm, L. and Paulson, J .C. (1981) Proc. Natl. Acad.
Sci., USA. 78, 5406-5410.

Clowes, A.W., Cherry, R.J. and Chapman, D. (1972) J. M01. Biol. 67, 49-57.
Critchley, DR. and Vicker, MG. (1977) Cell Surface Rev. 3, 307-370.

Van Heyningen, W.E., Carpenter, C.C.J., Pierce, NP. and Greenough, W.B.
(1971) J. Infect. Dis. 124, 415-418.

Fishman, RH. (1982) J. Membrane Biol. 69, 85-97.

Cuatrecasas, P. (1974) Annu. Rev. Biochem. 43, 169-214.

Besamcon, F. and Ankel, H. (1974) Nature 252, 478-480.

Kleinman, H.K., Martin, GR. and Fishman, PH. (1979) Proc. Natl. Acad.
Sci., USA. 76, 3367-3371.

Perkins , R.M., Kellie, S., Patel, B. and Critchley, DR. (1982) Exp. Cell Res.
141, 231—243.

30

19.

19.

20.

21.

22.

24.

25.

26.

27.

28.

29.

30.

31.
32.

Matyas, G.R., Evers, D.E., Radinsky, R. and Morre, DJ. (1986) Exp. Cell
Res. 162, 296-318.

Clausen, H., Levery, S.B., Nudelman, E., Baldwin, M. and Hakomori, S.
(1986) Biochem. 25, 7075-7085.

Tiemeyer, M., Swank-Hill, P., and Schnaar, R.L. (1990) J. Biol. Chem. 265,
11990-11999.

Spiegel, 3., Handler, J.S. and Fishman, P.H. (1.986) J. Biol. Chem. 261,
15755-15760.

Cheresh, D.A., Pierschbacher, M.A., Herzig, M.A. and Mujoo, K. (1986) J.
Cell Biol. 102, 689-696.

Bremmer, E. G., Hakomori, S., Bowen-Pope, D.F., Raines, E. and Ross, R.
(1984) J. Biol. Chem. 259, 6818-6825.

Bremmer, E.G., Schlessinger, J. and Hakomori, S. (1986) J. Biol. Chem. 261,
2334-2340.

Hanai, N. Nores, G., Torres-Mendez, C.R., and Hakomori, S. (1987)
Biochem. Biophys. Res. Commun. 127, 134.

Hanai, N., Nores, G.A., MacLeod, C., Torres-Mendez, C.-R. and Hakomori,
S. (1988) J. Biol. Chem. 263, 10915-10921.

Hanai, N., Dohi, T., Nores, GA. and Hakomori, S. (1988) J. Biol. Chem.
263, 6296-6301. _

Igarashi, Y., Kitanura, K., Toyokuni, T., Dean, B., Fenderson, B.A., Ogawa,
T., and Hakomori, S. (1990) J. Biol. Chem. 265, 5385-5389.

Hakomori, S. (1990) J. Biol. Chem. 265, 18713-18716.

Weis, F.M.B. and Davis, R.J. (1990) J. Biol. Chem. 265, 12059-12066
Ladisch, S. (1989) in Gangliosides and Cancer (Oettgen, H.F., ed.) VCH
Publishers, New York, NY, p.p. 217-238.

31

33.

34.
35.

36.

37.

38.
39.

40.

41.
42.

43.
44.
45.
46.
47.
48.

Portoukalian, J. (1989) in Gangliosides and Cancer (Oettgen, H.F., ed.) VCH
Publishers, New York, NY, p.p. 207-216.

Marcus, MD. (1984) Molec. Immunol. XXX, 1083-1091.

Stults, C.L.M., Sweeley, C.C., and Macher, BA. (1989) Methods Enzymol.
179, 167-214.

Sweeley, CC. (1985) in Biochemistry of Lipids and Membranes, (Vance, DE.
and Vance, J.E., eds.) The Benjamin/Cummings Publishing Co., Inc., p.p.
361-403.

Reutter, W., Kottgen, E., Bauer, C. and Gerok, W. (1982) ”Biological
Significance of Sialic Acids" in Sialic Acids: Chemistry, Metabolism and
Function (Schauer, R., ed.) Springer-Verlag Wien, New York.

Komfeld, R. and Komfeld, S. (1984) Annu. Rev. Biochem. 54, 631-664.
Hassell, J. R., Kimura, J. H., and Hascall, V.C. (1986) Annu. Rev. Biochem.
55, 539-567.

Keenan, T.W., Huang, C.M. and Morre, DJ. (1972) Biochem. Biophys. Res.
Commun. 47, 1277-1283.

Keenan, T.W., Morre, DJ. and Huang, C.M. (1972) FEBS Lett. 24, 204-208.
Miller-Prodraza, H., Bradley, R.M. and Fishman, PH. (1982) Biochemistry
21, 3260-3265.

Matyas, GR. and Morre, DJ. ( 1987) Biochim. Biophys. Acta 921, 599-614.
Roseman, S. (1970) Chem. Phys. Lipids 5, 270-297.

Keenan, T.W., Morre, DJ. and Basu, S. (1974) J. Biol. Chem. 249, 310-315.
Komfeld, R. and Komfeld, S. (1985) Ann. Rev. Biochem. 54, 631-664.
Trinchera, M. and Ghidoni, R. (1989) J. Biol. Chem. 264, 15766-15769.
Trinchera, M., Pirovano, B. and Ghidoni, R. (1990) J. Biol. Chem. 265,
18242-18247.

32

49.

50.
51.

52.

53.

54.
55.

56.

57.

58.

59.

61.

62.

63.

Trinchera, M., Vallenzasca, C., Ghidoni, R. and Tettamanti, G. (1990) FEBS
Letters 260, 23-26.

Fishman, RH. and Brady, RD. (1976) Science 194, 906-915.

Morre, D.J., Kartenbeck, J. and Franke, W.W. (1979) Biochim. Biophys. Acta
559, 71-152.

Morre, D.J., Boss, W.F., Grimes, J. and Mollenhauer, H.H. (1983) Eur. J.
Cell Biol. 30, 25-32.

Miller-Podraza, H. and Fishman, RH. (1984) Biochim. Biophys. Acta 804, 44-
51.

Yamada, K. and Saski, T. (1982) J. Biochem. 92, 457-464.

Abe, S., Yamada, K. Sakagami, T. and Sasaki, T. (1984) Biochim. Biophys.
Acta 778, 239-244.

Wong, M., Brown, R.E., Barenholz, Y. and Thompson, TE. (1984)
Biochemistry 23, 6498-6505.

Kadowaki, H., Symanski, LA. and Koff, RS. (1988) J. Lipid Res. 29, 52-62.
Balch, W.B. (1989) J. Biol. Chem. 264, 16965-16968.

Rothman, J .E. and Orci, L (1990) FASEB J. 4, 1460-1468.

Shite, S., Seguchi, T., Mizoguchi, H., Ono, M., and Kuwano, M. (1990) J.
Biol. Chem. 265, 17385-17388.

Pohlentz, G., Klein, D., Schmitz, D., Schwarzmann, G., Peter-Katalinic, J.
and Sandhoff, K. ( 1988) Biol. Chem. Hoppe-Seyler, 369, 55-63.

Pohlentz, G., Klein, D., Schwarzmann, G., Schmitz, D. and Sandhoff, K.
(1988) Proc. Natl. Acad. Sci. U.S.A. 85, 7044-7048.

Senn, H.-J., Manke, C., Dieter, P., Tran-Thi, T.-A., Fitzke, E., Gerok, W.
and Decker, K. (1989) Arch. Biochem.. Biophys., in press.
Schuz-Henninger, R., Ullmer, E., Prinz, C. and Decker, K. (1989) Arch.

Biochem. Biophys., in press.

33

65.

67.

68.
69.

70.
71.

72

73.

74.

75.

76.
77.

78.

79.

80.

Yusuf, et al. (1987) Biol. Chem. Hoppe-seyler 368, 455-462.

Sandhoff, K. (1980) Naturwissenschaften 67, 431-444.

Schuz-Henninger, R., Prinz, C. and Decker, K. (1988) Arch. Biochem.
Biophys. 262, 49-58.

Hirschberg, C.B. and Snidem, D. (1987) Ann. Rev. Biochem. 56, 63-87.

Shao, M.-C., Krudy, G., Rosevear, RR. and Wold, F. (1989) Biochem. 28,
4077-4083.

Speigel, S. and Panagiotopoulos, C. ( 1988) Exp. Cell Res. 177, 414-427.
Mitranic, M., Sturgess, J.M. and Moscarello, M.A. (1976) Biochim. Biophys.
Acta 443, 190-197.

Paulson, J.C. and Colley, (1989) J. Biol. Chem. 264, 17615-17618.

Weinstein, J., Lee, E.U., MeEntee, K., Lai, P.-H. and Paulson, J.C. (1987) J.
Biol. Chem. 262, 17735-17743.

Narimatsu, H., Sinha, S., Brew, K., Okayama, H. and Qasba, P. (1986) Proc.
Natl. Acad. Sci. U.S.A. 83, 4720-4724.

Appert, H.E., Rutherford, T.J., Tarr, G.E., Wiest, J .S., Thomford, NR. and
McCroquodale, DJ. (1986) Biochem. Biophys. Res. Commun. 139, 163-168.
Smith, C.A. and Brew, K. (1977) J. Biol. Chem. 252, 7294-7299.

Shaper, N.L., Hollis, G.F., Douglas, J.G., Kirsch, LR. and Shaper, J.H.
(1988) J. Biol. Chem. 263, 1042-10428.

Nakazawa, K., Ando, T., Kimurs, T. and Narimatsu, H. (1988) J. Biochem.
(Tokyo) 104, 165-168.

Masri, K.A., Appert, HE. and Fukuda, MM. (1988) Biochem. Biophys. Res.
Commun. 157, 657-663.

D'Agostaro, G., Bendiak, B. and Tropak, M. (1989) Eur. J. Biochem. 183,
211-217.

34

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.
93.

94.

95.

Shaper, N.L., Shaper, J.H., Meuth, J.L., Fox, J.L., Chang, H., Kirsch, LR.
and Hollis, GP. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 1573-1577.
Joziasse, D.H., Shaper, J.H., van den Eijnden, D.H., van Tunen, A.J. and
Shaper, N.L. (1989) J. Biol. Chem. 264, 14290-14297.

Larsen, R.D., Rajan, V.P., Ruff, M.M., Kukowska-Latello, J., Cummings,
RD. and Lowe, LB. (1989) Proc. Natl. Acad. Sci. U.S.A., in press.

Lance, P., Lau, K.M. and Lau, J .T.Y. (1989) Biochem. Biophys. Res.
Commun. 164, 225-232.

Colley, K.J., Lee, E.U., Adler, B., Browne, J .K. and Paulson, J.C. (1989) J.
Biol. Chem. 264, 17619-17622.

Lammers, G. and Jamieson, J .C. (1988) Biochem. J. 256, 623- 631.

Roth, J. (1984) J. Cell Biol. 98, 399-406.

Roth, J. (1987) Biochim. Biophys. Acta 906, 405-436.

Lee, E.U., Roth, J. and Paulson, J.C. (1989) J. Biol. Chem. 264, 13848-
13855.

Taatjes, D.J., Roth, J., Weinstein, J. and Paulson, J.C. (1988) J. Biol. Chem.
263, 6302-6309.

Taniguchi, N., Yokosawa, N., Gasa, S. and Makita, A. (1982) J. Biol. Chem.
257, 10631-10637.

Taniguchi, N. and Makita, A. (1985) J. Biol. Chem. 259, 5637-5642.
Taniguchi, N., Yangisawa, K., Makita, A. and Naiki, M. (1985) J. Biol.
Chem. 260, 4908-4913.

Basu, M., Weng, S.-A., Das. K.K. and Zang, B.-H. (1989) Joint Meeting of
Society for Complex Carbohydrates and Midwest Connective Tissue Workshop
at University of Michigan, Ann Arbor, MI, November 8-11, 1989.

Gu, X.-B., Gu, T.-J. and Yu, R.K. (1990) Biochem. Biophys. Res. Commun.
166, 387-393.

35

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.
110.

Clausen H., White, T., Taiko, K., Titani, K. Stroud, M. Holmes, E., Karkov,
J. Thim, L. and Hakomori, S. (1990) J. Biol. Chem. 265, 1139-1145.
Melkerson-Watson, LJ. and Sweeley, CC. (1991) J. Biol. Chem., in press
(See Chapter 3).

Larsen, R.D., Rajan, V.P., Ruff, M.M., Kukowska-Latallo, J., Cummings,
RD. and Lowe, LB. (1989) Proc. Natl. Acad. Sci, U.S.A. 86, 8227-8231.
Larsen, R.D., Rivera-Manero, C.A., Ernst, L.K., Cummings, RD. and
Lowe, LB. (1990) J. Biol Chem. 265, 7055-7061.

Rajan, V.P., Larsen, R.D., Ajmera, S., Ernst, L.K. and Lowe, J.B. (19989) J.
Biol Chem. 264, 11158-11167.

Ernst, L.K., Rajan, V.P., Larsen R.D., Ruff, MM. and Lowe, LB. (1989) J.
Biol. Chem. 264, 3436-3447.

Larsen, R.D., Ernst, L.K., Nair, R.P. and Lowe, LB. (1990) Proc. Natl.
Acad. Sci., U.S.A., (personal commun.) in press.

Kukosska-Latallo, J .F., Larsen, R.D., Nair, R.P. and Lowe, LB. (1990) Genes
and Development, (personal commun.) in press.

Spiegel, S. and Fishman, PH. (1987) Proc. Natl. Acad. Sci., U.S.A. 84, 141-
145.

Kreutter, D., Kim, J.Y.H., Goldenring, J.R., Rasmussen, H., Ukomadu, C.,
de Lorenzo, R.J. and Yu, R.K. (1987) J. Biol. Chem. 262, 1633-1637.
Wiegandt, H. (1982) Adv. Neurochem. 4, 149.

Feizi, T. (1985) Nature 314, 53-57.

Hakomori, S. (1989) in Gangliosides and Cancer, lst ed., VCH Publisher,
New York, NY, p.p.57-68.

Ledeen, R.W. (1984) J. Neurosci. Res. 12, 147.

Nagai, Y. and Tsuji, S. (1989) in Gangliosides and Cancer, lst ed., (Oettgen,
H.F., ed.) VCH Publishers, New York, NY, p.p. 33-39.

36

111.

112.

113.

114.

115.

116.

117.

118.
119.

120.

121.
122.

123.

124.

125.

Goldenring, J.R., Otis, L.C., Yu, R.K. and De Lorenzo, R.J. (1985) J.
Neurochem. 44, 1229-1234.

Yates, A.J., Walters, J.D., Stock, S.M., Mcllroy, B.K. and Johnson, JD.
(1989) in Gangliosides and Cancer, lst ed., (Oettgen, H.F., ed.) VCH
Publishers, New York, NY, p.p. 41-56

Moskal, J.R., Lockney, M.W., Marvel, C.C., Trosko, J.E. and Sweeley, C.C.
(1987) Cancer Res. 47, 787-790.

Deutsch, V. and Lotan, R. (1983) Exp. Cell Res. 149, 237-245.

Lotan, R., Lotan, D. and Meromsky, L. (1984) Cancer Res. 44, 5805-5812.
Payatte, R., Biehl, J., Toyama, T. and Holtzer, H. (1980) Cancer Res. 40,
2465-2474. .

Koeffier, H.F., Bar-Eli, M. and Territo, M.C. (1981) Cancer, Res. 41, 919-
926.

Fukuda, M. (1981) Cancer Res. 41, 4621-4628.

Huberman, E., Heckman, C. and Langenbach, R. (1979) Cancer Res. 39,
2618-2624.

Burczak, J.D., Moskal, J.R., Trosko, J.E., Fairley, LL. and Sweeley, C.C.
(1983) Exp. Cell Res. 147, 281-286.

Leder, A. and Leder, P. (1975) Cell 5, 319-322.

Ginsburg, E., Salomon, D., Sreevalsan, T. and Freese, E. (1973) Proc. Natl.
Acad. Sci. U.S.A. 70, 2457-2461.

Fishman, P.H., Simmons, J.L., Brady, RD. and Freese, E. (1974) Biochem.
Biophys. Res. Commun. 59, 292-299.

Chatterjee, S., Sweeley, C.C. and Velicer, LP. (1975) J. Biol. Chem. 250,
61-66.

Macher, B.A., Lockney, M., Moskal, J.R., Fung, V.K. and Sweeley, C.C.
(1978) Exp. Cell Res. 117, 95-102.

37

126

127.
128.

129.

130

131.

132.
133.

134.
135.

136.
137.

138.

139.

140.
141 .

Moskal, J.R., Lockney, M.W., Marvel, C.C., Mason, RA. and Sweeley, C.C.
(1980) in ACS Symposium Series, No. 128, Cell Surface Glycolipids (Sweeley,
C.C., ed.) p.p. 241-263.

Schlessinger, et a1. (1983) Crit. Rev. Biochem. 14, 93-111.

Usuki, S., Lyu, S.-C. and Sweeley, C.C. (1988) J. Biol. Chem. 263, 6847-
6853,

Usuki, 8., Hoops, P. and Sweeley, C.C. (1988) J. Biol. Chem. 263, 10595-
10599.

Usuki, S. and Sweeley, C.C. (1988) Ind. J. Biochem. Biophys. 25, 102-105.
Swamy, M.J. and Sweeley, C.C. (1989) Biochem. Biophys. Res. Commun.
162, 1188-1193

Hampton, R.Y. and Morand, O.H. ( 1989) Science, 246, 1050.

Hannun, Y.A., Loomis, C.R., Merril, A.H., Jr. and Bell, R.M. (1986) J. Biol.
Chem. 261, 12604-12609.

Hannun, Y.A. and Bell, R.M. (1989) Science 243, 500-507.

Faucher, M.F., Girones, N., Hannun, Y.A., Bell, R.M. and Davis, R.J.
(1988) J. Biol. Chem. 263, 5319-5327.

Merrill, AH. and Stevens, V.L. (1989) Biochim. Biophys. Acta 101, 131-139.
Igarashi, Y. and Hakomori, S.-I. (1989) Biochem. Biophys. Res. Commun.
164, 1411-1416.

Bottega, R., Epand, R.M. and Ball, EH. (1989) Biochem. Biophys. Res.
Commun. 164, 102-107.

Burczak, J.D., Soltyziak, R.M. and Sweeley, C.C. (1984) J. Lipid Res. 25,
1541-1547.

Nishizuka, Y. (1984) Nature 308, 693-698.

Mitchell, B. (1983) Trends Biochem. Sci. 8, 263-265.

38

142.

143.

144.
145.

Dawson, G., McLawhon, R. and Miller, R.J. (1980) J. Biol. Chem. 255, 129-
137.

Capasso, J.M., Keenan, T.W., Abeijon, C. and Hirshberg, C.B. (1989) J.
Biol. Chem. 264, 5283-5240.

Svennerholm, L. (1964) J. Lipid Res. 5, 145-155.

IUPAC-IUB Recommendations ( 1977) Lipids 12, 455-468.

39

CHAPTER 2

A QUANTITATIVE METHOD FOR SEPARATING REACTION COMPONENTS
OF CMP-SIALIC ACID:LACTOSYLCERAMIDE SIALYLTRANSFERASE
USING SEP PAK C13 CARTRIDGES

L. J. Melkerson-Watson, K. Kanemitsu, and C. C. Sweeley

November 11, 1986

Department of Biochemistry
Michigan State University
East Lansing, MI 48824

Revised March 9, 1987
Glycoconjugate J. (1987) 4, 7-16

Running Title: Studies on the Purification and Regulation of SAT-l

ABSTRACT

A rapid procedure is described for the separation of CMP-sialic
acidzlactosylceramide sialyltransferase reaction components using Sep Pak C13
cartridges. The quantitative separation of the more polar nucleotide sugar, CMP-sialic
acid, and its free acid from the least polar GM3 ganglioside is simple and rapid relative
to previous described methods. Recovery of GM3 is optimized by the addition of
phosphatidylcholine to the reaction mixture prior to the chromatographic step. Using
rat live Golgi membranes as a source of CMP-sialic acidzlactosylceramide
sialyltransferase activity (GM3 synthase, SAT-1), the transfer of [14C]-sialic acid from
CMP-[14C]-sialic acid to lactosylceramide can be quantified by this assay. The
procedure is reliable and may be applicable to the isolation of ganglioside products in

other in vitro glycosyltransferase assays.

41

INTRODUCTION

The cell surface has been postulated to play an important role in the regulation
of cell growth and differentiation. Gangliosides, ubiquitous sialic acid-containing
glycosphingolipids of plasma membrane, have been implicated in these phenomena.
Changes in ganglioside composition and metabolism have been associated with
oncogenic transformation (1); during which there is often an accumulation of less
complex gangliosides than in the non-transformed cells, due to a block or impairment
of the ganglioside biosynthetic pathway. Alternatively, there may be synthesis of novel
gangliosides due to activation of specific quiescent embryonic or fetal
glycosyltransferases (2).

Several lines of evidence suggest that synthesis of GM3 ganglioside, catalyzed
by CMP-sialic acidzlactosylceramide sialyltransferase (SAT-1), may be critical in the
regulation of ganglioside biosynthesis and cell growth. SAT-1 catalyzes the first
committed step in the synthesis of gangliosides of the ganglio type. An especially
important finding was the inhibition of cell growth via platelet dependent growth factor
and epidermal growth factor mediated mitogenesis by the GM3 ganglioside
incorporated into the plasma membrane (3,4).

The present assay was developed for use in the purification and characterization
of SAT-1 to homogeneity from Golgi membranes and for studies of the mechanisms
involved in the regulation of GM3 synthesis by this enzyme. Like other
glycosyltransferases, SAT-1 is relatively unstable in solubilized preparations and its
specificin for lactosylceramide (LacCer) and the sugar nucleotide is low (5). In vivo
sialyltransferase activity may be dependent upon the composition of the surrounding
membrane, leading to difficulties in assays of its activity in partially purified
preparations. Separation and analysis of the labeled GM3 product from its labeled
precursors by most chromatographic methods (6-13) is time consuming. For example,

determination of GM3 synthase activity in our laboratory by the paper chromatographic

42

method described by Basu and coworkers (10) requires two to three days (14). On the
other hand, employing Sep Pak C13 cartridges (Waters Associates) for the analysis
requires only 2-3 h and gives yields of labeled GM3 that is completely separated from

the labeled sugar nucleotide precursors and free acid.

43

MATERIALS AND METHODS

Materials

Sep Pak C13 cartridges were obtained from Waters Associates (Milford, MA,
USA); Leur-lok tip syringes from Becton-Dickinson (Rutherford, NJ, USA); reagent
grade solvents from Fisher Scientific (Pittsburgh, PA, USA); cytidine 5'-
monophosphate [4-14C]-sialic acid, (NBC-636, 1.8 mCi/mmol) from New England
Nuclear (Boston, MA, USA); [4,5,6,7,8,9-14C]-sialic acid (20 mCi/mmol) from ICN
Pharmaceuticals (Irvine, CA, USA); phosphatidylcholine type III-L (from bovine liver)
and type V-E (from frozen egg yolk); sodium cacodylate, cytidine 5'-monophosphate
sialic acid (CMP-sialic acid) and Triton CF-54 from Sigma Chemical Co. (St. Louis,
MO, USA); and HPTLC silica gel 60 from EM Science (Cherry Hill, NJ, USA).

Three sources of lactosylceramide were employed: a product from human
erythrocytes prepared according to the method of Kundu (13), N-palmitoyl-DL-
dihydro-lactosylcerebroside purchased from Sigma Chemical Co., and synthetic
lactosylceramide prepared in our laboratory (15). The labeled substrates [3H]-GM3,
[3H1-LacCer, [3H]-GD13, [3H]-GD3], [3H]-FchM1 and [3H]-GgOse4Cer ([311]-
asialo-GMl) were prepared according to the potassium [3H]-borohydride reduction
method of Schwarzmann (16). The FchMl was a gift from Dr. S. Hakomori; asialo-
GM1 was derived by mild acid hydrolysis of GM1.

Rat Liver Golgi Membranes

CMP-sialic acidzlactosylcerarnide sialyltransferase (SAT-1) was partially
purified from rat liver Golgi membranes. The Golgi microsomal fraction was prepared
by a method similar to those previously described by Morre et al. (197), Fleischer and
Kervina (18) and Carey and Hirschberg (19). Briefly, livers were removed from 176-
200 g male CD rats (Charles River, Wilmington, MA, USA) and homogenized in 5 vol
of 25 mM sodium cacodylate buffer (pH 6.5) containing 0.25 M sucrose and 1 mM

44

phenylmethyl sulfonyl fluoride (PMSF). Golgi microsomes were collected by
differential centrifugation followed by sucrose density gradient centrifugation. Protein

determinations were by the method of Peterson (20).

CMP-Sialic Acid:Lactosylceramide Sialyltransferase Assay and Sep Pak C13
Separation Strategy

Assays of SAT-1 activity were performed as previously described (5,21). The
reaction was terminated by addition of 1000 pg phosphatidylcholine in 0.4 ml 100 mM
KCl and frozen. this reaction mixture, containing LacCer, [14C]-GM3, CMP-[14C]-
sialic acid and, possibly, free [14C]-sia1ic acid formed by the hydrolysis of the sugar
nucleotide, was transferred to a Sep Pak C18 cartridge (pre-conditioned as suggested by
the manufacturer). Typically, the cartridges were attached to the end of a 5 m1 Luer tip
syringe, and washed with 5 ml methanol followed by 10 ml water. The sample was
applied repeatedly with slight pressure until the eluate, after loading the sample
mixture, was clear. Typically 2-3 times was adequate. However, when analyzing
SAT-1 activity in crude preparations, the assay mixture may need to be applied to more
than one cartridge or the same cartridge may be reutilized. The cartridges are re-usable
after repeating the washing procedure, three times for assays containing crude
homogenates and about eight times in assays with Golgi vesicles. The radiolabeled
nucleotide sugar, CMP-[14C]-sialic acid, and free [14C]-sialic acid were eluted from
the cartridge with 10 ml water. Chloroform/methanol, 2/1 by vol, 0.5 ml, was added
to change the solvent system in the cartridge. This step reduced the amount of time
required to evaporated the organic solvent. The GM3 ganglioside was subsequently
eluted with 5.0 ml chloroform/methanol, 2/1 by vol, and collected directly into a glass
scintillation vial. The organic solvent was evaporated to dryness (traces of chloroform

quench the radioactive signal) and the amount of [14C]-GM3 was determined by liquid

45

scintillation spectrometry in a Packard Tri-Carb 460C in 5 ml Flo ScintTMlI
(Radiomatic Instruments and Chemical Co., Tampa, FL, USA).

Quantitative Recovery of SAT-l Reaction Components from Sep Pak C13
Cartridges

Solutions containing 40 nmol CMP-sialic acid (donor substrate), 100 nmol
LacCer (acceptor substrate), 5 nmol GM3 (product) and 5 nmol sialic acid (free acid)
in a final volume of 500 pl (see above) were applied to the reverse-phase cartridge.
The elution profiles of the various reaction components were ascertained for reaction
mixtures containing either CMP-[14C]-NeuAc, [14C]-NeuAc, or [3H]-GM3. The
radiobiologicals were eluted with sequential applications of 0.5 m1 aliquots of various
aqueous and organic solvents and processed as described above.

Recovery of each reaction component was also examined by GLC for sialic acid
content. A modification of the method using trimethylsilyl (TMS) ethers, first
developed by Sweeley and coworkers (22, 23) and later amended by Ledeen and Yu
(24) was employed. The reaction mixtures for these studies contained only one sialic
acid-containing component per assay. All other conditions were as described above.
All samples were taken to dryness, TMS derivatized, and fractions analyzed. Methanol

was added to aid in evaporation of the solvent.

Qualitative Recovery of the [14Cl-GM3 Product

Identification of the SAT-l product was verified by TLC. The SAT-1 activity
assay was performed and GM3 product recovered as described above. The assay was
repeated ten times and the organic phases pooled. The sample was taken to dryness
and the phosphatidylcholine and [14C]-GM3product were separated according to a
published procedure (25). The [14C]-GM3 was chromatographed on HPTLC silica gel
60 plates in chloroform/methanol/water, 60/30/5 by vol. The TLC plate was scraped

46

in 0.5 cm segments, transferred to glass scintillation vials containing 1.0 ml water and
sonicated. The amount of radioactivity was quantified by liquid scintillation

spectrometry in a Packard Tri-Carb 460C in 5 ml Flo scintTMn.

47

RESULTS

Separation of SAT-1 Reaction Components

Conditions for optimal recovery of the GM3 ganglioside from the SAT-1 assay
mixture were determined from experiments employing a variety of eluting solvents.
Under the conditions of glycosyltransferase assays, in addition to the expected transfer
of sialic acid to the LacCer acceptor substrate to form GM3, hydrolysis of the
radiolabeled sugar nucleotide to radiolabeled free sugar and unlabeled nucleotide may
occur if phosphodiesterases or phosphatases are present in crude preparations (26).
Therefore, it was necessary to compare recovery of CMP-[14C]-sialic acid, [14C]-
sialic acid and [14C]-GM3 from our applied sample.
The separation of the radiolabeled reactants and the GM3 product of the SAT-l
reaction mixture was performed as described in the Materials and Methods section.
The recoveries of CMP-[14C]-sialic acid, [14C]-sialic acid and [3H]-GM3 were
determined in H20, 100 mM KCl, phosphate-buffered saline (PBS, 137 mM NaCl,
2.68 mM KCL, 1.47 mM KH2PO4, 8.06 mM NaH2P04) and 25 mM sodium
cacodylate (pH 6.5): methanol, chloroform/methanol, 1/1 by vol;
chloroform/methanol, 2/1 by vol; chloroform/methanol/water, 60/30/5 by vol; and in
the absence and presence of phosphatidylcholine. Employing the strategy of Williams
and McCluer (27), application of GM3 gangliosides to the cartridge in 100 mM KCl
gave recoveries of only 39-53% with the aqueous solvents surveyed and 47-61 % with
the organic solvents from the SAT-1 reaction mixture. This method has been reported
to give 94% recovery of rat brain gangliosides from the theoretical upper phase
containing 100 mM KCL (27). Our inability to recover GM3 completely was
attributed to the presence of 0.3% Triton CF-54 in our SAT-1 assay buffer. To
enhance the recovery of the [3H]-GM3 by increasing the size of the lipid-detergent-
lipid micelles, phosphatidylcholine (PC) was added. PC enhances SAT-l activity and

48

is probably a membrane-associated lipid ( 14). PC‘from two different sources (bovine
liver and egg yolk) was examined and both gave similar results. Other phospholipids
were not assayed. The recovery of GM3 was dependent upon the amount of PC added
(10-1000 pg). Formation of the PC/Triton CF-54/GM3 micelles gave recoveries of
70-100% of the ganglioside from the SAT-1 reaction mixture in the organic was (Table
1). Complete recovery of GM3 was accomplished with the addition of 1000 pg PC in
100 mM KCl. Further, the detergent served to reduce the interaction of the nucleotide
sugar with thecartridge. Williams and McCluer (27) reported that 0.39% of added
CMP-NeuAc was absorbed on Sep Pak C13 cartridges and that with extensive washing
of the cartridge with water, retention of nucleotide sugars could be reduced. In our
experiments, the nucleotide sugar and free sialic acid were recovered in good yield (99—
100%) from the cartridge regardless of the aqueous solvent used. The recovery of
CMP-sialic acid and sialic acid was completed within 4-5 ml aqueous solvent; however,
10 ml was typically employed. Negligible amounts (0.01 - 0.04%) were recovered in
the organic solvent and as shown by GLC analysis for sialic content, the anomeric
mixture of sialic acid recovered from the CMP-sialic acid after methanolysis was
different from that obtained form GM3. Thus, critical differentiation can be made
between the amount of product formed and any possible contamination from nucleotide
sugars. Additionally, the recovery of the ganglioside correlated with the degree of
polarity of the solvent employed. With the exception of methanol (92% recovery of
GM3), all the organic solvents examined gave complete recovery of the GM3
ganglioside. The most reproducible results were obtained when chloroform/ methanol,
2/1 by vol, was the solvent. A typical elution profile for the recovery of the GM3
ganglioside according to our separation scheme, as described in the Materials and

Methods section, is shown in Fig. 1.

