Type, location and role of glycosaminoglycans in cloned differentiated chick retinal pigmented epithelium

TISSUE & CELL 1984, 16, (6) 885-908 @ 1984 Longman Group Ltd

B. J. CRAWFORD*

and T. J. CRAWFORDt

TYPE, LOCAT ION AND ROLE OF GLYCOSAMINOGLYCANS IN CLONED DIFFERENTIATED CHICK RETINAL PIGMENTED EPITHELIUM Key words:

Glycosaminoglycans,

clonal culture,

retinal pigmented

epithelium

ABSTRACT. In clonal culture, colonies of 3-4 week old chick retinal pigmented epithelial cells exhibit Alcian Blue positive extracellular matrix (ECM) material on the surface of the cells. Alcian blue positive ECM is located between undifferentiated cells at the edges of the disc-shaped colonies and beneath the differentiated cells in the colony center. The latter material is associated with the basement membrane. The staining properties suggest that glycosaminoglycans (GAG) are present in these regions. Extraction of GAG from homogenates of colonies, followed by electrophoresis on cellulose acetate strips, results in three bands with mobilities similar to those of hyaluronic acid, heparan sulfate, and chondroitin sulfate, respectively. All three bands label with [‘H]glucosamine, and the last two also label with [35S]sulfate. The composition appeared to differ when colonies were grown in different media. Digestion of the GAG preparations with various enzymes suggests that bands 11 and III represent heparan sulfate and chondroitin sulfate. respectively, in colonies grown in Ham’s Flog medium. The composition of band I is as yet undetermined. In minimal Eagle’s medium (MEM), bands I and 111 consisted of hyaluronic acid and chondroitin sulfate. respectively, while band II had properties suggestive of a copolymer of heparan sulfate and an unidentified GAG. Cells release only one [3H]glucosamine-labelled GAG into the medium. This material has a mobility similar to hyaluronic acid and is digested by Sfreptomyces hyaluronidase. suggesting that it is hyaluronic acid. Staining with Alcian Blue at different pH suggests that it may represent the material associated with the upper surface of the cells. Some of the ECM located between the undifferentiated cells and associated with the basement membrane in the differentiated regions of the colonies stains with Alcian Blue at pH 1.0 and 0.2 suggesting that it may contain GAGS found in bands I and II. Colonies treated with medium containing 6.diazo-5-oxo-L-norleucine (DON), an inhibitor of GAG synthesis, for 48 hr showed a reduced Alcian Blue staining of the ECM in the undifferentiated regions. After 72 hr of treatment with DON, the undifferentiated cells had detached from the plate, whereas the differentiated cells remained intact. The results suggest that the GAG may be involved in cellular adhesion, particularly of the undifferentiated cells.

Introduction Changes in composition of glycosaminoglycan (GAG) components of extracellular matrix (ECM) have been correlated with specific morphogenetic events in a number of systems (for review, see Toole, 1981). High levels of hyaluronate in the extracellular matrix have been associated with increased proliferation in several types of * Departments of Anatomy and *Botany and Zoology, University of British Columbia, Vancouver, B.C. V6T lW5, Canada. Received

22 June 1984.

cultured cells (Tomida et al., 1975; Cohn et al., 1976; Hopwood and Dorfman, 1977). In addition, increased hyaluronate production has often been noted when cellular proliferation was stimulated by growth factors (Tomida et al., 1975; Lembach, 1976). This association between hyaluronic acid and cell proliferation is by no means universal. Increased amounts of chondroitin sulfate have been associated with increased proliferation of embryonic rat fibroblasts induced by thrombin (Kittlick and Neupert, 1982) and sulfated GAG predominates in cultures of mitotically active differentiating chick neural retina (Morris et al., 1977).

CRAWFORD

Cell migration and the outgrowth of cellular processes are affected by GAG (Toole, 1981). Migration of neural crest cells (Pratt et al., 1975; Pintar. 1978: Derby, 1978; for reviews, see Le Douarin, 1982: Weston, 1982, 1983). sclerotome cells (Kvist and Finnegan. 197Oa, b), cornea1 keratocytes (Trelstad et cd., 1974; Toole and Trelstad, 1971). ectodermal cells at gastrulation (Solursh and Morriss, 1977) and endocardial cushion cells (Markwald et al.. 1978) occur through regions of the embryo containing high ratios of hyaluronate to sulfated GAG. Recently studies with cultures of neural tissue (Hawrot. 1980; Lander et al., 1982), leukocytes (Forrester and Wilkinson, 1981), and neural crest cells (Newgreen et al., 1982) have demonstrated that GAG alone or in combination with fibronectins affect cellular outgrowth and migration. although different cell types appear to respond differently to the various GAG components (Reichard-Brown and Akeson, 1983). The expression of cellular differentiation is often accompanied by changes in the type and/or amount of GAG present. Differentiation of several cell types is associated with increased levels of sulfated GAG, particularly chondroitin sulfates (Kvist and Finnegan. 1970a. b; Toole and Trelstad, 1971; Toole. 1981). These studies. in addition to numerous others. suggest that specific GAG components may play an important role in regulating cell proliferation, cell outgrowth and migration. and cellular differentiation during development. Three to four-week-old clones of chick retinal pigmented epitheliai (RPE) cells show a number of the GAG-associated activities listed above. Cells at the outer edge of the disc-shaped colony are flattened and unpigmented. Cells in the center of the colony are cuboidal and heavily pigmented (Cahn and Cahn. 1966; Crawford et d., lY72), while those in between exhibit different morphological stages leading to the re-expression of differentiation (Crawford, 1975). Squamous undifferentiated cells at the outer edges of the colonies and those in a stratified region further toward the center, have high rates of division (Crawford. 1979). Differentiated cells in the center of the colony undergo little or no cell division while partially differentiated cells lying be-

AND C‘RAWFORD

tween the central and the outer regions have an intermediate rate. The different regions also exhibit different levels of cellular movement. Slow contraction of groups of cells (focal contractions) which appear to be involved in developing the differentiated shape (Crawford. 1979) occur in regions containing undifferentiated or partially differentiated cells. Differentiated cells in the center do not undergo such movements. Large amounts of extracellular matrix (ECM) materials are present in the colonies. Extracellular matrix with the staining properties of GAG partially surrounds the lateral borders of cells in the squamous zone and is present between cell layers in the undifferentiated zone (Crawford et al., 1972; Crawford. 1975). Typical basement membranes containing both basal and reticular laminae are found beneath the cells in the inner (differentiated) zone of the colonies (Crawford et al.. 1972: Newsome and Kenyon. 1973) but no ECM is found between the differentiated cells. In addition. Alcian Blue positive material is present on all cell surfaces adjacent to the medium (Crawford et ul., 1980). Preliminary studies involving electrophoresis of GAG extracted from colonies grown in Ham’s Flog medium (Cahn and Cahn. 1966) demonstrated that three GAG components. which migrate at similar speeds to hyaluronic acid. heparan sulfate and chondroitin sulfate (Crawford et al., 1980). are present in the colonies. In the present study, identification of the composition of GAG preparations obtained from whole colonies grown in two different types of medium. from different regions of colonies, and from the culture medium was carried out using a combination of cellulose acetate electrophoresis. radioautography. and digestion with glycosaminoglycan degradative enzymes. Further. localization of GAG in relation to colony morphology was carried out using light microscopy. enzyme electron microscopy, and histochemistry. cytochemistry with the d,ye Alcian Blue. Finally. the results of inhlblting the synthesis of GAG with the drug 6-diazo-5-oxo-Lnorleucine (DON) was analyzed with light and electron microscopy in an attempt to determine some of the functions the GAG may perform during the re-expression of differentiation in culture.

