THESlS '

This is to certify that the
dissertation entitled

Turnover of Cell Surface Proteoglycans

In Cultured Fibroblasts

presented by

 

 

 

James H. Brauker

has been accepted towards fulfillment
of the requirements for

\M. D - degree in __B1_la_d“_‘““_" ’7 ‘

 

QM Z- [It—’1 9

/t/ Major professor )

Date Lj‘ 13 (78C

McIIIIII-‘mp—‘q‘ 1.: 11- .n . , . . 0.12771

 

TURNOVER OF CELL SURFACE PROTEOGLYCANS

IN CULTURED FIBROBLASTS

BY

James H. Brauker

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1986

( “979573

ABSTRACT
TURNOVER OF CELL SURFACE PROTEOGLYCANS
IN CULTURED F I BROBLASTS
BY
James H. Brauker

Human fibroblasts were cultured in 3580‘;2 to label
the glycosaminoglycans (GAGs); the extracellular matrix was
derived from these labeled cells by removal of cells with
the chelating agent ethylene glycol-bis-lfi-amino ethyl
ether)N,N'—tetra acetic acid (EGTA). When unlabeled cells
were plated onto these radiolabeled extracellular matrices,
two distinct events were observed: a) the cells actively
released [BSslPG from the matrix and; b) the cells
degraded a large fraction (20-40% in 24 hours) of the
matrix PG, releasing free sulfate. The latter degradation
event could be inhibited in a Specific dose-dependent
manner by addition of mannose 6—phoSphate (Ki - 60 uM) to
the culture medium. Analyses of this effect in terms of
the saccharide Specificity, NH4Cl sensitivity, and the
requirement for cells suggest that both an intracellular
compartment and the mannose 6-ph05phate receptor that binds

lysosomal enzymes at the cell surface may play important

roles in the turnover of PG of the extracellular matrix.

The turnover of PG of the cell surface was also
studied in I-cell fibroblasts. These cells have a
deficiency of lysosomal enzymes both at the cell surface
and within the cell, resulting in the intracellular storage
of partially degraded GAGs. Our analysis of these cells
indicated no apparent perturbation of the turnover of the
PC of the cell surface. The cell surface heparan sulfate
(HS) PG is internalized by the cells at a rate comparable
to that of normal cells. Once internalized, the H8 is
acted on by an endoglycosidase. This HS endoglycosidase
activity appears to be unaffected by the I—cell defect,
suggesting that it is a non—lysosomal enzyme.

Treatment of normal cells with NH4C1, which inhibits
intracellular lysosomal enzyme activity, resulted in
defective GAG degradation, mimicking that observed in I—
cells. However, the initial steps in the turnover of cell
surface PGs, including internalization, separation from
core protein, and endoglycosidase activity was unaffected
by treatment with NH4C1.

These data support the hypothesis that partial (HS
endoglycosidase) and complete (mannose 6-phosphate
inhibitable) degradation of GAGs may occur in compartments

other than lysosomes.

For Muggy

ACKNOWLEDGEMENTS

I owe thanks to John Wang for supporting me and giving
me the freedom to "noodle around", while at the same time
teaching me to set realistic goals. I would also like to
thank Dr. Richard Anderson for teaching me the importance
of attending to details (although I don't like doing it any
more than previously), and for his listening ear.

I wish to thank my wife, Diane, for always knowing
where my head was at, understanding why it was there, and
waiting for it to cycle around to ground zero occasionally.
She is also a good fishing partner.

To those lab mates who worked with me, whether they
were my teachers, students, or "worthy opponents", I owe a
debt which I h0pe I can repay over the coming years as we
all venture forth to various grant review committees. I
owe Special thanks to Cal Roff and Liz Cowles. They worked
with me, had faith in me, and will be remembered warmly as
coauthors of my first two papers.

To Craig Thomas, Lee Morehouse, and others who shared
time with me around pool tables I can only say; "I'm just

happy to be here".

 

TABLE OF CONTENTS

page

LISTOF TABLESooooooeeooooooeooeooe00000000000000.0000. Vii

LIST OF FIGURES...... .............. ....................viii

ABBREVIATIONS . O . 0 O O O C O O O O O O O O O O O O O O O O O O O O O O O C O O O O O O C O C . Xi

 

INTRODUCTIONOOOOOOOO0.0000IOOOOOOOOOOOOOOOOOOOOOOOOOOOO 1

CHAPTER 1 LITERATURE REVIEW

Structure and synthesis of proteoglycans..... 3
Secreted proteoglycans....................... 6
Cell surface proteoglycans. ..... .. ......... .. 8
Intracellular proteoglycans.................. ll
Degradation of GAG chains.................... 12

TranSport and distribution of lysosomal
enzymes..0...OOOOOOOOOOOOOOOOOOOOOO00...... l7

ReferenceSOOOOOOOIOO00.0.0.0...DOOOOOOOOOOOOOOOOOO 21

CHAPTER 2 THE EFFECT OF MANNOSE 6—PHOSPHATE ON THE
TURNOVER OF THE PROTEOGLYCANS IN THE
EXTRACELLULAR MATRIX OF HUMAN FIBROBLASTS... 30

summary. 0 O O O O O O O 0 O O O O O ..... 0 O 0 O O O O O O O O O I O O O O O O O O 0 O 31
Introduction. 9 O O O O O O O O O 0 O O O O O O O O O O O O O O 0 O O O O O 0 O O O O O 32
Materials and Methods

Cell lines and culture conditions............ 35
Preparation of radiolabeled SAM.............. 36
Digestion of SAM by unlabeled cells or
conditioned medium......................... 36
Harvest and bulk quantitation of
radioactivity.............................. 39
Quantitation of degraded material............ 40
Analysis of SAM fraction by column
chromatography............................. 41

iv

Results

Characterization of 3:’SO4—1abeled
SAM by column chromatography ............... 42
Solubilization of SAM components into
the medium...... ...... . .................... 45
Column chromatography of solubilized SAM
components ............................. .... 45
Kinetics, dose response curve and sacchar—
ide specificity of the MBP effect ...... . . 51
Evidence that M6P affects a cell— associated
compartment ................... ... .......... 58
Discussion ........................................ 60
Acknowledgements .................................. 64
References ........................................ 65

CHAPTER 3 TURNOVER OF CELL SURFACE PROTEOGLYCANS
IN CULTURED FIBROBLASTS: COMPARISON

BETWEEN NORMAL AND I—CELLS...... ........... . 67
Summary.............. ............. ........ ........ 68
Introduction... ...... . ...... . ..... ................ 70

Materials and Methods

Cell culture and radiolabeling procedures.... 73
Harvest of cell fractions during label—

chase experiments ....................... ... 74
Chemical and enzymatic degradation of GAGs... 76
Gel filtration ........ . ...................... 77

Results

Turnover of 35SO ‘2-1abeled components

. 4

in cells... .................... .. ...... .. 78
Molecular weight distribution of SO —

labeled components in the cell mono ayer... 81
Characterization of the cell surface PG...... 85
Effect of trypsin of the turnover of

pericellular PG ............. .. ............ 88
Shedding and uptake of pericellular PG ....... 92
Characterization of intracellular GAGs....... 101
Turnover of intracellular GAGs ..... . ...... ... 109

Discussion...... .................................. 113
References ........................................ 120
V

CHAPTER 4 TURNOVER OF CELL SURFACE PROTEOGLYCANS IN
CULTURED FIBROBLASTS: INTERNALIZATION AND

EARLY STEPS IN DEGRADATION ............. ..... 123
Summary...... .............. ......... ..... . ........ 124
Introduction..... ..... . ........................... 126

Materials and Methods

Cell culture and radiolabeling procedures.... 128
Harvest of cell fractions during label-

chase experiments. ........... ...... ....... . 128
Gel filtration .................. ............. 130
Chemical and enzymatic degradation of GAGs... 131
Cell fractionization ................. ........ 131

Results

Theseffect of NH Cl on the turnover of

SO —labe1ed components .............. ..... 134
Characterization of intracellular GAGs in
NH Cl treated cultures ..................... 137

The effect of NH Cl on the turnover of
the intracellu ar GAGs..................... 142
Analysis of the intracellular GAGs in

subcellular fractions ................... ... 146
Discussion ...... . ......... . ....................... 150
References .......... . ....... . ..................... 157

Vi

Table
Chapt

1.

LIST OF TABLES

er 2

Quantitation of degraded [3551PG

various incubations..............

Quantitation of degraded [3581PG
by SL66 cells in the presence of

saccharides...... ....... . ........

vii

Page
after
.00. ........ 0 O. 52
produced
various
................. 58

 

LIST OF FIGURES

Figure
Chapter 1
l. Subunit structures of the glycosaminoglycans of
cultured fibroblasts ..... ............. ........... .
2. Schematic representation of the steps in the
degradation of heparan sulfate ........ ............
Chapter 2
1. Protocol for the preparation of radiolabeled
SAM and subsequent digestion of SL66 cells........
2. Chromatographic profile of SAM components before
and after digestion with pronase ..... ........ .....
3. Time course of release of radioactivity from SAM..
4. Sepharose CL6B chromatography of soluble radio—
labeled material released from SAM by SL66 cells..
5. nge course of the generation of degraded
[ SlPG by SL66 cells from SAM....................
6. Dose respogge of M6P effect on the generation of
degraded [ SlPG from SAM by SL66 cells ......... ..
Chapter 3
1. Kinetics of the turnover of 3580 —labeled
components in the intracellular and pericellular
compartments of SL66 and I—cells ....... . ...... ....
. . 35 —2
2. Gel filtration of SO -1abeled
components derived from the whole cell monolayer
and the intracellular fraction of SL66 cells and
I—cells during a label and chase experiment.......
. . 35 -2
3. Gel filtration of SO -1abeled
components derived from the pericellulgr fraction
of I—cells labeled for 24 hours with SO _ ......

viii

 

Page

14

37

43

46

49

53

55

79

82

86

 

10.

11.

 

Kinetics of the turnover of 35SO ‘2—labeled

components derived from the intracellular and
pericellular compartments of SL66 and I—cells..... 89

Kinetics of the appearance of LMw radioactivity
and HMW radioactivity in Egeo medium of SL66
and I- cells labeled with O4 ............... ... 93
. . 35 -2
Gel filtration of SO —1abe1ed components
in the medium of SL66 cells and I—cells ...... .... 96
Gel filtration of 35SO _2-1abe1ed components
in the medium of SL66 and I—cells..... ...... ...... 99

ggantiaation of internalized and shed
-1abe1d macromolecules from the

pericellular fraction of SL66 and I—cells ....... ..102
. . 35 —2

Gel filtration of SO -1abe1ed components

of the intracellular fraction of I—cells,

Sanfilippo cells, and heterozygous I—cells ........ 105
. . 35 —2

Gel filtration of so —labeled

components correSponding to region B and C

(figure 9) of the intracellular fraction of

I-cells after chemical treatments ................. 107

. . 35 —2

Kinetics of the turnover of 80 -labeled

components corresponding to regions B and C

of the intracellular fraction of I—cells... ....... 110

Chapter 4

1.

Kinetics of the turnover of 3SSO4-labeled

components in the intracellular and peri-

cellular compartments of untreated and

NH4Cl treated cells ................ ......... ...... 135
. . 35 —2

Gel filtration of 80 —labeled components

derived from the intracellular fraction of

untreated and NH Cl treated cells during a

label and chase experiment ..... . ..................l38
. . 3S -2

Gel filtration of SO —1abeled components

corresponding to region B and C (figure 2)

of the intracellular fraction of NH Cl

treated cells after chemical treatments ........... 140

Kinetics of the turnover of 35SO ‘2—labeled

components correSponding to regions 3 and C

(fig 2) of the intracellular fraction of NH 4C1

treated c€Ils. ...... ..... ............ ............. 143

ix

 

 

Percoll density gradient fractionation of
NH4C1 treated cells. .............................. 147

Schematic diagram of prOposed pathways for the
degradation of GAGs in fibroblasts.... ...... ......153

 

C'ase
CM
Cpm
CS

DL
dpm
DS
DSPG

EGTA

EDTA

GAG
GalN
GlcN
GlcUA
HMW
HS
HSPG

IdUA

 

ABBREVIATIONS

chondroitin ABC 1yase
conditioned medium

counts per minute

chondroitin sulfate

dense lysosomes
disintegrations per minute
dermatan sulfate

dermatan sulfate proteoglycan

ethyleneglycol—bis—UG—aminoethyl ether)N,N' tetra
acetic acid

ethylenediaminetetraacetic acid
fetal bovine serum
glycosaminoglycan

galactosamine

glucosamine

glucuronic acid

high molecular weight

heparan sulfate

heparan sulfate proteoglycan

iduronic acid

xi

PG

SAM

SDS

 

light lysosomes

low molecular weight
modified Eagle's medium
mannose 6-ph05phate
muc0polysaccaridosis
phosphate buffered saline
proteoglycan

substratum attached material

sodium lauryl sulfate

 

 

INTRODUCTION

This thesis describes the turnover of proteoglycans
(PG) at the cell surface and in the extracellular matrix of
human fibroblasts. The literature review should serve as
an introduction to the distribution and turnover of the
types of PG studied in cultured cells, and the transport of
the enzymes involved in their degradation. We are mainly
interested in the enzymes involved in the catabolism of
pericellular PG, and whether some of the steps in their
degradation occur in extralysosomal compartments. We have
chosen two main approaches to address these issues.
One approach is to analyze the metabolism of the PG of
normal cells cultured in the presence and absence of
mannose 6—phosphate (M6P). M6P competes with lysosomal
enzymes for binding sites at the cell surface, but does not
inhibit the intracellular transport of the enzymes to
lysosomes. Therefore, if M6P perturbs the degradation or
turnover of extracellular matrix PG, it would most likely
be affecting the lysosomal enzyme action in a non-lysosomal
compartment. The other approach is to study cells that
have dysfunctional lysosomes (I-cells). Identification of
the characteristics of the undegraded fragments ”stored" in

the lysosomes of I—cells, in comparison to the precursers

 

for degradation internalized from the cell

surface/extracellular matrix, would allow us to deduce the
non-lysosomal steps of PG degradation.

The experimental data of this thesis provide evidence
that non-lysosomal compartments, and at least one non—
lysosomal enzyme (heparan sulfate endoglycosidase) are
involved in the degradation of the PG internalized from the

cell surface and extracellular matrix.

 

Chapter 1

LITERATURE REVIEW

§££E£$BE§ EEQ §XE£Q§§2§ 92 BEQEEQQJXE§£§

Proteoglycans (PG) are molecules composed of linear
polysaccharide chains covalently linked to glycoproteins
called core proteins (1—4). The polysaccharide chains,
called glycosaminoglycans (GAGs) are composed of repeating
disaccharide units consisting of a hexosamine (D—
glucosamine or D—galactosamine) and an uronic acid (D—
glucuronic acid or its 5—epimer, iduronic acid). Ester
sulfate groups are present at various locations on the
disaccharide units, and N-sulfate groups are found on
heparan sulfate (HS) molecules. Because little is known
about the various core proteins of proteoglycans, the
molecules are usually classified according to the chemical
nature of the GAG chain. The human fibroblast synthesizes
three classes of GAG chains, called dermatan sulfates,
(DS), chondroitin sulfates (CS) (DS and CS will be referred
to as galactosaminoglycans), and heparan sulfates (HS).
The subunit structures of these polysaccharides are
illustrated in figure 1. Each of these GAG chains is
linked to serine residues of the core protein by a

trisaccharide linkage region (1—4).

 

 

 

 

Figure 1. Subunit structures of the
glycosaminoglycans of cultured fibroblasts: a) Subunit
structures of the galactosaminoglycans. Chondroitin
sulfate is composed of galactosamine (GalN) and glucuronic
acid (glcUA) dimers, whereas dermatan sulfate dimers
contain galactosamine linked to either glucuronic acid or
iduronic acid (IdUA); b) The subunit structure of heparan
sulfate is a dimer of glucosamine (GlcN) and either
glucuronic acid or iduronic-acid. All three types of
glycosaminoglycans may be sulfated and/or acylated as
indicated by the R grOUps. Many combinations of dimers are
observed based on the carbohydrate components, linkage
between carbohydrates, and the degree of sulfation and

acylation.

 

 

COO"

3.

 

 

R = -c\
CHZOR' CH3
OH R,0
B-D-GlcUA
- HNR , _
a . B-D-GolN R ' ‘“ °' ‘503
OR’
a-L-IdUA
_ b
coo
O
o
/
° , R ‘ 'C<CH
CHZOR 3
0H 0 or ~80;
B-D-GIcUA m .
. HNR
@ . a-D-GlcN R'= -H or -803

OR’
a- L-IdUA

 

 

 

Proteoglycans are assembled in the Golgi apparatus

by addition of monosaccarides to Specific sites on core
protein glyCOproteins. The glyc0proteins, containing 0 a N
—linked oligosaccharides, are synthesized in the rough
endOplasmic reticulum and then tranSported to the Golgi
apparatus (1,3,5,6). Either in the Golgi or just prior to
reaching the Golgi (6,7), they are recognized by a specific
enzyme, xylose—transferase, which is responsible for
initiating the synthesis of the linkage region. The
xylose—transferase links a xylose residue to the protein at
the hydroxyl group of serine. Two galactose molecules are
added to the xylose by galactosyl transferase to make the
complete linkage region gal—gal—xyl—ser (1—4). The
disaccharide units are then added by enzymatic transfer of
hexosamines & uronic acids from the appropriate UDP-sugar
onto the non—reducing end of the chain. Chain elongation
proceeds to a Specified number of carbohydrate dimers, when
termination of chain synthesis occurs by an as yet
unspecified mechanism. Completed PG molecules may be
secreted from the cell, transported to the cell surface, or
tranSported directly to intracellular degradative
compartments (8). These three pathways are discussed in

detail below.

Secreted Proteoglycans

 

The dominant PG secreted into the medium of

fibroblasts (80% of secreted PG) is a low molecular weight

(LMW) iduronate rich dermatan sulfate (9—13). These
dermatan sulfate proteoglycan (DSPG) molecules are eXported
rapidly from the cell with a half life of only a few
minutes (14) and apparently have a low affinity for the
cell surface or for the extracellular matrix of
fibroblastic cells grown on plastic. Although the small
DSPG does not become associated with the matrix in cultures
of fibroblasts, it is an important component of the
extracellular matrix of many tissues. It has a core
protein Mr of 38,000 (9), and an overall Mr of 70,000 -
100,000, depending on the number and Size of DS GAG chains
(16-19), and the number of oligosaccharide chains (18,20).
Immunological studies revealed a relationship between the
fibroblast DSPG and the DSPG of sclera and tendon (21).
Histochemical and immunohistochemical studies indicate that
it binds to type I collagen from rat tail tendon (22).
Abnormal biosynthesis of the DSPG by fibroblasts may be the
underlying cause of Coffin-Lowry syndrome, a disorder
marked by mental retardation, skin disorders, and skeletal
abnormalities (23).

The reports cited above indicate that there is
secretion of the DSPG in fibroblast cultures, and that this
structure, but not important in maintenance of
extracellular matrix for adhesion of fibroblasts in
culture. This observation bears similarity to the case of

procollagen synthesis and secretion by cultured cells.

Although both procollagen and procollagen processing
enzymes are secreted by cultured fibroblasts (26, 27),

very little collagen is produced (25,26), nor is it an
important component of fibroblast adhesion sites (24).
Interestingly, when fibroblastic cells are grown on
collagen substrata, the secreted DSPG binds strongly to
collagen fibrils in the matrix (15). Apparently, in the
artificial conditions of the culture dish, both procollagen
and DSPG are secreted, but do not associate as they would

1D 3139 to produce an extracellular matrix.

Two other classes of PG are also found in the medium
of fibroblasts. One is a high molecular weight (HMw)
glucuronic acid rich DSPG and the other is a HMw HSPG.

Both of these molecules are prOposed to be cell surface PG
which have been shed into the medium by a process which is
slower than secretion of the LMw DSPG, thus they represent
only a minor fraction of the total extracellular PG
(10,13). Because these molecules are found at the cell
surface, they may play important roles as matrix components
or as cell surface receptors of fibroblasts (30-33). It
may be that they are released from the cell surface by
enzymes during such processes as cell migration and mitosis

(33,40).

Cell Surface Proteoglycans

 

Heparan sulfate is the predominant cell surface PG of
most cultured cells (34). Evidence is accumulating which

suggests that the majority of the cell surface PG of rat

 

hepatocytes (36,37), mammary epithelial cells (35), glial
cells and endothelial cells (38) are intercalated into the
plasma membrane by hydrophobic regions of the core protein.
The structures and sizes of these HSPG vary from one cell
type to the next, but the general pattern is the same. The
cell surface HS are membrane intercalated, but can be
released into the medium by protease activity. Protease
activity releases a soluble PG into the medium, leaving
behind the small, unglycosylated, membrane intercalated
fragment of the protein core. Treatment of cells with
trypsin releases a soluble fragment of similar size to that
which is "shed" into the medium during normal 13 33339 cell
processes. For example, when cells round up in mitosis
(40), or when stimulated to round up by treatment with EGTA
(33) HS molecules are released into the medium, apparently
by an enzymatic activity (33).

While at the cell surface, the HSPG may act as
adhesive molecules, mediating binding of cells to other
extracellular components such as fibronectin (41—43),
laminin (44,45), and collagens (46). While interacting
with extracellular matrix components, the PG may serve as a
linkage molecule between the matrix and the cytoskeleton.
Some evidence indicates co—distribution patterns between
cell surface PG and intracellular actin (47). Moreover,
the core protein of the cell surface PG of mammary
epithelial cells may bind actin (48,49). Interestingly,

this particular core protein is a hybrid, carrying both HS

10

and CS chains (50). As yet, there is no evidence as to
whether the small percentage of CS on the cell surface of
fibroblasts is covalently linked to the cell surface HSPG.

