STRUCTURE OF DOG INTESTINAL FORSSMAN HAPT EN
AND PURIFICATION AND PARTIAL CHARACTERIZATION
OF FORSSMAN‘ HAPTEN HYDROLASE
(31- N -ACETYLGALACTOSAMINIDASE EC. 3.2.1.49)
FROM PORCINE LIVER

Dissertation for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
SUN-SANG JOSEPH SUNG
1977

 
  

  
   
 

L r a}: A R. '7
Michigan State i
University

This is to certify that the

thesis entitled

Structure of Dog Intestinal Forssman Hapten and
Purification and Partial Charactenization of
Forssman Hapten Hydrolase
(a-fl:Acetylgalactosaminidase EC. 3.2.1.49)

from Porci e Hiver
presen ed y

Sun-Sang Joseph Sung

has been accepted towards fulfillment
of the requirements for

Ph. D. degree in Biochemistry

W. M.

 

 

Major profes

M. n, 1777
d ,

0-7639

 

 

 

 

 

ABSTRACT

STRUCTURE OF DOG INTESTINAL FORSSMAN HAPTEN AND
PURIFICATION AND PARTIAL CHARACTERIZATION OF
FORSSMAN HAPTEN HYDROLASE
(o-NfACETYLGALACTOSAMINIDASE EC. 3.2.1.49) FROM PORCINE LIVER

By
Sun-Sang Joseph Sung

The structure of a pentaglycosylceramide from canine intestines
was shown by gas chromatography, gas chromatography-mass spectrometry
and stepwise stereospecific glycosidase degradation to be fl:acetylgalac-
tosaminyl-(al+3)-flyacetylgalactosaminyl-(Bl+3)-galactosyl-(al+4)-galac-
tosyl-(Bl+4)-glucosyl-(l+l)-ceramide.

The hydrolysis of this glycosphingolipid by G-Nfacetylgalactosa-
minidase from porcine liver was studied. The enzyme was purified 3,300-
fold with respect to p-nitrophenyl-afiflracetylgalactosaminidase activity
and 19,600-fold with respect to Forssman hydrolysing activity. Steps
in the purification included acid precipitation, ammonium sulphate
precipitation, and chromatography on DEAE-cellulose, Con A-Sepharose,
DEAE-cellulose (ampholine elution), Sephadex 6-150 and hydroxylapatite.
Purity was judged to be greater than 90% when analysed by SOS gel elec-
trophoresis. The subunit molecular weight was 52,000 daltons and that
of the native enzyme was 102,000 daltons. Porcine afiflyacetylgalactosa-
minidase is a glycoprotein; the carbohydrate moiety was found to consist

of mannose and Nfacetylglucosamine which together accounted for 7% of

Sun-Sang Joseph Sung
the total weight of the enzyme. The amino acid composition of the enzyme
was obtained. Isoelectric focussing gave eight enzymatically active
peaks with isoelectric points between 5 and 6.5 pH units. The kinetic
behavior and enzyme mobility on native polyacrylamide gels of the eight
forms were similar.

The kinetic properties of purified porcine a-flfacetylgalactosamini-
dase towards p-nitrophenyl-a-flyacetylgalactosaminide, Forssman hapten,
Forssman oligosaccharide, fly[l-]4C]-acetyl-sphingosyl-Forssman-oligosac-
charide, porcine submaxillary mucin and GalNAc-(al+6)-l,2:3,4-di-iso-
propylidene-galactose were examined. The enzyme hydrolysed these sub-
strates with KM values of 2.9 x 10'3, 2.6 x l0'4, 1,0 x 10'2, 2,5 x 10‘4,
2.3 x 10-6, and 4.8 x 10'3 M, respectively. The Vmax values were 14,],
4.2, l.7, 0.81, 0.42 and l.l umoles/min/mg for each of these substra-
tes, respectively. p-Nitrophenyl-a-flfacetylgalactosaminide was hydroly-
sed at pH optimum of 4.5. The hydrolysis of Forssman hapten was optimal
at a taurocholate concentration of about 4 mg/ml and a pH of 3.9. The
hydrolysis 0f ErEI'14CJ-acetyl-sphingosyl-Forssman-oligosaccharide at
pH 3.9 was also optimal at a detergent concentration of 4 mg/ml but
with a much sharper peak. However, taurocholate at a concentration of
4 mg/ml inhibited the hydrolysis of Forssman-cligosaccharide. Hydro-
lysis of porcine submaxillary mucin was optimal at pH 4.2. The enzyme
also hydrolysed three different human red blood cell A+ glycosphingo-
lipids and a dog intestinal A+ blood group glycolipid.

The hydrolysis of Forssman hapten was slightly inhibited by the
alkaline earth metals. The transition metals inhibited the enzyme

hydrolysis much more strongly; the most potent of these inhibitors were

Sun-Sang Joseph Sung

2+ 2

+, Fe3+

Co , Cu , HgZ+, and Pb2+. flrAcetyltalosamine-(al+6)-l,2:3,4-
di-isopropylidene-galactose inhibited the hydrolysis of Forssman hapten
competitively with a K1 of l.5 mM. Half of the Forssman hapten hydro-

lysing activity was lost after 0.4 minutes at 60° and 6 minutes at 50°.

STRUCTURE OF 006 INTESTINAL FORSSMAN HAPTEN AND
PURIFICATION AND PARTIAL CHARACTERIZATION OF
FORSSMAN HAPTEN HYDROLASE
(a-flrACETYLGALACTOSAHINIDASE EC. 3.2.1.49) FROM PORCINE LIVER
By

Sun-Sang Joseph Sung

A DISSERTATION

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

Department of Biochemistry

1977

dedicated
to
YANG LI and my PARENTS
without WHOM

this work would not have been completed

ii

ACKNOWLEDGEMENTS

I would like to express my sincere gratitude to those who strived
to contribute to a better society, thus making this work possible. My
special thanks go to all my friends who have made my five and more
years of life here so thoroughly enjoyable and to Dr. Charles C. Sweeley,

whose patience, understanding and erudition have made this a most en-

lightening experience.

iii

TABLE OF CONTENTS

PAGE
LIST OF TABLES .......................... ...' ....................... vii
LIST OF FIGURES ................................................... viii
LIST OF ABBREVIATIONS ............................................. xii
INTRODUCTION ...................................................... 1
LITERATURE REVIEW ................................................. 3
Forssman Hapten .............................................. 3
Blood Group A Active Substances .............................. 8
Metabolism of Neutral glycosphingolipids ..................... 14
a-NrAcetylgalactosaminidases ............................ l4
B-N:Acety1hexosaminidases ............................... 16
a-Galactosidases ........................................ l9
B-Galactosidases ........................................ 21
Glucocerebrosidase ...................................... 24
In Viva Metabolism of Glycosphingolipids ..................... 25
Enzyme Therapy of Glycosphingolipidoses ...................... 27
Glycosidases in Transformed Cells ............................ 28
MATERIALS AND METHODS ............................................. 29
Materials .................................................... 29
Methods.. .................................................... 31
(I) Substrate Preparation ............................... 31
l Forssman Hapten( GL-S ) ....................... 31
2 Labelling of GL-S with the Galactose Oxidase-
Tritiated Sodium Borohydride Method ........... 32

(3) Synthesis of GalNAc-(al+6)-I,2:3,4-di-Qf
isopropylidene-galactose and TalNAc-(al+6)-
1,2:3,4-di-Q:isopropylidene-galactose ......... 35

(4) Preparation of A+-Porcine Submaxillary Mucin... 37

(5) Isolation of A+-Glycosphingolipid from

Human Red Blood Cell Membrane ................. 37
(6) Preparation of [3H]-GL-5-Pentasaccharide ....... 38
(7) Preparation of N;[l-I4C]-Acetyl-Sphingosy1-

Forssman-Pentasaccharide ...................... 39

iv

TABLE OF CONTENTS (cont'd)

PAGE
(II) Enzyme Assays

(1) Assay with Forssman Glycolipid ............... 40
(2) Assay with p-Nitrophenyl—a-N-acetylgalac—

tosaminide .................................. 40
(3) Assay with Forssman- Pentasaccharide .......... 40
(4 Assay of N-[l-14C]-—Acetyl-sphingosyl-Forssman

-pentasaccharide Hydrolysis ................. 41
(5) Assay of Porcine Submaxillary Mucin

Hydrolysis .................................. 42

(6) Assay of N—Acetylgalactosamine-(al+6)- l, 2. 3, 4
-di-0-isopropylidene-Galactose Hydrolysis. .42
(7) HydroTysis of Human Red Blood Cell and Dog

Intestinal Blood Group A Glycolipid ......... 43

(III) Enzyme Purification ............................. 43

(IV) Preparation of Con A-Sepharose ................... 45

(V) Methanolysis ...................................... 45

(VI) Permethylation for Linkage Studies ............... 46

(VII) Polyacrylamide Gel Electrophoresis .............. 47

(VIII) General Procedures ............................. 48

RESULTS .................. I. V ..... .i ...... . . . I ....................... so

Characterization of the Pentaglycosylceramide in Dog
Intestines ............................................ 50

Purification of a-NfAcetylgalactosaminidase from
Porcine Liver ......................................... 69

Physical and Chemical Charaterization of Enzyme
Preparation ........................................... 76

Isoelectric Focussing of the Purified Enzyme Preparation... 87

Kinetic Properties of the p-Nitrophenyl-a-Nyacetyl-
galactosaminidase Activity ............................ 9l

Hydrolysis of Forssman Hapten by a-N-Acetylgalactos-
aminidase ............................................. 100

(1) Characterization of Products of Hydrolysis of
Forssman Hapten .................................. 100

(2) Kinetics of Hydrolysis of Forssman Hapten ......... 100

TABLE OF CONTENTS (cont'd)

‘PAGE
(3) Inhibitors of Forssman Hydrolase Activity ......... 105
(4) Temperature Inactivation of Forssman Hapten
Hydrolase ........................................ 112
Hydrolysis of Forssman Pentasaccharide by a-N7Acetyl-
galactosaminidase ..................................... 112
(1 Characterization of Forssman Pentasaccharide ...... 112
(2 Characterization of Products After Hydrolysis
of Forssman Pentasaccharide with a-NfAcetyl-
galactosaminidase ................................ 116
(3) Kinetics of Hydrolysis of Forssman Pentasac-
charide .......................................... 116
Hydrolysis of N-[l-14CJ-Acetyl-Sphingosy1-Forssman-
Pentasaccharide ....................................... 116

(1) Characterization of Nf[1-I4C]-Acety1-sphingosyl-
Forssman-pentasaccharide ......................... 116

(2) Characterization of Hydrolysis Products from
a-N-Acetyl-sphingosyl-Forssman Pentasaccharide... 122

(3) KinEtics of the Hydrolysis of N-[l- CJ-Acetyl-
sphingosyl-Forssman PentasacCharide .............. 125
Hydrolysis of Porcine Submaxillary Mucin by afiNyAcetyl-
galactosaminidase ..................................... 125
Hydrolysis of N-Acetylgalactosaminyl-(a1+6)-l,2:3,4-di-Qf
isopropylidene-galactose .............................. 132
Hydrolysis of Human Red Blood Cell and Dog Intestinal
A -Glycosphingolipid .................................. 139
DISCUSSION ...................................................... 142
Structural Determination of Forssman Hapten ................ 142
Purification of a-NrAcetylgalactosaminidase ................ 147

Characterization of the Kinetic Properties of a-NrAcetyl-
galactosaminidase ..................................... 149

The Role of Carbohydrates on a-NfAcetylgalactosaminidase... 156
SUMMARY ......................................................... 161
BIBLIOGRAPHY .................................................... 163

vi

LIST OF TABLES

TABLE PAGE
1. Potential Substrates of agNrAcetylgalactosaminidase ......... 11
2. Yields of Glycosphingolipids from Canine Intestines ......... 53
3. Purification of a-NrAcetylgalactosaminidase from Porcine

Liver.... ................................................ 70
4. Carbohydrate Composition of a-N:Acetylgalactosaminidase ..... 88
5. Amino Acid Composition of a-NrAcetylgalactosaminidase ....... 89
6. Specificities of the Multiple Isoelectric Focussing

Forms of a-N:Acetylgalactosaminidase ..................... 96
7. Metal Inhibitors of Forssman Hydrolase ...................... 109
8. Carbohydrate Derivatives in the Inhibition of Forssman

Hydrolase ................................................ 110
9. Comparison of the Kinetic Parameters of the Hydrolysis of

Different Substrates by afiNyAcetylgalactosaminidase ...... 151
10. Structures of Oligosaccharides in Glycoproteins ............. 157

vii

LIST OF FIGURES

FIGURE PAGE

1. Metabolism of Glycosphingolipid Substrates of aij

Acetylgalactosaminidase .................................... 15
2. Preparation of [3HJ-eL-5 and Its Derivatives .................. 33
3. Synthesis of N-Acetylgalactosaminyl-(al+6)-1,2:3,4-di-0-

isopropylidEne-Galactose and N-Acetyltalosaminyl-(dT46)-

l,2:3,4-diagrisopropylidene-Gilactose ...................... 36
4. Thin-Layer Chromatography of Glycosphingolipids of

Canine Intestine and Kidney ................................ 51
5. Gas-Liquid Chromatography of Trimethylsilylated Methyl

Glycosides of Standard Sugars, Canine Forssman Hapten

and Porcine Globoside ...a ................................. 54
6. Thin-Layer Chromatography after Enzymatic Hydrolyses of

Forssman Hapten ............................................ 56
7. Gas-Liquid Chromatography of Neutral Partially Methylated

Alditol Acetates from Canine Forssman Hapten and

Porcine Globoside..... ..................................... 59

10.

ll.

12.

I3.

. Gas-Liquid Chromatography of Partially Methylated Alditol

Acetates of Neutral and Aminosugars from Forssman Hapten... 52

. Primary Fragments of Partially Methylated Alditol Acetates

of Aminosugars ............................................. 63

Mass Spectrum of Partially Methylated Alditol Acetate
of Terminal Aminosugar of Forssman Hapten Compared with
a Standard.. ........ . ............ . ......................... 54

Mass Spectrum of Partially Methylated Alditol Acetate of
the Internal Aminosugar of Forssman Hapten Compared
with Standards... .......................................... 55

A Typical Elution Pattern of a-NyAcetylgalactosaminidase
from Con A-Sepharose Column ................................ 71

DEAE-Ampholine Elution in the.Purification-of.a-N¢
'Acetylgalactosaminidase .................................... 72

viii

FIGURE PAGE

14. Native Gel Electrophoresis of EnZyme Preparations in.
the Purification of a-NfAcetylgalactosaminidase ........... 73

15. Sephadex G-150 Elution Profile in the Purification of
a-NrAcetylgalactoSaminidase ............................... 75

16. Repeated Sephadex G-150 Column Chromatography of aeNy
Acetylgalactosaminidase ................................... 77

17. Hydroxylapatite Chromatography in the Purification of
afiN7Acetylgalactosaminidase ............................... 78

18. Scans of Native Gel Electrophoresis of Purified a—N:
Acetylgalactosaminidase Preparation from Hydroxylapatite
Column .................................................... 79

19. Scan of sos Gel Electrophoresis. of Purified a-N—Acetyl-
galactosaminidase from Hydroxylapatite Column ............. 82

20. Molecular Height Determination of the Subunits of
Purified a-N-Acetylgalactosaminidase in SDS Poly-
acrylamide Gels ........................................... 83

21. Gas-Liquid Chromatography Trace of Carbohydrate Analysis
of a-NrAcetylgalactosaminidase ............................ 85

22. Isoelectric Focussing of Partially Purified a-N:Acety1-
galactosaminidase ......................................... 90

23. Native Polyacrylamide Gel Electrophoresis of Isoelectric
Focussing Peaks of aaNrAcetylgalactosaminidase ............ 92

24. Scan of Native Polyacrylamide Gel Electrophoresis of the
Isoelectric Foccussing Peak of a1NyAcetylgalactosaminidase 94

25. Enzyme Activity of pNP-u-NyAcetylgalactosaminidase vs
Enzyme Concentration ...................................... 97

26. Rate of Hydrolysis of pNP-aaN:Acetylgalactosaminide vs pH.... 98

27. Lineweaver-Burk Plot for the Hydrolysis of p-Nitrophenyl-
angAcetylgalactosaminide by aeNrAcetylgalactosaminidase.. 99

28. Radio-Thin-Layer Chromatogram and Thin-Layer Chromatography
, of the Folch Upper Phase Product after Hydrolysis of
Forssman Hapten by a-NyAcetylgalactosaminidase ............ 101

29. Radio-Thin-Layer Chromatogram and Thin-Layer Chromatography
of the FOlch Lower Phase Product after Hydrolysis of.
Forssman Hapten by aeNfAcetylgalactosaminidase ............ 102

30. Rate of Hydrolysis of Forssman Hapten vs Enzyme COncentra-
tion ...................................................... 104

ix

FIGURE PAGE

31.
32.

33.

34.

35.

36.
37.

39.

40.

41.

42.

43.

44.

45.
46.

Rate of Hydrolysis of Forssman Hapten vs pH ...................... 106

The Effect of Taurocholate Concentration on the Rate of
Hydrolysis of Forssman Hapten ................................. 107

Lineweaver-Burk Plat for the Effect of Change of Substrate
Concentration on the Rate of Hydrolysis of Forssman Hapten....108

Lineweaver-Burk Plot of Hydrolysis of Forssman Hapten
with NrAcetyltalosaminyl-(a1+6)-1,2:3,4—di-Qriso-

propylidene—Galactose as an inhibitor ......................... 111
Semi-logarithmic Plot of'Temperature Inactivation of

Forssman Hapten Hydrolysing Activity vs Time .................. 113
Thin-Layer Chromatography of Forssman Pentasaccharide ............ 115

Thin-Layer Chromatography of Hydrolysis Products Obtained after
a-N:Acetylga1actosaminidase Treatment of Forssman
Pentasaccharide ............................................... 117

. The Activity of Hydrolysis of Forssman Pentasaccharide vs

Amount of Enzyme .............................................. 118

Lineweaver-Burk Plot of the Hydrolysis of Forssman Penta-
saccharide by a-N:Acetylgalactosaminidase ..................... 119

Characterization of N:[1-14C]-Acety1—sphingosyl-Forssman-
pentasaccharide ............................................... 120

Thin-Layer Chromatography and Radioactivity Scans of the
Products of Hydrolysis of N7[l-14C]-Acetyl-Sphingosy1-
Forssman-pentasaccharide by atNyAcetylgalactosaminidase ....... 123

Detergent Dependence in the Hydrolysis of N¢[l-]4C]-
Acetyl-sphingosy1-Forssman-pentasaccharide by B:Nr
Acetylgalactosaminidase ....................................... 126

Hydrolysis of N-[1-14CJ-Acetyl-sphingosy1-Forssman-
pentasaccharide by Varying Amounts of a-NfAcetylgalac-
tosaminidase .................................................. 128

Lineweaver-Burk Plot of the Hydrolysis of Ny[1-]4C]-Acetyl-
sphingosyl-Forssman-pentasaccharide by afflfACEtYI'
galactosaminidase ............................................. 129

Effect of pH on the Hydrolysis of Porcine Submaxillary Mucin ..... 130

Hydrolysis of Porcine Submaxillary Mucin with Varying
Amount of afiNfAcetylgalactosaminidase ......................... 131

FIGURE PAGE

47. Lineweaver-Burk Plot of Porcine Submaxillary Mucin
Hydrolysis by a-NrAcetylgalactosaminidase .................. 133

48. Thin- -Layer Chromatography of N— -Acety1talosaminy1- (a1+6)-
1 ,2:3,4-di-0-isopropy1idene—Galactose, N- -Acetylga1actosaminyl-
(a1+6)- 1 ,2: 3, 4- di- -0-isopropylidene-Galactose and N—Acetyl-
galactosaminyl- (a1+6)- Galactose ............................ 134

49. Gas-Liquid Chromatography of Trimethylsilylated Methyl
Glycosides of Methanolysed N-Acetylgalactosaminyl- (a1+6)-
Galactose, N-AcetylgalactosaminyL -(a1+6)- l, 2: 3,4-di- 0-
isopropylidene-Galactose and N-Acetyltalosaminyl- (a1+6)-
1 2: 3, 4-di-0- -isopropy1idene-Galactose ...................... 136

50. Lineweaver-Burk Plot of the Hydrolysis of N-Acetylgalactosa-
minyl- -(a1+6)- l ,2: 3,4-di-0- -isopropy1idene-Galactose by
a-N-Acetylgalactosaminidase ................................ 138

51. Thin-Layer Chromatography of Products of Hydrolysis of Blood
Group A Glycosphingolipids by a-NrAcetylgalactosaminidase..140

52. Gas-Liquid Chromatography of the Partially Methylated
Alditol Acetate Derived from the Purified Fucodecasa-
ccharide A from the Urine of Fucosidosis Patients .......... 143

53. Mass Chromatograms of the Partially Methylated Alditol

Acetates of Decasaccharide A from the Urine of Fucosidosis
Patients ................................................... 145

xi

GC-MS
GLC

TLC

GL-5
6N1

GM2
GM3

LIST OF ABBREVIATIONS*

Chromatography
Gas-Liquid Chromatography-Mass Spectrometry
Gas-Liquid Chromatography
Mass Spectrometry

Thin-Layer Chromatography

Glycosphingolipids
Glucosylceramide, Glucocerebroside
Galactosylceramide, Galactocerebroside
Lactosylceramide

Galactosyl-(a1+4)-galactosyl-(81+4)-glucosyl-
(l+1)-ceramide

Globoside, N-Acetylgalactosaminyl-(81+3)-galactosy1-

(a1+4)-galactosyl-(Bl+4)-glucosyl-(1+1)-ceramide
Forssman hapten, Forssman antigen, Forssman glycolipid

Gal-(Bl+3)-GalNAc-(Bl+4)-Gal-[3+2aNeuAc]-(Bl+4)-
Glc-(1+l)-Cer

GalNAc-(Bl+4)-Ga1-[3+2aNeuAc]-(Bl+4)-Glc-(1+1)-Cer
NeuAc-(a243)-Ga1-(Bl+4)-G1c-(1+1)-Cer

* Standard abbreviations have been defined in J. Biol;Chem.251, 1-10

xii

LIST OF ABBREVIATIONS (cont'd)

 

Miscellaneous
Con A Concanavalin A
pNP p-Nitrophenyl
SDS Sodium dodecyl sulphate
RBC Red blood cell
BSA Bovine serum albumin
a.m.u. Atomic mass unit
DCE 1,2-dichloroethane
Cer Ceramide
TalNAc NrAcetyltalosamine,.Nyacetyltalosaminyl
PPD 2,5-Diphenyloxazole

Dimethyl-POPOP l,4-bis-[2-(4-Methy1-5-phenyloxazolyl)]-benzene

xiii

INTRODUCTION

A glycosphingolipid with a partial structure of Nfacylgalactosami-
nylaNyacylgalactosaminyl-ga1actosyl-ga1actosyl-glucosyl-ceramide was
reported to be a major glycolipid in dog intestines (1). Judging from
the sequence of sugar residues it was suggested that this glycosphingo-
lipid belonged to the globoside series, with an additional Nracetyl-
galactosamine moiety. When the structural studies were completed, it
was found that the glycosphingolipid indeed belonged to the globoside
series (2) and had the same structure as horse Spleen Forssman hapten (3).

Sweeley and Klionsky were the first to Show that a lipidosis (Fa-
bry's disease) was caused by the accumulation of complex neutral glyco-
sphingolipids (4): it was shown subsequently that accumulation of this
glycosphingolipid resulted from the deficiency of an acid a-galactosi-
dase (5,6). It is now recognized that deficiency of any of a variety
of lysosomal glycosidases gives rise to a specific glycosphingolipidosis
(7). Although these inherited diseases cannot be cured by present
technology, it may be possible to use enzyme replacement therapy, which
involves the injection of an active enzyme preparation into the patient
in a fbrm that could reduce the level of the accumulating glycosphingo-
lipids, to alleviate the symptoms of the disease. A thorough study of
the properties of lysosomal glycosidases should precede their introduc-
tion into patients so that the effectiveness of the therapy and possible

side effects, including immunological rejection, can be evaluated.

The human blood group A+ glycosphingolipids have terminal 9‘!7
acetylgalactosamine residues (8). A deficiency in the enzyme a-Nfacetyl-
galactosaminidase might therefore result in the accumulation of blood
group AI glycosphingolipids and expression of some clinical symptoms of
this lipidosis. No such disease has yet been described.

The enzyme properties of aaNyacetylgalactosaminidase were studied
in the hope that the development of a reliable assay for the enzyme
could render the disease more readily identified. Forssman hapten can
be purified readily from dog intestines. Its hydrolysis by an u-N:
acetylgalactosaminidase can be used as a model for studies of the
mechanism of enzymatic glycosphingolipid hydrolysis by lysosomal
hydrolases. The enzyme from pork liver was chosen because it was
partially purified previous to this work (9) and it gave a high speci-

fic activity in vitro with Forssman hapten as the substrate.

LITERATURE REVIEW

Forssman Hapten

In 1911, Forssman (10) demonstrated the formation of sheep blood
hemolysins after parenteral administration of extracts from guinea pig
organs into rabbits. The antigens that induce the formation of these
hemolysins have been called Forssman antigens in recognition of his
poineering efforts. They occur in many species of animals and bacteria
but not in plants (11). Landsteiner (12) showed that Forssman antigens
consist of a specific alcohol-soluble component which he called the
hapten and a nonspecific protein component. Brunius (11) observed sub-
sequently that the purified Forssman hapten from horse kidney was a
galactosamine-containing lipid and Papirmeister and Mallette (13) report-
ed that the Forssman hapten from sheep erythrocytes contained hexose,
hexosamine, fatty acid and a base. Makita at at. (14) were the first to
assign a structure, Nracetylgalactosaminyl-(a1+3)-ga1actosyl-(?1+4)-
galactosyl-(Bl+4)-glucosylceramide, for the Forssman hapten from horse
kidney and spleen. In agreement with these results, Mallette and Rush
(15) concluded that the Forssman hapten from sheep erythrocytes was also
a tetrahexosylceramide, with the same composition as that reported for
the horse Forssman hapten (14). Siddiqui and Hakomori (3) arrived at a
different structure on the basis of results with the Forssman hapten of
horse spleen, and preposed that the Forssman hapten has the structure
N:acetylga1actosaminyl-(al+3)ega1actosylaminyl-(31+3)—ga1actosyl-(al+4)-
galactosyl-(Bl+4)-glucosyl-(l+l)ceramide. This controversy was later

settled when Fraser and Mallette (16) and Makita, Yokoyama and Takahashi
(l7) reinvestigated their Forssman glycolipids and found that their
results agreed with the proposed structure of Siddiqui and Hakomori (3).

Vance, Shook and McKibbin (1) described a pentaglycosylceramide from
dog intestine which they proposed to be Nracetylgalactosaminyl:Nfacetyl-
galactosaminyl-galactosylega1actosyl-glucosy1-ceramide. Sung, Esselman
and Sweeley (2) Completed the anomerity and linkage studies of this gly-
colipid and showed that the oligosaccharide moiety was identical to that
of the horse spleen Forssman hapten. An identical structure was proposed
fbr Forssman hapten from goat erythrocyte (18) and guinea pig (19). To
date, there is only one exception to this widely accepted structure for
nammalian organs. This anomalous Forssman hapten was isolated from
hamster fibroblast NIL cells and has the structure of'N7acetylga1actos-
aminyl-(a1+3)-N:acety1galactosaminyl-(Bl+3)-galactosyl-(a1+4)-ga1actosyl-
(Bl+1)-ceramide (20). Bacterial Forssman hapten activity has not yet
been characterized chemically.

The immunological specificity of Forssman hapten probably resides
in the oligosaccharide moiety, since the oligosaccharide product after
ozonolysis and reduction of the N:acetylsphingosyl-Forssman hapten-
oligosaccharide retains its haptenic activity (21). The fact that
reduction of the oligosaccharide structure from a pentaglycosylceramide,
that of horse spleen, to a tetraglycosylceramide, as occurs in NIL cells,
does not affect the immunological specificity of the hapten indicates
that the nature of the sugar residue proximal to ceramide_may not play
a major role in the antibody-antigen interaction. Ceramide, with its
hydrophobic nature, is probably important in viva for the hapten to

anchor itself in cellular membranes and in the fbrmation of micelles.

