Atherosclerosis, 81 (1990) 255-265 Elsevier Scientific Publishers Ireland,

ATHERO

255 Ltd.

04451

Endothelial cell stimulation of smooth muscle glycosaminoglycan synthesis can be accounted for by transforming growth factor beta activity M.J. Merrilees

and L. Scott

Department of Anatomy, School of Medicine, University of Auckland, Auckland (New Zealand) (Received 23 October, 1989) (Accepted 22 December, 1989)

Summary Endothelial cell conditioned medium (ECCM) contains a factor which markedly stimulates smooth muscle cell (SMC) glycosaminoglycan (GAG) synthesis. We report here that the factor responsible is transforming growth factor beta (TGF-/3) as assessed by (1) protease and thiol sensitivity, (2) heat and acid enhancement of ECCM activity, and (3) neutralisation of ECCM activity by anti-TGF+immunoglobulin. Anti-TGF-fl-neutralisation was effective against increases in both sulphated and non-sulphated GAG. Previous studies showed that ECCM from EC of varying densities stimulated individual GAG to varying degrees. ECCM from low density EC preferentially stimulated hyaluronic acid (HA) whereas ECCM from intermediate and high density cultures stimulated increasing amounts of sulphated GAG. Exposure of SMC to varying concentrations of TGF-j3 produced a similar pattern of response. Very low amounts of TGF-fl (< lo-500 pg/lO cells) stimulated a marked and significant increase in HA synthesis. Increase in chondroitin sulphate 4/6 was most marked at TGF-/3 levels from 500-1000 pg/106 cells. At levels above 1000 pg/106 cells both HA and sulphated GAG synthesis decreased but still remained elevated above controls. These findings indicate that TGF-P alone can account for the changes in SMC GAG synthesis stimulated by ECCM. It was also found, however, that heat-treated SMC conditioned medium stimulated SMC GAG synthesis, thus SMC may contribute to the control of their own GAG synthesis through autocrine TGF-0 activity.

Key words:

Endothelial cell stimulation; ming growth factor beta

Smooth

muscle

cells;

Glycosaminoglycan

synthesis;

Transfor-

Introduction Correspondence to: Dr. M.J. Merrilees, Department of Anatomy, School of Medicine, University of Auckland, Auckland, New Zealand. 0021-9150/90/$03.50

0 1990 Elsevier Scientific

Publishers

Ireland,

It is now clearly established that endothelial cells (EC) in culture secrete into growth medium a Ltd.

256 factor which is a potent stimulator of smooth muscle cell (SMC) glycosaminoglycan (GAG) synthesis [1,2,4]. There is indirect evidence that such a factor operates in whole vessels in organ culture [3] and in vivo [5-81. The factor stimulates the synthesis of all GAG, including sulphated and non-sulphated species, whether cultured in the presence of EC or in medium conditioned over EC (ECCM). The factor also appears to alter the balance between sulphated GAG and non-sulphated hyaluronic acid (HA). EC at low density, or their ECCM, preferentially stimulate HA, whereas ECCM from confluent cultures stimulates mainly sulphated GAG, especially chondroitin sulphate 4/6 (CS) [2]. It has also been found recently that both contractile and synthetic state SMC respond to the factor [9]. The importance of these findings relates to the well documented correlation between plaque development, increased levels of intimal proteoglycans (particularly those containing sulphated GAG), and uptake and accumulation of lipoproteins [lo-151. In this paper we demonstrate that both the increased synthesis of GAG by SMC, and the changed proportions of individual GAG, can be accounted for by EC-derived transforming growth factor beta (TGF-P).

Methods CeN culture Porcine smooth muscle cells (SMC) were prepared for culture from explants of thoracic aortas of 6-month-old pigs according to the methods described by Merrilees and Scott [l]. Bovine endothelial cells (EC) were prepared from segments of bovine thoracic aorta. Segments were washed in serum free M199, split lengthwise, and the intima scraped with a scalpel blade to remove the EC. The EC were then shaken gently from the blade into 60-mm Falcon dishes and grown to confluency in (Y MEM supplemented with 15% FBS and 2 mM glutamine. These cells were then subcultured and experiments were conducted on cells from Sl to S18. All cultures were maintained at 37’C in an atmosphere of 95% air and 5% CO,.