49

Table 1

Recovery of [3m-GM3 from Sep Pak C1 Cartridges. The Effect of

 

 

Phosphatidylcholine on ecovery.
Sample 100mM KCl PCa H20 [3H]-GM3
no. (ml) (mg) (ml) 6pm %
Recovered Average Recovered
l - - - 209490 100.0
i 6060
2 0.4 - - 122270 129340 61.7
3 0.4 - - 136400 i7070
4 - 10 0.4 148960 144080 68.8
5 - 10 0.4 139210 i4880
6 - 100 0.4 181020 182930 87.3
7 - 100 0.4 184850 i 1920
8 - 1000 0.4 198010 206340 87.3
9 - 1000 0.4 214670 i8330
10 0.4 10 - 156870 165890 79.2
11 0.4 10 - 174900 i9020
12 0.4 100 - 175000 180000 85.9
13 0.4 100 - 185000 i5000
14 0.4 1000 - 210850b 210850 101.0 b
15 0.4 1000 - 228920 (219890 (106.0)
$9030)
16 - 10 0.4 186400 174810 83.4
17 - 10 0.4 163220 i11590
18 - 100 0.4 173690 184290 88.0
19 - 100 0.4 194890 i 10600
20 - 1000 0.4 209080 200730 95.8
21 - 1000 0.4 191730 i9000
22 0.4 10 - 125970 144280 68.9
23 0.4 10 - 162600 i 18320
24 0.4 100 - 184290 181360 86.6
25 0.4 100 - 178430 i2930
26 0.4 1000 - 201560 207045 98.8
27 0.4 1000 - 212530 i5500

 

a For samples 2-15, the PC was added to the 0.4 m1 additive solution. For samples 16-27, the PC was

b Value outside confidence limits of standard.

added to the SAT-1 reaction matrix.

50

Figure 1. Elution Profile of Radiolabeled SAT-l Reaction Components on Sep
Pak C13 Cartridges. Chromatography on Sep Pak C13 was performed on complete
reaction mixtures composed of 100 ml of a protein fraction containing SAT-1, and 400
ml 100 mM KCl containing 1000 pg PC (sonieated before addition to the reaction
mixture). The final concentrations of components contained in the SAT-1 reaction
mixture were 1 mM LacCer, 0.4 mM CMP-NeuAc, 0.05 mM GM3, 0.05 mM NeuAc,
112.5 mM sodium cacodylate (pH 6.5), 15 mM MnClz, 0.125 mM sucrose, 0.05 mM
PMSF and 0.3% (w/v) Triton CF-54. CMP-[14C]-NeuAc (o), [14C1-NeuAc (0) or
[3H]-GM3 (0) were individually substituted into the SAT-1 reaction mixture and their

recovery monitored by collecting 0.5 ml fractions from cartridges.

51

 

$5 m230> 202.3,...

fineflioeo m a. o m e m a F

A

 

«.2951 .
o<z=oz._Oez ..
o<z=oz.8ez.ns_o .

 

l

 

 

0..

ON

om

0V

om

031.013 WHSWOIGVH 0/o

52

The recovery of the SAT-1 reaction components, as investigated by GLC
analysis of both the polar and non-polar eluants for sialic acid content, gave similar,
results for the devised separation scheme of: 1000 mg in 100 mM KCI, water, and
CHC13/MeOH, 2/1 by vol (applied volume, aqueous solvent, and organic solvent,
reSpectively). The sugar nucleotide and free sialic acid eluted with the aqueous solvent
and the GM3 ganglioside was recovered with the organic solvent. Methanolysis of
GM3 gave the methyl ketal of N-acetylneuraminic acid in the B-anomeric configuration
which was recovered in the organic solvent without contamination from other reaction
components. The free sialic acid liberated from the other reactions components was
primarily in the a configuration and recoverable in the aqueous solvent. Some traces of
a-NeuAc (0.05 _+_ 0.02 pg) were detected in the organic fractions. However, these
amounts did not significantly increase the background in blanks for the analysis of the
GM3 by radioactivity.

The capacity of the Sep Pak C18 cartridge for GM3 was examined by analysis
of radioactivity and by gas chromatographic analysis. In analysis of radioactivity,
application of up to 50 nmol of [3H]-GM3 was completely recoverable. Gas
chromatographic analyses indicated that application of GM3 equivalent to 25 pg of
sialic acid (81 nmol) overloaded the Sep Pak C13 cartridge wheres 13.6 pg of the GM3
sialic acid (44 nmol) was retained by the cartridge. To determine the maximum
capacity of the cartridge with optimal recovery, the recovery of 0-15 pg GM3 sialic
acid was examined. For amounts of GM3 no exceeding 2.5 pg (8 nmol) of sialic acid,
96% of the GM3 was recovered. This is in contrast to the 5 pmol reported by
Williams and McCluer for ganglioside sialic acid recovery from theoretical upper phase
(27). When 5-15 pg (16-40 nmol) of GM3 sialic acid was applied to the cartridge
using our SAT-1 system, the recovery was reduced to 75-87%.

53

Application of SAT-l/SepPak C13 Strategy in Determining SAT-l Activity

The transfer of [14C]-NeuAc from CMP-[14C]-NeuAc to LacCer to form
[14C]-GM3 by SAT-1 was examined using the Sep Pak C13 separation strategy
described in the Materials and Methods section. The enzymatic synthesis of GM3 in
rat liver Golgi was studies as a functions of time and Golgi microsomal protein (Fig.
2). The synthesis of GM3 proceeded linearly for at least 180 min and the synthesis of
the product was proportional to the amount of protein in the reaction mixture up to
0.25 mg of microsomal protein. The cartridges were saturated by excess protein in the
crude enzyme preparation, resulting in non-linearity of the curve beyond 0.3 mg of
Golgi fraction.

The acid glycosphingolipid formed form LacCer and CMP-sialic acid was
identified as GM3 ganglioside by TLC. The SAT-1 activity assays were set up with
two different formats. Pooled eluates of glycosphingolipid were obtained from 10
assays employing CMP-[14C]-sialic acid and LacCer while separate products were
pooled from a second set of 10 assays employing CMP-sialic acid and [3H]-LacCer.
The organic eluants from each set were dried under nitrogen and the
phosphatidylcholine and salts removed by the procedure described by Ladisch and
Gillard (25). The radiolabeled products comigrated with a reference sample of [3H]-
GM3 on HPTLC plates (data not shown).

The applicability of this reverse-phase chromatography strategy on Sep Pal C18
cartridges for the recovery of other glycosyltransferase products was briefly
investigated. Table 2 shows that while the conditions have been optimized for the
recovery of GM3; other glycosphingolipids such as LacCer, Fuc-GMl , GgOse4Cer,
GD3 and GD1 a can be recovered form the detergent-containing SAT-1 reaction mixture
by this procedure. The recovery and analysis of SAT-l sialyltransferase with slight

modifications may be equally effective for assays of other glycosyltransferase activities.

54

Figure 2. Enzymatic Synthesis of [14C]-GM3 from CMP-[14C1-NenAc and
LacCer as a Function of Time and Protein. A. Reaction mixtures with 2.28 x 10’2
mg Golgi-enriched microsomal protein were incubated for the indicated period of time
at 37°C and the [14C]-GM3 product was recovered as described in the Materials and
Methods section. B. Reaction mixtures containing the indicated amount of Golgi-
enriched microsomal protein were incubated for 155 min at 37°C and processed as
described in the Materials and Methods section. These data were collected from two

different Golgi membrane preparations from rat liver.

55

Eamon. «.2983. See

 

 

 

 

 

 

0:1

 

 

0 0 0 0 0 0 0 0
0 O 0 0 O 0 0 m
3 2 1 7 6 5 4 3 AOL
3
2
o 0 ##
18 PP
1 EM
0 0 RP
D.
. 0 mm
.2..) e LL
n 00
.W GG
01‘ .0
..9E
m
T
o .m o o
o
.o .
e 3
A B o.
o. o. o. o. _ .
4 3 2 1

Z_m._.0mn_ Sammy—43$— mma

omsEon. osoeoez Eno
omsmOn eso $305624:

0.3

0.2
PROTEIN (mg)

56

Table 2

Recovery of Other Gangliosides from SAT-l Reaction Mixtures from Sep Pak C13

 

 

Cartridges.
Glycosphingolipid Amount cpm [3H]-GSL % Recovered

added Recovered

(cpm) Above Background
[3H11acCer 300,000 297460 $8680 99.2
[3H]-GM3 184,000 188580i5250 102.5
[3H]-FchM1 1600 1970_+_110 123.1
[3H]-GgOse4Cer 20,000 17350-1750 86.8
[3H]-GD3 9800 8380 :60 84.9
[13111-61512, 5700 49001L 610 85.9

 

57

DISCUSSION

The major purpose of this investigation was to simplify the analysis of CMP-
sialic acid:lactosylceramide sialyltransferase (SAT-1) activity. We have developed a
rapid and convenient system for the separation and analysis of GM3 ganglioside from
crude reaction mixtures. Sep Pak C13 cartridges (octadecylsilane bonded-phase
packings) are useful for the separation of problem sample preparations containing
components of differing polarities in complex aqueous solvents. A common problem in
studies of ganglioside synthesis and regulation is the need for complete separation of
radiolabeled sugar nucleotides from the radiolabeled ganglioside product. Several
methods have been developed for removal of non-ganglioside reactants (6-13).

Williams and McCluer (27) have previously reported that Sep Pak C18 reverse-
phase cartridges can be employed for the recovery of gangliosides from theoretical
upper phases of Folch-extracted tissues with the addition of 100 mM KCl to the sample
matrix. However, their modification is not applicable for the recovery and analysis of
gangliosides synthesized from detergent-solubilized Golgi membrane preparations.
With detergent in our SAT-l assay mixtures, the recovery of GM3 was 39-62%, which
is not acceptable for enzymatic analyses. Since purification and characterization of
membrane-bound SAT-1 requires detergent solubilization for in vitro analysis of
enzyme activity, we found it necessary to modify the Sep Pak C13 procedure to
accommodate the composition of our reaction mixtures. The Sep Pak C13 cartridges
provide a rapid, convenient and reliable system for determining SAT-l activity during
its purification and characterization. SAT-1 activity can be quantified accurately up to
the formation of 8 nmol of GM3. Enzyme activity can be determined by the recovery
of radiolabeled GM3 or by GLC analysis of sialic acid recovered from the GM3 by
methanolysis. This method can be applied to quantitative analysis of other
glycosyltransferase activities involved in the biosynthesis of gangliosides. The

described procedure has been modified in our laboratory for the analysis of GD3

58

synthase activity. The only problem we have encountered is that some detergents, e.g. ,
Triton X-100 (>0.5%, by vol), completely solubilize 1000 pg PC in the sample
matrix. Therefore, depending on the nature of the detergent employed and the nature
of the lipid environment of the glycosyltransferase, a different amount or type of

phospholipid may have to be used to obtain optimal recovery of the ganglioside
product.

59

ACKNOWLEDGEMENTS

This study was supported in part by a research grant from the National Institutes of
Health (AM12434).

We thank Kimihiro Kanemitsu for sharing his expertise and training on the GC analysis

of sialic acid.

10.
ll.
12.
13.
14.

15.
16.
17.

18.
19.

REFERENCES
Hakomori, S. (1981) Ann. Rev. Biochem. 50, 733-764.
Hakomori, S. and Kannagi, R. (1983) J. Natl. Cancer Inst. 71, 231-251.
Bremmer, E. G., Hakomori, 8., Bowen-Pope, D.F., Raines, E., and Ross, R.
(1984) J. Biol. Chem. 259, 6818-6825.
Bremmer, E.G., Schlessinger, J. and Hakomori, S. (1986) J. Biol. Chem. 261,
2434-2440.
Burczak, J.D., Fairley, J.L., and Sweeley, C.C. (1984) Biochim. Biophys.
Acta 804, 442-449.
Walton, K.M. and Schnaar, R.L. (1986) Anal. Biochem. 152, 154-159.
Byme, M.C., Sbaschnig-Agler, M., Aquino, D.A., Slcafani, J .R., and Ledeen,
R.W. (1985) Anal. Biochem. 148, 163-173.
Yates, H.C., Ueno, K., Chang, N.-C., Glaser, G.H., and Yu, R.K. (1980) J.
Neurochem. 34, 560-568.
Carter, T.P. and Kanfer, J. (1973) Lipids 8, 537-548.
Chien, J., Williams, T., and Basu, S. (1973) J. Biol. Chem. 245, 1778-1785.
Kanfer, J. (1969) Methods Enzymol. 3, 41-90.
Chien, J., Williams, T., and Basu, S. (1966) Methods Enzymol. 8, 365-368.
Kundu, S.K. (1981) Methods Enzymol. 62, 174-185.
Burczak, JD. (1984) Ph.D. Dissertation, Michigan State University, East
Lansing, MI, USA.
Kanemitsu, K. and Sweeley, C.C. (1986) Glycoconjugate J. 3, 143-151.
Schwarzmann, G. (1978) Biochim. Biophys. Acta 529, 106-114.
Morre, D.J., Merlin, L.M., and Keenan, T.W. (1969) Biochem. Biophys. Res.
Commun. 37, 813-819.
Fleischer, S. and Kervina, M. (1974) Methods Enzymol. 31, 6-41.
Carey, DJ. and Hirschberg, C.B. (1980) J. Biol. Chem. 255, 4348-4354.

61

20.
21.

22.
23.
24.
25.
26.

27.

Peterson, G.L. (1977) Anal. Biochem. 83, 346-356.

Kaufman, B., Basu, S., and Roseman, S. (1966) Methods Enzymol. 8, 365-
368.

Sweeley, C.C. and Walker, B. (1964) Anal. Chem. 36, 1461.

Vance DE. and Sweeley, C.C. (1967) J. Lipid Res. 8, 621.

Ledeen, R.W. and Yu, R.K. (1982) Methods Enzymol. 83, 179-183.

Ladisch, S. and Gillard, B. (1984) Anal. Biochem. 146, 220-231.

Yu, P.H., Barclay, S., Davis, B., and Boulton, A.A. (1981) Biochem.
Pharmacol. 30, 3089-3094.

Williams, M.A. and McCluer, RH. (1980) J. Neurochem. 35, 266-269.

62

CHAPTER 3

PURIFICATION TO APPARENT HOMOGENEITY BY
IMMUNOAFFINITY CHROMATOGRAPHY AND PARTIAL
CHARACTERIZATION OF THE GM3 GANGLIOSIDE FORMING ENZYME,
CMP-SIALIC ACID:LACTOSYLCERAMIDE n2-3 SIALYLTRANSFERASE
(SAT-1), FROM RAT LIVER GOLGI“

L. J. Melkerson-Watson and C. C. Sweeley+

July 30, 1990
Submitted August 22, 1990
Revised November 9, 1990

J. Biol. Chem. (1991), in press.

Department of Biochemistry
Michigan State University
East Lansing, MI 48824

Running Title: Studies on the Purification and Regulation of SAT-1

ABSTRACT

CMP-sialic acid:lactosylceramide a2-3 sialyltransferase (SAT-1) has been
purified approximately 40,000-fold to apparent homogeneity from rat liver Golgi. The
enzyme was solubilized from Golgi vesicles in 5% lauryl dimethylamine oxide
(LDAO), "partially” purified by affinity chromatography twice on CMP-hexanolamine
and once on lactosylceramide-(LacCer-) aldehyde Sepharose 4B. Final purification was
achieved by immunoaffinity chromatography on M12GC7-Affi Gel-10. The M12GC7
monoclonal antibody specifically inhibits and immunoprecipitates SAT-1 activity.
Identification of the protein, with an apparent molecular weight by SDS-PAGE of
about 60,000 daltons, was confirmed by Western Blot and immunodetection with
M12GC7. SAT-1 specifically catalyzes the transfer of N-acetylneuraminic acid
(NeuAc, sialic acid) to lactosylceramide (GalBl-4Gchl-O-Cer), forming GM3
ganglioside. Studies on substrate specificity indicate that the preferred acceptors have
the general structure saccharideBl-O—Cer, a disaccharide being preferred to a
monosaccharide. SAT-1 is a glycoprotein. The carbohydrate moieties are detected
with specific lectins. Deglycosylation of SAT-1 with N-glycanase results in an increase
in a 43,000 dalton band. The two-dimensional electrophoretogram of SAT-1 indicates

a pI range of 5.7 to 6.2 for the 60,000 dalton protein.

INTRODUCTION

Glycoconjugates have been implicated in a variety of cell surface interactions,
cellular differentiation, and oncogenic transformation (for reviews, 1-3 and references
therein). The synthesis of these complex carbohydrates is catalyzed by linkage-specific
glycosyltransferases, several of which have been purified and cloned (for review, 4 and
references therein). Many of these purified and characterized enzymes are involved in
the synthesis of various glycoproteins. Very limited success has been achieved in
purifying the glycosyltransferases specifically involved in glycosphingolipid (GSL)
biosynthesis. This may reflect the difficulty in solubilization, the relative instability of
the enzymes once extracted from their native lipid environment, low substrate
specificities (5), laborious activity assays and proteolytic degradation during
purification (6, for review 4).

The general approach for this group of glycosyltransferases has been the
affinity chromatography employing either an acid modification of the acceptor
glycolipid (5,7-10) or Sepharose 4B followed by FPLC MonoS ion exchange and
proRPC reverse-phase chromatography (1 1). While these protocols have proven to be
applicable for purifying some GSL glycosyltransferases, they are not optimal for those
glycosyltransferases involved in the early stages of the glycosphingolipid cascade. A
recent report (12) identifies two classes of GSL glycosyltransferases based on their
substrates specificities. The first class recognizes acceptor glycolipid substrates strictly
on the basis of carbohydrate-carbohydrate linkages, and the second recognizes
carbohydrate-lipid linkages. Thus, a glycolipid-acid Sepharose, or alternative, may not
provide the optimal glycolipid structure to successfully purify both classes of glycolipid
glycosyltransferases.

We report here the purification to apparent homogeneity of rat hepatic lactosle-
O-ceramide 012-3 sialyltransferase (SAT-1). Purification of this GM3 ganglioside

forming enzyme was accomplished using a glycolipid-aldehyde Sepharose affinity resin

65

and monoclonal antibody, MIZGC7, directed against SAT-1. We describe the
synthesis of this affinity ligand, production and characterization of a specific anti-SAT-

l monoclonal antibody, and immunoaffinity chromatography in verification of SAT-1

apparent homogeneity.

66

EXPERIMENTAL PROCEDURES

Materials - Cytidine 5'-monophosphate sialic acid, (CMP-[14C4’5,5,7,3,9]-
sialic acid, 286.5 mCi/mmol) was purchased from New England Nuclear (Boston, MA)
and its specific activity adjusted to 22,200 dpm/nmol with unlabeled CMP-sialic acid
from Sigma (St. Louis, MO). Lactosylceramide was purified from AB+ human red
blood cells obtained from the American Red Cross, Lansing, M1 or was synthesized as
described (13). The sphinganine-containing form of lactosylceramide was purchased
from Sigma and used in activity assays. Ammonyx LO (lauryldimethylamine oxide,
LDAO) was obtained from the Stepan Company (Chicago, IL). Sodium cacodylate,
CMP-hexanolamine-Sepharose 4B, Hexanolamine-Sepharose 4B, fetuin, asialofetuin,
mucin, asialomucin, glycophorin, asialoglycophorin, al'aC1d glycoprotein, lactose, 6-
aminohexyl-Sepharose 4B and pristane were purchased from Sigma Chemical Company
(St. Louis, MO). RPMI 1640, NCTC-109, penicillin, streptomycin sulfate, HEPES,
fetal bovine serum, donor horse serum, hypoxanthine, thymidine and L-glutamine were
purchased from GIBCO Laboratories (Gaithersburg, MD). Affi-Gel 10, Goat anti-
mouse IgG (H-l-L), goat anti-mouse IgG-alkaline phosphatase conjugate, SDS-PAGE
low molecular weight standards, prestained molecular weight standards, MAPS 11
Protein A HPLC column, MAPS II buffers and ultrapure electrophoresis reagents were
bought from BioRad (Richmond, CA). Mouse Isotyper test kit was from Amersham
(Arlington Heights, IL). Rabbit anti-mouse IgG's were from Chemicon (Temecula,
CA). Protein A-positive Staph A cells, polyethylene glycol-1500 (PEG-1500),
aminopterine, Nutridoma-SP, a2-6 sialyltransferase, PMSF, leupeptin, aprotinin,
pepstatin, TPCK, TLCK, E-64, and the glycan differentiation kit were purchased from
Boehringer Mannheim (Indianapolis, IN). Genzyme (Boston, MA) was the source of
N-glycanase, as well as a2-6 and a2-3 sialyltransferases. Sep Pak C13 cartridges were
purchased from Waters Associates (Milford, MA). HPTLC plates with fluorescent
indicator were from EM Science (Cherry Hill, NJ). Polyisobutylmethylacrylate

67

(PIBMA) was obtained through the Aldrich Chemical Co. (Milwaukee, WI).
Ampholines, Pharmalytes and carbamylated pl standards were from Pharmacia-LKB
(Piscataway, NJ). All other reagents were ultrapure or ACS grade from commercial
sources.

Animals and Cell Lines - Male CD rats (six to eight weeks old) and female
Swiss albino mice (four to six weeks old) were purchased from Charles River
(Wilmington, MA). Rats were anesthetized with C02 and the liver perfused with ice-
cold 25 mM sodium cacodylate (pH 6.5) containing 0.25 M sucrose and 1 mM PMSF
before excising the tissue. The mice were anesthetized with ether and cervically
dislocated. SP2/O-Agl4 cells (ATCC CRL 1581) were obtained from the American
Type Culture Collection (Rockville, MD). These cells were maintained between 1 x
105 to 1 x 106 cells/ml in RPMI 1640 containing 10% fetal bovine serum, 10% horse
serum, NCTC- 109, hypoxanthine, thymidine, penicillin, streptomycin and 1%
Nutridoma-SP at 7.5% C02.

1. ASSAYS.

Protein concentration. Protein was assayed by the procedure of Peterson (14)
or spectrophotometrically at A230 (15).

SAT-l Activity Assay. SAT-l assays, with radiolabeled CMP-sialic acid as the
donor substrate and lactosylceramide as the acceptor substrate, were carried out by a
modification of a previously described procedure (16). Assay mixtures contained the
following components in 100 pl: lactosylceramide, 0.2 pmoles; CMP-
[14C4,5,6,7,3,9]-sialic acid (22,200 dpm/nmole), 0.04 pmoles; sodium cacodylate
buffer (pH 6.5), 10 mmoles (100 mM final concentration); manganese chloride, 1
mmoles (10 mM final concentration); 5.0% (v/v) LDAO and 0005-01 mg of the
protein. GM3 reaction product was recovered by reverse phase chromatography on

Sep Pak C13 cartridges as previously described (17).

68

ST2-6 Sialyltransferase Assay. B-galactosyl 012-6 (ST2-6) sialyltransferase
was assayed according to the conditions previously described for this enzymes (18-21)
using 50 pg asialo-al-acid glycoprotein as acceptor. The al‘aCid glycoprotein was
desialylated by acid hydrolysis (22) in 0.05 M H2804 at 80°C for 60 min. The
reaction was neutralized with 0.1 M NaOH and dialyzed versus double-distilled H20
overnight and lyophilized.

Thiamine Pyrophosphatase Activity. The assay was performed as previously
described (24).

Quantitation of Lauryldimethylamine Oxide (LDAO). The amount of
LDAO in solution was quantitated as described by the manufacturer (Stepan Chemical
Co., Chicago IL).

2. AFFINITY PURIFICATION OF SAT-1.

Preparation of Golgi-enriched Microsomes. Procedures for the preparation of
a highly enriched fraction of Golgi vesicles from rat liver have been well-established.
Golgi-enriched microsomes were prepared by differential centrifugation similar to
methods previously described (23-27). All fractionation steps were carried out at 4°C.
Livers (100 g wet weight) were from six- to eight-week old (approximately 175 g) male
CD rat. After anesthetizing rats with C02 and decapitation, the livers were perfused
with 10-20 ml of 25 mM sodium cacodylate (pH 6.5) containing 0.25 M sucrose and 1
mM phenylmethylsulfonyl fluoride (PMSF), removed and immediately homogenized in
five volumes of the same buffer using a Polytron homogenizer (Brinkman Instruments,
Westbury, NY). After centrifugation at 4°C for 10 min at 5000 x g, the supernatant
fraction was centrifuged on 1.2 M sucrose cushions at 26,000 rpm (approximately
100,000 x g) for 30 min in a Beckman SW27 rotor. The microsomes were collected
and pelleted at 100,000 x g for 30 min at 4°C. The microsomal pellets were
resuspended in 25 mM sodium cacodylate (pH 6.5) containing 0.25 M sucrose and 1

mM PMSF to a final ratio of about 0.7 to 1.0 g liver per ml buffer. Discontinuous

69

sucrose gradients were routinely prepared by layering 10 ml of the microsomal
suspension on top of 5.0 ml of 10%, 20%, 25%, 32%, 35% and 40% (w/v) sucrose,
prepared in 25 mM sodium cacodylate (pH 6.5), and centrifuging at 26,000 rpm for 3
h in a Beckman SW27 rotor at 4°C. Employing this methodology, the Golgi were
recovered at the sucrose interfaces between 10-30%, as verified by increased thiamine
pyrophosphatase and SAT-l activity. The Golgi were concentrated by centrifugation at
100,000 x g and resuspended in a minimal volume of 25 mM sodium cacodylate (pH
6.5).

Detergent Extraction. The Golgi-enriched microsomal fraction was brought to
a final concentration of 15% (v/v) LDAO in 25 mM sodium cacodylate (pH 6.5)
containing 10 mM MnClz, 1 mM PMSF. The suspension was sonicated for 15-20
min. Following gentle stirring for 12-18 h (overnight) at 4°C, the detergent-
solubilized fraction was recovered by centrifugation at 150,000 x g for 90 min at 4°C.
The supernatant fraction contained the detergent-solubilzed SAT-1 activity. The pellet
was re-extracted and the supematants pooled and the final volume adjusted to give 5%
LDAO final concentration. Special considerations in using LDAO for SAT-l
purification are described elsewhere (29).

Affinity Chromatography on CMP-hexanolamine—Sepharose I. The
detergent-extracted proteins from Golgi vesicles were applied to a column (1.5 cm x 30
cm) of CMP-hexanolamine Sepharose (2-4 pmoles CMP per ml gel) which had been
equilibrated with 25 mM sodium cacodylate (pH 6.5) containing 5% (v/v) LDAO and
10 mM MnC12 (affinity chromatography buffer). The column was washed with a
minimum of 10 column volumes of the same buffer to remove excess proteins at a flow
rate of 40 ml/h. The absorbance at A280nm of the fractions was monitored. Elution
of the sialyltransferase activity was carried out with a linear salt gradient established
between 500 ml (total volume) of the affinity chromatography buffer and buffer
containing 0.5 M NaCl.

70

Rechromatography on CMP-hexanolamine Sepharose II. The SAT-1 active
fractions were pooled from CMP Sepharose I and concentrated on an Amicon PM10
membrane filter. The concentrate was dialyzed against 25 mM sodium cacodylate (pH
6.5) and applied on a second CMP-hexanolamine Sepharose 48 column (1.5 cm x 30
cm) equilibrated with the 25 mM sodium cacodylate (pH 6.5) affinity chromatography
buffer containing 5% (v/v) LDAO and 10 mM MnClz. The flow rate for loading the
sample was 2-4 ml/h. The column was washed with a minimum of 10 column volumes
of buffer at a flow rate of 40 ml/h until no protein was detected by A280nm- The
SAT-1 enzyme activity was eluted with a linear gradient (500 ml total volume) of
affinity chromatography buffer and buffer containing 1 mM CMP. SAT-1 active
fractions were pooled and concentrated. The CMP was removed during protein
concentration with PM10 membrane and chromatography on Sephadex G-25 (fine) (1.5
cm x 25 cm) in the affinity chromatography buffer.

Preparation of LacCer-aldehyde Sepharose. Steps in the synthesis of the
lactosylceramide (LacCer)-aldehyde affinity column are shown in Fig. 1. Thirty mg of
LacCer (I), purified from AB+ human red blood cells as previously described (28),
was solubilized in 1.2 ml pyridine to which 0.9 ml acetic anhydride was slowly added
with gentle stirring. The acetylation was allowed to proceed overnight at room
temperature. The reaction mixture was evaporated to dryness via addition of toluene.
The acetylated LacCer (II), AcLacCer, was resuspended in 5 .0 ml chloroform and the
allylic double bond of the sphing-4-enine residue was oxidized, in the presence of
excess ozone for 40 min at -50°C with 90 volts at 6 psi 02, to the ozonide derivative
(III). The AcLacCer ozonide was converted to the aldehyde (IV) by reduction of the
ozonide with excess Zn in acetic acid. This derivative was taken to dryness under a
stream of N2. The AcLacCer-aldehyde (IV) was recovered in the aqueous methanolic
phase by aqueous methanol/hexane (1:1, v/v) partition and verified by HPTLC in a

dual solvent system, dichloroethane/acetone (1:1, v/v) followed by

71

Figure 1. Chemical Synthesis of LacCer-aldehyde Sepharose. The synthesis of the
lactosylceramide-aldehyde-Sepharose 4B affinity resin is described under Experimental
Procedures. The compounds are identified as follows: I.) LacCer; II.) acetylated
lactosylceramide (AcLacCer); III.) AcLacCer ozonide; IV.) AcLacCer-aldehyde; V.)
Schiff base intermediate; VI.) AcIacCer-aldehyde—Sepahrose 4B; VII.) LacCer-
aldehyde-Sepharose 4B.