Some of the components of the pericellular matrix are
only loosely associated with the cell surface. These
molecules may be separated from the cells by treatment with
with heparin, which competes for GAG binding Sites on the
cell surface (51,52). Evidence from several laboratories
indicates that the cell can actively bind and endocytose
soluble and matrix associated molecules (53-59), via
receptors that recognize the GAG chains (55,57), or protein
cores (59).

The cell surface PGs of fibroblasts have a half-life
of between 10 and 20 hours. They leave the cell surface by
being either Shed into the medium or internalized. From
45% (60) to 75% (61) of the PG leaving the cell surface of
skin fibroblasts do so by internalization. This makes them
likely candidates as cell surface receptors which might
bind ligands and tranSport them to lysosomes (31).

Indirect evidence, i.e. the competitive inhibition of
binding by GAG chains, indicates that lipOprotein lipase
(62-65) and platelet factor 4 (66) may be bound to cell
surface GAGS on endothelial cells. More conclusive
evidence suggests that the core protein of the cell surface
HSPG of human fibroblasts may be similar to or identical to
transferrin receptors. Fransson et al (67,68) have

demonstrated transferrin binding activity in the HSPG, and

11

that the core protein can bind monoclonal antibodies that
recognize transferrin receptors (67). Moreover, the core
protein, like the transferrin receptor, occurs as a
disulfide linked dimer having a subunit Mr of 80,000 —
100,000. When cleaved by trypsin, the HSPG becomes
soluble, and the core protein is no longer a disulfide
bonded dimer, but is a monomer having an Mr of about
70,000. similarly, trypsin treatment of transferrin

receptor releases soluble monomers with Mr of 70,000 (69).

Intracellular Proteoglycans

Those cell surface PG which are internalized by
fibroblasts become associated mainly with a degradative
compartment (8). Some evidence exists, however, for a
small pool of intracellular GAG chains which have a long
half-life. Fedarko and Conrad (70) have identified a
nuclear fraction of HS, (representing about 6% of the total
cell associated HS) which has an unusual disaccharide
structure present in high amounts. They suggested that,
because of the long transit time to the nucleus (2 hours),
the nuclear HS may be a unique catabolite of the HSPG which
is endocytosed from the cell surface. HS may play a role
in the regulation of nuclear activity, based on the
observation that heparin, when added to isolated nuclei,
stimulates both DNA replication and transcription (71—75).
Heparin also affects the activities of a number of other

enzymes found in the nucleus, including cAMP dependent

12

protein kinase N II (76), and tOpoisomerase I (77).

The remainder of the intracellular GAG appears to be
in a degradative compartment. No function has been
assigned to these intermediate sized fragments of GAG
chains found in the cells. These molecules will be
discussed below in terms of the enzyme systems used by the

cells to degrade them.

Degradation g; GAG chains

The degradation of the GAG chains of internalized
proteoglycans has been studied extensively. However,
little information is available as to whether, or how, the
protein core of internalized PG is degraded, although
there is evidence for extracellular cleavage of core
proteins of PGS (78-81). In cartilage, release of PG
monomers from the matrix results from cleavage near the
hyaluronate binding region of the core protein (82-84).
Once endocytosed, soluble or cell surface core proteins are
rapidly separated from GAG chains. This leads to
difficulty in further analyzing the degradation of core
proteins, since the probes used to study PG usually involve
labeling the GAG chains and/or purification schemes based
on prOperties of GAGS (l-4).

The steps in the degradation of the internalized GAGS
are thought to occur predominantly in lysosomes, although
indirect evidence indicates the possiblity of non—lysosmal
endoglycosidase activity against HS (85). Klein et a1 (86)

have reported that there is an HS endoglucuronidase

13

activity in human fibroblasts, but presented no evidence as
to its cellular localization. Data from other studies
suggest that HS endoglucuronidases are common features of
several cell types and tissues including platelets (87-89),
activated T-lymphocytes (90), rat liver (91), and human
placenta (92,93). The pH optimum for the activity of the
HS endoglycosidase is around 6, which is high for lysosomal
enzymes but somewhat low for neutral enzymes.
Endoglycosidase activity also occurs against CS chains in
chick chondrocytes (94) and liver (95). Moreover, CS is a
substrate of lysosomal hyaluronidase (96,97). However, as
yet, no hyaluronidase activity has been observed in human
fibroblasts (98,99).

The GAG chains which are found in the lysosomes,
whether or not they have been previously acted on by
endoglycosidases, are substrates for the action of several
glycosidases and sulfatases which act sequentially as
exoenzymes at the non-reducing end of the GAG chains (100—
102). The scheme for the degradation of HS is shown in
figure 2. Each enzyme in the pathway changes the structure
of the GAG chain so that it becomes a substrate for the
next enzyme in the pathway. Thus, for HS degradation,
three glycosidases, two sulfatases, a sulfamatase, and an
N-acetyl transferase act in sequence to free monomers from
the polysaccharide (101). Degradation of the
galactosaminoglycans occurs in a similar manner, differing

only in the specificity of the participating enzymes.

||r

14

Figure 2. Schematic representation of the steps in
the degradation of heparan sulfate. The heparan sulfate
chain consists of repeating disaccharides of glucosamine
and either glucuronic acid or iduronic acid. A series of
lysosomal enzymes act at the non—reducing of the chain to
modify (usually by hydrolytic cleavage) the structure of
the chain, making it a substrate for the next enzyme in the
pathway. The diagram illustrates the actions of iduronate
sulfatase (A) which removes the O—sulfate at the 2 position
of iduronic acid. The next enzyme in the pathway is
u—iduronidase (B), which can then remove the iduronic acid
residue. These activities are followed by those of heparan
N—Sulfatase (C), and a Specific lysosomal acetyl coA
transferase (D) which links an acetyl group to the nitrogen
of glucosamine. This alteration makes the terminal monomer
a substrate for a—N—acetylglucosaminidase (E). Sites of
action of,G—g1ucuronidase (F) and N—acetyl glucosamine 6—

sulfatase (G) are also shown.

15

 

 

Hzcon 00H Hzco@ 00H
<42» 0'. O" O O
0(8) l N6)
T
A
H COH Hco© COOH

BJ/ZN N© NAc

H 200B. mm H 2coCs) c00H

C}© is 63‘ m.

H 2cor! mm H 2co© com:

-‘.0- ...... O

0/2E F

 

 

 

by

 

16

If one of the exoenzymes of the GAG degradation
pathway is missing, the GAG chain is not depolymerized
because the substrates for the sucessive enzymes in the
pathway are not uncovered. Much of our knowledge about the
exoenzymes of GAG catabolism has come from studies of human
genetic diseases called mucopolysaccharidoses (MPS). In
cells from MPS patients, the gene for production of a
lysosomal enzyme is defective, resulting in little or no
enzyme activity appearing in the lysosomes (100—104). When
GAG chains are transported to lysosomes in these cells they
are degraded only until the substrate for the missing
enzyme is produced at the reducing end of the GAG chain.
Since cells have a low capacity to clear polymers from
their lysosomes [some lysosomal enzymes have half-lives in
the cell of 30 days (105)], the undegraded GAGS are stored
intracellularly, resulting in a myriad of secondary
molecular, cellular and clinical defects (100-104).

Studies of the MPS diseases over the last 15 years have
facilitated the discovery and/or purification of most of
the exoenzymes involved in the degradation of GAGS. For
example, HS GAG chains require the activity of a number of
exoenzymes in the lysosomes. A group of MP8 diseases exist
(collectively called Sanfilippo disease) which are specific
for H3 (100,102). The several subtypes of Sanfilippo
disease arise from deficiencies of different lysosomal
enzymes, each of which are coded for by different genes,

but have in common the function of participating in the

degradation of HS. Absence of any one of these gene
products results in storage of HS in the cells of patients,
resulting in identical clinical phenotypes. Presently,
four subtypes of Sanfilippo syndrome have been identified;
Sanfilippo type A [heparan N—sulfatase deficiency (106—
108)], type B [6-N—acetylglucosidase (109,110)], type C [N—
acetyl transferase deficiency (lll—,112)], and type D [fi—N—
acetylglucosamine—6—sulfate sulfatase deficiency (113)]
(see figure 2 for details of HS degradation). The
identities of these gene products were discovered because
it was found that the secretions of normal cells contained
corrective factors which could be internalized by diseased
cells and correct the phenotype. Thus, the enzymes could
be purified based on assays of GAG degradation in cultured

fibroblasts (100—104).

252222255 and Distribution 2: Lysosomel EEEXEEE

 

The enzymes of the lysosome, including those
exoenzymes of the pathway of GAG degradation, receive a
chemical tranSport marker in the Golgi which allows them to
be segregated from secretory proteins and tranSported to
lysosomes (114,115). Much of our information about the
pathway for the tranSport of lysosomal enzymes comes from
studies of fibroblastic cells from patients with a genetic
disease known as I—cell disease (mucolipidosis II) (114-
116). Cultured fibroblasts of I—cell patients synthesize

precurser lysosomal enzymes, which, instead of being

18

localized into lysosomes, are secreted into the surrounding
medium (115,117). Other evidence indicated that lysosmal
enzymes from normal cells could be selectivly pinocytosed
by cells via specific cell surface receptors, and
subsequently transported to lysosomes (114,115). When
similar experiments were performed using the enzymes
secreted by I-cells, they were not taken up by normal
cells. In contrast, enzymes from normal cells could be
taken up by I—cells. These data indicated that the
receptors for transport of lysosomal enzymes were
unaffected in I-cell fibroblasts, but the enzymes
themselves were affected such that they were not recognized
by the receptors. Subsequent studies revealed that the
defect of lysosomal enzymes in I-cell patients was an
abnormal chemical structure of the oligosaccharide chains.
Specifically, the oligosaccharides lacked mannose 6-
phosphate (M6P) residues which are necessary components
(117-119) for recognition of the lysosomal enzymes by
specific membrane receptors called M6P receptors (120-122).
It has now been demonstrated that the primary gene defect
in I—cell disease is in the phosphotransferase required for
the construction of the NSF recognition marker on the
oligosaccharides of lysosomal enzymes (123). When the
transferase is missing, lysosomal enzymes do not find their
way to lysosomes, but are instead secreted from the cells.
frherefore, I-cells have a deficiency of the lysosomal

enzymes involved in the degradation of GAGs, and they

 

 

 

 

 

 

19

consequently store partially degraded GAG chains in their
lysosomes (124).

The observation, cited above, that fibroblasts can
bind and endocytose lysosomal enzymes via cell surface
receptors, raised the question whether lysosomal enzymes
are tranSported to lysosomes by first being secreted and
then recaptured (115,125). Long term treatment of cells
with M6P did not inhibit the localization of lysosomal
enzymes to lysosomes (126,127) deSpite the fact that M6P
could compete with lysosomal enzymes for binding sites on
the cell surface, even diSplacing previously bound enzymes
(128,129). These results argue against the secretion-
recapture hypothesis. Recent quantitation of the M6?
receptor Showed that only about 20% of the M6? receptors
were localized to the cell surface (130), whereas the rest
were located inside the cell where they play a role in
transporting lysosomal enzymes via an intracellular route
to the lysosomes.

The localization of lysosomal enzymes to the cell
surface is puzzling since the pH optimum of lysosomal
enzymes is too low for these enzymes to be active in the
neutral conditions of the culture medium (131). Recent
evidence suggests that cell surface lysosomal enzymes may
play a role in the degradation of the PG of the
extracellular matrix of human fibroblasts. EXposure of
cells to M6? [which removes lysosomal enzymes from the cell

surface (128,129)] disrUpted the kinetics of the turnover

 

20

of cell surface GAGS (132,133). Moreover, when cells were
35 -2
$04

labeled PGs, the cells internalized and degraded the P65,

plated on an extracellular matrix containing

and this degradation was inhibited by NSF (134). Thus, it
was proposed that some internalized extracellular matrix
PGs may be degraded in an acid compartment of the cell
which requires a secretion-recapture mechanism for delivery

of lysosomal enzymes.

I.

10.

11.

12.

13.

14.

REFERENCES

Roden, L. (1980) In The Biochemistry g;
Glycoproteins and Proteoglycans (Lennarz, W., ed.),
Plenum, New York. pp. 267—371.

Dorfman, A. (1981) In Cell Biology 9; the Extracellular
Matrix. (Hay, E.D., ed.) Academic Press, New York. pp.
115-138.

 

 

Rodin, L., and Horowitz, M.I. (1978) In The
Glycocongugatesi Vol. 2 (Horowitz, M.I., and Pigman, W.,
ed.) Academic Press, New York. pp. 3-71.

Hascall, V.C., and Hascall, G.K. (1981) In Cell Biology
pf the Extracellular Matrix (Hay, E.D., ed.) Plenum,
New York. pp. 39-63.

 

 

Fellini, S.A., Hascall, V.C., and Kimura, J.H. (1984) J.
Biol. Chem. 259, 4634-4641.

Kimura, J.H., Lohmander, L.S., and Hascall, V.C. (1984)
J. Cell. Biochem. gg, 261-278.

Hoffman, H.—P., Schwartz, N.B., Roden, L., and Prockop,
D.J. (1984) Connect. Tissue Res. 12, 151-163.

Fratantoni, J.C., Hall C.W., and Neufeld E.F. (1968)
Proc. Natl Acad. Sci. U.S.A. g9, 699—706.

Glossyl, J. Beck, M., and Kresse, H. (1984) J. Biol.
Chem. 259, 14144-14150.

Vogel, K.G., and Peterson, D.W. (1981) J. Biol. Chem.
256, 13235-13242.

Coster, L., Carlstedt, I., and Malmstrom, A. (1979)
Biochem. J. 183, 669-681.

Carlstedt, I., Coster, L., and Malmstrom, A. (1981)
Biochem. J. 197217—225.

Coster, 1., Carlstedt, I., Malmstrom, A., and
Sarnstrand, B. (1984) Biochem. J. 220, 575-582.

Vogel, K.G., and Kendall, V.F. (1980) J. Cell. Physiol.
103 475-487.

21

 

 

 

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

Gallagher, J.T., Gasiunas, N., and Schor, S.L. (1983)
Biochem. J. 215, 107—116.

Coster,L., and Fransson, L.~A. (1981) Biochem. J. 193,
143-153.

Damle, S.P., Coster, L., and Gregory, J.D. (1982) J.
Biol. Chem. 257, 5223-5527.

Nakamura, T., Matsunaga, E., and Shinkai, H. (1983)
Biochem. J. 213, 289-296.

Rosenberg, L.C., Choi, G.U., Tang, L.-H., Johnson, T.L.,
Pal, S., Webber, ., Reiner, A., and Poole, A.R. (1985)
J. Biol. Chem. 260, 6304—6313.

Shinkai, H., Nakamura, T., and Matsunaga, E. (1983)
Biochem. J. 213, 297-304.

Heinegard, D., Bjorne—Persson, A., Coster, L., Franzen,
A., Gardell, S., Malmstrom, A., Paulsson, M., Sandfalk,
R., and Vogel, K. (1985) Biochem. J. 230, 181—194.
Scott, J.R., and Orford, C.R. (1981) Biochem. J. 1 7,
213-216.

Beck, M., 610351, J., Ruter, R., and Kresse, H. (1983)
Pediatr. Res. 11, 926-929.

Culp, L.A., and Bensusan, G. (1978) Nature (Lond.)
273, 680-682.

Kleinman, H.K., Klebe, R.J. and Martin, C.R. (1981) J.
Cell Biol. _§, 473-485.

Hay, E.D., (1981) In Cell Biology 9: the Extracellular
Matrix (Hay, E.D., ed.) Plenum, New York. pp. 379—409.
Lichenstein, J.R., Martin, G.R., Kohn, L.D., Byers,
P.H., and McKusick, V.A. (1973) Science 182, 298-300.
Layman, D.L., and Ross, R. (1973) Arch. Biochem.

Biophys. 157,
Goldberg, B.,
A, 45-50.

Bernfield, M.,
(1984)

Development
New York pp.

Taubman, M.B.,

Banerjee, S.,
In Th3 Role pf the Extracellular Matrix 1p
(Trelsted, R.L., ed.) Alan R. Liss, Inc.
545-572. ‘

22

 

 

451-456.

and Radin, B. (1975) Cell

Koda, J., and Rapraeger, A.

 

 

 

 

 

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

23

Hook, M., Kjellan, L., Johansson, 8., and Robinson,
J. (1984) Ann. Rev. Biochem. 53, 847—869.

Rollins, B., Cathcart, M., and Culp, L. (1982) In The
Glycocongugates, Vol. 3 (Horowitz, M., ed.) Academic
Press, pp. 289-329.

Rapraeger, A., and Bernfield, M. (1985) In Diseases 9:
Connective Tissue: Molecular Pathology 93 Egg
Extracellular Matrix (Uitto, J. and Perjda,A., ed.)
Marcell Dekkar, Inc., New York. pp. 2-32.

 

Gallagher, J., Lyon, M., and Steward, W. (1986)
Biochem. J. 2 6, 313-325.

Rapraeger, A., and Bernfield, M. (1983) J. Biol. Chem.
258, 3632-3637.

Kjellen, L., Petterson, I., and Hook, M. (1981) Proc.
Natl. Acad. Sci. U.S.A. 18,5371-5375.

Kjellen, L., Oldberg, A., Hook, M. (1980) J. Biol.
Chem. 255, 10407—10413.

Norling, B., Glimelius, B., Wasteson, A. (1981)
Biochem. Biophys. Res. Comm. 103, 1265-1272.

Rapraeger, A., and Bernfield, M. (1985) J. Biol. Chem.
260, 4103-4109.

Kraemer, P., and Tobey, R. (1972) J. Cell Biol. 55,
713-717.

Ruoslahti, E., and Engvall, E. (1980) Biochem. Biophys.
Acta 631, 350-358.

Yamada, K.M., Kennedy, D.W., Kimata, K., and Pratt, R.M.
(1980) J. Biol. Chem. 2 5, 6055-6063.

Stamatoglou, S., and Keller, J. (1982). Biochim.
Biophys. Acta 7 9, 90—97.

Sakashita, S., Engvall, E., and Ruoslahti, E. (1980)
FEBS letters 116, 243-246.

Woodley, D.T., Rao, C.N., Hassell, J.R., Liotta, L.A.,
Martin, G.R., and Kleinman, H.K. (1983) Biochim.
Biophys, Acta 761, 278—283.

Koda, J., Rapraeger, A., and Bernfield, M. (1985) J.
Biol. Chem. ggp, 8147-8162.

 

 

47.

48.

49.

so.

51.

52.

53.

S4.

55.

56.

S70

58.

59.

60.

61.

62.

24

Woods, A., Hook, M., Kjellen, L., Smith, C.G., and
Rees, D.A. (1984) J. Cell Biol. 35, 1743-1753.

Rapraeger, A., and Bernfield, M. (1982) In 153
Extracellular Matrix (Hawkes, S.P., and Wang, J.L.,
ed.) Academic Press, New York. pp265-269.

Rapraeger, A., Jalkanen, M., and Bernfield, M. (1986)
In Biology 9; the Extracellular Matrix, Vol. VII:
Biology pf Proteoglycans (Wight, T.N., and Mecham,
R.P., ed.) Academic Press, New York. pp 265-269.

Rapraeger, A., Jalkanen, M., Endo, E., Koda J., and
Bernfield, M. (1985) J. Biol. Chem. 269, 11046-11052.

Kjellen, L., Oldberg, A., and Hook, M. (1980)
J. Biol. Chem. 255, 10407-10413.

Kraemer, P.M. (1977) Biochem. BiOphys. Res. Comm. 18,
1334-1340.

Kresse, H., Tekolf, W., von Figura, K., and Buddecke, E.
(1975) Hoppe-Seyler's Z. Physiol. Chem. 356, 943-952.

Truppe, W., and Kresse, H. (1978) Eur. J. Biochem. 85,
351—356.

Prinz, R., Schwermann, J., Buddecke, E., vonFigura, K.

vonFigura, K., Kresse, H., Meinhard, U., and
Holtfrerich, D. (1978) Biochem. J. 119, 313-320.

Glossl, J., Schubert-Prinz, R., Gregory, J.D., Damle,
S.P., vonFigura, K., and Kresse, H. (1983) Biochem. J.
215, 295-301.

Schmidt, A., Buddecke, E. (1985) Eur. J. Biochem. 153,
269-273.

Smedsrod, B., Kjellen, L. and Pertoft, H. (1985)

Kresse, H., von Figura, K., Buddecke, E., and Fromme,
H.G. (1975) HOppe—Seyler's Physiol. Chem. 356, 929-941.

Cowles, E.A., Brauker, J.H., and Anderson, R.L., (1986)
EXp. Cell Res. (In press).

Olivecrona, T., Bengtsson, G., Marklund, S.-E., Lindahl,
U., and Hood, M. (1977) Fed. Proc. 36, 60-65.

 

 

63.

64.

65.

66.

6'1.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

25

Brown, W.V., Wang-Iverson, P., and Paterniti, J.R.
(1981) In Chemistry and Biology 9: Heparin (Lundblad,
W.V., Brown, K. G., Mann, H. and Roberts, R., ed.)
Elsivier/North Holland, Amsterdam. pp. 175-185.

 

 

Cheng, C.-F., Oosta, G.M., Bensadoun, A., and Rosenberg,
R.D. (1981) J. Biol. Chem. 256, 12893-12898.