Conditioning of media EC which had been seeded 24 h earlier at densities ranging from 0.3 x lo6 to 3.5 x lo6 cells/25 cm2, were washed several times in serum-free M199. Conditioned medium was prepared by culturing EC in 6 ml Ml99 with either no serum present or with 1% FBS for 24 h. The conditioned medium was then collected, pooled where appropriate, and treated by the various methods described below before being placed on washed confluent SMC cultures. EC were trypsinised and cell numbers determined by coulter counting following collection of conditioned media. Trypsin treatment Serum-free Ml99 (SFM199) was conditioned on 0.5 X lo6 EC/25 cm2 flask (ECCM). Post conditioning, ECCM was collected and pooled, and along with control SFM199 was heat treated as described below. Control and ECCM media, were then each divided into three aliquots and treated as follows. (i) Incubated at 37” C for 2,5 h. (ii) Incubated at 37” C for 2.5 h, followed by treatment with 6.75 mg/ml soybean trypsin inhibitor (Sigma) for 30 min at 37 o C. (iii) Treated with 300 pgg/ml trypsin (Difco) for 2 h at 37” C, followed by treatment with 6.75 mg/ml soybean trypsin inhibitor for 30 min at 37 o C. After treatment, duplicate 5 ml samples of each aliquot were added to confluent cultures of SMC. All cultures received 200 PCi [3H]acetate and 50 ~1 FBS. Incorporation of [3H]acetate into GAG was measured after 24 h of culture. Heat treatment Fresh Ml99 containing 1% FBS (1% M199) and 1% ECCM from EC cultures ranging in density from 0.26 to 1.21 x lo6 cells/25 cm2, were heated for 5 min at 100°C in a boiling waterbath. The samples were then cooled to 37 o C before adding 5-ml aliquots to cultures of SMC and measuring incorporation of [3H]acetate over 24 h. Thiol treatment One percent Ml99 was conditioned on confluent EC. Aliquots of 1% control 1% Ml99 were treated as Maintained at room temperature for

(1% ECCM) ECCM and follows. (i) 30 min, then

251 stored for 24 h at 4” C. (ii) Heat-treated at 95 o C for 30 min, then stored for 24 h at 4°C. (iii) Heat-treated at 95” C for 30 min, then dialysed for 24 h at 4” C against 4 changes of 1% M199. (iv) Heat-treated at 95’ C for 30 min in the presence of dithioerythritol (final concentration 20 mM DTE), then dialysed for 24 h at 4°C against 4 changes of 1% M199. (v) Heat-treated at 95” C for 30 min in the presence of 2-mercaptoethanol (final concentration 30 mM), then dialysed for 24 h at 4” C against 4 changes of 1% M199. After these treatments, duplicate 4-ml samples were added to confluent washed SMC cultures. The cultures were then labelled with 200 PCi [ ‘HIacetate and incorporation of [ 3H]acetate measured over the next 24 h. Acid activation One percent Ml99 was conditioned on 0.2 x lo6 EC/25 cm* flask. Control 1% Ml99 and 1% ECCM were then treated either by incubation at 4” C for 1 h, or by acidification to pH 3 with 1 N HCl for 1 h at 4’ C. The acidified samples were then pH-adjusted to 7.4 using a 1 M NaHCO, solution. Following treatment, duplicate 5-ml samples were added to confluent washed SMC cultures. All cultures received 200 PCi [3H] acetate and incorporation measured over the next 24 h. TGF-/I dose response 100 ng porcine platelet TGF-& (R&D Systems, Minneapolis) was reconstituted in 0.5 ml 4 mM HCl. A 1.0 ng/ml TGF-& stock solution was then prepared using SFM199. Control SFM199 medium containing equivalent amounts of 4 mM HCl was then used to dilute the TGF-P, stock solution to give the following concentrations of TGF-& in the growth medium; 0.01, 0.1, 0.5, 1.0 and 2.0 ng/ml. Confluent cultures of EC and SMC were washed 3 times in SFM199. Duplicate 5-ml samples of growth medium, containing the appropriate concentrations of TGF-P, 50 ~1 FBS and 200 PCi [‘HIacetate, were added to cultures for 24 h and then analysed for freshly synthesised GAG. TGF-p neutralising antibody Prior to conditioning of culture medium, EC at different densities were washed and maintained in