Glycolipids are believed to be located on the surface of plasma
membranes (22-28). For example, Forssman hapten of erythrocytes is
probably located primarily on the erythrocyte membrane because it has
been isolated in quantity from washed stroma (15). However, glycosphin-
golipids are also found in substantial quantities in intracellular mem-
branes such as smooth endoplasmic reticulum (29), Golgi apparatus (30)
and primary and Secondary lysosomes (31). Smith and McKibbin (32) have
examined the subcellular distribution of glycolipids of dog intestinal
mucosal glycolipid and found that Forssman hapten is located in a crude
nuclear fraction, which also contains the cellular debris and in the
microsomal fraction, which also contains the plasma membrane.

Some possible functions of Forssman hapten in plasma membranes are
cell-cell interaction, adhesion and recognition, cellular membrane
structural organization, and action as a cell surface receptor. None of
these functions has yet been definitely proven, but a vast amount of
work has been done which implicates this glycolipid in some way in cell
transformation and malignancy. Forssman hapten, while absent in normal
human organs, has been reported to appear in a case of human metastatic
tumour of biliary adenocarcinoma in liver (33). Polyoma-induced hamster
kidney tumours were also shown by complement fixation to have an
increased concentration of Forssman antigen (34), and a number of cul-
tured cell lines were shown to have altered levels of cell surface
Forssman antigen after infection with different viruses.

O'Neill (35) has shown that Forssman antigen reactivity was induced
in polyoma and Rous virus-transformed baby hamster (BHK) cell lines, and
fbund that the presence of this antigen was dependent upon the maine
tenance of the transfonmed state. In another cell line (NIL) , however,

6
both normal and polyoma-transformed cell lines exhibited Forssman re-

activity. Utilizing heterologous rabbit antisera and immunofluorescent
techniques, Robertson and Black (36) detected a new antigenic activity
on the surfaces of SV40- and polyoma-transfbrmed BHK cells, which was
characterized as that of the Forssman antigen; no Such activity was
detected on adenovirus-transformed BHK cells. The increase in Forssman
reactivity on virus-transformed cell surfaces may result from induced
synthesis of the antigen or by the unveiling of a previously cryptic
antigen by changes in membrane conformation. Studies were performed to
distinguish between these processes in transformation. Burger (37)
treated normal BHK cells briefly with proteases and showed that these
cells have the same amount of Forssman antigen as was found in polyoma-
transformed-BHK cells not treated with protease. He concluded that the
Forssman receptor was not the result of new synthesis or induction of
antigenic activity in the tumour cells but rather represented a receptor
which is present in a cryptic form in certain cell lines and is exposed
in the course of transformation. Makita and Seyama (38) similarly
demonstrated that the trypsin-treated plasma membrane fraction of non-
transfbrmed BHK cells contains previously undetectable Forssman activity
which is similar in intensity to the activity of the polyoma-transformed
BHK plasma membrane fraction that can be obtained without trypsin treat-
ment. Trypsin treatment of polyoma-transformed BHK plasma membranes
does enhance the reactivity to a small extent, however.

The carbohydrate content of the plasma membrane was decreased as a

result of transformation. These results are consistent with the hypothe-

sis that there is decreased carbohydrate incorporation into membrane but

increased exposure of cryptic Forssman active sites following polyoaa

transformation.

By using NIL 2 hamster cell lines, Robbins and Macpherson (39,40)
and Sakiyama, Gross and Robbins (41) studied the db novo synthesis of
glycolipids by normal and virus-transfonmed cell lines based on the
incorporation of [‘4CJ-palmitate. In general, it was found that the
more complex neutral glycosphingolipids, including Forssman hapten,
were no longer syntheSized after transtrmation with adeno 7/SV40 hybrid
virus or hamster sarcoma virus. The synthesis of Forssman hapten in
tissue culture was also density-dependent. Its rate,of synthesis in
NIL cell lines increased as the cells grew from sparse to dense cultures
(42, 43). However, when these cells were transfbrmed by hamster sarcoma
or polyoma virus, the density dependent synthesis of the more complex
neutral glycolipids ceased (41—44). Moreover, when cell contact was
prevented by growing normal NIL cells in a spinner culture (45), there
was no appreciable increase in Forssman hapten. Dibutyryl adenosine-B'.
5'-cyclic monophosphate was employed to arrest cell growth, to examine
.whether the failure of hamster sarcoma virus transfbrmed NIL cells to
synthesize Forssman hapten was a result of continuous cell growth or a
direct result of transfbrmation (46). The data did not support the notion
that lack of Forssman hapten synthesis was due to continuous cell growth,
since transformed cells did not regain their ability to synthesize
Forssman hapten after arrest of cell growth. The cryptic nature of
Forssman hapten in cultured cells may be related to the cell cycle,
because transformed cells resemble mitotic cells in that they are also
rapidly dividing cells. Their analogy extends to exposed Forssman hapten
(47), which is found to be fully exposed in mitotic cells and becomes

cryptic after division.

' Blood Group A Active Substances

The A80 system is the first blood group system discovered. Land-
steiner (48) published an article on human blood group typing in 1900
and the importance of this system for blood transfusion was soon recog-
nized. A whole family of blood group substances has since been isolated,
purified and characterized from various soUrces. They are identified as
glycosphingolipids, glycoproteins, or oligoSaccharides. All blood group
A substances have the following immunological determinant,

BalNAc-(o1+3)-Gal[-(2+la)-Fuc]-(8144)-GlcNAc.
while in blood group B substances a gal-(a1+3) residue substitutes for
the N:acetylgalactosamine at the nonreducing end. Blood group 0 substan-
ces are antigens known as H substances that are thought to be precursors
of the A and B blood group substances, the immunological determinant is
Fuc-(o1+2)-Gal-(Bl+4)-GldlAc.

Among the blood group A substances, the most thoroughly studied
are probably those of the human red blood cell. Hakomori, Stellner and
Hatanabe (49) have characterized four antigenic variants of blood group
A-active glycosphingolipid, with the structures shown in Table 1.

Recently, an A-active megaloglycolipid (polyglycosylceramide) with
20-40 sugar residues per ceramide was isolated from human RBC (50). This
megaloglycolipid was thought to comprise a major portion of the A-active
antigenic sites on the red cell suuface. 'This view is contrary to the
commonly accepted concept that the red cell A-antigens are mostly glyco-
proteins (ST-63) but is supported by the work of Yamato, Handa and Yama-
kawa (64) who have determined that the blood group A activity of the gly-
coprotein preparation on a dry weight basis was only about one-two
hundredth of that of highly purified glycolipids. The A activity of the

9

glycoprotein fraction was mainly associated with PAS I (54, 64).

The blood group A-active glycosphingolipids in hog gastric mucosa
have also been Characterized (65-70). They vary in complexity from a
hexa- to an octadeca- glycosylceramide and are similar in structure to
human RBC Aa glycosphingolipid (49).

The saliva of certain individuals, known as secretors, are also A+.
The immunologically active mucin was purified from hog submaxillary mucin
(71, 72) and subjected to alkaline borohydride treatment (73). The
resulting A+ oligosaccharide was shown to have the structure shown in
, Table 1. Dogs also possess a blood group A active glycolipid (74), the
structure of which was prOposed to be

GalNAc-(a1+3)-Gal-(Bl»4)-GlcNAc-(Bl+3)-Ga1-(Bl+4)-Glc-Cer (75)
Fuca1+2
An oligosaccharide with blood group A activity has also been demonstrated
to occur in the human urine (76).

The occurrence of ABO blood group substances is widespread. They
are present as cell surface components of many tissue cells, including
the red blood cell, and in saliva, gastric juice, duodenal juice, bile,
spermatic fluid, vaginal secretions, amniotic fluid, milk, sweat, tears,
urine and meconium (77). Speculations about the functions of these
potent antigens are still premature, although there are quantitative and
qualitative changes of blood group activity in various carcinomas and
transfbrmed cells. Complete or partial loss of blood group A and 8 anti-
gens has been observed frequently in tumours arising from urinary epithe-
lium (78). Correlation with other properties of the tumour revealed
increased frequency of antigen loss, particularly among the more pleo-

morphic, anaplastic, infiltrating and rapidly fatal tumours. In a large

10
number of cases of gastrointestinal carcinoma, A, B and H blood group
activity was not detectable in all anaplastic cells that ceased to
secrete mucus while in mucinous carcinomas, there were both ABO
positive and negative neoplastic cell types (79). The same phenomenon of
progressively increasing loss of ABH antigens from carcinomas in situ to
anaplastic, invasive, and metastatic carcinomas was also demonstrated in
355 primary carcinomas of the uterine cervix, lung, pancreas and stomach,
578 metastatic carcinomas of these organs (80) and 12 cases of oral
squamous cell carcinomas (81). The incorporation of fucose into normal
and virus-transformed cells has also been examined. By the use of RE2
Sprague-Dawley rat and the murine sarcoma-murine lukemia transfbrmed REZ
(MSV-REZ) cell lines, Steiner, Brennar, and Melnick demonstrated that
there was a sharp increase in fucose incorporation into the larger
fUcosylglycolipid and a corresponding rise in radioactivity in the more
mobile and probably simpler fucosylglycolipids (82). Similar results
were obtained with human cell lines transfbrmed with malignant melanoma,
rhabdomyosarcoma and lung carcinoma (83). By the use of a cold-sensitive
mutant of murine sarcoma virus, it was shown that alterations in fucosyl-
glycolipid metabolism were related directly to the expression of the
transformed state, and were not simply the result of murine sarcoma virus
infection (84). Although Stellner and Hakomori (85) showed that the loss
of A and B glycolipids in carcinomas may be due to the loss of the trans-
ferase activities that convert H to A and B glycolipids,the relationship

between carcinoma and loss of'ABH antigens is still not clear.

11

 

“a.aa.

 

 

rooapaopumon
ween“.
massages roacwcamouxpe
om oursacou ea opos\mo=vvmms camsm omrom ouauoeguzgo cuss: < cacao uoopm
oezauasum ucvgucnsn
Fecovuvuum an ear: new u<zo_u m>_mmouxu nor: “an u< op cap.svm u<
mh~+_Vu=a
- xfTézaaAzTPa
c Lou-upwufietpV ...u<z—uw
Hpae- AMTFV-a<za_aV- AarpV- Paw
/ “mapV-a<z~aa
5565.2??? 3.4
«rpVoau u<
Lao-upe-m¢+PV-=APaaw-am+~V-u<zu.e
-A¢._V-Paa-hm+PV-u<zapw-A¢+.V- aaa-fie+~V- PaeJAm+.V-u<zpae a< a
a. .p
me sourupw ocusaeoe tonnesamooapm
-AerpV- ApauT AmrpV-u<za_e-AaapT hose- Ae+~V -Vpa¢-Am+_V-u<zpae a< auauaaguaga sass; < agate eaapm
ow cmurAPtpVpaarnet—aVrpawramtpmVru<z-m1~m+FaVru<zp~m ppau 4H2 gunman;
op mooxuosguauo googm
mp men oucpam
mp unauocguzgm noon
N mcvumou=_ mow
m Lao-“_+.V-a~a-aa+_mV-PauraaapaV-~ae-am+FaV-u<zpaa-am+PaV-u<zpaa caapaa amen; sauna; emanated

 

.eom oaauuacum oueaom eczemamm
omuugc_somouuepamquou<umra mo monogamazm puvucouos .9 «pack

12

Lao-AP+PmV-qu .ke.paV.Paa-flm+~aV-u<zFaw

-sraTzwééTzo
Hm.amV-na=a.AaPINV-VFaw-Am+FaV-a<zFaw
HH a_aapoa=a

o“ A V emu A V A V
- _+.m -a.w e+_a -_ao- m+_a -a<zpaw
-A¢+_QV-Fa¢\\\\
AmIAQV-.ao-he+PmV-Hu=a-AaP+~V-VFae-am+PaV-u<zpae

H asaepaaza
we Lao-”P+FV-aPu-me+pa~-Fae-Am+paV
Am+FaV-Ha=a-APrmV-Va<zan-A¢+PaV-masa-aa,+~V- Paw- m+paV-a<z_ae

venVPoozpm Famooaeva

aau-AP+FV-a_a
ea -Ae+.aV-_ao-A¢+PaV-Paw-fim+FaV-Ha=a-AaptuV-VPae-Am+.aV-u<z.ae
ou_gmzuuam~xm= ovvsacou

Lau-aPapaV-uPu-M¢+aaV-Pau-ae+eaV
-Paa-fim+~aV-a<zaFe-fie+paV-Ho=a-AaF+~V- Fae-am+PaV-a<zFam A=V

 

no emu-“P+PmVruporme+FuVrFomIA¢+PmV cenVP
-Fae-Am+PaV-a<zupu-am+.aV-ma=a-AaptuV- Pae-am+.aV-u<zFae AAV -amcegaaaua_o
ouremguummauao: onwsasou mucosa crewman no; < macaw voopm
.eom oaauuasum magnum ucaoasoo

Aea==Fu=auV _ apes»

 

meco~w_»ua_p.xos cease auoppwxas

 

 

pop new so seer au<zpmw ream och>o razm m=m>c
ms sum so schrao<zpew

. mucapa agapppxes cvozs zen—prxusnam

me H<zaz-A~+aV-V_au-am+PaV-Huza-naF+~V-V_a¢-am+~aV-a<z_aw -aaa a=_uaoa aeeuaaa r<

acreaguummomrpo

mu HozmrflapthVruporfietpmV-mernuzmrhF+~V1mequm+fiaVru<zpuw were: cuss: < macaw voopm

 

aau-apa-ae+pV-mu<zapa
-A_+eV-VPao-Am+eV-a<zuFe .xkataV-a<zupe-fie+FV-a<zaPo

-A¢+~V-u<zu_e-A¢+_V-_agH
mu m+FV-a<zupe-nu=a-aFtuV-VPao-Am+_aV-o<zpac
c rope .
nu -Ae+_V-Hu<za.w as aaa,_aaae xa_a2oo
-ApreV-Pao-AP+e oezapm
-uazupw-AF+mV- “Frag
-Paa- m+P aezupa
-a<za_a- 4+_V APtaere+sV-a<za.e-A¢+FV-Hu=a-APtNV-VFae-Am+FaV-a<zFaa
-aezapa-ae+_V-Paa.lrr
Am+PV-aezaF¢-A¢+PV-Paa-fim+FaV-a<z.aa
H aaa.Fou=a xapasau
aathe ,
-Ap+saV-aFa. area -Fao- cap -aezupa
so -A44eaV-paa\\
Am.+aV-Fae-fia+aaV-fia=a-AaP+~V-VFae-am+PaV-u<z_aa
“Ha a.aepoa=a

 

.mma 2:323 338 252.58
Aua==_»=ouV P «Pane

Metabolism of Neutral Glycosphingolipids
The glycosphingolipids are believed to be metabolized in a stepwise
fashion by a family of specific exoglycosidases each of which catalyzes
the hydrolysis of one sugar residue at the nonreducing end (86). The
degradation of some of the glycosphingolipid substrates of a-N:acetyl-
galactosaminidase is shown in Figure 1.
The sources and properties of some of the glycosidases are

summarized below.

afoAcetylgalactosaminidases

 

a-NrAcetylgalactosaminidases have been demonstrated in extracts of
Mchomnaa foetus (87,88), Helix pomatia (89), Busycon liver (90),
Aspergilluo niger (91), rat, porcine, and bovine livers (92) and rat
brain and kidney (93). A partially purified preparation of a-Nfacetyl-
galactosaminidase from human liver has also been characterized (94).
Cleavage of the Qfglycosidic linkage between Nracetylgalactosamine and
the hydroxy amino acids in ovine and bovine submaxillary mucins is also
due to a-Nracetylgalactosaminidases (95-101), which have been isolated
from ox spleen, Lumbricus terrestris, and rat liver and kidney. None
of these enzyme preparations has been purified to homogeneity and
consequently little information is available about the physical and
chemical nature of the a-N:acetylga1actosaminidases, nor of their sub-
strate Specificities. It has been noted that o-Nracetylgalactosaminidase
from bovine liver dissociates in dilute enzyme solutions (102); reasso-
ciated form of the enzyme has a molecular weight of 155,000 and the
monomer has a molecular weight between 30,000 and 42,000.

Using phenyl-e-N§acetyl-Q-galactoSaminide as substrate, it was

14

15

.mmauV:VEmmouummepxumu<umra mo mmaacpmaam cmawpomcwcnmouapo mo EmmFonmuo: .P acumen

  

 

mmaomra a.manmcx one Fauna
mcvmognocmuopumpmo
courpmw
mascara n.3uamm one figure
cmorpmorpmw
mmm xmzam

smoupmwupmwru<zpww
one u<zpoura

Amppou VHzV nouns: cmsmmgom
L3484812551253

aaanaspaa
9

Perprcmumumxo:
.9
arprpawoucpgflm
mcwmomcvsnm

*
Lou
ommomwo m.am;u=mw one upurm

ovwmogawgmooozpo
Lau-ape

one pmwrm

oursmcmupamouuug
gmurupmrpou

mmmmmwo m.zgaa one pours
amoru~orpaur—ew
«mm xmzrm

auVaoaope
Lau-upe-saw-_ae-a<z_au

own u<zpmwna

:manug :mEmmLom
coorupw-Peuru<zpeuru<z~am

own xoxru
cmurupwupawru<zu~m

one Fauna
Louuupw1-u1u<zu—wrpaw
mmm.u=u-JTa

one

= macaw coopm _
cmuropmrpmwru<zupwrpmw

one v<zpmwua

use
< qaoem voopm _
Laure—urpawru<zupwrFawru<zpaw

16

shown that the pig enzyme had a pH Optimum at 4.3 and an apparent KM of
6.4-6.6 TIM. N-Acetylgalactosamine was a competitive inhibitor with a K1
of 10.1 TIM. Copper sulphate and cysteine hydrochloride at 1 11M produced
10% inhibition. The beef enzyme had‘a pH optimum of 4.7 and a KM of
25 11M at 38°C. Crude preparations from rat brain and kidney were shown
to be able to hydrolyse Forssman hapten with a KM of 1.0 x 10"4 M and
3.5 x10"4 M at pH 4.4 with a taurocholate cancentration of 1.5-2 mg/ml.
The human a-N:acetylgalactosaminidase was found to be markedly thermo-
stable. It maintained its maximum activity at 50° for at least 4 hours.
The optimum for hydrolysing p-nitrophenyl-a-N:acety1galactosaminide
was at pH 4.3 with KM values of 3.5 mM for the liver enzyme. Both N:
acetylgalactosamine and galactose are inhibitors of this enzyme with a

Ki of 2.5 mM for the former monosaccharide.

B-NyAcetylhexosaminidaSes

 

B-NfAcetylhexosaminidase is the most studied glycosphingolipid
hydrolase because of its deficiency in Tay-Sach's and Sandhoff's diseases.
It has been shown to exist in four forms, A, B, C, and D. The A and B
fbrms have been purified to homogeneity from human placenta (103,104),
beef spleen (105), human plasma (106) and hen oviduct (107). Partially
purified fractions were also obtained from rat cerebral cortex (108, 109)
calf and human brain (110), human serum (111), rat testis (112), ram
testis and epididymis (ll3),equine kidney (114), porcine kidney (115),
rat kidney (116, 117) and bovine liver, spleen and kidney (118). Hexos-
aminidase C has been obtained in a partially purified form from human

placenta (119).
Human placental B-Q-Nfacetylhexosaminidase (103) A has a pI of

17

5.4 while.the B fbrm has a pI of 7.9. The A form is distinguishable
fran the B form in terms of heat lability, and the two forms can also-be
distinguished by immunological means.’ Both fOrms hydrolyse 4amethylum—
belliferyl-B-D-Nfacetylgalactosaminide with a pH optimum of 4.4, KM of
0.5 mM, and heat of activation of 10,500 cal (120). Cupric ion (5 mM),
p-hydroxymercuribenzoate (1 11M) and acetate ('30 11M) inhibit the hydroly-
sis of the synthetic substrate. The molecular weights of the hexosami-
nidases A and B are between 100,000 and 140,000. Hexosaminidase A
dissociates into one major subunit corresponding to a molecular weight
of 17,000 to 18,000, while hexosaminidase B dissociates in sodium
dodecyl sulphate into three subunit species with molecular weights of
,l7,000 to 18,000, 35,000 and 55,000 daltons. Urea-starch gel electro-
phoresis of the subunits suggests that hexosaminidases A and B may share
one common subunit. Both hexosaminidases A and B have blocked NHZ-ter-
minal groups and differ in the COOH- terminal amino acid, hexosamini-
dase A having serine as the COOH-terminal amino acid while hexosamini-
dase B has aspartic acid or asparagine. The two forms are somewhat
different in amino acid composition.

Geiger and Arnon (104) have shown that the human placental hexosami-
nidase A contains 1.65 residues of sialic acid while the B fbrm has none.
By reductiOn, alkylation and dissociation in 5 M guanidine hydrochloride,
the A form can be separated into 2 types of subunits by ion-exchange
chromatography, while the B form produces only 1 type of subunit. Based
on these data, it was postulated that hexosaminidase A is a tetramer,

a282, where a and 3 donate the subunit species, and hexoaminidase B is

composed of azaz tetramers.

18

The beef hexosaminidases were found to be similar to those of
human placenta, with optimum fer pNP-B-acetylgalactosaminide
hydrolysis at pH 4.5, KM values of about 0.9 mM andVmax of 30 umoles/
min/mg fbr both forms. The molecular weights were about 137,000 to
146,000 daltons, with the molecular weight of the smallest subunit at
56,000 daltons. The hen oviduct hexosaminidases had molecular weights
of 118,000 for the A form and 158,000 for the B form.

Hexosaminidase C has been purified to apparent homogeneity from
.human placenta (120). It is found in the cytosol, with an estimated
molecular weight of 190,000 and a pI of 5.7, compared to pI values of
5.0 and 7.3 for hexosaminidases A and B, respectively. (The pH optimum
fbr the hydrolysis of 4-methy1umbelliferyl-8-N:acetyl«Q-galactosaminide
was found to be 7.0, with a KM of about 0.83 mM. This fbrm is not
thermostabl e .

Three forms of 8-hexosaminidases denoted A, A' and B were purified
from equine kidney (114). GMZ ganglioside was hydrolysed by either A
or A' but not by B. On the other hand, globoside I was hydrolysed by
A' or B but not by A. When the Oligosaccharides of the glycolipids
were used it was found that A or A' hydrolysed 6M2 oligosaccharide but
not B while GHz and globoside 2 Oligosaccharides were hydrolysed by all
three components, A, A' or B.

A hexosaminidase C with only B-Nkacetyl-Q-glucosaminidase activity
has been shown to be present in bovine brain (121). It did not elute
from Sephadex G-200 or DEAE-Sephadex, but came off in the void volume
from Bio-gel P-200, indicating a molecular weight equal to or larger
than 200,000 daltons. Another form, denoted hexosaminidase D, was shown

to be present in bovine brain preparations. It was specific for

19

Beflracetylgalactosaminides. The report is in agreement to that of
Frohwein and Gatt (122) and Hooghwinkel at al. (123), which described
the presence of a BeflraCetylglucosaminidase and a B—Nyacetylgalactos-
'aminidase.

A partially purified fraction of BeNfacetylhexosaminidase from hog
epididymus was also shown to hydrolyse globoside I and asialo-GM2 while
no activity toward 6M2 was observed. The best detergent eXamined was
sodium cholate at an optimum concentration of 2 mg/ml. The pH Optimum
for the hydrolysis of globoside I was 4.2 and the KM was 3.7 x 10'3M
(124).

a—GalactosidaSes

 

a-Galactosidase was shown to occur in dog, human, ox, pig, rabbit
and rat (125). Ceramide trihexosidase (globotriaosylceramide hydrolase)
was purified from normal human plasma by affinity chromatography (126-
128). This enzyme has also been purified to near homogeneity by cone
ventional methods from human placenta (129,130). Partially purified
fractions have been obtained from human liver (131,132) and kidney (133).
Ceramide trihexosidase activity has been found in concentrated human
urine (134).

The plasma enzyme was shown to consist of two forms called A-1
and A-2. They had a molecular weight of about 95,000, with subunit
molecular weights of 22,000. The A fbrm had a K" value toward GL-3
of 4.5 x 10'4 M with 0.03 M sodium taurocholate and 0.15 M sodium
chloride. In the presence of optimal sodium taurocholate and sodium
chloride, the enzyme was inhibited competitively by digalactosylceramide
and GL-3 oligosaccharide. The A-2 fOrm had a KM of 5 x 10'4 M toward

GL-3, was inhibited by galactose, GL-2, mwoinositol and

20

digalactosylceramide but not by pNP-o-galactoside or 4~methv1umbellife-
ryl-o-galactoside (127). The human placental o-galactosidase also con-
sists of two forms, A and B (129). The A farm probably corresponds to
the A-l fonm and the B-fOrm may be equivalent to the A-2 farm of Mapes
and Sweeley. The A form had a molecular weight of 150,000, KM of 3.4 mM
towards 4~methy1umbe11iferyl-o-Q-galactopyranoside, and 40.6 mM for
melibiose. The 8 form did not hydrolyse melibiose. The pH optimum for
both forms was 4.5, with the A farm showing a broader pH activity curve.
The A form was inhibited by myoinositol, while the 8 form was not.
The isoelectric points were at pH 4.7 and 4.42 for the A and B forms,
respectively. Antibodies towards these two placental farms of o-galac-
tosidase did not cross react.

The human liver a-galactosidase hydrolysed GL-3 with a pH optium
of about 3.5, was activated by a mixture of sodium taurocholate and
Triton X-100 and was inhibited by high concentrations of bovine serum
albumin. It had a KM of about 5 x 10"5 M (131) towards GL-3. The human
kidney o-galactosidase had pI values of 4.5 and 4.3 fbr the A and 8
forms, respectively. In agreement with Beutler and Kuhl (130), it was
Shown that the A form was thermolabile while the 8 form was thermostable.
The KM values for the A and 8 forms were 6.0 x 10-3 M and 13.3 x 10"3 M
towards pNP-o-Q-galactopyranoside and 5.7 x 10‘4 M and 4.7 x 10'4 M
toward GL-3, respectively. The activity of ceramide trihexosidase was
shown to be depressed by high concentrations of protein (133). The
thermostabilities of the human liver o-galactosidase B (130,131) and
oi-N-acetylgalactosaminidase ('94) in addition to the inhibition of human
liver afiNfacetylgalactosaminidase by both Nyacetylgalactosamine and

galactose support the hypothesis of Dean, Sung and Sweeley (135) that

21

oegalactosidase B and o-Nyacetylgalactosaminidase in human liver are the
same enzyme.

It was suggested that neuraminidase treatment converted the A form
into a mixture of 14 enzymatically active products as revealed by iso-
electric fbcussing, while a crude kidney sialyltransferase preparation
could be used to incorporate CMP-[l-14C]-N§acetylneuraminic'acid or
UDPaNyacetyl-[1-]4C]-glucosamine into the B form, resulting in the
fbrmation of a fbrm Similar in electrophoretic behaviour to the A fOrm
(136). This observation was disputed by Romeo et at. (132) who showed
by the combined use of isoelectric focussing, DEAE-chromatography and
enzyme kinetic parameters that no conversion of o-galactosidase A into
B or vice veraa could be detected after neuraminidase treatment.

In Fabry's disease, only the B fbrm was found, accounting for
about 2 Z of the normal 4-methy1umbelliferyl-o-Q-galactopyranoside
hydrolysing activity in the kidney and 20% of the normal activity in the
liver (137, 138). Fibroblast cultures from patients with Fabry's
disease also exhibited the presence of the B fbrm while the A form was
absent (139): the residual activity accounted fbr 10-30% of the normal
activity towards the artificial substrate-(140)..

B-Galactosidases

The action of B-galactosidase on four natural glycosphingolipid
substrates has been examined. The substrates are lactosylceramide,
galactosylceramide, GMl-ganglioside and asialo-GMl-ganglioside.
There is evidence that two of these B-galactosidases have different
substrate Specificities towards the glycosphingolipids. One fOrm is
Specific for galactosylceramide and the other for GMIeganglioside,

while both fbrms can hydrolyse lactosylceramide under certain assay

22
conditions. Suzuki and his group have examined the B-galactosidases
of rat brain and human liver.