Shimada, K., Gill, P.-J., Silbert, J.E., Douglas,
W.H.J., and Fanburg B.L. (1981) J. Clin. Invest. 68,
995-1002.

Busch, 0., Dawes, J., Pepper, D.S., and Wasteson, A.
(1980) Thromb. Res. 19, 129-135.

Fransson, L.-A., Carlstedt, I., Coster, L, and
Malmstrom, A. (1984) Proc. Natl. Acad. Sci.
U.S.A. 81, 5657-5661.

Fransson, L.A., Coster, L., Carlstedt, I., and
Malmstrom, A. (1985) Biochem. J. 231, 683-687.

Schnieder, C., Sutherland, R., Newman, R., and Greaves,
M. (1982) J. Biol. Chem. 257, 8516-8522.

Fedarko, N.S., and Conrad, H.E. (1986) J. Cell Biol.
102, 587-599.

Arnold, S.A., Yawn, D.H., Brown, D.S., Wyllie, R.C., and
Coffey, D.S. (1972) J. Cell Biol. 53, 737-757.

Dynan, W.S., and Burgess, R.R. (1981) Biochemistry 18,
4581-4588.

Furakawa, K., and Bhavanandan, V.P. (1983) Biochim.
Biophys. Acta 740, 466-474.

Kovacs, J., Frey, A., and Seifert, K.H. (1981) Biochem
Int. p, 645-653.

Kraemer, R.J., and Coffey, D.S. (1970) Biochim.
Biophys. Acta 224, 553-567.

Hara, T., Takahashi, K., and Endo, H. (1981) FEBS lett.
128, 33-36.

Ishii, K., Katase, A., Andoh, T., and Seno, M. (1982)
Biochem. Biophys. Res. Comm. 104, 541-547.

Caputo, C.B., MacCallum, D.K., Kimura, J.H., Schrode,
J., and Hascall, V.C. (1980) Arch. Biochem. Biophys.
204, 220-233.

 

 

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

26

Galloway, W.A., Murphy, G., Sandy, J.D., Gavrilovic, J
Cawston, T.E. and Reynolds, J.J. (1983) Biochem. J.
209, 741-752.

Heinegard, D., and Hascall, V.C. (1974) J. Biol. Chem.
249, 4250-4256.

Roughly, P.J. (1977) Biochem. J. 157, 639-646.

Handley, C.J., and Lowther, D.A. (1977) Biochim.
Biophys. Acta 500, 132-139.

Sandy, J.D., Brown, H.L.G., and Lowther, D.A. (1978)
Biochim. Biophys. Acta 543, 536-544.

Tyler, J.A. (1985) Biochem. J. 25, 493-507.

Yanagishita, M., and Hascall, V.C. (1984) J. Biol.
Chem. 259, 10270-10283.

Klein, U., Kresse, H., and von Figura, K. (1976)
Biochem. Biophys. Res. Comm. 66, 158-166.

Oosta, G.M., Favreau, L. V., Beeler, D.L., and

Rosenberg, R.D. (1982) J. Biol. Chem. 257, 11248—11255.

Wasteson, A., Hook, M. and Westermark, B. (1976) FEBS
lett. 66, 218-221.

Yahalom, J., Elder, A., Fuks, Z., and Vlodavsky, I.
(1984) J. Clin. Invest. 11, 1842-1849.

Naparstek, Y., Cohen, I.R., Guks, Z., and Vlodavsky, I.
(1984) Nature (London) 310, 241—244.

Hook, M., Wasteson, A., and Oldberg. A. (1975)

Biochem. Biophys. Res. Comm. 61, 1422-1428.

Klein, U., and von Figura, K. (1976) Biochem. Biophys.
Res. Comm. 16, 569-576.

Klein, U., and von Figura, K. (1979) Hoppe-Seyler's Z.
Physiol. Chem. 360, 1465-1471.

Glaser, J.H., and Conrad, H.E. (1979) J. Biol. Chem.
254 2316-2325.

Takagaki, K., Nakamura, T., Majima, M., and Endo, M.
(1985) FEBS lett. 181, 271-274.

Hoffman, P., Meyer, K., and Linker, A. (1956) J. Biol.
Chem. 219, 653-663.

 

 

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111.

112 O

27

Meyer, K., and Rapport, M.M. (1950) Arch. Biochem. 11,
287-293.

Klein, U., and von Figura, K. (1980) Biochim. et
Biophys. Acta 666, 10-14.

Arbogast, B., Hopwood, J.J., and Dorfman, A. (1975)
Biochem. Biophys. Res. Comm. 61, 376-382.

Sly, W.S. (1980) In Metabolic Control and Disease (Brady
P.K., and Rosenberg, L.E., ed.) W.B. Saunders,
Philadelphia, pp. 545-581.

von Figura, K., and Klein, U. (1981) In Lysosomes and
Lysosomal Storage Diseases (Callahan J.W., and Lowden,
J.A., ed.) Raven Press, New York. pp. 229-248.

McKusick, V.A., Neufeld, E.F., and Kelly, T.E. (1983)
In The Metabolic Basis 91 Human Disease (Stanbury,
J.B., Wyngaarden, J.B., Fredrickson, D.S., Goldstein,
J.L., and Brown, M.S., ed.) McGraw-Hill pp.751-777.

 

 

Neufeld, E.F., and Fratantoni, J.C. (1970) Science 169,
141-146.

Neufeld, E.F., and Timple, W.L., and Shapiro, L.J.
(1975) Ann. Rev. Biochem. 44, 357-376.

Skudlarek, M.D., Novak, H.K., and Swank, R.T. (1984) In
lysosomes 1E Biology and Patholggy (Dingle, J.T., Dean,
R.T., and Sly, W., ed.) Elsevier, New York. pp. 17-43.

Kresse, H., and Neufeld, E.F. (1972) J. Biol.
Chem. 247, 2164-2170.

Kresse, H. (1973) Biochem. Biophys. Res. Comm. 66,
1111-1118.

Matalon, R., and Dorfmen, A. (1974) J. Clin. Invest.
22, 907-912.

O'Brien, J.S. (1972) Proc. Natl. Acad. Sci. U.S.A. 66,
1720-1722.

von Figura, K., and Kresse, H. (1972) Biochem. Biophys.
Res. Comm. 66, 262-269.

Kresse, H., von Figure, K., and Klein, U. (1978) Eur.
J. Biochem. 61, 333-339.

Klein, U., Kresse, H., and von Figure, K. (1978) Proc.
Natl. Acad. Sci. U.S.A. 16, 5185—5188.

113.

114

115.

116.

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

28

Kresse, H., Paschke, B., von Figura, K., Gilberf, W.,
Fuchs, W. (1980) Proc. Natl. Acad. Sci. U.S.A. 11,
6822-6826.

Hickman, S., and Neufeld, E.F. (1972) Biochem. Biophys.
Res. Comm. _6, 992-999.

Neufeld, E.F., and Mckusick, V.A., (1983) In 166
Metabolic Basis 21 Human Disease (Stanbury, J.B.,
Wyngaarden, J.B., Fredrickson, D.S., Goldstein, J.L.,
and Brown, M.S., ed.) McGraw-Hill. pp 778-787.

Miller, A.L., Kress, B.C., Lewis, L., Stein, R.,
and Kinnon, C. (1980) Biochem. J. 6, 971-975.

Kaplan, A., Achord, D.T., and Sly, W.S. (1977) Proc.
Natl. Acad. Sci. U.S.A. _6. 2026-2030.

Hasilik, A., Rome, L.H., and Neufeld, E.F. (1979) Fed.
Proc. 66, 467.

Natowitz, M.R., Chi. M.-M., Lowry, D.H., and Sly, W.S.
(1979) Proc. Natl. Acad. Sci. 16, 4322-4326.

Sahagian, G.G., Distler, J., and Jourdian, G.W. (1981)
Proc. Natl. Acad. Sci. U.S.A. 16, 4289-4293.

Steiner, A.W., and Rome, L.H. (1982) Arch. Biochem.
Biophys. 214, 681-687.

Fischer, H.D., Creek, K.H., and Sly, W.S. (1982) J.
Biol. Chem. 257, 9938-9943.

Reitman, M.L., Varki, A., and Kornfeld, S. (1981) J.
Clin. Invest. 61, 1574-1579.

Schmickel, R.D., Distler, J.J., and Jourdian, G.W.
(1975) J. Lab. Clin. Med. 86, 672-682.

Neufeld, E.F., Sando, G.N., Garvin, A.J., and Rome, L.H.
(1977) J. Supra. Struct. 6, 95-101.

von Figura, K., and Weber, E. (1978) Biochem. J. 1 6,
943-956.

Vladutiu, G.D., and Rattazze, M. (1979) J. Clin.
Invest. 66, 595.

Gonzalez-Noriega, A., Grubb, J.H., Talkad, V., and Sly,
W.S. (1980) J. Cell Biol. 66, 839-852.

Rome, L.H., Weissman, B., and Neufeld, E.F. (1979)
Proc. Natl. Acad. Sci. U.S.A. 16, 2331-2334.

130.

131.

132.

133.

134.

29

Fischer, H.D., Gonzalez-Noriega, A., and Sly, W.S.
(1980) J. Biol. Chem. 255, 5069.

Dean, R.T., and Barrett, A.J. (1976) Essays Biochem.
lg: 1'40.

Roff, C.F., Wozniak, R.W., Blenis, J., and Wang, J.L.
(1983) Exp. Cell Res. 166. 333—344.

Roff, C.F., and Wozniak, R.W. (1982) In Extracellular
Matrix (Hawkes, S.P., and Wang, J.L., ed.) Academic
Press, New York. pp. 329-333.

Brauker, J.H., Roff, C.F., and Wang, J.L. (1986) Exp.
Cell Res. 166, 115-126.

 

 

 

Chapter 2

The Effect of Mannose 6—phosphate on the Turnover
of the Proteoglycans in the Extracellular

Matrix of Human Fibroblasts

 

James H. Brauker, Calvin F. Roff and John L. Wang

Department of Biochemistry
Michigan State University

East lansing, MI 48824

30

 

 

SUMMARY

Human fibroblasts (SL66) were cultured in medium containing 358042”

 

to label the glycosaminoglycans. The cells were then detached from the
culture dish to leave radioactively-labeled components of the extracellu-

lar matrix, hereafter termed 35S-labeled substrate-attached material.

 

When unlabeled SL66 fibroblasts were plated onto this 35S-labeled
substrateeattached material, the cells mediated two distinct events:
(a) release of radioactivity from the substrate-attached material into
the medium; and (b) degradation of certain glycosaminoglycans into
radioactive components of very low molecular weight, including free
radioactive sulfate. In the presence of mannose 6ephosphate, however,
the degradation of the substrate-attached material by SL66 cells was
partially inhibited. Analyses of this effect in terms of the dose-
response curve, saccharide specificity, ammonium chloride sensitivity,
and the requirement for cells suggest that both an intracellular com-
partment and the mannose 6-phosphate receptor that binds lysosomal
enzymes at the cell surface may play important roles in the turnover and

degradation of certain proteoglycans in substrate-attached material.

31

 

INTRODUCTION

Fibroblasts adhere to tissue culture substrata at certain regions
of the cell undersurface called "foot pads" (1). Upon cell movement
over the substratum, these foot pads are pinched off from the posterior
of the cell and remain bound to the substratum. Culp and co-workers
showed that the chemical components of adhesion sites from 3T3 fibro-
blasts could be isolated as substratum attached material (SAM) upon
EGTA-mediated detachment of the cells (2—“). Biochemical analysis of
the 3T3 cell adhesive material indicates that it is composed of the
glycosaminoglycans (GAGS) heparan sulfate, chondroitin sulfate, and
hyaluronic acid, as well as cytoskeletal proteins and cellular fibro-
nectin. Both the quantity and the types of proteoglycans found in SAM
change with time in cell culture, possibly as a result of cell movement
and reattachment of new adhesion sites (5,6). Therefore, the turnover
and metabolism of the GAGS may play an important role in determining the
anchorage and motility of cells.

Although it is generally thought that the primary site of the
breakdown of the GAGS is the lysosome (7), some evidence has accumulated
that partial degradation of the GAGs might occur in the extracellular
space. Lark and Culp have reported the existence of a low molecular
weight, single chain heparan sulfate in the SAM of cultured Balb/cSVT2
cells (6). It has also been reported that 816 melanoma cells release

heparan sulfate from an isolated extracellular matrix as fragments

32

 

 

 

33

approximately one-third their original size (8). Apparently, this ,
degradation is extra-lysosomal since passage of the GAGS through the

lysosome would be expected to result in complete degradation of the

molecules. This process is nevertheless thought to involve lysosomal

enzymes inasmuch as the pH optimum for this activity was found to be

approximately pH 5. Extracellular acid hydrolase activity is not an

 

unprecedented idea; several tumor cell lines have been shown to secrete
active lysosomal enzymes (9-12).

Recent work by several laboratories has clearly shown that the

 

cellular distribution of membrane bound lysosomal enzymes includes the
cell surface (13-15). Moreover, it has been shown that these enzymes
contain D-mannopyranoside 6-phosphate (M6P) residues (16-18) and can
bind to a specific carbohydrate binding protein at the cell surface
called the mannose 6-phosphate receptor (19,20). In previous studies,
we had observed that treatment of normal human fibroblasts with 10 mM
M6P decreased the turnover of sulfated GAGs from the cell surface
(pericellular compartment) into the surrounding fluid medium (extra-
cellular compartment) of the culture dish (21,22). It was suggested
that normal turnover of some pericellular GAGS requires cell surface
associated lysosomal enzymes and that dissociation of lysosomal enzymes
by M6P caused an increased retention of GAGS at the cell surface. Since
the labeling conditions of these experiments did not differentiate
between plasma membrane and SAM associated GAGs, the present effort was
undertaken to determine if the turnover of sulfated GAGS in SAM is
affected when lysosomal enzymes are dissociated from the cell surface by

M6P. The data presented in the present communication show that M6P

34

inhibits the degradation by human fibroblasts of a fraction of the ,

sulfated GAGs in isolated SAM.

 

 

MATERIALS AND METHODS

Chemicals

Carrier free H235SOu was obtained from ICN. All monosaccharides
were purchased from Sigma and were of the D-configuration. Phosphory-
lated sugars were purchased as mono- or disodium salts. Pronase was

from Calbiochem~Behring Corp. EGTA was from Sigma.

Cell Lines and culture conditions

 

The normal human foreskin fibroblast cell line SL66 (23) was a gift
of Dr.V. Maher and Dr. J. McCormick (Carcinogenesis Laboratory, Michigan
State University). Fibroblast lines from mucolipidosis II subjects (GM
2933, I~cells) were from the Human Mutant Cell Respository. All cells
were maintained in Eagle's MEM (K.C. Biologicals) with 20% fetal bovine
serum (FBS, Gibco), 100 ug/ml streptomycin (Sigma), and 159 IU/ml peni~
cillin (Sigma), at 35°C in humidified air containing 5% C02. Stock SL66
cells were not allowed to reach confluence and medium was changed two
times per week. Medium for the GM 2933 cell line was changed every two
days and was supplemented with two times the normal concentrations of
amino acids and vitamins. GM 2933 cells were split 1:2 every 6-9 days.

Cells were used between passages 10 and 30.

35

 

 

36

Preparation of Radiolabeled SAM

 

Procedures for preparation and digestion of labeled SAM are illus-
trated in fig. 1. Cultures were labeled by seeding 1.0 X 106 cells in
streptomycin-free MEM (20% FBS, uo-so uCi/ml H235sou), with MgC12
substituted for MgSOq, onto 100 mm tissue culture dishes. When M6P was
included in the radiolabeling medium, the medium was brought to 10 mM
M6P by addition of concentrated M6P in MEM. Some experiments reported
here used SAM prepared in the presence of M6P; similar results were
obtained irrespective of whether the SAM was prepared in the presence or
absence of the saccharide. After 72 hours, the radiolabeling medium was
removed, the monolayers were washed two times with 5 ml of 5 mM EGTA in
phosphate buffered saline (PBS), and incubated for 30-H5 minutes in 10
ml of the 5 mM EGTA in PBS on a rotary shaker (60 cycles/minute) at 35°C
to remove the cells. Undetached cells were removed by gentle pipeting.
The isolated SAM was washed two times with 5 m1 of 5 mM EGTA in PBS,
followed by two washes with 5 ml sterile, glass distilled water to lyse
any remaining cells, and stored at 35°C in 5 ml MEM for not more than 30

minutes.

Digestion of SAM by Unlabeled Cells or Conditioned Medium

 

Unlabeled cells used to digest SAM were removed from stock flasks
by adding EDTA (.02% in PBS) for 10 minutes followed by 10 minutes in
0.0M% trypsin (Gibco). After the cells were completely detached, the
trypsin solution was brought to 20% FBS by addition of PBS and centri-
fuged for 3 minutes at 1330 X g. The cell pellet was suspended in a
small volume of MEM containing 5% FBS and recentrifuged. The pellet was

resuspended in MEM and 3.0 x 106 cells were added to each dish of

 

 

 

37

Figure 1: Protocol for preparation of radiolabeled SAM and subse-
quent digestion by SL66 cells. Cells (1 x 106) were incubated with
“0-50 uCi H23580u for 72 hours and then detached with EGTA. Unlabeled
cells were plated on the radiolabeled SAM in the presence or absence of
M6P. After 2“ hours, the medium was collected and analyzed by column

chromatography.

 

38

35
CELLS, H2 so4

 
      

H o o o 000000000000
>.M.o.o.o.o.o.u.uo.o.o.o.o.o.o.o.«

 

EGTA

 

\/

SAM

SL66

 

 

REMOVE
MEDIUM

 

W

Assay by
Gel Filtration

39

isolated SAM in A ml MEM (without FBS). Saccharides and NHuCl were 3
prepared as stock solutions in MEM at 50 mM or 5 mM concentrations,
passed through 0.45 um filters (Millipore), and added to the culture
dish to give the desired final concentration in a total volume of 5 ml.
After completion of the digestion period, the soluble radioactivity was
harvested for analysis by gel filtration (fig. 1).
Medium was conditioned by exposure to confluent monolayers of SL66
fibroblasts. Cells were grown to confluence in 150 cm2 flasks in 20 ml
MEM containing 20% F88. The medium was removed and 10 ml of fresh

medium was added. After 2” hours, the medium was decanted into 10 ml

 

plastic centrifuge tubes and centrifuged for 3 minutes at 1330 X g to
remove cellular debris. This preparation was called conditioned medium
(CM) and was used to digest 35SOu-blabeled SAM in the absence of cells.
Five ml of the CM were added to [35$]SAM for 2H hours after which the

solubilized radioactivity was analyzed by gel filtration.

Harvest and Bulk Quantitation of Radioactivity

 

The culture dishes were divided into three compartments called the
SAM, cellular, and soluble compartments. The soluble material was
defined as the radioactivity removable with the medium. The cellular
compartment was defined as the radioactivity in the EGTA solution used
to remove the cells. The SAM was defined as the radioactivity left on
the dish after removal of the cells by EGTA. SAM was harvested with
0.2% sodium dodecyl sulfate (SDS, Sigma) in glass distilled water.
Alternatively, the SAM and cellular compartments were harvested together
using 0.2% SDS and 0.05 M NaOH in glass distilled water, and called the

insoluble compartment. Total radioactivity consisted of soluble radio-

 

 

 

 

 

 

40

activity plus insoluble (cellular and SAM) radioactivity. Aliquots from
each fraction were solubilized at a ratio of 1:9 (sample: scintillation
cocktail) in scintillation fluid consisting of 7 g 2,5-diphenyloxazole
(Research Products International Corp.), 338 ml Triton X-100 and 667 m1
toluene. Radioactivity was determined on a Packard Tricarb 300C liquid
scintillation counter and expressed as counts per minute (cpm) or dis-

integrations per minute (dpm).

Quantitation of Degraded Material

 

The soluble fraction was harvested, boiled for 5 minutes, and
stored at ~20°C until analyzed. After thawing, 1.5 m1 of the soluble
fraction were boiled for 1 minute; 0.5 ml of 0.6 M Tris, pH 7.“, 0.8%
SDS and A mM NaN3 were added and the sample boiled for 1 minute more.
After cooling, 0.5 ml were diluted in A.5 ml of scintillation cocktail
and counted to determine total dpm applied to the column. The remaining
1.5 ml were loaded on a Sephadex G-25 column (1.3 X 35 cm) equilibrated
with 0.15 M Tris, pH 7.4, 0.2% SDS and 1 mM NaN3. Fractions of 1.0 ml
were collected directly in scintillation vials. The radioactivity in
each fraction was determined by liquid scintillation counting. The
radioactivity that was not excluded from the column was called degraded
[35$]proteoglycan ([35SJPG). This material eluted at the same position
as free 3580u2‘. When free 3580u2‘ containing a known amount of radio-
activity was chromatographed in the presence of SL66 conditioned medium,
greater than 95% of the radioactivity was recovered. Percent degraded
[35SJPG was determined by the calulation:

dpm degraded

% Degraded = ~~~~~~~~~~~~~ X 100
dpm loaded

 

 

41

Analysis of SAM Fractions by Column Chromatography

 

The SAM fraction was harvested after completion of the radiolabel-
ing period, boiled for 5 minutes and lyophilized. The dried material
was suspended in 1 ml of 0.1 M acetic acidéNHAOH, pH 5.7, 0.013 M EDTA,
and 0.5 mM NaN3 and the pH was adjusted to 5.7 with acetic acid.
Pronase (A mg) was added and the mixture was incubated at 50°C for 2H
hours. More pronase (N mg) was added, and the incubation was continued
for an additional 2H hours. During the digestion, the pH was adjusted
to 5.7 when necessary. The samples were neutralized with 1 M NaOH and
stored at ~20°C.