SFM199 for 24 h. Conditioned medium was prepared on these cultures using either 1% Ml99 or SFM199. This conditioned medium and the control medium at the appropriate serum level were heat treated as described above and treated as follows: (i) Maintained at room temperature for 1 h. (ii) Incubated with 20 pg/ml (1% ECCM) or 40 pgg/ml (ECCM) of TGF-fi neutralising antibody (Lot 5834 or 5808 British Biotechnology Ltd.) for 1 h at room temperature. Confluent SMC cultures were maintained for 48 h in Ml99 supplemented with either 1% FBS or 0.2% FBS, then washed 3 times in SM199 prior to the addition of 4 ml treated growth medium. Duplicate cultures of each condition were labelled with 200 PCi [‘HIacetate and incorporation into GAG measured for the next 24 h. Autocrine activity One percent Ml99 was conditioned on confluent EC or SMC for 24 h. Conditioned medium and 1% Ml99 control medium were then treated as follows: (i) No further treatment. (ii) Heat treatment as described. (iii) Heat treatment as described, followed by Millipore filtration (SMC only). The confluent cultures used for conditioning the medium were washed 3 times in SM199 and then 5-ml aliquots of the treated samples were added back to these cultures. All cultures were labelled with 200 PCi [jH]acetate and incorporation into GAG measured for the next 24 h. Glycosaminoglycan analysis At the termination of all experiments, the growth medium was collected and the cell layer harvested using 0.05% trypsin/O.Ol M Na EDTA in Tris-buffered saline. The growth medium and cell layers were combined for GAG analyses. Newly synthesised GAG were measured as previously described [1,9]. Results In previous studies we showed that cultured vascular endothelial cells secrete a factor into growth medium which increases the synthesis and alters the proportions of GAG types produced by cultured vascular smooth muscle cells [1,2]. We

258 TABLE

1

EFFECT OF TRYPSIN TED STIMULATION TURED SMC

AND THIOLS ON ECCM MEDIAOF GAG SYNTHESIS BY CUL-

Values are means + range of duplicates. Treatment

% stimulation HA

removed

cs

Total GAG

Trypsin 300pg/m12hat37°C

89* 10

83+17

90+

Dithioerythritol 20mM,95”C30min

72*18

66*21

61+23

Mercaptoethanol 30mM,95°C30min

1

!

600

100

100

100

show here that both the increased synthesis and the change in GAG proportions can be accounted for by TGF-/?. Initial studies showed that the factor was both trypsin and thiol sensitive (Table 1). Trypsin treatment of ECCM resulted in a 90% reduction of the stimulatory activity, whilst there was complete removal of ECCM activity with 30 mM 2mercaptoethanol (2 ME) treatment and 60% reduction with 20 mM DTE treatment. The thiol treatment was unsuccessful when carried out at 37O C for 5 h (data not shown), but significantly reduced the stimulatory activity with harsher treatment of 95 o C for 30 min, a finding consistent with the large number of disulphide bonds present in TGF-j3 [16,21]. The activity of ECCM was not only heat stable but enhanced by heat treatment (loo0 C/5 min) (Fig. 1). The effect of heat treatment on GAG synthesis was variable with the greatest enhance‘ment of synthesis often occurring when the stimulation by non-heat-treated medium was small or moderate. Where SMC showed a good response to ECCM the effect of heat treatment was less marked (Fig. 1). It was also noted that GAG synthesis increased slightly but significantly in heat treated fresh growth medium containing 1% serum (Fig. 1). Enhancement of stimulatory activity also occurred after transient acidification of ECCM to pH 3 (Fig. 2). As with heat treatment, the synthesis of GAG was enhanced, and again the increased