The human liver preparation, purified 250-fold, was active towards
CHI-ganglioside, asialo-GM1-ganglioside, and lactosylceramide but in-
active towards galactosylceramide when the Suzuki assay system was used
(141). The enzyme preparation was stimulated by chloride ion and
unstable upon dilution. When a liver preparation was subjected to
isoelectric focussing, three B-galactosidase peaks, denoted a, B,and 1,
were resolved; their pI values were at pH 4.1-4.2, 4.5-4.6 and 4.8-4.9,
respectively. Gel filtration also gave three peaks of activity, the
second of which corresponded to the B fbrm and the third to the 7 form.
Lactosylceramidase activity was present in all three peaks, a, B, and 7,
while only a and y had CHI-ganglioside and galactosylceramide hydrolyo
sing activity (142). In globoid cell leukodystrophy, a disease with
accumulation of galactosylcermide, isoelectric focussing patterns of
tissue B-galactosidase activities toward lactosylceramide and asialo-
GMl were normal, while the 4-methy1umbelliferyl-B-Q-ga1actopyranoside
and galactosylceramide hydrolysing activities existed as a single, broad
but diminished peakin the range of pH 4.6 to 4.8 (143). There were
two types of isoelectric focussing patterns for B-galactosidase in
tissues from patients with GM] gangliosidosis. One had a shifted p1 to
a lower pH for the two galactosylceramide p-galactosidase activity
peaks while in the other type these peaks were normal. In both types
of 6M1 gangliosidosis, there was no residual asialo-GMI-ganglioside
hydrolysing activity and only one 4-methylumbe1liferyl-B-galactosidase
peak that appeared between the B and y ferms (144). The above data
with globoid cell leukodystrOphy and G“1_gangliosidosis did not give

23
definitive results as to the nature of the.deficiency. It was hinted

that there may be structural and fUnctional interrelationShips between
galactosylceramide- and GHl-ganglioside— B-galactosidases, perhaps
similar to but more complex than the structures of the human placental
hexosaminidases, which were ozsz'for the A form and 8282 for the B
fbrm (104). Rat brain B-galactosidase was partially purified and shown
to have the following properties: pH optimum 3.9-4.2, 4.9, 3.9 and 4
fbr 4-methy1umbelliferyl-B-ga1actopyranoside, lactosylceramide, asialo-
6M1 ganglioside and 6M1 ganglioside hydrolysing activities, reSpectively.
The optimal taurocholate cOncentration was 1.5 mg/ml for the asialo-GH1
ganglioside and GMl-ganglioside B—galactosidases. The apparent KM at
optimal activities were 2.2 x 10'5 M, 2.0 x 10'4 M and 4.3 x 10'4 M for
lactosylceramide-, asialo-Gm ganglioside- and GMI-ganglioside- B-ga-
lactosidases. Huge amounts of Oligosaccharides were required to cause
appreciable inhibition of asialo-GNI-ganglioside- and GMl-ganglioside-
B-galactosidases (145). A rat brain preparation also hydrolysed galac-

5 M and pH optimum of 4.2 to 4.5

5

tosylsphingosine with a KM of 1.1 x 10'
(146). Human brain preparations gave a KH of 2 x 10' M fbr both
galactosylceramide and lactosylceramide (147). A GMl-ganglioside-B-
galactosidase A was purified to homogeneity from human liver (148). It
had a molecular weight of 65,000 to 75,000 and 72,000 by SDS gel fil-
tration of the reduced and carboxymethylated enzyme. This enzyme also
cleaved B-D-fucoside and o-L-arabinoside linkages. H0 et a2. isolated
two fbrms of Gal-ganglioside-a-galactosidases, one of which was membrane
bound and the other soluble (149). The apparent KH was 2.8 x 10"5 M and
' 7.7 5: '10'5 n for the two forms, respectively. (Sm-ganglioside-B-galac-
tosidase was partially purified from rabbit brain. It eluted in one

24

single peak from Sephadex G-200. The pI was.at.pH 6.3. With GM]-
ganglioside as substrate,.the pH optimum was 4.3 and the KM 7.8 x 10‘5M
while with lactosylceramide the pH optimum was 4.5 and the KM was 1.7

x 10"5 M (150). y-Q-Galactonolactone, lactose and galactose (151) were
competitive inhibitors of the hydrolysis of the natural substrates.

Rat and pig brains were shown to contain galactosylceramide—B-galactosi-
dase activity (152). The activity was activated by the Nydecanoyl-
derivative of 2-amino-sZ-met1wl-l-propanol by 34% at 0.15 TIM cOncentra-
tion (153) while Nfdecanoyl-pL-erythro-3-phenyl-2-aminopropanediol
inhibited competitively with a K1 of 0.4 mM (154). A lactosylceramide-
B-galactosidase that did not exhibit any GMI-ganglioside or galactosyl-
ceramide hydrolysing activity was also purified from porcine thymus
tissue (155). Two activity peaks could be resolved by isoelectric
fbcussing, with pI values at pH 6.3 and 7.0. The pH optimum was at 4.6
and the apparent KM was 2.3 x lO'SM.

GlucOcerebrosidase

Glucocerebrosidase activity was demonstrated to be present in
spleen (156). The enzyme consisted of a heat labile, membrane bound
factor C and a heat stable soluble acidic glycoprotein called factor P.
In the absence of factor P, factor C could be activated by sodium
taurocholate to hydrolyse glucocerebroside (157). The activation
of factor C by factor P also required the presence of acidic phospho-'
lipid to give an active enzyme (158). A model was postulated in which
the effector (factor P) and catalytic protein (factor C) were depicted
to combine in the presence of 0.15 an acidic phospholipid in a ratio of
1:1 at low effector concentration, and exist in an equilibrium of ch’

PC and P when the effector was in excess (159).

25
.The.enzyme has also.been purified.to apparent.homogeneity from
human placenta. It was.composed of four catalytically active subunits

514

of 60,000 daltons each. The KM for glucosylceramide was 6.5 x 10'
at an optimal pH of 5.0 (160).
Deficiency of gluCocerebrosidase activity was shown to be

characteristic in Gaucher's disease (161-163).

In Vivo'Metab01ism of Glycosphingolipids

The glycolipid hydrolysing enzymes have been shown to be lysosomal
in origin (86). They have pH optima in the range of 4-5, which is known
to be characteristic of lysosomal enzymes. It is believed that lyso-
somal hydrolases are synthesized on membrane-associated polysomes and
gain access to the endoplasmic reticulum cisternae, through which they
are transported to other areas of the cell (164, 165). Primary lysomes
package the lysosomal hydrolases and eject them into digestive vacuoles
to fbrm secondary lysosomes. They are thought to originate either from
the Golgi apparatus or GERL, a region of the smooth endoplasmic reticu-
lum that has an intimate structural relationship with the innermost
Golgi element. In the neurons of dorsal root ganglia the innermost
Golgi element is composed of anastomosing tubules that enclose polygonal
compartment, in each of which a tubule of endoplasmic reticulum,
probably GERL, is found. There are three ways that lysosomes digest
endogenous material. This process is known as autophagy. The primary
lysosomes fuse with intracellular vacuoles and the hydrolases are
released into the vacuole form a so called type 1 vacuole. In another
type of autophagy. vacuole 2 appears to form by enlargement and twisting
of GERL. The membrane internalizes to form inner tubules as the vacuole

enlarges, enclosing bits of cytoplasm in the process. The third process.

26
known as.microautophd9y. is believed to be occurring in a similar manner

as the type 1 vacuole, only at a much lower scale and with less easily
identifiable manifestations. Exogenous materials can also be digested
in a similar manner of fusion with the primary lysoSomes. They enter
the cells by either phagocytosis or pihocytoSis. All the above digestive
vacuoles are known as secondary lysosomes. They can also fuse with
other Vacuoles to form another Secondary lysosome. When the process of
digestion is completed, electron opaque materials are,left enclosed in
the original lysosomal membrane. These structures are then called re-
sidual bodies.

The process of digestion of glycosphingolipids may be similar to
the endocytosis processes cited above. The glycolipids may be digested
as part of a membrane fragment, or by association with certain lipid
carrier proteins. The detergent requirement of hydrolysis of glyco-
sphingolipid in vitro may mimick the hydrolysis of membrane associated
glycosphingolipids in viva.

An activator fur the hydrolysis of glucocerebrosidase was found
in human spleen (166). It is a heat stable glycoprotein and associates
with the catalytic subunit of the enzyme in the presence of phospholipids
to form an enzymatically active entity (157,158). It was contended
that a similar activator from human spleen was present in molar quanti-

ties 0f 001! 1/25 0f the amount of gludocerebrosidase. The presence of
an eight-fold in excess of the physiological amount did not result in

activation of glucocerebrosidase activity. An activator that can sub-
stitute fbr detergents was also demonstrated (l68, 169). A heat-stable
protein factor that activates the hydrolysis of cerebroside sulphatase
was purified 2,000-fold and found to have a molecular weight of 21,500

27
with the pI at pH 4.3. The activating effect of this protein was
about half as effective as taurodeoxycholate at their Optimal concentra-
tions. The amount of activator'protein required for‘maXimum enzyme
“activity was however only 1/300 that of taurodeoxycholate in molar
amounts (170). It is likely that activator proteins along with phos-
pholipid may play significant roles in the ia viva metabolism of glyco-
sphingolipids.

Enzyme Therapy of‘Glycosphingolipidoses

In cases of genetic defects where certain acid hydrolase activities
are absent, glycosphingolipid accumulation in residual bodies results.
A whole family of such disease called sphingolipidoses has been dis-
covered (7). The strategy for alleviating the symptoms is to purify
the enzyme and administer it in a form that would result in minimal
immune responses and yet active in hydrolysing the accumulated glyco-
sphingolipids. The packaging of the enzyme in lysosomes and the coating
of the lysosome surface with organ specific marker molecules may be a
plausible approach to enzyme replacement therapy. Pioneering work in
genetic engineering to introduce B-galactosidase genes into GMl’
gangliosidosis fibroblasts were also done (171). Such work should be
approached with prudence because any unforeseeable mistake would result
in the irreparable damage to the genetic make-up of the individual.

Animal models for studying the physiological impairment and its
possible cure are used. Numerous reports of animal gangliosidoses have
appeared (172). Artificial induction of lipidoses by injection of’gly~
cosphingolipids into small animals were successful in producing cells
thathavelipid containing residual bodies (173-175). The addition of

anti-lysosomal enZyme antibodies to cell cultures also resulted in

28

dramatic increases in the.number and size of residual bodies (176, 177).
The inhibitory effect of flfheXylfigfglucosylsphingosine0n gluCocere-
brosidase was used to produce cell strains that exhibit the accumulation
of gluCocerebroside (168,178,179). TheSe cell cultures can also be very
useful in studying the physiological effects of glycolipid acCumulations.

The studies of storage diseases have made possible the early
diagnoses of these diseases by amniocentesis. Tay-Sachs disease was
diagnosed by hexosaminidase A deficiency in amniotic fluid and cells
(180). This represents a great step forward in combating inherited
sphingolipidoses.

Glycosidases in Transformed Cells

The role of glycosidases in transformed cells is not well under-
stood. Their changes in activities may be a primary trigger of the
transformed state, or more likely, one of the secondary events that
sets in after the cell is transformed from the normal state. fifAcetyl-
B-glucosaminidase and B-glucuronidase were increased in platelet-free
plasma from patients with malignancies (181). B-Glucosidase and fl;
acetyl-B-glucosaminidase activities were likewise increased in oncogenic
virus-transformed fibroblasts (182). Glycosidase levels of polyoma
virus-transformed BHK cells and murine sarcoma virus- and Rous sarcoma
virus- transformed 3T3 cells were also shown to be increased over the
levels in normal cells (183-185). Elevated levels of proteases and
glycosidases were found in human breast and colon tumours (186). They

are thought to be released to modify the cell surface so that the

transformed state can be maintained.

MATERIALS AND METHODS

”Materials

The following solvents were of reagent grade: methanol, chlorofomm,
pyridine, dimethylformmmide, toluene,acetic anhydride, dichloromethane,
cyclohexane, l-pentanol, l-propanol, 2-propanol, benzene, l-butanol,
'acetic acid, acetone, hexanes, tetrahydrofuran (Nallinckrodt. St. Louis.
"0.); dimethylsulphoxide, l-octanol, 1,2-dichloroethane (Fisher, Fair
Lawn, N. J.); acetonitrile (Aldrich, Milwaukee, His.). All solvents
were redistilled before use.

Chemicals and their sources are as fellows: NaH from Alfa Inorganics
(Beverly, Mass.); G-150, 6425, LH-ZO and Sepharose 48 from Pharmacia
(Piscataway, N. J.); 02-52 from Nhatman (Chifton, N. J.); DEAE-cellulose,
Bio—bead P-2 from Bio-Rad (Rockville Center, N. Y.); silicic acid and
hydroxylapatite from Clarkson (Hilliamsport, Pa.); N:acetylgalactosamine,
a-methyl glucoside, a-methyl mannoside, glucosaminic acid, galactose,
N:acetylglucosamine, glucose, glucosamine hydrochloride, galactosamine
hydrochloride from Pfanstiehl (Naukegan, 111.); nitrosyl chloride adduct
of-galactal from Raylo (Edmonton, Alberta, Canada); percine submakillary
glands from Pal-Freez (Rogers, Ark;); ampholytes from LKB_(Chicago,
111.); Concanavalin A and chondrosine from Males '(Kankakee, 111.);
sodium taurocholate from Calbiochem (La Jolla, Ca.); silica gel G TLC
plates fromAnaltech (Newark, Del .); [3HJ-NaBH4, ‘and‘acetyl-l-[uc]-
'acetic anhydride from New England Nuclear (Boston, Mass.); galactose

29

3O

oxidase andhorseradish peroxidase from Worthington (freehold, N. J.);
molecular weight calibration'proteins from Boehringer-Nannheim' (In-
dianapolis, Ind.); anti-A and anti-B‘antisera from Ortho Diagnostics
(Raritan, N. J.); Bio-Solv BBS-3 and'scintillation grade'toluene from
Beckman (Fullerton, can; PPO and dimethyl-POPOP from Research Product
International '(Elk Grove Village, 111.); 'p-nitrophenyl-o-N-acetyl-
galactosaminide, p-nitrophenyl-8-N-acetylgalactoSaminide, 'p-nitrophenyl-
o-N-acetylglucosaminide, and p-nitrophenyl-B-fl-acetyl-glucosaminide
from Koch Light (Colnbrook, Buckinghamshire, England); 3% SE-30 on
SupelcOport (80-100 mesh), 3% SP-2100 on Supelcoport '(80-1‘00 mesh),

0.2% EGS, 0.2% EGA and 1.4% XE-‘60 on Gas Chrom 0 (80-100 mesh), 3%
ECNCSS-H on Supelcoport ‘(80-100 mesh), 3% 011-210 on Gas Chrom Q (80-

100 mesh) from Supelco (Bellefonte, Pa.); mercaptoethanol , and borane-
tetrahydrofuran from Aldrich (Milwaukee, Nis.); sodium borohydride,
Dowex 1, Dower 50, galactose, glucose, fucose, melibiose. stachyose,
raffinose‘, N—acetylneuraminic acid, hexamethyldichlorosilazane, tri-
methylchlorosilane, trifluoroacetic acid, neuraminidase, cetyltrimethyl
annionium bromide, tris-(hydromiethyl)-aminomethane, glycine, o-naphthol,
bromophenol blue and chitin from Sigma (St. Louis, No.); dialysis
tubings from A. H. Thomas (Philadelphia, Pa.); iodomethane, cyanogen
bromide, phenylmethyl sulphomyl fluoride and sodiun cyanoborohydride
from Pflatz and Bauer (Stamford, Conn.); 90% nitric'acid, nitric acid,
bromine, amoniun persulphate and orcinol from Fisher (Fair Lawn, N. J.);
acrylamide, and _N_, [if-methylene bisacrylamide from Canalco (Rockville,
Nd.): LL [1' ,_l_i_'-tetramethyl-ethylenediamine and Photo-Flow '600 from
Eastman Kodak (Rochester, N. Y.); silver carbonate, phenol, citric acid,
EDTA, urea, and sodium citrate frcmi Nallinckrodt (St. Louis, 140.);

31
.xylene brilliant cyanine G from ICN Pharmaceuticals (Cleveland, Ohio);
100% ethanol from Cdmmericial Solvents (Terre Haute, Ind.); Hydrochloric
'acid, sulphuric acid and iodine, from Baker (Philipsburg, N. J.); and

Folin Ciocalteu reagent from Harleco (Phila., Pa.).

‘”Neth0ds
(I) Substrate Preparation

(1)Aforssmanflapten

 

Forssman hapten from dog intestines was purified by the method of
Saito and Hakomori (187). Dog intestines (6 Kg) were rinsed in dis-
tilled water, cut up into small pieces with scissors, and homogenized
with 12 liters of chloroform/methanol 2/1 in a Haring blender. The
homogenate‘was filtered‘with a Buchner funnel. Homogenization and fil-
tration were repeated with 12 liters each of chloroformvmethanol 2/1,
111 and 1/2. The pooled filtrate was evaporated to dryness in vacuo on
a rotary evaporator. The lipid was dried by evaporation of traces of
water with 100% alcohol and the residue was acetylated overnight in a
mixture of 450 ml of dry pyridine and 300 ml of acetic anhydride. The
acetylation solution was evaporated to dryness in uaeuo with repeated
addition of large amounts of toluene. The acetylated lipid was loaded
onto a 1000 g Florisil column (100 x 7.5 cm) packed with 1,2-dichloro-
ethane (DCE). After washing the column with ten column volumes of DCE.
the acetylated glycolipids were eluted with l0 column volumes of DCE/
'acetone (l/l). The eluate was evaporated to dryness in dunno and
dissolved in 500 ml of chloroform/methanol 2/1. A solution (250 ml) of
0.5% sodium methoxide in methanol was added and the mixture was allowed

to stand at ream temperature for several hours. A large volume of

32
ethyl acetate was added_to.neutralize the base and the.solution was

evaporated in vacuo .. The'de-Q-acetylated glycolipid was then dispersed
in water by sonication, dialysed overnight against distilled water, and
lyophilized. The lyophilized glycolipids were dissolved in a small
volume of chlorofonh/methanol 2/1 and loaded’onto a 100 g DEAE-cellulose
column (70 x 4.5 cm) prepared in chloroform. The neutral glycolipid
fraction was eluted with 10 column volumes of chloroform/methanol (7/3),
evaporated to dryness, and loaded onto a 1.2 Kg silicic acid column

(100 x 7.5 cm) packed in chloroform. A gradient of 4 liters of chloro-
form and 3 liters of methanol was used to elute the column. Fractions
of 20 ml were collected. The glycolipids were monitored using TLC with
the solvent system chloroform/methanol[water 100/42/6. The Forssman
hapten (GL-S) fractions were pooled into 5 fractions and evaporated to
dryness, the residues were dissolved in known volumes of chloroform/

methanol 2/1. About 1.8 g of crude GL-5 was thus isolated.

“(2 'Labellin fof GL-5 with the Galactose Oxidase-Tritiated Sodium
Borohydride Method (Figure 2).

About 100 mg of crude GL-S was purified with preparative thin layer
chromatography using 500 micron plates developed with solvent system
chloroform/methanol/water (100/42/6). The GL-5 bands were visualized
with iodine vapor, scraped off and eluted with 20 volumes each of chloro-
form/methanol 2/1, 1/2, methanol and finally chlorofdrm/methanol/water
'50/50/15. The pooled eluate was dried in am, dissolved in 10 ml of
chloroform/methanol 2/1 and washed with 2.5 ml of 20: sodium thiosulphate,
after which the lower phase was washed three times with water, and dried
under nitrogen. The glycolipid was labeled with galactose'oxidase-sodium
borotritiide’according to the method of Suzuki and Suzuki with

33

mm>vue>mema mum ecu midwinzmH mo :ovueeenmee .N «Lam—e

.u‘uoe-Joio a... 04>0008.xta nabhuo¢onov.o. u as.

4w.

88
Pun.~xo.xo.xo:woxwuxuo

casing-nu .2

:c 3 a
x. :o :3

nxuamzo.xu.xuzwxmuxoo§

gems»

 

gonzo .. ousuznfiz

 

ecu-ea
.3323 on; .2
cc «1 x c
.. x me ~.... «.5 M
.. o nxua.~..o...o...ox .3339 o o o n..o~..~..8..o...o.. so»:
unu- E e e .4 3393:
o z. m...» .. x .

{guise 29:28:58..

Q id.

34
modifications.(188). The purified GL-S'was dissolvedin 4 ml of 0.1 M
phosphate buffer, pH' 7.0. Approximately 330 units of h0rser'adish
peroxidase (3324 I.'U./mg) and 0.5 ml galactose oxidase (854 uhitS/ml)
were added and shaken for 4 hours at 37°C and 0.5 ml more of the
galactose oxidase solution was added. The incubation was perfdrmed
overnight. The oxidation reaction was terminated with '20 ml of
chlorofonh/methanol 2/1. The lower phase was dried and the GL-5 dis-
solved in 5 ml of tetrahydrofuran with 5 ul 5 N potassium hydroxide.
The oxidized GL-S was reduced with 2.3 mg NaBT4 (14 mCi) overnight. The
excess [3HJ-Na8ll4 was destroyed with several drops of glacial acetic
acid added in a fume hood. The solution was dried under nitrogen and
the residue was partitioned in 10 ml chloroform/methanol/water 8/4/3.
After the lower phase was dried under nitrogen, '30 mg of NaBH4 in 3 ml
water was added to the GL-S to ensure complete reduction. After
allowing to stand overnight at 4°, the excess NaBH4-was again decomposed
with several drops of glacial acetic acid. Solvent (12 ml of chloro-
form/methanol 2/1) was added and the lower phase was removed and washed
once again with 6 ml of theoretical upper phase. The [3H]-GL-5 was
further purified by TLC developed with chlorofonn/methanol/water 65/
45/8. The GL-5 was visualized with iodine vapor, scraped, eluted with
chloroform/methanol 2/1, 1/2, methanol and chloroform/methanol/water 50/
50/15 as before. The eluate was dried, dissolved in ‘10 ml chloroform]
methanol 2/1, washed with 20% Na25203. twice with theoretical upper
phase, and dried again. The [3MJ-GL-5 product was dissolved in 325 m1
’ chloroform/methanol. 2/1 and quantitated by GLC analysis of methanolysis
products. The '48 mg of 'I3Mj-GL-5 thus obtained was diluted with cold
purified GL-S to a final yield of 105.5‘pmoles of diluted '[3H].si.-5

35
with a specific.activity of 446,000 cpm[pmole. To find.the specific
activity of the non-redUcing terminal N-acetylgalactosamine residue
of the [3HJ-GL-5,.150 nmoles of GL-S was hydrolysed with excess purified
porcine liver o-N-acetylgalactoSaminidase until the release of radio-
activity counts levelled off. This count was taken as the total
count on the terminal N-acetylgalactosamine of [3HJ-GL-5. The specific
‘activity (terminal residue) was thus calculated to be 360,000 cpm/umole.

(3).Synthesis of GalNAc-(3146)-l,2:3,4-di-Q-isopropylidene-galactose

 

and TalNAc-(ol+6)-l,2:3,4-di-Q-isopropylidene—galactose

 

N-Acetylgalactosamine-(d1+6)-l,2:3,4-di-Q-isopropylidene-galactose
was synthesized by the method of Lemieux, James and Nagabhushan (189).
The reaction is shown in Figure 3. Dimeric tri-Q-acetyl-2-deoxy-2-ni-
troso-a-p-galactopyranosyl chloride (13.0 g) was coupled with 8.0 g of
l,2:3,4-diHQ-isopropylidene—galactose in dimethylformamide as described.
After working up the mixture, 12.0 g of 6-9-(tri-Q-acetyl-2-oximino-o-
Q-lyxo-hexopyranosyl)-l,2:3,4-di-Q-isopropylidene-o-Q-galactopyranose
was obtained; m.p., l7l-l72°. The compound (20 mmoles) was acetylated
and reduced in borane-tetrahydrofuran, acetylated and de-Q-acetylated.
The reduced disaccharides were applied to a 80 x 3.8 (i.d.) cm column
with 453 g of Dowex l-X2 (200-400 mesh, OH' form) which was prewashed
with l N NaOH, water, acetone and l N_HC1. Elution was performed with
water, and 18.4 ml fractions were collected. An aliquot of each frac-
tion (10 pl) was spotted on a TLC plate and sprayed with 2% o-naphthol
(in ethanol) and sulphuric acid. The carbohydrate-positive fractions
were analysed by GLC after methanolysis. Fractions 44 to 51 were pooled
to give GalNAc-(ol+6)-l,2:3,4-di-isopropylidene-Gal and fractions 57
to 107 for TalNAc-(ol46)-l,2:3,4-di-isopropylidene-Gal. The pooled

36

 

(DAc
| H
AcO
.4)

£1, +

(lllAcéo

(ZlNoOMe

H
E aNiAc
H -4)
c
GolNAc(al-6) di lP-Gol TalNAc“! l-GldiIP-Gol

Figure 3. Synthesis of N-Acetylgalactosaminyl- -(0146)- 1,2: 3,4-di-0-iso-
$.8ylidene-Galactose and N-AcetyltaloSaminyl- (31+6)- 1,2: 3,4-
_-isopropylidene-Galactose.

37

fractions were lyophilized and recrystallized, the farmer from ethyl
'acetate-hexane and.the latter from water. 'The yield of the GalNAc
dissacharide was 0.85 g. The ratio Galactose:N-acetylgalactOSamine was
”1:0.91. The isopropylidene groups were hydrolysed off by trifluoro-
'acetic acid. TalNAc-(ol+6)-l,2:3,4-di-Q-isopropylidene-galactose was
obtained from Dowex l-X2 dolumn as a by product. The yield was 2.97 g.

(4) Preparation Of'A+-Porcine Submaxillary Mutin

Blood group A activity of individual glands was determined by the
haemagglutination inhibition reaction (64). Small pieces of the glands
were minced in 100 pl of saline and dilutions of the homogenate
was allowed to react with 4 haemagglutination units of anti-A antiserum
for 30 minutes, followed by addition of 50 pl of 2% washed human A+-
red blood cells in isotonic saline solution. Haemagglutination inhibi-
tion was read after some wells showed a tight button of red cells.

A+-Porcine submaxillary mucin was extracted from the pooled A+-
glands by a method described previously (72). Pooled Af-glands (446 g)
were used and the total amount of lyophilized major mucin isolated
was 4.7 g.
(5)’Is01ation'0f‘A+-Glycosphingolipid'fromifluman;Red‘BlOod’Cell‘Membrane

Twenty-four units of outdated A+ human red blood cells were cen-
trifuged at 5,000 x g to sediment the red blood cells, which were then
lysed with an equal volume of 0.2% acetic acid. The packed red cell and
membrane fractions were homogenized with 3 liters each of the following
solvents, chloroform/methanol 2/1 (twice), 1/1, 1/2 and then refluxed
wdth 3 liters methanol overnight. The pooled solvents were dried in
flacua and the.glycolipids were acetylated with 150 ml of acetic

38

anhydride/pyridine 1/2 at room temperature overnight.. The solution was
dried with the aid of toluene and the residue applied in DCE to a 100 9
column of Florisil packed in DCE. After washing with 1 liter of DCE,
the glycolipids were eluted with 1 liter of DCE/acetone l/l, dried,
dissolved in 50 ml chloroform/methanol 2/1, deacetylated with 10 ml 2%

sodium methoxide, dried_again and dialysed in water. The lyophilized
glycolipids were then loaded onto a 5 9 column of washed DEAE-cellulose
packed in chloroform/methanol 7/3, and eluted batchwise with 200 ml each
of chloroform/methanol 7/3, 1/2 and methanol. The majority of the blood
group Af-glycolipids were obtained in the chloroform/methanol 1/2
fraction. The glycolipids in this fraction were separated and purified
by TLC using the solvent system chloroform/methanol/water/acetic'acid
55/45/5/5. Blood group A activity of the individual bands was deter-
mined by haemagglutination inhibition. The carbohydrate composition of
each of the glycolipids was determined by GLC after methanolysis.
'(61‘Preparation‘of'[3H1-GL-5-Pentasaccharide (Figure 2)

A solution of 2.8 pmole of [3H]-GL-5 in 2 ml of methanol was ozono-
lyzed in a Supelco ozonolyzer. The evolving gas from the glycolipid
solution was monitored by a 2% solution of KI. when the K1 solution
turned brown, the reaction was stopped and the GL-5 solution was eva-
porated under nitrogen. One milliliter of 0.2 N_Na2C03 solution was
added and the mixture was left at room temperature for 14 hours. After
neutralization with acetic acid and evaporation of the solvent under
nitrogen, the residue was partitioned in 5am] of chlorofdrmeethanol/
water 8/4/3. The lower phase was washed with 2.5 ml of theoretical
upper phase.. The pooled upper phase was dried and applied to a 50 ml
Bio-Gel P-2 column (1 x'60 cm) packed in water. The fractions

39
(0.83 nil/fraction) were monitored with radioactivityand TLC with

solvent systembutanol/acetic acid/water 100/50/50 using galactose,
melibiose, raffinose and stachyose as reference compounds. A yield

of 1.17'pnioles of [3HJ-GL-5-pentasaccharide was obtained.