Samples were chromatographed on columns (110 x 1 cm) of Sepharose
CL6B equilibrated with 0.2% SDS, 0.15 M Tris, pH 7.4, 1 mM NaN3.
Fractions of 1.0 ml were collected and the radioactivity was determined

by scintillation counting.

RESULTS

Characterization of 3580u-labeled SAM by Column Chromatography

 

The main goal of this study was to test the hypothesis that
fibroblasts can degrade radioactive components in SAM and whether this
digestion is sensitive to M6P. To determine if the turnover of compo-
nents of SAM is affected by M6P, the chromatographic properties of SAM
labeled with 35$Oh2“ were analyzed. Radioactive SAM produced by SL66
cells in the presence or absence of M6? was chromatographed on Sepharose
CL6B columns. The chromatographic profile of SAM prepared in the
presence of M6P showed a large peak of radioactivity, representing
35S-labeled GAGs of high molecular weight, near the void volume of the
column (fig. 2). Pronase treatment of the same material resulted in a
reduction of the molecular weight of the components. Therefore, the
majority of the GAGs in SAM were associated with protein, presumably as
proteoglycans (PG). There was little or no radioactivity eluted at the
total volume of the column.

Similar chromatographic profiles were obtained when SAM prepared in
the absence of M6? was analyzed on Sepharose CL6B columns. Moreover,
I~cells, which produce lysosomal enzymes lacking the M6? marker required
for intracellular transport and therefore are deficient in certain

hydrolases in the lysosome (2”), also yielded similar results.

42

43

Figure 2: Chromatographic profile of SAM components before
(o—-——o) and after (o-e-eo) digestion with pronase. Fibroblasts (SL66)
were labeled with 35SOu2“ for 72 hours in the presence of M6P (10 mM),
after which the cells were removed using EGTA. The radioactive SAM was
solubilized in 0.2% SDS. Samples were incubated with and without
pronase and then analyzed on a Sepharose CL6B column (106 x 1 cm)
equilibrated with 0.2% SDS in 0.15 M Tris Buffer, pH 7.”, 1 mM NaN3.

Fractions of 1.0 ml were collected. Void volume, V; total volume, T.

Radioactivity (dpm x IO'Z)

44

 

20—

5
i

 

 

 

H Before
Pronase

O- -0 After
Pronase

 

50
Fraction

 

45

Solubilization of SAM Components into the Medium

 

Perturbation by M6P of the turnover of PC by fibroblasts might be
expected to affect the quantity or composition of molecules released
from the SAM into the surrounding medium. To test this possibility,
353042“ labeled SAM was subjected to digestion by SL66 cells. At vari-
ous times, the medium was harvested and the amount of bulk radioactivity
released from the SAM into the medium was determined (fig. 3).

Over a period of 10 hours, SL66 cells solubilized more than 40%
(15,000 dpm) of the radioactivity originally associated with SAM and
released it into the medium (fig. 3). This effect required the presence
of cells since only 15% (6,000 dpm) of the radioactivity can be released
over the same time period when SAM was incubated with culture medium
alone (MEM). After 24 hours, the level of radioactivity released by the
cells did not change significantly. The kinetics of the release of
radioactivity from SAM shown in fig. 3 is representative of all condi-
tions employed. The amount of radioactivity released depended only on
the presence of cells; it was not affected by the addition of sac-

charides.

Column Chromatography of Solubilized SAM Components

 

Analysis of the solubilized SAM components by Sepharose CL6B
chromatography separated the radioactivity into two distinct peaks (fig.
4). The first peak (component A) eluted at the void volume of the
column (Mr > 1 x 106). This peak is believed to consist of high
molecular weight PG since pronase treatment of the sample reduced the
molecular weight of component A (data not shown). The second peak

(component B) eluted at the total volume of the column. Several lines

by

46

Figure 3: Time course of release of radioactivity from SAM.
Radiolabeled SAM was prepared as described in the legend to Figure 2.
It was then treated with medium with (o————o) or without (o-seso) SL66
cells and the radioactivity released into the medium was analyzed at

various times as described in Materials and Methods.

 

Soluble 35$ Radioactivity (dpm x IO'S)

47

 

l8r

E
l

m
l

 

 

___.._..———-O
5’”...—

H cells present
o—-o cells absent

1 I l 1
IO IS 20 25

Time(Hours)

 

 

 

48

of evidence indicate that component B consists of 35S-labeled material
of very low molecular weight, including free 358013". First, cetylpy-
ridinium chloride failed to precipitate the radioactivity, indicating
that the material was not polyanionic (25). Second, the radioactive
material in component B was dialyzable in dialysis tubing with a
molecular weight cut off limit of 12,000-14,000. Third, Sephadex G~25
columns, which are used to remove unincorporated 3580h2’ from prepara-
tions of metabolically labeled GAGs (26), also resolved the material
released from SAM into two peaks having relative proportions identical
to those observed on Sepharose CL6B columns. Finally, free 35SOu2'
eluted at the same position as Component B on both Sephadex 0-25 and
Biogel P2 columns. Component B was observed only in the material
released from SAM when SL66 cells are present; incubation of SAM with
medium alone failed to yield this low molecular weight material (see
below). Therefore, Component B represents SAM material extensively
degraded by cells and will be referred to as degraded [35SJPG.

When SL66 cells were plated on labeled SAM in the absence of M6P,
the amount of degraded [35$]PG was always greater than the corresponding
fraction from M6P-treated cultures (fig. 4). This difference was offset
by a corresponding decrease in the quantity of Component A in control
cultures compared to M6P treated cultures. However, there was no net
difference in total amount of radioactivity released, as pointed out
previously.

Degraded [3SSJPG was quantitated by analysis of the radioactive
material released from SAM on Sephadex G-25 columns and is expressed as
a percent of the total radioactivity loaded onto the column. Analysis

of samples from 24-hour incubations showed that the material released

 

49

Figure 4: Sepharose CL6B chromatography of soluble radiolabeled
material released from SAM by SL66 cells. Radiolabeled SAM was prepared
and subjected to digestion by SL66 cells in the presence (o~———c) or
absence (o----o) of 10 mM M6P. After 24 hours the medium was removed
and chromatographed on a Sepharose CL6B column (110 x 1 cm) equilibrated
with 0.2% SDS, 0.15 M Tris, pH 7.4, 1 mM NaN3. Fractions of 1.0 ml were

collected. The arrow indicates the total volume of the column.

 

50

 

 

0—4 M6P

°'“<> CONTROL

15-

L
N O) ‘9
...

lz-Ol x del Aimuoeoipeu sst

 

 

\
0

Fraction

51

from cultures without M6P contained about 20% degraded [3SSJPG, whereas
that from M6P treated cultures contained about 10% degraded [3SSJPG
(table 1, line A). This represents about a 50% inhibition in conversion

of PG to low molecular weight degradation products.

Kinetics, Dose Response Curve and Saccharide Specificity of the M6P

 

Effect

The amount of radioactivity found in the medium in the form of
degraded [3SSJPG was quantitated at various times after the addition of
SL66 fibroblasts (fig. 5). A rapid increase in degraded [35SJPG was
observed during the first 10 hours, followed by a slower, gradual
increase. At all time points, digestion of SAM by SL66 cells in the
absence of M6P always produced a higher percentage of degraded [35$]PG
than digestion of SAM by the cells in the presence of M6P.

Analysis of the doseeresponse curve of the M6? effect on the pro~
duction of degraded [3581PG by SL66 cells showed that the effect was
saturable at a saccharide concentration of 1 mM or above (fig. 6). The
concentration of M6P required for half~maximal inhibition was estimated
to be 60 uM.

The saccharide specificity of the effect of M6P on the production
of degraded [35$]PG from SAM by SL66 fibroblasts was studied by quanti-
tating the percentage of degraded [3SSJPG in cultures containing various
saccharides (table 2). Cultures containing fructose 1,6édiphosphate and
M6? showed the effect (38 and 45% inhibition, respectively) to a much
greater extent than both phosphorylated (mannose lephosphate) and

non~phosphorylated (mannose) analogs of M6P (15*19% inhibition).

 

 

 

 

52

 

TABLE 1 Quantitation of degraded [35$]PG after various incubations.a

 

DEGRADED [35SJPC

 

 

INCUBATIONS n93 CONTROL
A. SL66 9.9 1 0.35b 19.7 1 1.80b
B. MEM 0.0 0.0

C. CM 0.0 0.0

D. I-cell 1.0 4.0

E. SL66 + NHhCl 9.1 : 2.3b 9.9 :0.21b

 

a35S~labeled SAM was prepared as described in materials and methods.
The SAM was incubated in the presence of the various solutions for 24
hours, after which the medium was harvested and degraded [35$]PG
determined by gel filtration. The concentrations of NHuCl and M6P were
10 mM.

bStandard deviations calculated for triplicate determinations.

53

Figure 5: Time course of the generation of degraded [35SJPG by
SL66 cells from SAM. Radiolabeled SAM was subjected to digestion by
SL66 cells for various times in the presence (6————o) or absence
(osé~*o) of 10 mM M6P, after which the medium was removed and chroma-
tographed on Sephadex G~25 and the % degraded [35SJPG was determined as

described in Materials and Methods.

54

 

°/. Degraded (”3) PG

 

    

H M6P

 

O--O CONTROL
1 1 H 1
IO 20 ll 40

TIME (Hours)

 

55

Figure 6: Dose response of M6P effect on the generation of
degraded [35SJPG from SAM by SL66 cells. Radiolabeled SAM was digested
by SL66 cells in the presence of various concentrations of M6P. After
24 hours, the medium was removed and chromatographed on Sephadex Ge25
and the % degraded [35SJPG was determined as described in Materials and

Methods.

56

 

 

 

 

40F

chant

 

O
3

.8858 as.

IO

 

(M6P)(mM)

57

 

TABLE 2 Quantitation of degraded [35SJPG produced by SL66 cells in
the presence of various saccharides.a

 

 

 

 

Saccharides Degraded [35SJPG (%)b % Inhibitionc
None 24.2 ~
mannose 20.6 14.8
mannose 1~phosphate 19.5 19.4
glucose 1~phosphate 19.7 18.5
fructose 1,6ediphosphate 15.0 I 38.0

M6P 13.4 44.6

 

a35S-labeled SAM was prepared as described in materials and methods.
The radiolabeled SAM was incubated in the presence of SL66 cells for 8
hours, after which the medium was collected and the % degraded [35SJPG
determined by gel filtration. The concentration of saccharides was 1
mM.

bMean of duplicate cultures.
0% inhibition was calculated using the formula:

% degraded (saccharide)
% degraded (control)

 

 

 

58

Evidence that M6P Affects a CellsAssociated Compartment

 

When 358*labeled SAM was incubated with culture medium alone, there
was a release of radioactivity from the SAM into the medium (fig. 3).
However, all of this released radioactivity consisted of high molecular
weight material; no degraded [35SJPG was produced (table 1, line.B).

The SAM used in these experiments contained no radioactive material
eluting at the position of degraded [3SSJPG on Sepharose CL6B columns
(fig. 2). Therefore, the lack of degraded [3SSJPG after SAM had been
incubated with culture medium suggests the generation of degraded
[35SJPG was not due to an autolytic process within the SAM.

When 35Selabeled SAM was incubated with CM, about 30-40% of the
total radioactivity was released into the medium in 24 hours. This is
similar to the level of radioactivity released from SAM by SL66 cells.
Presumably, this is the result of either enzymatic activity in the CM or
the exchange with unlabeled SAM components accumulated in the CM. In
any case, no degraded [35SJPG was observed under these conditions (table
1, line C). These results Suggest that when SAM is incubated with SL66
cells, the resultant production of degraded [3SSJPG is not due to
soluble factors secreted into the medium.

When Iécells were plated on 35Sélabeled SAM, only low amounts of
degraded [3SSJPG were produced (table 1, line D). The Iecells are known
to lack hydrolases in the lysosomes and to have a high level of secre-
tion of the same enzymes (24). Therefore, the present results indicate
that even in the presence of elevated extracellular lysosomal enzymes,
very little degraded [3SSJPG is produced.

GonzalezsNoriega et al. (27) reported that treatment of cells with

NHuCl caused elevated secretion of lysosomal enzymes. Moreover, NHuCl

59

is a potent inhibitor of intracellular lysosomal hydrolase activity
because it increases intralysosomal pH (28). When SL66 cells were
plated on 358~labeled SAM in the presence of NHuCl (10 mM), the degrada-
tion and release of [35$]PG was reduced in a fashion similar to that
observed with M6? (Table 1, line E). The presence of both NHuCl and M6P

did not have any effects beyond that observed with either reagent alone.

DISCUSSION

The experiments presented here document several key features con-
cerning the turnover of GAG components in SAM. First, 35SOu2'-labeled
molecules from SAM produced by SL66 cells in the presence and absence of
M6? consist mainly of proteoglycans, similar in molecular weight to
those isolated from the SAM of mouse 3T3 fibroblasts (1). Second, when
SL66 fibroblasts are plated onto SAM, GAG containing components are
released from the SAM. Some of these remain in the medium as macro-
molecules and others are degraded. Degradation of the GAGS is mediated
by cells and not by secreted acid hydrolases, since incubation with
conditioned media or with cells oversecreting lysosomal enzymes (I-cells
and NHuCl treated SL66 cells) actually results in decreased degradation
of GAGS. Third, degradation of GAGS from SAM is inhibited by 40-50%
when M6? is included in the medium.

Several lines of evidence indicate that the effect of M6P is
specifically mediated through the cell surface M6P-receptor that binds
to lysosomal enzymes. Analysis of the dose-response curve of M6?
inhibition showed that it was a saturable effect and the value of K1,
estimated at half-maximal inhibition, was 60 uM. This is in close
agreement with that reported for the inhibition of binding and uptake
via the M6? receptor of a-L-iduronidase by iduronidase deficient cells
(13,16) and with that reported for the bindng of B-galactosidase to the

M6P receptor purified from liver plasma membranes (19). In previous

60

 

61

experiments (21,22), we had estimated a value of Ki for the effect of
M6? on the turnover of the GAGS of the cell surface of SL66 fibroblasts.
This value (60 uM) is in agreement with the Ki estimated by the present
assay.

In addition, the specificity of the effect was restricted to M6P
and fructose 1,6~diphosphate. Both of these phoshorylated saccharides
have low Ki values and are strong inhibitors of the binding and uptake
of lysosomal enzymes via the cell surface M6P receptor (13). In con-
trast, phosphorylated and non-phosphorylated analogs of M6P failed to
yield the same effect, in accord with the specificity of M6? receptor-
lysosomal enzyme interaction (13,14,16). Glucose 1-phosphate, which
does inhibit lysosomal enzyme binding and uptake, but with a high K1
value (13), did not mimic the effect of M6P at the concentration (1 mM)
tested.

Finally, when the radioactive SAM was subjected to digestion by
I-cells, which lack lysosomal enzymes carrying the M6P marker (18,24),
there was little degradation of high molecular weight GAGS into material
of lower molecular weight. The Iécells secrete high levels of lysosomal
enzymes that do not bind to cell surface M6P receptors. Similarly,
NHACl, which raises the pH of intracellular compartments, inhibits
lysosomal function, disrupts turnover of cell surface receptors, and
increases secretion of lysosomal enzymes (27,28), reduced the degrada-
tion of proteoglycans to the same extent as M6P. The presence of both
NHuCl and M6? yielded effects no different than those of either reagent
alone. Therefore, the results indicate that the generation of degraded
[35$]PG from high molecular weight molecules is not mediated by elevated

levels of extracellular, soluble lysosomal enzymes.

62

It should be noted that M6P inhibits the degradation of some, but
not all, of the sulfated PG derived from SAM. In the presence of M6P, a
significant quantity of [35$]PG is still degraded by cells to low
molecular weight components. This represents ~60% of the total degraded
[35SJPG produced by control cells. It is likely that the effect of M6?
is exerted at the population of M6? receptor-acid hydrolase complexes
residing at the cell surface. In fibroblasts, sorting of newly synthe-
sized lysosomal enzymes into a pathway leading to lysosomes is mediated
by M6? receptors (14-19). This appears to be an intracellular process,
unaffected by incubating cells in the presence of M6P. Therefore, the
majority of receptor-enzyme complexes most likely do not reach the
plasma membrane. Nor does M6P enter the cell and dissociate receptor~
enzyme complexes at sites prior to commitment of the enzymes into a
pathway leading to lysosomes. Consistent with this is the finding that
M6P treatment of cells does not cause intracellular accumulation of
undegraded GAGS (21). In fibroblasts, some acid hydrolasesM6P receptor
complexes are found at the cell surface, due either to transport of
complexes from internal membranes or binding of secreted enzymes to
unoccupied surface receptors (14,15). This binding is prevented and
dissociation is facilitated by inclusion of M6P in the medium.

How M6P prevents degradation of a population of GAGS derived from
SAM is unknown. The saccharide does not inhibit endocytosis of GAGS
derived from SAM (Brauker, J.H., Roff, C.F. and Wang, J.L., unpublished
observation). Instead, M6P may cause accumulation of lysosomal enzymes
in the medium; some of these activities may result in the removal of

recognition markers on PGs required for binding and endocytosis by cells.

 

 

63

Alternatively, surface bound hydrolases may be required to solubilize
GAGS in a form that is endocytosed by fibroblasts.

Degradation requires internalization of the GAGS. Recently,
compartments other than lysosomes, possibly including endocytotic
vesicles, have been demonstrated to be involved in degradation of GAGS
(29,30). Endocytic vesicles contain acid hydrolases and become acidic.
Therefore, if endocytosed GAGS enter these vesicles, conditions for
degradation may exist. In fibroblasts, a major pathway of entry of acid
hydrolases into endocytic vesicles may be via cell surface receptors.

If this and the concept that endocytic vesicles function as sites of
degradation are true, then displacement of lysosomal enzymes from the
cell surface M6P receptors would be the basis for the observed reduction
in the rate of degradation of PG in the presence of M6P. Furthermore,
this would support the original notion of the secretionsrecapture
hypothesis for the transport of lysosomal enzymes proposed by Neufeld et
al. (31). The only modification is that transport and sorting vesicles,
rather than lysosomes, are the primary target for delivery of lysosomal

enzymes by a cell surface pathway.

ACKNOWLEDGEMENTS

This work was previously published in Experimental Cell
Research, Volume 164 (1986) pp. 115-126, and is reproduced with
permission fromthe publisher. Copywright © 1986 by Academic
Press Inc. All rights of reproduction in any form reserved

0014—4827/86 $03.00.

64

10.
ll.

12.

13.
14.

15.
16.

REFERENCES

Culp, LA, Rollins, BA & Rollins, BJ, J supramol struct 11 (1979)
281.

Culp, LA, Biochemistry 15 (1976) 4094.

Rollins, BA & Culp, LA, Biochemistry 18 (1979) 141.

Rollins, BA & Culp, LA, Biochemistry 18 (1979) 5627.

Lark, MW & Culp, LA, Biochemistry 22 (1983) 2289.

Lark, MW & Culp, LA, J biol chem 259 (1984) 212.

Roden, L, The biochemistry of glycoproteins and proteoglycans (ed
WJ Lennarz) p. 267. Plenum Press, New York (1980).

Nakajima, M, Irimura, T, DiFerrante, D, DiFerrante, N, & Nicolson,
GL, Science 220 (1983) 611.

Recklies, AD, Poole, AR & Mort, JS, Biochem j 207 (1982) 633.
Sylven, B, Eur j cancer 4 (1968) 463.

Weiss, L, Int j cancer 22 (1978) 196.

Poole, AR, Lysosomes in biology and pathology (ed JT Dingle) vol 3
p. 303. Elsevier/North Holland, Amsterstam (1978).

Sando, GN & Neufeld, EF, Cell 12 (1977) 619.

Fischer, HD, Gonzalez~Noriega, A, Sly, WS, & Moore, DJ, J biol chem
255 (1980) 9608.

von Figura, D & Voss, B, Exp cell res 121 (1979) 267.

Kaplan, A, Achord, DT & Sly, WS, Proc natl acad sci US 74 (1977)

2026.

65

17.
18.

19.

20.

21.

22.

23.

24.
25.
26.

27.

28.
29.
30.
31.

66

von Figura, D & Weber, F, Biochm j 176 (1978) 943.

Tabas, I & Kornfeld, S, J biol chem 255 (1980) 6633.

Sahagian, GG, Distler, J & Jourdian, GW, Proc natl acad sci US 78
(1981) 4289.

Steiner, AW & Rome, LG, Arch biochem biophys 214 (1982) 681.

Roff, CF, Wozniak, RW, Blenis, J & Wang, JL, Exp cell res 144
(1983) 333.

Roff, CF & Wozniak, RW, The extracellular matrix (eds SP Hawkes &
JL Wang) p. 329. Academic Press, New York (1982).

Drinkwater, NR, Corner, RC, McCormick, JJ & Maher, VM, Mutat res
106 (1982) 277.

Reitman, ML, Varki, A & Kornfeld, S, J clin invest 67 (1981) 1574.
Saarni, H, & Tammi, M, Anal. biochem 81 (1977) 40.

Oohira, A, Wright, T & Bornstein, P, J biol chem 258 (1983) 2014.
Gonzalez-Noriega, A, Grubb, JH, Talkad, V & Sly, WS, J cell biol 85
(1980) 839.