GLYCOSAMINOGLYCANS

IN RETINAL

887

CELLS

Materials and Methods Cell culture Chick retinal pigmented epithelial (RPE) cells were dissociated and cultured according to methods described previously (Cahn and Cahn, 1966; Crawford et al., 1972). The cells were grown in a 37°C gassing incubator in 7% CO* in air in 60 mm Falcon plastic Petri dishes in either Ham’s Flog HO.1 medium (Flog) (Cahn and Cahn, 1966) or minimal Eagle’s medium (MEM) with 10% foetal calf serum. Isolation and characterization of GAG Extraction. Three to five-week-old (RPE) clones were washed twice in Saline G (Puck et al., 1958) and scraped from the plate in 3-S ml of Saline G. Following this, the cells were centrifuged gently (speed 3 for 34 min in an International clinical centrifuge) and the supernatant was removed. The cells were then resuspended in approximately two volumes of distilled water and stored at -20°C. GAG were extracted after the technique of Solursh and Morriss (1977), modified as follows: the thawed cell suspensions were homogenized for l-2 min with a Dounce homogenizer, and the tube was placed in boiling water for 3 min to destroy endogenous polysaccharidases. The homogenates were digested with 5 mgiml pronase (Calbiochem) in 0.1 M Tris buffer, pH 8.0, at 55°C for 20 hr. Merthiolate (0.01%) was added to inhibit bacterial growth during this period. Cold trichloroacetic acid (TCA) added at a final concentration of lo%, in order to precipitate proteins. The homogenate was centrifuged at 27,000 g for 30 min, and the supernatant was dialyzed against three to four changes of distilled water at 0-4”C for 2% days. The GAG were precipitated overnight at 0-4”C by the addition of three volumes of cold 95% ethanol containing 1% potassium acetate and 1% acetic acid. The precipitate was collected by centrifugation and redissolved in distilled water (ca. 0.1 ml for an initial sample of 0.1 ml packed cells.) Electrophoresis

Aliquots (10-15 ~1) of isolated GAG were electrophoresed on cellulose acetate strips (Sepraphore III, Gelman) in 0.15 M zinc

acetate, pH 5.9 (Haruki and Kirk, 1967; Pearce et al., 1972; Hsu et al., 1972), for 2%-3 hr with a current of 2 mA/strip. A mixture of known polysaccharides containing 2 mg/ml chondroitin sulfate C (Miles, special grade), 2 mgiml dermatan sulfate (chondroitin sulfate B, Miles, special grade), 2 mgiml hyaluronic acid (Sigma), and 6 mg/ml heparin sulfate (a gift from Dr R. H. Pearce) were run on the same strip to facilitate identification of the unknown GAG. The strips were stained for 15 min with 0.1% Alcian Blue in 5% acetic acid and 10% ethanol, destained in 5% acetic acid, and photographed. Enzyme digestions

Samples of the unknown GAG isolated as above were digested with either 20 mgiml bovine testicular hyaluronidase (Worthington HSER) in 0.1 M sodium acetate buffer, pH 6.0 for 16 hr at 37°C (Culling, 1974); 200 TRUiml of Streptomyces hyaluronidase (Sakagakei Fine Biochemicals, Miles) in 0.1 M phosphate buffer, pH 5.0, for 4 hr at 37°C (Ohya and Kaneko, 1970), 0.X units/ ml of chondroitinase AC or ABC (Sigma) in 0.1 M Tris, pH 8.0. for 1 hr at 37°C (Yamagata et al., 1968); 1 mg/ml crude heparinase (a gift from Dr A. Linker) in 0.1 M sodium acetate buffer, pH 7.0, containing 1 mM calcium acetate for 16 hr at 30°C (Dr A. Linker, pers. comm.); or 0.015 U/ml of Hepartinase (Miles) in 0.1 M sodium acetate containing 1 mM calcium acetate, pH 7.0, for 18-20 hr at 43°C (Linker and Hovingh, 1972). Samples of known substrates were digested with the same enzyme preparations, spotted on filter paper, and stained as above, in order to test for enzyme activity. Each enzyme digestion was performed at least three times. Labelling of GAG

Three to five-week-old colonies were incubated in medium containing either 20 ,uCi/ ml [35S]sulfate or 2 &i/ml [‘Hlglucosamine for 24-48 hr. The GAG were isolated from the colonies, electrophoresed, and stained. In addition, the medium in which labelled colonies were grown was collected, digested with pronase, and the proteins precipitated with TCA as above. After the supernatant was dialyzed, it was concentrated to half the starting volume with C-bag concentrators

CRAWFORD

xxx cells

(Pierce Chemical). The GAG were then precipitated with ethanol, separated by electrophoresis and stained. After staining, the strips containing labelled GAG were taped to glass plates, and a reference mark was made with radioactive ink. Material labelled with [“SS]sulfate was radioautographed on Kodak No Screen X-ray film. Samples labelled with [“Hlglucosamine were radioautographed with LKB Ultrofilm “H. Films were processed according to the manufacturer’s instructions.

were treated as follows: clones on 60 mm plates were fixed at room temperature with 2.5% glutaraldehyde in Sorensen’s phosphate buffer, pH 7.4, containing 1% Alcian Blue. Clones were either left in situ for 1 hr or fixed for 15-30 min in situ and carefully dissected free of the plate and fixed for a further 31/2 hr to allow more time for the dye to penetrate. The colonies were then washed twice in 2.5% NaHC03 buffer. pH 7.2. and post-fixed in 1% OsOJ in 1.25% NaHC03 buffer. pH 7.4, for 1 hr at room temperature. This was followed by a rinse in 2.5% NaHCOi buffer. dehydration in ethanol and embedding in Epon 812 (Luft, 1961). Clones which remained in situ were transferred directly from 100% ethanol to Epon (Crawford, 1972). Clones which had been removed from the plates were embedded in the normal manner using propylene oxide before transfer to EPON. The EPON was polymerized for 24 hr at 60°C. In order to determine the effect of pH on Alcian Blue staining at the light and electron microscopic level. two different methods were used. Some colonies were fixed in 2.5% glutaraldehyde containing 1% Alcian Blue buffered at either pH 7.4, 3.2. 1.0 or 0.2 (see above), both in situ and after removal from the plate. Other colonies

Whole mounts Whole mounts of 3+I-week-old colonies, grown on coverslips, were fixed in glutaraldehyde in Sorensen’s phosphate buffer, pH 7.4, and washed in buffer. Three colonies were stained with Alcian Blue (Eastman Co.) at either pH 7.4 (Sorensen’s phosphate buffer), pH 3.2 (0.5% acetic acid), pH 1.0 (0.1 N HCI) and pH 0.2 (10% sulphuric acid) for 1 hr. Some colonies were stained as above at pH 7.4 or 2.5 and counterstained with PAS. The coverslips were then dehydrated in alcohol, cleared in xylene, and mounted on glass slides. Sectioned material for microscopy Three to four-week-old Table

light and electron clones

AND CRAWFORV

of (RPE)

1, Results of Aician Blue staining at different pH at the electrorl microscope level PH

ECM location

Fixation Alcian Blue in fixative

(a) Apex (b) Between (c) Between cells Cells removed from plate (a) Apex (b) BM (c) Between cells

Alcian Blue post-fixation

Colony zones

7.4

3.1

1.5

0.2

Cells in situ

Cells (a) (b) (c)

in

All zones

+ +++

(+)

-

-

++

+

+

+

(+)

-

-

Inner

two zones

++

t

(+)

~

Outer

fwo 7ones

+++

++

+

+

All zone Inner two zones Outer two zones

+++

++

+

+

All zones Inner two zones Outer two zones

++ +++

+ ++

(+) +

+

situ

Apex BM

Between cells

Cells removed (a) Apex (b) BM (c) Between BM, basal membrane. * Undifferentiated regions. t Differentiated regions.

All zones Inner two zonest Outer two zones-

-

from plate

cells

-

GLYCOSAMINOGLYCANS

IN RETINAL

CELLS

were fixed in situ and either left in situ or removed from the plate and then stained with 1% Alcian Blue at the different pHs described above. The clones were washed in buffer at the appropriate pH to remove the excess dye, post fixed in 1% 0~0~ in 1.25% NaHC03 pH 7.2-7.4, dehydrated in a graded series of ethanols, embedded in Epon and sectioned for LM and TEM. At least three colonies were examined at each pH in both experimental groups. Effects of DON on colonies

Three to five-week-old clones on 60 mm plates were incubated as follows: (i) five plates were incubated with 3 ml of normal growth medium (normal control); (ii) four plates were incubated with 3 ml of medium containing 50 pgiml DON and 100 mgiml glucosamine (glucosamine controls); (iii) four plates were incubated with 3 ml of medium containing 50 pg/ml DON; (iv) one plate was incubated with 3 ml of medium containing 10 pg/ml DON; (v) two plates were incubated with 3 ml of medium containing 20 ,ug/ml DON. Both the undifferentiated cells at the edge of the colony and the differentiated ones in the center of specially marked clones were photographed on a Wild phase contrast microscope on days 0, 1, 2, and 3. On days 1 and 2, one plate containing 50 pg/ml DON, one normal control, and one glucosamine control were fixed for light and electron microscopy in 2.5% glutaraldehyde containing 1% Alcian Blue as above. On day 3, plates treated with 10, 20 and 50 pgiml DON as well as one normal and one glucosamine control were also fixed as above. One plate which had been treated with 50 pug/ml for 3 days was washed 2~ in growth medium and allowed to recover for a further 4 days. The marked colonies were photographed each day. After this period this plate and another normal control were fixed as above. The entire experiment was repeated twice. The effect of 50 pg/ml with both normal and glucosamine controls was repeated a further four times. Results