72

822...? :zo.~:o.xz.~:o\:of_t/o go: go: Alxos‘l

#8 e36. 2:
JH zeozo :ooxo

 

 

H
03 o<O
o<O
goiowIIZDANIOV :2 INIU\/7 20 O 0‘0 IOOZ\¢IQOZ 083.381.? DWI/ANIOVIZIIO o<O O 20
:8 e319 0 o o 02 \Jztzxo on «8888i: :Auxoe :2

M c

2:

JV oaouxo gonzo ...o 36./W San zo o<o~zooo
o

H H

010§0<0
ave/“avgesw so Azlsoo<a~ n..o2.o..o.e\A Y/joNllle/esms nallelio
nzo {row/a nxo emzojflz:

o<oozo onoozoo o<o~xo e<o~zooe<

 

N
t ...... / sQéQll ...... Q.......
8.1.5.3. o a. 141...?

a a
30 :o 63 :o zo~xo zouroo:

73

chloroform/methanol/water (60:30:5, v/v/v). The AcLacCer-aldehyde (IV) was linked
to a methanolic suspension of prewashed 6-aminohexyl-Sepharose 4B. The Sepharose
was prewashed twice with double distilled H20 and three times with MeOH. The
AcLacCer-aldehyde and gel were allowed to react for 18 h at 4°C. The Schiff base (V)
was reduced with a 10-fold excess of NaBH4 in methanol for 1 h, and the acetyl groups
(VI) were removed by mild alkali-catalyzed hydrolysis to give the LacCer-aldehyde
affinity ligand (VII). The affinity resin was equilibrated with affinity chromatography
buffer.

Chromatography on LacCer-aldehyde Sepharose. Active fractions of SAT-l
from the CMP-Sepharose 11 step were directly subjected to LacCer-Sepharose affinity
separation on a 1.0 cm x 10 cm column. The column was previously equilibrated with
25 mM sodium cacodylate (pH 6.5) containing 5% (v/v) LDAO and 10 mM MnC12
affinity chromatography buffer at a flow rate of 2.5 ml/h. After washing the column
with five volumes of affinity buffer at a flow rate of 10 ml/h, SAT-1 was eluted from
the column with affinity buffer containing 1 mM CMP. Active fractions were pooled,
concentrated and the nucleotide removed by gel filtration on Sephadex G-25 (fine).
The enzyme was stored at -20°C in the 25 mM sodium cacodylate containing 5% (v/V)
LDAO for 6-12 months without appreciable loss of activity (29).

3. STRATEGY FOR DEVELOPING A MONOCLONAL ANTIBODY
TO SAT-1 AND ITS APPLICATION IN SAT-1 PURIFICATION

Production of Monoclonal Antibodies to SAT-1. Monoclonal antibodies to
rat liver SAT-1 were raised to the 40,000-fold CMP/LacCer affinity-purified enzyme.
Fifty pg of active affinity-purified SAT-1 was suspended in 2.0 ml of Ribi Probe MPL-
TDM Emulsion (Ribi Immunochem Research Co.) by sonication. The fusion protocol
used for the production of anti-SAT—l hybridomas and the Western blot and ELISA
screening protocols were performed as described (30). Hybrids were produced by the

fusion between SP2/0-Ag14 mouse myeloma cells (2-3 x 10° cells/ml) and mouse

74

spleen cells (2-3 x 107 cells/ml) three days post booster. Hybridomas were cultured
and maintained in HT-containing RPMI 1640 with 10% (v/v) fetal bovine and 10%
(v/V) donor horse sera at a cell density of 1 x 105 to 1 x 106 cells/ml. Hybridomas
were screened by ELISA and immunoprecipitation for anti-SAT-l activity. The anti-
SAT-l positives were cloned by the method of limiting dilution (30) and the resulting
monoclonal hybridomas were grown and the culture media screened as described before
for the hybridomas. Anti-SAT-l producing monoclonal hybridomas were established,
expanded and cultured in serum-free RPMI 1640 containing 1% (WV) Nutridoma-SP,
10% NCTC- 109, hypoxanthine, thymidine, penicillin and streptomycin with 7.5%
C02. The medium containing the secreted monoclonal antibody (MAb M12GC7) to
SAT-1 was collected, clarified by centrifugation at 2000 x g for 10 min at 22°C and
concentrated using an Amicon YM100 membrane. Alternatively, MlZGC7 anti-SAT-l
antibodies were recovered from ascites fluid.

The isotype of MIZGC7 was determined using the Amersham's mouse antibody
isotyper kit according to the manufacturer's directions.

Monoclonal antibodies were purified using the BioRad MAPS 11 Protein A
HPLC system according to the manufacturer with some modification. M12GC7 was
bound to the MAPS ' 11 column in MAPS II binding buffer according to the
manufacturer. The column was washed with a 1:10 dilution of MAPS II elution
buffer. The monoclonal antibodies were eluted on a linear gradient established between
the diluted elution buffer and the MAPS II elution buffer.

Inhibition of SAT-1 Activity by M12GC7. The effect of monoclonal antibody
M12GC7 on SAT-1 activity was determined by preincubating 10 pg of affinity-purified
sialyltransferase with 1.2, 2.4, 6.0 and 12 pg antibody, in duplicate, for 30 min prior
to performing the standard SAT-l activity assay described above. Controls were run

with heat-inactivated enzyme and no enzyme.

75

Specificity of M12GC7. The specificity of M126C7 to a variety of Proteins
was examined by Western Blot and ELISA of M12GC7 . 1-10 pg of mouse IgG,
purified SAT-1, detergent-solubilized Golgi, 812-3 and 812-6 (Genzyme), a2-6
sialyltransferase (BMB), human transferrin, human fetuin, human glycophorin, soybean
lectin, and E. coli B-galactosidase, and E. coli acetate kinase were examined.
M12GC7 in PBS, pH 7.2, (for ELISA) or in TBS, pH 7.4, (for Western) containing
1% (w/v) BSA at a final concentration of 5 pg/ml was incubated under the appropriate
conditions. Immunodetection, with the appropriate reagents, was accomplished with
goat anti-mouse IgG alkaline phosphatase as a secondary antibody.

Immunoaffinity Purification of SAT-1. Anti-SAT-l monoclonal antibody (3
mg), isolated and characterized as described above, was bound to 6 ml of Affi-Gel 10
(BioRad) according to the manufacturer's instructions. The anti-SAT-l-Affi-Gel 10
affinity column was equilibrated with 25 mM sodium cacodylate (pH 6.5) containing
5% (v/V) LDAO, 10 mM MnClz, 1 mM PMSF, and 0.1 M NaCl. Rat liver Golgi
vesicles were prepared and detergent-solubilized, as described above except that several
cathepsin and thiol protease inhibitors (1 mM PMSF, 0.3 mM aprotinin, 1 mM
pepstatin, 1 mM leupeptin, 135 mM TLCK-HCl, 284 mM TPCK, and 2.8 mM E-64)
were added. SAT-1, from the detergent extract, was batch-adsorbed to the
immunoaffinity ligand overnight with gentle rocking at 4°C. The column was washed
with 10-20 column volumes of the same salt-containing buffer until no protein could be
detected by A280nm- The SAT-1 sialyltransferase was then eluted with 25 mM sodium
cacodylate (pH 6.5) containing 5% (v/v) LDAO, 10 mM MnC12 and 1.0 M NaCl.

4. ASSESSMENT OF PURITY AND BIOLOGICAL ACTIVITY

Verification of SAT-1 Homogeneity. SDS polyacrylamide gel electrophoresis
(31) was carried out on 5-20 pg of SAT-1 and silver-stained (32). Prior to
electrophoresis, the enzyme was removed from its detergent-containing buffer first by

precipitation in absolute ethanol (1:9 ratio sample to ethanol) at -20°C for 24-48 h

76

followed by acetone (HPLC grade) precipitation 12-18 h. With each precipitation step,
the sample was centrifuged at 14,000 x g for 30 min, all but 25 pl of supernatant
removed so as not to disrupt the protein pellet, and the sample taken to dryness in a
Savant speedvac. The standard proteins (BioRad) for SDS-PAGE were lysozyme
(14,400), soybean trypsin inhibitor (21,500), carbonic anhydrase (31,000), ovalbumin
(42,700), bovine serum albumin (66,200) and phosphorylase B (97,400).

Localization of SAT-1 Activity with Silver-Stained Protein. Duplicate SDS-
PAGE gels containing 5 pg SAT-1 were run for detection of CMP/LacCer affinity-
purified and M12GC7 immunoaffinity-purified SAT-1. Detection of SAT-1 was made
using either of two parallel approaches. First, enzyme activity of the protein was
assessed by cutting the gel lane into slices and incubating the samples in SAT-l assay
mixture to localize the enzyme activity with the silver-stained protein. Alternatively,
Western blots (described above) of SAT-1 on Immobilon (Millipore Corp.) were
detected with anti-SAT-l monoclonal antibody M12GC7 (5 pg/ml) and Goat anti-
mouse IgG alkaline phosphatase-conjugated secondary antibodies (BioRad).

SAT-1 Specificity. The specificity of SAT-1 to various glycoconjugates was
examined. The methods are described below.

Method #1: Sfiificity for Various Glyggsphingolipid Substrates. The
specificity of SAT-1 toward various glycosphingolipid substrates was examined by
direct assay of the sialyltransferase activity on substrates chromatographed on high-
performance-thin-layer-plates (HPTLC) as described (33), with some modifications.
Briefly, 10 nmol of each glycosphingolipid substrate was chromatographed on Merck
HPTLC plates containing fluorescence indicator. (Note: being of a harder
composition, this type of plate withstands the washing process better than plates
without indicator. Also, lipids can be detected prior to fixing by UV illumination or
iodination.) The lipids were chromatographed with chloroform/methanol/0.2% CaC12
(60:40:9, v/v/v). The HPTLC plate was dried in vacuo 18 h and then developed with

77

iodine to clearly define and mark the glycolipid boundaries. Once the lipid boundaries
were marked lightly with a pencil, the iodine was evaporated from the plate by
incubating the plate at 55°C for a few min. This treatment to remove the iodine did
not appear to affect the glycosphingolipids or their ability to serve as receptor substrate
(C. Sranka and R. Laine, Glycomed, Alameda, CA, personal communication). The
dried plates were immersed in n-hexane for about 30 sec, fixed in 0.1% (w/v)
polyisobutylmethylacrylate, PIBMA, (Aldrich Chemical Co., lot #03639KT) in n-
hexane for 30 sec, and then dried completely and stored at -20°C until used. The
PIBMA fixing soldtion was prepared from a stock solution of 10% (w/v) PIBMA in
chloroform. Prepared HPTLC plates were incubated with 0.1 mg/ ml purified SAT-1
enzyme in 25 mM sodium cacodylate (pH 6.5) containing 10 mM MnClz, 5% (v/v)
LDAO and 240 nmol of CMP-[14C4,5,6,7,3,9]-sialic acid with a specific activity of
22,200 dpm/nmol (NEN lot #2655-018, CMP-NeuNAc from Sigma lot #86F-793l). A
final volume of 1.0 ml was pipetted onto a 2.0 x 4.0 cm plate and incubated for 60 min
at 37°C in a 6 cm tissue culture dish. Following incubation, the plate was carefully
transferred to a clean dish and was washed five times with 5.0 ml phosphate-buffered
saline (PBS) containing 0.05% Tween 20 (ELISA grade). The positions of the
glycosphingolipids were detected with orcinol and the radioactivity quantitated by
scraping the glycosphingolipids into scintillation vials containing 0.2 ml double distilled
H20. Following sonication for 10 min, 5 ml Safety-Solve (Research Products
International Corp.) was added and the amount of [14C]-sialic acid transferred
determined on a Packard Tri-Carb 460C Liquid Scintillation System. Three controls
were with no acceptor, with no enzyme, and with boiled enzyme.

Method 492: Wes. SAT-1 Specificity
to various glycoprotein substrates was determined following incubation of 50 pg of
each glycoprotein with 10 pg immunoaffinity-purified SAT-1, 50 nmol of CMP-
[14C4,5,6,7,3,9]-sialic acid (NEN lot #2655-018, specific activity adjusted to 22,200

78

dpm/nmol with CMP-sialic acid Sigma lot #86F-7931) in a final volume of 0.1 ml in
25 mM sodium cacodylate (pH 6.5) containing 10 mM MnC12 and 5% (v/v) LDAO.
The reactions were terminated by the addition of 0.9 ml absolute ethanol (final
ethanol/H20 ,90: 10). The proteins were precipitated at -20°C for 48 h in 90% ethanol
followed by 18 h in acetone, as described. The amount of radiolabeled sialic acid
incorporated was determined by counting the protein pellets in scintillant.
Additionally, duplicates of the glycoproteins were subjected to SDS-PAGE followed by
staining with Coomassie Blue. The radioactivity associated with each glycoprotein was
determined by cutting the protein from the gel and counting the amount of [14C]-sialic
acid transferred in 5 m1 of Safety-Solve (Research Products International), sonicated 5
min and counted on a Packard Tri-Carb 460C Liquid Scintillation System. Controls
were run with a2-6 B-galactosyl sialyltransferase and the same glycoprotein acceptors.

Glycan Detection. Immunoaffinity-purified SAT-1 was electrotransferred onto
Immobilon and the Western blot was cut into 0.5 cm strips, each containing
approximately 0.7 pg/cm SAT-1. The SAT-1 was incubated with biotinylated lectins
obtained from E-Y Labs (San Mateo, CA) and digoxigenin-labeled lectins from
Boehringer (Indianapolis, IN) under the conditions suggested by the manufacturer for
analysis of the carbohydrate moieties of SAT-1.

Two-Dimensional Gel Electrophoresis. Analysis of rat liver Golgi SAT-1 by
2D SDS-PAGE was performed as described by Dunbar (34). The purified
sialyltransferase, 10 pg, was isoelectrofocussed in the first dimension in tube gels
according to the O'Farrell method (35) in the second dimension on 10-20% SDS-PAGE
according to the Laemmli system (31), as described by Dunbar (34) using the BioRad
mini 2D Protean II system. The carrier ampholytes were a mixture of 4 parts
Pharmacia Pharmalytes pH 5-8 and 1 part (LKB Ampholines pH 3.5-10/Pharmacia
Pharmalytes pH 3-10, 1:1). The 2D SDS-PAGE patterns were analyzed by the Bio

79

Image Visage 110 computerized digital image analysis system (Millipore, Milford,
MA).

80

RESULTS

Detergent Extraction. As described under "Materials and Methods,” Golgi
were prepared from rat liver homogenates by differential centrifugation and
discontinuous sucrose density gradient centrifugation. Analysis of the preparation for
enzyme markers showed that SAT-1 and thiamine pyrophosphatase, a Golgi enzyme
marker, were recovered from the same fractions. The rat liver Golgi, prepared by
established TEM procedures (36,37), were observed to be pure intact vesicles.
Sonication and gentle agitation of the Golgi vesicles in 25 mM sodium cacodylate
containing 10 mM MnClz and 15% (v/v) lauryl dimethylamine oxide (LDAO) gave the
most effective solubilization. The merits of this surfactant for the purification of SAT-
1 are described (29).

Preliminary Characterization of SAT-1. For optimal enzyme assay
conditions and affinity purification, SAT-l activity from rat liver Golgi vesicles was
characterized with respect to substrate binding, stability, requirements for maximal
activity and potential inhibitors. The data correlated well with those reported for SAT-
1 in studies using homogenates of embryonic chicken (38) and cultured hamster
fibroblasts (16). The apparent Km values for the CMP-sialic acid and LacCer
substrates were 0.26 and 0.11 mM, respectively; the Vmax values were 26.3 and 18.7
pmoles of GM3 formed per min per mg protein. SAT-1 was inhibited to the same
extent by either 1.0 mM CMP, 0.5 mM CDP or 0.1 mM CTP. The enzyme required
10 mM MnC12 for optimal activity . SAT-1 was most active at pH 6.5 in sodium
cacodylate with detectable activity in the pH range of 6.1 to 7.6.

Affinity Purification of SAT-l from Rat Liver Golgi. The affinity
purification of SAT-1 on CMP- and LacCer-Sepharoses is summarized in Table l.
SAT-1 was purified 43,000-fold over crude homogenate from detergent-extracted rat

liver Golgi. Conditions were Optimized for affinity purification based on preliminary

81

Table 1

AFFINITY PURIFICATION OF SAT-1

 

 

Fraction Volume Protein Units Specific Purification
(ml) (mg) (pmol/ min) Activity (-fold)
(units/ mg)
100g rat liver
Crude
Homogenate 325 34100 3750 0. 1 1 1.00
Post
Mitochondrial 365 17700 4500 0.25 2.28
Supernatant
Detergent
Extracted 57.0 73.0 1050 14.4 130
Golgi
CMPI(NaCl) 12.5 4.0 350 87.5 795
CMPII(CMP) 2.6 0.5 82 160 1450
LacCer(CMP) 1 1.0 1.0 4660 4710 42800

 

82

characterization of the enzyme activity in rat liver Golgi. Throughout the purification
procedure, SAT-1 sialyltransferase activity wasassayed with human RBC LacCer or the
commercially available LacCer containing N-palmitoylsphinganine.

SAT-1 activity was recovered from CMP I with a linear salt gradient between
0.36-0.4 M NaCl. The SAT-1 active fraction was pooled, desalted and
rechromatographed onto the CMP 11 column. SAT-1 was recovered at 0.1-0.15 mM
CMP using a linear CMP gradient. Further washing of the CMP 11 column with buffer
containing 5 mM and 10 mM CMP resulted in no further recovery of enzyme. The
final sialyltransferase affinity step was IacCer-aldehyde Sepharose. The enzyme was
eluted with 1 mM CMP.

Assessment of Affinity-Purified SAT-1 by Gel Electrophoresis. The 43,000-
fold purified enzyme gave nine polypeptides on a 12.5% SDS-PAGE (Fig. 2) with five
predominant molecular weight species at 68, 60, 43, 25 and 12 Kd. Similar size
heterogeneity, attributed to proteolytic degradation (6,21, for review 4), has been
reported for other purified sialyltransferases (9,10,18-21). Further, the electrophoretic
profile of the 40,000-fold affinity-purified enzyme could be varied with time and
temperature of solubilization in SDS-PAGE sample buffer (data not shown). Only four
polypeptides were resolved following solubilization of SAT-1 in sample buffer for 2
min at 95°C, while nine polypeptides were present in gels when the SAT-1 preparation
had been treated for 30 min at 37°C.

The enzyme activity of each of these polypeptides was examined following
electroelution from the gel. As judged by specific activity, 27% of the total activity
applied to the gel was recovered. The pooled 68 and 60 Kd polypeptides exhibited the
most activity (57% of the total recovered activity). A small amount of activity was
associated with the 25 and 12 Kd polypeptides. No activity was associated with the 43
Kd band.

83

Figure 2. Gel Electrophoresis of Affinity-Purified SAT-l. SDS-PAGE gel
electrophoresis was performed on 15 pg of SAT-l, treated as described under Methods
to remove the LDAO detergent and then electrophoresed under the conditions of
Laemmli (31) on a 12% SDS-PAGE gel. Five predominant molecular weight species
were resolved at 68, 60, 43, 25 and 12 Kd. Similar size heterogeneity has been
reported for other purified sialyltransferases (9,10,18-21).

84

 

 

 

w
. 't e_
.0
U8
. \I/ I _N
.O (m -...
..6 E ..f
._
F.03 m. In. _m
.0
.4 m
m D t we
m - oo
m0 IUXFm
l. . . . _o r.
w m m m m m _éno 85 32
O. O. O. O. O. O.

 

 

 

>._._mzmn_ ..EO omkommmoo
_t._.<m QMEEDQ >t_._Z_lt_l._<

85

Treatment of SAT-1 with cathepsin D gave an increases in the 56 Kd
polypeptide. Digestion of both the 60 Kd and 56 Kd species with cathepsin D and N-
glycanase gave an increase in the 43 Kd band, suggesting that they were the same
activity and that SAT-1 was a glycoprotein. A Cathepsin D-like activity has been
found to be associated with the conversion of the "mature" 47 Kd membrane-bound
form of the B-galactoside 02-6 sialyltransferase to its soluble 41 Kd form (39, for
review 4).

The ability to manipulate the electrophoretic pattern of the sialyltransferase
suggested there may be aggregation of the membrane protein when solubilized at high
temperature in SDS. Likely too is enhanced proteolytic degradation at 37°C even in
the denaturing conditions of SDS sample buffer. The loss of activity following elution
to verify the SAT-1 protein may have been due to loss of protein during electroelution,
failure to renature or production of an inactive proteolytic fragment. The 60 Kd
polypeptide was suspected to be the mature form of the enzyme and the other peptides
were either subunits of the mature form or proteolytic fragments.

Since the possibility of protein contaminants could not be ruled out, monoclonal
antibodies were raised to the 40,000-fold affinity-purified SAT-l. The advantages of
immunoaffinity purification of SAT-1 were that SAT-1 could be isolated from
contaminants and minimal time was required for the purification (4 days total),
lessening the amount of time the enzyme would be subjected to conditions allowing
proteolytic degradation.

Specificity of M12GC7 Anti-SAT-l Monoclonal Antibody. With these
several possible explanations for the heterogeneity of the 43,000-fold affinity purified
SAT-1 (i.e., contaminating protein(s), proteolytic degradation, subunits, etc.),
monoclonal antibodies were raised against this ”partially-purified" SAT-1 to resolve the
issue and purify SAT-l. As described under "Materials and Methods” mouse anti-rat

liver Golgi SAT-1 monoclonal antibodies were obtained. One monoclonal (M12GC7),

86

on the basis of its ability to immunoprecipitate and inactivate SAT-1, was chosen for
immunoaffinity purification of SAT-1. M12GC7 markedly inhibited SAT-1 (Fig. 3).

The specificity of MlZGC? was examined by Western Blot analysis. MIZGC7

was found to react only with SAT-l and Goat anti-mouse IgG. No reaction was
observed with ST2-6, ST2-3, human transferrin, human albumin, human fetuin, E. coli
B-galactosidase and acetate kinase (data not shown). The results were confirmed by
ELISA (data not shown).
Purification of SAT-1 to Apparent Homogeneity by Immunoaffinity
Chromatography. Anti-SAT-l monoclonal antibody, M12GC7, coupled to Affi-Gel
10 (BioRad) was used to purify SAT-l to apparent homogeneity. To reduce the
amount of fragmentation possibly caused during the mechanical disruption of the
enzyme, sonication of the detergent extract was reduced to 5 min. To limit endogenous
proteolysis of SAT-1 during purification, possibly by a cathepsin-like activity similar to
that identified for B-galactoside a2-6ST (40, reviewed 4), several cathepsin and thiol
protease inhibitors were kept in mixtures during the purification. All of these protease
inhibitors, with the exception of PMSF, inactivated the enzyme (data not shown).
However, they allowed recovery of a homogeneous protein, as assessed by SDS-PAGE
and verified by Western Blot to Immobilon and immunodetection with MIZGC7. The
apparent molecular weight of SAT-1 was approximated at 60,000 daltons as judged by
SDS-PAGE. This analysis is shown in Fig. 4.

The capacity of the anti-SAT-l Affi-Gel 10 column for SAT-l from detergent-
extracted Golgi, as tested so far, is about 100-150 pg SAT-1 from 20 g rat liver and
purification data from a subsequent preparation with only PMSF and without the
cathepsin and thiol protease inhibitors is summarized in Table 2. Purification of SAT-1
by immunoaffinity chromatography to 37,000-fold was found to be comparable to that
obtained by affinity chromatography on CMP- and LacCer-aldehyde- Sepharoses.

Employing the same immunoaffinity column, SAT-1 has also been isolated from human

87

Figure 3. Inhibition of SAT-1 activity by M12GC7 monoclonal antibody. A
monoclonal antibody was raised to the 40,000—fold affinity-purified sialyltransferase as
described under the Methods section. Increasing amounts of MlZGC7 were
preincubated for 30 min with 10 pg of the "partially-purified“ SAT-l enzyme. The
open squares represent SAT-l activity present in reaction samples without added
M12GC7. The inhibition of SAT-l by the addition of increasing amounts of M12GC7
is shown with closed diamonds. All SAT-l activity was abolished in samples containing

2 or more pg of MlZGC7. Controls were run with heat-inactivated enzyme and no

enzyme.

88

3000

2000

1 000

dpm [14C1-GM3 Formed

M1 2607

INHIBITION OF SAT-1

 

 

 

 

 

v r Y I

6 8
ul M12GC7 added

0'0

12

Figure 4. Assessment of Immunoaffinity-Purified SAT-l Homogeneity by SDS-
PAGE Gel Electrophoresis. Electrophoresis and Western blotting of 5 pg of the
immunoaffinity-purified SAT-1 was described in the methods. Verification of the
protein as SAT-l was by immunodetection with M12GC7 and alkaline phosphatase-
conjugated secondary antibody. Immunolocalization of SAT-l was necessary because
addition of protease inhibitors inhibited the enzyme both in standard assay protocols
and activity assays preformed on gel slices from a duplicate gel. Following

immunodetection, only one band was resolved at 60,000 daltons.

90

 

91

Table 2

IMMUNOAFFINITY PURIFICATION OF SAT-1

 

 

Fraction Volume Protein Units Yield Specific Purification
(m1) (mg) (pmol/min) (%) Activity (-fold)
(units/ mg)
20g rat liver
Crude 106 27880 5940 100 0.15 1.00
Homogenate
Post
Mitochondrial 88 3476 5100 86 0.8 5.33
Supernatant
Detergent
Extracted 17.2 ND. 1170 31 ND. -
Golgi
aSAT-l 1.2 0.1 440 7.4 5500 36670

 

92

KB cells. One note of caution, the proteases associated with these sialyltransferases can
also degrade the Affi-Gel bound antibody.

Specificity of SAT-l. The preferred substrate for the purified SAT-1 is LacCer
with glucosylceramide (GlcCer), galactosylceramide (GalCer) and asialoGMl also
serving as substrates, but to a lesser extent (Table 3). SAT-l appears to recognize the
saccharide Bl-O-Cer linkage, with the disaccharide being preferred over the
monosaccharide. There was not any significant amount of sialic acid (NeuAc)
incorporated into any other glycoconjugate examined. The radioactivity recovered with
GM1 and GDla was due to a minor asialoGMl contaminant in the glycosphingolipid
substrate. These data were verified following sialyltransferase activity assays with
glycosphingolipid immobilized on polystyrene (Falcon) microtiter plates (40) or by
SDS-PAGE for glycoprotein substrates. Control assays with ST2-6 were performed.

SAT-1 Glycan Detection. The carbohydrate moieties of SAT-l, bound to
Immobilon, were characterized with biotin- and digoxigenin-labeled lectins. These data
are summarized in Table 4. Positive binding of the biotinylated lectins and
digoxigenin-conjugated lectins enables the immunological detection of the carbohydrate
structures. There may be both N- and O-linked carbohydrate side chains as indicated
by the binding of DSA to the SAT-1 glycoprotein. The carbohydrate side chain(s) on
SAT-1 contain galactose, indicated by the positive reaction of BPA, GS-I, and PNA.
The carbohydrate structures may also contain GlcNAc as indicated by the strong
reaction with ConA and WGA. Further, the positive reaction of ConA indicates the
carbohydrate structures are branched N-glycans containing mannose. It is unlikely that
there are terminally linked mannose residues as the reaction with GNA was negative.
The presence of N-linked sugars is also indicated by a positive reaction of SAT-1 with
WGA, which recognizes N-acetyl chitobiose units of asparagine-linked
oligosaccharides. Further, binding of WGA could result from the presence of NeuAc.

Positive reaction with the LPA, SNA and MAA lectins, which recognize sialic acid

93

 

o.o

'—

 

~.o .eo..fi.o..eo<z_eoo.a_eo 5.289.853
o... 0 ed eeoeneozo
0.0 o o... e2.2a.e<zo.o.a.ee2...o<zgosaeo. 529893
28.5 228
0.0 u m.o 5203820 28.6
.ooefi.o..8<z_eoo.a_eo
N. a me 395.0..eo<z_eoo.a_eo.~. Beebeagzao £85292
2 S 3. £82
.oo..£.o..eo<2_eoo.a_eo
ea<.z.~.o<zo_o.~.e«2.22266...aao.
...” me I: 92.2...o<22o.~.e§.n.e<22or3.8. 522292
A... m ... seen
a. 85.22
3. .8 o. : 8.0.36.9...._eo.38<82.sao<z_eoo.588.8282. .6o
3 a. m... 30.0.ao_oe.a_eo8.«8<eez.r3.2238298 ....o
E 3. e... zoo.ago.......eo.e.~eo<az.sao<z_eo N...o
3 8 ma 8.0.2.29.5.88.8282. n....o
0.0 o o... zoo.32o.......eo.e.~eo<=e2e.~eo§ez.e. 2.22.892.83.8282. 5o
0... o o... .eoo.ago?...aoaoeoéezoéeogez?32239...aao £8
2.. 9a 2 no.0.ago...5393628298282. floo
8.8 one So .eo.o.ao.oe.a_eor3.228398 fogs?
2.. 88 3... 390228 soao
Se 8: 3m no.0. ago .866
0.8. 08... m8 30.02.29.258 soon.
3 8522
x oE x SESOEQ .oEq
€22... :28... 8.28.8... 8:825 .288...
3:23.. 2.8on 22.52

 

CLO—Ecumm 7.5m
a Eng...

94

Table 4

GLYCAN DIFFERENTIATION OF SAT-1 CARBOHYDRATE SIDECHAINS

 

 

Lectin Carbohydrate Reaction
Specificity
E-Y Labs - Biotinylated Lectins
BPA Gal & GalNAc +
ConA a-Man > oz-Glc > GlcNAc +
GS-I melibiose & a-D-Gal +l-
LPA NeuNAc +
PNA GalBl-BGalNAc > a—Gal & B-Gal +/-
SBA a- & B-GaINAC > 01- & B-Gal +
UEA-I a-L—Fucose +/-
WGA (GlcNAcBl-4GlcNAc)1-4 > B-GlcNAc > NeuSAc +
Controls
M12GC7 SAT-1 +
DBA Me—2-acetamide-2-deoxy-D-Gal + 1
BMB - Digoxigenin-labeled Lectins
GNA Mannose (terminally linked) -
SNA NeuAca2-6Gal or NeuAca2-6GalNAc +
MAA NeuAca2-3Gal +
PNA GalB(1-3)-GalNAc -
DSA GalB(1-4)GlcNAc-(in N—glycans) & +

GlcNAc-Ser/Thr

 

1 Detection of LDAO by lectin.

95

73“ .i-‘w'

residues, substantiate the presence of NeuAc on SAT-1 and suggest both a2-3 and a2-6
NeuAc linkages as defined by MAA and SNA, respectively. Reaction with DBA may
have been due to the presence of some LDAO in the SAT-1 preparation.
Two-Dimensional Gel Electrophoresis of SAT-l. The two-dimensional
protein pattern of rat liver SAT-l was established by standard 2D SDS—PAGE methods
(34,35,31,32). The 60 Kd SAT-1 was estimated to be in the pI range of 5.7 to 6.2
following computerized digital image analysis on the Bio Image Visage 110 using
internal molecular weight (BioRad) and carbamylated pI standards (Pharmacia) (Fig.
5A and B). The high degree of glycosylation of SAT-1 is indicated by the lateral
spread of the protein due to greater molecular and charge heterogeneity. Establishment
of the 2D pattern of SAT-l is a necessary step for analyzing post-translational covalent

modification of SAT-l by metabolic labeling experiments.