Dean, RT, Jessup, W & Roberts, CR, Biochem j 217 (1984) 27.
Yanagishita, M & Hascall, VC, J biol chem 249 (1984) 10270.
Yanagishita, M & Hascall, vc, J biol Chem 260 (1985) 5445.
Neufeld, EF, Sando, GN, Garvin, AJ & Rome, HJ, J supra struct 6

(1977) 95.

Chapter 3

Turnover of Cell Surface Proteoglycans
in Cultured Fibroblasts: Comparison

between Normal and I-cells

James H. Brauker and John L. Wang

Department of Biochemistry
Michigan State University

East Lansing, MI 48824

67

 

SUMMARY

The metabolism of cell-associated proteoglycans (PG)

was studied in I-cell fibroblasts and normal (SL66) skin

5804-2 to label the

glycosaminoglycans (GAGS). Kinetic data from label-chase

fibroblasts cultured in 3

experiments and gel filtration analysis of the molecular
weight distribution of the radiolabeled GAGs yielded the
following main conclusions. a) The amount and GAG
composition (75% heparan sulfate) of the 35SO4-Z-labeled
cell surface PG of I-cells do not differ significantly from
those of normal cells. b) The rates of shedding into the
medium and internalization of cell surface PG are
unaffected in I—cells. c) I—cells retain partially
degraded GAG chains intracellularly during a chase in
unlabeled medium, reflecting their known deficiency of
lysosomal enzyme activities. d) Two classes of GAG chains
are stored in I—cells. One group of large molecules is
composed mainly (90%) of galactosaminoglycans, and makes up
2/3 of the total intracellular radioactivity. The other
group of small molecules is composed of mainly (90%) heparan
sulfate (HS) and makes up 1/3 of the total intracellular

radioactivity. e) The galactosaminoglycan GAG chains are

delivered to the degradative compartment only during the

68

69

radiolabeling period, and not during the chase, consistent
with the observation that few galactosaminoglycan PG are
found at the cell surface. f) The HS GAG chains are
internalized continuously from the cell surface during the
chase; g) The internalized HS GAG chains are acted on by a
HS endoglycosidase which reduces the size of the GAG to
about 1/6 of its original size.

It is likely that this HS endoglycosidase activity is
mediated by an enzyme that is delivered to its site of
action by a pathway which is not dependent on mannose 6-
phosphate recognition markers. Thus, it is unlikely that

it is a lysosomal enzyme.

INTRODUCTION

Cultured human fibroblasts synthesize sulfated
proteoglycans (PG) which are secreted into the medium.
tranSported to the cell surface. or tranSported directly to
degradative compartments of the cell (1). The cell surface
is rich in proteoglycans which contain heparan sulfate (HS)
glycosaminoglycan (GAG) chains (2—6). The PG may leave the
cell surface by two pathways; they may be shed into the
medium. or they may be internalized by the cells and
tranSported to degradative compartments (3-7). After
internalization of the PG. the GAG chains are acted on by
endoglycosidase and exoglycosidase enzymes to completely
degrade them to their constituent monomers (8.9). This
activity has been generally hypothesized to occur in
lysosomes. Recently. non-lysosomal protease and endo—
glycosidase activity against internalized HSPG have been
postulated to occur in rat ovarian granulosa cells (10.11).
This hypothesis is based on evidence that the protease and
endoglycosidase activities are not interUpted when cells
are treated with the lysosomotr0pic agent chloroquine.
However. this evidence is inconclusive because some
lysosomal enzymes may still be active at pH 6 (9). which is
the approximate pH of the intracellular acid compartments

of cells treated with chloroquine (12).

7O

 

71

Fibroblasts from I-cell (mucolipidosis II) patients
have a deficiency of lysosomal enzyme activity in their
lysosomes (13.14). The basic biochemical mechanism has
been characterized as an inability of I-cells to
phOSphorylate mannose residues on the oligosaccharide
chains of lysosomal enzymes. The normal mode of tranSport
of lysosomal enzymes requires that the oligosaccharide
chains of the enzyme be phOSphorylated. so that they can be
recognized by Specific receptors in the Golgi called
mannose 6-phosphate receptors (15.16). The enzymes bind to
these receptors and are tranSported directly to lysosomes
via an intracellular route (17-19). Lack of phosphorylated
mannose residues on lysosomal enzymes in I-cells results in
the enzymes being secreted into the medium. [where. at
neutral pH they do not degrade GAGs (24)]. rather than
being tranSported to lysosomes (13.14). Thus. although I-
cells synthesize precurser lysosomal enzymes. they have a
deficiency of the enzymes needed for degradation of GAGs.
Undegraded GAGs are stored in the lysosomes of these cells.

Like the genetic disorders involving the deficiency of
single enzymes (muc0polysaccharidoses) (9.20). it would be
eXpected that the GAGs which are stored in the
intracellular degradative compartment would consist
predominantly of catabolites which are substrates for the
first missing enzyme in the pathway of degradation.
Therefore. we studied the internalization and degradation

of cell surface PC in I—cells. in comparison to normal

72

cells. to determine whether some of the early steps in PG
degradation are affected by the lysosomal dysfunction. We
show evidence that the steps of internalization. separation
of GAG chains from protein cores. and heparan sulfate (HS)
endoglycosidase action are all unaffected in I-cell
fibroblasts. These data SUpport the hypothesis that the
early steps in the degradation of PG occur in non—lysosomal

compartments.

MATERIALS AND METHODS

Cgll culture and radiolabeling procedures

Fibroblast lines from I—cell (GM 22738, GM 2013A),
heterozygous I-cell (GM 2014) and Sanfilippo (GM 0312)
subjects were from the Coriell Institute for Medical
Research (Camden, NJ). SL66 cells, derived from normal
skin fibroblasts, were from Dre. J. McCormick and V. Maher
(21). All cells were maintained in Eagle's Minimal
Essential Medium (MEM) (Gibco, Grand Island, New York) with
20% fetal bovine serum (FBS) (Hazelton Research Products,
Denver, PA), 100 ng/ml streptomycin and 159 IU/ml
penicillin (Sigma, St. Louis, M0) at 35°C in humidified
air. Medium for stock SL66 cultures was changed three
times a week; these cells were passaged before the cultures
reached confluence. Medium for the other cell lines was
changed every two days and the cultures were split 1:2 or
1:4 upon reaching confluence. Confluent monolayers of
cells were removed from the dish with trypsin (400 ng/ml)
in buffer A [11.5 mM sodium—potassium phosphate (pH 7.2),
140 mM NaCl, 2.68 mM KCl, and 0.01 mM EDTA], and plated in
6 well dishes (8 cmz/well, Costar, Cambridge, MA) for 1-2
ciays before labeling. All cultures were confluent at the

'time of labeling.

73

74

Cells were radiolabeled in streptomycin-free MEM with

M9012 substituted for M9804, 20% dialyzed FBS (Gibco), 50

35

uCi/ml H2 SO4 (carrier free, ICN. Irvine, CA).

Harvest 9; cell fractions during label-chase experiments

 

The radioactivity was harvested either at the end of
the radiolabeling period (0 hour) or after "chasing" in
unlabeled medium for various times, and designated as
extracellular, pericellular, or intracellular radioactivity
(22). The extracellular compartment was the medium,
including cell secretions, which was removable by
pipetting. The pericellular material was the radioactivity
in the cell monolayer which could be removed from the cells
by trypsinization. The intracellular fraction was the
trypsin resistant radioactivity remaining with the cells
after trypsinization. To harvest these cell compartments,
the medium (extracellular fraction ) was removed and
centrifuged for 3 minutes at 1330 g to remove cellular
debris. The cell monolayers were washed several times with
MEM; the washes were discarded. Pericellular and
intracellular material were obtained by removal of the
washed monolayer with trypsin (400 ng/ml in buffer A for 20
minutes at 35°C); this removed all the radioactivity from
the dish. Cells which remained loosely attached to the
dish were dislodged by gentle pipetting. The cell
suspension was centrifuged for 3 minutes at 1330 g. The
supernatant, defined as the pericellular fraction, was

decanted. The remaining cell pellet, or intracellular

75

fraction, was dissolved in H20 and frozen at —20°C. The
intracellular material was analyzed for cell protein by the
Bradford method (23). Alternatively, for gel filtration
experiments to compare the size of pericellular and
intracellular 35SO4-labeled material, the whole cell
monolayer (sum of intracellular and pericellular), was
harvested at once, after removal of medium and washing, by
addition of 0.2% sodium dodecyl sulfate (SDS), 1 mM NaNa.
Harvest by this method does not affect the molecular weight
of the pericellular PG, whereas trypsin reduces their
molecular weight. Aliquots from extracellular,
pericellular, intracellular, or whole cell samples were
solubilized at a ratio of 1:9 (sample:scintillation
cocktail) in scintillation cocktail consisting of 7 g 2,5—
diphenyloxazole (Research Products International Corp.),
338 ml Triton X-100 and 667 ml toluene. Radioactivity was
determined on a Packard Tricarb SOOCD liquid scintillation
counter and expressed as disintegrations per minute (dpm).
Protease inhibitors were added to all of the cell fractions
to make final concentrations of lmM benzamidine HCl, 0.5 mM
EDTA, 5 mM N-ethyl maleimide, 25 mM 6-aminohexanoic acid,
and 1 mM phenyl methyl sulfonyl fluoride.

For "chase" experiments, the monolayers were washed
after removal of radiolabeling medium and incubated with
MEM containing 20% FBS, and harvest procedures were

performed at various times thereafter. Alternatively, the

cells were pretreated for 15 minutes with PBS alone, or PBS

76

containing 30 pg/ml trypsin. to remove pericellular GAGe
without removing cells from the dish. In some eXperiments
the cells were removed from the dish with trypsin (400
pg/ml in buffer A). and replated in growth medium for one
hour. The adherent cells were then washed and "chased" in
unlabeled medium. At various times thereafter. the
extracellular. pericellular. and intracellular compartments
were harvested as above. Low molecular weight (LMW) and
high molecular weight (HMW) components of medium were
quantitated by loading 200 pl aliquots on columns
containing 2 ml of DEAE cellulose (Sigma) equilibrated and
eluted with 0.05 M sodium acetate. 0.15 M NaCl. 0.5% Triton
x-loo. pH 6. Under these conditions the HMW components
bound to the gel. whereas the LMW components eluted in the
first 2 ml of buffer. The LMW components migrated on
Sephadex G—25 columns at the position of free sulfate and

therefore were considered to be degraded material (24).

Chemical and enzymatic degradation 9; GAGs

 

Samples were used directly from pericellular
fractions. whereas intracellular fractions were first
eluted on CLGB columns in 4 M guanidine HCl (see below).
The radioactivity eluting in regions B or C of the CL6B
column was pooled and precipitated in 2 volumes of cold 90%
ethanol. heated for 2 minutes in a 95°C water bath.
followed by 5 minutes on ice. The samples were then

centrifuged for 10 minutes at 12.000 X g. The supernatant

77

was removed and the precipitate was suspended in H20.
Aliquots of 0.3 ml of the pericellular and intracellular
samples were treated as follows: a) For digestion with
chondroitin ABC 1yase (C'ase . Miles. Elkhart. IN). 0.3 ml
of a solution containing 0.1 M Tris. 0.1 M Na acetate. 0105
unit/ml C'ase (pH 7.3). were added and incubated for 3
hours at 37°C; b) For nitrous acid degradations. 37.5 pl of
2.4 M NaNOZ. and 37.5 pl glacial acetic acid were added for
80 minutes at 25°C. followed by addition of 75 pl of 3 M
sodium sulfamate to stOp the reaction; c) For alkaline
degradations. 0.3 ml of 0.4 N NaOH were added for 24 hours
at 37°C. followed by addition of acetic acid to neutralize
the sample. The degradation products were analyzed by gel

filtration on Sepharose CLGB or Sephadex G-50 columns (see

below).

Gel filtration

Samples were chromatographed on columns of Sepharose
CLSB (0.7 x 110 cm) or Sephadex G-50 (0.7 x 50 cm) at a
flow rate of 5 ml/hour. The elution buffer was 0.2 M

lithium acetate. 0.2 % SDS. and 1 mM NaN pH 5. or 0.05 M

3!

Tris. 0.05 M Na 80 4M guanidine HCl. 0.5 % triton—X

2 4"

100. pH 7. Fractions of 0.8 ml were collected and analyzed

for radioactivity.

 

 

RESULTS

Egrnoveg 9f 35§Q -z-labeled components 13 cells

4

I-cells and SL66 cells were labeled with 3580‘;2 for
24 hours. After washing to remove unincorporated
radioactivity. the labeled cultures were chased in
unlabeled medium. The intracellular and pericellular
fractions were isolated and the radioactivity quantitated
(fig. 1). At the beginning of the chase period (0 hours)
the level of intracellular radioactivity in I-cells was
more than 3 fold higher than that in normal cells (fig.
la). Over a 48 hour chase period. the quantity of 3SS-
label in I-cells decreased to 63% of the initial amount.
In contrast. the normal cells rapidly lost most of the
initial radioactivity during the first 12 hours (24% of
initial value). The accumulation of [BSSIGAGs is
consistent with the observation of Schmickel et al (25).
and is likely due to dysfunctional lysosomal activity in
these cells (26).

In both cell types. radioactivity in the pericellular
compartment decreased with time (fig. lb). No difference
was observed at any time point between the two cell types._
Thus. it appears that the turnover of the GAGs in the
pericellular compartment of I-cells is not affected by the

lysosomal dysfunction.

78

 

79

35

Figure 1. Kinetics of the turnover of $04-1abeled

components in the (a) intracellular and (b) pericellular

compartments of SL66 (0) and I (0) cells. Cells were

-2
4

chased in unlabeled culture medium. The intracellular and

labeled with 35$0 (50 uCi/m1) for 24 hours, washed and
pericellular fraction were isolated as described in
Materials and Methods and their radioactivity quantitated.
The data represent the averages of duplicate determinations

(error bars indicate standard deviations where larger than

symbols).

80

3.52.: 92:.

w? VN

 

 

 

004m 0:0
zool— I
3.2.8::

_

 

0' —

 

@015 0.10
:2. 1H I

33:38::

I

(..oixwdp) iii/410001903 [39g]

 

 

81

Molecular weight distribution 9f 35§Q4-labeled components

 

 

$3 the cell monolayer

Radioactively labeled cells were removed from the
culture dish with trypsin (400 pg/ml. 15 minutes. 37 C) and
the trypsin resistant pellet (intracellular) fraction was
collected. Parallel cultures were extracted with 0.2% SDS
which removed all of the radioactivity from the dish. This
material was termed the whole cell monolayer. Gel
filtration on columns of Sepharose CLGB was carried out
under dissociative conditions for the intracellular
fraction and the whole cell monolayer. A comparison of the
profiles was made between material derived from various
times during the label-chase eXperiment for normal SL66
cells (fig. 2 a-d) and for I-cells (fig. 2 e-h). The
intracellular GAGS of both SL66 and I-cells have
predominant peaks of radioactivity centered around 0.42 Kd
on Sepharose CL6B (see. for example. region B of fig. 2a
and 2e). In addition. I-cells develOp a second "shoulder"
of radioactivity during the chase period. centered around
Kd - 0.72 (see. for example. region C of fig. 2 e-h).

At any given time. the radioactivity of the PCs of the
pericellular fraction is represented by the difference
between the profiles derived from the whole cell monolayer
(solid line in fig. 2) and the intracellular fraction
(dotted line in fig. 2). The pericellular material

chromatographs near the void volume of a CL6B column. This

82

Figure 2. Gel filtration of 35SO4-z-labeled
components derived from the whole cell monolayer (SDS
extract. --) and the intracellular fraction (trypsin
resistant cell pellet. --—) of SL66 cells (a - d) and I-
cells (e - h) during a label (50 pCi/ml. 24 hour) and chase
eXperiment. Radioactivity correSponding to 10 pg of cell
protein (using the intracellular fractions for protein
determinations) was chromatographed on a column of
Sepharose CLGB (0.7 X 110 cm) equilibrated with 0.2 M

lithium acetate. 0.2% SDS. 1 mM NaN pH 5. The horizontal

3!
bars designated A.B. and C highlight regions of interest

discussed in the text.

83

 

0.. 3.

«-

 

 

30¢

 

 

 

....VN

 

 

:vN

 

“.

 

   

 

 

 

 

 

 

 

a a N. .1 N.
111 8.3.8935 .11... 3.2.82.5
.. I :8 22.3 I :8 22.3 1
P p n L P p u L
0 . 0 . m u a S o a U u m u ( z o
:3 .. H mwuw

1,. 1 x wdp) 4114110001908 [8“]

84

material is in the form of proteoglycan, because its
molecular weight is reduced after treatment with 0.2 M NaOH
(see below). Thus, the bulk of the radioactivity of the
pericellular fraction (region A, fig.2) has a higher
molecular weight than that of the intracellular fraction.
Similar results were obtained for both SL66 and I—cells.
For the purpose of clarity of subsequent description and
discussion, we define the regions A - C (fig. 2) of the
chromatographic profile as follows: (A) high molecular
weight material corresponding to the pericellular PG of the
cell; (B) material with Kd 0.42 corresponding to the
majority of the intracellular GAGs; (C) low molecular
weight material (Kd = 0.72) corresponding to a class of
GAGs that is accumulated in I—cells but not in normal
fibroblasts.

The results of both quantitative (fig. 1) and
qualitative (fig. 2) experiments indicate that cell surface
PG of I-cells turnover in a fashion comparable to that in
normal fibroblasts. In addition, the gel filtration
profiles reveal that, if I—cells internalize cell surface
proteoglycans, these molecules are rapidly reduced in size,
since the peak corresponding to cell surface PG does not
seem to accumulate (fig. 2 e—h) as "storage" material in
the intracellular compartment (1,9). Thus, it is likely
that the first steps of degradation of cell surface PG by

I-cells are not impaired.

85

Characterization g: the cell surface gg

3550 "2 for 24

I-cell fibroblasts were labeled with 4

hours. The radiolabeled cells were washed and then
trypsinized. The radioactive material released by trypsin,
representing the PG of the cell surface, was subjected to
gel filtration on a column of Sepharose CLGB. This yielded
a radioactive profile (fig. 3a) consisting mostly of high

molecular weight material, with a broad shoulder extending

 

into the more included regions of the column. The bulk of
this material migrated at a position on the column (region
A) consistent with our previous conclusion that the
pericellular PGs are represented by the difference between
the profiles derived from the whole cell layer (solid line,
fig. 2a) and the intracellular fraction (dotted line, fig.
2a). However, trypsin treatment appears to reduce the
molecular weight of some of the pericellular PG, resulting
in the "shoulder" of radioactivity in fig. 3a.

When the pericellular radioactive material was
subjected to treatment with 0.2 M NaOH, there was a
reduction in the molecular weight (compare fig. 3a and 3b).
This suggests that the majority of the pericellular label
was in the form of P65 which lose the core proteins upon/6-
elimination. Approximately 25% of the pericellular PG was
sensitive to C'ase (fig. 3c), suggesting that these are
galactosaminoglycans. The majority of the pericellular PG

was sensitive to nitrous acid (fig. 3d). Therefore, HSPG

86

35304‘2—labeled

Figure 3. Gel filtration of
components derived from the pericellular fraction (trypsin
released) of I-cells labeled for 24 hours with 50 uCi/ml of
35$0472. (a) untreated pericellular fraction; (b)
pericellular fraction treated with 0.2 N NaOH for 24 hours
at 37°C; (c) pericellular fraction digested for 3 hours at
37°C with 0.025 unit/ml C'ase; and (d) pericellular
fraction treated with 0.24 M nitrous acid for 80 minutes at

25°C.

.b

N

[358] Radioactivity (dpmx 103)
N

g

87

 

- 0 Untreated

 

 

 

 

F' c C’ase

 

 

 

d Nitrous Acid

 

 

 

 

 

88

represents the bulk of the 35

SO4-labeled GAGS of the cell
surface.

The results described above were from I-cells.
Essentially identical results were obtained when the same

analyses were carried out on the pericellular material of

SL66 cells.

Effect 9: trypsin 9n the turnover 9: pericellular gg

In order to examine the relationship between the cell
surface pool of PG and the intracellular (degradative)
compartment, the turnover of cell associated GAGS was

analyzed with and without pretreatment of the cells with

3580 -2

4 for 24 hours,

trypsin. The cells were labeled with
and then treated with trypsin (30 pg/ml) in PBS, or with
PBS alone for 15 minutes. Under these conditions of
trypsin treatment, the cells were beginning to round up but
they did not become detached from the dish. The cell
layers were then washed and chased in unlabeled growth
medium. The trypsinized cells reestablished flattened
morphology within 1 hour.

Treatment with trypsin removed about 80% of the
pericellular PG from both I—cells and normal cells (fig.
4b). This conclusion is derived by comparing the levels of
radioactivity at the initial time point (0 hour) in fig. 4b
(compare, for example, open circles and closed circles for
SL66 cells). Some radioactivity was lost from the
intracellular pool of both cell types when treated with

trypsin (note lower radioactivity at each time point in

89

35SO4-z—labeled

Figure 4. Kinetics of the turnover of
components derived from the (a) intracellular and (b)
pericellular compartments of SL66 and I—cells. Cells were
labeled with 35804—2 (50 pCi/ml. 24 hours). Parallel
cultures were treated with trypsin (30 pg/ml in PBS. 15
minutes) or incubated in PBS as controls. The wells were
then chased in unlabeled culture medium and the
radioactivity in the intracellular and pericellular
fractions was quantitated. The data represent the averages

of duplicate determinations (error bars indicate standard

deviations where larger than symbols).