(a) Identification of extracted glycosaminoglyclans Electrophoresis

of GAG

extracted

from

whole colonies grown in either Ham’s Flog or MEM demonstrated three Alcian Bluepositive bands (Figs. la, 2a). The first band (I) co-migrated with the hyaluronic acid standard. The second (II) exhibited a migration rate that was similar to, but slightly slower than, both the dermatan sulfate and heparan sulfate standards. These migrated at the same rate in this system and formed a single broad band. The fastest migrating band (III) had a slightly slower migration rate than a mixture of chondroitin sulfates A and C. Bands II and III labelled when [35S]sulphate was included in either type of medium (Figs. lb, 2c). All three bands labelled with [3H]glucosamine when the colonies were grown in MEM (Fig. 2b). When GAG extracted from cultures grown in Flog was digested by bovine testicular hyaluronidase and electrophoresed, bands I and III were missing while band II remained undigested (Fig. lc). Streptomyces hyaluronidase did not appear to cause a complete loss of band I, although it was difficult to tell as band I overlapped the origin in these strips (Fig. Id). Chondroitinase ABC, like testicular hyaluronidase, appeared to eliminate bands I and III but did not affect band II (Fig. le). Crude heparanase eliminated all three bands (Fig. If). Material isolated from colonies grown in MEM containing [“Hlglucosamine demonstrated a different electrophoretic pattern when digested with the degradative enzymes. After digestion with Streptomyces hyaluronidase, only band I was missing (Fig. 2d), a result which was confirmed by radioautography of the strips (not shown), while digestion with chondroitinase AC (Fig. 2e) and ABC (not shown) caused the loss of all three bands. Digestion with heparitinase demonstrated that the enzyme caused the loss of band II and had little or no effect on band III (Fig. 2f). Its effect on band I could not be determined. The results of these enzyme digestions are summarized in Table 2. In addition, all three bands were strongly represented in GAG isolated from the differentiated cells located in the colony center (Fig. 2g), but band I appeared to decrease and band III to increase relative to band II in material isolated from the un-

x90

CRAWFORD

AND

c

CRAWFORD

s

‘nE

8

01 I,

Fig. Ham’s sulfate strip.

Sepraphore

Flog

second

to that that

(II)

and third

hands

(III)

and

Swqfo!n~ce~

sulfatasc

(HE).

rcmovcd

hand

Sveprornwe~

labelled.

that

hovine

hyaluronidasc

Fig.

The

material.

black

(b)

[‘H]glucosamine

in (a) (F)

digested

wth

hcparitinase

while

only (d)

(IlS).

radioautography

is relatively Fig

\trongcr

3.

showing

are

.Srre@v?~w~v AC

(AC),

lahellcd and

(k.

I)

matcnal

heparitinaae i\ removed

chondroitinase

position

of the hand

AC

plus

after

hut

strip

by black

hyaluromdase (f)

Crude

(H)

heparan

ABC

have

all hand5 while

the

while

(g.

(HS). ib not

lines.

with h)

wtb

R

removed is marked

III

GAG

extracted

The

lrom

it\ I-adwautogrnph i\ lahrlled. testicular ABC

ban.

4c

h>

ABC

mcdtunr

and

(b)

of photograph\

hyaluromdase (I.

I

hand one

aupcrlmposed.

mdlcatca the radioautograph. hyaluromdax. Sfrqzovuw~

by chondroltmasc

the cell\

(S) and ongin\

Pair\

(ABC).

II the

from

ME>1

(c-l)

radioautographs

hy the black

m

in hoth tax\.

standard\

and

hand

standards

the undlfferentiatcd

cxtractcd

a\

(AC‘)

m the umhffcrentintcd

grown.

label

same material

heparitinase.

Matenal

arc prcsxt

hovmc

tcstvxlar

(g)

m

(c) A

bands

AC.

the

handr

cbondro;ltmase

The

01

from

one hand

wperimposcd

hy hovine

in the control

(C)

only

The

I and

extracted

ot

all three

(d-f)

hand

grown undigested

[‘Hlglucosaminc

whdc

positions

wcrc

colonw

Control.

(c) chondroitinaae

and hand

control

with

that

rcmovcs

all three

the colonic\ that

labelled

(SH). The

from (a)

[“S]sulfatc.

black

radioautograph\

dIgestion

(SH).

wth

@on

extracted standards.

Note

(h) Matenal

that

the

to (a)

hands.

by

center.

and

hyaluromdase

strip Note

arc marked

removed

of GAG

label

all

Undigested

d)

of a Gmilar

chondroitmase ha\

(S)

hyaluromdaae

in which

strips

(c.

cnzymc

(‘%jrulfate.

(ABC).

and

hyaluronidase

marked

showing

(a)

ABC

[“S]sulfate.

II and III

of (a) showing

Scpraphore

radioautographs:

with

remove,

stnps

as above.

radioautograph

on each

without

with

testicular

sulfatasc

of a similar

bands

[‘H]gluco~amme

bovine

origins,

m the diffcrcntmtcd

Scpraph<~re

arc marked The

the

labelled

cells m the colony

at the cdgc of the colony

in

1s hght and close to the origin. labelled

radioautographs

Sfrepioqwr~

controls

ditferentiatcd

contaming

and

AC

grown

and dcrmatan

or no effect.

Strrpromwes

chondroltinasc

colonies

of the standards

with

hcparan

colome~

sulfate

A radioautograph

hyaluronidasc

represent

of material

(h)

Chondroltinase

A radioautograph

radioautograph with

strips

first (I)

positton

I. Crude

arrows

from

in medium

digestion

(e)

has had littlc

The

from

A and C (3) arc pretcnt

extracted

rtaining.

testicular

band

2. Sepraphorc

sulfates

grown

The

(SH).

111 and possibly

MEM.

hcavicr

had been

cxlractcd

(I ). heparan

acid

Matenal

as in (a) after

hyaluronidaae Note

(a)

are present.

show

111 are

same matenal

of GAG

hyaluronic

and chondroitin

hands

the colonies

II

(S).

the origin.

that three

where

(c) The

B) (2).

represent Note

rn (a)

only

lines. (d)

(C).

radioautographs

standards

sulfate

arrow

treatment

and

Four

(chondroitin The

The

stnp\

medium.

\idc

wrh (H).

the (c.

I)

i) chondrwtmaw Nott: that the hyaluromdaw heparltinase.

The

S:HS

3 L,

2 *‘ 1

Is

-

C-CR

31 0

al

SZEJE

s

XYZ

CRAWFORD

AND CRAWFORD

Table 2. Summury of results of radioactive lahelling and digestion wth Laariousenzymes of GAG exlructed from colonies and culrure medium Band I (MEM)

3epto.pzyce.s hyaluronidase

Removed

Chondroitinase

AC

Removed

Chondroitinase

ABC

sulfatase

Removed

No effect

label

]‘H]glucosamine

(MEW

No effect

Not labelled

Labelled

Labelled

differentiated regions of the colonies (Fig. 2h). Digestion of purified standards (2 mgiml) with these enzymes under the same conditions as for the unknowns confirmed that chondroitinase AC and ABC digested chondroitin sulfate and hyaluronic acid, and that chondroitinase ABC also digested dermatan sulfate, while neither digested heparan sulfate. Heparitinase digested only heparan sulfate. Electrophoresis of GAG isolated from MEM in which the colonies had been grown for 4X hr demonstrated two Alcian Bluepositive bands (Fig. 3a). The first band, which migrated more slowly than the hyaluronic acid standard, did not label when [‘Hlglucosamine was included in the medium suggesting that it was not synthesized by the colonies. The second, wide band, migrated slightly faster than the hyaluronic acid marker but slower than the

Band III Removed

13H]Glucosamine labelled band from medium (MEM) Removed

No effect

No effect

Removed

Removed

Removed

Removed

Removed

Removed

Slightly

Removed

(crude)

Heparitinase [‘5S]sulfate

Band II

No effect

Bovine testicular hyaluronidase

Heparan

Band I1 (Ham’s Flog)

reduced

Removed Removed

No effect

Labelled

Lahelled

Lahellcd

Lahellcd

No effect Labelled

heparan sulfate marker. The slower portion of this band labelled when [‘Hlglucosamine was included in the medium. This labelled material exhibited a migration rate similar to that of the hyaluronic acid marker. The labelled band was lost when the extract was treated with bovine testicular hyaluronidase (Fig. 3c, e), Streptomyces hyaluronidase (Fig. 3e, f), and chrondroitinase AC (Fig. 3i. j). Neither chondroitinase ABC (Fig. 3g, h) nor heparitinase (Fig. 3k, I) removed the labelled band although chondroitinase ABC may have reduced it slightly. The results of these experiments are also summarized in Table 2. (b) Histochemistry

und cytochemistq

When whole colonies were fixed and stained with Alcian Blue in situ at pH 7.4 and mounted as whole mounts, extracellular material stained only in the squamous and outer edges of the stratified zones.