. (7) . Preparation . ofN-Il-‘MCJ-Acetyl -S'ph'i_ngosyl -Fors man-Pentasaccharide

 

Forssman hapten (354 mg) was deacylated by the method of Taketomi
et oz. (18). The glycolipid was dissolved in 30 ml of 90% aqueous
butanol containing 1 N KOH and the solution was refluxed for 2.5 hours.
The solution was then dialysed for 2 days evaporated in vacuo with the
aid of absolute ethanol, dissolved in 3 ml chlorofomn/methanol 2/1 and
centrifuged. The supernatant fraction was poured into '30 ml of‘acetone.
The precipitate was collected by centrifugation at 1,200 x g for 3 -
minutes and dissolved in a known voltmie of chloroform/methanol 2/l.

Deacylated GL-S (30 mg) was dissolved in‘ 4 ml of dry methanol.
Acetic-[i-“cJ-ahhydride (5.8 mg, 4.4 mCi/nfl) in 1.5 ml of thiophene-free
dry benzene was added to the solution and the mixture was left at room
temperature overnight. After evaporation of the acetic anhydride
under nitrogen in a fume hood, the glycolipid was analysed by radio-
scanning TLC. Complete acetylation was achieved with the addition of
0.5 ml of acetic anhydride to the glycolipid dissolved in 5 m1 of
methanol and allowed to stand at room temperature overnight. A yield
of 11.2 mg of N-[l-'“C]-acetyl-sphingosyl-GL-5-pentasaccharide was
obtained. It was further purified by preparative TLC using the solvent
system butanol/acetic acid/water 1’00/50/50.

(II) Enzyme Assays

(l) Assay with_Forssman Glycolipid

Forssman glycolipid was assayed by dissolving 150 nmoles of [3H]-
glycolipid in 50pl of sodium taurocholate (12 mg/ml) with sonication.
To this solution were added 50 p1 of 0.3 M sodium citrate buffer, pH
3.9, and 50 pl of enzyme solution; incubation was for 30 minutes at 37°.
The reaction was terminated by adding 4‘m1 of chloroform/methanol 2/1
and 0.85 ml water. After thorough mixing, the samples were centrifuged.
The lower phase was pipetted out and the upper phase was washed with
2.5 ml of theoretical lower phase. The upper phase was then pipetted
into scintillation vials, dried and counted in 0.5 ml of water and
10 ml of scintillation Cocktail (7 g PPO, 0.6 g dimethylPOPOP, 100 ml
Bio-Solv BBS-3 and 1 liter of toluene).

(2) Assay with p-Nitrophenyl-o-N-.acetylgalactOSaminide

 

The assay mixture contained 100 pl of 2 mM p-nitrophenyl-o-N-acetyl-
galactosaminide, 50 pl of 0.3 M sodium citrate buffer pH 4.3 and 50 pl
of enzyme solution. After incubation at 37° for 15-30 minutes, the
reaction was terminated with 3 ml of 0.6 M potassium borate pH 10.4.

Absorbance was read at 410 nm.

(3) Assay withyGL-S-pentasaccharide aggrolysis

The incubation mixture contained in a final volume of 25 pl, 25
nmoles of GL-5-pentasaccharide in 5 pl of 0.3 M sodium citrate buffer,
pH 3.9 and 10 pl of enzyme solution. The solution was incubated at 37°
for 30 minutes and terminated by boiling in a water bath for 3 minutes.
The samples were spotted on silica gel G plates with 100 pg each of

_ galactose and stachyose standards on the same plate. After
40

41
chromatography in.butanol/acetic acid/water 100/50/50, the.standards
were Sprayed with orcinol-sulphuric acid and heated at 100° fer 5-10
‘minutes until the carbohydrate spots were visualized.. The area corres-
ponding to the monosaccharide region were scraped off and counted
directly in 0.5 ml water and 10 ml toluene-based scintillation fluid.
Samples were hydrolysed with pure enzyme until the amount of liberated
monoSaccharide levelled off and the total amount of radicactivity
liberated by this exhaustive digestion was taken as the total counts of
25 moles of N-acetylgalactosamine liberated. The specific activity
thus found was 400 cpmvnmole.

(4).Assay.of N-[l-]4C]-Acetyl-sphingosyl-GL-5-pehtasaccharide Hydrolysis

 

Hydrolysis of 50 nmoles of N-[l-‘4CJ-acetyl-sphingosyl-GL-5-penta-
saccharide was measured in a final volume of 50 p]; components were
10 u] sodium taurocholate (20 mgflml), 10 pl of 0.3 M sodium citrate
buffer, pH 3.9, and 10 fil enzyme solution. After at 37° for 30 minutes,
the reaction was terminated in a boiling water bath for 3 minutes. The
incubated mixtures were spotted on silica gel G plates and chromatogra-
phed with solvent system chloroform/methanol[acetic acid/ water 55/45/5/
5. At the end of the TLC run, the lanes were scanned by a Berthold
radio-scanner. The radioactive N-[l-l4C]-acetyl-galactosamine spots on
the plates were located by matching the thin layer plates with the radio-
active scanning traces. The areas corresponding to the liberated N-
'acetylgalactosmmine were marked, scraped into scintillation vials and
counted with 0.5 ml ofwater and 10 ml of toluene-based scintillation
cocktail. The total radioactivity of 50 nmoles of liberated N-acetyl-
galactosamine was found by exhauStive hydrolysis of the substrate with

pure.ofiN-acetylgalactosaminidase and counting the area corresponding to

42
the liberated monosaccharide. .The specific activity of the substrate

thus found was 499 cum/nmole.
(5) ASSax of‘PorCine sumaxillary MuCin HydrOLysis

A solution (200-p1) of major A+ percine submaxillary mucin (10 mg/
ml) was hydrolysed by 25 pl of enzyme with 25 pl 0.3 M sodium citrate
buffer, pH 4.2, at 37° for 30 minutes. The reaction was terminated by
the addition of 750 pl of absolute ethanol. The precipitated porcine
submaxillary mucin was centrifuged at 3 minutes at 1200 x g in a Sorvall
GLC-l centrifuge. The supernatant fraction was pipetted out and the
precipitate was rinsed with 1 m1 of absolute ethanol. The pooled
supernatant was dried in a vial under nitrogen, with the addition of 20
nmoles of'mannitol as the internal standard. A mixture (50 pl) of pyri-
dine/hexamethyldisilazane/trimethylsilane 10/4/2 was then added to
trimethylsilylate the sugars. After sonication and standing for at
least 30 minutes, 2-5 pl aliquots of the samples were injected into a
' 6 ft 5% 52-30 column on Supelcoport (80-100 mesh) at 185°. The amount
of liberated N-acetylgalactosamine was quantitated by comparing its
area with the area of the peak for the mannitol. The relative detector
response of N-acetylgalactosamine as compared to mannitol was determined
by comparing the areas of known amounts of the two sugars and was fbund
to be 0.84:1.
(6) assay 0fngAcetyljalactosamine-(d146);l,2:3,4-di+Q§isOpropy1idene-
A on hydrolysis ‘ fi h V - w

The hydrolysis of GalNAc-(ol-6)-l,2:3,4-di-Q-isopr0pylidene-Gal was

 

performed in a final volume of 150 p1, consisting of 100 pl substrate,
25 pl of 0.3‘M.sbdium citrate buffer, pH 4.3; and 25 pl of an appropriate
dilution of the purified enzyme. The solution was incubated at 37° for

43

one hour, terminated by boiling in water for 2 minutes, evaporated to
dryness after the addition of known amounts of mannitol and trimethyl-
silylated with 50 pl of pyridine/hexamethyldisilazane/trimethylsilane
‘10/4/2 for at least 30 minutes. The product ( l to 2 ul aliquot was
analysed by GLC with a 3% SP-2100 column on Supelcoport (100-200 mesh)
temperature programmed from 180° to 230° at 3°lmin. The liberated N-
acetylgalactosamine was quantitated as in the percine submaxillary mucin
hydrolysis assay.
(7) Hydrelysis of Human Red Blood Cell and Dog Intestinal Bleed Group A
Glycolipids

Blood group A-active glycolipids were hydrolysed in a final volume
of 150 pl, a solution with 4 mg/ml sodium taurocholate and 0.1 M sodium
citrate, pH 3.9. After incubating at 37° overnight the reaction was
stopped by boiling for 2 minutes in water and chromatographed by TLC
using chloroformMmethanol/acetic acid/water 55/45/5/5 as the solvent
system. The products were visualized by iodine vapor or orcinol-

sulphuric acid.

(III) Enzyme Purification

Fresh liver was put on ice after excision from the animals. All
subsequent steps were perfbrmed at 4° unless otherwise indicated. The
liver '(6 Kg) was homogenized in a Haring blender twice for intervals of
15 Seconds in 3 volumes of’l mM phenylmethylsulphonyl fluoride and 0.1
M EDTA. The henogenate was centrifuged and adjusted to pH 3.7 with
l‘M citric acid. After stirring for 15 minutes the solution was
readjusted to pH 4.8 with l M sodium citrate and quickly centrifuged.
Ammonium sulphate was added to the supernatant fraction to 30% of

44
saturation and stirred.for one hour before.centrifugation...The 30% super-

natant fraction was then mixedtwith ammonium sulphate to 55% of satura-
tion and allowedto stand for another hour. The 30'-55% amonium
sulphate precipitate was redissolved in water and dialysed exhaustively
against 5 mM sodium citrate buffer, pH 6.0, with several changes of
water. The supernatant fraction of the dialysed, centrifuged enzyme
solution was then loaded onto a DEAE-cellulose column (40 x 4.5 cm i.d.)
packed with 200 g of 0E-52 and preequilibrated with 5 mM sodium citrate
buffer, pH 6.0. After the column was washed overnight with 5 mM sodium
citrate buffer, pH 6.0, the enzyme activity was eluted with 2 liters

of 0.6 M NaCl in the starting buffer and quickly precipitated with
ammonium sulphate to 70% saturation. The precipitate was dialysed
against 0.05 M sodium citrate buffer, pH 6.0, and applied at reom
temperature to a 200 m1 Con A-Sepharose column (30 x 3.5 cm i.d.) which
was equilibrated in 0.05 M sodium citrate buffer, pH 6.0. The column*r
was washed with five column volumes of the equilibrating buffer after
which the activity was eluted with 500 ml of 0.5 M o-methyl mannoside
in 0.05 M citrate buffer, pH 6.0. The solution was immediately
concentrated and dialysed exhaustively and applied to a small 10 9 DE-
52 column (40 x 1.5 cm i.d.) equilibrated with 5 mM sodium phosphate
buffer pH 6.8. o-N-Acetylgalactosaminidase activity was eluted with
200 m1 of 1% ampholine pH 4-6 in water. The eluted enzyme activity
was cOncentrated, dialysed against 0.05M sodium citrate, pH 6.0, and
applied to a 700 ml Sephadex G-150 column (1200 x 25 cm i.d.) equili-
brated with 0.05 a sodium citrate buffer, pH 5.0. Fractions of 12.7 ml
were collected. The enZyme.activity peak (fractions 24-28) was pooled
and dialysed against l 1114 sodium phosphate, pH 6.8. A 10 g

45

hydroxylapatite column(l6.x 1 cm i.d.) equilibrated.in 1 mM sodium
phosphate buffer, pH 6.8, was used for.the final purification step.

The enzyme was loaded and eluted with a linear gradient of sodium phos-
phate buffer, with 220 ml of 1 mM and 200‘m1 of 0.4 M sodium phosphate
buffers, pH 6.8, in the gradient chambers. The final pure enzyme
preparation was concentrated and stored frozen in 0.05‘M sodium citrate

buffer, pH 6.0.

(IV)'Preparation'of'con'A-sepharose

 

Con A-Sepharose was prepared according to the method of Lloyd (190).
Packed Separose 4B (250 ml) was activated according to the method of
March, Parikh and Cuatrecasas (191). Con A (2 g) was added to a 500 ml
suspension of the activated Sepharose beads with 2.2 g of NaHCO3 at
4° and after two days with occasional swirling, the beads were packed
into a column and washed with 1 liter 0.07 M NaHC03, and 4 liters each
of 0.1 M sodium borate, pH 8.5, in l M NaCl, 0.1 M sodium acetate, pH
4.1, in l M NaCl and finally 1 M NaCl. The Sepharose was then suspended
in l M NaCl in a final volume of 700 ml. ‘The.amount of Con A coupled to
Sepharose was 761 mg, as determined by the direct Lowry determination
(192) of an aliquot of the beads.

'(V) Methanolysis

Methanolysis of glycolipids and carbohydrates were performed by
the method of‘Vance and Sweeley (193). Samples were hydrolysed in 3 ml
of 0.75 H methanolic HCl in sealed ampules for 24 hours at 82°. Fatty
acid methyl esters were extracted three times with equal volumes of
hexane. Silver carbonate was added until the solution is neutral to

a wetted pH paper, 'Acetic anhydride (0.3‘m1) was added to re-N-acetylate

46
the amino-sugars for at least 6 hours at reom temperature. After the

addition of mannitol and 2 drops of water, the sample was centrifuged and
the supernatant fraction removed. The residue was washed four times with
2 ml aliquots of methanol. The pooled supernatant fraction was dried
under nitrogen and derivatized with about 100 pl of pyridine/hexamethyl—
dichlorosilanzane/trimethylchlorosilane 10/4/2. Trimethylsilylated
methyl glycosides were analysed on an F a M Model 402 gas chromatograph
with a 3% SE-30 or SP-2100 on Supelcoport (80-100 mesh).

(VI) Pennetmuation ‘ for, Linkage Studies

 

Permethylation of carbohydrates was perfbrmed by the method of
Hakomori (194). All the penmethylation operations were done under
dry nitrogen. Hexane was dried by refluxing with Ba0 (20 g/liter) for
2 hours. The redistilled dried hexane was stored over sodium. Di-
methylsulphoxide was dried by refluxing with Ba0'(50 g/liter) for 2
hours, redistilled and stored over molecular sieves. All the other
solvents used were redistilled.. A sample of NaH (0.9 g of 57% oil
emulsion) was washed 7 times with 15 ml aliquots of dried redistilled
hexane. Dry redistilled dimethylsulphoxide (10 ml) was added and
allowed to react at 65-70° until bubbling of hydrogen ceased (approxima-
tely 90 minutes). The dimethylsulphinyl ion solution (0.5 ml) was
added to samples (0.5 mg) dissolved in 0.5 ml dimethylsulphoxide and
allowed to react for 30.minutes with periodic sonications. Redistilled
iodomethane (2 ml) was then slowly added and allowed to stand for 2
hours at room temperature. The solutions were then dissolved in 5 ml
of chloroform and washed twice with 5‘ml water, Once with 5 ml 20%
Na25203 and thrice with water. The samples were dried under nitrogen

with the aid of absolute alcohol and hydrolysed in 0.5 ml 0.5 N H2504in

47
95% acetic acid for 24 hours at 85°. Hater (0.5 ml) was then added and

allowed to react for an additional 5 hours at 85°. A small column with
2‘ml of Dowex l x 8, acetate form (50-100 mesh) was used to adsorb the
sulphate. The column was washed with 2-3 ml of acetic acid. The
hydrolyzate was evaporated to dryness under nitrogen and reduced with
0.5 m1 of NaBH4 (10 mg/ml) for 2 hours. Reduction was terminated with
the addition of several drops of glacial acetic acid. The solution was
dried under nitrogen. Borate was removed as its methyl ester by 1-2
drops of acetic acid and 2 ml methanol, heating in a boiling water

bath for 5 minutes, and evaporating under nitrogen. The esterification
procedure was repeated three more times. The dried sample was acetyla-
ted in 0.5-1 m1 acetic anhydride for 60-90 minutes at 100°. After
drying under nitrogen with the aid of toluene, dissolved in 2 ml of
CH2C12, washed three times with 1-2 ml water and drying under nitrogen
again, the partially methylated alditol acetates were ready for analyses
by GLC or GC-MS. Column packings used for these analyses were 3%
ECNSS-M on Supelcoport (100-200 mesh), 3% 0V-210 on Gas Chrom Q (80-

100 mesh) or 0.2% EGS/0.2% EGA/1.4% XE-60 on Chromosorb P (100-200 mesh).

The mass spectra were interpreted according to Bjorndal at oz. (195).

(YII) Polyacrylamide Gel Electrophoresis
The method of Gabriel was used for native gel electrophoresis (196).
Gels were stained with coomassie blue fast stain (197). p-Nitrophenyl-
o-N-acetylgalactosaminidase activity on gels was monitored by incuba-
tion of the gel in 3 ml of p-nitrophenyl-o-N-acetylgalactosaminide
solution in a final concentration of 2 mm substrate and 0.1 M sodium
citrate, pH 4.3, at 37° for 15-30 minutes. Liberated p-nitrophenol was
visualized with 0.6 M potassium borate, pH 10.4. The activity was

48

scanned immediately at 410 nm on a Gilford spectrophotometer equipped

with a linear transport unit.

"(yIII)°General:Procedures

Isoelectric focussing was performed by the method of Vesterberg
(198) using ampholytes at pH 3-10, 4-6 and 5-8 and an LKB isoelectric
focussing unit with a capacity of 110 ml.

A Sorvall RC 2-B centrifuge was used for enzyme purification
procedures. Either a GSA or 58-34 rotor was spun at 23,000 x g for 20
minutes. A Sorvall GLC-1 centrifuge was used for all centrifugation
procedures that did not require high speed. A centrifugal ferce of
1200 x g for 3 minutes was used. Protein was determined by the method
of Lowry at al. (192) using BSA as standards. pH measurements were
performed at 25°. Solutions used in enzyme purifications were prepared
in glass-distilled water. The purified enzyme used for assays was
diluted with 10 mg/ml BSA solutions. The final protein concentrations
in enzyme assays were maintained above 0.4 mg/ml. Dialysis tubing was
boiled with ethanol and washed by a published procedure before use (199).
o-Galactosidase from ficin and B-galactosidase from Jack bean were
purified by procedures described previously (200-202). The conditions '
for enzymatic and partial acid hydrolyses have been published (203,204).
Periodate oxidation was performed by the method of Siddiqui and Hakomori
(3). Amino acid composition was analysed as described (205). The
carbohydrate composition was determined by the method of Chambers and
Clamp (206). Molecular weights were detenmined by $05 polyacrylmmide
gel electrophoresis as described by Segrest and Jackson (207) using
cytochrome C, aldolase, catalase, chymotrypsinogen A, BSA, ovalbumin,

and ferritin as standards. The molecular weight of the native enzyme

49
was estimated by Sephadex G-150 chromatography using the same protein
standards (208) 3.

Sialic acid in the purified a-N-acetylgalactosaminidase was analysed
by incubating 59 pg of enzyme with 0.05 units of neuraminidase from
czootridtum perfringens in a final volume of 100 pl, containing 0.05 M
acetate buffer,pH 5.0, for 2 hours at 37°. After terminating the
reaction with 250 pl of 5% phosphotungstic acid the liberated sialic
acid was quantitated by the thiobarbituric acid assay (209).

Glycolipids and carbohydrates on TLC plates were visualized by
heating in a 110° oven for 10 minutes after spraying with 2% u-naphthol
in 95% alcohol and sulphuric acid or 0.5% orcinol in 4 N_sulphuric acid.
Glycolipids were also visualized by exposing the plates to iodine vapour
in a ch tank. '

RESULTS

CharacterizatiOn'of the;Pentaglycosylceramide in Dog Intestines

A thin-layer chromatogram of the mixture of glycosphingolipids from
canine kidney and intestine is shown in Figure 4. Monoglycosylceramide
(GL-l), triglycosylceramide (GL-3), and hematoside (6M3) were the major
components of kidney, while GL-3 and the pentaglycosylceramide (GL-5)
were the major constituents of intestine (Table 2), Gas chromatographic
analysis of the trimethylsilyl derivatives of methyl glycosides from
GL-5 (Figure 5) indicated that the oligosaccharide moiety consisted of
glucose, galactose, and N-acetylgalactosamine in a molar ratio of 1.0:
2.1:2.0. Colorimetric analysis of the long-chain base fractions,
liberated from GL-5 by methanolysis. indicated approximately 1 mole of
base per mole of glucose.

Partial acid hydrolysis of GL-S liberated glucosylceramide and a
diglycosylceramide with a 1:1 molar ratio of galactose to glucose, in-
dicating a probable partial internal sequence of galactosylglucosyl-
ceramide. Enzymatic hydrolysis of GL-5 was carried out by the following
mixtures of glycosidases on a scale (200 pg each) sufficient for analy-
sis by thin-layer chromatography (Figure 6): (a) o-N-acetylhexosamini-
dase alone, and (b) a-N-acetylgalactosaminidase plus B-N-acetylhexosami-
nidase. The GL-3 derived from partial acid hydrolysis or enzymatic
hydrolysis was treated with (a) o-galactosidase, and (b) o-galactosidase

plus a-galactosidase. The products were a tetraglycosylceramide, a

50

Figure 4.

51

Thin-Layer Chromatography of Glycosphingolipids of
Canine Intestine (Lane I) and Kidney (Lane K). Lane
5: Standards from Porcine Erythrocytes.

The solvent system was chloroform/methanol/water 100/
42/6 and 0.25 mm Silica gel H plates were used. Plates
were visualized by iodine vapor.

52

 

Figure 4.

FRONT

GL-l

GL-2

(51.-3

GL-4
GL-S-
GM3
ORIGIN

53

Table 2.~ Yields of Glycosphingolipids from Canine Intestines.

 

,Glycosphingolipids. Amount (pmoles/g wet weight)
Glucosylceramide 0.052
Lactosylceramide 0.050
Triglycosylceramide 0.160
Tetraglycosylceramide trace
Forssman hapten 0.110

Hematoside 0.028

Figure 5.

54

Gas-Liquid Chromatography of Trimethylsilylated Methyl
Glycosides of Standard Sugars (Top). Canine Forssman
Hapten (Center) and Porcine Globoside (BottOm).

A: methyl galactosides; B: methyl glucosides; and
C: methyl N-acetylgalactosaminides.

Analyses were performed on a 2 m x 2 mm column packed
with 3% SE-30 on 80-100’mesh Supelcoport and programmed
from l60°-210° at 2°/min with nitrogen as carrier gas.

RESPONSE

DETECTOR

 

 

 

STANDARDS

A. 13 C
F‘jfil—lfl 1 l l

 

 

 

 

A GL-5

 

 

J

PORCINE GL-4

 

 

\li

 

 

 

,5 l0 is 20

RETENTION TIME (min)

Figure 5.

 

Figure 6.

56

Thin-Layer Chromatography after Enzymatic Hydrolyses of
Forssman Hapten.

Lane 1: Canine intestinal GL-S;

Lane 2: GL-5 treated with o-N-acetylgalactosaminidase;

Lane 3: GL-5 treated with d-N-acetylgalactosaminidase
and B-N-acetylgalactosaminidase, could be
visuaszed by o-naphthol-sulphuric acid;

Lane 4: GL-3 obtained from partial hydrolysis of GL-S;

Lane 5: GL-3 treated with a-galactosidase;

Lane 6: GL-3 treated with o-galactosidase and B-ga-
lactosidase;

Lane 7: Horse kidney glycolipid standards.

TLC conditions are the same as Figure 4.

57

 

Figure 6.

58
triglycosylceramide, a diglycosylceramide, and a monoglycosylceramide,
respectively. The sequence and stereOChemistry of the glyCosidic
linkages were therefore (o-N-acetylgalactbSaminyl)-(B-N-acetylga1ac-
toSaminyl)-(e-galactosyl)-(3-galactosyl)-(glucosyl)-ceramide. These
results confirm the arrangement of carbohydrate units proposed by
Vance at at. (1).

Gas chromatograms of partially methylated alditol acetates,
obtained from the neutral sugars of canine GL-5 and porcine globoside,
are shown in Figure 7. The three peaks from GL-5 corresponded in
their relative areas and retention times to those of porcine globoside.
Mass spectra provided evidence that they were l,3,5-tri-Q-acetyl-2,4,6-
tri-Q-methylgalactitol (peak A), 1,4,5-tri-Q-acetyl-2,3,6-tri-Q-methyl-
galactitol (peak a), and l,4,5-tri-Q_-acetyl-2,3,6-tri-_0_-methylglucitol
(peak C), respectively. Enzymatic degradation of 5 mg of GL-5 by a
mixture of a-N-acetylgalactosaminidase and 3-N-acetylhexosaminidase
(as described above) gave a triglycosylceramide which was isolated by
preparative thin-layer chromatography. Permethylation and mass spectral
analysis of the products after acid hydrolysis, borohydride reduction,
and acetylation gave evidence for the presence of l,5-di-Q-acetyl-2,3,4,
6-tetra-Q-methylgalactitol, l,4,5-tri-Q-acetyl-2,3,6-tri-Q-acetyl-
galactitol, and l,4,5-tri-Q-acetyl-2,3,6-tri-Q-methylglucitol. The
internal likages were therefore galactosyl-(l+4)-galactosyl-(144)-glu-
cosy1(l+1)-ceramide. Hhen considered along with the results of permethy-
lation of the intact GL-S presented above, it was cencluded that the
penultimate N-acetylgalactosamine residue must have a 1+3 glycosidic
linkage to galactose.

Gas-liquid chromatography of the partially methylated alditol

Figure 7.

59

Gas-Liquid Chromatography of Neutral Partially Heth lated
Alditol Acetates from Canine Forssman Hapten (upper)
and Porcine Globoside (lower).

The carbohydrate derivatives were analysed on a 2 m x
2 mm column with a mixture of liquid phases containing
0.2% ethylene glycol succinate, 0.2% ethylene glycol '
adipate and 1.4% XE-60 on 100-120 mesh Gas-Chrom P.
The temperature was programmed from 155° to 210° at
2°/min. Retention times are in minutes.

RESPONSE

DETECTOR

60

 

 

 

 

 

 

 

 

 

Ni-

o a l0 l2 l4

RETENTION TIME

Figure 7.

‘16

 

61

acetatesderived fromthe neutral and aminosugars from GL-5 is

shown in Figure 8. Peak C is identical in retention time with the
alninosugar derived fran globoside; N-acetyl-N-methyl-l‘,5-di-N-’acetyl-
3,4,6-tri-Q-methylgalactosllminitol (Structure 1,,‘Figure '9). Mass
spectra of the authentic sample, from globoside, and that from the
unknown are shown in Figure 10. Primary fragnent ions at m/e 158,161,
202 and 205 can be found in both specta. The base peak at m/e 116 is
probably derived from m/e 158 by the loss of ketene (42 a.m.u.). Pro-
minent peaks at We 98, 142 and 145 are assumed to be formed from m/e
158, 202 and 205, respectively, by the loss of'acetic‘acid (60 a.m.u.).
A significant peak at m/e 129 is related to m/e 202 by the loss of
either CHZOAc or N-methylacetamide (73 a.m.u.).

Peak 0 (Figure 8) had the same retention time as N—‘acetyl -N-methy1-
l,3,5-tri-9_-acetyl-4,6-di-Q_-methylgalactosaminitol (Structure 2, Figure
9) derived from N-acety'lchondrosine, and mass spectra of the two peaks.
shown in Figure 11, B and C, indicate their identity. Significant
peaks from primary ions were found at m/e 158 and 318 (M+-'73). The base
peak was at m/e 116, as observed with other partially methylated
alditol acetates of aminosugars~ permethylated by the Hakomori proce-
dure as outlined by ijrndal at at. (195). Secondary fragments were
found at m/e 170, assumed to be formed from the ion at m/e 230 by the
loss of acetic acid; m/e 242 derived from m/e 346 by the loss of acetic
"acid and ketene; and m/e 272, derived from m/e 346 by the loss of ketene
and methanol.