96

Figure 5. Analysis of SAT-l by 2D SDS-PAGE. SAT-1, 10 pg, was analyzed by
2D SDS-PAGE (34) following ethanol and acetone precipitation as described in the
text. The first dimensional IEF gel was according to the method of O'Farrell (35).
The ampholyte mixture was 1 part pH 3-10 and 4 parts pH 5-8. A standard 12%
polyacrylamide electrophoretic gel was run according to Laemmli (31) and silver-
stained by the method of Merril et al. (32). The approximate molecular weight and p1
range of SAT-1, 60,000 daltons and 5.7 to 6.2, respectively, were estimated following
a computerized digital image analysis on the Bio Image Visage 110 relative to internal
molecular weight (BioRad) and carbamylated pI (Pharmacia) marker. Panel A is the
SAT-1 protein pattern obtained and Panel B is the computer diagrammatic

representation of the detected protein and standards.

97

2-D SDS-PAGE (pH5-8) SAT-l

 

 

 

SDS IEF—-b pHS -------------------------- 3
l 97Kd- ‘j—j-——-
66 "' '"A"“SAT-1
MW 43 " 3 o a
3| - c
2| -
I I
6.7 “(J
pI
A=CPK
BtGADPH
086A
8 lEF—o pH5 .................................. 3
805
l 97Kd-
as - ...-45..-.
A Wcooooooooo
MW 43 - Ban..ooc°oooooo° O 0
3| - o c Ooooooooco O 0
2| -
| l
6.7 7.l
pI

DISCUSSION

The cell surface plays an important role in the regulation of mammalian cell
growth and differentiation. Gangliosides, ubiquitous sialic acid-containing
glycosphingolipids of the plasma membrane, are implicated in these phenomena.
Changes in ganglioside composition and metabolism are associated with cell
proliferation and oncogenic transformation (1-3). Further, marked variations in
ganglioside composition found during differentiation and aging (1), following
transformation (2) or treatment of cells with hormones (3) or chemicals (41-52) are
largely due to alterations of the glycosyltransferase activities. Purification of the
enzymes regulating specific glycosylation steps within the lumen of the endoplasmic
reticulum and Golgi apparatus may lead to an understanding of the functional role of
gangliosides in cellular proliferation.

It has been postulated that several glycosyltransferases may be involved in the
main level of regulation of ganglioside biosynthesis (53-57). Whether at the
transcriptional or post-translational level, regulation appears to be primarily dependent
on one galactosyltransferase, GalT-l, and three sialyltransferases, SAT-l, SAT-2 and
SAT-3 (53-57). These enzymes function at branch points between the asialo, a, b and
c series of glycosphingolipids and are involved in the conversion of GlcCer to LacCer,
LacCer to GM3, GM3 to GD3, and GD3 to GT3, respectively.

There has been limited success in the purification of these glycosphingolipid
glycosyltransferases. The acceptor glycolipid acid-Sepharoses first synthesized by
Makita and coworkers (5,7-8) have been found to be ineffective in the purification of
GalT-l (S. Chatterjee, personal communication) and SAT-1. Gu et al. (10) have
recently reported a 10,000-fold purification of SAT-2 using a glycolipid-acid column;
however, no substrate specificity was reported.

Purification of SAT-1 to apparent homogeneity was achieved by affinity
chromatography on CMP-Sepharose and LacCer aldehyde-Sepharose. The size

09

heterogeneity, as originally observed by SDS-PAGE, may have been due to proteolytic
degradation during purification by a cathepsin-like activity, as described by others
(21,39, for review 4). Digestion of SAT-l with cathepsin and N-glycanase indicated
this possibility.

Verification that SAT-1 was purified to apparent homogeneity required the
production of an anti-SAT-l specific monoclonal antibody, MIZGC7, capable of
immunoprecipitating and inactivating SAT-1. Using an M126C7-Affi Gel 10
immunoaffinity column, in the presence of several thiol and cathepsin protease
inhibitors, SAT-l was resolved as a homogeneous band by SDS-PAGE. The apparent
molecular weight of the 37,000-fold purified SAT-l was 60,000 daltons as judged by
SDS-PAGE. Immunoresolution of a minor band at 56,000 daltons suggests some
endogenous proteolytic activity may have been associated with the enzyme even after
antibody chromatography.

SAT-l catalyzes the transglycosylation reaction between its sugar nucleotide
donor, CMP-NeuAc, and oligosaccharide-lipid acceptor substrate GalBl—4Gchl-O—Cer.
SAT-1 appears to be specific for the saccharide-Bl—O-Cer linkage. LacCer is the
preferred substrate although GlcCer (Gchl-O—Cer), GalCer (GalBl-O-Cer), and asialo
GMl (GalBl-BGalNAcB1-4GalBl-4Gchl-O-Cer) serve as substrate to a lesser extent.
Sandhoff and coworkers (53-55) have suggested that SAT-1, SAT-2 and SAT-3 are
unique regulatory enzymes in the ganglioside biosynthetic cascade in that they
recognize the carbohydrate lipid linkage and not the carbohydrate "backbones" (53). In
contrast, SAT-4, the a2-3 sialyltransferase purified by Joziasse et al. (9), exhibits
specificity for the particular carbohydrate sequence, GalBl-3GalNAcBl-4Ga131-4Glc
and recognizes this sequence in gangliosides (GM1 and Gle) and glycoproteins.
Basu et al. (12) have categorized several glycosyltransferases into two distinct groups
based on their ability to recognize specific acceptor substrate sequences. These

recognition sites are either for a specific hydrophobic and carbohydrate sequence or for

100

a particular carbohydrate acceptor sequence. They termed these two classes of
glycosyltransferases HY-CAR and CAR, respectively (12).

The ability of SAT-1 to recognize GalCer is supported by early studies of Yu
and Lee (58) which suggest SAT-l catalyzes the transfer of sialic acid to both GalCer
and LacCer, forming 6M4 and GM3, respectively. In their experiments with a mouse
brain microsomal fraction, both substrates served equally well. The activity of our
purified SAT-1 is two—fold greater for LacCer than for GalCer (or GlcCer). It is
presumed that the carbohydrate/hydrophobic recognition site of SAT-1 prefers the
conformation of the disaccharide linked to ceramide over the monosaccharide-ceramide
linkage. This does not, however, discount the possibility of tissue specificity of the
SAT-1 sialyltransferase for these two acceptor substrates; GM4 is found almost
exclusively in the brain and GM3 is found in liver and many other tissues. Tissue-
specific expression of B-galactoside 012-6 and 012-3 sialyltransferases mRNA has been
observed by Paulson and coworkers (59) and by O'Hanlon et al. (60).

The observation that asialo 6M1 is also a substrate for SAT-1 is more difficult
to rationalize. While we expect SAT-l to be involved in the sialylation of terminal Gal
moieties, we have not ruled out the possibility that SAT-1 is capable of sialylating
either one or both Gal residues of asialo GMl, forming GM 1b: GMla, or GDla- We
are investigating this possibility through substrate competition analyses, specific
neuraminidase digestion of the labeled product and ability of asialo 6M2 to serve as
acceptor;

Two modes of regulation of SAT-1 have been proposed, (1) negative feedback
inhibition (53-55) and (2) covalent modification of SAT-l by a
phosphorylation/dephosphorylation mechanism (61). SAT-l activity is cell cycle-
dependent, increasing during late M/early GI (48). During early 61, both SAT-1
activity and GM3 expression are maximal (48,52,62). Apparently, the cyclic

accumulation of GM3 at 61 (53) may serve to inhibit progression through the cell cycle

101

since exogenously added GM3 has been shown to inhibit growth factor-induced
mitogenesis via EGF-receptor tyrosine kinase autophosphorylation (for review 3), an
event associated with cellular proliferation. HOpefully, the availability of SAT-1 will
make it possible to investigate the involvement of SAT-1 activity and GM3 ganglioside

in cellular proliferation and to identify the regulatory mechanism involved.

102

ACKNOWLEDGEMENTS

We thank Dr. Kimihiro Kanemitsu for assisting with the ozonolysis of LacCer.

We thank Carol Smith for her excellent secretarial support.

103

FOOTNOTES
* This work was supported in part by a research grant (DK12434) from the
National Institutes of Health.
+ To whom correspondence should be addressed: Dept. of Biochemistry,

Michigan State University, Rm. 401 Biochem. Bldg., E. Lansing, MI 48824-1319.

104

ABBREVIATIONS
2D, two-dimensional
BPA, Bauhinia purpurea agglutinin
CMP, cytidine 5-monophosphate
Con A, Canavalia ensiformis agglutinin
DBA, Colichos biflorus agglutinin
DSA, Datum stramonium agglutinin
Glycosphingolipid,“
GM3 , NeuAca2-3GalBl-4Gchl-1Cer
GNA, Galanthus nivalis agglutinin
GS-I, Gnfionia simplicifolia agglutinin
HPLC, high pressure liquid chromatography
HPTLC, high performance thin layer chromatography
IEF, isoelectric focusing
LDAO, lauryl dimethylamine oxide
LacCer, GalBl-4Gchl-1Cer
LPA, Limulus polyphemus agglutinin
MAA, Maackia amurensis agglutinin
MAPSII, monoclonal antibody purification system 11
NeuAc, neuraminic acid, also known as sialic acid
PAGE, polyacrylamide gel electrophoresis
PBS, Phosphate-buffered saline
PIBMA, polyisobutylmethylacrylate
PMSF, phenylmethane-sulfonyl fluoride
PNA, Arachis hypogaea agglutinin
SAT-1, CMP-sialic acid:lactosylceramide 012-3 sialyltransferase, also known as GM3

synthase

105

SDS, sodium dodecyl sulfate

SNA, Sambucus nigra agglutinin

ST2-3, B—galactoside 012-3 sialyltransferase

ST2-6, B-galactoside 02-6 sialyltransferase

TBS, TRIS-buffered saline

TLCK-HCl, (L-1—chloro-3-[4-tosylamido]-7-amino-2-heptanone-HC1
TPCK, L-l-chloro-3-[4-tosylamido]-4-phenyl-2-butanone

WGA, wheat germ agglutinin

*All abbreviations for Glycosphingolipids are according to the

Svennerholm nomenclature (63) and the IUPAC - IUB recommendations (64).

106

9'95”!"

99$?"

10.

ll.

12.

13.

14.

15.

16.

17.

REFERENCES

Hakomori, S. (1984) Ann. Rev. Immunol. 2, 103-126.

Hakomori, S. (1985) Cancer Res. 45, 2405-2414.

Hakomori, S. (1990) J. Biol. Chem. 265,.18713-18716.

Paulson, J. C. and Colley, K. J. (1989) J. Biol. Chem. 264, 17615-17618.
Taniguchi, N., Yokosawa, N., Gasa, S. and Makita, A. (1982) J. Biol. Chem.
257, 10631-10637.

Paulson, J. C. (1987) Trends Biochem. Sci. 14, 272-276.

Taniguchi, N. and Makita, A (1984) J. Biol. Chem. 259, 5637, 5642.
Taniguchi, N. and Makita, A. (1985) J. Biol. Chem. 260, 4908-4913.

Joziasse, D.H., Bergh, M.L.E., ter Hart, H.G.J., Koppen, P.L., Hooghwinkel,
G.J.M. and Van den Eijnden, DH. (1985) J. Biol. Chem. 260, 4941-4951.
Gu, X.-B., Gu, T.-J. and Yu, R.K. (1990) Biochem. Biophys. Res. Commun.
166, 387-393.

Clausen, H., White, T., Takio, K., Titani, K., Stroud, M., Holmes, 13.,
Karkov, J., Thim, L. and Hakomori, S. (1990) J. Biol. Chem. 265, 1139-1145.
Basu, S., Ghosh, S., Basu, M., Wang, 8., Das, K.K., Hawes, J.W., Li, Z.,
and Zhang, B. (1990) FASEB J. Vol. 4, No. 7, Abstract #2584.

Kanemitsu, K. and Sweeley, C. C. (1986) Glycoconjugate J. 3, 143-151.
Peterson, G.L. (1977) Anal Biochem. 83, 346-356.

Layne, E. (1957) Methods Enzymol. 3, 447-454.

Burczak, J.D., Fairley, J.L., and Sweeley, C.C. (1984) Biochim. Biophys.
Acta 804, 442-449.

Melkerson-Watson, L.J., Kanemitsu, K., and Sweeley, C.C. (1987)
Glycoconjugate J. 4, 7-16

107

18.

19.

20.

21.

22.

23.

24.

25 .

26.

27.

28.
29.

30.

31.
32.

Paulson, J.C., Beranek, W.B. and Hill, R.L. (1977) J. Biol. Chem. 252, 2356-
2362.

Sadler, J.E., Rearick, J.I., Paulson, J.C., and Hill, R.L. (1979) J. Biol. Chem.
254, 4434-4443.

Sadler, J.E., Rearick, 1.1., and Hill, R.L. (1979) J. Biol. Chem. 254, 5934-
5941.

Weinstein, J., de Souza-e-Silva, U., and Paulson, J .C. (1982) J. Biol. Chem.
257, 13835-13844.

Baubichon-Cortay, H., Broquet, P., George, P. and Louisot, P. (1989)
Glycoconjugate J. 6, 115-127.

Morre, D.J., Merlin, L.M., and Keenan, T.W. (1969) Biochem. Biophys. Res.
Commun. 37, 813-819.

Morre, DJ. (1971) Methods Enzymol. 22, 130—148.

Fleischer, S., and Kervina, M. (1974) Methods Enzymol. 31, 6-41.

Carey, D.J., and Hirschberg, C.B. (1980) J. Biol. Chem. 255, 4348-4354.
Leelavathi, D.E., Estes, L.W., Feingold, BS. and Lombardi, B. (1970)
Biochim. Biophys. Acta 211, 124-138.

Kundu, S.K. (1981) Methods Enzymol. 72, 174-204.

Melkerson-Watson, L. J. and Sweeley, C. C. (1990) Biochem. Biophys. Res.
Commun. 172, 165-171.

Fuller, S.A., Takahashi, M., and Hurrell, J.G.R. (1987) in Current Protocols
in Molecular Biology, Vol. #2, Ch. 11 (Ausubel, F. M., Brent, R., Kingston,
R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and Struhl, K., eds.)
Greene Publ. Asso. & Wiley-Interscience, NY, NY.

Laemmli, U.K. (1970) Nature 227, 680-685.

Merril, C.R., Goldman, D., Van Keuren, M.L. (1984) Methods Enzymol. 104,
441-447.

108

33.
34.

35.

36.

37.

38.

39.
40.

41.

42.

43.

45.

46.
47.

48.

49.

Gu, X., Gu, T., and Yu, R. K. (1990) Anal. Biochem. 185, 151-155.

Dunbar, B. S. (1987) in Two-Dimensional Electrophoresis and Immunological
Techniques, pp. 227-271. Plenum Press, New York.

O'Farrell, RH. (1975) J. Biol. Chem. 250, 4007-4021.

Roth, J., Bendayan, M., and Orci, L. (1978) J. Histochem. 26, 1074-1081.
Roth, J., Taatjes, D.J., Lucocq, J.M., Weinstein, J., and Paulson, J.C. (1985)
Cell 43, 287-295.

Kaufman, B., Basu, S., and Roseman, S. (1966) Methods Enzymol. 8, 365-
368.

Lammers, G. and Jamiesson, J .C. (1988) Biochem. J. 256, 623-631.
Holmgren, J ., Svennerholm, L., Elwing, H., Fredman, P. and Strannegard, 0.
Proc. Natl. Acad. Sci., U.S.A., (1980) 77, 1947-1950.

Moskal, J.R., Lockney, M.W., Marvel, C.C., Trosko, J.E., and Sweeley,
C.C. (1987) Cancer Res. 47, 787-790.

Deutsch, V. and Lotan, R. (1983) Exp. Cell Res. 149, 237-245.

Lotan, R., Lotan, D. and Meromsky, L. (1984) Cancer Res. 44, 5805-5812.
Payatte, R., Biehl, J., Toyama, T. and Holtzer, H. (1980) Cancer Res. 40,
2465-2474.

Koeffler, H.F., Bar-Eli, M. and Territo, M.C. (1981) Cancer, Res. 41, 919-
926.

Fukuda, M. (1981) Cancer Res. 41, 4621-4628.

Huberman, E., Heckman, C. and Langenbach, R. (1979) Cancer Res. 39,
2618-2624.

Burczak, J.D., Moskal, J.R., Trosko, J.E., Fairley, LL. and Sweeley, C.C.
(1983) Exp. Cell Res. 147, 281-286.

Leder, A. and Leder, P. (1975) Cell 5, 319-322.

109

50.

51.

52.

53.

54.

55.

56.

57.

58.
59.

61.

62.

63.

Ginsburg, E., Salomon, D., Sreevalsan, T. and Freese, E. (1973) Proc. Natl.
Acad. Sci., U.S.A. 70, 2457-2461.
Fishman, P.H., Simmons, J.L., Brady, RD. and Freese, E. (1974) Biochem.
Biophys. Res. Commun. 59, 292-299.
Macher, B.A., Lockney, M., Moskal, J.R., Fung, V.K. and Sweeley, C.C.
(1978) Exp. Cell Res. 117, 95-102.
Pohlentz, G., Klein, D., Schwarzmann, G., Schimtz, D. and Sandhoff, K.
(1988) Proc. Natl. Acad. Sci., U.S.A., 85, 7044-7048.
Yusuf, H.K.M., Schwarzmann, G., Pohlentz, G. and Sandhoff, K. (1987) Biol.
Chem. Hoppe-Seyeler 368, 455-462.
Yusuf, H.K.M., Schwarzmann, G., Pohlentz, G. and Sandhoff, K. (1987) Biol.
Chem. Hoppe-seyler 368, 455-462.
Schuz-Henninger, R., Ullmer, E., Prinz, C. and Decker, K. (1989) Eur. J.
Biochem. 185, 327-330.
Schuz-Henninger, R., Prinz, C. and Decker, K. (1988) Arch. Biochem.
Biophys. 262, 49-58.

Yu, R. K., and Lee, SH. (1976) J. Biol. Chem. 251, 198-203.
Paulson, J. C., Weinstein, J., and Schauer, A. (1989) J. Biol. Chem. 264,
10931-10934.
O'Hanlon, T. P., Lau, K. M., Wang, X. C., and Lau, J. T. Y- (1989) J. Biol.
Chem. 264 17389-17394.
Burczak, J.D., Soltyziak, R.M. and Sweeley, C.C. (1984) J. Lipid Res. 25,
1541-1547.
Chatterjee, 8., Sweeley, C.C. and Velicer, L.F. (1975) J. Biol. Chem. 250,
61-66.

Svennerholm, L. (1964) J. Lipid Res. 5, 145-155.
IUPAC-IUB Recommendations (1977) Lipids 12, 455-468.

110

CHAPTER 4

SPECIAL CONSIDERATIONS IN THE PURIFICATION OF THE GM3
GANGLIOSIDE FORMING ENZYIVIE, CMP-SIALIC
ACID:LACTOSYLCERAMIDE 012-3 SIALYLTRANSFERASE (SAT-l):
SOLUBILIZATION OF SAT-1 WITH LAURYLDIMETHYLAMINE OXIDE’

L.J. Melkerson-Watson and C.C. Sweeley1

Department of Biochemistry
Michigan State University
East Lansing, MI 48824

August 16, 1990

Biochem. Biophys. Res. Commun. (1990) 172, 165-171.

Running Title: Studies on the Purification and Regulation of SAT-1

SUMMARY

Lauryldimethylamine oxide (LDAO)2 was employed in the purification of the
GM3 ganglioside forming enzyme, CMP-sialic acid:lactosylceramide 012-3
sialyltransferase (SAT-l) (4). This detergent has advantages over the typically
employed Triton detergents in the solubilization and stabilization of this
sialyltransferase. Crude protein fractions solubilized from rat liver Golgi by several
such detergents are very similar in composition as determined by two-dimensional gel
electrophoresis. However, LDAO appears to activate and stabilize SAT-l activity. It
is possible that SAT-l activation involves the structurally similar hydrophobic moieties

and quaternary amino groups of LDAO and phosphatidylcholine.

INTRODUCTION

Detergent solubilization is essential for the purification of integral membrane
proteins. Detergents serve to replace the native lipids of the membrane bilayer about
the hydrophobic domain(s) of the protein so that routine biochemical and
chromatographic methods can be employed for purification. Many times the choice of
detergents is by trial and error, optimized to stabilize a particular enzyme activity.
Although the stability of a protein in detergent in not clearly understood, it is likely
related to the artificial conformation imposed on the protein by a particular detergent
environment. These properties of detergents have been reviewed (1,2) and recently
summarized (3). We report here the application of a nonionic/cationic detergent,
lauryldimethylamine oxide (LDAO) in the purification of a sialyltransferase, the GM3-
forming Golgi enzyme, CMP-sialic acid:lactosylceramide 012-3 sialyltransferase (SAT-

1).

METHODS

Materials: Lauryldimethylamine oxide (LDAO), also called Ammonyx LO,
was obtained from the Stepan Chemical Co, (Northfield, IL). Triton CF-54, B-
octylglucosidc and myosin were purchased from Sigma Chemical Co. (St. Louis, MO).
Triton X-100, Gelcode molecular weight markers and Extracti-Gel D were purchased
from Pierce Chemical Co. (Rockford, IL). CMP [14C4,5,6’7,3,9] sialic acid was
obtained from New England Nuclear (Boston, MA). Sep Pak C13 cartridges were
purchased from Waters (Milford, MA). Carbamylated p1 standard markers were from
Pharmacia (Piscataway, NJ). The reverse-phase RP-300, C3 HPLC column was
obtained from Applied Biosystems (Foster City, CA). All other reagents and chemicals
were of reagent grade.

Rat Liver Golgi: Rat liver Golgi was prepared as described previously (4)
using well-established procedures (5-7).

CMP-sialic acid:lactosylceramide 012-3 Sialyltransferase (SAT-l) Activity
Assay: SAT-l activity was assayed as previously described (8,9).

Two-Dimensional Gel Electrophoresis: Rat liver Golgi, 0.3% Triton X-100
extracted Golgi, 1% Triton CF-54 extracted Golgi, and 15% LDAO extracted Golgi
proteins were isoelectrofocussed in the first dimension in tube gels according to the
O'Farrell method (25) and in the second dimension on 10-20% SDS-PAGE according
to the Laemmli system as described by Dunbar (10). The carrier ampholytes were a
mixture of 2 parts Pharmacia Pharmalytes pH 5-8 and 1 part (LKB Ampholines pH
3.5-10/Pharmacia Pharmalytes pH 3-10, 1:1). The 2D SDS-PAGE patterns were
analyzed by the Bio Image Visage 110 computerized digital image analysis system
(Millipore, Milford, MA).

RESULTS AND DISCUSSION

LDAO was employed in the purification to homogeneity of CMP-sialic
acid:lactosylceramide 012-3 sialyltransferase (SAT-1) from rat liver Golgi (4). LDAO
is a nonionic/cationic detergent which protonates under acidic conditions (pH < 7)
conferring a net positive charge to its hydrophilic head group (3). It's use has been
primarily for the solubilization of photoreactive centers from chloroplasts (11,12).
This is the first report of the use of LDAO for the solubilization of a
glycosyltransferase.

Typically, Triton X-100 (13-19) or Triton CF-54 (20) have been the surfactants
employed in sialyltransferase solubilizations from Golgi vesicles. Two-dimensional
computerized digital image analyses of the proteins solubilized from rat liver Golgi
vesicles by either 0.3% Triton X-100, 1% (w/v) Triton CF-54 or 15% (w/v) LDAO
gave very similar protein patterns under the conditions employed relative to the
reference image obtained with unextracted rat liver Golgi. Each detergent resolved
about 300 silver-stained polypeptides within the pl range of 4.7 to 8.3. LDAO
resolved 284 polypeptides, Triton CF-54 318, and Triton X-100 295. The reference
pattern of rat liver Golgi protein (no detergent digestion) contained 315 proteins. The
Bio Image Visage 110 match-paired 156, 175 and 191 (for LDAO, Triton X-100 and
Triton CF-54, respectively) of these proteins to the reference image. The two-
dimensional map of the LDAO-soluble Golgi proteins is shown in Figure 1.

LDAO offers several advantages over the Triton surfactants. First, LDAO
enhances SAT-l activity from rat liver Golgi 12 to lS-fold higher than either Triton
CF-54, Triton X-100, or B-octylglucoside (Figure 2). This activation of SAT-1 may
reflect a stabilization of the solubilized enzyme in a structural motif similar to its native
conformation in a phosphatidylcholine (PC) membrane. PC, a major phospholipid of

the Golgi, has been shown previously to give a l6-fold activation of SAT-1 (21).

Figure 1. Two-Dimensional SDS-PAGE of Detergent—Solubilized Rat Liver Golgi
Proteins. Protein from rat liver Golgi was extracted in the presence of 3.0% (w/v)
Triton X-100, 1.0% (w/v) Triton CF-54 or 15% (v/v) lauryldimethylamine oxide. The
proteins solubilized from the Golgi membranes under these conditions were analyzed
by two-dimensional gel electrophoresis. First-dimension tube gels were performed
according to the method of O'Farrell (25) with an ampholyte mixture of 1 part pH 3-10
and 2 parts 5-8 on sixty micrograms of total protein. The second-dimension slab gels
were standard 10-20% Laemmli SDS-polyacrylamide gels. The two-dimensional
patterns were analyzed following silver staining (26) by computerized digital imaging
' using the Bio Image Visage 110 (Millipore/Bio Image, Ann Arbor, MI). Bio Image
parameter settings were filter width 15, spot threshold 6, minimum spot width 4,
minimum filter width volume 15, and minimum spot size 60. A 2D SDS-PAGE
reference image was from sixty micrograms of unextracted Golgi membrane. Panel A
shows the 2D pattern from 13.1 11g 15% LDAO soluble Golgi proteins. A composite
image of the analysis of unextracted rat liver Golgi proteins, LDAO extracted protein,
Triton X-100 extracted proteins and Triton CF-54 extracted proteins gave very similar
2D patterns (for discussion see text). Panel B illustrates a typical diagrammatic
representation plotted for these samples obtained from the digital image analysis on the
Bio Image Visage 110. This pattern is from the LDAO extracted Golgi proteins. The
boxed area on the plot indicates the p1 and M.W. range of the purified SAT-1, a
60,000 dalton glycoprotein in the pl range of approximately 5 .7 to 6.2 (4). The
molecular weight markers 200 Kd, myosin; 81 Kd transketolase; 40.5 Kd, creatinine
phosphokinase; 29 Kd, phosphoglucomutase; and 17.5 Kd myoglobin. The
carbamylation standards for determining pI were creatinine phosphokinase (pI 4.9-7.1);
glyceraldehyde-3-Phosphatedehydrogenase (pI 4.8-8.3) and carbonic anhydrase (pI 4.7-

6.7).

115

2203 5.3222
2902, 638.02

200 Kd

 

-200 kd
—40.5

-|75
o
o
o
c:

.Id
_

9
2
_

I

7.1

pl
67
1
oo
o 0
000
@000
- %°
8%0
.- oo
oo 8
3900 0
egg
0 o c
O 6)
oo
0::
08°
co
00

 

 

116

 

 

Figure 2. Effects of Various Surfactants on SAT-l Activity in Rat Liver Golgi.
Golgi vesicles were prepared from rat liver by well-established procedures (5 -7) and is
described in detail elsewhere (4). Increasing concentrations of LDAO, Triton CF-54,
Triton X-100 and B-octylglucoside were added to standard reaction mixtures (4,8-9)
with Golgi-enriched microsomes as the enzyme source and assayed immediately. The
GM3 product was recovered as described (8) and the SAT-l activity determined. The
inserted panel serves for comparison of SAT-1 activated by LDAO and PC and is
reproduced with permission from Journal Lipid Research 25, 1541-1547 and the
authors (21).

117

SPECIFIC ACTIVITY

(pmole .min".mg")

no
l60
150
I40
130
I20
I I0
IOO
90
so
70
so
50
4o
30
20
IO

 

 

 

 

 

 

 

0.7 1:25 215
Percent phospholipid (w/v)

0| 5 IO

°/. (W/v) DETERGENT

118

20

 

A second advantage of LDAO is that it does not absorb at A280nm1 as do the
Tritons, permitting the monitoring of protein elution at A280nm during purification
steps involving column chromatography. LDAO exhibits end-group absorption below
215 nm.

The hydrophobic nature and positive charge of LDAO below pH 7.0 confers

considerable stability on SAT-l for purification and storage. SAT-1 has been stored in
25 mM sodium cacodylate (pH 6.5) containing 15 % (w/v) LDAO at -80°C for periods
of 6—12 months without appreciable loss of activity (Table 1). We attribute this to the
hydrophobicity of the detergent and the structural similarity between LDAO and PC.
Glew and coworkers (private communication) have observed a similar stabilization by
charge with glucocerebrosidase, a membrane-bound lysosomal protein, and
phosphatidylserine (PS). Acyl CoA, a negatively charged amphipatic molecule,
confers the same effect as PS.
LDAO remains tightly associated with the enzyme, a factor which may complicate
some protein analyses. Several methods have been explored for the removal of LDAO.
LDAO will dialyze, but the process is slow. Extracti-Gel D (Pierce Chemical Co.) is
not a viable alternative for SAT-l purification because both LDAO and the
hydrophobic SAT-l sialyltransferase are adsorbed. TCA precipitation is ineffective.
Further, analysis of the glycan residues on SAT-1 (4) indicate SAT-1 is a glycoprotein
containing sialic acid. The use of TCA to precipitate sialylated proteins for
carbohydrate analysis is not recommended as sialic acid hydrolysis occurs under the
acidic conditions. Combined ethanol and acetone precipitations, which serve to
precipitate proteins by changing their solvation properties, removes sufficient LDAO
from SAT-l sialyltransferase to allow SDS-PAGE gel electrophoresis. For
electrophoretic analysis of SAT-1, 5-20 11g of the sialyltransferase is precipitated in
90% (v/v) ethanol at -20°C for 36-48 hrs followed by centrifugation at 14,000 x g for
30 min. The ethanol is carefully removed and the pellet taken to dryness in a

TABLE 1

STABILITY OF LDAO SOLUBILIZED SAT-l IN 15%(V/V) LDAO AT -20°C

 

 

FOR 6 MONTHS

Fraction Specific Specific % Activity

Activity Activity Remaining

to t6 after 6 mos. at -20°C

CH 0.43 0.07 16.5 %
PMS 0.74 0.23 31.3%
DE 0.52 0.56 106.3%
CMPI(NaCl) 1036 1 170 113 %
CMPII(CMP) 51 10 4980 97.5 %
LacCer(CMP) 15350 15640 102 %

 

1AA

Speedvac. The sialyltransferase pellet is resuspended in HPLC-grade acetone to desalt.
Precipitation of SAT-1 in acetone is allowed to procwd at -20°C for 18 hrs. Good
resolution of SAT-1 by 2D SDS-PAGE is achieved (4). However, the hydrophobic
nature of LDAO allows some of it to co-precipitate with SAT-1 even in
ethanol/acetone.