Fig. 4. Two layers of undlftcrentiated cells tram the outer region ot a control c~~lwq preserved in a fixatwe containing 1% Al&n Blue at pH 7.4. Stamed cxtracellular material forms beads on the apical surface of the cells (large arrows). In addition, stained extracellular matrix (E) consisting of strands and amorphous material is present between the cell layers. Only short regions of stained material are present between the cells and the plate (small arrows). x 8000. Ftg. 5. Undifferenttated squamous cells near the lateral border ot a control colony tixcd with Alcian Blue at pH 74. The cells are partially surrounded by heavily stained regwns of extracellular material (E) which often lies immediately adjacent to regions of intracellular microfilaments (M). x5000.

CRAWFORD

Occasional patches of stained material were seen in regions close to the center of the colonies but, for the most part. the inner part of the stratified zone plus the differentiated regions remained unstained. Examinations of sections of colonies fixed in situ with light microscopy and transmission electron microscopy revealed a fine line of stained material over the entire surface of the colony adjacent to the medium. At the EM level, the material adjacent to the medium exhibited a beaded appearance in both the undifferentiated (Fig. 4) and the differentiated regions (Fig. 6). The beads of stained material were usually associated with the outer leaflet of the membrane (Fig. 6 inset), although in some cases the staining appeared to penetrate the membrane completely and to extend into the adjacent cytoplasm. In order to visualize this material, it was necessary to have the dye present in the fixative. Colonies stained in 1% Alcian Blue at pH 7.4 immediately after fixation in phosphate-buffered glutaraldehyde alone failed to exhibit any material on their free surfaces. In the undifferentiated regions of colonies fixed in situ with Alcian Blue at pH 7.4, TEM revealed that the large amounts of Alcian Blue stained ECM was located between the lateral cell borders (Fig. 5). Where undifferentiated cells were multi-

AND CRAWFORD

layered, ECM was also located between the cell layers (Fig. 4). Near the outer edge of the clones in the squamous and outer regions of the stratified zones where the cells were loosely packed and exhibited punctate intercellular junctions, the material was densely stained. Further toward the colony center in the inner half of the stratified zone, the majority of the material located between the cell layers often failed to stain. In such regions. small amounts of stained ECM were often located immediately beneath gaps between adjacent cells suggesting that lack of staining ECM in these regions may have been due to a failure of dye penetration. When the colonies were removed from the plate during fixation at pH 7.3 in the presence of Alcian Blue or when they were stained with Alcian Blue after fixation. the entire colony stained. The heaviest staining was again localized in the undifferentiated zones; however, I pm sections of these colonies revealed that regions beneath the differentiated cells were also stained. The sections also revealed that the cytoplasm of some of the cells had stained in the samples that had been fixed prior to staining. This was probably due to damage to the cell membranes which allowed penetration of the dye into the cells. TEM of this region revealed that Alcian

Fig. 6. The apical protrusions of cells from a differentiated rcgwn of a colony fixed wth Alcian Blue at pH 7.4 showing beads of Alcian Blue stained material on the apical surface (arrows). ~26,700. Inset:A higher magnification showing the lower leaflet of the cellular membrane (arrows) beneath the stained material. x 150.000. Fig. 7. The base of differentiated cells fixed for 15 mm m glutaraldchydc and Alclan Blue m siru. followed bv removal from the nlate and further fixation for 3’1~ hr. The entire basal lamina is stained &d gives a felt-like’ appearance (B). In addition. a network of ataincd material lies beneath this structure in a region where only collagen fibres arc seen with routine fixation. Fibres showing the banding pattern of collagen (C) are often associated with stained material. x49.000. Fig. 8. Cells in an undifferentiated zone of the colony qtaincd with Alcian Bloc at pf1 3.2. after fixation in 1% gluteraldehyde at pH 7.4. These cells lie closer to the ccntcr of the colony than those seen in Fie. 1. No beads of stained material arc seen on the aoical cellular surfaces. Extracellular material (E) between the cell laycrr exhibits patchy \tammg. Iuart Higher magnification of extracellular material to show the densely stained S-20 nm fibres (F) and the heavily stained amorphous patches (A) m more detail. Note that majority of this region is only moderately stained. ~32.000.

CRAWFORD

Blue stained ECM in the lamina lucida externa and interna and the lamina densa of the basal lamina (Fig. 7). In addition, a fibrous network of densely stained ECM was present in the region occupied by the reticular lamina. Stained material was also associated with the edge of the collagen fibres (Fig. 7). Whole mounts of pre-fixed whole colonies, stained with Alcian Blue at pH 3.2, 1-O and 0.2, demonstrated a gradual decrease in staining in all regions, although material between the cells in the squamous and stratified regions was still stained at pH 0.2; 1.5 pm sections of colonies stained at pH 0.2 showed that some stain was also visible beneath the cells in the differentiated zones. Incubation of whole mounts with bovine testicular hyaluronidase for 24 hr at 37°C (Culling, 1974) essentially eliminated the Alcian Blue staining at pH 3.2. Studies of material stained and fixed at lower pHs at the TEM level revealed that although a few beads of material were seen on the surface when Alcian Blue was present in the fixative at pH 3.2 (Fig. 9), they were not seen in material fixed at pH 1.0 or 0.2. At pH 3.2, the regions of ECM between cells at the outside edge of the colonies were densely stained and similar to those seen at pH 7.4. In the stratified regions of the undifferentiated zone some areas demonstrated ‘patchy staining’ (Fig. 8) with some regions of the ECM staining heavily and others much more lightly. Closer examination revealed that some of the more densely stained regions appeared to contain fibres with a size ranging from 5 to 20 nm (Fig. 8 inset). Indeed, the majority of the fibrous material between cells in the stratified zone and at the outer edge of the colony retained a light stain at pH l-5 and 0.2 (Fig. 11). The basal lamina was lightly stained and had the appearance of loose felt-work cut in section at pH 3.2 (Fig. 10) and 1.5. It did not appear to be stained at pH 0.2. The collagen of the reticular lamina demonstrated similar but less intense staining at pH 1.5 and O-2 to that seen at pH 3.2 (Fig. 10). These results are summarized in Table 1.

(c) Effect of DON Colonies grown in

media

containing

50

AND CRAWFORD

,ug/ml DON or 50 pgiml DON plus 100 mgiml glucosamine for 24 hr were essentially the same as the controls at the light microscope level (Figs. 12-15). Differentiated cells in the center of the colony retained their polygonal shape and their characteristic arrangement, and the undifferentiated cells at the colony edge remained as a single sheet. In colonies grown in media containing 50 pgiml DON for 48 hr. the pigmented cells at the center of the colonies also demonstrated no detectable differences from controls at the light microscope level. At the outer edge of these colonies, however, large holes were present in the cell sheet (Fig. 16). The cells seemed less well spread and numerous large apical protrusions were visible. After 72 hr in 50 pg/ml DON plus 100 mg/ml glucosamine, cells in both regions of the colonies were similar in appearance to controls at the light microscope level (Fig. 17). In colonies treated with media containing 50 pg/ml DON alone for 72 hr. the differentiated region in the colony center still demonstrated little or no visible changes. The undifferentiated regions. on the other hand, were essentially missing. and a thick ridge of cells was visible at the edge of the differentiated zone (Fig. 18). When the drug was removed from such colonies, cells with a morphology similar to that seen at the colony margins of controls were present within 24 hr. Ninety-six hours after the drug had been removed (Fig. 19). a halo of undifferentiated cells was again present at the outer edge of the colonies. A ridge usually marked the position of the colony edge after 72 hr of treatment. Electron micrographs of DON-treated colonies and glucosamine controls after 24 hr showed only minor changes (no! shown). Beads of stained material were still present on the apical surfaces of both the differentiated and the undifferentiated cells. Differentiated cells in the center of the colony exhibited no visible changes from the controls, and the basal and reticular laminae appeared intact. In both cases, the undifferentiated cells were associated with large amounts of densely stained extracellular material. However, in DON-treated colonies, the cells did show numerous small vacuoles which were not present in controls. Undifferentiated cells treated with

897

GLYCOSAMINOGLYCANS IN RETINAL CELLS

Fig. 9. The apex of differentiated cells fixed with Al&n Blue at pH 3.2. Some stained material is still present on the apical surface (arrows). Melanin granules can be seen within the cytoplasm (M). x 15,500. Fig. 10. The base of differentiated cells fixed at pH 7.4 and stained at pH 3.2 after removal from the plate. Note the decreased staining of the lamina lucida (LL) over that seen in Fig. 4 and the felt-like appearance of the Iamina densa (LD). A region of densely stained amorphous material is seen beneath the reticular lamina (RL). C, collagen fibres; M, melanin granule. X 19,003. Fig. 11. Poorly differentiated cells from the outside edge of a colony fiied at pH 7.4 and stained with Alcian Blue at pH 0.2. Note that the fibrous extracellular material (E) shows moderate staining at this pH.