The mass spectrum (Figure 11) of N-acetyl-N-methyl-l,4,5-tri-_Q-
'acetyl-a.edig-mthylglucosminitol (structure 3, Figure 9). prepared
fruachitin, iskclearly different from Peak 0 derived from GL-S. A
similar pattern of ionswas observed at We 74, '98, 1'16 and 158, but

DETECTOR RESPONSE

62

 

 

 

 

 

 

 

 

 

 

I

i

I A B

I I

I

u 5 lo 1.5+2l5 3'0 35

RETENTION TIME

Figure 8. Gas-Liquid Chromatography of Partially Methylated Alditol

Acetates of Neutral and Aminosugars from Forssman Hapten.

Analysis was performed on a 2 m x 2 lllll column of 3% 0V-210 on
80-100 mesh Supelcoport operated isothermally for 15 minutes
at 160° and then programmed from 160° to 250° at 5° per minu-
tes with nitrogen as carrier gas. Retention times are in
minutes. A: l ,3 ,5-tri-0-acetyl- 2,4, 6-tri-Q-methylgalactitol;
B, a mixture of 1,4, 5-tri-9_-acety1-2,3 6-tri-0-methyl-galac-
titol and glucitol. The mass spectra of peaks c and D are
shown in Figures 10 and 11.

63

.3355833Seeoeeée. M. $83.25. e. 7 $552-53; .m 2385
2; ..3eeeamoeofiefleeoe-oé-e. $383.2»- m. «.7 ”€52-53; .N 232:3
.3:558»:322559.3e. e. m- 38343. 7 355538.... ._ 2:323.

532855 ee 5:52 F332 notifies 5:29. to 3:253“. 232..

m
ozoezo
20 I
I .. .. memo I
Now rumor: m8
Mex/.201- - .3”
oqomzo 2

N
mEONInw me
oqoonllea
.. _..m.._ .. modelaMM
one I83
04» T .. .. sew...
$.sz .. new
oqomzo e.

.m meamvm

secure
o<ovI
5. IOOmE

l"'l

mow 10022 mow

oq, Til..-
. asewF -....._

oqoezo a.

Figure 10.

64

iMass Spectrum of Partially Methylated Alditol Acetate

of Terminal Aminosugar of Forssman Hapten Compared
with a Standard.

(A) N-acetyl-N-methyl- 1 ,5-di-0-acetyl -3, 4, 6-tri-0-
methylgalactosaminitol (Strucure l in Figure 9)
from pOrcine globoside;

(8) peak C in Figure 8.

INTENSITY

RELATIVE

65

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

1“" ' 4
I 116 §
' A .
301 43 ‘ 'x 10 ;L
z i
I
80" j {I
158 / i
(.01 I45 / f
74 '42\ / !
205 I I
.. J 129 202 / 2
-“ 98 ,’ I
l I / I i
U mth'm'L‘ .‘uWHhir‘J
loo
1 6
43 ' B
30‘ /-xlo L
w / _
74 l45
40, 98 ME
129 202
204 I
II
/ r'L [LLM 1 '7‘ m'r'“
.pv 100 130 200 250 300 330
nn/m:

Figure 10.

ll)

66

Figure 11. Mass Spectrum of Partially‘Nethylated Alditol Acetate
of the Internal Aminosugar of Forssman Hapten Compared
with Standards.

(A) N-acetyl-N-methyl-l,4,5-tri-0-acetyl-3,6-di-Q-
‘methylglucosaminitol from chitin;
(B)‘N-acetyl-N-methyl-l,3,5-tri-0-acetyl-4,6-di-0-
‘methylgalactosaminitol from N}acetylchondrosine;
(C) Peak 0 of Figure 8.

INTENSITY

RELATIVE

67

1‘") j 10

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

43 116
.- .X-W
“."I 233 A ’
158
sol I
74 98
mi
30 202
ILA-1- W
100 - 12
I16
100 . 13
43 II6 I58
.. C
604
74 242
40‘ 98 272 3l8
”OJ I70
I
so
so 100 150 200 . 250 300
m / e

Figure 11.

68
significant ions characteristic of this amino sugar.were observed at
m/e 202 and 233'. Furthermore, fragment ions at m/e 161, 170, 242, 272
and 318 in the spectrum of Structure 2 were either absent or of very
low intensity in this spectrum.

Glucose, galactose, and N-acetylgalactoSamine in a molar ratio
of 1:1:1 were recovered after periodate oxidation of GL-S, showing
that the terminal N-acetylgalactosamine residue is not linked 1+6 t0 the
penultimate sugar. The glucose of GL-S is periodate-resistant under
these particular condition (3). It can thus be concluded from analyses
of permethylation and periodate oxidation products that GL-5 has a
1+3 glycosidic linkage between the two N-acetylgalactosamine residues.

The pentaglycosylceramide from kidney and intestine had strong
Forssman activity when tested by the complement fixation test (210).
Hydrolysis with a-N-acetylgalactosaminidase gave a product which showed
globoside activity by Ouchterlony double diffusion and provided addi-'
tional evidence for the assignnent 0f the configurations of the internal
glycosidic linkages of GL-5. A single precipitation band was observed
when GL-5 was tested by Ouchterlony double diffusion with anti-Forssman
antiserun prepared against horse spleen Forssman antigen. Furthermore,
the precipitation band was fused with the horse spleen Forssman preci-
pitation band with no overlap.

Hith the data presented above, it was demonstrated that canine
GL-5 isolated from both kidney and intestine is a true Forssman hapten,
which is chenically and immunologically identical with that of horse
spleen Forssman hapten (3), that is, 'N-acetylgalactosaminyl-(ol-r3)-N-

‘ acetyl gal actosami nyl -( 5143)-galact0syl -(o'l-4)-galactosyl-(el+4)-glucosy1-

(1+1 )-ceralllide.'

"Purification of o-N-AcetylgalactoSaminidase from'Porcine Liver

.The.method and results of purification are summarized in Table 3.
The pNP-o-N-acetylgalactosaminidase activity was purified 3,300-fold with
a 2% recovery while the Forssman hapten hydrolysing activity was purified
20,000-fold with a 10% recovery. A protease inhibitor,phenylmethylsul-
phonyl fluoride, was used to prevent proteolytic degradation of the
enzyme (211). In the acid precipitation step, the ForSsman hydrolysing
activity increased almost two-fold. This may be due to precipitation of
unknown inhibitors in the homogenate. Steps subsequent to DEAE-cellulose
chromatography should be performed quickly so that the enzyme is not
inactivated due to dilution. The Con A-Sepharose step is the most
potent single step in the purification scheme, giving a 23-fold purifi-
cation for the pNP-o-N-acetylgalactosaminidase activity and a 47-fold
purification fer Forssman hapten hydrolase activity. The enzyme was
eluted by o-methyl mannoside in a skewed peak, as shown in Figure 12 for
a typical run. A large number of fractions had to be pooled because of
the tailing of the peak and consequently a batchwise method was adopted.
The DEAE-ampholine elution step is shown in Figure 13. There were
shoulders of’o-N-acetylgalactosaminidase activities, indicating not only
that the enzyme preparation was chromatographically heterogeneous, but
also the high resolution of the ampholine elution procedure. The amount
of purification obtained by this step is also evident in the native gel
electrophoresis of enzyme preparations in Figure 14. A major contamina-
tion band which ran slightly slower than o-N-acetylgalactosaminidase on
native gel electrophoresis (Figure 14) was removed by Sephadex G-150
column chromatography. This major impurity band could also be seen in
the Sephadex G-150 elution profile shown in Figure 15. It eluted before

69

7O

 

e.op m~.P oommp opmm owa— case o.m a.mp u.~ mupuuaupaxosux:

 

.em m.m owns coup mus ocmu ¢.~p m.am a.mp om—ia xuvosnmm
¢.mm m.n omov new man «NN— 0.0— a.mm a.mc caveapu ocwpogasi
a.mm 5.5 ompw mmm N.ch vwh N.m— N.wo hm umosusamml¢ coo
m.pv NN cc a.mw m¢.¢ o.¢m m.ap Nmp the omopa—Puu m<mo

mup Nm w.m N.m mn.o vm.¢ a.mm mom o—oom ummiom comwavzzv

cm— an N.@ m.N Vam.o m.m m.om , n¢m oopwm— :uan uvu¢
cc p cop — F mac .0 Ne . _. m . me was oummmc manganese:
E 2.33 355er mazes 3.523 33
assuage“. ezfi c2528. can - .5533“. Azn zmfimeeh. a? 532.. 538....—
huu>ooo¢ gcuuam .m—Lag auv>ruu< owmvuonm haw>puu< pouch punch

.33.. 0593.. 32m omegfifimouuapuuEvenings mo 53332.... .m «Se»

B:

S

 

g

119 PROTEIN me-o-o-
8

Figure 12.

71

 

 

 

 

 

a-MM' -5 I
’ l
B
‘4 N ’2:
h E
3 8 5
i 2313
«2 L9 3
‘ E
‘l U)
(D
D
. , m
._ __.__.__._~_._._.__T__',__‘.__‘Qi , .___;_._._a_‘
D 20 30 40 50

ELUTION VOLUME (ml)

A Typica] Elution Pattern of d-fl-Acetyigalactosaminidase
from Con A-Sepharose Comm.

9 ng of DEAE eluate was loaded onto 2 m1 of Con A-Sepharose.
Arrow indicates elutionuith 0.5 M comethy‘lmannoside. The
substrate used was pNP-a-GaINAc.

72

2:295 a: £3 83:. we. 5:5: 2:

.ose za .gcuu: :*

.3 .3 5&3 3239... .558 E. m 5:.

vanguarpraom A.u.m so m.p x ocv gaspou «mums u op a so awesomsun ma: agnngmoaascggu as»

33.5533:33325-5 t 538:23 2: 5 53:: 252.3%“: .2 9:5:

O
(\IIO

iVHBBFIde
O
T

@001

)

In OmL

UU

m. 8...
wow. ..
II

 

w
no:L

(

 

 

 

 
 

 

 

2c: m2340> 20....qu
NOW. . 0.0. 00. Om

0.0 r a» i _. .. No d
. . _. H
F/ ‘ IVO m
om- t .w. imoe
IQ . .. 10.0 N
ON: Em , ix .. o. N
. hot-Q lull-I. _ . .. 0
.~. mw._ “It
Hung .u n .anla. . . mm
z_cm -v_fi
82.006 cl... .. m._ M

 

 

 

Figure 14.

73

Native Gel Electrophoresis of Enzyme Preparations in
the Purification of a-flfAcetylgalactosaminidase.

Electrophoresis was performed at pH 8.3 in 7% gels at
2 mA/tube.

A) Blank gel;

B 100 ug of Con A-Sepharose eluate;

C ‘60 ug of DEAE-ampholine elution eluate;
D 27 pg of Sephadex G-lSO eluate.

* —_—__—-————_——I_P_

‘_.—‘ fi__-.._.

74

 

 

 

Figure 14

75

638:8 28: E 5.2 we 23502...

.22.... 2 .8 32 so: a
an .o.o 1n .mewzn mansu'u Eswuom : mc.o saw: causpa ma: xup>puun weancu
mo 25:; van a 52.: 5:38 6.. .5 mu...“ 82. a 53 3.5353 mm: £33355

.psacos

$832333F338$m3 t 8585...... 23 5 2:2,. 8:5 83 58.38

(Iw/Ju/salow'n’ )

palmaqn de

0.
ON
On

0?.

On
Om

.5962. c288...

 

 

I

 

 

. 50.0}; olo t
82.8.1.1 .
62.86 «L

 

 

J o

 

 

to st r9 N =-
O O O o O
(IN/5W)'ONOO NIL-LLOHd

0

.mp ug=m*m

76
the enzyme activity.peak.

‘BfiflfACetylgalactosaminidase eluted slightly befbre the a-flfacetyl-
_galacto$aminidase activity. An attempt to separate.the two enzyme
further by repeating the chromatography on the Sephadei 6-l50 column
was not successful. In Figure 16, aeflfacetylgalactosaminidase activity
eluted with the protein peak and the B-!:acetylgalactosaminidase eluted
wdth it as well. If the efflyacetylgalactoSaminidase'activity were eluted
in a Gaussian distribution, it would have been eluted before instead of
coinciding with the agflfacetylgalactosaminidase activity; This implied
that the a-flyacetylgalactosaminidase in pdrcine liver was heterogeneous
in eiclusion behaviour.

Somelslight contamination persisted after Sephadex GAlSO column
chromatography and can be seen as a faint band slightly above the two
enzyme bands in Figure 14. This faint band was removed by hydroxyl-
apatite chromatography (Figure 17). This step gave a two-fold purifi-
cation of the pNP-a-GalNAc hydrolysing activity and 2.6-fold purifica-
tion of the Forssman hydrolysing activities but only 30% of the applied
pNP-afigracetylgalactosaminidase and 43.5% of the Forssman hydrolysing
activity were recovered. The hydroxylapatite step was also necessary
for the elimination of the final traces of 87!:acetylglucosaminidase

activities.

' PhySicalfiand;§hemical'Characterization‘of'Enzyme'Preparation
The final enzyme preparation was electrophoresed as the native
enzyme (Figure 18). The scan of protein stain at 550 nm indicated that
there were twoprotein bands (lower scan); The activity stain (upper
scan) showed that both of these bands are enZymatically active against

77

H ot-GolNoc ose
H {d-GolNoc ose
H Protein

 

 

 

 

 

06 L. '. G‘I5O II 70E
E fiGOg
\ O5” #3
2‘ ‘50 e
‘f 04- :3
£2) l 440 8
0 O3— ‘1’ I—
: «so:
__ _ Lu
:2 02 ~20?
o I“. \\ .0-
E 0"” j ‘. p .. IO 5.

O A O

25 360 350
ELUTION VOLUME (ml)

 

Figure 16. Repeated Sephadex Sol 50 Column Chronatography of d-E-ACEty‘I-
, galactosaminidase.

The conditions were the same as Figure 15.

78

.:n mean as» an commas ouogamoza

.553. 2 ed me E SN 23 £6 :a .3232; 53.8» z... p we FE emu mo «5:53

.32.: a :23. 3528 was 53:5 .md matcotan 3232... t.— _. 5:; 3322339...
33323.3ch 2. 2 5.... 55396.. .5 F x 3 a 5 3.5923 mm: 233325.28

.332233923332$... B 533:2... as» 5 23339529 35325.23.- .2 0.53“.

2:: w2340> 2025.5
00. Om
- . .

 

Om.
_

    

no.0...

om~< moo

 

5L
«5

J
(I)
eso ooNlog-o °—‘
( MIN/5810017”) oaavaae n chl

 

 

 

"""" (HV'IOW) 'ONOO BlVHdSOHd

Figure 18.

79

Scans of Native Gel ElectrOphoresis of Purified o-N-

'Acetylgalactosaminidase Preparation from Hydroxylapa-

tite Column.

22.5 ug of the purified enzyme preparation was used for

each gel. The electrophoresis conditions are the same

as Figure 14. The upper scan is the activity scan at

:10 nm and the lower scan the protein stain scanned at
50 nm.

80

 

 

 

 

 

 

 

O! J PNP
moos-
t:
Z
3
LIJ ,
‘2’ 06' PROTEIN
E.
(I: ()H‘; P
O
U)
m 02
<[
O 1 1

 

 

O 2 4 6 8
MIGRATION (cm)

Figure 18.

81
the p-nitrophenyleo-GalflAc.substrate. .The higher molecular weight
species may be an aggregation of the lower molecular weight form.

The enzyme was reduced with dithiothreitol and electrophoresis
was carried out in 505 with 5:. 7%. 10% and 12.57; gels. Figure 19 shows
the SDS gel electrophoresis scan of the purified a-flyacetylgalactosamini-
dase in 10% gel. One major protein peak was found. A small shoulder
that migrated slightly faster than the enzyme band can be seen. The
molecular weight was estimated by Comparing the relative migration of
the reduced enzyme with that of standard proteins. In Figure 20 only
the curves for 10% and 12.5 x gels are shown. The enzyme was found to
have an apparent molecular weight in SDS gels of about 52,000 daltons.

The molecular weight of the native enzyme was determined by gel
filtration on Sephadex 8-150. A molecular weight of about 102,000
daltons was found. The native enzyme is probably composed of 2 sub-
units of the 52,000 molecular weight species.

The carbohydrate composition was found by GLC after methanolysis
of 500 ug of the purified enzyme in l lmethanolic HCl. The GLC trace
of the carbohydrates from o-fl-acetylgalactosaminidase is shown in Figure
21. The top trace (1) is of the trimethylsilylated methyl glycosides
derived from the enzyme: A, B and c are methyl mannosides. mannitol
and methyl fl-acetylglucosaminides. respectively. Although A did not
coincide exactly with peaks A in trace II. they had almost identical
retention times relative to mannitol. 0.621 compared to 0.612. Like-
wise. the retention times of C relative to B were 1.24 and 1.26 for
traces '1 and II. respectively. x was probably derived from citric
acid. Y and 2 were not found in the traces when the sample was in-
Jected 30 minutesafter derivatization. indicating that they had. groups

Q
4:.
l

ABSORBAN CE UNITS

Figure 19.

Q
N
l

82

 

 

 

 

 

 

l l I

o 2 4 6 8
MIGRATION (cm)

Scan of SDS Gel Electrophoresis of Purified o-fl-Acetyl-
galactosaminidase .fraa Hydrwlapatite Colum.

22.5 319 of the purified enzyme was electrophoresed in 10%
gel in'O.l‘H-sodium phosphate. pH 7.1, with 0.1%;SDS at
Bah/tube. The'Scan was at 550m with a scan speed of

0.5"anlm'in.

83

Figure 20. Molecular Height Determination of the Subunits of
Purified o-Né-Acetylgalactosaminidase in SDS Poly-.-

acrylamide Eels; a sani-log plot of molecular weight
vs the relative mobility of the protein.

The electrophoresis conditions were the same as Figure
19. The crosses on the curves indicate the relative
mobility of the enzyme.

MOL. WT. x i0"4

.Sb

  
 
 
 
 
  

 

84

0 BSA
Coiolose

Ovolbumin
Aldolose

Chymotrypsin

IO%

Cytochrome C

02 0-4 0-6 0-8
RELATIVE MOBILITY

Figure 20.

Figure 21.

85

Gas-Liquid Chromatography Trace of Carbohydrate Analysis
of a—fl-Acetylgalactosaminidase.

I is the trace of sugars from degracetylgalactosamini-
dase, II and III are standard sugar runs.

A: mannose; B: mannitOl: C: li-Lacetylglucosamine:
D: NeuAc; E: galactose:'F: gTucose; G: flfacetyl-
gal actosami ne .

The sugars are trimethylsilylated methyl glycosides.
X. Y and Z are unknown impurities. GLC was performed
on a 3% SP-ZlOO on Supelcoport (80-100 mesh) with
nitrogen at a flow rate of 45 mllmin. Temperature was
prograivnied from 150°-230° at 3°lmin.

DETECTOR RESPONSE

86

 

 

 

 

 

A. Q 9. I
Y Z
I h
A EL 3.
II
F:

   

 

 

 

5 IO I5 20
RETENTION TIME
(min)

Figure 21.

 

87
(e.g. COOH orNHz) thatwere more difficult to derivatize than OH.
groups. Besides. these peaks did not coincide with peaks of known
trimethylsilylated methyl glycosides. The carbohydrate composition
data also agreed with the fact that the enZyme was bound by Con A,
which specifically binds with o-mannoside or o-glucoside residues (212).
Mass spectrometry of peaks at A. B and C also confirmed the identities
of these peaks. The carbohydrate Composition of o-fl-acetylgalactosami-
nidase is shown in Table 4. llannose and fl-acetylglucoSamine were
present in a ratio of 2.4:1 and the two sugars constitute about 7.4%
of the total weight of the enzyme. It is worth noting that there is
no galactose. fucose, glucose. fi-acetylgalactos’amine or sialic acid
in this enzyme.

The amino acid composition of the enzyme is shown in Table 5.
Aspartic acid and glutamic acid comrised 19.6%. while serine and
threonine constituted 14.2% of the total amino acid. These two types
of amino acids are known to be involved in glycosidic linkages with
01 igosaccharides in glycoproteins. The non-polar amino 'acids made up
about 42% of the amino acids and basic amino acids shared about 14%
of the total amino acids.

Isoel ectric ‘ Focussi 111 of the ‘ Purified Enzyme Preparation
Heissmann and Hinrichsen (9) found that a partially purified

preparation of d-fl-acetylgalactosaminidase from pdrcine liver could be
separated by isoelectric focussing into eight bands. .all of which were
active toward phenyl m-fi-acetylgalactosninide. These fractions had
isoelectric points which varied from pH 5 to 6.5. Isoelectric focussing
of the more highly purified enzyme obtained by our'procedure showed the
same general patternof multiple peaks (Figure 22). Forssmanhapten

88

Table 4. Carbohydrate Composition of’aeflyhcetylgalactoSaminidase

.Sugar. .AmouhtfthmOTeslhg-proteih)

 

‘Hannose 0.290
!:Acetylglucosamine 0.l2l

 

89

Table 5. Amino Acid Composition of deflfAcetylgalactosaminidase

Amino Acid Percentage

Cysteic Acid ~-
Hethionine Sulfone --

Aspartic Acid 8.68
Threonine 6.06
Serine 8.14
Glutu'iic Acid 10.91
Proline 0.33
Glycine 12.27
A1anine 10.19
Valine 0.71
Methionine 0,24
Isoleucine 6.20
Leucine 12.94
Tyrosine 3.68
Phenylalanine 5.92
Lysine 4.78
Histidine 2.99
Arginine 5.97

Total 100

 

90

.33 mm: 558.... mnemfiumé =8 «5 we a... «6 «:22 .> 08 an 2.5.. we s8
m use m 5253 :q 5:. E5283 2. a: a 5 3.5923 an: 35338 8.50333

632553328pxuou<umue vet—hi 2.33.5. mo 9.3.0.38 258.33 .3 953m

:5. “.239
ow an 8 0.

ill 0010,0101.

9
8
18

O

0

g

MW WmmIWww-Nmm

§
V
O

  

if

at”
Munoez .Lv aouveuosew unnamed
(D

 

 

§ §

 

«‘7‘
8
&
#Hd

 

«(w/W) mom-l mm
Ail/01W 3mz~3

§§§

 

 

r

 

§

91

hydrolase'activity generally coincided with the peaks of activity toward
artificial substrate. although the ratio‘of glycolipid to artificial
substrate activity increased slight1y with‘increasing pI of the multiple
forms.

Native polyacrylamide gel electrophoresis of the fractions separa-
ted by preparative isoelectric focussing is shown in Figure 23.. Two
protein bands were observed. which coincided with the pNP-a-fl-acetyl-
galactosaminidaseactivities (Figure 24). The mobilities of the forms
were a11 similar and thus no differences cou1d be shown by polyacrylamide
gel e1 ectrophoresis.

The K” values of three of the peaks with pNP-a-i-Lacetylgalactosami-
nides and Forssman hapten were Compared (Table 6). The K,1 values with
pNP-a-CalMc increased with increasing pI of the enzyme while the 1%

values with Forssman hapten as substrate were almost the same for the

three peaks.

. . Kinetic . Properties .of . the . p-Ni trophenyl eaegeacetyl ga1 actosuninide Activity

 

The hydrolysis of pNP-g-fi-acetylgalactosaminide proceeded linearly
for at least 60 minutes. when the final concentration of enzyme was
less than 0.6 ng/pl the enzyme activity was less than that at higher
concentrations (Figure 25). At concentrations above 0.8 _ng/,,1. the
enzyme activity was linear up to at least a concentration of 4 ng/pl.
The pH optium for the hydrolysis of pNP-a-fl-acetylgalactos‘aminide was
broad (Figure 26). The highest activity was at pH 4.5 and the range of
the optium was from 4.3 to 4.7. A lineweaver-Burk plot of UV vs US
is shown in Figure 27. The K! value was found to be 2.9 all! and vmax

as 14.1 nmoles/mi n/mg.

Figure 23.

92

Native Polyacrylamide Gel E1ectrophoresis of Isoe1ectric
Focussing Peaksof e-flfAcetylgalactosaminidase.

The enzymes were electrophoresed in 71 gel in tris-
glycine buffer, pH 8.3. The numerals indicate the
number of the enzymatic peak in Figure 22, starting
from the peak with the lowest pI.

93

 

 

 

 

 

Figure 23.

Figure 24.

94

Scan of Native Polyacrylamide Ge1 Electrophoresis of
the Isoelectric Focussing Peak of d-flfAcetylgalactos-
aminidase.

Scan.at 550 nm are for protein stains and 410 nm for
pNP-aggracetylgalactosaminidaseactivity stains. The
electrophoresis conditions are the Same as in Figure

23.

 

 

 

 

mmmmmmmmmm

 

96

.¢ eucogomum m .n .comgurscvz was .m .ccusmm*uz sag» sauna

 

Ps.o ca .. om.o
mm.o m.~P .. m~.m
mm .o m m6 55.c
sauna: casemcom u<zpaw ud-p»:u;nosa_:tn au<zpuwtaupacesn
“3.! . 2

3.2553828533... .3 econ. 2:38... 2.38.3: 32:2. 2: .3 8.23:2...» .e as:

97

.13: «.2 2! 9.55 we 5522028 2..—
.=o32u=3=8 255 9 omnuwcefimsunFunguognmnaéza no .3332 855 .3 95m:

2.: ms;sz do 53024
on 0d 0m 8 o.

I?

 

—

Low.

 

 

c3
(1)
(Au/salowUMu/umv BWAZNB

Lom.

98

$33 gone 5 new: we: gauze we on «.0
in m> 322338pampaaogumtauaza mo a.mapfv»: mo 33. .8 253“.

 

In
on om me od 5
.1 _ . _ a . O
- ow %
w
v
.. ow . .X
V
O
om H.
s m
om M.

 

 

00.

0'4 " PNP
0-3 -

0-2 -

I /v (min nmoles")

O-l -

 

\

 

0 0'2 0'4 06
US (mM")

Figure 27. Lineweaver-Burk Plot for the Hydrolysis of p-Nitrophenyl-a-
' 1-Acetyl gal actosaminide by a—l-‘Acetylgalactosaminidase.

The amount of Enzyme used for each assay was 0.54 ug.

Hydrolysis of .Forssman Hapten . by. O9fl-rACEty] gal actosaminidase

 

(I) CharaCterization‘of’ProdyLCts Of'Hy'df‘OUSB'gf Fortsman Hapten

The hydrolysis product of Forssman hapten by a-fl-‘acetylgalactosami-
nidase was characterized. The radioactivity'scan of the TLC'run of
the Folch upper phase after hydrolysis of Forssman hapten by d—_N_-acety1-
galactosaminidase is shown in Lane 1 of Figure 28. The radioactivity
had the same migration as the monosaccharide (Lane 6), indicating that
the radioactive production of hydrolysis in the Folch upper phase was
a monosaccharide which was expected to be fl-acetylgalactoSamine. The
heavy spot in Lane 2 turned yellow when sprayed by drcinol-sulphuric
acid and was shown to be identical in migration to sodium taurocholate.

The lower phase was analysed by TLC in the solvent system chloroe
form/methanol/water 100/4276. Lane 1 in Figure 29 shows the lower
phase of the control sample with no enzyme added. The radioactivity
scan of this lane is shown in scan a. When a-fl-acetylgalactosaminidase
was added to the incubation mixture. the mobility of the glycolipid
product increased and was similar to that of standard GL-4. The
radioactivity scan. shown in lane b. was lost from the glycolipid.

The results indicated that the [3HJ-GL-5 was labelled in the non-
reducing terminal fl-acetylgalactosamine. n-fl-Acetylgalactosaminidase
acted as an exoglycosidase. with the hydrolysis products identified as
(it-4 and l-acetylgalactosuine.

(gi'kjnetic's'o‘f'fly'drolfi' is 0f :orssmaniiap ten

The hydrolysis of Forssman hapten by a-fl-acetylgalactosminidase

 

was linear for at least 3 hours. It was linear with respect to enzyme

cOncentration except at very low cOncentrations (Figure 30) .The optimal

1'00

101

 

Froni -i -

 

 

 

Origin->1 5-
l 2 3 4 5 6

Figure 28. Radio-Thin—Layer Chromatogram and Thin-Layer Chromatography
of the Folch Upper Phase Product after Hydrolysis of
Forssman Hapten by a-fl-Acetylgalactosaminidase.

Lane 1: Radio-thin-layer chromatogran of Lane 2;

Lane 2: hydrolysis product; Lane 3: stachyose.100 :19:

Lane 4: raffinose. 100 ug: Lane 5: melibiose. 100 ug:

Lane 6: galactose. 100 tag. The solvent system was butanol]
inter/acetic acid 100/50750.