0.0

I

I

0.5

1.0

1.5

EC density x 16925 cmaflask

Fig. 1. Effect of heat treatment (100°C/5 min) on total GAG synthesis by confluent SMC cultured for 24 h in fresh Ml99 containing 1% FBS (zero EC density values) and in ECCM prepared over EC cultures ranging in density from 0.26 to 1.21 X lo6 cells/25 cm2 flask. The bars about the means in this and all subsequent figures are the range of duplicate cultures.

response was most often evident when the initial stimulatory activity in untreated ECCM was low or occasionally even absent (Fig. 2). Control cultures in 1% Ml99 also showed enhanced synthesis after acidification but not to the same degree as for the ECCM (Fig. 1). For both acid and heat enhancement of activity both sulphated and nonsulphated GAG were further stimulated (data not shown). These findings, that the ECCM factor was protease and thiol sensitive, heat stable, and acid and 3000

u) = aI 0 9 z ii 8

2000

1000

0

No Treatment

Acid Treatment

Fig. 2. Effect of acid treatment (pH 3, 30 min) on total GAG synthesis by confluent SMC cultured in fresh Ml99 (FGM) containing 1% FBS and in 1% ECCM prepared over EC cultures at a density of 0.5 X lo6 cells/25 cm2 flask.

259 300

1 1

.looll 0.0

-100 ( 0.0

1.0

0.5 TGF-8 ng/ml

TGF-8

Fig. 3. Effect of TGF-j3 at concentrations of 0.01, 0.1, 0.5 and 1.0 ng/ml on total GAG synthesis by EC at densities of 0) and 2.4 x lo6 (m -W) cells/25 cm2 1.3x106 (Oflask.

heat activated, indicated that TGF-/3 was the most likely candidate. The effect of TGF-& on both EC and SMC was investigated. These studies demonstrated that SMC respond to TGF-fi in the same way as they respond to ECCM by increasing their synthesis of GAG. EC on the other hand showed very little response to TGF-j?. Treatment of EC with varying concentrations of TGF-fi resulted in either no increase or only a slight increase in the synthesis of GAG (Fig. 3)

I 2.0

I 1.o rmlml

Fig. 5. Effect of TGF-/I on the synthesis of HA, CS, DS and HS by SMC at a cell density of 0.8x106/25 cm2 flask. The change in synthesis is expressed as % change over control.

and had no effect on the proportions of individual GAG. In contrast SMC increased markedly their synthesis of GAG, especially at low levels of TGF-/3. A typical dose-response curve of SMC at a high density (Fig. 4) shows that all GAG were stimulated although the greatest increases were for hyaluronic acid (HA) and chondroitin sulphate 4/6 (CS). Similar increases occurred in other experiments but where SMC were of intermediate (0.8 X 106/25 cm2) as opposed to high (1.9 X 106) =

10001

0

2000 -

u) = 8 Q Q Z E B

1500-

0

1000 TGF-B

0: 0.0

.

I

1.o

0.5 TGF-8

1.5

2.0

ng/ml

Fig. 4. Effect of TGF-8 at concentrations ranging from 0.01 to 2.0 ng/mf on the synthesis of hyahtronic acid (HA), chondroitin sulphate 4/6 (CS), dermatan sulphate (DS) and heparan sulphate (HS) by SMC at a cell density of 1.9 x 106/25 cm’ flask.

2000

3000

pa/106cslls

Fig. 6. Relationship between change in SMC GAG synthesis of HA and CS and TGF-8 levels expressed as pg/106 cells. Data are pooled from 4 experiments with SMC densities ranging from 0.8 to 1.9X106 cells/25 cm* flask. The upper (HA) and lower (CS) curved lines are second order polynomial curve fits with r values of 0.40 (P < 0.02) and 0.67 (P < 0.001) respectively.