DON plus glucosamine also demonstrated these vacuoles but were otherwise similar to the controls. After 48 hr of exposure to 50 pg/ml DON without glucosamine, differentiated cells in

the center of the colonies seemed relatively unaffected. The cell morphology and arrangement were similar to the controls at the electron microscopic level, and the staining of the basal and reticular laminae

898

CRAWFORD

Fig. 12. A phase contrast colony. x70.

micrograph

showing undifferentiated

AND CRAWFORD

cells at the edge ot a control

Fig. 13. A phase cotrast micrograph equivalent to that shown medium containing 50 @g/ml DON plus 100 mgiml of glucosamine.

in Fig. x70.

11. after 24 hr in

Fig. 14. A phase contrast micrograph equivalent to that shown in Fig. medium containing 50 rgiml DON. Little or no effect is apparent. x70. Fig. 15. A phase contrast micrograph equivalent to that shown m Fig. medium containing 50 @g/ml DON plus 100 mgiml glucosamine. x70.

Il.

after

11.atter

24 hr

4X hr m

Fig. 16. A phase contrast micrograph cqmvalent to that shown m Fig. I I. after 48 hr in medium containing 50 pgiml DON. Note that the cells on the edge of the colony arc less spread and the edge itself is irregular with small isolated Islands of cells which are often separated from the main mass of the colony. x70. Fig. 17. A phase contrast micrograph of a colony grown for 72 hr in medium containmg 50 pgiml of DON plus 100 mgiml of glucosamine. Note that the morphology of the undifferentiated region at the colony edge is similar to that of the other controls although a few more holes exist in the cell sheet (arrows). Differentiated cells are visible at the lower right (D). x70. Fig. 18. A phase contrast micrograph of a colony similar to that seen in Fig. 11, after 72 hr in medium containing 50 &ml DON. Only the differentiated cells in the center of the colony are visible (D) bordered by a rolled edge (R). Scattered undifferentiated cells (U) remain attached to the plate. x70. Fig. 19. A phase contrast micrograph of a similar region of the after the medium containing DON had been removed and medium. Undifferentiated cells (U) arc again present at the edge which marked the colony boundary in Fig. 17 is still visible. visible to the left of the photograph. x5000.

colony seen in Fig 17. 72 hr replaced by normal growth of the colony. The ridge (R) Differentiated cells (D) arc

Fig. 20. An electron micrograph of a section taken parallel to the plate ot cells at the edge of a colony which had been grown for 48 hr in medium containing 50 pg/ml DON. The extracellular material (E) appears leached out (compare with Fig. 2). In addition, the strands of stained material appear to be separating from the cells in numerous places (arrows), M. microfilament; N, nuclei. x7500. Fig. 21. A cross-section of undifferentiated cells from a colony which had been grown m medium containing 50 pg/ml of DON plus 100 mgiml glucosamine for 72 hr. Extracellular material (E) consisting of strands and amorphous regions are seen above the cells, and the cells are firmly attached to the plate. The apical cellular membrane is more densely stained than material which has not been fixed m Alcian Blue although beads of material on the apical surface are not present and the cells are filled with small clear vacuoles (V). x lO.ooO. Fig. 22. Cells forming the ridge at the rdge of a colony, grown tor 72 hr m medmm containing 50 Kg/ml DON and fixed in siru in fixative contammg 1% Alcian Blue. The cells have detached from the plate and rolled up surrounding some extracellular material (E) which is unstained except where it was in direct contact with the fixative (large arrows). Stained material is present on the apical surface although it does not exhibit the beaded appearance seen in the controls and large apical blisters are seen associated with the apical surfaces of the rolled up cells (small arrows). Stained extracellular material (E) is also present beneath the cells and on what was the lower surface of the rolled area. Cell processes (P) extend through this material. x4050. Inset: A higher magnification of the region beneath the diferentiated cells near the rolled edge of the colony. The extracellular matrix (E) is patchy in apperance, and a distinct basal and reticular lamina are not visible. x 14,000

I

-’

CRAWI-ORD

was unchanged. However, the amount of stained material on the apical surface was reduced. Instead of the beaded appearance present in the controls, the surface exhibited strands of weakly stained material giving the apical surfaces a fuzzy appareance. Undifferentiated cells in the outer zones of the colony demonstrated much more effect. The cells contained numerous small vacuoles. The apical surfaces of these cells had also lost their beaded appearance and were similar to those in the differentiated cells in the center of the colony. In addition, the ECM surrounding them was more lightly stained (Fig. 20). Numerous well-stained extracellular fibrous elements were present. but the regions between the fibres now appeared washed out. The association between the extracellular fibres and intracellular filaments was less obvious than that seen in the controls, and in some cases the appearance suggested that contact of the intracellular filaments with the cell membrane had been lost. After 72 hr in SO pgiml DON plus 100 mgiml glucosamine, the cells at the outer edge were still attached and were surrounded by extracellular matrix with a staining density similar to that present in the controls (Fig. 21). The cells were not totally unaffected as their surfaces had lost their beaded appearance, and numerous vacuoles were apparent. Differentiated cells in the colony center appeared identical to those in the untreated controls. Examination of material grown for 72 hr in medium containing 50 pgiml DON, demonstrated that the rolled-up margin seen in the light microscope consisted of poorly differentiated cells which had remained attached to the differentiated cells in the center of the colony. On occasion some extracellular matrix was located between these cells (Fig. 22). Examination of three single cells and three small groups of undifferentiated cells which were still attached to the plate revealed that these cells were associated with extracellular material in every case. The cells were vacuolated, and the apical membranes had lost their beaded appearance. The differentiated cells in the center of the colonies retained their characteristic shape and arrangement and exhibited intact basal and reticular laminae.

ANL) C‘RAWFORI>

Towards the rolled edge of the colony. the cell shape was distorted, and the basal and reticular laminae were discontinuous (Fig. 22 inset) with numerous apical protrusions projecting through them (Fig. 22). The apical surfaces of these cells were similar in appearance to those described after 48 hr of exposure to the drug. In addition, the apical surfaces of cells close to the edge demonstrated numerous clear blebs which were often associated with stained material (Fig. 22). Electron micrographs of material incubated in media containing SO ,ug/ml DON for 3 days and then allowed to recover for a further 3 days showed that the differentiated cells in the center were morphologically undistinguishable from the controls. This included the apical surfaces which had a beaded appearance. The flattened unpigmented cells located at the margin of such colonies were surrounded by densely staining extracellular material with a similar appearance to that seen in the controls, and there were few, if any. vacuoles present in them, A thickened ridge consisting of several layers of cells occupied the region which marked the position of the rolled edge of the colony described above. The apical surfaces of the undifferentiated cells also exhibited a beaded apperance similar to that found in the control cells.

Discussion The three classes of GAG, which were present in and synthesized by clones of retinal pigmented epithelial cells, and the single labelled GAG extracted from the medium were partially identified by migration rates, labelling characteristics, and susceptibility to digestion with various enzymes. The single type of GAG which is synthesized by the colonies grown in MEM and is released into the medium appears to be hyaluronic acid. This is shown by its migration rate and susceptibility to Streptomyces hyaluronidase. an enzyme specific for hyaluronic acid (Ohya and Kaneko, 1970), as well as to the less specific GAGdegrading enzymes. Lack of total degradation by chondroitinase ABC may be due to inhibitory compounds in the GAG preparation, such as GAG originating from the serum. It has been observed that digestion