LDAO at high concentration is toxic. The LD50 of LDAO is 3.6 g/kg (value
supplied by the manufacturer). For immunological work, LDAO can be exchanged
with PC at a minimum of a 1:10 ratio of detergent to phospholipid and dialyzed. The
exchange of PC for LDAO decreases the toxicity of the immunogen. Further,
incorporating SAT-l antigen into PC liposomes has the advantage that the half-life of
the antigen in circulation is increased. This methodology permitted us to raise a
monoclonal antibody to rat liver SAT-l (4).

The most effective means of LDAO removal is reverse-phase HPLC.
Resolution of purified SAT-1 sialyltransferase from LDAO is achieved on a C3
microbore HPLC system (Applied BioSystems). SAT-l in 25 mM sodium cacodylate
(pH 6.5) containing 5% (v/v) LDAO is applied onto an Aquapore Cg, RP300 (250 x
1.0 mm x 7 am) reverse-phase microbore column and eluted on a 90 min linear
gradient established between 0.1% (v/v) trifluoroacetic acid and 90% (v/v) acetonitrile.
The differences in hydrophobicity permit the separation of SAT-1 from LDAO (Figures
3a and 3b). SAT-1 sialyltransferase exhibits greater hydrophobicity than LDAO and
elutes at 76% and 82 % of the B solvent (i.e., 68.4% and 73.8% acetonitrile). LDAO
resolves into several peaks, with the majority of the detergent eluting from 47% to 54%
of the B solvent (i.e., 42.3 to 48.6% acetonitrile). The two SAT-1 peaks, 56 Kd and
60 Kd by SDS-PAGE, respectively, were verified to be immunologically reactive on
Western Blots with M126C7, a specific anti-SAT-l monoclonal antibody (4). The 56
Kd polypeptide may be a proteolytic product of the 60 Kd SAT-l. SAT-1, like other

purified sialyltransferases, is subject to degradation by an associated endogenous

1’31

proteolytic activity (for review, 22 and references therein). Further, this activity is
present even after one week of storage at -80°C following SAT-1 purification. One
possible explanation is that a brush border protease associates with SAT-1 via the
LDAO detergent, enhancing its stability and activity. Preliminary investigation by
limited digestion of the HPLC 56 Kd and 60 Kd, M126C7 positive, SAT-1
polypeptides with 0.2 units cathepsin D and 2 units N-glycanase for 30 minutes at 37°C
prior to SDS gel electrophoresis resulted in formation of 56 Kd proteolytic and 43 Kd
deglycosylated polypeptide/products from both protein peaks obtained by HPLC. The
endogenous proteolytic activity is suspected to be a cathepsin D-like activity (23) and is
believed to be important in the release of the soluble catalytic domain of the
glycosyltransferase from the membrane anchor (22).

One other consideration for the use of LDAO is important. Since it is an
amphipathic compound, some is recoverable from SAT-l product assays on Sep Pak
C18 cartridges. This does not present a problem for optimal recovery of the GM3
product under the conditions described by Melkerson-Watson et al. (9). We
recommend, however, that an alternative method of analysis be employed for studies of
enzyme specificity as the LDAO and PC, used in the Sep Pak C13 recovery, have
relative mobilities on HPTLC in the range of some glycolipid substrates. Modifications
of the method of Yu et al. (24), reported elsewhere (4), were well-suited for analyzing

sialyltransferase activity with glycolipids chromatographed on HPTLC plates.

1flfi

Figure 3a and 3b. Reverse-phase C3 HPLC of SAT-1 vs. Reference Buffer
Containing 5% (v/v) Lauryldimethylamine Oxide (LDAO). Resolution of purified
SAT-1 sialyltransferase from LDAO was achieved on a C8 microbore HPLC system
(Applied Biosystems). SAT-1 in 25 mM sodium cacodylate (pH 6.5) containing 5%
(v/v) LDAO was applied onto an Aquapore C3, RP300 (250 x 1.0 mm x 7 11m)
reverse-phase microbore column and eluted on a 90 min linear gradient established
between 0.1% (v/v) trifluoroacetic acid and 90% (v/v) acetonitrile. Sample volume
injected was 250 111. Panel A illustrates a typical chromatogram obtained from buffer
containing LDAO. B shows the resolution of unique SAT-l proteins from the same

LDAO containing buffer.

123

 

 

 

 

 

If)

I?

REFERENCE

25mM sod. cacodylate (pH6.5)
with l%("/v)LDAO

 

 

 

 

 

 

 

0.02
E
C
E
9’,
a) h ‘7!
E ‘3
§ 0.0l "
.1:
<1 .-
0.00 l I . l I I I
40.0 50.0 60.0 70.0 80.0
Time (minutes)
r I9 SAT-l
9 g 56 KG
05 Protoolytic
u: Fragment
A 0.02 8 .
g ‘8
E
9‘, com
3 3 Mature
c [(5 Form
5: “ 1.
8 0.0l B
.0
<
0'00 I I . I I I
400 50.0 60.0 70.0 80.0

Time (minutes)

124

 

 

FOOTNOTES

1To whom correspondence should be addressed.

*This work was supported in part by a research grant from the National Institutes of
Health (DK12434).

2The abbreviations used are:

SAT-l, CMP-sialic acid:lactosylceramide a2-3 sialyltransferase, also known as GM3
synthase;

LDAO, lauryl dimethylamine oxide, also known as Ammonyx LO;

GM3, NeuAca2-3GalB 1-4Gch 1- lCer;

PC, phosphatidylcholine;

2D, two-dimensional;

SDS, sodium dodecyl sulfate;

PAGE, polyacrylaminde gel electrophoresis;

LD50, lethal dose with 50% survival;

HPLC, high performance liquid chromatography;

Kd, kilodalton.

1":

ACKNOWLEDGEMENTS

We thank Carol Smith for her excellent secretarial support.

We thank David Hicks, formerly of Integrated Separation Systems (assisting with the
2D SDS-PAGE of the detergent-solubilized proteins from Golgi vesicles (Fig. 2A).

1fl/

9°.‘I9‘S‘

10.

11.

12.

13.

14.

REFERENCES
Hjelmeland, L. M., and Chrambach, A. (1984) Methods Enzymol. 104, 305.
Neugebauer, J .M. (1990) Methods Enzymol. 189, 239-264.
Neugebauer, J .M. "Guide to the Properties and Uses of Detergents in Biol. and
Biochem.”, Calbiochem, San Diego, CA.
Melkerson-Watson, L. J., and Sweeley, C. C. (1990) "Purification to
Homogeneity by Immunoaffinity Chromatography and Partial Characterization
of the GM3 Ganglioside Forming Enzyme, CMP-Sialic Acid:Lactosylceramide
012-3 Sialyltransferase(SAT-l), From Rat Liver Golgi", submitted for
publication.
Morre, D. J. (1971) Methods Enzymol. 22, 130-148.
Fleischer and Kervina, M. (1974) Methods Enzymol. 31, 6-41.
Carey, D. J ., and Hirschberg, C. B. (1980) J. Biol. Chem. 255, 4348-4354.
Burczak, J. D., Fairley, J. L., and Sweeley, C. C. (1984) Exp. Cell Res. 147,
281-286. .
Melkerson-Watson, L. J ., Kanemitsu, K., Sweeley, C. C. (1987)
Glycoconjugate J. 4, 7-16.
Dunbar, B. S. (1987) in Two-Dimensional Electrophoresis and Immunological
Techniques, pp. 227-271. Plenum Press, New York.
Kendall-Tobias, M. W., and Seibert, M. (1982) Arch. Biochem. Biophys. 216,
255-258.
Bowes, J. M., Stewart, A. C., Bendall, D. S. (1983) Biochim. Biophys. Acta
725, 210-219.
Sadler, J. E., Paulson, J. C., and Hill, R. L. (1979) J. Biol. Chem. 254, 2112-
2119.
Sadler, J. E., Rearick, J. I., Paulson, J. C., and Hill, R. L. (1979) J. Biol.
Chem. 254, 4434-4443.

1"”!

15.

16.

17.

18.

19.

20.

21.

22.
23.
24.
25.
26.

Taniguchi, N., Yokosawa, N., Gasa, S., and Makita, A. (1982) J. Biol. Chem.
257, 10631-10637.

Taniguchi, N. and Makita, A.(1984) J. Biol. Chem. 259, 5637-5642.

Joziasse, D. H., Bergh, M. L. E., ter Hart, H G. J., Koppen, P. L.,
Hooghwinkel, G. J. M., and Van den Eijnden, D. H. (1985) J. Biol. Chem.
260, 4941-4951.

Clausen, H., White, T., Takio, K., Titani, K., Stroud, M., Holmes, E.,
Karkov, J., Thim, L., and Hakomori, S. (1990) J. Biol. Chem. 265, 1139-
1145.

Gu, X.-B., Gu, T.-J., and Yu, R.K. (1990) Biochem. Biophys. Res. Commun.
166, 387-393.

Weinstein, J., de Souza-e-Silva, U., and Paulson, J. C. (1982) J. Biol. Chem.
257, 13835-13844.

Burczak, J. D., Soltyziak, R. M., and Sweeley, C. C. (1984) J. Lipid Res. 25,
1541-1547.

Paulson, J. C., and Colley, K. J. (1989) J. Biol. Chem. 264, 17615-17618.
Lammers, G. and Jamiesson, J. C. (1988) Biochem. J. 256, 623-631.

Gu, X., Gu, T., Yu, R.K. (1990) Anal. Biochem. 185, 151-155.

O'Farrell, P.H. (1975) J. Biol. Chem. 250, 4007.

Merril, C.R., Goldman, D. and Van Keuren, M.L. (1982) Electrophoresis 3,
17.

1G0

CHAPTER 5
SPECIAL CONSIDERATIONS IN THE PURIFICATION OF THE GM3
GAN GLIOSIDE-FORNIIN G EN ZYIVIE, CMP-SIALIC

ACID:LACTOSYLCERAMIDE 022-3 SIALYLTRANSFERASE (SAT-l):
EFFECTS OF PROTEASE INHIBITORS ON RAT HEPATIC SAT-1 ACTIVITY

L. J. Melkerson-Watson and C. C. Sweeley

Department of Biochemistry

Michigan State University
East Lansing, MI 48824-1319

November 6, 1990

Running Title: Studies on the Purification and Regulation of SAT-1

SUMMARY

Co-purification of an endogenous proteolytic activity has been proposed as the
eause for the size heterogeneity of sialyltransferases. Reported herein are results on the
effects of various protease inhibitors, sulfhydryl-reducing agents and antimicrobial
agents on SAT-1 activity. Addition of protease inhibitors to immunoaffinity-purified rat
liver SAT-1 dramatically affects its activity. All protease inhibitors examined, with the
exception of PMSF, inhibited the purified enzyme. The most inhibitory were the
cysteine (thiol) protease inhibitors. This effect is less spectacular when the effect of
these inhibitors was studied on SAT-l activity in Golgi—enriched microsomes, although
the inhibition was greatest by the cysteine protease inhibitors. One dramatic effect,
found in both cases, was the apparent activation of SAT-1 activity in the presence of B-

mercaptoethanol.

130

INTRODUCTION

Intracellular proteases perform a variety of highly controlled, necessary
biologieal functions, such as the regulation of cellular proliferation (for review, 1).
For example, in glycosphingolipid metabolism, specific hydrolases cleave bioreactive
groups from the carbohydrate or lipid moieties. The catabolites formed alter the
modulation of the cell surface. For example, the cell-dependent degradation of GM3
ganglioside by Sialidase to lactosylceramide (LacCer) (2-5) or by cerebrosidase to lyso-
GM3 (6), may relieve GM3 inhibition of EGF-receptor tyrosine kinase
autophosphorylation, thus allowing cell proliferation to continue (7,8). GM3 and its
catabolites also affect the protein kinase C (PK-C) signal transduction mechanism in a
manner antagonistic to that observed for EGF-receptor (for review, 9). Catabolites of
GM3 ganglioside are postulated to be internalized and recycled back to the endoplasmic
reticulum and Golgi for resynthesis of GM3 (4,10).

Another function for proteolytic processing in complex carbohydrate
metabolism has been proposed as a mechanism for the release of soluble, active forms
of glycosyltransferases to the extracellular matrix through the site-specific cleavage of
these enzymes within the ”stem" region (11,12). Glycosyltransferases are the enzymes
responsible for extending the carbohydrate chains of glycoproteins and glycolipids.
The presence of endogenous proteases within the Golgi, responsible for these
cleavages, complicate purification of intact glycosyltransferases from the Golgi. We
report here our observations made during the purification of CMP-sialic
acid:lactosylceramide 012-3 sialyltransferase (SAT-l), on the effects of various protease
inhibitors on SAT-1 activity.

131

MATERIALS AND METHODS

Materials - Cytidine 5'-monophosphate sialic acid, (CMP-[14C4,5,6,7,3,9]-
sialic acid, 286.5 mCi/mmol) was purchased from New England Nuclear (Boston, MA)
and its specific activity adjusted to 22,200 dpm/nmol with unlabeled CMP-sialic acid
from Sigma (St. Louis, MO). The sphinganine-containing form of lactosylceramide
was purchased from Sigma and used in activity assays. Ammonyx LO
(lauryldimethylamine oxide, LDAO) was obtained from the Stepan Company
(Northfield, IL). EDTA, azide and sodium cacodylate were purchased from Sigma
Chemical Company (St. Louis, MO). PMSF, APMSF, Leupeptin, aprotinin,
pepstatin, TPCK, TLCK-HCl, 02-macroglobulin (and 02-macroglobulin carrier-fixed),
E-64 and dithiothreitol (D'IT) were purchased from Boehringer Mannheim
(Indianapolis, IN). Ep-459 and Ep-475 were gifts from Dr. Hanada of Taisho
Pharmaceutical (Tokyo, Japan). Ultrapure electrophoretic-grade B-mercaptoethanol
was from Bio Rad (Richmond, CA). Sep Pak C13 cartridges were purchased from
Waters Associates (Milford, MA). All other reagents were ultrapure or ACS grade
from commercial sources.

Preparation of Golgi-enriched Microsomes. Golgi vesicles were isolated
from rat liver by well-established procedures (13-17) as described in detail (18) in 25
mM sodium cacodylate (pH 6.5) containing 0.25 M sucrose in the absence of protease
inhibitors.

Purification of CMP-sialic acid:lactosylceramide 02-3 Sialyltransferase.
CMP-sialic acid:lactosylceramide 02-3 sialyltransferase (SAT-1) was purified from rat
liver Golgi by immunoaffinity chromatography using MIZGC7 anti-SAT-l monoclonal
antibody coupled to Affi Gel 10 as previously described for the purification of SAT-1
from detergent-extracted rat liver Golgi vesicles (18).

CMP-sialic acid:lactosylceramide 012-3 sialyltransferase (SAT-1) Activity
Assay. SAT-1 assays, with radiolabeled CMP-sialic acid as the donor substrate and

132

lactosylceramide as the acceptor substrate, were carried out by adsorbing 5 nmol
LacCer to the microtiter plates in 50% ethanol as described (19). The plates were
washed with PBS and blocked with 5% (w/v) bovine serum albumin (BSA)Vfor 30 min.
After rinsing, 20 nmol of CMP-[14C4,5,6,7,3,9]-sialic acid (New England Nuclear,
lot If 2655-018, specific activity adjusted to 20,000 dpm/nmol) was added. To the
reaction wells were added sonicated mixtures containing the following components: 25
111 of assay buffer (200 mM sodium cacodylate (pH 6.5), 20 mM MnClz and 0.3%
(v/v) lauryldimethylamine oxide (LDAO) and 25 [.11 of SAT-1 enzyme or Golgi (0.005-
0.1 mg protein). The plate was covered and incubated in a humidified environment at
37°C for 2 hr. Following the incubation period, the reaction mixture was removed and
the microtiter wells washed 3 - 5 times with PBS. The wells were put into 5 ml Safety-
Solve liquid scintillant and the amount of [14C]-GM3 formed counted in a Packard
(model #460C) liquid scintillation counter.

133

RESULTS AND DISCUSSION

The homogeneity of the affinity purified SAT-l was extremely difficult to
verify because of the size heterogeneity of the enzyme in electrophoretic patterns (18
(see chapter 3)). The pattern generated could be varied depending on the temperature
and time of solubilization in SDS-PAGE sample buffer.

While there are several explanations for these data, similar size heterogeneity
has been observed with other purified sialyltransferases (Table 1). Proteolytic
degradation and mechanical disruption of the enzymes from the Golgi during their
purifications were considered as possible causes for the heterogeneity of these purified
enzymes. Recently, this concept of proteolytic processing was addressed in a review
by Paulson and Colley (12). The cDNA's from four known glycosyltransferases
exhibit some homology with regard to gross structure. They all possess a cytosolic
NH2 terminus, a hydrophobic transmembrane domain, a ”stem" region, and a Golgi
lumenal COOH terminus. The ”stem" region between the transmembrane and catalytic
domain has been implicated as the necessary element in the anchoring and targeting of
the glycosyltransferase in the protein transport between the endoplasmic reticulum and
the Golgi. It is also the apparent site of proteolysis (26). The endogenous proteolytic
activity, suggested to be a cathepsin D-like activity (27), releases the catalytically
active C-terminus of GalBl-4GlcNAc 02-6 sialyltransferase (CMP-N-
acetylneuraminatezll-galactoside 02-6 sialyltransferase, EC 2.4.99.1) from the lumenal
face of the trans-Golgi for transport out of the cell during acute-phase response.

Therefore, to minimize proteolytic degradation during the purification of SAT-
1, anti-SAT-l specific monoclonal antibody, M120C7, (18) was used to

immunoaffinity-purify the enzyme from lauryldimethylamine oxide-extracted

134

TABLE 1

SIZE HETEROGENEITY OF SIALYLTRANSFERASE

 

 

Reference Specificity Source Fold Molecular
Purification weights
(Kd)
Paulson et al. (1977) GalBl-4GlcNAc Bovine 440,ooo-x 561
JBC 252, 2356. «new Colostrum 43
Sadler 11 at. (1979) Gal 1.2.351 Porcine 92,2oo-x 491
JBC 254, 4434. Submaxillary 44
Gland
Sadler et al. (1979) GalNAc 02-681‘ Porcine 117,ooo—x 172
BC 254, 5934. Submaxillary 160
Gland 100
so
69
56
Weimtein et al. (1982) GalBl-4GlcNAc Rat 23,ooo-x 47l
JBC 257, 13335. M Liver 43
Weinstein et al. (1932) GalBl-3(4)GlcNAc Rat 860,000-X 561
JBC 257, 13335. 1.2-351‘ Liver 44
Joziasse mt. (1985) GalBl-3GlcNAc Human 20.00611 651
JBC 260, 4941. a2-3sr and Placenta 43
02-3SAT (SAT-4) 41
40
Ga et al. (1990) NeuAc02-3GalBl-4Gchl-1Cer Rat ro,ooo-x 632
BBRC 166, 337-393. a2-88AT (SAT-2) Liver 59%
55
512
432
Mailman-Watson GalBl-4Gchl-1Cer Rat 42,soo-x 6t)1
and Sweeley 02-3SAT (SAT-l) Liver 56
(1990) BC, in press 50
47

 

lApparent molecular weights of the purified sialyltransferases.

2Molecular weight values were estimated form the SDS-PAGE pattern in Figure 3 of Gu, et al (25).

135

Golgi membrane proteins (28) in the presence of several protease inhibitors including
leupeptin, pepstatin and E64 (and its analogs Ep-459 and Ep-475, gifts from Dr.
Hanada of Taisho Pharmaceuticals, Japan (29)), all of which are potent cathepsin
inhibitors. The enzyme was purified to homogeneity with an apparent molecular
weight of 60,000. The enzyme was inactivated, but was immunologically reactive on
Western blots with M12GC7. We speculated that this inactivation was due to one or
more of the cysteine (thiol) protease inhibitors (TLCK-HCI, TPCK, E-64, or Ep-459),
which is consistent with the finding that a thiol (R-SH) group has been found in the
region of the CMP-NeuNAc binding site of B—galactoside 02-6 sialyltransferase (30).

A summary of the effects of these and other protease inhibitors, as well as some
sulfhydryl reducing and antibacterial agents used in SDS-PAGE and protein
chromatography column preservation, are listed in Table 2 along with their effect on
SAT-l activity. SAT-1 was immunoaffinity-purified from LDAO-extracted rat liver
Golgi-enriched fraction as previously described elsewhere (18). No protease inhibitors
were added to the purification buffer. Various protease inhibitors, at the concentrations
specified, were added to immunoaffinity-purified SAT-l and assayed for activity (Table
2). Of the inhibitors tested, only the serine protease inhibitor PMSF, common in most
glycosyltransferase purifications (18,20-25), did not inhibit the enzyme. All other
protease inhibitors significantly inactivated the purified SAT-1 48-87% under the
conditions employed. The most inhibitory substances were the cysteine (thiol) protease
inhibitors (leupeptin, TLCK-HCl, TPCK, E-64, Ep—459, and Ep—475), suggesting that
SAT-1 may also contain a thiol group in or near its sugar nucleotide-binding site, as
had been reported for B-galactoside 02-6 sialyltransferase (30). Some sequence
homology has been reported for the sugar-nucleotide binding region in other

glycosyltransferase (for review, 12).

136

EFFECTS OF PROTEASE INHIBITORS ON IMlWUNOAFFINITY-PURIFIED

TABLE 2

SAT-1 ACTIVITY

 

 

Inhibitor Inhibitor Specificity Concentration 14 dpm % of
[ Cl-GM3 Control
Control none 0.0 850 j; 30 100 %
PMSF serine proteases 1000 pM 1120 132 %
APMSF serine proteases 20 11M 440 52 %
Aprotinin serine proteases 0.3 11M 220 26 %
Leupeptin serine and thiol proteases
(e.g., cathepsin B & L) 1 M 240 28 %

TLCK-HC] trypsin & thiol proteases 135 11M 230 27 %
TPCK chymotrypsin & thiol proteases 284 11M 280 33 %
E-64 thiol proteases

(e.g., cathepsin B) 2.8 nM 250 29 %
Ep-459 thiol proteases

(e. g., cathepsin D) 2.8 11M 210 25 %
lip-475 thiol proteases 2.8 11M 140 I6 %
Pepstatin A acid proteases

(e. g., cathepsin D) 1 11M 240 28 %
az-Macroglobulin general endoproteases 1 unit 280 33 %
EDTA metalloproteases 100 11M 250 29 %
D'I'I‘ sulfhydryl reducing agent 100 11M 90 11%
B-ME sulfhydryl reducing agent 180 14M 3330 392 %
Azide antimicrobrial agent 1000 11M 110 13 %

 

137

Pepstatin, a potent inhibitor of cathepsin D, also inhibited SAT-1. the enzyme.
Cathepsin D, a thiol protease with a heavy metal requirement, exhibits a preference for
peptides flanked by hydrophobic amino acid residues (1). Addition of EDTA can
inhibit cathepsin D proteolytic degradation, but EDTA also inhibits SAT-1 activity,
since SAT-1 has a divalent cation requirement for activity (18,31). A 10,000-fold
purification of a related glycolipid sialyltransferase, GD3 synthase (NeuAC02-3Ga181-
4Gchl-1Ceramide 02-8 sialyltransferase, SAT-2), has recently been reported (25); this
purification was carried out in the presence of 1 mM EDTA and 10 mM 8-
mercaptoethanol (ll-ME). Analysis of SAT-2 by SDS-PAGE showed size
heterogeneity. The predominant molecular weight was reported to be 55,000 daltons
(25). The combination of EDTA and B-ME in their purification buffer apparentlydid
not inhibit proteolysis.

Our investigation of the effects of sulflrydryl reducing agents, B-ME and
dithiothreitol (DTT), commonly used in SDS-PAGE sample buffer, indicates that the
addition of B—ME at a final concentration of 0.18 mM enhances SAT-1 activity about 4-
fold relative to control. Thus, there is potential use of B-ME in concert with the
appropriate concentration of protease inhibitors (e.g. , EDTA or pepsatin) for the
inhibition of cathepsin D proteolytic degradation of sialyltransferases during
purification while maintaining active fractions for monitoring purification of these
enzymes.

In a previous study, summarized in Table 3, are the results of adding the same
concentrations of protease inhibitors to Golgi-enriched microsomal fraction. The
inhibition of SAT-1 activity was not as dramatic. SAT-1 activity was stable in
microsomes treated with 1 mM PMSF, 1 IIM leupeptin, 1 11M pepstatin, 1 unit 02-
macroglobulin, and 2.8 11M E—64. The cysteine (thiol proteinase inhibitors, TLCK-
HCl, TPCK, Ep-459 and Ep-475 inhibited SAT-1

138

TABLE 3

EFFECTS OF PROTEASE INHIBITORS ON SAT-l ACTIVITY IN GOLGI-

 

 

ENRICHED MICROSOMES
IIflIibitor Inhibitor Specificity Concentration 14 dpm % of
[ Cl-GM3 Control
Control none 0.0 104 j; 18 100 %
PMSF serine proteases 1000 11M 108 j; 19 104 96
APMSF serine proteases 20 11M 86 j; 11 83 %
Aprotinin serine proteases 0.3 11M 71 j; 12 68 %
Leupeptin serine and thiol proteases 1 11M 220 j; 36 212 %
(e.g., cathepsin B & L)
TLCK-HCI trypsin & thiol proteases 135 11M 76 11:17 73 %
TPCK chymotrypsin & thiol proteases 284 11M 82 j; 18 79 %
E-64 thiol proteases 2.8 11M 155 i 87 149 %
(e.g., cathepsin B)
Ep-459 thiol proteases 2.8 11M 65 j: 13 63 %
(e. g., cathepsin D)
lip-475 thiol proteases 2. 8 11M 90 j; 13 86 %
Pepstatin A acid proteases 1 11M 154:1:63 148%
(e. g., cathepsin D)
02-Macroglobulin general endoproteases 1 unit 112 i 42 108 %
EDTA metalloproteases 100 11M 66 j; 6 63 %
D'IT sulfhydryl reducing agent 100 11M 66 i 6 63 %
B-ME sulfhydryl reducing agent 180 11M 1 170 i 155 1 120 %

 

139

20—40%. EDTA (100 IIM) inhibited SAT-1. The greatest inhibition of SAT-1 activity
when assayed in Golgi-enriched microsomes was observed at 63% of control following
the addition of either Ep-459 or EDTA. Addition of 0.1 mM B-ME resulted in an 11-
fold enhancement of SAT-l activity in intact Golgi.

Not only should the effects of protease inhibitors on the inhibition of proteolytic
degradation and enzyme activity be considered, but so should the differential effects of
detergents on both proteinase activity(ies) during purification as well as the detergent
effects on the glycosyltransferase being purified. We recently reported the effects of
lauryldimethylamine oxide (LDAO) on SAT-l (28 (see chapter4)). SAT-l activity and
stability are dependent upon the concentration of LDAO used. Detergents have also
been shown to stabilize and activate proteinase activity. Arribas and Castano (32)
found that low concentrations of detergents (0.01%) such as Triton X-100, SDS and
CHAPS activate the hydrolysis of protein substrates by proteases, while higher
concentrations (0.1%) inhibit degradation. A similar argument may apply to LDAO.
LDAO may stabilize the sialyltransferase activity and, in addition, stabilize (and
potentially activate) the endogenous proteolytic activity associated with the purification
of these enzymes. Another possibility is that the associated endogenous protease is a
lysosomal proteinase (like cathepsin D) which co-purifies with SAT-1 in Golgi-enriched
microsomes through the disruption and mixing of the Golgi vesicles with light
lysosomes during sonication and detergent extraction. Dawson and coworkers (33)
have studied the subcellular distribution of cathepsin D, as well as its pre-pro and pro
forms, in human fibroblasts using Percoll density gradients. Subcellular fractionation
of cathepsin D and its intermediates were found uniformly distributed throughout the
gradient with a slight enrichment of the mature active form in more buoyant fractions.
Further, treatment of pre-pro-forrns of Cathepsin D with cysteine (thiol) proteinase
inhibitors, Ep-459 or leupeptin, caused inhibition of cathepsin D processing.
Treatment with Ep-475 had no effect.

140

Another important effect on proteinase activity in relation to SAT-l and its
purifieation is the stimulation of some proteases by ATP (for review, 1). Two classes
of ”ATP—dependent" proteases have been characterized; one requires ATP for
stabilization, the second requires ATP for hydrolysis. ATP activation of cathepsins D
and L has been reported (34,35). One mechanism reported for cathepsin L requires
binding of ATP to the protein substrate, which increases susceptibility of the protein to
proteolytic degradation (36).

Activation of SAT-1 through the phosphorylation of tyrosine residue(s) has
been proposed as a mechanism of regulation of cell growth by the increase of cell
surface GM3 ganglioside during 61 of the cell cycle (37 (see chapter 6)). ATP has
been demonstrated to enter the Golgi (38). Potentially, ATP-stabilization of protease
activity or ATP—dependent hydrolysis of enzyme by proteases may play a role in the
mechanism for the proteolytic degradation associated with SAT-1 sialyltransferase

during its purification.

141

CONCLUSIONS

From these observations and considerations, SAT-1 purifications are now
carried out under the following conditions to inhibit proteolytic degradation during its
purification. SAT-1 is typically immunoaffinity purified from 20 g (wet weight) rat
liver which has been perfused with 25 mM sodium cacodylate (pH 6.5) containing 0.25
M sucrose, 1 mM PMSF, 1 11M leupeptin and 1 11M pepstatin A. Following
homogenization in five volumes of the same buffer, the nuclei, mitochondria and
cellular debris are removed by centrifugation at 5000 x g for 10 min. The Golgi-
enriched microsomes are collected by centrifugation at 100,000 x g and detergent-
extracted with LDAO as previously described (18). The LDAO-soluble proteins are
reacted (batch method) with 2 ml carrier-fixed 02-macroglobulin for 30 min and
proteases liberated during detergent extraction are coupled to carrier-fixed 02-
macroglobulin which is removed following centrifugation (as described by the
manufacturer, BMB).