260 400 ,

B E b

s

300

2500

I

0

2000 I = 8 Q 0 c E * 0

1500

1000

500

CP 0

F”

0

Ia

I

1000 TGF-B

Fig. 7. Percentage

change

I

0

,

FGM

3000

2000

in the HA/CS Fig. 6.

ratios

of data

the

sponse occurred at a higher concentration of TGF-/3. This difference was reflected in an altered HA/CS ratio. In general, at low levels of TGF-P, the HA/CS ratio increased markedly compared with control cultures, but at higher concentrations (P- 1000 pg/106 cells) the ratio then decreased reflecting both a decrease in HA and the shift towards CS (Fig. 7). These changes were similar to those found for ECCM [2]. These results further indicated that the factor in ECCM could be TGF-j3. This was tested directly using TGF-j3 neutralising antibody. The activity in ECCM was neutralised by the addition of TGF-j3 neutralising antibody to the medium (Table 2). Initially, experiments were carried out with 1% Ml99 conditioned on both high and intermediate density endothelial cells. In these

centrations

TABLE

ECCtf (3.4 x 100)

Fig. 8. Effect of anti-TGF-fi immunoglobulin (20 pg/ml) on synthesis of GAG by SMC cultured in fresh Ml99 (FGM) and ECCM both containing 1% serum. ECCM are from EC cultures of two densities.

in

degree of stimulation at TGF-P conof l-2 ng/ml was less than at lower concentrations of TGF-fi (Fig. 5). Thus in pooling the results of several experiments, concentrations of TGF-P were expressed as pg/106 cells and increases in GAG as % change from control. A consistent finding was that the increase in HA was particularly marked at low concentrations of TGF-fi (< 500 pg/106 cells) and significant increases were clearly seen with very low levels of TGF-P ( < 10 pg/106 cells) (Fig. 6). At concentrations of TGF-P above 1000 pg/106 cells, however, the level of stimulation of HA decreased compared with the maximum response seen at lower levels of TGF-P. The pattern of stimulation for CS was similar (Fig. 6) except that in general the response at low levels of TGF-j3 was not as pronounced as that for HA and the maximum redensity,

ECCM_ (0.5 x 100)

~gdcells

2

EFFECT OF TGF+ CULTURED SMC

NEUTRALISING

ANTIBODY

TGF-fi neutralising antibody concentration (pg/ml)

ECCM endothelial cell density/ 25 cm2 flask

Serum concentration (%)

20 20 20 40 40

0.4x 3.4x 3.5 x 1.5 x 0.4x

1 1 1 0 0

lo6 lo6 lo6 lo6 lo6

ON ECCM

MEDIATED

t%stimulation

STIMULATION

over control

OF GAG

SYNTHESIS

% stimulation neutralised by anti TGF-P

HA

cs

Total

HA

cs

Total

7 111 130 19 54

78 116 94 60 96

47 69 109 33 65

4 121 30 102 126

60 38 9 120 84

60 22 20 119 80

BY

261 600

-

0

NoTreatment

1000 600 v) =

I =

8 Q a

8 a 0

400

t:

600

600

z

k 0

8

400

200 200

C

-

0

ECCM

FGM

ECCM

FGM

Fig. 9. Effect of anti TGF-B immunoglobulin (40 pg/ml) on synthesis of GAG by SMC cultured in FGM and ECCM (from EC at 0.3 x lo6 cells/25 cm2 flask) in the absence of serum.

Fig. 10. Effect of heat treatment on synthesis cultured in FGM and ECCM.

experiments, 20 pg/ml of neutralising antibody removed between 20% and 60% of the ECCM stimulatory activity depending upon the magnitude of the initial response (Fig. 8). The smaller

the initial stimulation, the larger the percentage removal of that stimulation by the antibody (Table 2). With increased stimulation the antibody removed a lower proportion of the increase al-

HA

of GAG

by EC

HS

0 Nolrealmenl a HT m HT. millpore

FGM

WC conditioned

medium

FGM

SMC conditioned

medium

FGM

SMC conditioned

madium

DS

FGM

Fig. Il.