GLYCOSAMINOGLYCANS

IN RETINAL

CELLS

of hyaluronate by chondroitinase ABC can be inhibited in the presence of chondroitin sulfate (M. Adams, pers. comm.). In material extracted from the colonies, band I was an unsulfated GAG with a migration rate close to that of hyaluronic acid. Initial studies with material grown in Ham’s Flog medium seemed to indicate that this band was not totally susceptible to Streptomyces hyaluronidase but overlap of the band with a residue at the origin obscured the result. When better migration of all bands was obtained with material isolated from colonies grown in MEM, this band was entirely lost after digestion with Streptomyces hyaluronidase as well as GAG-degradative enzymes with broad specificities such as bovine testicular hyaluronidase and chondroitinases AC and ABC. In colonies grown in MEM, band I therefore appears to consist of hyaluronic acid. Band III, a sulfated GAG, exhibits a slightly slower migration rate compared with a mixture of chondroitin sulfates A and C (chondroitin 4- and 6-sulfates). It is also digested by chondroitinase AC, which digests only hyaluronic acid, chondroitin and chondroitin sulfates A and C and by chondroitinase ABC, which digests these plus dermatan sulfate (Yamagata et al., 1968). This suggests that band III is composed of chondroitin sulfate A and/or C. Since partially desulfated chondroitin sulfate migrates at a slower rate than the fully sulfated form in this electrophoresis system (results not shown), it is possible that band III is composed of slightly undersulfated molecules. Alternatively band III may represent a chondroitin sulfate/dermatan sulfate hybrid (CSIDS). CSiDS hybrids, which are found in a number of adult tissues (for review, see Pearson and Gibson, 1982), have a greater ratio of D-glucuronate to L-iduronate residues than dermatan sulfate which contains primarily L-iduronate (Buddecke and von Figura, 1975). They would be expected to have migration rates between dermatan sulfate and chondroitin sulfate. In GAG from colonies grown in Flog, band II, a sulfated GAG, was only removed by digestion with heparanase, an enzyme which digests all known GAG including heparan sulfate (A. Linker, pers. comm.). This suggests that band II is heparan sul-

903

fate. However, when the colonies were grown in MEM, this band became susceptible to chondroitinases AC and ABC (which do not digest heparan sulfate) as well as to purified heparitinase, which digests only heparan sulfate (Linker and Hovingh, 1972). These characteristics are not consistent with any known GAG, but suggest that, when the colonies are grown in MEM, they synthesize a copolymer of heparan sulfate and another GAG which is sensitive to chondroitinases AC and ABC, such as chondroitin sulfate or hyaluronic acid. Katagiri and Yamagata (1981) demonstrated that a change in culture conditions as subtle as changing the feeding schedule from every day to every other day, can dramatically influence the types of ECM components that are synthesized with little accompanying change in cellular morphology. Webber et al. (1984a, b) have shown that alterations in the culture medium of fibrocartilage cells can result in changes in the properties of ECM components, and these biochemical alterations may or may not be accompanied by changes in cellular morphology. Our results suggest that in clones of retinal pigmented epithelial cells, some changes in the character of the GAG components synthesized by the colonies occur when different culture media are used, even though the colonies seem morphologically identical. This suggests that some variations in the characteristics of GAG molecules synthesized by the colonies do not necessarily interfere greatly with cellular growth and differentiation. Although the characteristics of the major GAG synthesized in the two media differed somewhat, the charge densities of the three components remained the same, as shown by their identical relative migration rates as compared with the standards. Turley (1980) demonstrated that changes in cell rnorphology brought about by changes in ECM could be mimicked by changing the charge density on a plastic plate. Perhaps, as suggested by Turley (1980), the most important factor affecting morphogenesis is the charge on the GAG rather than the specific type. In RPE clones differentiated cells in the center of the colonies appeared to exhibit a high concentration of band I (hyaluronic acid) relative to bands II and III whereas

CRAWFORD

the undifferentiated cells exhibited relatively more of band III (chondroitin sulfate). This suggests that in these cells there is a relative decrease in chondroitin sulfate and an increase in hyaluronic acid during differentiation. These results are in conflict with those found during the differentiation of somite mesodermal cells into cartilage (Kvist and Finnegan, 1970a, b), and the differentiation of the avian cornea (Toole and Trelstand, 1971) where the amount of hyaluronic acid is high during cell migration but decreases during differentiation, which is associated with increasing levels of chondroitin sulfate. However, in chick neural retina (Morris et al., 1977) the ratio of chondroitin sulfate to heparan sulfate increased in 5-14-day-old chick embryos, while the synthesis of hyaluronic acid remained low throughout. Our results further suggest that the patterns of GAG production during tissue differentiation may be characteristic of the individual tissues. Alcian Blue is an anionic dye which stains negatively charged groups on large molecules. When used in the fixative at physiological pH. it binds to extracellular glycoproteins and GAG. The dye/GAG complex then appears to form an insoluble precipitate with osmium tetroxide giving a staining identical with that seen when pattern Ruthenium Red is used in the fixative (Behnke and Zelander, 1971). Alcian Blue can also be used to differentiate between the presence of highly charged sulfated GAG such as heparan and chondroitin sulfates, and weakly charged ones, such as hyaluronic acid, in conventionally fixed material by controlling the pH. At pH 1.0 and below only sulfate groups take up the stain. Weakly acidic groups such as hyaluronic acid stain at pH 2.5-3.0 and above (Pearse, 1968). At physiological pH (7.(X7.4), Alcian Blue-positive materials are present in both the differentiated and undifferentiated regions of the colonies. Bovine testicular hyaluronidase digests most GAG except heparan sulfate. The nearly complete removal of stained material from the whole mounts after digestion with bovine testicular hyaluronidase suggests that the majority of it consists of GAG. At the TEM level, the stained material is found on the surface of all the cells, surrounding cells in the

AND CRAWFORD

undifferentiated zones, and in the basement membrane. In the latter structure. it is associated with the lamina rara externa and interna of the basal lamina and also with the collagen fibres of the reticular lamina. The material located on the apical cellular surfaces does not appear to stain at a pH less than 1.0, suggesting that it is an unsulfated GAG. In addition, it appears to be easily removed from the cells. Since only one unsulfated GAG. hyaluronic acid. is present in the colonies and this is also found in the medium. we suspect that the material located on the apex of the cells contains hyaluronic acid. Material located between the cells in the undifferentiated regions of the colonies demonstrates a gradual decrease in staining as the pH is decreased, but remains lightly stained at pH I.0 and 0.2 at both the light and electron microscopic level. This suggests that it contains a sulfated GAG. Material isolated from these regions demonstrates large amounts of band 111 (chondroitin sulfate) and band II (heparan sulfate and/or a heparan sulfate copolymer) suggesting that these materials may be a part of the ECM located in this region. The Alcian Blue-positive material located between the cells in the undifferentiated regions of the colonies. often appears in association with extracellular 5-20 nm fibres. These fibres are associated with intracellular microfilaments at the cell membrane at the electron microscope level and are similar in structure to fibronectin which is also associated with intracellular microfilaments across the cellular membrane (Singer, 1979). In addition these regions stain strongly with antifibronectin at the light microscopic level (Crawford and Vielkind, 1984a, b), giving evidence that they contain fibronectin. Fibronectins bind GAG, particularly hyaluronic acid and heparan sulfate (Yamada, 1981). both of which are found in this region of the colony. It is possible that some of the Alcian Blue-positive ECM associated with the extracellular fibres contains these GAG components. Heparan sulfate proteoglycan has been found in basement membranes of chick retinal pigmented epithelial cell explants (Turksen er al.. 1983), and both heparan sulfate (Laurie er al., 1982) and chondroitin

GLYCOSAMINOGLYCANS

IN RETINAL

CELLS

sulfates have been localized in basal laminae of other epithelia (for review, see Toole, 1981). Material in the basal and reticular lamina of the basement membrane of cloned retinal pigmented epithelial cells stains with Alcian Blue at pH down to 0.2 although staining is weak at this pH. This suggests that highly charged acidic groups are present in these regions, groups which are usually found only on sulfated GAG. It is possible that the heparan sulfate and/or heparan sulfate copolymer and chondroitin sulfate are also found in these regions. DON acts by inhibiting the formation of glucosamine, a GAG precursor, causing a significant reduction in [35S0,] uptake into GAG within 3-5 hr in cultured embryonic heart and cartilage cells (Spooner and Conrad, 1975). In REP cells inhibition of GAG synthesis with the drug DON has little or no effect on the presence of GAG for the first 24 hr. This is similar to results from other systems (Lang, 1984). These results suggest that RPE cells may have a reserve of either glucosamine or of newly synthesized GAG sufficient for at least 24 hr. After 48 hr, there is a loss of Alcian Blue-stained material, and holes begin to appear in the undifferentiated regions of the cell sheet suggesting that the cells are losing their ability to adhere to each other. Further, after 72 hr, the undifferentiated cells have lost contact with the dish and many of them appear to have been lost into the medium suggesting that both cell/cell and cell/substrate adhesion are lost. Examination of this material treated for 48 hr with the electron microscope discloses that the extracellular fibrous strands thought to contain fibronectin are losing their attachments to the membrane opposite microfilaments within the cell. Fibronectins are involved in cell/cell and cell/substratum adhesion (Hynes, 1981; Yamada, 1981). Hyaluronate, chondroitin sulfate and heparan sulfate are associated with the footpads of murine cells suggesting they may play a role in cellular adhesion, perhaps by interacting with the fibronectin located there (Gulp et al., 1979) and hyaluronate appears to be necessary for the adhesion of virus transformed chondrocytes to plastic dishes (Mikuni-Takagaki and Toole, 1980). Fibronectin molecules contain carbohydrates (Yamada et al., 1977; Alexander et al.,