102

£55 2 an 33... 58.25 E. 2. x p an 533 :pm .333 2 an
ucfimcou 2.5 6:83. can 3.53 2 x . an 335—33 .3 an Sum $.82 «a .38

53:823. 53.: a P3 8:; 5 3:8, eaeazzeefiesceeeze 5 8.:

.p can; be noun au.>.uuoo*v~s “a as»;

"N ago; no scum xa—>*uucopuus ”a was;

338-. amp .3933...» midinzmu “m 2.3

3269:. mama—ecu? ”.5255 .3 32g .833 sup n_ “N «:3
Sobcou .3 32.. .333 cope“. up 23..

§o€§ are: .8...— muESoubn 23:3» 8 2:...

$335233:pegbogimia 3 :33: 5.520... mo «Expose»... gate «032..
one... .533 cope“. 05 mo 233355 $234-55. 93 533385 5.24.55-395

.mu oeam.m

103

.mu ossn.m

3.225485%

2.630

 

 

 

min—o
vii—0

who

 

bzomm

104

 

OJ
0

N
9

'25

 

 

 

GL-S HYDROLYZED (nmoles/hr)

J 4 L . l L

‘ 20 so 40' 50
VOLUME OF ENZYME (LII)

O _

IO

Figure 30. Rate of Hydrolysis of Forssman Hapten vs Enzyme Concentration.
The concentration of enzyme used was 2.4 ng/nl.

105

activity was at pH 3.9 (Figure 31). No hydrolysis of Forssman hapten
was found in the absence of detergent. lihen taurochdlatewas used. the
activity showed a maxinium at a final concentration of- 4 mg/ml (Figure
32'). The Liheweaver-Burk plot for the effect of change in substrate
concentration on the rate of hydrolysis was linear. The phenomenon of
inhibition at high substrate concentration was not observed. The KM
value of the enzyme was found to be 0.26 nii and the 11m was 4.2
mles/min/mg (Figure 33).

(3) Ihh‘l bitors ' 0f ' ForSSmah.‘flYdr01 ase ' Adj '1 1:]

letal salts at a cOncentration of 10 Iii were used to study the
effects of metal ions on the hydrolysis of Forssman hapten by d-fl-
acetylgalactosaminidase. whenever available in a soluble form, chloride
salts were used. The results shown in Table 7 indicate that all of
transition metals studied inhibited the hydrolysis to a certain extent.
the most potent being Ag”. n33". "92* and Pb”. Alkali metals did not
affect the enzymatic hydrolysis while alkaline earth metals were slightly
inhibitory.

lihen Cl' was used as an inhibitor with increasing concentrations of
NaCl from 0.01 14 to 0.1 14. no inhibition of the hydrolysis of 8L-5 was
observed.

01' various sugars and carbohydrate derivatives tested. only 11-
acetyl gal actosamine. 6alllAc-(d1+6)-1,2:3,4—di-0_-isopropylidene-6al . and
surprisingly. TalNAc-(nl'-n6)-l,2:3.4-di-0_-isopropylidene-Gal (Table 8)
were inhibitors. The hydrolysis of Forssman hapten was inhibited
capetitively by TalNAc-(a146)-l .2:3.a-di-§_-isopropyiidene-eai with a
K‘ of 1.5 an. The Lineweaver—Burk plot of the effects on'inh'ibition is
shown in Figure 34.. The results indicated the configuration of the

106

 

 

60 r l i
t O
4: o
B 50- o ‘
2 00
(I)
E. . ° 0 °
5 40» 0 ~
CD
a C
5 30" o .
O
93 i -
>_ 20 ‘
I o
m )- 0 an
___'_1 IO .
L9 .

 

 

3.0 35 ‘ 40‘ “ 4.5 50
pH
Figure 31. Rate of Hydrolysis of Ferssman Hapten vs pH.

Citrate buffer was used for the entire range and 0.24 ug I
of enzyme was used for each assay.

107

100-

80-

60'

40*

"la MAXIMUM ACTIVITY

20-

 

 

‘ 215‘ so a ‘ ‘ “.070
[TAUROCHOLATE ] MG/ML

Figure 32. The Effect of Taurochol ate Concentration on the Rate
of Hydrolysis of Forssman Hapten

0.24 ug of enzyme was used for each assay.

108

   

KM=ZI3 xi0‘4 M

  

05
04
.l.
V 03
(hr nmolé')

 

0 1001 002 003 004

[$1 ()1 M")

Figure 33. Lineweaver-Burk Plot for the Effect of Change of Substrate
Concentration on the Rate of Hydrolysis of Forssman Hapten.

A standard assay procedure wdth 0.36 ug of enzyme was used
for each assay.

109

Table 7. lite Inhibitors of ForSsman Hydrolase

 

.heta1 . (10.1104101' .‘ salt) .Z'.’Act'iv1tyl
89* 1 o
Ba2+ '86
Ca2+ so
edz“ as
Co2+ 59
Cu2+ ' 35'
“3+ 7
it+ 101
1.1“ '1'08
"92“ 25
1192* so
Mn“ 69
11114“ 103
Na+ 100
Pb2+ 31
Sr2+ 84

110

 

 

a... 2 258
R e Eben...Seasiaeeéafi. 19.3.35.
3 e .m 3.28. paaeeea...o..%-e.n&. 19.3 -323

3:52 a $3 . 6:8 $392.28 muse—.2339

3223. 5.52.: ea 5:32.: e5 5 85:28 32.5.3.8 .e as:

111

 

 

0-6 - .18
0'5 *-
{7
33 0.4 ..
o
E
C2 o
._ 0'3 *-
c 0
2»
_\_ 0-2 - 0
0| - 1°
, --1 -1- s I
O O l 2 3
US (mM")

Figure 34. Lineweaver-Burk Plot of Hydrol sis of Forssman Hapten
with H-Acetyltalosaminyl-(0146 -1.2:3,4-diegyiso-
propyTidene-Galactose as an inhibitor.

0.24 ug of enzyme was used for each assay. 18’ 8 mM
inhibitor: 10’ no inhibitor.

112

entire geacetylgalactosamine residue was important for inhibition,
except for the .C2-position. where an epimer could also inhibit the
hydrolysis of Ferssman ~ hapten.
(41‘ Temerature ' Inacti vati 0h ' 0f 'forssman ' Hapten 'dei‘dl a'se
The enzyme activity of d-fl-acetylgalactosaminidase on ForSsman
hapten decreased very rapidly at 60° (Figure 35). Half of the'activity
was lost after incubating the enzyme at that temperature for about
1.4 minutes. The activity reached zero after 20 minutes of incubation.
At 50°, the decline in activity was not so rapid. A semi-log-
aritilnic plot of the percent activity remaining vs time did not give
a straight line, implying that the enzyme activity might have existed
in more than one species. which showed different degrees of stability
toward temperature inactivation. The half life of the total enzyme
preparation was. about 6 minutes at this temperature.

Hydrolysis .0f Forssman Pentasaccharide . by . aegeAcetyl gal actosami nidase

 

(.1) Characterization of 51.-5 Pentasaccharide

The product after ozonolysis. hydrolysis and P-2 chromatography
showed a product not visualized by iodine. indicating that it was not
lipoidal in nature. The spot turned pink after orcinol-sulphuric acid
spray. and migrated below the standard stachyose. a tetrasaccharide
(Figure '36). GLC analysis of the methanolysate of an aliquot of the
carbohydrate product showed that it is composed of galactose. glucose
and H-acetylgalactosamine in the ratio of 2:1:2. The mode of formation
of the capound with ozonolysis and alkaline hydrolysis. combined with
the above results. characterized the product as GL-S-pentasaccharide
with the structure:'8alNAc-(dl-v3)‘-8alHAc—(a'l43)-8al-(dl+4)-8al-(al+4)-81c.

113

Figure 35. Semi~1ogarithmic Plot of Temperature Inactivation of
ForSsman Hapten Hydrolysing'Activity vs Time.

The enzyme was incubated in a water bath and aliquots
were taken out at different time intervals and assayed

for Forssman hydrolysing activity.

114

100

50°

p _
O O O 5 3

oz_z_<$_mm >._._>_._.o<

60°

o\o

‘10

 

3O

20
TlME(min)

Figure 35.

From-t“ I 4":

Origins

Figure 36.

115

 

 

..._...._... ._. J
A B c 0 E

Thin-Layer Chromatography of Forssman Pentasaccharide.

   

 

e.|||| l Il|a

 

 

 

GL-S Pentasaccharide (Lane 8) was scanned with a Berthold
radio scanner (Lane A) and compared with the migrations of
stachyose (Lane C), raffinose (Lane 0) and melibiose (Lane
E) in butanol/methanol/acetic acid 100/50/50 on silica

gel 6 plates. The spots were visualized with orcinol-
sulphuric acid. ’

(2) Characterization of Products after'Hydrolysis 0f'GL45 Pentasaccharide

with aeH-rAcetyl galactosaminidase

_L.u_.J_A

The hydrolysis products after the action of’o-flyacetylgalactosamini-
dase on GL-S pentasaccharide are shown in Figure 37 (Lane 8). There
were orcinol-positive spots that migrated just a little below the
standard stachyose. A faint spot charred by sulphuric acid can be
found migrating in the region of the standard monosaccharide_(ga1actose).
This monosaccharide spot coincided with the liberated radioactivity
(Lane A). Very 1itt1e radioactivity remained in the tetrasaccharide-
pentasaccharide region. The results showed that the radioactive fly
acetylgalactosamine at the non-reducing end of GL-S petasaccharide was
liberated by the enzyme with the globoside-tetrasaccharide remaining
intact.

‘ ' (31' kinetics 'of‘ Hydrolysis; of ' 813-5 ‘ Pentasac'Charide

The hydrolysis of GL-S pentasaccharide was linear for at least
‘90 minutes with 1 119 of enzyme per assay. The hydrolysis was liear with
enzyme concentration up to 1.6 ng of enzyme per assay (Figure 38). The
K" was found to be 10.4 mH and the vmax was 1.7 nmoles/min/mg (Figure
39). The addition of taurocholate to a final concentration of 4 mg/ml
inhibited the enzyme activity by 132.

Hydrolysis of Hr11r14CJrAcetyl sphingosylrForssmanepentasaccharide

42.44 __ _._ _._L

(1) Characterization of Hrlfil-rMCJrAcetyl sphingosylrForssman-rpenta-

LLL

vvaui-V‘W-V fl V

saccharide

The mobility of the alkaline degradation product of GL-s by NOH‘is
shown in Lane :i-oir Figure 40. The deacy1ated'caupound. when re-fl-ace-
tylated (Lane 2) by [1-1‘C]eacetic anhydride. migrated faster than the

116

Front > v * i

Origin-s

Figure 37.

117

 

 

 

 

 

 

Thin-Layer Chromatography of Hydrolysis Products Obtained
after a-H-Acetylgalactosaminidase Treatment of Forssman
Pentasaccharide.

B: hydrolysis product of a-N-acetylgalactosaminidase on
GL-S pentasaccharide: A: radioactivity scan of Lane 8:
C: stachyose: D: raffinose: E: melibiose: F: galactose.

The chromatography conditions were the same as Figure 35.
25 moles of substrate was incubated with 0.47 119 of
enzyme for 12 hours.

ACTIVITY (nmoles/hr)

 

Figure 38.

 

118

5 IO I5
VOLUME(,u.I)

The Activity of l-Udrolysis of Forssman Pentasaccharide vs
Amount of Enzyme.

The concentration of enzyme was 0.11 ug/ml. Incubation
was for 30 minutes.

119

  
 

1-0 L _
GL 5 OLIGOSAC.
Km' '0'4 mM

 

 

 

‘59 0-8 I-
CD
a
= 0-6
2
S 0-4

0-2

0 l 2 3 4 '5
US mm")

Figure 39. Lineweaver-Burk Plot of the Hydrolysis of Forssman Penta-
saccharide by o-fl-Acetylgal actosaminidase.

0.24 mg of enzyme was used for each assay.

120

.Am «away 5 ecu; we soon

auv>vuufimvuoe as» use mxmxmcxmm sougzxvvun u.uouoxpocm;uesseeoposopgu c— miaw
capepaumumizimeivmuapxuuov as» can an oceav —a.euuge_mcvusoum ea ugh mg: a meme;
.3143 $232. 9.3.33 3 0:3 .mid so: 33:3»; 23:3..." 353:8

m use _. 35.. .mkexobp soughbcafiosxenomoeazu no: Emu—mam "—52.3. 0.:

.AN acmdv migu vuuepauoumiziez ago an ecugv wouupauaev mo cap on» was < usage

.outmsuummmucmaicammeomi339.23..33218315....“ mo 55323022... .3 0.53...

.o¢ ogzmwm

 

 

121

 

11"-

W

 

.m-.._o
who
T40

 
 

 

 

122

deacylated GL-S but slower than a GL-S standard (Lane l, 5) and the
starting material (Lane 4). The fl-[l -"C]-acetyl-sphingosyl-GL-5-
pentasaccharide was further purified and the product (Lane 7. Figure 40)
was chromatographyically homogeneous on TLC using the solvent system
chloroform/methanol/water/acetic acid 55/45/5/5; its mobility coincided
with the radioactivity (Lane 8) and was a little below that of the
starting material (Lane 6). The carbohydrate composition of the com-
pomd was analysed by GLC of the trimethylsilylated methyl glycosides.

A ratio of 2:2:l for galactose:[i_-aceiylgalactosamine:glucose was found.

(2) Characterization of Hydrolysis Products from a-g-Acetylgalactosamini-

dase Treatment of _N_-[l:flt]-Acetyl-sphingosyl~Gl.-S-pentasaccharide
lihen fl-[l~fiCJ-sphingosyl-GL-5-pentasaccharide was hydrolysed by

 

' a-fi-acetylgalactosaminidase. a lipid band (visualized by iodine vapor)
was produced that migrated faster than the parent compound (Figure 41).
A radioactivity scan showed two radioactive peaks (Lane 5), the faster
of which corresponded in migration to standard fl-acetylgalactosamine
(Lane 6). The second radioactive peak had the same migration as the
iodine-posi tive band. There were no other iodine-positive bands except
the one that corresponded to taurocholic acid. GLC analysis of the
iodine-positive product showed that it contained galactose. glucose
and g-acetylgalactosamine in an approximate ratio of 2:1:1. These
results provided evidence that the hydrolysis products were [L-acetyl-
galactosamine (the faster radioactive band) and fl-[l-"Cl-acetyl-sphin-
gosyl-GL-ll-tetrasacdiaride (the slower radioactive band).

123

.3550 2.0 .335 05 08 now: 0.33 000.509... .3 «035: cm— .Afleaga 00...: 5
0050.:— 30—5 03:03.? no: 0003 002022.33 05!: 5:. 0003.03; 23 0000:.
w _.0a 3:3. =0 m\m\mv\mm guanigua 0339\3550533235 5..) 33 u._._.

.p 0:2 05 50m
3.. 0:2 .3 :30
to. 0...: no Sum

“05:0 035:. 30.53:»
50:09:. 310.322..
«05503020350072 0: 8—

pun-tuna

.08 E
-Eomouuflump502ouaé W3 0vIufiuamscaégmflou-303:23Ab0ufnuznfi.mvuo
«Wax—Eu»: .3 30.69:. 05 we Scum 5330325. 23 202032.25 5033-5...—

. 3 026 w...

124

C. C

Wcoam

.F¢

0L=mvm

 

 

 

 

 

 

 

 

 

 

 

(3) Kinetics of the Hydrolysis 3f [[-[lJ4CJ-Acetyl-sphingosyl-Forssman
‘ Elntasacgaride

The hydrolysis of 1-[1-“C]-acetyl-sphingosyl-Forssman-pentasaccha-
ride was optimal at a final sodium taurocholate of 4 mg/ml (Figure 42).
The optimun was much sharper than the sodium taurocholate requirement
curve for the hydrolysis of Forssman hapten (Figure 32). Hydrolysis of
this substrate was linear for at least 30 minutes when 0.27 ug of the
enzyme was used for each assay. The hydrolysis was linear up to 1.08 as
of enzyme used for each assay (Figure 43). There was no apparent lag
of activity at low enzyme concentration. probably because the total
volume of the assay was smaller and consequently the final concentration
of the enzyse was higher. The K” value was found to be 2.5 x lo"4 M for
this swstrate and V max was 0.81 umles/min/mg(Figure 44). No substrate
inhibition was found in the Lineweaver-Burk plot of the action of

a—fl-aoetylgalactosaminidase on this enzyme.

Hydrolysis of Porcine Submaxillary Mucin by a—fi-Acetylgalactosaminidase

 

The lit-porcine submaxillary mucin was prepared by a method described
previously (64). The product liberated from the mucin by wig-acetylga-
lactosaminidase was identical in retention time on GLC to an authentic
trimethylsilylated fi-aoetylgalactosamine samle. The hydrolysis of
porcine sibmaxillary mucin was optimal at pH 4.2 (Figure 45). At this
pH the hydrolysis was linear with respect to time for at least 60 minutes
when 2.l7 ug, of enzyme was used for each assay. The activity was linear
up to 5.34 ug of enzyme added per assay when 30 minutes incubations were
used but there was a lag in the hydrolysis when the amount of enzyme used
was below l pg per assay (Figure 46). The KM of the enzyme in mg/ml was

125

126

Figure 42. Detergent Dependence in the Hydrolysis of N-[i-"CJ-
Acetyl-sphingosyl-Forssman-pentasaccharide"by a-fl:
Acetylgalactosaminidase.

The amount of enzyme used for each assay was 0.27 ug.

% MAX. ACTIVITY

127

  
   

lOOr-
De-Acyl-GL-S

80-

 

 

 

o)w A lo 26‘
[TAUROCHOLATE]
(mg/m0

Figure 42.

128

 

 

 

A80?

E

\

0‘)

260*-

o

E

C

>-40r-

t:

2

l— 20.-

o

<
O A- ,l... l I J
0 IO 20 3O 4O

AMOUNT OF ENZYME(/J.l)

Figure 43. Hydrolysis of fi-[l-”C]-Sphingosyl-Forssman—pentasaccharide
by Varying Amounts of o-fl-Acetylgalactosaminidase.

The concentration of enzyme used was 27 ng/ul.

129

    

 

 

.20 De-Acyl-GL-5 °
1“
2 'l5
C)
E
c:
g, '|O
.c:
> /
> -05 ..
1 We 1 WW. I l a .l.
O 2 4 6 8 IO
I/S(mM")
Figure 44. Lineweaver-Burk Plot of the Hydrolysis of flr[l-I4C]-Acetyl-

Spaingosyl-Forssman-pentasaccharide by a-flyAcetylgalactosami-
n ase.

The amount of enzyme used in each assay was 0.27 ug.

130

[00

80

60

4O

20

"/0 MAX. ACTIVITY

 

 

. L a l . l

(:) . J
39 4| 4-3 4.5 4-7

Figure 45. Effect of pH on the Hydrolysis of Porcine Submaxillary Mucin.

2.l7 ug of enzyme was used for each assay and incubation
was at 37° for 30 minutes.

131

 

TIOO"
.C:
E
280'
‘e’
560“
>.
:40“-
._>_.
l”"20"
0
<
C) - I

Figure 46.

 

 

05 IOU IS 20 25
AMOUNT OF ENZYMEHLI)

Hydrolysis of Porcine Submaxillary Mucin with Varying
Amount of aeflrAcetylgalactosaminidase.

The concentration of.enzyme‘was 0.22 ug/ul. Incubation
was at 37° fbr 30 minutes.

132

found to be 2.31 (2.3 x 10-6 M) and the \imax was found to be 0.42 umoles
min/ug (Figure 47).

Hydrolysis of GalHAc-(ol+6)-l,2:3.4-di-9_-isopropylidene-Gal

 

The substrate accounted for only 60% of the weight of the compound
when quantitated by GLC, and on TLC, it was not homogeneous. One major
and two minor bands were found (Figure 48. lane 2). Hhen the diisopro-
pylidene group on the coupound was removed by trifluoroacetic acid,
(Lane 3), the same pattern was found, with one major and bio minor
bands. The major band coincided with the mobility of a disaccharide.
while the impurities migrated in the monosaccharide and trisaccharide
regions. respectively. Analyses by GLC of the methanolysed trimethyl-
silylated methyl glycosides of GalNAc-(ol-r6)-l ,2:3,4-di-_0_-isopropylidene-
galactose (I) and GalHAc-(ol+6)-Gal (II) are shown in Figure 49. The
ratio of galactose to fi-acetylgalactosamine was approximately lzl.

One of the products obtained from the enzyme hydrolysis of GalNAc-
(ol+6)-l,2:3.4-di-_0_-isopropylidene-galactose was identical in retention
time to an authentic sample of [acetylgalactosamine when the trimethyl-
silylated derivatives of the hydrolysis products were anlysed by GLC.

The Lineweaver-Burk plot of the hydrolysis of the disaccharide by
o-H-acewlgalactosaminidase is shown in Figure 50. The K" value was
4.8 x 10"3 H and the me was 1.1 umoles/min/mg protein.

The linearity of activity with respect to enzyme concentration and
time was not examined. The hydrolysis of the substrate. however.) was
limited to less than l0: of the total initial substrate present in the

assay.

133

 

 

_:~ -|2 . PSM
'2

O

E 08 -

E

> -04 -

_\..

Mo A) :0 A 52-0 30‘

US (ml mg" or FM")

Figure 47. Lineweaver-Burk Plot of Porcine Submaxillary Mucin
Hydrolysis by odflyAcetylgalactosaminidase.

2.l7 ug of enzyme was used for each assay.

134

Figure 48. Thin-Layer Chromatography of H-Acetyltalosaminyl-(ol-r6)-
l,2:3,4-di-_0_-is0prop lidene-Galactose (lane l), fl-Acetyl-
galactosaminyl-(ol-v6)-l ,2:3.4-di-0-isopropyl idene-
Galactose (lane 2) and _N_-AcetylgaTactosaminyl-(al+6)-
Galactose (lane 3).

The solvent system used was butanol/acetic acid/water
l00/50/50. 100 ug of sanple was used for each compound.
The standards used were stachyose (lane 4), raffinose
(lane 5), melibiose (lane 6), and galactose (lane 7).

135

 

Figure 48.

Figure 49.

136

Gas-Liquid Chromatography of Trimethylsilylated Methyl
Glycosides of Methanolysed HrAcetylgalactosaminyl-(al+6)-
Ga actose, flyAcetylgalactosaminyl-(ol+6)-l.2:3,4-di-Q:iso-
prooylidene-Galactose and N-Acetyltalosaminyl-(ul46)-l,2:
3,4-di-Qrisopropylidene-Galactose.

They are shown in I, II and III. respectively. GLC was
performed isothermally at l85° and the samples were
chromatographed on 3% SP-2l00 on Supelcoport (80-l00 mesh)
with 6 ft columns using nitrogen at 45 ml/min as carrier
gas. The peaks are A: galactose; 8: mannitol; C: Hyacetyl-
galactosamine and peaks in D are probably firacetyltalosamine.

RESPONSE

DETECTOR

137

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W Unfit II.
I. ULfl D II

 

 

 

 

 

I
J a. W

O 4 8 l2
RETENTION TIME (min)

;

 

Figure 49.

138

.ma om.p no: 50000 guo0 a? 000: 0sx~c0 mo ucaoso 0:»

.0movremanmouuapomhau00<umio so 000000peoi0c0vvpxaocaomwumivu
imam.78:3-EeEooBozomEooTz to $522: one co do: {32.28385 .8 2.3:

3.25 m:
0.0 v.0 m.o NO _.0 O

u A u q q

 

     

l
N ._
<9 <9

1
if)
C?

(l_a|owu Jq)A/|

 

Hydrolysis of Human Red Blood Cell
and Dog Intestinal AT-Glycosphingolipids

When the human RBC blood group A-active glycolipid (lane l) was
hydrolysed by the enzyme. TLC gave an iodine- and orcinol-sulphuric
acid- positive band between the taurocholic acid and blood group A+
band (lane 2, Figure Sl).

Human red blood cell A2 glycolipid (lane 4, Figure Sl) was hydro-
lysed by aefl:acetylgalactosaminidase to give a product which migrated
on TLC higher than the starting material (lane 3, Figure Sl).

Human red blood cell A3 glycolipid (lane 5) was also hydrolysed
by the enzyme to give an orcinol-sulphuric acid- positive band (lane
6) on TLC.

Dog intestinal AT-glycosphingolipid (lane 8) was hydrolysed by
the enzyme to a product which reacted positively on TLC with orcinol as
well as iodine and which migrated slightly higher than the starting
material .

139

140

.oWoo oWPoeooeooe a. eo_c_oeoe. no: o.o am o.

0000 00:00 0:» .m0:.um0a=. not act» mu»:»»00»»m aaocu uoo»:-e< «a: uni»: mm o» » sou»
xu»—»:os mcpmo0cuoc :0»: »»00 000»: 00: cuss: soc» mu»n»»oua»m < 030cm woo»: 0L0: mien:
e5 .39. .709. 082553323Sofia... 5:. $5222 tote 32.8.3 oz:
0:» «000:0: a van »o»:0uoe m=_ucoum m0uo=0u m .v»uo u»:=:a»=m-»oc_uco x: e0~ppo=mp>
no: :0»:: .miumx un0ux0 goao> 0:»00» a: u0~»»u:m»> 0:03 0000—: 0:» .m\m\m¢\mm 00002
\uWuo u~a0uu\»o=u:u0e\zcomoco»:0 :u»x_m _0u ou»»»m co v0:aacmaageoc:u 0:0: 00—3500 0:»

.0muu»c—Eumouuu»ampau0u<umro a:
muvnw»om=»:nmoux»w < aaocw woo—m mo m.m»»ocuxz mo 0003009: co xgaucaouoeoc:u :0»«4-=.:»

.pm oeso.a

141

 

 

RBC—2

 

 

 

 

 

Figure 51.

DISSUSSION

Structural Determination of Forssman Hapten

The structure of Forssman hapten‘was shown to be smilar to that
found in horse spleen (3), sheep erythrocytes (l6) and goat erythrocytes
(18) but different from Forssman-positive tetraglycosylceramide purifi-
ed from NIL cells (20). These structural studies involved permethyla-
tion of the intact glycolipid by Hakomori's procedure (194) and sub-
sequent reduction, acetylation and determination by GC-HS by the
Lindberg technique (195) for the assignment of linkages. This per-
methylation technique is a very versatile procedure and has been used
for linkage studies of glycolipids (2,3,20,203), oligosaccharides (213,
2l4). glycopeptides (215.216). and bacterial (195) and yeast (217) cell
walls. Not only will the procedure indicate the linkages of the saccha-
ride units present in an oligosaccharide. but it also gives valuable
information regarding the ratios of different sugars with different
linkages and the homogeneity of the sample. The stationary phase for
GLC is crucial for the best resolution of the various partially
methylated alditol acetates present in the sample. especially with some
biological samples that are exceedingly complex. An oligosaccharide
with only four kinds of sugar linked in 1,2 or 3 positions can give
up to 44 peaks in a single GLC chromatogram. An example of the com-
plexity of the chromatograms is illustrated in Figure 51, which is
[a GLC analysis of an accumulating decasaccharide in the urine of fu-

cosidosis patients (214). The decasaccharide contains fucose, mannose.

142

143

A

 

 

 

 

 

 

 

 

m . ALDITOL ACETATES FROM ”

‘2 PERMETHYLATED GLYCOPEPTIDE

o

O.

(0

LL]

0:

0:

O \I

'- W

C)

LIJ

*-

LIJ

O
.__l_ L L 1
IO 20 30

RETENTION. TIME ( min)

Figure 52. Gas-Liquid Chromatography of the Partially Methylated
Alditol Acetate Derived from the Purified Fucodecasacchari-
de A from the Urine of Fucosidosis Patients.

The sugar derivatives was analysed on a 2 mm x 2 m cslw
with 3% 0V- 210 on Chromosorb Q (100- 120 mesh) at l25

250° programmed at 2°/min with nitrogen at a flow rateto
of 45 ml/min.

144
galactose and gracetylglucosamine.