90

 

 

   

 

 

 

 

 

 

 

 

F
O
"' 0 Lee"
3: Intracellular Without Trypsin H
.5 With Trypsin H
2 SL66
8 Without Trypsin 0"0
:1 With Trypsin O—O
f, Pericellular b
0
0'
it — 3
E» 3.
\\\
3:): 2" \\\\ 1 2
>
“:3 \CA.‘
8 b‘\“‘~
.. a g
«o | \ \\ _‘ |
o \ ‘0
05 \
\
l-_l
(D ——_
s :9 R 4-
IO 20 IO 20

Time (Hours)

91

fig. 4a). This was thought to be due to a small loss of
cells during the washing procedures following the trypsin
treatment. No difference was observed in the Sepharose
CLSB gel filtration profiles of trypsinized and
untrypsinized cells at the initial time point (data not
shown). When the cells were chased in unlabeled medium.
the radioactivity disappeared from the pericellular pool of
untrypsinized I-cells and normal cells at approximately the
same rate (fig. 4b). as described before. (fig. lb). There

was little loss of radioactivity from the pericellular pool

 

of trypsinized cells. compared to that of untrypsinized
cells.

The rate of loss of radioactivity from the
intracellular pool of trypsinized normal cells paralleled
that of untrypsinized cells (fig. 4a). suggesting that
internalized cell surface molecules do not affect the
steady state levels of intracellular PCs in these normal
cells. In contrast. the rate of disappearance of
radioactivity from the intracellular pool of untrypsinized
I-cells paralelled that of trypsinized I-cells during the
first 12 hours . but then diverged. resulting in higher
intracellular levels of radioactivity after 22 hours (fig.
4a). We interpreted this to be due to the storage of
internalized cell surface PC in the untrypsinized cells.
This hypothesis is further SUpported by analyzing the rates
of accumulation of low molecular weight degradation

products (presumed to be free sulfate) in the medium. which

92

is an estimate of the overall degradation rate of the 355—

labeled PGs (fig. 5).

The untrypsinized SL66 cells released LMW degradation
products at a higher rate than trypsinized normal cells
during the first 12 hours. After 12 hours. little
additional degradation occurred in trypsinized cells.
while the untrypsinized cells continued to release degraded
material (fig. 5a). In contrast. trypsinization had little
effect on the rate of production of degraded (LMW) material
in I-cell cultures (fig. 5a). The higher levels of LMW
radioactive sulfate from untrypsinized normal cells
compared to trypsinized normal cells probably is due to the
internalization and degradation of pericellular PG
initially at the cell surface of the untrypsinized cells.
The corresponding material had been removed from the
trypsinized cells. In contrast. I-cells internalized PG
from the pericellular pool (note the higher radioactivity
in the intracellular pool of untrypsinized I-cells at 22
hours. fig. 4a). but this additional substrate did not
increase the rate of degradation. because in I-cells. the
rate of degradation is limited by the intracellular

lysosomal enzyme concentration.

Shedding and Uptake 9f pericellular g9

Analysis of the HMW material in the medium of
trypsinized and untrypsinized cells revealed that both I-
cells and normal cells released substantial amounts of

macromolecules into the medium during the chase (fig. 5b).

 

93

Figure 5. Kinetics of the appearance of (a) LMW
radioactivity and (b) HMW radioactivity in the medium of
SL66 and I-cells labeled with 35504‘2 (so pCi/ml. 24
hours). Parallel cultures were treated with trypsin (30
pg/ml. 15 minutes) or incubated in PBS as controls. The
cells were than chased in unlabeled culture medium and the
radioactive components of the medium were fractionated on a
column of DEAE cellulose (2ml total volume) equilibrated
with 0.05 M sodium acetate. 0.5% Triton x-loo. pH 6. LMW
components represent material that did not bind to DEAE-
cellulose (free sulfate) whereas HMW components represent

material bound to the column. as defined in Materials and

Methods.

94

(Pm x uieimd "as bn/iudp) Kighiioooipou [9“]

42 iii
0 O
I I

 

23
00
H

2!)

 

ypsln

"3

Without Trypsin
With Trypsin
Time (Hours)

Without Trypsin
With Tr

I- call
SL66

 

 

2C)

"3

 

 

 

N -
(”’0' x ugeimd ”as bfi/wdp) KiiAgioooipoa [$92]

 

95

Pretreatment of the cells with trypsin resulted in a
reduction of the total radioactivity in the medium compared
to untrypsinized cells. After 22 hours. I-cells
accumulated higher amounts of macromolecules~in the medium
than did normal cells (fig. 5b). This was true of both
untrypsinized and trypsinized cells. Gel filtration
analysis of the radioactivity in the medium revealed a
possible explanation for this difference. The bulk of the
radioactivity released into the medium by untrypsinized
normal cells (solid lines. fig. 6 a a b) chromatographed at
the total volume of the column (large peak in region C).
indicating that it consists of degraded material (24). The
remaining radioactivity was of the same size class as
pericellular GAGS (compare. for example. fig 6a with fig
2a). suggesting they were released from the cell surface.
While trypsinized cells also yielded a high level of low
molecular radioactivity (dotted lines. fig. 6 a a b) the
macromolecules in the medium of trypsinized cells were
predominantely of the size class of intracellular molecules
(region B). suggesting that they were leaked or excreted
from the cells.

Analysis of untrypsinized I-cell medium revealed
similar data to that of normal cells. except that a broader
shoulder of radioactivity was seen in the intermediate
molecular weight range (region B. fig. 6 c a d). Analysis
of the column profiles of the medium from trypsinized cells

yielded higher amounts of radioactivity in region B of I—

 

 

96

35SO -2-labeled

Figure 6. Gel filtration of 4

components in the medium of SL66 cells (a and b) and I-
cells (c and d). Cells were labeled with 3580“"2 (50
pCi/ml. 24 hours). Parallel cultures were treated with
trypsin (30 pg/ml. 15 minutes) or incubated in PBS as
controls. The cells were then chased in unlabeled culture
medium and the radioactive components of the medium were
chromatographed on a column of Sepharose CLSB (0.7 X 110
cm) equilibrated with 0.2 M lithium acetate. 0.2% SDS. 1 mM
sodium azide. pH 5. (-——0 material derived from cultures
without trypsin treatment; (---) material derived from
cultures with trypsin treatment. ,The horizontal bars

designated A.B. and C highlight regions discussed in the

text.

97

I-ceH

SL66

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A- II o.
C
.... d
B K
II . O
r. A fl. W In” N .... 0.
u ll- ) o
.m h
2
m. 2
_ q q
41
C .... m
d
B K
I‘ 00
AI?- IT 0

10 hour

 

 

22 hour

 

 

.99 x 5%. £28853. we...

98

cells than in the correSponding region of normal cells.
This observation supports the hypothesis that the region B
molecules in the medium came from the degradative
compartment. However. it is possible that the light
trypsin concentration used in these experiments caused
partial degradation of pericellular PG. which were then
shed into the medium during the chase. Evidence against
this possibility came from experiments in which cells were
labeled. removed from the dish with trypsin. and replated
(fig 7). The cells were trypsinized under the same
conditions used to harvest cell compartments (400 pg/ml).
which completely removed pericellular PG. Analysis of the
Sepharose CLGB column profiles of the medium after a 24
hour chase revealed that the bulk of the released
radioactivity migrates as degraded material (solid lines.
fig. 7) (note. the same data are plotted at two scales to
emphasize degraded radioactivity (closed circles) and
macromolecules (open circles)]. The remaining
radioactivity elutes in region B. and correSponds in size
to the material that was inside the cells at the beginning
of the chase period (open circles. fig. 7). About 25% of
the radioactivity in I—cell medium migrates in region B
(fig. 7 b). compared to only 5% in SL66 medium (fig. 7a).
In normal cells. this represents a minor amount. compared
to the amount of degraded material. as pointed out by
Fratantoni and Neufeld (1). However. it represents a

substantial fraction of the total macromolecules in the

 

99

35SO -2-labeled

Figure 7. Gel filtration of 4

components in the medium of (a) SL66 and (b) I-cells.

Cells were labeled with 35504—2 (50 pCi/ml. 24 hours). The
cells were then removed from the original dish by trypsin
(400 pg/ml in buffer A). They were replated for 1 hour in
culture medium and then chased for another 24 hours in
unlabeled culture medium. The medium was then
chromatographed on a column of Sepharose CLGB (0.7 x 100
cm) equilibrated with 0.2 M lithium acetate. 0.2% SDS. 1 mM
sodium azide. pH 5. The same radioactivity profile is
plotted on two different scales to emphasize degraded

radioactive material (0——4) and macromolecules @——«».

100

(H.,-Ol X uldp) KiiAiioooipoa [592]
I) q. '0 N

 

 

' l

 

 

 

 

 

 

 

 

 

 

 

.0
—————————— —O-
—____——_—————-0
,o———--—-—-o-—--0'
53-»..-
.0—‘—_. 0 ~
"0'
0
I
H
l l I I
co 1 l
.- 0 °
+
. -..O
+ -
——_—————_———-—————— .
'f.
as:
.c:
.6
o' .0
.2 . x
:0
0..
v"
5”.
0°. to.
o .1
‘° ~o. “ o
co .;;
-’ Q
U)

 

... _
(0.0 $0. x uidp) 4114119001903 [See]

101

medium after a chase period without prior trypsinization
(26% in normal cells and 40% in I-cells). These data
indicate that quantitation of total macromolecules in the
medium is not sufficient for an accurate measure of
"shedding" from the cell surface. In light of these data
"shedding" was defined as the difference in the quantity of
macromolecules in the medium of untreated cells and those
pretreated with 30 pg/ml trypsin. Using this method.
shedding from I-cells did not differ significantly from
that in normal cells (fig. 8). The rate of Uptake of
pericellular PG could then be determined as the
radioactivity lost from the pericellular pool but not shed
(fig. 8). Uptake was not altered in I-cells compared to
normal cells. and was about two fold higher than shedding

in both cell types. (fig. 8).

Characterization Q; intracellular GAGs

We had observed in previous label—chase experiments
that the intracellular GAGs of both SL66 and I-cells have.
on gel filtration on Sepharose CLGB. predominant peaks
around Kd 0.42 (region B. fig. 2a and 2e). In addition. I-
cells develOp a shoulder of radioactivity during the chase
period-(region C. fig. 2.e-h). A hint toward the chemical
nature of the components in regions B and C of the
Sepharose CL6B column was revealed by comparing the gel
filtration profiles of the intracellular PG (derived from
cells labeled for 24 hours and then chased for 24 hours)

with the corresponding material derived from Sanfilippo

102

Figure 8. Quantitation of internalized (0,0) and shed

(4.4) 35304.2—labeld macromolecules from the pericellular
fraction of SL66 (—-—) and I—cells (——-). Cells were
35 —2

labeled with 804 (50 pCi/ml. 24 hours). Parallel
cultures were treated with trypsin (30pg/ml in PBS. 15
minutes) or incubated with PBS as controls. The cells were
then chased in unlabeled culture medium and the radioactive
components of the medium were fractionated on a column of
DEAE cellulose to quantitate the HMW material (as described
in fig.5) Shed material (A,A) was derived by subtracting
the amount of HMW material in the medium of trypsinized
cultures from that in untrypsinized cultures. Uptake was
derived by subtracting the amount of shed material from the
total amount of radioactivity lost from the pericellular

pool at each time point.

[35 S] Radioactivity (dpm /p.g cell protein x l0'3)

 

I-cell Uptake H
Shedding H

SL66 Uptake O-~O .6
Shedding A—-—-A //

 

 

 

 

Time (Hours)

104

cells (fig. 9). The intracellular GAGS of Sanfilippo cells
yielded a predominant peak in region C of the Sepharose
CL6B column. Inasmuch as a previous study (27) had shown
that Sanfilippo cells accumulate only HS GAGs. due to a
defect in an enzyme in the degradation pathway of HS. these
results suggest that the region C molecules stored in the
intracellular pool of I-cells may be HS. We also analyzed
the intracellular GAGs derived from cells of a subject
characterized as an "I—cell heterozygote" (GM 2014). The
Sepharose CL6B profile from these cells yielded a peak in
region B of the column. The level of radioactivity was
much lower than that of I-cells. Moreover. this profile
was similar to that of normal (SL66) cells (compare with
the dotted lines in fig. 2 a—c).

Additional information concerning the chemical nature
of the intracellular material chromatographing in regions B
and C of the Sepharose CLGB column was derived from studies
on their sensitivities to chemical treatments. The
intracellular material in region B (derived from I-cells
labeled for 24 hours and chased for 24 hours) was sensitive
to C'ase but resistant to nitrous acid (fig. 10 a-c). This
suggests that the material consisted of
galactosaminoglycans. In contrast. the bulk of the
intracellular material from region C was resistant to C'ase
but was reduced in molecular weight by treatment with
nitrous acid. COUpled with the results derived from

Sanfilippo cells. these data indicate that the material in

105

3S804.2-labeled

components of the intracellular fraction of I-cells (*-0.

Figure 9. Gel filtration of

Sanfilippo cells Urwo). and heterozygous I-cells (---).
Cells were labeled with 35304-2 (so pCi/ml. 24 hours) and
chased in unlabeled culture medium for 24 hours. The
intracellular fraction was chromotographed on a column of
Sepharose CL6B (0.7 x 110 cm) equilibrated with 0.2 M

lithium acetate. 0.2% SDS. 1 mM sodium azide. pH 5.

106

 

 

C
HI-cell

O--O Sanfilippo

— — -- heteroz

 

 

-

.70. x Eat.

  

 
 

figoue

I-ce

.
3

228862 We”.

 

 
    

I

 

 

do

107

35$04-2—labeled

Figure 10. Gel filtration of
components corresponding to region B and C (figure 9) of
the intracellular fraction of I—cells after chemical
treatments. Cells were labeled with 35804—2 (50 uCi/m1, 24
hours) and chased for 24 hours. The intracellular
radioactivity was fractionated on columns of Sepharose CLGB
as described in legend to fig. 9. The pooled fractions
corresponding to regions B and C were chromatographed
without treatment (a, d), after C'ase digestion (0.025
unit/ml, 3 hours, 37°C) (b, e,), and after nitrous acid
treatment (0.24 M, 80 minutes, 25°C) on columns of

Sepharose G—5O (0.7 X 50 cm) equilibrated with 0.2 M

lithium acetate, 0.2% SDS, 1 mM sodium azide, pH 5.

108

Region 8 Region C

 

at
1

Control

#1 at
I 1
‘5.

 

 

Radioactivity (dpm x 10‘2)

53] Radioactivity (dpm x 10.3)
at m
l
U'

I:

 

Control

 

 

 

 

 

  

Nitrous
Acid

 

lo 20 30 to
Fraction #

20

 

f

 

30

[355]

109

region B of the column profile consists mainly of
galactosaminoglycans and the material in region C of the

column profile is predominantly HS.

Turnover g; intracellular GAGs

The intracellular 35SO4—labeled GAGs from cultures
of I-cells labeled for 24 hours and chased in unlabeled
medium. was harvested and analyzed on Sepharose CLGB
columns. The radioactivity migrating in regions B and C
was quantitated at various times (fig. 11).. To determine
the impact of the pericellular PG on the size of these
pools. the cultures were trypsinized and replated before
the chase (fig. llb) or were left untreated (fig. 11a).
The radioactivity disappeared from region B to around 60%
of the initial (0 hour) value during a 48 hour chase
period. and this disappearance of radioactivity was
unaffected by pretreatment of the cells with trypsin
(compare Open circles in fig. 11a and llb). In contrast.
the rate of loss of radioactivity from region C was slower
in untreated cultures compared to cultures pretreated with
trypsin (compare closed circles in fig. 11a and b). In the
trypsinized cultures. region C diminished to less than 70%
of the initial value. whereas. in the untrypsinized
culture. the value at 48 hours was greater than 85% of the
initial amount (fig. lla). Similar data were obtained when
cultures were treated with lower concentrations of trypsin

(30 pg/ml). so that they were not removed from the dish

(data not shown). These data suggest that the cell surface

110

Figure 11. Kinetics of the turnover of 3550472—
labeled components corresponding to regions B and C (fig.
9) of the intracellular fraction of I—cells. Cells were
labeled with 35804—2 (50 uCi/ml, 24 hours). Parallel
cultures were trypsinized (400 ng/ml in buffer A) or left
untreated as controls. The cells were then chased in
unlabeled culture medium. The intracellular fraction was
fractionated on columns of Sepharose CLSB as described in
legends to fig. 9. The radioactivity migrating in the
fractions corresponding to region B and C was quantitated.

The data represents the averages of duplicate

determinations (error bars indicate standard deviations).

111

.23.... 9......

 

 

we ...N we cw
_ . _ _ .
o 8.3.. I 0 833. I
i m 532 0.10 m 532 0.1.0
bonsai... “.8022:
O//
/
I /
/
/
I // 1
/
/ b H /
/ H /
i / /
/ /
/ /
a x
. . . .

 

 

\

0v

Ow

ow

 

00.

 

“biuug ,o ox.) Kiihiiooowos [8“]

112

HSPG were internalized and that the GAO chains. which
originally migrated at 0.42 Kd on Sepharose CLSB columns.
were degraded to fragments migrating in region C of the
column. Apparently. few galactosaminoglycans were
internalized from the cell surface because pretreatment of
the cells with trypsin does not alter the rate of turnover
of the intracellular galactosaminoglycan (region B) pool.
These data support the hypothesis that an HS degrading
endoglycosidase activity is present in I-cells. and this
activity is unaffected in the cells that lack functional

lysosomes.

DISCUSSION

The data reported here describe some characteristics
of the PG of the cell surface and the degradative
intermediates found in the interior of human fibroblasts.
Comparison of the degradative intermediates of I-cells and
normal cells illuminates our understanding of the early
steps of degradation of cell surface PG, since those steps
of PG degradation which are unaffected in I—cells are
presumably mediated by non—lysosomal enzymes. Five key
conclusions have been reached: a) The rates of shedding
and internalization, and the composition of cell surface PG
are not affected by the I-cell defect; b) The separation
of GAG chains from protein cores is also unaffected; c)
The endoglycosidic fragmentation of HS GAG chains is not
affected by the I—cell defect; d) Internalization from the
cell surface is a major pathway for delivery of HS GAGs to
lysosomes; and e) The majority of intracellular storage
GAGs are galactosaminoglycans. In contrast to HS GAGs,
internalization from the cell surface pool is not a major
pathway of entry for galactosaminoglycans into the
degradative compartment.

In addition to the major conclusions concerning the

degradation of GAGs in fibroblasts, we show evidence that a

113

 

 

114

substantial fraction (25% in normal cells and 40% in I-

cells) of the 35

SO4-labeled macromolecules appearing in the
medium during a chase (after a 24 hour labeling period) are
derived from the intracellular (degradative) pool. This
conclusion was derived by comparing the molecular weight
distribution of macromolecules in the medium of trypsinized
and untrysinized cells (figs.6 and 7). It is important to
note that not all of the macromolecules in the medium of
cultures after label-chase arrive by "shedding". An early
report by Fratantoni and Neufeld (1), showed that 5% of the
35804—labeled material liberated by normal fibroblasts
(cells were labeled for 24 hours, replated, and chased)
were macromolecules. The other 95%, under these
conditions, consisted of degraded (free sulfate) molecules.
We concur with these findings. However, our data also
indicate that, under conditions where cells are chased
without prior trypsinization, intracellular GAGs are also
released without being degraded. This material makes up a
small part of the total extracellular radioactivity, but
makes up a relatively large part (>25%) of the
macromolecules in the medium. Since these molecules are
free GAG chains, they may need to be accounted for in
interpreting the processes involved in shedding of PG from
the cell surface (i.e. are GAG chains or whole PG shed into
the medium from the cell surface?), A recent report by

Bienkowski and Conrad (28) showed that free GAG chains were

released from the intracellular compartment of a rat

 

 

 

115

hepatocyte cell line during a chase after a 24 hour label

3580 -2.

with 4

The presence of undegraded intracellular GAGs in the
chase medium raises the question whether lysosomes fuse
with the cell surface and release their contents into the
culture medium. Alternatively, the GAG chains may have
been liberated through cell death. Neither of these
possibilities have been ruled out by our present
experiments.

When the amounts of intracellular GAG chains released
into the medium were taken into account, no difference was
observed in rates of shedding of PG from the cell surface
of I-cells when compared to normal cells (fig. 8). It has
been previously suggested by others (4,11) that the bulk of

35

the 80 —labeled material released into the medium during

4
a chase is by a mechanism involving proteolytic cleavage of
the core protein of cell surface PG. If this is indeed the
mechanism occurring in the fibroblasts, it is unlikely that
lysosomal enzymes are involved, since the process is
unaffected in I-cells.