905

1979; Yamada, 1981) and also exhibit binding sites for GAGS (Yamada et al., 1980; Yamada, 1981). The results of inhibition of GAG synthesis with DON described above suggest that in RPE clones the GAG may aid in cell/cell and cell/substrate adhesion in the undifferentiated regions of the colonies either directly or by interaction with fibronectins to preserve the attachments of the fibronectin filaments to the cells and to each another. Cells in the center of the colonies remain attached and appear to be relatively normal after 72 hr of growth in medium containing 50 pg/ml of DON, although those at the edge of the differentiated zones are showing signs of loosening. This suggests that the differentiated cells either require less GAG, have a lower GAG turnover rate. or both. In the undifferentiated region of the colonies, the cells either lack junctional complexes or exhibit incomplete ones (Crawford, 1980), and their microfilaments seem to attach to ECM and particularly to fibronectin strands surrounding them. In the differentiated regions of the colonies, little ECM is present between the cells, and the cells are joined by zonular junctional complexes to which the internal microfilaments are attached (Crawford et al., 1972; Crawford, 1979, 1980). It is possible that one reason the differentiated cells are less susceptible to DON is that they adhere strongly to one another via junctional complexes and that ECM is only required for the adhesion of cells to the substrate. Differentiated cells rest on an organized basal lamina containing GAG and a thick reticular lamina of collagen (Crawford et al., 1972; Newsome and Kenyon, 1973 which appears to contain fibronectin (Crawford and Vielkind, 1984a, b) and GAG. Undifferentiated cells rest on a thin layer of ECM derived from serum components of the medium (Crawford et al., 1981; Crawford, 1983), although their lateral borders are often adjacent to ECM containing fibronectin and GAG. In the undifferentiated regions, the ECM is lost into the medium after 24 hr of DON treatment while that associated with the organized basal lamina is retained. In cultured mammary epithelial cells the presence of type I collagen reduces the degradation of basal lamina proteoglycans (Guido and Bernfield, 1979, 1981).

906

CRAWFORD

Perhaps the resistance of the differentiated cultured RPE cells to DON is also due to the stabilization of GAG by the collagen located beneath them.

AND

CRAWFOKD

Acknowledgement This work was supported Canada (grant MA-6337).

by

MRC

of

References Alexander.

S. S.. Co~lonna,

glohulm Behnke,

J. Bio/.

0.

and

preparative Buddecke,

and

van

D.

and

I he structure

and stahthty

of human

plasma

msoluhle

cold

150&1505.

1Y71.

Preservation

m electron

microscopy.

Figura.

IY7Y.

K.

1974.

of

mtcrccllular

substances

by the

cationic

dye

alclan

blue

in

J. Uhrustrucr. Res.. 31, 324-438.

Different

types

of chondroitm

sulfate-dermatan

sulfate

and Related Subjecfs. 22, Prorids of biological Fhids (ed. H. Peeters).

Cahn,

differentiation R.

T.

H.

hybrids

in arterial

pp. 2 19-226.

Pergamon

Oxford.

R.

Cohn.

tdelhoch.

254,

Zclander,

In Proreins

Press, Cahn,

Chem.,

procedures

E.

tissue.

Ci. and

H..

control

M.

in retinal

Cassiman,

University

Herttability

Bernfield.

ultrastructure

Crawford.

B. J. 1075. The B. J. 1979.

cellular

R

ditferentlation:

1976.

Relationship

synthcaued

and dlfferentiatton

clonal

growth

and

expresston

ot

Acud. Sci. U.S A.. 55, 106144.

Nafn.

of transformation.

glycosaminoglycan

J

of chick

retinal

pigmented

cell denstty

and growth

Bioi.. 71. X0!-2Y4.

Cell

cells m ckrnal

culture.

Ph.D.

Seattle.

Structure

Cloned

M.

of newly

of Washington.

Crawford.

of

cells m VIITO. Proc.

distribution

B. J. 1972. The

thesix.

1966.

J. .I. and

to the cellular

Crawford.

B.

pigment

and development

pigmented

retinal

of the pigmented

epithehum.

The

retinal

clone.

C’rw

role of microfilament\

J

Zoo/..

53, 5h(L570.

in the differentiation

of cell

J. Cell Biol., 81, 301-315.

shape. Crawford,

B. J. 1980.

cells in clonal Crawford,

Development

culture.

/nv.

B. J. 1983. Some

of the lunctional

factors

complex

during

dittcrcntlatton

of chtck

plgmeuted

epithehal

Vrs. Sci.. 19, 223-237

O>$tih.

controlling

cell polarity

in chick

retmal

pigment

cplthehal

cells m clonal

culture.

Tissur & Cc//. 15, YY3-1005. Crawford.

B. J.. Cloney,

on ultrastructure Crawford.

R. A.

and Cahn,

and behavior.

B. J. and Vielkind.

U.

Z.

R

D.

lY71.

19%

Cloned

Mikrosk.

Zellforsck. Flhronectin

plgmentcd

retinal

cells;

oi cytochalasm

the eftects

B

Anar.. 130, 135-151.

and laminin

in clonal

cultures

of chtck pigmcntcd

retmal

cell\.

Anur. Rec. 208, 37A. Crawford.

B. .I

cells. Crawford,

and

Vim

in

U.

T. and Crawford,

Amt.

B.

during

J.,

Yan.

A.

and

C.

cptthehal

F

A.

Buttcrworth. L. A..

M.

M.

and function

in clones

of chick

rctmal

pigmcntzd

epithclial

(GAG)

in clone\

of chick

pigmented

rctmal

cplthelwm.

1981.

Changes

m surface

morphology

L. lY80.

Changes

m glycosaminoglycans

Am Zoo/.. 20, 1044. of Hi.~foparho~ogical und Hbfochernrccd

Handbook

of ptgmented

retmal

cells

J. ZooL.. 59, 61-69.

Can.

T. and Horspool,

1974.

B. J.. Bunicl.

during

dlftcruntlation

of chick

culture.

1978.

J. and Hltri,

site of murme

Analyrls

crest cell\. J

culture.

cells in clonal

adhesion

A.

neural Forrcatcr.

locattw

7erhique.s.

3rd

cdn.

pp.

2SY-311.

Toronto.

Rollins.

substrate

MacDonald.

in clonal

B. J.. Crawford,

pigmented

Derby,

The

B. J. 19x4. Glycosaminoglycans

differentlatton

Crawford,

Culling,

IYX4h.

Rec.. 208, %A.

Crawford.

Gulp.

Viclkind.

(in press).

cells.

ot glycosaminoglycans

Devl

Sfol..

Two

tunctmnally

dlstmct

pool\

of glycosam~noglyc~~n

m the

within

the extracellular

cnwronment

encountered

hy mtgratlng

66, X-336.

I mm. P C. IYXI.

V.

S. lY7Y

J. Suprumol. Struck.. II. 401-417

and W’lk’

lnhihitttrn

ot Icukocytc

Iocomotton

reduces

glycoummoglycan

hy hyaluromc

acid

J C r/l Hwl.. 48.

315-331. Guide.

D. and Bernfield.

cells:

powhle

M.

GUI&~. D. and Bernfield, mammary Haruki,

cpithelial

F. and Ktrk.

clectrophoresis 39 I-3Y?.

1979. Collagen

mechanism M.

19X1.

cells. J. Cell

J. E. wth

for basal

1967.

lamma Type

I

No/..

91, 281-2X6

A method

~mc wlphate

formatwn. collagen

reduce\

tor \cparatwn

or zmc acctatc

Proc.

dcgradatwn

Arnd. So

N&n. the

degradatmn

of chondrtntm

wlutmn

a\ electrolyte

wlfatc

by cultured

mammary

ep~thehal

U.S.A.. 76, 7X&700 of

haul

B from

medium

lamma

~wmen

Biocl~inl

protcogl~can

h!

A and c‘ hy paper hro$!.\.