Identification and quantitation of the individual peaks can be
very difficult because of the inability of liquid phases to completely
resolve the individual partially methylated alditol acetates. Single
scans of the peaks sometimes give mass spectra of mixtures, which are
very confusing and may lead to erroneous interpretations. This problem
was resolved by taking repeated mass spectral scans of the GLC eluant
at intervals of about 4 seconds and storing the data in a POP/81 com-
puter. When the GLC run was completed, the data were displayed by
monitoring specific ions characteristic of different alditol acetates,
usually at m/e 45, 59, 71, 87, 89, 99, 101. 113, 115, 116, 117, 127,
129, 131. 139, 143, 145, 161, 175, 189, 203. 205, 233, and 261. The
powerful resolution of this method can be illustrated by Figure 53,
where ions at m/e 189 and 117 were monitored along with the total ion
intensity, which approximates the detector response of GLC analysis.
The ion at m/e 189 is characteristic of 1.2-substituted hexopyranosides
while the ion at m/e 117 monitors l-substituted, Z-unsubstituted hexo-
pyranosides. By comparing the two mass chromatograms, it can be
seen that the peaks at scans 112 and 117, indicated by arrows. have an
ion at mVe 189, but have little intensity at m/e 117, while an extra
peak at scan 114 can be found in the chromatogram for m/e 117. The
peak at scan numbers 112 and 114 were not resolved in the total ion
intensity chromatogram and exist as a broad peak.

The ratio of each of the peaks can be quantitated by measuring
the areas of the peaks for a specific ion and multiplying each peak by
a factor which is determined by the amount of that specific ion formed

from a particular alditol acetate.

145

2-suesn'rureo
HEXOSE . 1L

We.

 

CHéMm
HCOAC
189 HCOMe

OMe
H OAc

Hylh

HCLOMe 189

 

 

INTENSITY
I
a
O
3:
I

 

 

 

RELATIVE

 

 

 

SCAN NUMBER

Figure 53. Mass Chromatograms of the Partially Methylated Alditol

Acetates of Decasaccharide A from the Urine of Fucosidosis
Patients.

The mass spectra of eluants of the GLC run of Figure 50
were taken at 5 second intervals and displayed for relative
ion intensities of'mve 189 and 117. The lower scan is

a chromatogram of the relative intensities of the total ion
in each mass spectrum.

146
The most oomonly used liquid phases for the GLC analysis of

partially methylated alditol acetates is ECNSS-M (l95). This liquid
phase is not stable at temperatures above 200°C and various alternative
polar liquid phases have been introduced. They are 3% OV-ZlO. 31 0V—-
225, 3% 0V-l7 and EGS/EGA/XE-éo (0.2/0.2/l.4%). Analysis on 0V-225 has
been found to give a low yield of partially methylated aminosugars

and is therefore not useful for general oligosaccharide analysis. We
favour the use of OV—Zlo because this liquid phase gives a reasonable
resolution of the aoditol acetates, is thermostable up to 275°C, does
not absorb aminosugars, and can be used for general analysis without
a change of column. For higher resolution of disubstituted hexopy-
ranosides a polar liquid phase mixture consisting of 0.2% EGS/O.2% EGA/
l.4% XE-GO has also been used with success (2). 0V-l7 has been utilized
primarily for the analysis of partially methylated alditol acetates of
amino sugars (2l8).

Three methods can be used to determined anomeric configuration.
Stereospecific enzymatic analysis is the most useful because precise
anomerity and sequence assignments can be obtained with microgram
amounts of material. Purification of specific glycosidases for these
investigations can be very tedious, however. Anomerities can also be
determined by cma oxidation (2l9) or proton magnetic resonance (213).
These two methods are more appropriate for simple oligosaccharides
because only ratios of a and 8 sugar residues can be obtained for one
species of monosaccharide present in the compound while the exact
location of the a— or 8- linkages cannot be determined. The proton
magnetic resonance method suffers further from the disadvantage that

mg amounts of material are required for the determination.

Purification of a-fl-Acetylgalactosaminidase

 

A major difficulty in the purification of a-fl-acetylgalactosami-
nidase from porcine liver lies in the lability of this enzyme at low
protein concentrations. Losses of 60-801/day at 4° is not uncomnon
at protein concentrations below 0.1 mg/ml (9). The strategy in the
purification scheme was to enploy large-scale preparations in such a
way that protein concentrations could be maintained above 0.l mg/ml,
even when highly purified. Steps that would result in low protein
concentrations were carried out quickly and the enzyme activity was
pooled and concentrated inlnediately after the fractions were assayed.
Extensive efforts by Heissmann and Hinrichsen (9) to purifiy the
enzyme further with conventional methods failed to yield a preparation
with a specific activity greater than ”50 mU/mg of protein. To
refine their purification sheme further, the use of affinity chroma-
tography seemed to be a plausible solution. The problem of finding an
appropriate adsorbent for coupling to a chromatographic matrix was
solved when the disaccharide GalNAc-(al+6)-Gal was synthesized by the
reduction and fi-acetylation of the coupled product of the nitrosyl
chloride adduct of galactal with l,2:3.4-di-isopropylidene-galactose
(Figure 3). when this disaccharide was coupled to dextran-coated
glass beads. a-fl-acetylgalactosaminidase bound to the affinity adsor-
bent with such tenacity that the enzyme activity could not be eluted
with 0.1% Triton X-l00. 0.2 M potassium borate buffer, pH l0.0. or 50
mil GalNAc in 0.05 M potassium phosphate buffer, pH 7.0. Using a pH
gradient of 0.l H disodium citrate and 0.1 _N_ hydrochloric acid, the
activity was finally eluted at pii 3.2. At this low pH the enzyme was
rapidly inactivated and the method was consequently discarded.

147

148
Nonspecific affinity chromatography on Con A-Sepharose, however,

proved to be very useful in the purification of a-flracetylgalactosami—
nidase. The enzyme preparation from DEAE-cellulose was purified 30-
fold by this step with about 50% recovery. The discovery by Leaback

and Robinson (220) that B-hexosaminidase A could be separated into a..
and 8- forms by elution of the enzyme from CM-cellulose by ampholytes
prompted the use of a similar ampholine elution procedure to purify
a-flfacetylgalactosaminidase, and the result was a further increase in
specific activity of the enzyme preparation by two-fold. The

powerful resolution of hydroxylapatite column chromatography was
illustrated by the work Tulsiani, Keller and Touster (22]). A
chromatographically and electrophoretically homogeneous s-glucuroni-
dase preparation was separated into eight enzymatically active peaks

by this procedure. Indeed when this method was applied to the purifi-
cation of a-fl:acetylgalactosaminidase, various shoulder similar to

those found on the profile of ampholine elution of a-flracetylgalactosami-
nidase from DEAE-cellulose (Figure l3) was observed (not shown in data),
and the preparation from this column was almost electrophoretically
homogeneous.

The previous purification of a-fl:acetylgalactosaminidase from
porcine liver was not successful in removing the contaminating 8:37
acetylgalactosaminidase activities. This would cause errors in the
interpretation of results if the enzyme preparation were used in the
study of anomerities of gracetylgalactosamine residues on oligosacchari-
des. The enzyme preparation in this work did not contain any detecti-
ble B-flfacetylgalactosaminidase activity and is therefore more appro-

priate for structural work. Further improvement in the purification

149

scheme to remove the last traces of impurity is necessary, but it also
risks the danger of further diluting the enzyme and resulting in in-
activating the enzyme activity and subsequent decrease in the specific
activity of the enzyme preparation.

Phenylmethylsulphonyl fluoride was introduced into the purifica-
tion scheme to eliminate the possibility of forming artifactual isozyme
ferns by the action of proteases. With the use of this protease
inhibitor, the three most acidic forms were attenuated in size but
never eliminated, indicating that protease action may not be the sole
source of the multiple isoelectric focussing forms. The acid preci-
pitation step was performed in as short an interval as possible so that
the possibility degradation by the lysosomal enzymes as well as

inactivation or acid hydrolysis can be reduced.

Characterization of

the Kinetic Properties of asu:Acetylgalactosaminidase

 

In this work. a-flracetylgalactosaminidase from pig liver was
purified to near homogeneity so that more insight could be obtained
into the kinetic and structural aspects, substrate specificity and
mode of action of the enzyme on glycolipid substrates. This enzyme
preparation was able to act as an exo-a-fl:acetylgalactosaminidase on
all the potential substrates examined, namely pNP-a-flracetyl-Q-galac-
tosaminide. Forssman hapten. fl:acetyl-sphingosyl-Forssman pentasaccha-
ride, Forssman pentasaccharide. human blood group A glycolipids, dog
blood group A glycolipids. porcine submaxillary mucin, and fl;acetyl-
galactosamine-(al+6)-l,2:3.4:-di-Q;150propylidene-galactose. The

kinetic parameters were examined except for cases where availability

150

of the potential substrate was the limiting factor.

The hydrolysis of a glycolipid substrate by porcine a-fl:acetyl-
galactosaminidase is best examined by comparing the hydrolysis of
Forssman hapten, !:acetyl-sphingosyl-Forssman pentasaccharide and
Forssman pentasaccharide. Forssman hapten has a large hydrophilic
head group with five sugar residues and a forked hydrophobic tail
group made up of the sphingosine and the fatty acid chains. No
hydrolysis of this molecule by the enzyme preparation was found without
the use of detergents. The same detergent requirement was also observed
with fl:acetyl-sphingosyl—Forssman pentasaccharide, which had the same
head group but a tail group with only one prong made up of sphingosine
while the other prong was replaced by a short acetyl group. The
optimum taurocholate concentration was much sharper for the latter
substrate, and this optimum fell approximately on the critical micellar
concentration of taurocholate. When the hydrophobic tail was eli-
minated altogether by ozonolysis, detergent not only was not necessary
for catalytic activity, but in fact it inhibited the hydrolysis of the
Forssman hapten pentasaccharide. The comparison of Vmax and KM values
indicated that this enzyme preparation had a higher affinity for the
lipids than the oligosaccharides when taurocholate was present (Table
9). The results imply that the detergnet plays an important role in
forming a mixed micelle with the Forssman hapten so that the substrate
is in a dispersed state that is suitable for formation of an enzyme-
substrate complex. At a concentration of 4 mg/ml taurocholate and
l50 nmoles of substrate, the ratio of substrate to number of micelles
(micellar weight = 2200 (222)) is approximately 1:2. There is an

excess of pure micelles and thus at substrate concentrations below 1 mM

151

 

 

-- N.e Ne.o o-o~ x m.~ . :Puaz xgmppwxasnam ucpueoa
-- -- P._ m-op x m.¢ Fau-aenm-Pu-¢.m"~.F-Am+PuV-u<zFaw
-- m.¢ _.ep mqu x m.~ u<zpmwuaupacmsaoepvzua
oureuguumm
e .. Fm.o «lop x m.~ -mueoalmqu-Famomcpgamuqumu<
.. .. ~.F «no, x o.p moweuguuumeuzmm mudu
e a.m ~.¢ euo— x m.~ mudu
ape\msv Ams\c*s\mmpo24v sz
ucwuempoa Eaewuao :a er> xx muscumnam

an mmuaeumgam acmemeevo mo mmmzposv»: ecu mo

.mmacvcvsumouumpumqumu<umra
memuosmemm ovumcrx ms» mo comweuaeou .m «Fame

152

the enzyme activity obeys Michaelis-Menten kinetics if the enzyme
activity is proportional to the mixed-micellar concentration (the
micelles that contain both Forssman glycolipid and taurocholate mole-
cule are mixed micelles) which in turn approximates the substrate
concentration if the equilibrium between monomeric and micellar
Forssman hapten favours the formation of the latter form. This is
indeed the case for most lipids and detergents (222). The stimulation
of taurocholate on Forssman hapten hydrolysis can hardly be explained
by this simple micelle picture. The micellar weight is too small to
be able to provide a surface on which the enzyme can adhere unless the
mixed micelles form large globular micelle clusters of secondary mi-
celles (223). The relatively sharp taurocholate optimum at 4-5 mg/ml
fbr‘fl:acetyl-sphingosyl-Forssman pentasaccharide is not understood.
The possibility of appreciable activation of the enzyme by its direct
interaction with the detergent can be ruled out by the inhibition of
the hydrolysis of Forssman hapten oligosaccharide in the presence of
taurocholate at 4 mg/ml. The gradual decrease in activity to hydrolyse
Forssman hapten at taurocholate concentrations beyond 7 mg/ml probably
represents increasing inactivation of the enzyme at the higher deter-
gent concentrations. Substrate inhibition (224) similar to the action
of sphingomyelinase on sphingomyelin in Triton X-lOO has not been
observed.

The action of this enzyme seems to be better for glycolipids in the
presence of detergent than for hydrophilic substrates because it hydro-
lyses Forssman with a much smaller K".

The KM for phenyl-oaflracetylgalactosaminide (6.4 mM) compares
favorably with the KM of p-nitrophenyl-a-§:acetylgalactosaminide found

153

in this work (2.9 m). Both artificial substrates are aromatic glyco-
sides and their KM values are in the same order of magnitude. The pH
optimum for the hydrolysis of the former substrate was found to be
at pH 4.3 as compared to pH 4.5 for the latter. This difference may
have arisen from the charge present in the nitrophenyl group in the
nitrophenyl glycoside.

It is interesting to note that agflracetylgalactosaminidase hydro-
lysed GL-5 pentasaccharide with a very high KM’ about 50-fold higher
than the hydrolysis of GL-S (Table 9). Both the pNP-a-GalNAc and
GalNAc-(al+6)-l,2:3,4-di-Q:isopropylidene-galactose. which have more
nonpolar groups adjacent to the flfacetylgalactosamine residue than the
GL-S pentasaccharide, were hydrolysed with lower KM values. The
hydrolyses of GL-5 and acetyl-sphingosyl-GL-5-pentasaccharide required
the presence of detergents that would also be able to reduce the pola-
rity of the environment near the active site. The high Vmax of the
hydrolysis of pNP-o-GalNAc may be due to the nature of the linkage of
the carbohydrate to an aromatic ring, which is capable of stabilizing
both a carbonium ion and a carbanion intermediate, thus perhaps reducing
the activation energy of the reaction. The fact that p-nitrophenyl
glycoside substrates degrade on storage. even at 0°. is evidence for
the easier degradation of p-nitrophenyl-glycosides. In contrast.
spontaneous hydrolysis of oligosaccharides is not usually observed
unless the solutions are contaminated by bacteria or fungi.

The kinetic parameters for the hydrolysis of porcine submaxillary
mucin are more difficult to evaluate. There are about 38 blood group
Af pentasaccharidegroups on each molecule of blood group A+ mucin

(73). There are also a-linked Hyacetylgalactosamine residues linked

154
directly to the protein molecule that can also be hydrolysed by “-3;

acetylgalactosaminidase. The KM of the mucin must be increased at
least 38-fold if all the pentasaccharide groups are accessible to the
enzyme. It will be further decreased by the gracetylgalactosamine
residues directly linked to the protein. The pH optimum in the hydro-
lysis of the mucin showed a biphasic curve with increased hydrolysis
activity below pH 3.7 (not shown in the pH optimum curve). This may be
reflective of the hydrolysis of two types of gracetylgalactosamine
residues on the mucin molecule.

The competitive inhibition by fl:acetyl-talosamine»(al+6)-l,2:3,4-
diggrisopropylidene-galactose is unique in that fl;acetyl-talosamine
differs from gracetylgalactosamine in the configuration at C-2 of
the monosaccharide. It implies that the catalytic site of the enzyme
cannot distinguish between the R or S configurations at C-2. It is
surprising though, that galactose and galactosamine hydrochloride at
10 mM did not inhibit the hydrolysis of Forssman hapten. The presence
of a bulky gracetylgroup on the terminal sugar is probably a requirement
for recognition by the enzyme.

The high yield of TalNAc-(ol+6)-di-Qyisopropylidene—galactose in
its synthesis and its inhibition properties make it potentially
important as a ligand fbr affinity chromatography in the purification
of afiflyacetylgalactosaminidase.

The presence of two enzymatically active protein peaks on native
gel electrophoresis of the porcine oeflracetylgalactosaminidase is
probably due to aggregation of the enzyme, which chromatographed on
Sephadex 0-150 as a single symmetrical peak. The identity of the

isoelectric focussing forms in migration on native gel electrophoresis

155

probably indicates that the difference in these two forms arises from
minute differences in the protein molecule. It has been suggested that
neuraminidase treatment changes the mobilities of a-galactosidases (132,
136). No sialic acid residue was found in this enzyme. The change in
charge may be due to deamidation of glutamine, asparagine. 4-L—aspartyl-
glycosylamine, or S-L-glutamylglycosylamine, resulting in the loss of
an amino group or an oligosaccharide and an amino group. This would
shift the pI of the enzyme to a more acidic form. Indeed a glycosyl
asparaginase has been purified from hog kidney (225).

Porcine a-fl:acetylgalactosaminidase has been shown to be an exo-
glycosidase. Endoglycosidases are not uncommon. Endo-a-fl:acetylgalac-
tosaminidases were found in S$reptomyces griseus (226), Diplococcus
pneumoniae (227, 228) and Clostridium‘perfrigens (229). Endo—Bfiflf
acetylglucosaminidase were demonstrated to be present in Streptomyces
griseus (230-232), Diplococcus pneumoniae (232), and hen oviduct (233).
Endo-B-galactosidases were purified from Ebeherichia freundri (234)
and Diplococcus pneumoniae (235). These enzymes are very useful
tools in liberating oligosaccharides from glycoproteins and glycopep-
tides (215, 236, 237). No mammalian endoglycosidase has yet been
characterized.

The natural substrates of the pig liver aefl7acetylgalactosaminidase
are probably the blood group A active glycolipids, which have been
shown to be present in porcine gastric mucosa and porcine mucins (Table
l). The kinetic parameters of hydrolysis of the blood group A substan-
ces were not examined because of the lack of material. Under the condi-
tions used for the hydrolysis of Forssman hapten, the enzyme did
hydrolyse the blood group A-active glycosphingolipids to give

156

glycolipids of higher mobilities.

The Role of Carbohydrates on aaurACetylgalaCtosaminidase

The carbohydrate analysis of the enzyme indicated that the only
sugars present were mannose and gracetylglucosamine. This is different
from hexosaminidase which also has sialic acid and presumably galactose
(104).

As shown in Table 10. most glycopeptides have an inner core
composed of mannose and Hyacetylglucosamine while the periperal carbo-
hydrates can be H;acetyl-neuraminic acid, galactose and fucose. The
structure of the oligosaccharide residues in this u-fl:acetylgalactosa-
minidase is probably similar to the inner core of the oligosaccharide
of ovalbumin, which contains only mannose and uracetylglucosamine. The
role of carbohydrates on glycoproteins is not well understood except
perhaps for suspected roles as cell surface receptors, freezing point
depression, change in viscosity of solutions or hepatic recognition of
non:fl:acetylneuraminic acid-containing glycoproteins (238).

Carbohydrates play an important role in the metabolism of glycosi»
dases in cell cultures as well as in animals. Bovine testicular B-ga-
lactosidase. which contains only mannose and gracetylglucosamflne. is
rapidly assimilated by generalized gangliosidosis skin fibroblasts.
Pretreatment of the enzyme with mannosidase reduced the rate of assi-
milation by 97% (239). Similarly, treatment of rat liver lysosomal
B—glucuronidase. !:acetyl-8-glucosaminidase. a-L-fucosidase and a-Q-
mannosidase. as well as rat preputial gland e-glucuronidase, with sodium
periodate resulted in a near abolition of rapid clearance of the enzymes

from the circulation when injected into rats (240). The isolation of

157

Am+Aav-caz-A~+Auv-=az
=a<-u<zqu- Ae+AmV-u<zuAu- -A¢+Amv-eaa

 

 

 

 

/A 8:35:
m+Aav-=aa
,Ae+AuV-=a= AA aeAuaaaouAAu
Am+Aav-=az
=m<- u<zuAu- Ao+Amv-u<zuAu- AqumV-=a= . Am+AaV-=az
Ao+PavucmaA
mAN Am+AaV.saz caqueu > aewpaaaoaAAw
cvsnnpa>o
03a .Am+Auv- ces- A~+Amv-u<zqu- Ao+AmT Paw- Ae+~av -u<=az A
=m<-u<zqu-u<zuApA > aeAuaaaouxAu
emu o+Aav-=az-A~+Amv-u<zuAm-Am+Amv-Aao-Ae+~av-u<=az cuss; z cAAanoAmocaseA
a eaz-m-u<zuAu-Aae-u<=az
¢-N \.
cm<- Au<zupuvucas
New aleaz-m-u<zuAu-Aau-u<=az acesoa eAneoagpoaa
=a<-u<zuAu
-A¢+Amv-0MuuAu\Ao+Aav-caz- AN+AmV -u<zuAu Ae+Amv- Ago- Ae+~av-u<=az
-A¢+A -ee
oAN aAm+Auv-caz-AN+AmV-u<zu_u-A¢+Amv-Aam-Ao+~av-u<=az cuss; =Aaaaem=~auoaam
.emm assuuagum muesom cvmuoeqoua—w

mcvmuognooapu :A mmuvgacuuomomvpo mo moeauuaeum .op mpaah

158

 

caffuéuSATE-3-2+;vastGATE-33.33-05:
caffuezuaATE-A3-Am+~3-a§az

Z

A
2.75 «23552.30qu

 

 

 

 

 

mew :25; -3232 3532;38me
=a<-u<zuAm-A~+Amv-u<zuAu-AA+AmV-=~=-a-A-oA=azv AA aeAuaaaousAw
cm<uu<zuGA~+A3 \AiETuAAzuS- ~+E AaquméaTugmz
-u<zu$..A¢+ET§= Am+paTc£uM~+AmTu<zuA9 win upmw-2+~uv-u<=mz I
./ ~+Amv-u<zuAu-A~+Aav-Aac-A~+~av-u<=az A aquaaaouaAw
«AN “Page; seepage chvcrm orange:
=m<-m-u<zu8-?+av
-Huaa-AaA+mv-uu<z Am+Aav-caz-A~+AmV-u<zuAm
AGATAeém
new Amiav-cazAulmvno<zuEAe£mT$w...$+~3..u<=mz 2:23 o3
\813-cazquimYuAAzuSAeisufioAmlaTuE $2283
u<zupuu$+~3u§z .633 $8288....
EN Ania-cmziAwimTuAAzuSAcimTZcAwiav-2; 525; mats
.An+AuV-caz-AN+Amv-u<zuAa-A¢+AmV-Ana. aeAaaguuamomAAo A
o<zu5A¢+P3icm . A on»... 2323223 2m
m 5 field -5??va -u<zu SAeAB - :5 :25; .82..
com 9.32:3 magnum 53°38me

.Au.»=cuv .oA aAQaA

159

 

\on+Aav-eaz-a-=az-Am+Aav-eaz
aA

-u zu uehcgucmtuo LAmlaTcuz

 

 

Accompocnnna u uwavp
mew Am+Aav-=az- .caz AAau auAaaguuamomAAo ewoeAgp
. 98¢ 232:3 3.58 539.338

.Ae.»=ouv .oA aAnae

160

carbohydrate-binding proteins from specific organs Such as the B-
galactoside binding lectin of calf heart and lung (241) seems to imply
that the carbohydrates on proteins play an important role in being

recognized by specific organs in animals.

SUMMARY

The structure of a glycosphingolipid with Forssman reactivity was
characterized by the use of GLC analyses of the carbohydrate composition,
permethylation to determine the linkages and stereospecific enzymes to
determine the sequence of the carbohydrate residues and their anomeric
configurations.

In the course of the purification of a-fl:acetylgalactosaminidase
from porcine liver, methods like Con A-Sepharose, ampholine elution of
enzyme activities from DEAE-cellulose, and hydroxylapatite were found
to be very useful. The enzyme is difficult to purify because it is
very unstable when present in dilute concentrations. A large-scale
purification-has to be perfonmed so that even when highly purified,
the enzyme can still be dissolved in a volume that can be handled
and yet is not inactivate due to high dilution.

When the enzyme has been purified, it appears to consist of two
subunits of 52,000 daltons each. 0nly mannose and glucosamine are
present in the oligosaccharide residues. The enzyme is not heat-stable
and is rapidly inactivated at temperatures above 60°.

The substrate concentrations in many of the assays were not
saturating because of the availability of substrate. Even pNP-a-H:
acetylgalactosaminide was in short,supply because the manufacturer
discontinued its production for a long period. One of the reasons was

the difficulty in synthesizing a-linked amino sugar derivatives. The

161

162

a-anomerity was hindered because of the hinderance of the bulky gracetyl
group present on the molecule. The failure in synthesis of an affinity
adsorbant was also related to the difficulty in synthesing the di-
saccharide. The difficulty was increased because of the relatively high
cost of flracetylgalactosamine.‘

Many of the assays in this work are tedious and time consuming and
thus some of the kinetic aspects of the enzyme toward certain substrates
were not exhaustive evaluations. The work did shed some light on the
efficiencies of the enzyme to hydrolyse different substrates by finding
the KM and vmax of the enzyme with these substrates. In general, the
enzyme seemed to hydrolyse the lipid substrates with a higher affinity.
The pNP substrate was hydrolysed with the highest Vmax'

This enzyme should be a useful tool in studying the lipid storage
diseases. The presence of the antibody against this enzyme or the
inhibition gracetyltalosamine-(al+6)-Gal in cell culture should lead to
induction of the accumulation of blood group A+ glycosphingolipids.

This would facilitate the study of the metabolism of blood group
glycolipids by making them more readily available.

Previous preparations of this same enzyme were always contaminated
with B-flracetylgalactosaminidase. The purification procedure in this
work eliminated all the detectible B-flracetylgalactosaminidase activi-
ties. This will be very useful for the unambiguous assignment of ano-
merities in oligosaccharides.

This enzyme can also be used in the treatment of cell surfaces and
thus altering the antigenicity of the cells (e.g; A+ RBC can be treated
to become an A' RBC). This may be one of the ways to elucidate the

function of blood group substances.

 

BIBLIOGRAPHY

10.
11.

12.

13.

14.

15.
16.
17.

BIBLIOGRAPHY

. Vance, H. R., Shook, D. P., III, and McKibbin, J. M. (1966)

Biochemistry g, 435-445

 

. Sung, S. J., Esselman, H. J., and Sweeley, C. C. (1973) J. Biol.

Chem. _2_4§_. 6528-6533

. Siddiqui, B., and Hakomori, S.-I. (1971) J. Biol. Chem. 246, 5766-

5769

. Sweeley, C. C., and Klionsky, B. (1963) J. Biol. Chem. 238, 3148-

3150

. Kint, J. A. (1970) Science 167, 1268-1269
. Romeo, 6., and Migeon, B. R. (1970) Science 170, 180-181
. Stanbury, J. B., Hyngaarden, J. B., and Frederickson, D. S. (1972)

in The Metabolic Basis of Inherited Disease, 3rd Edition,
McGraw-Hill, N. Y.

. Watkins, N. M. (1966) Science 152, 172-181
. Heissmann, B., and Hinrichsen, D. F. (1969) Biochemistry g, 2034-

2043

Forssman, J. (1911) Biochem. Z. 31, 78-115

Brunius, F. E. (1936) in Chemical Studies on_the True Forssman
Hapten, the Cgrresponding_Antibody_and_lheir Interaction,
Aktiebolaget Fahlcrantz, Bo ryckeri, Stockholm

Landsteiner, K. (1921) Bi0chem. Z. 119, 294-306

Papirmeister, B. and Mallette, M. F. (1955) Arch. Biochem. Biophys.
51. 94-105

Makita, A., Suzuki, C., and Yosizawa, Z. (1966) J. Biochem. (Tokyo)
p_o_, 502-513

Mallette, M. F., and Rush, R. L. (1972) Immunochemistry 2, 809-820
Fraser, B. A., andMallette, M. F. (1974) Immunochemistrx 11, 581-593

 

 

 

 

Makita, A., Yokoyama, T., and Takahashi, H. (1973) Hoppe-Seyler's Z.
PHYS101. Chem. 354, 1149-1152

163

18.

19.

20.

21.

22.

23.

24.

25.

26.
27.

28.

29.

30.

31.

32.

33.
34.
35.
36.

37.

164

Taketomi, T., Hara, A., Kawamura, N., and_Hayashi, M. (1974) J,
Biochem. (Tokyo) 75, 197-199

Kijimoto, S., Ishibashi, T., and Makita, A. (1974) BioChem. Biophys.
Res. commun. 56, 177-184

 

 

Gahmberg, C. 6., and Hakomori, S.-I. (1975) J. Biol. Chem. 250.