The relationship between the effects of protease inhibitors on the size
heterogeneity of SAT-1 and on SAT-l glycosyltransferase activity is unknown. The
data are consistent, to a first approximation, with the hypothesis that proteolytic
cleavage of SAT-l during purification is primarily due to the action of a cathepsin-like
thiol protease. Inhibition of SAT-l glycosyltransferase activity by protease inhibitors,
and activation by B-mercaptoethanol, may occur by modification of thiol residues
necessary for SAT-l activity. Thus, attempts to inhibit protease activity generally lead
to coincident inhibition of SAT-l catalytic activity. The inhibition of SAT-l catalytic
activity by 02 macroglobulin, however, does not fit this pattern, and remains to be
explained.

The correlation between inhibition of SAT-1 proteolysis and catalysis may also
imply a physiologically significant interaction between SAT-1 and a protease activity

for which several speculative mechanisms can be envisioned: a) a specific protease

142

activity may be required for SAT-l activity, and may be involved in the regulation of
SAT-1 acivity during the cell cycle, b) an endogenous protease may form part of a

complex with SAT-1, or c) SAT-l may possess endogenous protease activity.

143

FOOTNOTES
1To whom correspondence should be addressed.

*This work was supported in part by a research grant from the National Institutes of
Health (DK12434).

2The abbreviations used are:

APMSF, (4-amidino-phenyl)-methane-sulfonyl fluoride;

B-ME, 2-mercaptonethanol;

CMP, cytidine 5' monophosphate;

D'IT, dithiotheitol;

EGF , epidermal growth factor;

GD3, NeuA002-8NeuAca2-3Galfl1-4Gchl-1Ceramide;

GM3, NeuAc02-3GalBl-4Gchl-1Ceramide;

GM1, GalBl-BGalNAcB1-4(NeuAcaZ-3)GalBl-4Gch1-1Cerarnide;

GDla, NeuAc02-3GalBI-3GalNAcB1-4(NeuAca2-3)GalBl-4Gchl-lCeramide;

Glycosphingolipid“

Kd, kilodalton;

LDAO, lauryl dimethylamine oxide, also known as Ammonyx LO;

PK-C, protein ln'nase C;

PMSF, phenylmethane-sulfonyl fluoride;

SAT-1, CMP-sialic acid:lactosylceramide 02-3 sialyltransferase, also known as GM3
synthase;

SAT-2, CMP-sialic acideM3 02-8 sialyltransferase, GD3 synthase;

SAT-4, CMP-sialic acid:GM1 02-3 sialyltransferase, GD] a synthase;

SDS-PAGE, sodium dodecyl sulfate polyacrylaminde gel electrophoresis;

ST, sialyltransferase;

144

TLCK-HCI, (L-1-chloro-3-[4-tosylamido]-7-amino—2-heptanone-HC1
TPCK, L-1-chloro-3-[4-tosylamido]-4-phenyl-2-butanone

*All abbreviations for Glycosphingolipids are according to the
Svennerholm nomenclature (39) and the IU PAC - IU B recommendations (40).

ACKNOWLEDGEMENTS

We thank Dr. Tom Deits for his comments and assistance in reviewing and preparing

this manuscript.

We thank Carol Smith for her excellent secretarial support.

145

10.
11.
12.
13.

14.
15.
16.
17.

REFERENCES
Bond, J. S. and Bulter, RE. (1987) Ann. Rev. Biochem. 56, 333-364.
Usuki, S., Lyu, S.-C., and Sweeley, C.C. (1988) J. Biol. Chem. 263, 6847-
6853.
Usuki, S., Hoops, P., and Sweeley, C.C. (1988) J. Biol. Chem. 263, 10595-
10599.
Swamy, M.J. and Sweeley, C.C. (1989) Biochem. Biophys. Res. Commun.
162, 1188-1193.
Melkerson-Watson, L.J., Ogura, K., Wiesner, D., and Sweeley, C.C. (1990)
manuscript submitted.
Hanai, N., Nores, G.A., MacLeod, C., Torres-Mendez, C.-R. and Hakomori,
S., (1988) J. Biol. Chem. 263, 10915-10921.
Bremmer, E., Hakomori, S., Bowen-Pope, D.F., Raines, E., and Ross, R.
(1984) J. Biol. Chem. 259, 6818-6825.
Bremmer, E., Schlessinger, J., and Hakomori, S. (1986) J. Biol. Chem. 261,
2434-2440.
Hakomori, S. (1990) J. Biol. Chem., in press.
Usuki, S. and Sweeley, C.C. (1988) Ind. J. Biochem. Biophys. 25, 102-105.
Paulson, J. C. (1987) Trends Biochem. Sci. 14, 272-276.
Paulson, J. C. and Colley, K. J. (1989) J. Biol. Chem. 264, 17615-17618.
Morre, D.J., Merlin, L.M., and Keenan, T.W. (1969) Biochem. Biophys. Res.
Commun. 37, 813-819.
Morre, DJ. (1971) Methods Enzymol. 22, 130-148.
Fleischer, S., and Kervina, M. (1974) Methods Enzymol. 31, 6-41.
Carey, D.J., and Hirschberg, C.B. (1980) J. Biol. Chem. 255, 4348-4354.
Leelavathi, D.E., Estes, L.W., Feingold, D.S. and Lombardi, B. ( 1970)
Biochim. Biophys. Acta 211, 124-138.

146

18.
19.

20.

21.

22.

23.

24.

25.

26.

27.
28.

29.

30.

Melkerson-Watson. LL and Sweeley, C.C. (1990) JBC, in press.

Holmgren, J., Svennerholm, L., Elwing, H., Fredman, P. and Strannegard, O.
(1982) Proc. Natl. Acad. Sci. 77, 1947-1950.

Paulson, J.C., Beranek, W.B. and Hill, R.L. (1977) J. Biol. Chem. 252, 2356-
2362.

Sadler, J.E., Rearick, J.I., Paulson, J.C., and Hill, R.L. (1979) J. Biol. Chem.
254, 4434-4443.

Sadler, J.E., Rearick, LL, and Hill, R.L. (1979) J. Biol. Chem. 254, 5934-
5941.

Weinstein, J ., de Souza-e-Silva, U., and Paulson, J .C. (1982) J. Biol. Chem.
257, 13835-13844.

Joziasse, D.H., Bergh, M.L.E., ter Hart, H.G.J., Koppen, P.L., Hooghwinkel,
G.J.M. and Van den Eijnden, DH. (1985) J. Biol. Chem. 260, 4941-4951.

Gu, X.-B., Gu, T.-J. and Yu, R.K. (1990) Biochem. Biophys. Res. Commun.
166, 387-393.

Colley, K.J., Lee, E.U., Adler, B. Browne, J.K. and Paulson, J.C. (1989) J.
Biol. Chem. 264, 17619-17622.

Lammers, G. and Jamieson, J. C. (1988) Biochem. J. 256, 623-631.
Melkerson-Watson and Sweeley, C.C. (1990) Biochem Biophys. Res. Comm.
172, 165-171.

Hanada, K., Tamai, M., Adachi, T., Oguma, K., Kashiwagi, K., Ohmura, S.,
Kominami, E., Towatari, T., and Katunuma, N. (1983) in Proteinase
Inhibitors: Medical and Biological Aspects (Katunuma, N ., ed.) Japan Soc.
Press, Tokyo, p.p. 25-36.

Baubichon-Cortay, H., Broquet, P., George, P. and Louisot, P. (1989)
Glycoconjugate J. 6, 115-127.

147

31.

32.

33.

34.

35.

36.

37.

38.

39.
40.

Burczak, J.D., Fairley, J.L., and Sweeley, C.C. (1984) Biochim. Biophys.
Acta 804, 442-449.

Arribas, J. and Castano, J. G. (1990) J. Biol. Chem. 265, 13969-13973.
Johnson, K.F., Dawson, G. and Hancock, L.W. (JBC manuscript submitted 4-
10-89 entitled “Proteolytic Processing of Lysosomal Enzymes in Cultured
Human Fibroblasts”.

Pillai, S. and Zull J.E. (1985) J. Biol. Chem. 260, 8384-8389.

Kirschke, H., Langer, J., Wiederanders, B., Ansorge, S., Bohley, P. (1977)
Eur. J. Biochem. 74, 293-301

McKay, M. J., Marsh, M.W., Kirschke, H. and Bond, J.S. (1984) Biochim.
Biophys. Acta 784, 9-15

Melkerson-Watson, L.J., Ogura, K., Wiesner, DA. and Sweeley, C.C. (1990)
manuscript in preparation, see dissertation chapter 6.

Capasso, J.M., Keenan, T.W., Abeijon, C. and Hirshberg, C.B. (1989) J.
Biol. Chem. 264, 5283-5240.

Svennerholm, L. (1964) J. Lipid Res. 5, 145-155.

IUPAC-IUB Recommendations (1977) Lipids 12, 455-468.

148

CHAPTER 6
EVIDENCE FOR THE REGULATION OF GM3 GANGLIOSIDE
BIOSYNTHESIS IN KB CELLS BY PHOSPHORYLATION OF
CMP-SIALIC ACID:LACTOSYLCERAMIDE 012-3 SIALYLTRANSFERASE
(SAT-l)

A Model for the Regulation of Cell Growth Via GM3 Ganglioside Metabolism

LJ. Melkerson-Watson, K. Ogura, D.A. Wiesner and C.C. Sweeley

Dept. of Biochemistry
Michigan State University
E. Lansing, MI 48824-1319

October 29, 1990

Running Title: Studies on the Purification and Regulation of SAT-l

ABSTRACT

Treatment of KB cells with 4 mM butyrate synchronizes the cells in late
M/early Gl phases of the cell cycle. In monolayer, butyrate changes the morphology
of these cells, which display long processes. Turnover of CMP-sialic
acid:lactosylceramide 02-3 sialyltransferase (SAT-l), GM3 and GM3 Sialidase in KB
cells grown in monolayer culture appear to be cell-cycle dependent. The highest level
of SAT-1 activity coincides with maximum levels of the GM3 ganglioside. GM3 levels
are minimal when GM3 Sialidase activity is optimum. SAT-1 activity peaks 6-10 hr
after the release of the KB cells from a butyrate block. This increase of SAT-1 activity
occurs during early G1, about 4-6 hr before the cells begin DNA synthesis. In
monolayer, Sialidase activity was found to be at a maximum late G] phase just prior to
DNA synthesis and again throughouth phase.

Regulation of GM3 synthesis may involve the phosphorylation of KB cell SAT-
1. Analysis of immunoaffinity-purified KB cell SAT-1, during different time points of
expression within the G1 phase, by SDS-PAGE and its immunodetection on Western
blots with monoclonal specific for phosphotyrosine residues, indicates SAT-1 is a
phosphotyrosine-containing protein. The expression of this phosphotyrosine form may

regulate SAT-l activity.

150

INTRODUCTION

Involvement of glycosphingolipids as tissue-specific antigens in differentiation
and oncogenic transformation has been established ( 1,2). Several lines of evidence
suggest a functional role for gangliosides in cellular proliferation as well. Specifically,
exogenous addition of GM3 (3,4) and the lyso and de-N-acetyl derivatives of GM3
(5,6) have been shown to alter the mitogenic response to EGF-receptor kinase activity.
Synthesis and degradation of endogenous GM3 may be dependent upon cell cycle
activation of SAT-1 (7-10), altered expression of the acidic and neutral GM3 sialidase
activities (11,12), and recycling of the LacCer (13). If incorporation of exogenous
GM3 into the plasma membrane can regulate EGF-receptor function, a similar role for
endogenous GM3 may be expected (13).

We have proposed a model for the regulation of cell growth based on the
synthesis and degradation of endogenous GM3 (13) which accommodates our results
(7-14) and those of Hakomori and coworkers (3-6). According to our model, the
inhibitory effect of GM3 on the EGF-receptor kinase activity is abolished by a cell
cycle-dependent sialidase—catalyzed conversion of GM3 to LacCer (14). LacCer may
be internalized (l3) and recycled back to GM3 within the Golgi by cell-cycle dependent
activation of SAT-l (7-10,15), possibly by a phosphorylation/dephosphorylation
mechanism (15,16). GM3 expression (7) and SAT-1 activity (8-10,15) are maximal
late M/early G1 of the cell cycle. Apparently, the cyclic accumulation of GM3 at 01,
prior to DNA synthesis (9,15) may inhibit progression through the cell cycle (14).

KB cell growth can be modulated chemically with butyrate (7-10,15), phorbol
esters, and retinoic acid (17,18). These compounds dramatically and reversibly alter
the activity of SAT-1, the enzyme which is responsible for the synthesis of the GM3
ganglioside and representing the first committed step of a multi-step pathway for the
biosynthesis of more complex gangliosides. KB cells can be blocked chemically with

butyrate in late M/early G1, synchronizing the cells at a point (12-15 hrs) prior to

151

DNA synthesis (9). Butyrate treatment results in a 5-fold (9) up to a 10 to 15-fold (8)
increase in SAT-1 activity; however, it remains to be determined whether this effect
results in an increase of GM3 per cell or an increase of cells expressing GM3. We
report here the cell-cycle dependent expression of GM3, SAT-1 and sialidase for
human KB cells in monolayer cultures. Determination of the cell cycle dependent
expression of SAT-l, GM3 sialidase and the level of GM3 is necessary in
substantiating our model of cell growth (14) through the recycling of GM3 and the
regulation of GM3 synthesis by cell-cycle-dependent phosphorylation of SAT-l.

152

MATERIALS AND METHODS

Materials. Cytidine 5'-monophosphate sialic acid, (CMP-[14C4,5,6,7,3’9]-
sialic acid, 286.5 mCi/mmol) was purchased from New England Nuclear (Boston, MA)
and its specific activity adjusted to 22,200 dpm/nmol with unlabeled CMP-sialic acid
from Sigma (St. Louis, MO). The sphinganine-containing form of lactosylceramide
and bovine brain gangliosides, type II, were purchased from Sigma Chemical Co. (St.
Louis, MO) and used in enzyme activity assays. Ammonyx LO (lauryldimethylamine
oxide, LDAO) was obtained from the Stepan Company (Northfield, IL). Sodium
cacodylate and thymidine were purchased from Sigma Chemical Company (St. Louis,
MO). Eagle's modified minimal medium (EMEM), EMEM for suspension culture
(EMEM-S), antibiotic/antimycotic solution composed of penicillin, streptomycin
sulfate, and fungizone, HEPES, fetal bovine serum, non-essential amino acid solution,
and L-glutamine were purchased from GIBCO (BRL) Laboratories (Grand Island, NY).
[3H]-Thymidine (thymidine [methyl-3H], specific activity 60-90 Ci/mmol) was
purchased from ICN Biomedicals, Inc. (Costa Mesa, CA). Anti-phosphotyrosine
monoclonal antibody was obtained from Upstate Biotechnology, Inc. (Lake Placid,
NY). Goat anti-mouse IgG-alkaline phosphatase conjugate was from Boehringer
Mannheim (Indianapolis, IN). Prestained molecular weight standards and ultrapure
electrOphoresis reagents were bought from BioRad (Richmond, CA). All other
reagents were ultrapure or ACS grade from commercial sources.

Cells. Human epiderrnoid carcinoma strain KB cells (American Type Tissue
Culture, Rockville, MD, ATCC #CCL17 ) were cultured in monolayer on plastic
culture flasks (Corning, 150 cm2, #25120) or in suspension in glass spinner flasks
(Bellco Glass Co., Vineland, NJ) using Eagle's minimal essential medium (EMEM) (or
EMEM-S for suspension) supplemented with 10% fetal bovine serum (FBS)
(Gibco/BRL, Grand Island, NY), 100 ug/ml penicillin, 100 units/ml streptomycin,
fungizone (antibiotic/antimycotic solution from Gibco/BRL (Grand Island, NY), MEM

153

nonessential amino acid solution (Gibco/BRL, Grand Island, NY) (KB/EMEM
complete). The cells were grown in a humidified environment with 7.5 % C02 in air at
37°C. The stock cultures of KB cells were subcultured (1:8) every 2-3 days by
trypsinization with 0.25% trypsin/EDTA (Versene) solution for 1-2 min at 22°C. An
equal volume of KB/EMEM complete medium was added to inhibit the trypsin. The
seeding density for the KB cells for monolayer cultures was 6.6 x 104 cells per cm2
(with 92% survival following trypsinization). The passage numbers of the KB cells
used for these experiments were from P331-P394. All chemical treatments were
administered along with fresh medium at the time the KB cells were subcultured in the
following concentrations: butyrate, 4 mM; and thymidine, 2 mM. Drug treatments
were for 24 and 20 hr, respectively. For ganglioside and enzyme assays, KB cells
were harvested by trypsinization followed by three washes in PBS, pH 7.2, (calcium
and magnesium free) and stored at -70°C.

Methods

Cell Synchronization

Butyrate Block of KB Cells. KB cells (maintained as described above) were
treated with fresh complete medium containing a final concentration of 4 mM butyrate
(sodium butyrate, Sigma Chemical Co., St. Louis, MO lot #74F5063 or butyric acid
(lot #105F-0613) neutralized with NaOH) for 24 hr as previously described by
Chatterjee et al. (7), Macher et al., (8) and Moskal et al. (9). Cells were released
from butyrate by washing the cells three times with Hanks buffered saline (HBSS),
centrifuging the cells at 500 x g for 10 min following each wash, and incubating with
KB/EMEM complete.

Double-Thymidine Block of KB Cells. Synchronization of monolayer cultures
of KB cells by a double thymidine block procedure (7,9) was performed as follows.
KB cells at 6.6 x 104 cells/cm2 were incubated in KB/EMEM complete for 20 hr in the
presence of 2 mM thymidine (Sigma Chemical Co., St Louis, MO). The first-round

154

thymidine-treated cells were removed from the excess thymidine-containing media by
trypsinization followed by centrifugation. These cells were resuspended in the
appropriate fresh medium and cultured for 12 hr. Again, the cells were harvested and
resuspended in fresh medium containing 2 mM thymidine for 20 hr. Following the
second treatment with thymidine, the cells were harvested and the cell pellets washed
three times with PBS. Viable cell counts in triplicate were made with trypan blue
(Gibco/BRL, Grand Island, NY) and a hemocytometer.

Determination of DNA Synthesis.

KB cells were swded at a concentration of 6.6 x 104 cells/cm2 in monolayer
culture. The cells were synchronized with 4 mM butyrate for 24 hr, after which
butyrate was removed (as described above) and the cells were resuspended in
KB/EMEM complete. Viable cells were counted in triplicate on a hemocytometer with
trypan blue at two hour intervals following the release of the cells from the butyrate
block. Each sample (a 150 cm2 tissue flask) was washed three times with 20 ml of
HBSS and incubated 1 hr at 37°C in HBSS containing 1 mCi/ml [3H]-thymidine (ICN,
lot # 4498167). The [3H]-thymidine was removed. The cells were washed three times
with phosphate-buffered saline (PBS), treated with 10 ml ice-cold 5 % TCA at room
temperature for 30 min, and then washed twice with 10 ml 5% TCA. The cells were
treated overnight with 5 ml 1 N NaOH at room temperature. The amount of [3H]-
thyrnidine incorporated into DNA was determined on each time point in triplicate. 5.0
ml of scintillant (Safety-Solve, Research products, Inc. Mount Prospect, IL) was added
to each 50 Ill aliquot for counting in a Packard Model #640C Liquid Scintillation
Counter.

Preparation of Cell Membranes

Cell lysates were prepared by equilibrating the cells (1 x lo6 cells/ml) in 1 mM

TRIS-HCl (pH 7.2) in a 100% N2 atmosphere at 40 psi for 5 min, then lysing the cells

by returning them to normal atmospheric conditions (N2 bomb). Subcellular

155

membranes were prepared by differential centrifugation. The cell lysate (5-10 ml) was
first centrifuged at 1000 x g for 10 min. The postnuclear supernatant was centrifuged
at 10,000 x g for 20 min. The postrnitochondrial supernatant was centrifuged at
100,000 x g for 1 hr. The cellular membrane pellet was resuspended in 0.5 ml of 1
mM TRIS-HCI (pH 7.2) and assayed for SAT-l or GM3 sialidase activity and
membrane protein.

Purification of CMP-sialic acid:lactosylceramide 02-3 Sialyltransferase.

CMP-sialic acid:lactosylceramide 02-3 sialyltransferase (SAT-1) was purified
from KB cells by immunoaffinity chromatography using M12GC7 anti-SAT-l
monoclonal antibody coupled to Affi Gel 10 as previously described for the purification
of rat hepatic SAT-l (19). KB cells were cultured as described above and synchronized
with either a butyrate or double-thymidine block. Control cultures were asynchronous
cultures. KB cells were harvested and membranes were prepared in 1 mM TRIS-HCl
(pH 7.2) containing both protease inhibitors (pepstatin (0.5 rig/m1), leupeptin (0.7
rig/ml), aprotinin (0.25 ug/ml)) and phosphatase inhibitors (0.2 mM orthovandate, 8
mM sodium pyrophosphate, 1 mM 0-DL-glycerol phosphate, 4 mM sodium
molybdate, 6 mM sodium fluoride and 5 mM dithiothreitol).

Assays

Protein Concentration. The amount of membrane protein associated with each
fraction was determined by the Bradford method (20) relative to a bovine serum
albumin (BSA) standard.

SAT-I Activity Assay. SAT-1 assays, with radiolabeled CMP-sialic acid as the
donor substrate and lactosylceramide as the acceptor substrate, were carried out by a
modification of a previously described procedure (21). Assay mixtures contained the
following components in 100 Ill: lactosylceramide, 0.1 pmoles; CMP-
[14C4,5,6,7,3,9]-sialic acid (22,200 dpm/nmole), 0.04 pmoles; sodium cacodylate
buffer (pH 6.5), 10 mmoles (100 mM final concentration); manganese chloride, 1

156

mmole (10 mM final concentration); 5.0% (v/v) LDAO and 0.00501 mg of the
protein. GM3 reaction product was recovered by reverse phase chromatography on
Sep Pak C13 cartridges as previously described (22).

GM3 Sialidase Activity Assay. GM3 sialidase activity present at each time
point was quantitated by a colorimetric assay detecting free sialic acid. The assay
mixtures contained 0.25 mg of bovine brain gangliosides, type II (Sigma Chemical Co.
St. Louis, MO), 100 111 of the cellular membrane fraction and 100 111 of sialidase assay
buffer (0.2 M sodium acetate (pH 4.6) containing 1% (w/v) Triton CF-54. Following
incubation of the ganglioside with the sialidase for 3 hr at 37°C, 100 111 of 25 mM
periodic acid in 0.125 N H2804 was added. After 30 min at 37°C, 1.0 m1 of TBA
reagent was added and the solution incubated at 96°C for 7.5 min. The tubes were
cooled and 2.0 ml acetone/HCI (40:1, v/v) was added. The absorbance at 551 nm was
read and the amount of sialic acid released was quantitated relative to a sialic acid
standard.

Ganglioside Analysis.

Extraction of Gangliosides. Gangliosides were extracted by a modification of
the method of Ladisch and Gillard (24). Briefly, the frozen cell pellets were placed in
a 10 ml conical centrifuge tube and 5 ml methanol was added. The samples were
sonieated and vortexed to break up the cell pellet, then 10 ml chloroform was added
and the samples were again sonicated and vortexed. The extraction was carried out for
6 hr at 5°C. The mixtures were then centrifuged at 2000 x g for 15 min and the
supernatants removed. The pellets were resuspended in chloroform-methanol (1:1,
v/v) and the extraction was again carried out for 6 hr. These mixtures were centrifuged
and the supernatants combined and reduced to one quarter volume in 8 ml conical
centrifuge tubes and kept at -20°C overnight. After centrifugation at 2000 x g for 15
min, the supernatants were removed and taken to dryness under N2 and any remaining

solvent removed by lyophilization.

157

The total lipid extracts were taken up in 3 ml of diisopropyl ether/butanol (6:4,
v/v) and dispersed by several minutes of sonication and vortexing. 1.5 m1 of 100 mM
NaCl was added and the solutions were sonicated and vortexed for approximately 2 min
then centrifuged at 500 x g for 10 min. The upper organic phases were removed and
the lower ganglioside-containing aqueous phases were re-extracted with 3 ml of fresh
organic solvent. After sonication and vortexing and centrifugation, the organic phases
were removed and the aqueous phases lyophilized overnight.

3 To remove salts and other low molecular contaminants, the samples were run
over a 10 ml Sephadex G-50 column. Voided compounds were collected and solvent
removed by lyophilization.. The samples were taken up in 200 111 chloroform-methanol
(1:1, v/v) and centrifuged to remove residual insoluble impurities.

Ganglioside Quantitation. The gangliosides, extracted from the KB cells at
each time interval following the release of the cells from butyrate, were applied in total
onto a pre-run, pre-activated HPTLC plate in 5 mm bands. Known concentrations of
GM2 (by resorcinol assay (25) for lipid bound sialic acid, LBSA) were simultaneously
chromatographed for quantitation of the gangliosides. The gangliosides were
chromatographed on the HPTLC plate in a chloroform/methanol/0.2% CaC12°2H20
(60:40:9, v/v/v) solvent system. Gangliosides were visualized by spraying with
resorcinol-HG] reagent (25) and heating at 110°C for 12 min.

The plate was immediately scanned using a BioImage Visage 110 digital imager
in the transmissive mode at an integration time of 126. Gangliosides were quantified
by comparison with the 6M2 standard curve, where one fg of LBSA/cell equals

integrated intensity/cell x 11.

158

RESULTS
Determination of SAT-1 Expression in KB Cells

Turnover of GM 3 Ganglioside, SAT-1 and Sialidase Following Release of
Butyrate Treatment. Under the conditions employed, KB cell SAT-l was expressed
maximally 6-10 hr following the release of the cells from the butyrate block.
Monolayer cultures of KB cells contained maximal levels of GM3 immediately upon
release from butyrate decreasing to a minimum during early 61 (8 hr after release of
butyrate block) (Figure 1a). GM3 increased to maximum levels mid to late G1 (8-12
hr after release) coinciding with increased SAT-l activity (Figure 1a and 1b). The
amount of GM3 decreased upon the activation of sialidase prior to S (Figure 1a and
1c). The sialidase activity was broad and showed two maxima. The first peak in
sialidase activity was observed at 12 hr coincident with decreased GM3 and just prior
to the cells entering S phase (1c). The second peak was observed 16-22 hr after
removal of butyrate and coincided with the appearance of more complex gangliosides
(data not shown).

Additionally, GM2 levels peaked with S phase and decreased during M phase
and throughout the remainder of S phase there was a significant increase in 6M2
ganglioside as well as a proportional increase in the higher order gangliosides (data not
shown). Thus, at the point in the cell cycle when monolayer KB cells commit to
another round of replication or become quiescent, GM3 is catabolized by GM3
sialidase or GM3 is metabolized to GM2. Either pathway relieves the constraint of the
cell for another round of replication.

Determination of DNA Synthesis. KB cells treated with 4 mM butyrate for 24
hr followed by growth in fresh KB/EMEM complete culture medium in the absence of
butyrate, showed an increase in [3H]-thymidine incorporation 12-14 hr after removal of
the butyrate (Figure 1e). The peak was broad yet symmetrical, indicating that the
butyrate treatment of these cultures synchronized the KB cells under culturing

159

Figure 1. KB Monolayer Culture Metabolism of GM3 Ganglioside.

Human epidermoid carcinoma cells (KB cells) were grown in EMEM
(Gibco/BRL, Grand Island NY) supplemented with 10% fetal bovine serum, non-
essential amino acids, and Gibco's antibiotic/antimycotic solution at 37°C and 7.5%
C02. KB cells were swded at 6.6 x 104 cells/cm2 in 150 cm2 tissue culture flasks
(Corning) containing 25 ml of fresh KB/EMEM complete medium. After establishing
the cells for 24 hr, the cells were treated with 4 mM sodium butyrate in fresh
KB/EMEM complete medium (pH 7.4) for 24 hr. The cells were released from the
butyrate block by trypsinizing the cells and centrifuging at 500 x g for 10 min, washing
three times with PBS (pH 7.2). The cells were cultured in fresh KB/EMEM complete
medium. At each time interval indicated, the cells were harvested by trypsinization and
centrifuging at 500 x g for 10 min, washing three times with PBS (pH 7.2). The viable
cell density was determined by counting a 1:2 dilution of KB cells in trypan blue vital
stain. The cell pellets were frozen at -70°C until assayed for. either GM3 content by
the Ladisch procedure (24), or the microsomes assayed for SAT-1 and GM3 sialidase
activities (as described in Materials and Methods). Determination of DNA synthesis by
incorporation of [3H]-thymidine in KB cells grown in monolayer culture after release
. from butyrate is described in detail under methods. The data presented in this figure
are: (A) GM3 (nmol/cell); (B) Specific activity of SAT-1 (pmol/min/mg); (C) Specific
activity of GM3 sialidase (nmol/min/mg); (D) Cell density (cells/cm2 x 104); (E)
[3H]-thymidine incorporation (dpm/cell) and number of cells per sample ( x 107).

160

 

 

 

 

 

 

 

 

 

 

6.00 -

4.00 a

2.00 4

 

 

 

0.00

 

1.2!)
100i

J
035-
0301

0.25 '4

 

 

0.00

161

 

26

 

 

 

 

conditions employed.

Morphological Changes Induced by Butyrate and 171 ymidine Blacks in
Monolayer Cultures of KB Cells. The morphology of KB cells treated with butyrate
exhibit a drastic change in their morphology (8,9). As illustrated in Figure 2, KB cells
change from round or polygonal shape (Figure 2a and 2b) to one that is more
fibroblastic with long membranous processes (Figure 2c and 2d). This alteration of
morphology is reversible as can be seen in photos taken of KB cells 6 hr following the
removal of butyrate from the culture medium (Figure 2e and 20. These observation
correlate with previous reports from this laboratory (8,9). Treat of monolayer cultures
of KB cells with a double thymidine block also cause distinct morphological changes.
The cells, just prior to S phase, become more dense and the nuclei and nucleosomes
very pronounced (Figure 2g and 2h). It is expected that these morphological changes
of KB cells are associated with changes of the complex carbohydrates on the cell
surface and the enzymes responsible for their synthesis. For example, GM3 levels on
the cell surface correlate well with the expression of SAT-l activity.

Regulation of SAT-l Activity in KB Cells by Phosphorylation.

SAT-1 Activity in Synchronized KB Cells. To investigate phosphorylation as a
mode of regulating SAT-l, butyrate and thymidine were used to block KB cells at two
different points in the G1 phase of cell cycle. Butyrate treatment blocks KB cells just
prior to maximal activity and thynridine blocks the cells when SAT-1 is inactive. The
distribution of SAT-1 activity in KB cells treated with 4 mM butyrate, 6 hr after
butyrate release, and 2 mM thymidine (double block) is illustrated in Figure 3.
Cellular fractions of KB cells blocked 24 hr with butyrate gave SAT-1 activity levels
slightly higher than control asynchronous cultures. Six hours following removal of
butyrate SAT-1 activity was about the same (120% of control) as that of control

cultures. KB cell SAT-1 activity from cells 6 hr after release from butyrate was three

162

Figure 2. Morphological Changes Induced by Butyrate and Thymidine Blocked
Monolayer Cultures of KB Cells. KB Cells were grown in KB/EMEM complete
medium without butyrate (A and B), with butyrate (C and D), six hours after the
release of butyrate (E and F) and with thymidine (G and H) as described under
Materials and Methods at a seeding density of 6.6 x 104 cells/cm2 into 150 cm2 tissue
culture flasks. Panels A, C, E and G are 10X magnification of the cells and panels B,
D, F and H show the cells at 20X.