Acrrr. 136,

GLYCOSAMINOGLYCANS

IN RETINAL

CELLS

907

cellulose polyacetate strip electrophoresis for the determination of acid glycosaminoglycans. Arm. Biochem.. 46, lS&l63. Hynes, R. 0. 1981. Fibronectin and its relation to cellular structure and behavior. In Cell Bio@y of Extrncellular Matrix fed. E. D. Hay), pp. 295-334. Plenum Press. New York, London. Katagiri. Y. U. and Yamagata. T. 1981. The persistence in the synthesis of type H proteoglycan and type II collagen by chondrocytes cultured in the presence of chick embryo extract. Deal Growfh Differ.. 23, 335-348 Kittlick, P. D. and Neupert. G. 1982. Effects of thrombin on glycosaminoglycans in fibroblast cultures. Exp/ Parhol. (Jena), 21, 26>267. Kvist. T. N. and Finnegan, C. V. 197Oa. The distribution of glycosaminoglycans in the axial region of the developing chick embryo. I. Histochemical analysis. J. Exp. Zoo/.. 175, 221-240. Kvist, T. N. and Finnegan, C. V. 197Ob. The distribution of glycosaminoglycans in the axial region of the developing chick embryo. II. Biochemical analysis. J. Exp. Zool.. 175, 241-258. Lander, A. D., Fujii. D.. Gospodarowicz, D. and Reichardt, L. F. 1982. Characterization of a factor that promotes neurite outgrowth: evidence linking activity to a heparan sulfate proteoglycan. J. Cell Biol., 94, 57&5X5. Lang, L. 1984. Effects of 6-diazo-5-oxo-L-norleucine (DON) on the growth and migration patterns of two human cervical carcinoma cell lines: role of extracellular materials. M.Sc. thesis, University of British Columbia, Vancouver, B.C., Canada. Laurie. G. W.. Leblond. C. P. and Martin, G. R. 1982. Localization of type IV collagen. laminin. heparan sulfate proteoglycan, and fibronectin to the basal lamina of basement membranes. J. Cell BioL, 95, 34&344. Le Douarin. N. 1982. The Neural Cresr. Cambridge University Press, Cambridge, London, New York. Lembach, K. .I. 1976. Enhanced synthesis and extracellular accumulation of hyaluronic acid during stimulation of quiescent human fibroblasts by mouse epidermal growth factor. J. Cell Physrol.. 89, 277-288. Linker. A. and Hovingh, P. 1972. Heparinase and hcparitinase from flavobacteria. Methods in Enzymology. 28. 902-911. Luft,. J. H. 1961. Improvements in epoxy resin embedding methods. J. Biophys. Biochem. Cyrol.. 9, 400-414. Markwald, R. R., Fitzharris, T. P.. Bank, H. and Bernanke. D. H. 1978. Structural analysis of the matrical organization of glycosaminoglycans in developing endocardial cushions. Devl Eioi.. 62, 292-316. Mikuni-Takagaki, Y. and Toole, B. P. 1980. Cell-substratum attachment and ccl1 surface hyaluronate of Rous szarcoma virus-transformed chondrocytes. J. Cell Biol.. 85, 481-488. Morris, J. E.. Hopwood. J. J. and Dorfman. A. lY77. Biosynthesis of glycosaminoglycans in the developing retina. Devl Biol.. 58, 313-327. Newgreen, D. F., Glbbins. I. L., Sauter. J., Wallenfels. B. and Wutz, R. 1982. Ultrastructural and tissue-culture studies on the role of fibronectin, collagen and glycosaminoglycans in the migration of neural crest cells m the fowl embryo. Cell Tiss. Res., 221, 521-549. Newsome. D. A. and Kenyon. K. R. 1973. Collagen production in vitro by pigmented retinal epithelium of the chick embryo. Dev[ Biol.. 32, 387-400. Ohya, T. and Kaneko, Y. 1970. Novel hyaluronidase from Srreptomyces. Biochim. Biophys. Acta, 198, 607-609. Pearce. R. H.. Grimmer. B. J. and Matheson, J. M. 1972. Fractionation of rat cutaneous glycosaminoglycans using an anron-exchange resin. Analyr. Biochem., 50, 63-72. Pearsc, A. G. E. 1968. Hisrochemiary. Theoretical and Applied. 3rd edn. Little, Brown and Co., Boston. Pearson, C. H. and Gibson, G. J. 1982. Proteoglycans of bovine periodontal ligament and skin. Occurrence of different hybrid-sulphated galactosaminoglycans in distinct proteoglycans. Biochem. J., 201, 27-37. Pintar, J. E. 1978. Distribution and synthesis of glycosaminoglycans during quail neural crest morphogenesis. Devl Biol.. 67, 444364. Pratt. R. M.. Larsen, M. A. and Johnston. M. C. 1975. Migration of cranial neural crest cells in a cell-free hyaluronate-rich matrix. Devl BioI.. 44, 289-305. Puck. T. T.. Cieciura, S. J. and Robinson, A. 1058. Genetics of somatic mammahan cells. III. Long-term cultivation of euploid cells from human and animal subjects. J. Exp. Med.. 108, 945-956. Reichard-Brown. J. L. and Akeson, R. 1983. Correlation of the cell phenotype of cultured cell lines with their adhesion to components of the extracclfular matrix. Devl Biol., 96, 304-316. Singer, 1. I. 1979. The fibronexus: a transmembrane association of fibronectm-containing fibres and bundles of 5 nm microfilamcnts in hamster and human fibroblasts. Cell, 16, 675-685. Solursh. M. and Morriss. G. M. 1977. Glycosaminoglycan synthesis in rat embryos during the formation of the primary mesenchyme and neural folds. DevI Bml., 57, 75-86. Spooner. B. S. and Conrad, G. W. 1975. The role of extracellular materials in cell movement. I. Inhibition of mucopolysaccharide synthesis does not stop ruffling membrane activity of cell movement. J. Cell Biol.. 65, 28b297. Tomida, M.. Koyama. Ii. and Ono. T. 1Y75. Induction of hyaluronic acid synthetase activity in rat fibroblasts by medium change of confluent cultures. J. Cell Physiol.. 86, 121-130. Toole, B. P. 1981. Glycosaminoglycans in morphogcncsis. In Ce0 Biology of Exrracellulur Marrix (cd. E. D. Hay), pp. 25%294. Plenum Press. New York. London.

908 To&,

CRAWFORD B. P. and Trclstad. R. L. lY71. Hyaluronate chick. Del.1 Bid.. 26, 2X-35.

production

and removal

durmg

corncal

AND CRAWFORD development

in the

Trelstad, R. L.. Hayashi. K. and Toole. B. P. 1974. Eplthelial collagens and gl~cosam~noglycanz m the embryonic cornea. Macromolecular order and morphogenesis m the basement membranes. J. Cell Bid.. 62, X15-X30. Turley. E. A. 19X0. The control of adrenocortical cytodiffcrcntiation by cxtraccllular matrix. Differenriunon. 17. 93-103. Turkscn. K . Aubin, J. E.. Sodek. J. and Kalnina. V I. 1983. Diatrlbution 01 lammm. fibroncctm. type IV collagen and heparan sulfate proteoglycan in colonies of chick retinal pigmented epithelial cclla. J. Cell Bml.. 97, 232a. Webber. R. J.. Harris, M. and Hough. Jr, A. J. 1984a. Proteoglyclans 5ynthesired by rabbit memscal tibrochondrocytes in monolayer culture: Modulation by medium and ascorbic acid ‘Trum. 30th Annual Orthop. Rec. Sot.. At/nntn. Gu.. p. 33X W&her, R J.. Harris. M. and Hough. Jr. A. J. iY84h. Intrtnslc repair capabdltx3 of rabbit mcniscal fihrocartilagc: A cell culture model. 7iat1.7. 30th Annuul Orthop. RN. Sot.. Atlannc. (;a.. p. 27X. W&on. J. A. 19X2. Motile and social behavior of neural crc\t cells. In Crll Brhaviour (cds. A. Curt15 and G. Dunn). pp. 42%470 Cambridge University Press, Camhndgc. London, New York Weston. J. A. 1983. Regulatmn of neural crest ccl1 migration and dlffcrcntiation. In (‘e/l Iwwcfior~ unrl Derdyment~ Mokrulur Mechanisms (cd. K. M. Yamada). J. Wiley and Sons. New York. pp. 1.5.%1X3. Yamada. K. M. lYX1. Fibroncctin and other structural proteins. In C‘rll Biolo,q~ of E.rtruceUulur Mutrix (cd. E. 11. Hay), pp. Y&l14 Plenum Press. New York, London. Yamada. K. M.. Kennedy, D. W.. Kimata. K. and Pratt. R. M. 1YXt).Characterization 01 hbronectln intcractmn\ with glycosaminoglycans and identification of active proteolytic fragments. J. Bid Chem.. 255, 6055-6063. Yamada, K. M.. Schlesinger, D. H., Kcnncdy. D. W. and Pastan. I.. 1977. Characterization of a major fibrohlast cell surface glycoprotein Biochemistry 16, S552-555’). Yamagata, T.. Saito. H., Habuchi. 0. and Suzuki. S. lY68. Purificatwn and propertlcs of bacterial chondromnascs and chondroitin sulfatascs. J. Bid. Chern.. 243, 1523-1535.