An alternate fate of cell surface PG is
internalization. Our data indicate approximately 60% of
the total PG lost from the cell surface during a label-
chase experiment are internalized. Although electron
microscopic studies suggest that vesicular traffic in I—

cells may be disrupted (29), it does not appear that

traffic of cell surface PG to the cell interior is affected

116

inasmuch as no significant difference was observed in the
rate of internalization by I-cells compared to normal
cells. Analysis by gel filtration on Sepharose CL6B
reveals that. in I-cells. none of the stored GAGs have the
size characteristics of cell surface PG. In fact. none of
the intracellular GAGS are associated with protein.
whereas. virtually all of the cell surface GAGS are in the
form of PG. These data suggest that the GAG chains are
separated from core proteins. or that core proteins are
degraded to such an extent that free GAG chains are
produced. Furthermore. this process is unaffected in I-
cells. suggesting it is a non—lysosomal enzyme activity.
Analysis of intracellular (storage) GAGs from I-cells
on Sepharose CLGB columns reveals two major peaks of
radioactivity. one large size class (region B. fig. 9)
making Up about 2/3 of the intracellular radioactivity. and
one small size class (region C. fig. 9). The region B
molecules are galactosaminoglycans and the region C
molecules are HS. Kinetic evidence (fig. 11) shows that
region C molecules (HS) enter the intracellular pool from
the cell surface. whereas. region B molecules
(galactosaminoglycans) are not derived from the cell
surface. The size class. on Sepharose CLGB columns. of the
HS GAG chains of cell surface PG (analyzed after alkaline
fi-elimination of the PG) was the same as that of the region
B galactosaminoglycans stored inside the I-cell. However.

after internalization. HS GAGs could be recovered only as

117

region C sized molecules. These data indicate that the HS
GAG chains from the cell surface are reduced in size after
internalization by a HS endoglycosidase. This observation
is SUpported by analysis of Sanfilippo cells. which have a
deficiency of a single lysosomal exoenzyme reSponsible for
degradation of HS. and therefore store only HS GAGs in
their lysosomes (8.27). The storage molecules of these
cells migrate in region C of Sepharose CLGB columns.
indicating the presence of a HS endoglycosidase activity
which must have reduced the molecular weight of the GAGs
(fig. 9). Hurler cells. which have a deficiency of a
single lysosomal enzyme involved in the degradation of both
galactosaminoglycans and HS. have two size classes of
storage molecules identical to I-cells (Cowles. E.A..
Brauker. J.H.. Anderson. R.L.. and Wang. J.L.. unpublished
data). Because the HS endoglycosidase activity is not
affected by the I-cell defect. it is likely that it is a
non-lysosomal enzyme activity.

Yanagishita and Hascall (10). studied the degradation
of cell surface PG in rat ovarian granulosa cells. Their
data indicate that-there is a HS endoglycosidase activity
which is not inhibited by the lysosomotrOpic amine
chloroquine. Based on these observations it was suggested
that the HS endoglycosidase was a non-lysosomal enzyme.
Similar results were obtained by us. treating normal

fibroblasts with NH4C1 (Brauker. J.H.. Laing. J.C.. and

 

118

Wang. J.L.. chapter 4. this thesis. manuscript in
preparation).

HS endoglycosidase activity has been reported in liver
cells (30). platelets (31—33). activated T—lymphocytes
(34). and placenta (35.36). The pH Optimum for the enzyme
is around 6 (30.31). However. it has generally been
described as a lysosomal enzyme. At least two enzymes
(acid phosphatase and B—glucosidase) are present in I-cell
fibroblasts in normal quantities (37.38). It is possible
that the activity observed in our experiments is mediated
by a lysosomal enzyme which can be tranSported to lysosomes
by some mechanism other than the mannose 6-phosphate
receptor transport pathway. It also can be argued. that
the HS endoglycosidase is simply a lysosomal enzyme with a
high pH Optimum. and therefore was not affected by
chloroquine or NH4C1 treatments. Our present data. showing
unimpaired HS endoglycosidase activity in I-cells. argues
against this possibility. The enzyme would have to have
two characteristics unlike other lysosomal enzymes (i.e..
high pH Optimum and an alternate tranSport mechanism). In
addition. if the enzyme is a lysosomal enzyme having a high
pH Optimum. then it would not be exPected to carry out the
HS degradation efficiently in the low pH mileu Of the
lysosome even if it were tranSported to the lysosome of I—
cells. The eXpected result would be that the amount Of
intracellular HS (migrating in region B) derived from I—

cells and untreated normal cells would exceed the

119

corresponding amount in NH4C1 treated normal cells. This
was not the case (data not shown).

In summary. these data show that fibroblasts
internalize HSPG from their cell surface. separate the GAG
chains from core protein by an unknown mechanism. and
degrade them to smaller fragments via an HS
endoglycosidase. We hypothesize that. because these

activities are unaffected by the I—cell defect. they occur

in compartments Of the cell other than lysosomes.

 

10.

11.

12.

13.

14.

REFERENCES

Fratantoni. J.C.. Hall. c.w.. and Neufeld. E.F.. (1968)
Proc. Natl. Acad. Sci. QQ. 699 706.

Hook. M.. Kjellen. L.. Johansson. 5.. and Robinson. J.
(1984) Ann. Rev. Biochem. 22. 847—869.

Kresse. H.. TeklOf. w.. von Figura. K.. and Buddecke. E.
(1975) HOppe-Seyler's z. Physiol. Chem. 22g. 943—952.

Johansson. 8.. Hedman. K.. Kjellen. L.. Christner. J..
Vaheri. A.. and Hook. M. (1985) Biochem. J. 222.
161-168.

Glimelius. B.. Westermark. B.. and Wasteson. A. (1977)
EXp. Cell Res. ng. 23—30.

Vogel. K.G.. and Peterson. D.W. (1981) J. Biol. Chem.
2§§. 13235 13242.

Cowles. S.A., Brauker. J.H.. and Anderson. R.L. (1986)
Exp. Cell Res. (In press).

Sly. W-S- (1980) In EEEEEQLIE EQDEEQl EDS Qlfifiéfifl
(Brady. P.K.. and Rosenberg. L.E. ed.) W.B. Saunders.
Philadelphia. pp.545—58l.

von Figura. K.. and Klein. U. (1981) In Eyggggmgg Egg
Lysosomgl Storage Diseases (Dingle. J.T.. Dean. R.T..
ed.) Elsivier. New York. pp. 229-248.

 

Yanagishita. H.. and Hascall. V.C. (1984) J. Biol.
Chem. 2§g. 10270-10283.

Yanagishita. M. (1985) J. Biol. Chem. 2gp 11075-11082.

Dean. R.T.. Jessup. w.. and Roberts. C.R. (1984)
Biochem. J. 222. 27—40.

Hickman. 8.. and Neufeld. E.F. (1972) Biochem.
BiOphys. Res. Comm. 12. 992—999.

Miller. A.L.. Kress. B.C.. Lewis. L.. Stein. R.. and
Kinnon. C. (1980) Biochem. J. 2§§.97l-975.

120

I¥i .

 

 

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

121

Sahagian. G.G.. Distler. J.. and Jourdian. G.W. (1981)
Proc. Natl. Acad. Sci. lg. 4289.

Steiner. A.W.. and Rome. L.G. (1982) Arch. Biochem.
Biophys. 221. 681-687.

von Figura. K.. and Weber. E. (1978) Biochem. J. lzg.
943-956.

Neufeld. E.F.. Sando. G.N.. Garvin. A.J.. and Rome.

Invest. g3. 595-600.

McKusick. V.A.. and Neufeld. E.F. (1983) In Egg
metabolic basis 92 inherited disease (Stanbury. J.B..
Wyngaarden. J.B.. Fredrickson. D.S.. Goldstein. J.L..
and Brown. M.S.. ed.) McGraw—Hill. New York. p.751-776.

 

Drinkwater. M.R.. Corner. R.C.. McCormick. J.J.. and
Maher. V.M. (1982) Mutat. Res. igg. 277.

ROff. C.F.. Wozniak. R.W.. Blenis. J.. and Wang. J.L.
(1983) EXp. Cell Res. 213. 333—344.

Bradford. M.M. (1976) Anal. Biochem. 12. 248-254.

Brauker. J.H.. ROff.. C.F.. and Wang. J.L. (1986) Exp.
Cell Res. lgg. 115—126.

Schmickel. R.D.. Distler. J.J.. and Jourdian. G.W.

Neufeld. E.F.. and McKusick. V.A. (1983) In 332
Metabolic Basis 92 Inherited Disease (Stanbury. J.B..
Wyngaarden. J.B.. Fredrickson. D.S.. Goldstein. J.L..
and Brown. M.S.. ed.) McGraw-Hill. New York. pp
778-787.

Klein. 0.. Kresse. H.. and vonFigura. K. (1976)
Biochem. BiOphys. Res. Comm. g3. 158-166.

Bienkowski. M.J.. and Conrad. E. (1984) J. Biol. Chem.
223. 12989—12996.

Brown. W.G.. and Farquhar. M.G. (1984) Proc. Natl.
Acad. SCie U.S.A. _8_}_’ 5135-51390

Hook. M.. Wasteson. A.. and Oldberg. A. (1975)
Biochem. BiOphys. Res. Comm. g1. 1422-1428.

 

31.

32.

33.

34.

35.

36.

37.

38.

122

Costa. G.M. Favreau. L.v.. Beller. D.L.. and Rosenberg.
R.D. (1982) J.BiOl. Chem. 221. 11248-11255.

Wasteson. A.. HOOk. M.. and Westermark. B. (1976).
FEBS lett. g1. 218—221.

Yahalom. J.. Elder. A.. Fuks. z.. and Vlodavsky. I.
(1984) J. Clin. Invest. 11. 1842-1849.

Naparstek. Y.. Cohen. I.R., Guks. 2.. and Vlodavsky. I.
(1984) Nature (London) 229. 241—244.

Klein. 0.. and von Figura. K.. (1979) HOppe-Seyler's
z. Physiol. Chem. 229. 1465—1471.

Klein. U.. and von Figura. K. (1976) Biochem. Biophys.
Res. Comm. 1;. 569—576.

Leroy. J.C.. Ho. M.W.. Mac Brinn. M.C.. zielke. K..
Jacob. J.. O'Brien. J.S. (1972) Pediatr. Re. g. 752—
757.

Berman. E.R.. Kohn. G.. Yatziv. 8.. and Stein. H.
(1974) Clin. Chim. Acta 22. 115-124.

 

Chapter 4

Turnover of Cell Surface Proteoglycans in Cultured
Fibroblasts: Internalization and

Early Steps in Degradation

James H. Brauker. James G. Laing and John L. Wang

Department of Biochemistry

Michigan State University

East Lansing. MI 48824

123

SUMMARY

3580 _2—labeled glycosaminoglycans was

The turnover of 4

studied in human fibroblasts labeled for 24 hours and
chased in unlabeled medium with and without the
lysosomotropic amine NH4C1. In the presence of NH4Cl, the
release of intracellular radioactivity from labeled cells
was reduced, reflecting the known inhibitory action of
NH4Cl on intralysosomal enzyme activity. Analysis of the
intracellular radiolabeled molecules revealed: a) NH4Cl
treated cells store two classes of GAG chains; one
fraction of large molecules is composed mainly of
galactosaminoglycans (75%), while another fraction of
smaller molecules is composed mainly of heparan sulfates
(HS); b) the large galactosaminoglycan GAG chains are
delivered to the degradative compartment only during the
labeling period, and not during the chase, consistent with
the observation that there are few galactosaminoglycan PG
at the cell surface; c) the small HS chains are
internalized from the cell surface continuously during the
chase; d) the internalized HS GAG chains are acted on by a
HS endoglycosidase which reduces the GAG chain to about 1/6
of its original size. It is likely that the HS

endoglycosidase is a nonlysosomal enzyme, since it is

124

125

active in cells treated with NH Cl,

4 which raises the pH of

lysosomes to pH 6 or higher.

Subcellular fractionation, on gradients of Percoll, of

cells labeled with 35804"2 in the presence of NH4Cl yielded

two species of GAG containing particles; 3 light fraction,

which contained mainly the large GAG fragments, and a dense

fraction, which contained smaller GAG fragments. Both the
light and dense fractions contained galactosaminoglycans

(GO-70%) and HS (SO-49%) as well as the lysosomal enzyme

marker B-hexosaminidase. These results suggest that the

initial steps in degradation of cell surface HSPG may not

depend on an acidic lysosomal compartment, but may, in

fact, involve a nonlysosomal endoglycosidase residing in an

endocytotic vesicle.

 

 

I NTRODU CT I ON

Human fibroblasts synthesize glycosaminoglycans (GAGs)
which are covalently linked to core proteins to make
proteoglycans (PG). The PG are either secreted into the
medium. transported to the cell surface (pericellular
pool). or tranSported directly to a degradative compartment
in the cell (1). Secreted PG contain mainly GAG chains of
the class known as galactosaminoglycans [dermatan sulfate
(08) and Chondroitin sulfate (C3)]. whereas cell surface PG
contain mainly heparan sulfate (HS) GAGs (2-7). The
intracellular (degradative) pool contains about 60—80%
galactosaminoglycans and the remainder HS (2.5.6). In
practically all cases studied. the intracellular GAGs are
free polysaccharide chains. devoid of core proteins
(2.5-7).

In the preceding paper (6)(chapter 3. this thesis). we
have reported the turnover of cell surface PG in cultured
fibroblasts from I-cell patients. which have a deficiency of
hydrolases in the lysosomes (8). These I-cells store
undegraded or partially degraded GAG chains in
intracellular compartments (6.9). We have found that the
early steps of degradation of the cell surface HS of I-

cells are not defective even though the cells have a

126

 

 

127

generalized deficiency of intracellular lysosomal hydrolase
activity. These early steps in the degradation of HSPG in
I—cell include (6): (a) internalization of cell surface PG;
(b) separation of GAG chains from core proteins (presumably
by either endoglycosidase or protease activity); and (c)
reduction in the size of the HS GAG chains by an
endoglycosidase activity. We have also analyzed the
degradation of galactosaminoglycan PGs. These PGs reach
the cell surface either by a direct intracellular route
from the Golgi, or by internalization from the soluble
secreted pool. They are not a major component of the
pericellular pool and are not internalized to any
measurable extent from the pericellular pool. The
galactosaminoglycans are recovered from the intracellular
pool as free GAG chains of approximately the same size as
newly synthesized GAG chains. Therefore, in I—cells, they
are probably not acted on by an endoglycosidase.

In this paper, we provide evidence that the HS
endoglycosidase activity of human fibroblasts is not
affected by treatment with the lysosomotropic amine NH4Cl.
This, in conjunction with the evidence that the enzyme
activity is not defective in I-cells (6), implies that the
HS endoglycosidase activity occurs in a compartment other

than lysosomes.

MATERIALS AND METHODS

Cell culture and radiolabeling procedures

Normal skin fibroblasts (SL66) were from Drs. J.
McCormick and V. Maher [10]. Cells were maintained in
Eagle's Minimal Essential Medium (MEM) (Gibco, Grand
Island, New York) with 20% fetal bovine serum (FBS)
(Hazelton Research Products, Denver, PA), 100 ug/ml
streptomycin and 159 IU/ml penicillin (Sigma, St. Louis,
M0) at 35°C in humidified air. Medium for stock cultures
was changed three times a week; these cells were passaged
before the cultures reached confluence. Confluent
monolayers of cells were removed from the dish with trypsin
(400 ug/ml) in buffer A [11.5 mM sodium-potassium phosphate
(pH 7.2), 140 mM NaCl, 2.68 mM KCl, and 0.01 mM EDTA], and
plated in 6 well dishes (8 cmz/well, Costar, Cambridge, MA)
for 1-2 days before labeling. All cultures were confluent
at the time of labeling.

Cells were radiolabeled in streptomycin—free MEM with

MgCl2 substituted for MgSO4, 20% dialyzed FBS (Gibco), 50

35

uCi/ml H2

SO4 (carrier free, ICN, Irvine, CA).

Harvest 9; cell fractions during label-chase experiments
The radioactivity was harvested either at the end of

the radiolabeling period (0 hour) or after "chasing" in

128

129

unlabeled medium for various times, and designated as
extracellular, pericellular, or intracellular radioactivity
(11). The extracellular compartment was the medium,
including cell secretions, which was removable by
pipetting. The pericellular material was the radioactivity
in the cell monolayer which could be removed from the cells
by trypsinization. The intracellular fraction was the
trypsin resistant radioactivity remaining with the cells
after enzyme treatment. To harvest the radioactivity from
these cell compartments, the medium (extracellular
fraction) was removed and centrifuged for 3 minutes at 1330
g to remove cellular debris. The cell monolayers were
washed several times with MEM; the washes were discarded.
Pericellular and intracellular material were obtained by
removal of the washed monolayer from the dish with trypsin
(400 ng/ml in buffer A for 20 minutes at 35 °C). Cells
which remained loosely attached to the dish were dislodged
by gentle pipetting. This procedure removed all the
radioactivity from the dish. The cell suspension was
centrifuged for 3 minutes at 1330 g. The supernatant,
defined as the pericellular fraction, was decanted. The
remaining cell pellet, or intracellular fraction, was
dissolved in H20 and frozen at -20°C. Protease inhibitors
were added to all of the cell fractions to make final
concentrations of lmM benzamidine HCl, 0.5 mM EDTA, 5 mM N—
ethyl maleimide, 25 mM G-aminohexanoic acid, and 1 mM

phenyl methyl sulfonyl fluoride. Aliquots from

130

extracellular. pericellular. intracellular samples were in
scintillation cocktail (1:9 v/v). which consisted of 7 g
2.5—diphenyloxazole (Research Products International
Corp.). 338 ml Triton X-100 and 667 ml toluene.
Radioactivity was determined on a Packard Tricarb 300CD
liquid scintillation counter and expressed as
disintegrations per minute (dpm).

For "chase“ experiments. the monolayers were washed
after removal of radiolabeling medium and incubated with
MEM containing 20% PBS. NH4C1 (1.0 M) in PBS was added to
make a final concentration of 10 mM. Controls received PBS
alone. Harvest procedures were performed at various times
thereafter. In some experiments the cells were removed
(prior to the chase) from the dish with trypsin (400 pg/ml
in buffer A). and replated in growth medium for one hour
(6). The adherent cells were then washed and "chased" in
unlabeled medium. At various times thereafter. the
extracellular. pericellular. and intracellular compartments

were harvested as above.

Gel filtration

Samples were chromatographed on columns of Sepharose
CLSB (0.7 x 110 cm) or Sephadex G-50 (0.7 x 50 cm) at a
flow rate of 5 ml/hour. The elution buffer was 0.2 M

lithium acetate. 0.2 % SDS. and 1 mM NaN pH 5. or 0.05 M

3’

Tris. 0.05 M Nazso 4M guanidine HCl. 0.5% Triton—X 100.

4!

pH 7. Fractions of 1 ml were collected and analyzed for

131

radioactivity. Chromatographic profiles in these two

buffers were identical.

The radioactivity eluting in regions B or C (see
fig. 1) of the Sepharose CLGB column (eluted with 0.05 M
Tris. 0.05 M Na2804. 4 M guanidine HCl. 0.5% Triton X—100)
was pooled and precipitated in 2 volumes of cold 90%
ethanol. heated for 2 minutes in a QS’C water bath. and
then cooled for 5 minutes on ice. The samples were then
centrifuged for 10 minutes at 12.000 g. The SUpernatant
was removed and the precipitate was su3pended in H20.
Aliquots of 0.3 ml of these samples were treated as
follows: a) For digestion with chondroitin ABC 1yase
(C'ase. Miles. Elkhart. IN). 0.3 ml of a solution
containing 0.1 M Tris. 0.1 M Na acetate. 0.05 unit/ml C'ase
(pH 7.3). were added and incubated for 3 hours at 37°C; b)

For nitrous acid degradations. 37.5 pl of 2.4 M NaNO and

2!
37.5 pl glacial acetic acid were added for 80 minutes at 25°
C. followed by addition of 75 pl of 3 M sodium sulfamate to

stOp the reaction. The degradation products were analyzed

by gel filtration on Sephadex G—50 columns.

Cell fractionation

 

Post nuclear SUpernates were prepared from cells grown

2 flasks as previously described by

to confluency in 150 cm
Rome et al (12). Cells were removed from the flask with

trypsin (400pg/ml in buffer A) pelleted at 1330 g for 3

 

132
minutes. suspended in 0.25 M sucrose. 10 mM EDTA.
recentrifuged. and resu5pended in 4 ml of the same buffer.
This sample was disrupted in a nitrogen cavitation bomb
(Parr Instrument Co.. Moline. Ill.) for 15 minutes at 30
lbs/inz. The sample was then further homogenized with 3
strokes of a pestle in a 5 ml glass cell homogenizer. The
nuclei were pelleted by centrifugation for 10 minutes at
1330 g. the supernatant was saved and the nuclear pellet
was resu5pended in 1.5 ml of the 0.25 M sucrose solution.
rehomogenized with three strokes of the pestle. and
recentrifuged. The second supernatant was combined with
the first. This post nuclear supernatant was separated on
a Percoll gradient. 27% Percoll gradients (13) were
prepared as follows: 9 volumes of Percoll (as supplied by
Pharmacia. Piscataway. NJ) were mixed with 1 volume of 2.5
M sucrose containing 10 mM EDTA. pH 6.8. Thirty ml of 27%
Percoll was layered onto a cushion of 5 ml of sucrose (2.5
M containing 10 mM EDTA. pH 6.8) in a 39 ml quick seal
centrifuge tube (Beckman Instruments. Fullerton. CA); 4 ml
of sample was layered onto the Percoll. Centrifugation was
performed in a Beckman vTi 50 vertical rotor for 1 hour at
33.000 g at 4°C. At the end of the centrifugation. a glass
tube was inserted from the tOp of the centrifuge tube
through the gradient to the bottom of the tube. Fractions
(1 ml) were collected by drawing from the bottom of the
tube with a peristaltic pump. Samples were stored at

-20 C. The lysosomal enzyme 6—hexosaminidase was assayed

133

as follows: Samples were thawed. and 50 pl were added to
250 pl 4—methylumbelliferyl—A—D—glucoside (Sigma) (1.2 mM
in 0.1 M sodium acetate. 0.1% Triton X—100. pH 4.4) for 30
minutes at 37°C. The reaction was Stopped by addition of 1
ml 0.5 M glycine. 0.5 M Na2C03. Fluorescence was measured

on a Perkin—Elmer 650-40 fluorescence spectrophotometer.