163

 

164

times control (279%) KB cell SAT-l from double-thymidine blocked cells showed
diminished levels of activity (21% of control).

Immunoqffinity Purification of KB Cell SAT-1. Recently, we reported the
purification of SAT-1 from rat liver Golgi to apparent homogeneity by immunoaffinity
chromatography using a monoclonal antibody to rat liver SAT-l (19). This
monoclonal, M12GC7, specifically inhibited and could immunoprecipitate rat hepatic
SAT-1 activity. M126C7 can also immunodetect and immunoprecipitate KB cell SAT-
1 (Figure 4). The multiple bands resolved were likely the result of proteolytic
degradation of the sialyltransferase. Proteolytic activity and degradation has been
reported associated with other glycosyltransferases (26-29) and has been postulated to
play a mechanistic role in the release of glycosyltransferases from the Golgi (28).

Immunodetection of SAT-1 from KB Cells with Anti-Phosphotyrosine
Monoclonal. SAT-1 from KB cells treated with 4 mM butyrate, 6 hr after butyrate
treatment, or 2 mM double thymidine and control cells (asynchronous culture) was
immunoaffinity-purified from detergent-extracted Golgi-enriched membranes as
described previously for the rat liver enzyme (19). An antiphosphotyrosine monoclonal
antibody detected a pattern of phosphotyrosine-containing polypeptides unique to each
sample of immunoaffinity-purified SAT-1 (Figure 5). This may represent a turnover of
the phosphorylated enzyme corresponding to different times of SAT-1 expression
during the GI phase.

SAT-1 from the control cells contained a phosphotyrosine in the band
corresponding to the 68 Kd mature form of SAT-1 with phosphotyrosine also being
detected at the 60, 54, 37, 33 and 25 Kd. Butyrate-synchronized cells contained
phosphotyrosine in the 68, 60, and 54 Kd polypeptides. Six hours following release of
the butyrate from the cells, phosphotyrosine was immunodetected at 68, 60, 47, 37,
35, 33, 31, 25 and 18 Kd. SAT-l immunoaffinity-purified from thymidine-blocked
KB cells contained phosphotyrosine predominantly in the 68 Kd and 25 Kd bands with

165

 

fl

Figure 3. SAT-l Activity Associated with KB Cells. KB cells were synchronized
with either butyrate or thymidine and their membranes recovered as described under the
methods. SAT-l activity was assayed as described under the experimental protocol.
Controls were cellular fractions from asynchronous cultures of KB cells. CL is the
specific activity associated with the cell lysate and M is the specific acitivity associated
with the Golgi-enriched membrane fraction prepared as described under the methods

section.

166

dpm (“IQ-Gm Formed / Cell (x toe-7)

 

l
X

l
\

i

J
A

 

 

CLM

“thaw

UCOV‘ ‘
NH"

0 6
mm w ma; tam-

0.0.0.0.! 0.0.0.0;
root- 00000-
01401 6600
...?" 0.0.0 ’0‘ .1
antral o o o o n o w
9000' or o
0 0 0 0 ~ 0 0 0 0 1 f
533V ’Ififi
cool 33%? $700
toto« root: 9060
too. too. or...
0000' 0000: root
1006- too. or. rooot
3333 5333'_ hth“ fVJfi'
oooo loco. root» 0000
339? ohhh‘ fafifii7 hhhhv
06061 4400* or A 00091
4460 tottv toot. oooo~
96441 0000 0000 totot
too 060 too. 0
0990.1 new, 50945 In; 0.0.00”
’V03 3333 hth 333V {’00
3’03 ”703$. %%%%' I’fifii- rihh‘I
to 3' too. 000 o o: 000' or
9’35 93%? 3%Vé' IVIJ‘V I?fifi" I333§7
too. I 0000‘ root 0449 00901

HOOP,» ‘ 0 I—. 0000'1300000 I—uOOOO! I. 0000 l

CL M CL M
KBCellularfiactlon

167

WW

'1'??? ’ . . . .
I
O t’
113°. m.
2 0600 I490.“ —

OLM

 

Figure 4. Immunoprecipitation of SAT-1 from KB Cells. KB cell SAT-l was
immunoprecipitated from asynchronous cultures of KB cells with MlZGC7 anti-SAT-l
specific monoclonal antibody bound to Staph aureus protein .A-conj ugated with rabbit
anti-mouse IgG as previously described (31). The Staph aureus - rabbit anti-mouse IgG
- M12GC7 - SAT-l conjugate was pelleted by centrifugation, washed, and the SAT-1
released and analyzed by SDS-PAGE on 12% polyacrylamide. The Staph aureus cells
were a gift from Dr. William Smith (Michigan State University, East Lansing, MI).
The closed circles represent the molecular weight of the SDS-PAGE standards (BioRad)
and the arrows indicate the molecular weights of the KB cell SAT-l polypeptides.

168

KB Cell Polypeptides Detected by anti-SAT—l IZGC7 MAb in Western
Blots of Total Cell Lysates

 

 
 
  
 
  

 

 

 

IO5 - Phosphorylase b _
2 BSA 2
: 68:; Ovalbumin :
60K 54K
P Carbonic anhydrase
_ 35K Soybean trypsin "
33K inhibitor
— 25K "
lxlOSKB Cells I7.5K Lysozyme
I04

169

Figure 5. Immunodetection of Purified KB Cell SAT-l by an Antiphosphotyrosine
Monoclonal Antibody. SAT-l was immunoaffinity purified with M 126C7-Affi Gel
10 from detergent-extracted Golgi—enriched membrane fraction (19) of KB cells that
were blocked in G1 phase with butyrate, 6 hours after release of butyrate, or blocked
pre-S phase by a double-thymidine block, and control (asynchronous cells without
chemical treatment). SAT-l (10 pg) from each treatment was electrophoresed and
transferred to Immobilon using 10 mM CAPS (pH 11.0) methanolic electrotransfer
buffer with the BioRad Mini-Protean II system. The blot was blocked with 2% (w/v)
gelatin in TBS (pH 7.4) for 1 hr at 37 °C and incubated with a monoclonal antibody
specific for antiphosphotyrosine (Upstate BioTechnology). Alkaline-phosphatase
secondary antibody with NBT/BCIP reagent served as the detection method. Lane A
shows phosphotyrosine-containing polypeptides of the asynchronous KB cells. lane B
shows the phosphotyrosine-containing SAT-l polypeptides from butyrate-synchronized
KB cells. Lane t6 are the phosphotyrosine-containing SAT-1 polypeptides from KB
cells six hours after they were released from butyrate. Lane T contains the SAT-1

polypeptides recovered from KB cells following a double-thymidine block.

170

Kd
97
66

43

31

21

14

Kd ,

97.." I ;~ 1—
66-"- -
43" -

 

 

   

171

by
[hi

minor bands detected at 60 and 47 Kd. These data may indicate that SAT-l is activated
by phosphorylation of Specific tyrosine residue(s) and that this phosphorylated form of
the enzyme may be proteolytically degraded as the KB cell progresses through the 61
phase.

Control blots, performed following the pre-incubation of the
antiphosphotyrosine monoclonal antibody with 50 mM soluble phosphotyrosine,
resulted in no detection of SAT-l with the antibody. Preincubation with 50 mM
phosphoserine or phosphothreonine had no effect on the binding of the monoclonal
(data not shown). These control experiments indicate this is a specific interaction
between the monoclonal and the purified SAT-1 and the results demonstrate the
presence of a phosphorylated form of SAT-1 corresponding to the active form of SAT-
1, namely 6 hr following the release of butyrate from the cells. Control cultures also
contained some detectable phosphotyrosine-containing SAT-1 at the 60,000 molecular

weight.

172

DISCUSSION

KB cells treated with butyrate become synchronized late M / early 61 of the
cell cycle. These data and observations correlated well with previous reports from this
laboratory (7-9). The morphology of these cells changes dramatically when in the
presence of butyrate. There is an alteration of their cell surface complex carbohydrates
and a change from polygonal-shaped cells to cells with a more fibroblastic appearance.
Butyrate-treated cells characteristically have long membranous processes.

The turnover of GM3 on the cell surface of KB cells was monitored as a
function of time following the release of KB cells from butyrate and as a function of
the enzymes responsible for its synthesis (CMP-sialic acid:lactosylceramide
sialyltransferase, SAT-l) and its degradation (GM3 sialidase). KB cells were chosen
because they are easily and consistently synchronized with either a butyrate block (7-9)
or a double-thymidine block (7,9). Previously, our laboratory reported an elevation is
the levels of GM3 and elevations in SAT-1 activity associated with treatment of these
cells with butyrate (7-9). While cell culture-dependent expression of gangliosides has
been demonstrated (7), neither the cell-cycle-dependent turnover of GM3, nor the
mechanism regulating GM3 ganglioside biosynthesis and degradation were known.

These results obtained for monolayer culture KB cells differ from those
previously reported for KB cells in monolayer. Previously, in monolayer cultures, the
SAT-l activity and GM3 levels of KB cells were highest immediately following release
from butyrate (8,9). The data obtained for optimal GM3 sialidase activity prior to S
phase and during G2 phase of the cell cycle agree with those reported for monolayers
of normal human fibroblasts ( 11,12). The observed difference for SAT-1 activity and
GM3 levels may reflect the nature of the microtubule disruption caused by
trypsinization and immediate treatment with butyrate.

While the regulation of glycosphingolipid metabolism and inhibition of cell

proliferation with increased levels of GM3 are largely correlational. GM3 has been

173

demonstrated to modulate the protein kinase C and EGF-receptor tyrosine kinase signal
transduction systems (3-6). We speculate that the regulation of GM3 synthesis may
involve phosphorylation of SAT-1. Previously, we reported increased SAT-1 activity
following incubation of crude homogenates with a cAMP—dependent protein kinase and
decreased activity following treatment with alkaline phosphatase (16).
Immunodetection of immunoaffinity-purified SAT-1 with an anti-phosphotyrosine
monoclonal indicates SAT-1 may contain one or more phosphotyrosine residues. We
would like to propose an extension of our model ( 14) for the regulation of cellular
proliferation by GM3 ganglioside through the regulation of its synthesis by a
phosphorylation mechanism as it potentially pertains to the EGF-R signal transduction
(Figure 6). The model shown in Figure 6 is only hypothetical at this time, but it does
accommodate the results reported by our laboratory (7-19) and those of Hakomori and
coworkers (3-6, for review 30). Efforts in our laboratory continues to address the
resolution of this cell growth regulatory mechanism involving the synthesis and

degradation of GM3 by SAT-1 and sialidase.

174

Figure 6. Pmposed Model of Cell Growth Regulation. A working model for the
regulation of cell growth based on the metabolism of GM3, SAT-l phosphorylation,
and EGF-receptor phosphorylation. EGF-R phosphorylation is controlled by the
presence of the GM3 ganglioside. GM3 is a negative effector of EGF-R
autophosphorylation (3,4,30) and inhibits cell growth. The level of cell surface GM3
and its turnover is regulated by extracellular sialidase (ll-14), de-acetylase (5) and by
SAT-1 (present study). SAT-1 (tyrosine) phosphorylation may be integrally related to

the EGF-R tyrosine kinase signal transduction system.

175

 

$283 “25025

 

 

 

 

 

03:: £320 A 03:3 5820
> a“5.03.2.3
9% 88.33 «so
«20 A Y .0003
+ wuflm
I +

 

 

03 a:
$283 ( .2502:
A 7 \

02032 00m 8.0002 00m
4

“how

Sam 3 920 .o «.8556 2.00822 05 a;
£320 :00 .0 80.5235. b08303. a ..8 .0005. 030020

176

FOOTNOTES
1To whom correspondence should be addressed.

leris work was supported in part by a research grant from the National Institutes of
Health (DK12434).

177

ACKNOWLEDGEMENITS

We thank Douglas Wiesner and Kiyoshi Ogura for their assistance with the quantitation

of the KB cell GM3 ganglioside and sialidase activity, respectively.

We thank Carol Smith for her excellent secretarial support.

178

ABBREVIATIONS

3The abbreviations used are:

CMP, cytidine 5'-monophosphate;

EGF, epidermal growth factor;

EMEM, Eagle's Minimal Essential Medium;

EMEM-S, Eagle's Minimal Essential Medium for Suspension Culture;

FBS, fetal bovine serum;

GD 1a, (NeuAC02-3)GalB 1-4GalNAcBl-4(NeuAC02-3)GalBl-4Glcll 1 - l Ceramide;

GD3, NeuAC02-8NeuAC02-3Galll1-4Gchl-1Ceramide;

GM] , GalB1-4GalNAcBl-4(NeuAC02-3)GalBl-4Glc81-lCeramide;

6M2, GalNAcBl-4(NeuAC02-3)Gallll-4Gchl-1Ceramide;

GM3, NeuAc02-3Ga131-4Gchl-lCeramide;

GT 1a: (NeuAC02-8NeuAC02-3)GalB l -4Ga1NAcB 1-4(NeuAC02-3)GalB l-4Gch 1 -
lCeramide;

Gle, (NeuAC02-3)GalB1-4GalNAcB1-4(NeuAC02-8NeuAC02-3)Galll1-4Gch 1-
lCeramide; »

Gangliosides and Glycosphingolipid"

HBSS, Hank Buffered Salts Solution;

Kd, kilodalton;

LDAO, lauryldimethylamine oxide, also known as Ammonyx LO;

SAT-1, CMP-sialic acid:lactosylceramide 02-3 sialyltransferase, also known as GM3
synthase;

SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis;

*All abbreviations for Glycosphingolipids are according to the Svennerholm

nomenclature (32) and IUPAC-IUB recommendations (33).

179

10.

11.
12.

13.

14.

REFERENCES
Hakomori, S. (1981) Ann. Rev. Biochem. 50, 733-764.
Ledeen, R.W. (1989) in Neurobiology of Glycoconjugates, (Margolis, R.K. and
Margolis, R.U., eds.) Plenum Press, NY, p.p.43-83.
Bremer, E.G., Hakomori, S., Bowen-Pope, D.F., Raines, E. and Ross, R.
(1984) J. Biol. Chem. 259, 6818-6825.
Bremer, E.G., Schlessinger, J. and Hakomori, S. (1986) J. Biol. Chem. 261,
2434-2440.
Hanai, N., Dohi, T., Nores, G.A. and Hakomori, S. (1988) J. Biol. Chem.
263, 6296-6301.
Hanai, N., Nores, G.A., MacLeod, C., Torres-Mendez, C.-R. and Hakomori
S. (1988) J. Biol. Chem. 263, 10915-10921.
Chatterjee, S., Sweeley, C.C. and Velicer, L.F. (1975) J. Biol. Chem. 250,
61-66.
Macher, B.A., Lockney, M., Moskal, J.R., Fung, Y.K. and Sweeley, C.C.
(1978) Exp. Cell Res. 117, 95-102.
Moskal, J.R., Lockney, M.W., Marvel, C.C., Mason, RA. and Sweeley, C.C.
(1980) in ACS Symposium Series, No. 128, Cell Surface Glycolipids (C.C.
Sweeley, ed.), p.p. 241-263.
Burczak, J.D., Fairley, LL. and Sweeley, C.C. (1984) Biochem. Biophys. Acta
804, 442-449.
Usuki, S., Lyu, S.-C., Sweeley, C.C. (1988) J. Biol. Chem. 263, 6847-6853.
Usuki, S., Hoops, P. and Sweeley, C.C. (1988) J. biol. Chem. 263, 10595-
10599,
Swamy, M.J. and Sweeley, C.C. (1989) Biochem. Biophys. Res. Commun.
162, 1188-1193.
Usuki, S. and Sweeley, C.C. (1988) Ind. J. Biochem. Biophys. 25, 102-105.

180

15.

16.

17.

18.

19.

20.
21.

22.

23.

24.

25 .

26.

27.

28.
29.

Melkerson-Watson, L.J., Ogura, K., Wiesner, D.A. and Sweeley, C.C. (1990),
submitted for publication.

Burczak, J.D., Soltysiak, R.M. and Sweeley, C.C. (1984) J. Lipid Res. 25,
1541-1547.

Burczak, J.D., Moskal, J.R., Trosko, J.E., Fairley, J.L. and Sweeley, C.C.
(1983) Exp. Cell Res. 147, 281-286.

Moskal, J.R., Lockney, M.W., Marvel, C.C., Trosko, J.E. and Sweeley, C.C.
(1987) Cancer Res. 47, 787-790.

Melkerson-Watson, LI. and Sweeley, C.C. (1990) "Purification to Apparent
Homogeneity by Immunoaffinity Chromatography and Partial Characterization
of the GM3 Ganglioside Forming Enzyme, CMP-Sialic acid:lactosylceramide
02-3 Sialyltransferase (SAT-1), From Rat Liver Golgi." submitted for
publication (Chapter 3).

Bradford, M.M. (1976) Anal. Biochem. 72, 248-254.

Burczak, J.D., Fairley, J.L., and Sweeley, C.C. (1984) Biochim. Biophys.
Acta 804, 442-449.

Melkerson-Watson, L.J., Kanemitsu, K. and Sweeley (1987) Glycoconjugate J.
4, 7-16.

Holmgren. J., Svennerholm, L., Elwing, H., Fredman, P. and Strannegard, O.
(19 ) Proc. Natl. Acad. Sci. 77, 1947-1950.

Ladisch, S. and Gillard, B. (1987) Methods Enzymol. 138, 300-306.

Ledeen, R.W. and Yu, R.K. (1982) Methods Enzymol. 83, 139-191.
Weinstein, J., de Souza-e-Silva, U., and Paulson, J.C. (1982) J. Biol. Chem.
257, 13835-13844.

Paulson, J.C. (1987) Trends Biochem. Sci. 14, 272-276.

Paulson, J .C. and Colley, KL (1989) J. Biol. Chem. 264, 17615-17618.
Lammers, G. and Jamieson, J.C. (1988) Biochem. J. 256, 623-631.

181

30.
31.
32.
33.

Hakomori, S. (1990) J. Biol. Chem. 265, 18713-18718.

DeWitt, BL. and Smith, W.L. (1982) Methods Enzymol. 86, 240-246.
Svennerholm, L. (1964) J. Lipid Res. 5, 145-155.

IUPAC-IUB Recommendations (1977) Lipids 12, 455-468.

182

CHAPTER 7
DISCUSSION

History and Perspectives

During the last five to ten years, several studies on glycosphingolipid function
have suggested several possibilities: (a) as modulators of membrane receptors and
transmembrane signal inducers, (b) as mediators of cell-cell recognition in development
and differentiation, and (c) as membrane receptors. The term ”bimodal” has been used
to describe the role of one particular glycosphingolipid, GM3 ganglioside (1) as it
primarily functions as a membrane modulator of membrane-associated protein kinases
and mediator of cell adhesion. This duality of function by GM3 at the cell surface
appears to be dependent on whether the cell is quiescent or in a proliferative state (for
review, 1,2).

The concept that GM3 ganglioside, as well as other gangliosides, regulate cell
growth was initially based on changes in glycosphingolipid synthesis associated with
oncogenic transformation, cell cycle and ”contact inhibition" of cell growth (for
review, 2). The idea that GM3 may influence cell proliferation was suggested by the
close association between the level of GM3 on the cell surface, the level of SAT-l
activity and cell growth (3-10; for review, 1).

The effect of gangliosides on cell growth modulation through. modified signal
transduction has been shown to occur in two of the three major transmembrane
signaling systems operating in most eukaryotic cells, growth-factor receptor-associated
protein kinases, i.e. , the EGF-R autophosphorylation (11,12) and protein kinase C
(13,14). Modulation of these two mechanisms by GM3 ganglioside and its naturally
occurring catabolites, lyso-GM3 (13, 14), de-N-acetyl-GM3 (15) and N,N-
dimethylsphingosine (16), was work accomplished through the efforts of Hakomori and

coworkers (for review, 1). They have demonstrated that exogenously added GM3

(11,12) and lyso GM3 (13,14) strongly inhibit EGF-R tyrosine autophosphorylation,
while de-N-acetyl GM3 (15) and N ,N-dimethylsphingosine (16) increase this EGF-R
kinase activity. Lyso-GM3 and N ,N-dimethylsphingosine inhibit protein kinase C
activity. A major criticism of these studies has been whether or not these observations
with exogenously added GM3, or its derivatives, is merely a pharmacological effect
( 17) or reflects an actual role of endogenous GM3 (9).

Recently, Weis and Davis (17) designed a series of experiments to test the
hypothesis that expression of GM3 at physiological levels modulates the signal
transduction of the EGF-R in mutant CHO cells which possess a reversible defect in
glycosylation. In the absence of galactose these cells cannot synthesize gangliosides
cannot perform terminal glycosylations of glycoprotein (an experimental problem).
Weis and Davis (17) demonstrated that the state of glycosylation did not affect the
ability of the EGF-R to bind EGF and function as a receptor. Using these cells, Weis
and Davis (17) demonstrated that decreased expression of gangliosides there was
increased EGF-R autophosphorylation and increased EGF-associated cellular
proliferation. The converse was also true. Thus, their findings for endogenously
synthesized ganglioside modulation of EGF-R were consistent with those reported for
exogenously added GM3 (11,12), adding support to the hypothesis that EGF-R function
is modulated by gangliosides. However, the level of in vitro autophosphorylation of
EGF-R isolated from the mutant CHO cell membranes was independent of galactose.
They concluded from these results that the inhibition of tyrosine autophosphorylation of
the EGF-R in these cells was not due to the direct interaction of the EGF-R with GM3.
They proposed alternative mechanisms for the observed inhibition of EGF-R tyrosine
kinase activity in these galactose-dependent cells: i.e. , other galactose-regulated
molecules may modulate EGF-R autophosphorylation; inhibition by GM3 on EGF-R
autophosphorylation may involve the state of threonine phosphorylation of the receptor

or the oligomeric state of the receptor, as previously suggested by Schlessinger (18).

184

This study by Weis and Davis (17) is in contrast to findings by Wedegaertner
and Gill (19), who examined the effect sphingosine on the activation of an
immunoaffinity-purified protein tyrosine kinase domain of the EGF-R. Their results
demonstrated that the addition of sphingosine induced a conformational change that
mimics the effect of the conformational change of the holo-receptor when EGF binds,
inducing a fully active tyrosine kinase activity. Further, the GM3 ganglioside has been
shown to co-precipitate with the EGF-R using EGF-R specific monoclonal antibody
(20). Clearly, the mechanism of the functional role played by GM3, and its
catabolites, on the autophosphorylation of the EGF-R, and concomitantly cell growth
regulation, is controversial and needs to be defined.

Work in our laboratory on the modulation of cell proliferation by GM3 focuses
on the two enzymes responsible for the synthesis and degradation of GM3. Our model
(9) suggests that when cells are in a proliferative state, the inhibition of cell
proliferation by GM3 is relieved by the dissociation of GM3 from the EGF-R through
the catabolism of GM3 by GM3 sialidase. Loss of the sialic acid from GM3 gives
lactosylceramide (LacCer) and relieves the constraint on the EGF-R, permitting
subsequent phosphorylation reactions and continued cell growth. We speculate that
LacCer, or its catabolites (glycosylceramide, ceramide, sphingosine) is internalized by
the cell to the endoplasmic reticulum or Golgi (10), where it (or they) serve as
substrates in resynthesizing GM3. SAT-1 catalyzes the addition of sialic acid to
LacCer forming GM3 ganglioside.

A logical point where SAT-1 may be regulated is post-translational
modification. Early evidence by Dawson and coworkers (21) suggested a
phosphorylation/dephosphorylation mechanism in the regulation of GM2 synthase.
Supportive are the findings of various phosphoprotein and protein kinases have been
found within the Golgi, as well as evidence for ATP translocation (22). Further,

Strous et al. (23) have identified a serine-linked phosphate on the cytoplasmic amino

185

terminus of a Golgi galactosyltransferase in HeLa and HepG2 cells. They speculated
that this posttranslational modification may be involved in membrane targeting of
Golgi-resident glycosyltransferases.

Considerations on the Regulation of SAT-l

Recently, Dumont et al. (24) speculated about the coexistence of cAMP-
dependent and independent mitogenic pathways. Two-dimensional electrophoretograms
of 32P-labeled proteins in dog thymocytes indicated that the proteins phosphorylated
due to the growth factor receptor stimulated tyrosine kinases or phorbol esters or other
activators of the phosphatidylinositol cascade, which in turn stimulated PK-C and
resulted in the phosphorylation of some common proteins. In contrast, cAMP acting
via protein kinase A resulted in a different phosphoprotein pattern, indicating that even
though transmembrane signaling systems are initially distinct they do seem to act in a
convergent manner to control the phosphorylation/dephosphorylation of proteins
involved during the progression of a cell through G1.

SAT-l activity is cell cycle-dependent (see Chapter 6). SAT-1 activity peaks
about 4-6 hours before DNA synthesis, during early 61 of the cell cycle. SAT-l
activity can be stimulated chemically by butyrate (3-5), retinoic acid and phorbol esters
(25). Phorbol esters are known modulators of phosphorylation acting through a DAG
mechanism on PKC. Other agents such as Prostaglandin E (21), enkephalins (21), the
B-subunit of cholera toxin (26), and GO lb gangliosides (27) are activators of cAMP-
dependent protein kinases (cAK), which also stimulate glycosyltransferase activity.
Further, stimulation of SAT-l activity correlates with increased CAMP-dependent
protein kinase and decreases in the presence of alkaline phosphatase (28).

I have proposed that the regulation of SAT-1 is integrally involved with
membrane signal transduction and have suggested a model whereby SAT-1 activity is
regulated by a bimodal system with protein kinase C and epidermal growth factor
receptor signal transduction systems (See Chapter 1 Figure 7). Much like the

186

"bimodal” function of GM3 (the SAT-l enzyme product) the proposed "bimodal"
regulation of SAT-1 is dependent on whether the cells are proliferating or quiescent and
that the site of phosphorylation is dependent on which signal transduction mechanism
operates to regulate the production of GM3. As a model for cell growth regulation
(Chapter 6 Figure 8), the receptor-mediated tyrosine kinase, operating possibly via a
kinase cascade, may act as a positive effector on SAT-1 through phosphorylation of
specific tyrosine residue(s). An antiphosphotyrosine antibody specifically blots purified
SAT-1 from rat liver (Figures 1 and 2) and from KB cells (Chapter 6). The
phosphorylation pattern of phospho-tyrosine(s) on KB cell SAT-1 differs among
samples taken during different times within the 61 phase of the cell cycle. Over the
last year the several studies on the effects of GM3 catabolites on PK-C and EGF-R
show an antagonist/protagonist effect (reviewed, 1). Thus, it seems plausible that
SAT-1 is similarly affected by PK-C and EGF-R mediated kinase activities. Clearly,
such a proposed requires detailed investigation into the protein sequence and isolation
of the gene transcribing SAT-1.

In vitro activation of PK-C and of SAT-1 can be stimulated by phorbol esters.
Phorbol esters presumably act as substitutes for diacylglycerol (DAG) (29). Hydrolysis
of membrane-bound enzymes, anchored by glycosyl-phosphatidylinositol (GPI), by
phospholipase C (GPI-PLC) releases DAG (30). SAT-1 is a membrane-associated
enzyme. The mechanism through which it is anchored in the Golgi is unknown. We
investigated the possibility that SAT-1 was GPI-anchored and postulated a potential
mechanism for PK-C activation of SAT-1 through its release from the Golgi by GPI-
PLC.

Intact and "leaky" rat liver Golgi vesicles were treated with GPI-PLC (a gift
from Dr. Martin Low) and the membrane pellets and supernatants were examined for
SAT-1 activity. These data are summarized in Table l. SAT-1 activity was found to

be associated with the supernatants in both the intact and permeabilized Golgi. This

187

Figure l. Immunodetection of SAT-1 with Antiphosphotyrosine Monoclonal
Antibody. Immunoaffinity-purified rat hepatic SAT-1 (10 pg) was analyzed by two
dimensional gel electrophoresis (2D SDS-PAGE), electrotransferred to Immobilon
(Millipore), and immunodetected with an antiphosphotyrosine monoclonal antibody
(UpState Biotechnology, Inc.). First dimension tube gels were performed according to
the method of O'Farrell (42) with an ampholyte mixture of 1 part pH 3-10 and 4 parts
pH 5-8. The second-dimension (2D) slab gels were standard 12% Laemmli (43) SDS-
PAGE. The 2D blot containing 10 ,ug SAT-l was immunodetected with
antiphosphotyrosine monoclonal (1:1000, 100g /ml), alkaline phosphatase-conjugated
secondary antibody (BMB, 1:6000) and BCIP/NBT reagent and the blot analyzed by
the BioImage Visage 110 computerized digital imager (BioImage/Millipore, Ann
Arbor, MI) to estimate the molecular weight and pl of the protein. SAT-l is indicated
by the arrow. Its apparent molecular weight was about 60,000 daltons and the pI in the
range of pH 6.2. The carbamylated IEF standards are (A) creatine phosphokinase,
M.W. 40 Kd and p1 range: pH 4.9-7.1, (B) glyceraldehyde-P dehydrogenase, M.W.
36 Kd and p1 range: pH 4.7-8.3, and (C) carbonic anhydrase, M.W. 30 Kd and p1
range: pH 4.8-6.7. The SDS-PAGE molecular weight standards are phosphorylase b
(97 Kd), bovine serum albumin (66 Kd), ovalbumin (43 Kd), carbonic anhydrase (31
Kd), soybean trypsin inhibitor (21 Kd) and lysozyme (14 Kd).

188

ms 5 ------------------------------------ 8
m -

97— ~
56 — w

43—

MW
31 —

21-\
M—\

 

 

6.7 7.1 8.3

pl

A=CPK
B = GADPH
C: CA

189

Figure 2. Specificity of the Immunodetection of SAT-1 with Antiphosphotyrosine
Monoclonal Antibody. Immunoaffinity-purified SAT-1 (10 ng) was analyzed
following SDS-PAGE (43) and Western blot with an antiphosphotyrosine monoclonal
antibody (UpState Biotechnology, Inc.). Unbound sites on the Western blots were
blocked with TRIS-buffered saline (pH 7.4) containing 2% (w/v) gelatin for 30 min at
37°C. Prior to immunodetection of SAT-1 with the antiphosphotyrosine monoclonal
antibody, the monoclonal antibody was preincubated with 50 mM phosphotyrosine (A)
and 50 mM phosphothreonine (B) or no pretreatment (C). The immunodetection was
carried out in TRIS-buffered saline (pH 7.4) containing 1% gelatin and 0.05% Tween
20 (ELISA grade). Specific detection of rat hepatic SAT-1 (lane 2) was demonstrated
at an apparent molecular weight of about 60 Kd (denoted with an arrow). The
molecular weight standards (lane 1) were phosphorylase b (97 Kd), bovine serum
albumin (66 Kd), ovalbumin (43 Kd), carbonic anhydrase (31 Kd), soybean tryspin
inhibitor (21 Kd) and lysozyme (14 Kd).