 

 

 

2438-2446

Taketomi, T., Hara, A., and Uemura, K. (1975) Japan J. Exp. Med.
15,. 293-293

Klenk, J. 0., and Choppin, P. W. (1970)'Proc. Natl. Acad. Sci. U.S.
_A_. pp. 57-64

Weinstein,D. B., Marsh, J. 8., Click, M. C., and Warren, L. (1970)
J. BiOI. Chem. 245, 3928-3937

 

Dod, B. J., and Gray, G. M. (1968) Biochem. Biophy§.Acta 150, 397-
404 '

Renkonen, 6., Gahmberg, C. 6., Simons, I., and Kaarkainen, L. (1970)
Acta Chem. Scand.'gg, 733-735

 

Gahmberg, C. G. (1971) Bi0chim. BiOphy§,Acta 249, 81-95

Yogeerwaran, C., Sheinin, R., Wherrett, J. R., and Murray, R. K.
(1972) J. Bial. Chem. 247, 5146-5158

 

Keenan, T. W., Huang, C. M., and Morre, D. J. (1972) Biochem. Biophys.
Res. Commun. 41, 1277-1283

Critchley, 0. R., Graham, J. M., and Macpherson, I. (1973) FEBS Lett.
32, 37-40

Keenan, T. W., Morre, D. J., and Huang, C. M. (1972) FEBS Lett. 23,
204-207

Huterer, 5., and Wherrett, J. R. (1974) Canad. J. Biochem. 53, 507-
513

Smith, E. L., and McKibbin, J. M. (1973) A13. J. Med. Sci. 19, 417-
42

Kawanami J. (1972) J. Biochem. (Tokyo) 12, 783-785
Fogel, M., and Sachs, L. (1962) J. Nat. Cancer Inst. 22, 239-252
0'Nei11, C. H. (1968) J. Ce11 Sci. 3, 405-422

Robertson, H. T., and Black, P. H. (1969) Prac. Soc. Exp. Biol. Med.
139, 363-370

Burger, M. M. (1971) Nature New Biol. 231, 125-126

 

38.

39.
40.

41.

42.

43.
44.
45.
46.

47.

49.

50.
51.
52.
53.

54.
55.

56.
57.

58.
59.
60.

165

Makita, A., and Seyama, Y. (1971) Biochim. Biophys.'Acta 241, 403-
411

o...

 

Robbins, P. W., and Macpherson, I. (1971) Proc.'qu; Soc. Lond. B.
_1_7_7_. 49-58

Sakiyama, H., Gross, S. J., and Robbins, R. W. (1972) PrOc. Nat.
Acad. Sci. U.S.A. g2, 872-875

 

Critchley, D. R., and Macpherson, I. (1973) Biochim. Bigphys.Acta
593, 145-159

Kijimoto, 5., and Hakomori, S. (1972) FEBS Lett. 33, 38-42
Sakiyama, H., and Robbins, P. W. (1973) Federation Proc. 33, 86-90

 

Sakiyama, H., and Terasima. T. (1975) Cancer Research 33, 1723-1726

 

Sakiyama, H., and Robbins, P. W. (1973) Arch. Biochem. Biophxsfllgg,
407-414

Wo1f, B. A., and Robbins, P. W. (1974) J. Ce11 Biol. 31, 676-687

. Lansteiner, K. (1900) Zentr. Bakterio. Parasitenk. 31, 357

Hakomori, S.-I., Stellner, K., and Watanabe, K. (1972) Biochem.
Biophys. Res. Commun. 33, 1061-1068

Gardas, A., and Koscielak, J. (1974) FEBS Lett. 33, 101-104
P0u1ik, M. 0., Lauf, K. (1965) Nature 208, 874-876
Green, F. A. (1967) J. Immunol. 33, 56-60

Whittemore, N. B., Trabold, N. C., Reed, C. F., and Weed, R. I.
(1969) Vox. Sang. 11, 289-299

Marchesi, V. T., and Andrews, E. P. (1971) Science 174, 1247-1248

Akiyama, Y., and Osawa, T. (1972) Hoppe-Seylers Z. Physiol. Chem.
_3§_3, 323-331

Gardas, A., and Koscielak, J. (1971) Vox Sang. 39, 137-149

Hamaguchi, H., and Cleve, H. (1972) Biochim. Biophys. Acta 278, 271-
280

Anstee, D, J., and Tanner, M. J. A. (1974) BiOChem. J. 138, 381-386
Takasaki, 5., and Kobata, A. (1976) J. 8101. Chem. 251, 3610-3615

Economidou, J., Hughes-Jones, N. C., and Gardner, B. (1967) Vox Sang.
‘ 2, 321-328

 

I'll-I'll llillia-Illlllv‘l 1.111

61.
62.

63.
64.

65.

66.

67.

68.

69.
70.
71.

72.

73.
74.

75.

76.
77.
78.

79.

80.
81.

166

Zahler, P. (1968) Vox Sang. 13, 81-101

Liotta, 1., Quintiliani, M., Quintiliani, L., Buzzonetti, A., and
Giuliani, E. (1972) Vox Sang.3g, 171-182

Fukuda, M., and Osawa, T. (1973) J. 3101. Chem. 248, 5100-5105

 

Yamato, K., Handa, S., and Yamakawa, T. (1975) J. Biochem. (Tokyo)
33;, 1207-1214

 

Slomiany, A., and Horowitz, M. I. (1973) J. Bi01. Chem. 248, 6232-
6238

 

Slomiany, A., Slomiany, B. L., and Horowitz, M. I. (1974) J. Biol.
Chem. 333, 1225-1230

Slomiany, B. L., Slomiany, A., and Horowitz, M. I. (1973) Biochim.
Biophys. Acta 326, 224-231

Slomiany, A., and Slomiany, B. L. (1975) Biochim. Biophys. Acta 388,
135-145

Slomiany, B. L., and Slomiany, A. FEBS Lett. in press
Slomiany, B. L., and Slomiany, A. Biochim. Biophys. Acta in press

Tettamanti, G., and Pigman, W. (1968) Arch. Biochem. Biophys. 124,
41-50

De Salegui, M., and Plonska, J. (1969) Arch. Biochem. Biophys. 129,
49-56

Carlson, D. M. (1968) J. Biol. Chem. 243, 616-626

 

Smith, E. L., Bowdler, A. J., Bull. R. W., and McKibbin, J. M.
(1973) Immunology 33, 621-629

Smith, E. L., McKibbin, J. M., Karlsson, K.-A., Pascher, I., and
Samuelsson. B. E. (1975) Biochemistry 19, 2120-2124

Lundblad, A., Svensson, S. (1973) Carbohyd. Res. 39, 187-189
Szulman, A. E. (1960) J. Expet. Med. 111, 785-800

Kay, H. E. M., and Wallace, D. M. (1961) J. Nat. Cancer Inst. 33,
1349-1365

Davidsohn, 1., Kovarik, S., and Lee, C. L. (1966) Arch. Pathol 31,
381-390

Davidsohn, I. (1972) Amer. J. Clin. Path. 31, 715-730

 

Dabelsteen, E.. and Pindborg, J. J. (1973) Acta Pathol. Microbiol.
"Scand. sect.A;§1, 435-444

82.

83.
84.

85.

86.
87.
. Harry, G. J., and Watkins, W. M. (1964) Biochem. J. 33, 9P
89.
90.
91.
92.
93.
94.
95.
96.

97.

98.

99

100

101

167

Steiner, 5., Brennan, P. H., and Melnick, J. L. (1973) Nature New
3101. _2__4_5_. 19-31

 

Steiner, S., and Melnick, J. L. (1974) Nature 251, 717-718

 

Steiner, S. M., Melnick, J. L., Kit, 5., and Somers, K. D. (1974)
Nature 248, 682-684

 

Stellner, K., and Hakomori, S.-I. (1973) Biochem. Biophys. Res.
C01m’1un.‘ ' _5_§_, 439-445

Morell, P. and Braun, P. (1972) J. Lipid Res. 13, 293-310
Watkins, W. M. (1959) Biochem. J. 11, 261-274

 

 

Tuppy, H., and Staudenbauer, W. L. (1966) BioChemiStry 3, 1742-1747

 

Howe, C., and Kabat, E. A. (1953) J. Amer. Chem. Soc. 13, 5542-5547

 

McDonald, M. J., and Bahl, 0. P. (1972) in Methods in Enzymol.
. (Ginsburg, V., ed.) Vol.XXVIII. pp. 734-738, Academic Press,
N. Y.

Weissmann, B., and Friederici, D. (1966) Biochim. Biophys. Acta
_1_1_z_, 493-499

 

Isreal, M., Bach, G., Miyatake, T., Naiki,M., and Suzuki, K.
(1974) J. Neurochem. 33, 803-809

Callahan, J. W., Lassila, E. L., Dentandt, W., and Philippart, M.
(1973)Biochem. Med. 1, 424-431

Bhavanandan, V. P., Buddecke, E.. Carubelli, R., and Gottschalk, A.
(1964) Biochem. Biophs. Res. Commun. 13, 353-357

Bhargava, A. S., Buddecke, E.. Wernes, E.. and Gottschalk, A.
(1966) Biochem. Biophys. Acta 127, 457-467

Schaner, H., and Gottschalk, A. (1968) Biochim. Biophy. Acta 156,
304-310

Mahadevan, S., and Tappel, A. L. (1968) Arch. Biochem. Biophys.
LL2_8_. 129-132

Buddecke, E.. Schaner, H., Werries, E.. and Gottschalk, A. (1969)
Biothem. Biophys. Res. Commun. 33, 517-521

Werries, E.. Wollek, E.. Gottschalk, A., and Buddecke, E. (1969)
Eur. J. Biothem. 13, 445-449

Gottschalk, A., and Buddecke, E. (1972) in'GlycoprOteins, Their

'C"'o$ition' StrUCture'and‘FUnction (Gottschalk, A. ed[)
pp. 1201-1210 Elsevier, Amsterdam

102.
103.

104.
105.
106.
107.

108.
109.

110.

111.

112.

113.
114.
115.

116.
117.

118.

119.

120.

121.

122.

168

Wang, C. T., and Weissmann, B. (1971) BiochemiStry 3, 1067-1072

 

Srivastava, S. K., Awasthi, Y. C., Yoshida, A., and Beutler, E.
(1974) J. Biol. Chem. 249, 2043-2048

Geiger, B., and Arnon, R. (1976)Biochemi5try 13, 3484-3493
Verpoorte, J. A. (1972) J. Biol. Chem; 247, 4787-4793

 

 

Verpoorte, J. A. (1974) BioChemistry 13, 793-798

 

Tarentino, A. L., and Haley, F. (1971) Arch. Biochem. Biophys. 147,
446-456

Sinha, A. K., and Rose, 3. P. R. (1973) J. NeuroChem.‘gQ, 39-44

Sellinger, 0. Z., Santiago, J. C., Sands, M. A., and Furinsloat, B.
(1973) Biochim. Biophys. Acta 315, 128-146

 

Robinson, 0., Jordan, T. W., and Horsburgh, T. (1972) J. Neurochem.
_1_9, 1975 - 1935

Price, R. G., and Dance, N. (1972) Bibchim. Bi0phys. Acta 271, 145-
153 .

 

 

Majumder, G. C., and Turkington, R. W. (1974)'BiOChemistry.13, 2857-
2864

Bullock, S., and Winchester, B. (1973) Biothem. J. 133, 593-599

 

 

Seyama, Y., and Yamakawa, T. (1974) J. Biochem. 13, 495-507

 

Wetmore, S. J., and Verpoorte, J. A. (1972) Can. J. Biochem. 33.
563-573

Fuchimi, H., and Tarui, S. (1976) J. Biochem. (Tokyo) 13, 265-270

 

 

Pugh, 0., Leaback, 0. H., and Walker, P. G. (1957) Biochem. J. 33,
464-469

Yoshikawa, J., Koide, H., and Yamanaka, M. (1972) Biochem. Biophys.
Res. Gammon. 33, 11-15

Penton, E.. Poenaru, L., and Dreyfus, J. C. (1975) Biochim. Biophys.
Acta 391, 152-139

Srivastava, S. K., Yoshida, A., Awasthi, Y. C., and Beutler, B.
(1974) J.'Bi01. Chem. 249. 2049-2053

Overdijk, B., van der Kroef, W. M. J., Veltkamp, W. A., and
Hooghwinkel, G. J. M. (1975) BibChem. J; 151. 257-261

Frohwein, Y. Z., and Gatt, S. (1967)'BiOChemiStgy‘3, 2775-2782

 

 

123.

124.

125.

126.

127.

128.

129.
130.
131.
132.

133.
134.

135.

136.

137.
138.

139.

140.

141.

142.

169

Hooghwinkel, G. J. M., Veltkamp, W. A., Overdijk,.B., and Lisman,
J. J. W. (1972) Hoppe-Seyler's Z. PhySiOl.'Chem. 349, 1095-1098

Abe,-T., Wanda, S., and Yamakawa, T. (1971) J;'BiOChem. (Tokyo)
'19. 1027-1033

 

Dey, P. M., Pridham, J. B. (1971) in Advances in Enzymol. (Meister,
A. ed.) Vol. 36. pp. 91-130, John Wiley & Sons, N. Y.

 

Mapes, C. A., and Sweeley, C. C. (1973) BiOChem.‘Bi0phys. Res.
Commun. 33, 1317-1324

 

Mapes, C. A., Suelter, C. H., and Sweeley, C. C. (1973) J. Biol.
Chem. 333, 2471-2479

Mapes, C. A., and Sweeley, C. C. (1973) J. Biol. Chem. 248, 2461-
2470

Beut1er, E., and Kuhl, W. (1972) J. Biol. Chem. 247, 7195-7200

 

 

Beut1er, E., and Kuh1, W. (1972) NatUre New 8101. 239, 207-208
Ho, M. W. (1973) BioChem. J. 133, 1-10

Romeo, G., Matteo, G. 0., 0'urso, M., Li, S.-C., and Li, Y.-T.
(1975) Biochim. Biophys. Acta 391. 349-360

 

Kano, 1., and Yamakawa, T. (1974) J. Biochem. (Tokyo) 33, 347-354

Rietra, P. J. G. M., Tager, J. M., and Borst, P. (1972) Biochim.
Bipphys. Acta 279, 436-445

Dean, K. J., Sung, S. J., and Sweeley, C. C. (1977) Fed. Proc.
in press

Mapes, C. A., and Sweeley, C. C. (1973) Arch. Biochem. Biophys.
133, 297-304

Kano, 1., and Yamakawa, T. (1974) Chem. Phys. Lipids 13, 283-291

Ho, M. W., Beutler, S., Tannant, L., and O'Brien, J. S. (1972)
Amer. J. Hum. Genet. 33, 256-266

Wood, 5., and Nadler, H. L. (1972) Amer. J. Hum. Genet. 33, 250-
255

Romeo, G., Childs, B., and Migeon, B. R. (1972) FEBS Lett. 31,
161-166

Tanaka, H., Meisler, M., and Suzuki, K. (1975)‘Bio¢him. Biophys.
'Atta‘333, 452-463

Suzuki, Y., and Suzuki, K. (1974) J; 3131L'cnem;'249, 2093-2104

 

143.

144.

145.
146.
147.
148.

149.

150.

151.

152.

153.
154.
155.

156.
157.
158.
159.

160.

161.

162.

163.

170

Suzuki, Y., and Suzuki, K. (1974) J. Biol. Chem. 249, 2105-2108
Suzuki, Y., and Suzuki, K. (1974) J. 8101. Chem. 249, 2113-2117

 

Miyatake, 1., and Suzuki, K. (1975) J. 3101. Chem. 250, 535-592

 

Miyatake, T., and Suzuki, K. (1972) J. Biol. Chem. 247, 5398-5403

 

Wenger, D. A. (1974) Chem. Phys. Lipids 13, 327-339

 

Norden, A. G. W., Tennant, L. L., and O'Brien, J. S. (1974) 3,
Bio1. Chem. 249, 7969-7976

Ho, M. W., Cheetham, P., and Robinson, 0. (1973) Biochem. J. 136,
351-359

 

Callahan, J. W., and Gerrie, J. (1975) Biochim. Biophys. Acta 391,
141-153

 

Callahan, J. W., and Gerrie, J. (1976) Can. J.'Biochem. 33, 803-812

 

Hajra, A. K., Bowen, 0. M., Kishimoto, Y., and Radin, N. S. (1966)
J. Lipid Res. 2, 379-386

Arora, R. C., and Radin, N. S. (1971) Ligids Z, 56-59
Arora, R. C., and Radin, N. S. (1972) J. Lipid Res. 13,86-91

Nishimura, K., and Amano, R. (1976) J. Biochem. (Tokyo) 33, 209-
215

Ho, M. W. (1973) Biochem. J. 136, 721-729

 

Ho, M. W., and O'Brien, J. S. (1973) Biochem. J. 131, 173-176
Ho, M. W., and Light, N. D. (1973) Biochem. J. 136, 821-823

 

H0, M. W., and Rigby, M. (1975) Biochim. Biophys. Acta 397, 267-
273

Pentchev, P. G., Brad , R. 0., Hibbert, S. R., Gal, A. E., and
Shapiro, 0. (1973 J. Biol. Chem. 248. 5256-5261

Brady, R. 0., Kanfer, J. N., and Shapiro, 0. (1965) Biochem. Biophys.
Res. Comun.‘_1_3, 221-225

Brady, R. 0., Kanfer, J. N., Bradley, R. M., and Shapiro, 0. (1966)
'J. Clin. InVest. 33, 1112-1115

Ho, M. W., Seck, J., Schmidt, 0., Veath, M. L., Johnson, W.,
Brady, R. 0., and O'Brien, J. S. (1972) Amer. J. Hum. Genet.
24, 37-45

164.

165.
166.
167.

168.

169.

170.

171.

172.

173.
174.
175.
176.

177.

178.

179.

180.

181.
182.
183.
184.

171

Novikoff, A. G. (1973) in Lysosomes and Storgge DiseaSes (Hers,
H. G., and van Hoof, F., edi) pp. 1-41,7Academic Press, N. Y.

Novikoff, A. 3. (1976) 3...; nat1.46ad.’sei. 0.5.4.113, 2731-2737
Ho, M. w. (1975) FEBS Lett. 33, 243-247

Pentchev, P. G., and Brady, R. 0. (1973)'Bi0¢him.Biophys.‘ACta
’32], 491-496

 

Li, S.-C., Wan, C.-C., Mazzotta, M. Y., and Li, Y.-T. (1974)
Carbohyd. Res. 33, 189-193

Li, Y.-T., Mazzotta, M. V., Wan, C.-C., Orth, R., and Li, S.-C.
(1973) J. Bi01. Chem. 248, 7512-7515

 

Fisher, G., and Jatzkewitz, H. (1975) Hoppe-Seyler's Z. Physiol.
Chem. 333, 605-613

Horst, J., Kluge, F., Beyreuther, K., and Gerok, W. (1975)
Proc. Natl. ACad. Sci. U.S.A. 33, 3531-3535

Baker, H. J., Mole, J. A., Lindsay, J. R., and Creel, R. M. (1976)
Fed. Proc. 33, 1193-1201

Andrews, J. M., and Menkes, J. H. (1970) ggper. NeUrol. 33, 483-493
Suzuki, K. (1970) Lab. Invest. 33, 612-619
Stern, J. (1972) Lab. Invest. 33, 509-514

Tulkens, P., Trouet, A., and van Hoof, F. (1970) Nature (London)
233,_1232 "“"'

Weston, P. 0., Barrett, A. J., and Dingle, J. T. (1969) Nature
(London) ggg, 235-236 """'

Warren, K. R., Schafer, I. A., Sullivan, J. C., Petrelli, M.,
and Radin, N. S. (1976) J. Lipid Res. 13, 132-138

Dawson, G., Stoolmiller, A. C., and Radin, N. S. (1974) J. Biol.
Chem. 333, 4638-4646 "““"

Schneck, L., Valenti, C., Amsterdam. 0., Friedland, J., Adachi, M.,
and Volk, B. W. (1970) Lancet 1, 582

Woollen, J. W., and Turner, P. (1965) Clin. Chim. Acta 12, 671-683
Bosmann, H. B. (1969) Experi. Cell Res. 33, 217-221
Bosmann, H. B., and Pike, 6. Z. (1970) Life sciences 3, 1433-1440

 

Bosmann, H. 3. (1972) Biochim. Biophys. Acta 264, 339-343

172

185. Bosmann, H. B., Lockwood, T., and Morgan, H. R. (1974) Exper. Cell
' _R_e_§_.' 33, 25-30

 

186. Bosmann, H. 3.. and Hall, 1. c. (1974) Proc. Natl. Acad. Sci. U.S.A.
11, 1333-1337

137. Saito, T., and Hakomori, S.-I. (1971) J. Lipid Res. 13, 257-259
133. Suzuki, Y., and Suzuki, K. (1972) J.'Lipid Res. 13, 637-690

 

189. Lemieux, R. V., James, K., and Nagabhushan, T. L. (1973) Canad.
J. Chem. 31, 48-52

190. L1oyd, K. 0. (1970) Arch. BiOChem. Biophyg, 137, 460-468

191. March, 5. C., Parikh, 1., and Cuatrecasas, P. (1974) Anal. Biochem.
69, 149-152

 

192. Lowry y, 0. H. Rosenbrough, N. J. Farr, A. L. , and Rose, J. R.
(1951) J. Biol. Chem. 193, 265- 275

193. Vance, D. E., and Sweeley, C. C. (1967) J. Lipid Res. 3, 621-630

 

194. Hakomori, S.-I. (1964) J. Biochem. (Tokyo) 33, 205-208

 

195. Bjorndal, ,Hellerqvist, C. G. , Lindberg, B., and Svensson, S.
(1970)H Angew. Chem. Internal. Edit. 3, 610-619

196. Gabriel, 0. (1971) in Egghods in Enzymg1, (Jakoby, W. B., ed.)
Vol. XXII, pp. 565-578, Academic Press, N. Y.

197. Ma1ik, N., and Berrie, A. (1972) Anal. BiochEM.332, 173-176

198. Vesterberg, 0. (1971) in Methods in Enzymol. (Jakoby, W. B. ed.)
Vol. XXII, pp. 389-412, Academic Press, N. Y.

199. McPhie, P. (1971) in Methods in Enz l. (Jakoby, W. 8. ed.)
Vol. XXII, pp. 23-32, Academic ress N. Y.

200. Li. S.-C., and Li, Y.-T. (1970) J. Bio1. Chem. 245, 5153-5160
201. Li, Y.-T., and Li, S.-C. (1968) J. Bio1. CheM. 243, 3994-3996

202. Hakomori, S.-I., Siddiqui, 8., Li, Y.-T., Li, S.-C., and Heller-
qvist, C. G. (1971) J. Biol. Chem. 246, 2271-2277

203. Laine, R. A., Sweeley, C. C. , Li, Y. T, Kisic, A., and Rapport,
M. M. (1972) J. Lipid Res. 13, 519-524

204. Sweeley, C. C., Snyder, P. 0., Jr., and Griffin, C. E. (1970)
‘ Chem. Phys. Lipids 3, 393-408

205. Moore, S., and Stein, H. H. (1954) J. 8101. Chem. 211, 907-913

206.

207.

208.

209.
210.

211.

212.

213.

214.

215.

216.

217.
218.

219.

220.

221.

222.

223.

224.

173

Chambers, R. E., and Clamp, J. R. (1971) Bi0chem. J. 125, 1009-
1018

 

Segrest, J. P., and Jackson, R. L. (1972) in MethodS'in Enzymol.
(Ginsburg, V., ed.) Vol. XXVIII, pp. 54-63, Academic Press,
N. Y.

 

Andrews, P. (1965) Biochem. J. '23, 595-606

 

Warren, L. (1959) J. Biol. Chem. 234, 1971-1975

 

 

Rapport, M. M., and Graf, L. (1957) Ann. N. Y. Acad. Sci. 33, 608-'
632

Fahrney, D. E., and Gold, A. M. (1963) J. Amer. Chem. Soc. 33, 997-
1000

 

Ogata, S.-I., Muramatsu, T., and Kobata, A. (1975) J. Biochem.
(Tokyo) 13 . 637-696

Wolfe, L. 5., Senior, R. G., and N9, N. M. K. Ying Kin. (1974)
J. Bio1. Chem. 249, 1828-1838

Tsay, G. C., Dawson, 6., and Sung, S.-S. J. (1976) J. Biol. Chem.
351,5352-5359

 

Tai, T., Yamashita, K., Ogata-Arakawa, M., Koide, N., Muramatsu, T.
Iwashita, S., Inoue, Y., and Kobata, A. (1975) J. Biol. Chem.
ZEQ, 8569-8575 .

 

Spik, G., Bayard, B., Forunet, B., Strecker, G., Bouquelet, S.,
and Montreuil, J. (1975) FEBS Lett. 33, 296-299

Nakajima, T., and Ba11ou, C. E. (1974) J. 8101. Chem. 249, 7685-7694

 

Tai, T., Yamashita, K., and Kobata, A. (1975) J. Biochem. (Tokyo)
19, 679-636

 

Hoffman, J., Lindberg, B., and Svensson, S. (1972) Acta Chem.
Scand. 26, 661-666

Leaback, 0. H., and Robinson, H. K. (1975) Biochem. Biophys.
Res. Commun. 33, 248-254

Tulsiani, E. R. P., Keller, R. K., and Touster, 0. (1975) J. Biol.
Chem. 339, 4770-4776

Helenius, A., and Simons, K. (1975) Biochim. Biophys. Acta 415,
29-79

Carey, M. C., and Small, 0. M. (1972) ArCh. Int. Med. 130, 506-527

 

 

Yedgar, S., and Gatt, S. (1976) Biochemistry 13, 2570-2573

 

225.

226.

227.

228.
229.
230.
231.

232.

233.

234.

235.
236.

237.

238.

239.

240.

241.

242.

243.

174

Kohno, M., and Yamashina, I. (1972) in Mefihods in Enzymol. (Ginsburg,
V. ed.) Vol. XXVIII, pp. 786-792, Academic Press, N. Y.

Reissig, J. L., Lai, W.-H., and Glasgow, J. E. (1975) Can. J.
Biochem. 33. 1237-1249 '

Bhavanandan, V. P., Umemoto, J., and Davidson, E. A. (1976)
Bi0chem. BiOphys. Res. comun. 19, 733-745

 

Endo, Y., and Kobata, A. (1976) J. Bi0chem. (Tokyo) 33, 1-8
Huang, C. C., and Aminoff, D. (1972) J. Biol. Chem. 247, 6737-6742
Tarentine, A. L., and Maley, F. (1974) J. BiOl. Chem. 249, 811-817

Tarentino, A. L., Plummer, T. H., Jr., and Maley, F. (1974) 3,
8101. Chem. 249, 818-824

Arakawa, M., and Muramatsu, T. (1974) J. Biochem. (Tokyo) 13, 307-
317

 

Tarentino, A. L., and Maley, F. (1976) J. Biol. Chem. 251, 6537-
6543

Fukuda, M. N., and Matsumura, G. (1976) J. Bi01. Chem. 251, 6218-
6225

Takasaki, S., and Kobata, A. (1976) J. Biol. Chem. 251, 3603-3609

Muramatsu, T., Koide, N., and Ogata-Arakawa, M. (1975) Bi0chem.
Biophys. Res. Commun. 33, 881-888

Ito, S., Muramatsu, T., and Kobata, A. (1975) Biochem. Biophys. Res.
Commun. 33, 938-944

Heide, K., and Schwick, H.-G. (1973) Angew. Chem. Internat. Edit.
12, 721-733

Hieber, V., Distler, J., Myerowitz, R., Schmickel, R. 0., and
Jourdian, G. W. (1976) Biochem. Biophys. Res. Commun. 13, 710-
717

Stahl, P., Six, H., Rodman, J. S., Schlesinger, P., Tulsiani, 0.
R. P., and Touster, 0. (1976) Proc. Natl. Acad. Sci. U.S.A.
23, 4045-4049

Waard, A. 0., Hickman, S., and Kornfeld,s . (1976) J. Biol. Chem.
_2_5_1_, 7531-7537

Nelsestuen, G. L., and Suttie, J. W. (1972) J. Bi01. Chem. 247,
6096-6102

Rao, A. J., Garver, F., and Mendicino, J. (1976) Biathemistry 13,
5001-5009

175

244. Kawasaki, T., and Ashwell, G. (1976) J. Bi01. Chem. 251, 5292-5299

 

245. Sugahara, K., Funakoshi, S., Funakoshi, I., Aula, P., and
Yamashina, I. (1976) J. BiOChem. (Tokyo) 39, 195-201

246. Spiro, R. G., Spiro, M. J., and Bhoyroo, V. D. (1976) J. Biol.
Chem. 2_51, 6409-6419