   

 

RESULTS

The effect 9: N§4§1 9n the turnover 9: 35§Q4-labeled

components

SL66 cells were labeled with 35

$04-2 for 24 hours.

The labeled cultures were chased in unlabeled medium in the
presence and absence of NH4Cl. The intracellular and
pericellular fractions were isolated and the radioactivity

quantitated (fig. 1). Cells in the absence of NH Cl

4
rapidly lost most of the initial radioactivity during the
first 12 hours (approximately 30% of initial value) (fig.
1a). In contrast, cells treated with NH4C1 lost
intracellular radioactivity at a slow rate; the overall
loss during the 48 hour chase period was about 25%. This
difference in the turnover of [358]GAG is consistent with
the known effect of NH4Cl, a potent inhibitor of
intracellular lysosomal hydrolase activity because it
increases intralysosomal pH (14).

The radioactivity in the pericellular compartment
decreased with time in cells treated both with and without
NH4Cl (fig. 1b). No difference was observed at any time
point between the two conditions. Thus, it appears that
the turnover of the GAGs in the pericellular compartment of

cells is not affected by lysosomal dysfunction. This

134

135

Figure 1. Kinetics of the turnover of 35304—labeled
components in the (a) intracellular and (b) pericellular
compartments of untreated (e ) and NH4Cl treated (e )
cells. Cells were labeled with 35804—2 (50 uCi/ml) for 24
hours, washed and chased in unlabeled culture medium. The
intracellular and pericellular fractions were isolated as
described in Materials and Methods and their radioactivity
quantitated. The data represent the averages of duplicate

determinations (error bars indicate standard deviations

where larger than symbols).

136

¢N

350$

 

 

63.2 I
.228 0....0
8...:8tom

 

 

 

8.42 I
.2200 0:10
to. a. .08ch

 

 

(..ou x "MOM nan bfl/wdp) Kuhuooowoa [$9.]

 

137

notion is consistent with the conclusion derived from our
parallel study comparing GAG turnover in I—cells with

normal cells (6).

Characterization 9f intracellular GAGs lg NH Cl

4

 

 

treated cultures

We had observed in previous label-chase eXperiments
that the intracellular PGs of SL66 cells have. on gel
filtration on Sepharose CL6B. a predominant peak around Kd
0.42 (region B. fig. 2a)(6). This peak decreases with time
of the chase period (Open circles. fig. 2 c.d). consistent
with the kinetic data shown in fig. 1. In contrast. NH4C1
treated cells (closed circles. fig. 2 b.c.d) retain
radioactivity in the intracellular pool during the chase
period. Moreover. in addition to the major 0.42 Kd peak. a
shoulder of radioactivity becomes prominent during the
chase period. This shoulder of radioactivity is centered
around 0.72 Kd. and is referred to as region C material.
This result mimics that obtained when labeled I—cells were
chased and the intracellular radioactivity was analyzed by
filtration on columns of Sepharose CL6B (6).

Radioactivity migrating in regions B and C of
Sepharose CLSB columns (derived from SL66 cells labeled for

24 hours and chased for 24 hours in the presence of NH Cl)

4
were pooled and subjected to treatment with C'ase. The
intracellular material in region B was sensitive to C'ase;

75% of the radioactivity was reduced in molecular weight by

the enzyme treatment (fig. 3a.b). In contrast. only 25% of

138

Figure 2. Gel filtration of 3SSO4—z-labeled
components derived from the intracellular fraction of
untreated cells (0—0) and NH4C1 treated cells (O—i) during
a label (50 uCi/ml. 24 hour) and chase experiment. Samples
representing 10% of the material from the respective
cultures were chromatographed on a column of Sepharose CLSB
(0.7 X 110 cm) equilibrated with 0.2 M lithium acetate.
0.2% SDS. 1 mM NaN3. pH 5. The horizontal bars designated
A.B. and C highlight regions of interest discussed in the
text. The 0 hour time point has open circles to point out
that the cells were not exposed to NH4Cl before the

beginning of the chase.

 

139

 

 

 

 

_ _

 

 

 

mb.

8 4.
x .53 33:828.. Hmong

 

to

Kd

0:0

140

35SO -2—labeled

Figure 3. Gel filtration of 4

components corresponding to region B and C (figure 2) of
the intracellular fraction of NH4C1 treated cells after
35S -2

O (50

chemical treatments. Cells were labeled with 4

pCi/ml. 24 hours) and chased for 24 hours. The
intracellular radioactivity was fractionated on columns of
Sepharose CLGB as described in legend to fig. 2. The
pooled fractions correSponding to regions B and C were
chromatographed without treatment (a. d). after C'ase
digestion (0.025 unit/m1. 3 hours. 37°C) (b. e.). and after
nitrous acid treatment (0.24 M. 80 minutes. ZS’C) (c.f) on
columns of Sephadex G-50 (0.7 X 50 cm) equilibrated with

0.2 M lithium acetate. 0.2% SDS. 1 mM sodium azide. pH 5.

141

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ Region B . Region C
”i l
3F F“ Control Control «- 2.0
e - l .. ..s
'5‘
A 4 - LO _
l0
'9 x
2 2 - as E
o. B
3
>4
> 3:.
E .2
;: C’ase *-
o 8
U o
.9 LS '{5
8 O
O: 0:
'3' 4 - (.0 '70"
3 8
‘__l I.._J
2 " 0.5
Ii; :% Nitrous Acid Nitrous Acid " "5
(.0
4 -
0.5
2 ..
1o 20 30 1o

 

F motion :11

 

 

 

142

the intracellular material from region C was reduced in
molecular weight by C'ase (fig. 3c.d). Conversly. the
material from region B was not affected by nitrous acid
treatment (fig. 3c). but the material from region C was
sensitive to the chemical treatment (fig. 3f). These
results suggest that region B material consisted mainly of
galactosaminoglycans while region C was predominantly HS.
We had previously shown that. for I—cells. the
intracellular PG migrating in region B were
galactosaminoglycans and the intracellular PG of region C
was predominantly HS (6). Therefore. the chemical
composition as well as the size distribution of storage
material in NH Cl treated normal cells is similar to that

4

of I-cells.

The effect gf NH Cl 92 the turnover gf Egg

 

 

 

4

intracellular GAGs

 

Cells were labeled with 35804.2 for 24 hours. The

radiolabeled cultures were trypsinized to remove the
pericellular PG and then chased in the presence of NH4C1.
Control cultures. without any trypsin treatment. were
chased with and without NH4Cl in parallel. The
intracellular GAGs in region B and C of Sepharose CLGB
columns were quantitated (fig. 4). The radioactivity in
region B of the Sepharose CL6B column decreased to

approximately 50% of the initial value during the 48 hour

chase period (fig. 4a). This loss of radioactivity was

 

 

35504—2-labeled

Figure 4. Kinetics of the turnover of
components corresponding to regions B and C (fig. 2) of the
intracellular fraction of NH4Cl treated cells. Cells were

labeled with 35804——2

(50 uCi/ml, 24 hours). Parallel
cultures were trypsinized (400 ug/ml in buffer A) and
replated or left untreated as controls. The cells were
then chased in unlabeled culture medium. The intracellular
fraction was fractionated on columns of Sepharose CL6B as
described in legend to fig. 2. The radioactivity migrating
in the fractions corresponding to region B and C was

quantitated. The data represents the averages of duplicate

determinations (error bars indicate standard deviations).

144

3.501: 06:.

 

we vw we vm
_ _ — _
I o .333. I o .333.
\ 0:0 m 83:. 0:0 m 53:. A
\\ 3353?... 38225 A
... loe

 

 

 

 

 

 

(mm! 40 °/.) Kmuooowoa [$9.]

145

unaffected by pretreatment with trypsin (compare Open
circles in fig. 4a and 4b). These results are similar to
those obtained when the turnover of region B material of
the intracellular GAGs of I—cells was quantitated (6).
Apparently. few galactosaminoglycans are internalized from
the cell surface because the intracellular
galactosaminoglycan pool (region B) is not affected by
prior trypsin treatment.

In contrast. the rate of loss of radioactivity from
region C of the Sepharose CL6B column was sensitive to
trypsin treatment (compare closed circles in fig. 4a and
4b). In untrypsinized cultures chased in the presence of
NH4C1. the radioactivity in region C actually increased to
a level beyond the initial value during the first 24 hours
of the chase period. Subsequently. the level of
radioactivity decreased (fig. 4a). In trypsinized
cultures. the radioactivity of region C decreased
monotonically to about 50% of the initial value (fig.4b).
These results suggest that intracellular GAGs migrating in
region C of the column are derived from the cell surface.
Trypsinization prior to the chase period deprived the
intracellular compartment of the pool of PG originally at
the cell surface. These results are again consistent with
previous findings. based on a comparative analysis of
normal and I—cells. that cell surface HSPG are internalized
and that the GAG chains. which originally (while in the

pericellular pool) migrated at 0.42 Kd on Sepharose CL6B

 

 

146

columns, were degraded to fragments migrating in region C

of the column.

fractions

35304—2, in the presence of

Cells were labeled with
NH4Cl for 24 hours. The cells were then harvested by
trypsinization and disrupted in a nitrogen decavitator and
subjected to subcellular fractionation in 27% Percoll

35804-2 was

gradients. When the radioactivity due to
monitored in the fractions of the gradient, two major peaks
of 35S-label were found (fig. 5, inset). When the
lysosomal enzyme marker, fi—hexosaminidase, was assayed in
the individual fractions, two peaks of enzyme activity were
found, corresponding in position to the radiolabeled peaks.
The results are consistent with the subcellular
fractionation studies of Rome and coworkers (12), who
characterized these peaks of lysosomal enzyme activity.
For this reason, these two peaks were designated as dense
lysosomes (DL) and light lysosomes (LL) in accordance with
the density of the Percoll fraction in which these two
peaks were found and the nomenclature of Rome et al. (12).
The radioactive fractions corresponding to LL were
pooled and analyzed on Sepharose CL6B columns (fig. 5).
The majority of the [35$]GAGS chromatographed at a position
corresponding to 0.42 Kd; there was also a shoulder at Kd =

0.62. In contrast, the pooled fractions derived from DL

 

147

Figure 5. Percoll density gradient fractionation of
NH4Cl treated cells. Cells were labeled with 35504—2 (50
uCi/ml, 24 hours) in the presence of NH4C1 . The
intracellular fraction was collected, and the cells were
disrupted by nitrogen cavitation at 30 lbs./in2 for 10
minutes. The nuclei were removed by centrifugation and the
post nuclear supernatent was layered onto 27% Percoll and
centrifuged at 33,000 g for 1 hour. Fractions of 1 ml were
collected and the radioactivity in each fraction was
quantitated (inset). The fractions were collected from
dense to light (left to right). Samples from dense (DL)
and light (LL) peaks were pooled (indicated by bars) and
chromatographed on columns of Sepharose CL6B (0.7 X 100 cm)
in 0.2 M lithium acetate, 0.2% SDS, 1 mM NaN3, pH 5. DL,
(---4; LL. (-—-)-

148

 

LL
——— DL

 

LL
i—l
20

 

30

  

Fraction #

l0

 

 

. “'1

 

 

2—

 

 

1 I l

 

| x ummd—bfl/uado
0

I" l 1
co

V’ N
(9.0I x wdp) Mimooowna [39;]

LC

0.5

0.0

Kd

149

yielded predominantly GAGs at 0.62 Kd. When samples of LL
and DL were subjected to C'ase treatment. about 42% of the
material in the LL. and about 32% of the material in the DL
were resistant to degradation. These data indicate that
the pooled LL fraction contains approximately 42% HS and
58% galactosaminoglycans and the pooled DL fraction

contains approximately 32% HS and 68% galactosaminoglycans.

DISCUSSION

The data presented here extend our previous results
derived from studies of I-cell fibroblasts (6), which
implicated a HS endoglycosidase activity that is functional
in cells which have dysfunctional lysosomes. We further
show here that: (a) the characteristic features of GAG
storage by the I-cell can be mimicked in normal cells (SL66
fibroblasts) by treatment of the cells with NH4Cl; (b) the
HSPG internalized from the cell surface are acted on by an
HS endoglycosidase, and this activity is not inhibited by
NH4Cl; and (c) two size classes of GAGs can be recovered at
different densities after Percoll gradient fractionation of
homogenized cells.

Both galactosaminoglycans and HS were recovered from
the intracellular (storage) pool of NH4Cl treated cells
after a label-chase experiment. The HSPG of the cell
surface contain GAG chains which migrate on Sepharose CLGB
columns at 0.42 Kd (6). In contrast, the HS GAG chains
found in the cell interior (storage pool) of cells treated
with NH4Cl are predominantly of a smaller size class (0.72
Kd). Thus, we concluded that the HS endoglycosidase
activity is not inhibited by NH4Cl, indicating that the

enzyme is active at pH 6 or higher [the pH of NH4Cl treated

150

151

lysosomes (14)], in contrast to acid hydrolases in the
lysosome.

We have previously reported that I—cell fibroblasts
[which are defective in the transport of lysosomal enzymes
to lysosomes (8)] store HS GAG chains as fragments similar
in size to those stored by NH4Cl treated normal cells shown
here (0.72 Kd) (6). Considered together, these data
suggest that the HS endoglycosidase is active at pH 6 or
higher and that it is transported to its site of action in
the cell by some mechanism other than the known pathway for
transport of lysosomal enzymes [mannose 6-phosphate
receptor transport (8)]. If the enzyme were a lysosomal
enzyme with a high pH optimum, as has been reported for HS
endoglucuronidase in other systems (15,16), it would be
expected to be active in NH4CI treated cells. In untreated
normal cells, however, the activity of such an enzyme at
low pH is expected to be reduced. Moreover, in I-cells,
such a lysosomal enzyme is expected to be misdirected
during transport and secreted. Analysis of the molecules
migrating in region B of Sepharose CLGB columns from
untreated cells, NH401 treated normal cells, and I-cells,
indicates no substantial difference in the amount of large
(0.42 Kd) HS chains. This can be determined by comparing
the area under the curve in region B of the Sepharose CLGB
column profiles of NH4C1 treated and untreated cells after
a 24 hour chase. The total radioactivity of untreated

cells is 32% of that of NH4C1 treated cells (fig 2c), while

152

25% of the radioactivity of NH4C1 treated cells is HS (fig.
3b). Thus, it is clear that HS GAG chains are not degraded
at a slower rate under conditions in which the lysosomes
are acidic. Otherwise, much higher amounts of
radioactivity would be recovered in region B of untreated
cells (fig. 2c).

Our observations are similar to those reported for
ovarian granulosa cells in culture (17). These cells
express both HSPG and DSPG on their cell surface. Both PGs
are internalized by the cells, and the GAG chains are
separated from protein cores. DS GAG chains are then acted
on by exoenzymes without further processing, whereas HS GAG
chains were further reduced in size by an apparent HS
endoglycosidase activity. This enzyme activity was
unaffected by the lysosomotropic amine chloroquine, which
inhibits intracellular lysosomal enzyme activity by raising
the pH of the lysosome (14).

The results of the present and previous studies
provide the basis for considering a schematic diagram to
describe the metabolic fate of the P65 of a cell (fig. 6).
Galactosaminoglycan PG are synthesized in the Golgi and
then secreted into the medium (pathway 1, fig. 6)
Substantial amounts of the GAG chains of these PG become
associated with intracellular degradative compartments.

Two possible pathways of entry into the degradative
compartments are proposed; (a) diversion of the GAGs from

the secretory pool into the degradative pool (pathway 2,

153

Figure 6. Schematic diagram of proposed pathways for
the degradation of GAGs in fibroblasts. Both
galactosaminoglycan and HS PGs are synthesized in the Golgi
(G). The galactosaminoglycan PGs are then secreted from
the cell (pathway 1). They reach light lysosomes (LL) by
diversion from the secretory pathway (pathway 2) or by
receptor mediated endocytosis after secretion (pathway 3).
HSPG are transported to the cell surface as membrane
components (pathway 4). They reach the LL by
internalization from the cell surface (pathway 5). GAGs in
the LL exist as free GAG chains. From here, they may be
degraded directly by exoenzymes (pathway 6), or they may be
transported (or the LL may mature) to dense lysosomes (DL)
(pathway 7). Exoenzyme activity in the DL completes the
process of degradation, releasing free monomers into the
medium (pathway 8). At some point, shortly after
internalization, HS GAG chains are cleaved by HS

endoglycosidase.

154

 

 

humanit‘

Wm

galactosaminoglycan 4

 

 

 

 

 

 

155

fig. 6) and (b) secretion and recapture via cell surface
receptors (pathway 3, fig. 6). In contrast, the HSPG are
synthesized mainly as membrane associated components which
are transported to the cell surface (pathway 4, fig. 6).
These PG must be internalized to reach degradative
compartments (pathway 5, fig.6).

The degradative phase of PG metabolism begins in a
group of vesicles collectively known as LL (fig. 6). This
LL is a light membrane fraction that contains cell
membranes, endosomes, multivesicular bodies, and light
lysosomes (C.F. Roff, personal communication).
Collectively, the contents of LL include both
galactosaminoglycans (58%) and HS (42%). The contents of
the LL may be degraded directly by exoenzymes (pathway 6,
fig. 6) with concomitant release of free sulfate (20), or
they may be transported to BL (pathway 7, fig.6) where
degradation is completed and free sulfate is released
(pathway 8, fig. 6). The DL is thought to represent mature
lysosomes (12). The pooled DL fraction yielded 32% HS and
68% galactosaminoglycans. Thus, it appears that both
galactosaminoglycans and HS are fairly evenly distributed
between the two Percoll gradient fractions.

It is not known whether a single endosome/vesicle
contains both types of GAGs. The intriguing possibility
exists that LL actually consists of two or more
subpopulations of vesicles. One type of vesicle may

segregate galactosaminoglycans, either recaptured from the

156

medium or diverted from the synthetic pool. Another type
of vesicle may contain mainly HSPG derived from the cell
surface. Implicit in this kind of notion is that
galactosaminoglycan PG and HSPG are initially degraded in
separate vesicles. Such segregation of PG derived from
different sources may account for the phenomenon. observed
in our earlier studies. that mannose 6-phosphate prevents
degradation of a population of GAGs derived from the
extracellular matrix (18). In any case. better
fractionation procedures [e.g. 27% Percoll followed by 17%
Percoll gradient fractionation (19)] need to be utilized to
separate the endosomes. multivesicular bodies. etc.. of the
LL fraction. With these separation procedures. the idea of
separate vesicles for galactosaminoglycan and HS

segregation and degradation can be more rigorously tested.

lo.

11.

12.

13.

REFERENCES

Fratantoni. J.C.. Hall. C.W.. and Neufeld. E.F..
(1968) Proc. Natl. Acad. Sci. 69. 699.

Vogel. K.G.. and Peterson. D.W. (1981) J. Biol.
Chem. 256. 13235.

Malmstrom. A.. Carlstedt. I.. Aberg. L.. and
Fransson. L.A. (1975) Biochem. J. 151. 477-489.

Sjoberg. I.. Carlstedt. I.. Coster. L.. Malmstrom.

and Fransson. L.A. (1979) Biochem. J. 118. 257-270.

Kresse. H.. von Figura. K.. Buddecke. E.. and Fromme.

H.G. (1975) HOppe-Seyler's z. Physiol. Chem. 356.

929941.

Brauker. J.H.. and Wang. J.L. (1986) J. Cell.
Physiol. (Submitted).

Carlstedt. I.. Coster. 8.. and Malmstrom. A. (1981)

Neufeld. E.F.. and McKusick. V.A. (1983) In The
Metabolic Basis pf Inherited Disease (Stanbury.

 

J.B.. Wyngaarden. J.B.. Fredrickson. D.S.. Goldstein.

J.L.. and Brown. M.S.. ed.) McGraw-Hill. New York.

Schmickel. R.D.. Distler. J.J.. and Jourdian. G.W.

Drinkwater. M.R.. Corner. R.C.. McCormick. J.J..
Maher. V.M. (1982) Mutat. Res. 196. 277.

Roff. C.F.. Wozniak. R.W.. Blenis. J.. and Wang.
(1983) EXp. Cell Res. 144. 333-344.

Rome. L.H.. Garvin. G.A.. Allieta. M.M.. and
Neufeld. E.F. (1979) Cell 11. 143153.

Merion. M.. and Poretz. R.D. (1981) J. Supramol.

Struct. 11. 337-346.

157

and

A0!

‘

 

14.

15.

16.

17.

18.

19.

20.

158

Dean. R.T.. Jessup. W.. and Roberts. C.R. (1984)

Hook. M.. Wasteson. A.. and Oldberg. A. (1975)
Biochem. Biophys. Res. Comm. 51. 1422—1428.

Oosta. G.M. Favreau. L.V.. Beller. D.L.. and
Rosenberg. R.D. (1982) J. Biol. Chem. 252. 11248-
11255.

Yanagishita. M.. and Hascall. V.C. (1984) J. Biol.
Chem. 255. 10270-10283.

Brauker. J.H.. Roff. C.F.. and Wang. J.L. (1986)
EXp. Cell Res. 251. 115-126.

Roff C.F.. Fuchs. R.. Mellman. I.. and Robbins.
A.R. (1986) J. Cell Biol. (in press).

Rome. L.H.. and Crain. L.R. (1981) J. Biol. Chem.
255. 10763-10768.

 

NV

lililllliililtil.iHliiliil“
3 96 494 7

1

MICHIGAN
1, “

STATE U
‘ mm
|~ ill)"
. i m:
1293 o