190

191

d SAT-1

Table l. Phosphatidyl Inositol as a Potential Membrane Anchor for SAT-l. Golgi
vesicles were prepared from rat liver (as described in Chapter 3) based on well-
established procedures. Like other glycosyltransferases, SAT-1 has a lumenal
topography. Assay of SAT-l activity requires the Golgi to be permeabilized with
detergent. The Golgi were made "leaky" with Triton CF-54 (final concentration,
0.3%) according to the method of Carey and Hirschberg (31). Intact and leaky Golgi
(1.4 mg) were incubated with S. aureus phosphatidylinositol phospholipase C (a gift
from Dr. Martin Low, Columbia University, New York, NY) at a final concentration
of 20 uglml for 2 hr at 37°C. Total volume was 2.0 ml. Golgi-enriched membranes
were recovered by centrifugation at 150,000 x g for 90 min at 4°C. The supernatants
and pellets were assayed for SAT-1 activity as previously described (32,33; see
Chapters 2 and 3). Verification of GPI-PLC activity was the hydrolysis of 1000 pg of '
PI resuspended in 25 mM sodium cacodylate (pH 7) and GPI—PLC (20 ug/ml), final
volume 1.0 ml. The reaction conditions were as described for treatment of the Golgi.
Aliquots from the control reactions for enzyme activity (25 111) were chromatographed
on HPTLC plates with a chloroform/method/HZO (65 :35:5 , v/v/v) solvent system or
on Clg-HPTLC in methanol/H20 (2:1, v/v). DAG standards, _n—l-
2,dioctanoylglycerol and _sn-1-oleoyl-2-acetylglycerol, were run in parallel. The lipids
were detected with 50% sulfuric acid. The experimental protocol was performed twice

and assayed in duplicate each time.

192

Table l

Phosphatidyl Inositol as a Potential
Membrane Anchor for SAT-1

 

Fraction GPI-PLC Specific Activity
(20 ng/ ml) SAT-1

(pmol-min'ng'l)

 

Intact Golgi
Supernatant + 10.0
Pellet + 2.7
Supernatant - 15.2
Pellet - 5.3
"My" Golgil
Supernatant + 4. 1
Pellet + 4.6
Supernatant - 14.6
Pellet - 7.0

 

l The term ”leaky“ refers to Golgi which were made permeable through the addition of Triton CF -54 to a
final concentration of 0.3 %.

193

may have resulted from the mechanical disruption and proteolytic degradation during
purification. This investigation on GPI as a potential membrane anchor for SAT-1 was
performed about one year prior to the detection of proteolytic degradation of SAT-1
and was performed without thiol protease inhibitors (see chapter 5). Further, GPI-PLC
has been demonstrated to release alkaline phosphatase (APase) and phosphodiesterase,
enzymes which are negative effectors of sialyltransferase activities (21,28). In
consideration of these factors, there is no direct evidence for or against GPI being
involved in SAT-1 anchoring within the Golgi. Therefore, to explore the relationship
between this novel lipid anchoring system and SAT-l, a repeat of the experiment is
nwded with appropriate protease and APase inhibitors.

Protein Sequencing of SAT-1

Studies employing antiphosphotyrosine monoclonal antibody are only suggestive
about the regulation of SAT-1 activity (see Chapter 6). However, this work adds to
other data implicating a role for phosphorylation in the regulation of SAT-1 (or other
glycosyltransferases). Identification of potential phosphorylation sites within the
protein sequence would support our findings. During the last year, many attempts have
been made to obtain protein sequence information. The approach has come full circle
and the end result to date, three residues: Ser-Tyr-Gly.

SAT-1 appears to be N-terrninally blocked. Therefore, SAT-1 was
electrophoresed and electrotransferred to Immobilon for enzymatic degradation with
trypsin. Digestion was incomplete, perhaps due to the high degree of glycosylation of
the enzyme (See Chapter 3), as evidenced by the broad peak in the Cg-HPLC profile
(Figure 3). This hydrophobic peak, estimated by A214nm to be 7-8 11g, gave no
sequence information. We also attempted cyanogen bromide (CNBr) and enzymatic
deglycosylation of SAT-1 with N-glycanase followed by trypsinization as alternative

approaches.

194

Figure 3. Reverse-Phase Cg-HPLC of SAT-l Tryptic Peptides. Reverse-phase
HPLC was performed on an Aquapore RP-300 C3 column (Applied Biosystems) (250 x
1.0 mm x 7 11m). Individual peptides were eluted from the microbore column on a 60
min linear gradient established between 0.1% trifluoroacetic acid and 90% acetonitrile.

Sample volume injected was 250 111. No sequence information was obtained for the

major peak.

195

0.00

 

0.32:2: m: C.

0.00 0.0V 0.00 0.0m 0.0..
_ . _ . _ I

5°73

 

a a
V 0.
9
a s a
a Qua—r
8
.88
SD 6
on v
0 <

 

m
0
m

0.0
_ 00.0

l «0.0.

I 00.0

r 00.0

 

 

Team “.0 20:30.0 2305:.

(WWW) SONVHHOSGV

196

Analysis of CNBr Digests. No internal amino acid sequence was obtained from
the major peaks of SAT-l (ethanol/ acetone precipitated) digestion with CNBr (Figure
4). The three major peaks were rechromatographed to resolve individual species. The
recovery of total protein was 10—20 % of starting material for each run. We expected
some loss during the CNBr digestion, but another probability was protein loss from
adsorption to the plastic collection tubes during Cg-HPLC. To minimize loss of SAT-l,
the sample was loaded directly onto the reverse-phase HPLC column along with
appropriate buffer controls containing LDAO. Comparison of the profiles gave two
unique peaks at the hydrophobic end of the run; the remainder of the peaks generated
were attributed to LDAO (Figures 5a and 5b). [These peaks were analyzed on SDS-
PAGE and verified to be the 60 Kd SAT-1 and a 56 Kd proteolytic fragment (See
Chapter 4). Both appeared to be N-terminally blocked. Comparison of these
chromatograms to the first CNBr digests indicated the peaks which generated "no
sequence” were in fact LDAO.

Two-dimensional SDS-PAGE was employed to recover SAT-1 without LDAO
contamination. LDAO protonates at pHs7. The pI of SAT-l was estimated in the
range of pH 5.7-6.2 by 2D-SDS-PAGE (See Chapter 3). Theoretically, under optimal
IEF conditions for SAT-1, LDAO should be charged and proceed through the tube gel.
SAT-l resolved by 2D SDS-PAGE (40 pg) was electrotransferred to Immobilon and
digested with CNBr (Figure 6). This has proved to be the most successful approach to
date. Amino acid sequencing of three peaks (A, B, E) has been tried but sequence
information could not be obtained from B because of a technical error. Peak A gave
the following internal sequence information: Ser-Tyr-Gly. No sequence information

was obtained from Peak E. The remaining peaks have not been tried.

197

Figure 4. Reverse-Phase Cs-HPLC Following CNBr Digestion of SAT-l. Reverse—
phase HPLC of SAT-l peptides generated following CNBr digestion was on a C3
microbore HPLC Aquapore RP-3OO column (Applied Biosystems) (250 mm x 1.0 mm
x 0.7 pm). Elution was preformed using a 90 min linear gradient established between

0.1% TFA and 90% acetonitrile. No sequence information was obtained from the

labeled peaks.

198

0.00

0.00

AmmSEEV wit.

00¢ 0.0m

0.0m

 

 

9'179

 

 

 

8'99

 

l'69

. F

 

 

6°81?

1729f

— -

 

 

9°62

 

 

 

 

9'92

 

 

0°92

_ . _ .

 

 

 

mmozhmm

.mzo

_...__.<m ...0 QJQItnE

00.0

50

No.0

«0.0

¢0.0

00.0

(“‘“V'Zw aoNveaosav

199

Figure 5. Reverse-Phase Cg-HPLC of SAT-l and Buffer Containing LDAO.
Resolution of purified SAT-l sialyltransferase from LDAO was achieved on a C3
microbore HPLC system (Applied Biosystems). SAT-l in 25 mM sodium cacodylate
(pH 6.5) containing 5% (v/v) LDAO was applied onto an Aquapore C3, RP-3OO (250 x
1.0 mm x 7 pm) reverse-phase microbore column and eluted on a 90 min linear
gradient established between 0. 1% (v/v) trifluoroacetic acid and 90% (v/v) acetonitrile.
Sample volume injected was 250 p1. Panel A illustrates a typical chromatogram
obtained from buffer containing LDAO. Panel B shows the resolution of unique SAT—
1 proteins from the same LDAO containing buffer. No sequence information was

obtained from the 56 or 60 Kd proteins.

200

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:2, 83, f3 REFERENCE
_ A. s a a
n 25mm sod. cacodylate (pH6.5)
with l%("/V)LDAO
0.02 -
E «‘13.
<_r :3
DJ
a) 5
8 i {3 “3
B , :5 °°
5‘2 0.0| — ,0
.0 I
<1 3 D)
U
0.00 A l l n l A l 1 1
40.0 50.0 60.0 70.0 80.0
Time (minutes)
- SAT-l
B. 19,
2} g 56Kd
as Proteolytic
(D Fragment
A 0.02 — .‘3
ID
2 g ‘9
3!. «s
91 . .' '08 ¢ 60Kd
a) ‘ trim Mattie
g 8 j. 10:; ?) Form
0 “5 ' tn " co
.0 V .' 'Q
a :2 : «o
B 0.0! "' 05 N
<r ¢ L
r r
000 1 I l . J A l . I
40.0 50.0 60.0 70.0 80.0

Time (minutes)

201

 

 

Figure 6. Reverse-Phase Cs-HPLC of SAT-l Following 2D SDS-PAGE and CNBr
Digestion. Reverse—phase HPLC of SAT-1 peptides generated following CNBr
digestion was on a C3 microbore HPLC Aquapore RP-BOO column (Applied
Biosystems) (250 mm x 1.0 mm x 0.7 um). Elution was preformed using a 90 min
linear gradient established between 0.1% TFA and 90% acetonitrile. No sequence
information could be obtained from B because of technical error. No sequence
information was obtained from E. Sequence obtained from peak A was Ser-Tyr-Gly.

The remaining peaks have not been tried.

202

0.00

 

(141-.

33:55 25.

10.0 0.0.“. 0.00 0.00 0.0a. 00» 0.0a

0.0. 0

 

01561. I

—n-.prrb
. . a l— - — u d —
:1
_ . ... .. a .. .
_ ___

 

 

was u.
*0

 

 

9

 

1299 m

 

98 '09 I”

.jll]

. 00.0

0.0

0N0

0nd

0V0

 

 

5% no zofimmoa .mzo

(mafia) aouquosqv

203

Analysis of Tryptic Digests. Deglycosylation of SAT-l with N -glycanase was
accomplished as described by the manufacturer (Genzyme). The deglycosylated SAT-1
was recovered by SDS-PAGE. The tyrpsinization was performed directly on SAT-1
within the gel according to the method of Matsudaira (34) and the resulting peptides
were resolved on by Cg-HPLC. The profile is shown in Figure 7. The chromatogram
from the deglycosylated enzyme gave better resolution than that generated from the
first tryptic digest (Figure 3). A series of extremely hydrophobic residues, eluting with
90% acetonitrile, were recovered and may correspond to the transmembrane domain.
None of these peaks have been analyzed for amino acid sequence.

Thus, after one year of obtaining no sequence information from these
conventional approaches, we have sent 4 nmol of affmity—purified rat liver SAT-1 to
Dr. Donald Hunt (University of Virginia, Charlottesville, VA) to have SAT-1
sequenced by protein mass-spectrometry.

Cloning a cDNA Encoding SAT-1

Rabbit anti-rat hepatic SAT-1 polyclonal antibodies have been produced (see
discussion below) and protein A-purified. Immunodetection of SAT-1 on Western
blots by these antibodies is sensitive in the 0.5 ng range. These antibodies have been
used in a primary screen of a human hepatoma cDNA expression library containing
about 1 x 106 clones. It has been reported that there is sufficient homology within the
catalytic domain between species to permit antibody (and oligonucleotide) probes to
detect glycosyltransferases in more that one species (35,36). About 30-40 putative
positives were detected on each of 5 x 150 mm blots (Figure 8). The positives need to
be plaque-purified, the phage isolates cultured and rescreened with antibody. While
false positive exist, these results indicate that sequence information of SAT-1 may be
obtained through the cloning of this cDNA expression library. Rescreening should be
done with polyclonal monospecific anti-SAT-l (deglycosylated form) antibodies.
These can be prepared by affinity purification using deglycosylated SAT-1 coupled to

204

figure 7. Reverse-Phase Cg-HPLC of Tryptic Peptides from Deglycosylated SAT-
1. Reverse-phase HPLC was performed on an Aquapore RP-3OO C8 column (Applied
Biosystems) (250 x 1.0 mm x 7 pm). Individual peptides were eluted from the
microbore column on a 60 min linear gradient established between 0.1% trifluoroacetic
acid and 90% acetonitrile. Sample volume injected was 250 #1- To date, none of these

peaks have been sequenced.

205

(2l4nm)

ABSORBANCE

0.35

0.28

0.2 I

0.I4

0.07

000

RP-HPLC of SAT-I Digested with N-glyconose 8 Trypsin

 

 

A

:- 2l.08 2

 

 

 

?- 3.sz

 

 

 

 

 

 

 

 

 

 

 

 

Tm
I
9 e
I. :i
1 I l
I0.0 20.0 30.0
TIME (minutes)
r cl ‘3:
i .3 E - 0.35
: 3 Fe
i m 9" ”53 I
E 3 3° 3 g - 0.28
I 3 V5 "
I a: o
I O
I 8
< I —
! 700 0.2I
TIME (min)
- 0J4
"‘ 0.07
.. /-’-‘W 000
I 1 ‘

 

4 0.0

206

TIME (minutes)

Figure 8. Primary Screen of a cDNA Expression Library with Rabbit Anti-SAT-l
Polyclonal Antibodies. About 1 x 106 clones from a human hepatoma cDNA
expression library were plated onto five 150 mm plates (A - E). The library filters
were screened with rabbit anti-SAT-l IgG polyclonal antibodies and detected by the
alkaline-phosphatase conjugated secondary antibody system. Approximately 30-40

putative positives were detected per plate.

207

 

Affi-Gel 10. About 500 pg of SAT-l has been deglycosylated with N- and O-
glycanase for this purpose.
Intracellular Localization of SAT-l

SAT-1 resides within the Golgi. SAT-1 activity and that of thiamine
pyrophosphatase, a Golgi membrane-marker enzyme, were recovered together in our
Golgi—enriched membrane fraction following differentiation centrifugation and
discontinuous sucrose gradient. SAT—1 distributed through the 10-32% interfaces and
most concentrated at 25-32% (see Chapter 2). Using a similar centrifugation protocol,
Trinchera and Ghindoni (37) localized SAT-1 activity within the cis Golgi using an
immunolocalized glycosyltransferase as an enzyme marker. The addition of sialic acid
is considered a terminal glycosylation and as such, SAT-1 would be localized within
the trans Golgi where other sialyltransferases have been immunolocalized (38).
However, GM3 also serves as percursor for other gangliosides and presuming a
precursor-product relationship which during glycolipid synthesis transverse would lend
support to the presence of SAT-l in the denser fractions of the Golgi stack. The best
method for identifying the location of SAT-1 will be immunolocalized with a rabbit
anti-SAT-l polyclonal antibody specific for the deglycosylated form of SAT-1.

I have enzymatically digested SAT-l with N- and O-glycanases (Genzyme) in
preparation for the immunolocalization of SAT-1 in rat liver thin sections.
Administration of this as an antigen was delayed pending an necropsy report on a rabbit
used for preparing anti-SAT-l polyclonal antibodies. The rabbit died as the result of a
systemic lymphosarcoma. The report from the veterinarian appears in Appendix A.

Speculation as to the cause of our rabbit's demise is intriguing. The potential
exists that this disease state may have resulted from administering SAT-1 to the animal.
Serum is rich in shed gangliosides (39,40) has been shown suppress the immune

response during tumorgenesis. LacCer, the acceptor substrate for SAT-1, is a major

209

component of the hemopoetic system. While it is generally presumed that antigen in
adjuvant is inactive. The possibility that SAT-1 is active exists. RIBI ImmunoChem
(Hamilton, MT) MPLTM + TDM (0.5 mg trehalose dimycolate (TDM) and 0.5 mg
monophosphoryl lipid A (MPL) in 2 % oil-Tween 80-H20; lot #060— 124) , which
employs a detergent-lipid emulsion, was used as the adjuvant. It is unknown if SAT-1
(at 50-200 ug/ m1) would be active in such an adjuvant system. SAT-1 exhibits
differential activity depending on the detergent used to solubilize the Golgi (see Chapter
4).

Another possible consideration was an autoimmune response.
Glycosyltransferases share greater than 85% sequence homology between species (35).
The rabbit elucidated a good immune response to the rat liver SAT-1. After the third
inoculation, her condition began to quickly decline after 7 days.

A third possibility is that the antigen may have contained some residual
lauryldimethylamine oxide (LDAO). LDAO can be toxic and HPLC analyses from our
protein work indicate that while ethanol and acetone precipitation minimize the amount
of detergent in a sample, some LDAO still remains (see Chapter 4). In raising
monoclonal antibodies to SAT-l in mice, it was necessary to do a
phosphatidylcholinezLDAO exchange (see Chapter 3). This approach was also
employed in raising the polyclonals.

To define a causal relationship, i.e. , active SAT-l, autoimmune response,
LDAO toxicity, or an inherent defect of the animals health, to the death of the rabbit.
Another control rabbit should be immunized with a control sample in RIBI. The
control would be a similar starting volume of 25 mM sodium cacodylate (pH 6.5)
containing 1.0 M NaCl and 0.3% (v/v) LDAO (approximately 20-25 m1), concentrated
using a centriprep 10 to about 1.0 ml, and ethanol and acetone precipitated (as
described, 41). The resulting pellet should be resuspended in PBS containing 1 mg PC
and dialyzed. (Note: The immunoaffinity purification of SAT-1 has been modified to

210

reduce the amount of LDAO in the enzyme fraction. Detergent extraction and
immunoaffinity purification is carried out as described (Chapter 3), but an additional
wash step changes the column buffer to 25 mM sodium cacodylate (pH 6.5) containing
0.1 M NaCl and 0.3% LDAO. The enzyme is eluted in the same buffer with 1 M
NaCl.)

Closing Statement

Research in our laboratory focuses on the chemistry and metabolism of
glycosphingolipids as it pertains to the modulation of cell proliferation by GM3
ganglioside. Our working hypothesis for cells in a proliferative state is that inhibition
of cell growth by GM3 ganglioside is relieved through the dissociation of GM3 from
the epidermal growth factor receptor (EGF-R) by degradation of GM3 to
lactosylceramide (LacCer) by GM3 sialidase. Loss of the sialic acid from GM3
relieves the constraint on EFG-R autophosphorylation and cell growth continues.
LacCer, once internalized to the Golgi, serves as substrate by CMP-sialic acidzLacCer
a2-3 sialyltransferase (SAT-1) to reform GM3 and repopulate the cell surface to
modulate another round of the cell growth cycle.

The studies described in this dissertation were specifically aimed at purifying
SAT-l to address the question of the regulation of GM3 ganglioside biosynthesis, more
specifically how SAT-1 is regulated. Purification and characterization of SAT-1
generates information which will prove most useful to others investigating glycolipid
glycosyltransferases. SAT-1 activity is unique among glycosyltransferases. It is
specific for a particular carbohydrate-lipid structure and as such reflects its potential as
a key regulatory step in ganglioside biosynthetic cascade. The results of the cell-cycle
dependent expression of SAT-l, potentially modulated by tyrosine phosphorylation,
when considered with a variety of other studies on GM3 metabolism and function, may
enable others to continue to define the mechanism played by GM3 ganglioside in the

regulation of cell growth and ultimately mechanisms controlling oncogenesis.

211

10.

11.

12.

13.

14.

REFERENCES
Hakomori, S. (1990) J. Biol. Chem. 265, 18713-18716.
Hakomori, (1985) Cancer Res. 45, 2405-2414.
Chatterjee, S., Sweeley, C.C. and Velicer, L.F. (1975) J. Biol. Chem. 250,
61-66.
Macher, B.A., Lockney, M., Moskal, J.R., Fung, V.K. and Sweeley, C.C.
(1978) Exp. Cell Res. 117, 95-102.
Moskal, J.R., Lockney, M.W., Marvel, C.C., Mason, P.A. and Sweeley, C.C.
(1980) in ACS Symposium Series, No. 128, Cell Surface Glycolipids (C.C.
Sweeley, Ed., p.p. 241-263.
Burczak, J.D., Soltyziak, R.M. and Sweeley, C.C. (1984) J. Lipid Res. 25,
1541-1547.
Usuki, S., Lyu, S.-C. and Sweeley, C.C. (1988) J. Biol. Chem. 263, 6847-
6853.
Usuki, S., Hoops, P. and Sweeley, C.C. (1988) I. Biol. Chem. 263, 10595-
10599.
Usuki, S. and Sweeley, C.C. (1988) Ind. J. Biochem. Biophys. 25, 102-105.
Swamy, M.J. and Sweeley, C.C. (1989) Biochem. Biophys. Res. Commun.
162, 1188-1193.
Bremmer, E.G., Hakomori, S., Bowen-Pope, D.F., Raines, E. and Ross, R.
(1984) J. Biol. Chem. 259, 6818-6825.
Bremmer, E.G., Schlessinger, J., and Hakomori, S. (1986) JBC 261, 2434-
2440.
Hanai, N., Nores, G., Torres-Mendez, C.R., and Hakomori, S. (1987)
Biochem. Biophys. Res. Commun. 127-134.
Hanai, N., Nores, G.A., MacLeod, C., Torres-Mendez, C.-R. and Hakomori,
S. (1988) J. Biol. Chem. 263,10915-10921.

212

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.
25 .

26.

27.

28.

29.

30

31.

Hanai, N., Dohi, T., Nores, G.A. and Hakomori, S. (1988) J. Biol. Chem.
263, 6296-6301.

Igarashi, Y., Kitanura, K., Toyokuni, T., Dean, B., Fenderson, B.A., Ogawa,
T., and Hakomori, S. (1990) J. Biol. Chem. 265, 5385-5389.

Weis, F.M.B. and Davis, R.J. (1990) JBC 265,12059-12066

Schlessinger, J. (1988) J. Biol Chem. 263, 3119-3123.

Wedegaertner, PB. and Gill, G.N. (1989) JBC 264, 11346—11353.

Hanai, N., Nores, G.A., MacLeod, C., Torres-Mendez, CR. and Hakomori,
S. (1988) J. Biol. Chem. 263, 10915-10921.

Dawson, G., McLawhon, R. and Miller, R.J. (1980) J. Biol. Chem. 255, 129-
137.

Capasso, J.M., Keenan, T.W., Abeijon, C. and Hirschberg, C.B. (1989) J.
Biol. Chem. 264, 5283-5240.

Strous, G.J., van Kerhoff, P., Fallon, R.J. and Schwartz, AL. (1987) Eur. J.
Biochem. 169, 307-311.

Dumont, J .13., Jauniaux, J.-C. and Roger, RP. (1989) TIBS 14, 67-71.
Moskal, J.R., Lockney, M.W., Marvel, C.C., Trosko, J.E. and Sweeley, C.C.
(1987) Cancer Res 47, 787-790.

Spiegel, S. (1989) in Gangliosides and Cancer (Oettgen, H.F., ed.) VCH
Publishers, New York, NY, p.p. 17-30. .

Nagai, Y. and Tsuji, S. (1989) in Gangliosides and Cancer (Oettgen, H.F.,
ed.) VCH Publishers, New York, NY, p.p. 33-39.

Burczak, J.D., Soltyziak, R.M., and Sweeley, C.C. (1984) J. Lipid Res. 25,
1541-1547.

Bishop, W.R. and Bell, R.M. (1988) Oncogene Res. 2, 205-218.

Low, M.G. and Saltiel, A.R. (1988) Science 239, 268-275.

Carey and Hirschberg, C. (1981) J. Biol. Chem. 256, 989-993.

213

32.

33.

34.
35.

36.

37.

38.

39.

40.

41.

42.
43.

Melkerson-Watson, L.J., Kanemitsu, K. and Sweeley, C.C. (1987)
Glycoconjugate J. 4, 7-16.

Melkerson-Watson, LI. and Sweeley, C.C. (1991) J. Biol. Chem., in press
(see Chapter 3).

Matsudaira, P. (1990) Methods Enzymol 182, 602-613.

Lance, P., Lau, K.M. and Lau, J .T.Y. (1989) Biochem. Biophys. Rec.
Commun. 164, 225-232.

Paulson, J.C. and Colley, K.J. (1989) J. Biol. Chem. 264, 17615-17618.
Trinchera, M. and Ghidoni, R. (1989) J. Biol. Chem. 264, 15766-15769.
Weinstein, J., Lee, E.U., McEntee, K., Lai, P.H. and Paulson, J.C. (1987) J.
Biol. Chem. 262, 17735-17743.

Iadisch, S. (1989) in Gangliosides and Cancer (Oettgen, H.F., ed.) VCH
Publishers, New York, NY, p.p. 217-230.

Portoukalian, J. (1989) in Gangliosides and Cancer (Oettgen, H.F., ed.) VCH
Publishers, New York, NY, p.p. 165-171.

Melkerson-Watson, LJ. and Sweeley, C.C. (1990) Biochem. Biophys. Res.
Commun. 172, 165-171.

O'Farrell, P.H. (1975) J. Biol. Chem. 250, 4007-4021.

Laemmli, U.K. (1970) Nature 277, 680-685,

 

214

APPENDIX A

215

REPORT OF LABORATORY EXAMINATION

P.O. Box 30076

Lansing, MI. 48909

I’IIIAL PAGE 1 OF 2 (I)

CROSS NECROPSY

“‘IC‘OIV4.
‘

ANIMAL HEALTH DIAGNOSTIC LABORATORY c‘
Case Number: 1008571
Reported : 08/09/90
Received : 07/30/90

Phone(517)3534683

Client Account:

STEIN, SUSAN

 

Pathologist: JAR

'1‘

PRIVILEGED INFORMATION . m" (..a‘ _ . .
NOT FOR PUBLICATION Case Or rgrn. NECROPSY

9016 Owner:

BIOCHEHISTRY

UNIVERSITY LAB ANIMAL RESOURCES
C-III CLINICAL CENTER
UNIV. ACCT. '21—3019

CANPUS

Name
Breed :

History :

Specimen:

HI A882“

: BETH Age: UNKNOWN

RABBIT DOMESTIC Sex: FEMALE

The rabbit arrived at HSU on 1/10/90. The animal was inoculated
three times with CH? sialic acid in an adjuvant of rat liver origin.
This enzyme catalyzes the synthesis of 6H3 ganglioside. The
inoculations began 6/18/90. Shortly after the inoculations began,
the rabbit was intermittently anorectic. On 7/27/90, the rabbit was
inappetent and was not defecating. Shortly afterwards, a purulent
ocular nasal discharge was observed. The rabbit was forcefed a
cereal-based diet on 7/27 through 7/29. The animal was euthanatized
by a terminal sanguination under anesthesia.

CARCASS

GROSS LESIONS, < 200 L85

The carcass
The stomach contained a poorly formed hairball.

jejunum and

weighed 2.1 kg and was judged to be in good condition.
Portions of the

cecum were distended with gas (postmortem change). The

gallbladder was distended, and the contents were inspissated

(alteration

possibly secondary to the presence of a gastric

hairball). Both kidneys were similar in appearance, and were tan

with multiple coalescing white, non-raised feci.

The foci were up to

a centimeter in diameter and appeared to be confined to the cortex.
All skeletal muscles contained yellow/white streaks (indication of

degeneration, necrosis and/or mineralization).
red muscles were affected, especially the white muscles.

Both the white and
The heart

was unremarkable.

LABORATORY FINDINGS
Nutritional examination:

Tests are in progress, and results will be

forwarded in a supplemental report when available.

*oENOTES ADDITIONAL TEST ncsuu

sec” 0C Ana ae‘reneea-Vuu’ nf‘v.(\.. (A.... ne.o.nuu.....‘u ...r-~. ..

1216

FINAL PAGE 2 OF 2 (1)

Case Number: 1008571

Specimen: FIXED TISSUES

HISTOPATHOLOGIC EXAMINATION
Hultiple tissues were examined. Skeletal muscle fibers consisted of

severe degeneration and necrosis. Affected muscle fibers were
scattered throughout muscle fascicles. Hyofiber alterations included
loss, vacuolation, hyalinization, and mineralization. In regions in
which myofibers were lost, there was an accumulation of macrophages.
The heart contained focal aggregates of lymphocytes within the
endocardium of the right ventricle and the myocardium of the left
ventricular papillary muscle. The lymphocytes were pleomorphic, and
occasional mitotic figures were observed. The kidney contained
numerous multifocal interstitial, coalescing aggregates of
pleomorphic lymphocytes. Nithin the region of interstitial
lymphocytic infiltration, there was loss of tubules. In other
tubules, there were protein casts. The glomeruli had no thickening
of the mesangium or basement membranes. The glomeruli did appear
slightly hypercellular, due to the presence of cells resembling
lymphocytes. The distribution of the alterations within the kidney
appeared to be cortical. Numerous lymphocytes were present within
the region adjacent to the renal transitional epithelium. The liver
had prominent portal areas, due to the presence of numerous
lymphocytes. The lymphocytes were encircling bile ducts and extended
into the adjacent hepatic parenchyma. The lymphocytes were markedly
pleomorphic, and mitotic figures were present. Centrilobular
hepatocytes had mild vacuolar change. The gallbladder contained
inspissated bile, which was amorphous and basophilic. Numerous
lymphocytes were present within the wall of the gallbladder. Mild
hemosiderosis was present in the spleen. Mild peribronchiolar
lymphocytic aggregation was present within the lung. The adrenal
cortex capsule and medulla were infiltrated by lymphocytes. The gut-
associated lymphoid tissue was prominent, and some macrophages were
present. The lamina propria of the duodenum contained a focal
accumulation of lymphocytes. Other tissues were unremarkable.

CONCLUSION
Disseminated lymphosarcoma; disseminated skeletal muscle degeneration

and necrosis, suggestive of a nutritional myopathy.

Comments:
James A. Render 8/9/90

Pathologist (517) 353—5275

538

Additional copies to: Dr. Render, +1

217

“I

fi-s’???
k