Effect of heat treatment

SMC conditioned

and Millipore

medium

filtration

on synthesis medium.

of GAG

by SMC cultured

in FGM

and SMC conditioned

262 though as expected the number of 3H counts which were neutralised in medium from high and low density EC cultures was similar (Fig. 8). It was also noted that in control SMC cultures, maintained in 1% Ml99 for 24 h, GAG production was reduced by about 20% in the presence of neutralising antibody (Fig. 8) suggesting either a small amount of TGF-/3 in serum or neutralization of SMC TGF-/? autocrine activity. Total blocking of the stimulator-y activity was achieved with higher levels of antibody (40 ~g/ml) in serum-free ECCM (Table 2) (Fig. 9). As for the experiments using the lower level of antibody, both HA and CS levels were reduced by antibody (Table 2). Dermatan sulphate (DS) levels were similarly reduced but results were less clear for heparan sulphate (HS). Since in these antibody experiments HS was not significantly stimulated by ECCM, it was not possible to detect blocking by anti TGF-j3. In other experiments, however, we noted that HS increased in the presence of ECCM and we did find that HS synthesis was stimulated by TGF-P (Figs. 4 and 5). Thus it is unlikely that HS responds any differently from the other GAG. In the experiments with 40 pg/ml of antibody in serum-free medium it was noted that there was a small, although not always significant, decrease in GAG in SMC cultured for 24 h in SFM199 in presence of antibody (Fig. 9). This suggested autocrine TGF-/3 activity by SMC. We therefore tested for autocrine activity by both SMC and EC by preparing serum-free conditioned media for replacement on their respective cell types. EC showed no significant autocrine activity with or without heat treatment (Fig. 10). In contrast SMC conditioned medium significantly stimulated all GAG activity after heat treatment (Fig. 11). GAG synthesis by SMC in non-heat-treated SMC conditioned medium, however, was markedly reduced compared with that in FGM, a finding we have observed previously, and reported for fibroblasts cultured in their own medium [41]. In these latter experiments with SMC conditioned medium we also tested for the effect of Millipore filtration on the stimulatory activity. As predicted, there was a marked loss of stimulator-y activity following filtration (Fig. ll), consistent with the adherent characteristics of TGF-P [42].

Discussion The evidence presented in this paper indicates that the SMC GAG stimulatory activity known to be present in ECCM [l-3] can be accounted for by TGF-/?. The evidence is based primarily on TGF-/3 antibody neutralisation of GAG stimulatory activity but is further substantiated by protease and thiol sensitivity of ECCM, by heat and acid enhancement of ECCM activity, and by the similarity in the pattern of GAG stimulation by ECCM and TGF-P. We have not determined which TGF-P subtype [17] is responsible for the stimulatory activity. It has been previously reported that TGF-fi is a potent stimulator of the synthesis of sulphated GAG by SMC but that it has no effect on GAG synthesis by EC [18]. We confirm these findings but in addition report that HA is particularly responsive to TGF-P, especially at very low concentrations (- 0.01 ng/ml) where there may be a 3-5-fold increase in synthesis. Interestingly the stimulation of sulphated GAG at low concentrations of TGF-P was generally less than that for HA but increased significantly as the concentration increased. Increases in both HA and sulphated GAG, however, were less marked at higher concentrations of TGF-/3. Thus not only did changed concentrations of TGF-P affect total GAG production but it also altered the balance between GAG types. These changed patterns and amounts of GAG were similar to those previously reported for ECCM [1,2] and indicate that TGF-/JI alone could play a role in modulating both proportions and amounts of GAG in the arterial wall. The importance of altered amounts and types of GAG in atherosclerotic vessels has been discussed by numerous investigators [19,20] and it is clear than an increase in total GAG, and particularly the sulphated GAG DS and CS, plays a key role in enhancing lipid accumulation [lo-151. It will now be important to determine, in vivo, if TGF-/3 has a role of stimulating intimal matrix formation and if endothelium and/or SMC are important sources of TGF-/% It is not clear if SMC produce TGF-P but some of the findings in this present study (Fig. 11) indicate that there is a strong possibility that

263

autocrine action by TGF-fl could modulate SMC GAG synthesis. On the other hand it has been demonstrated now in a number of organ systems that TGF-fi also acts in a paracrine manner, particularly between epithelial and stromal tissues [21--231. The idea that endothelium controls intimal SMC GAG production is consistent with this model and attractive in view of the demonstrated importance of endothelium in the control of smooth muscle growth and metabolism [l8,24-291. Others have also shown that EC produce TGF-/3 130-321 and interestingly it has been shown that activated TGF-P is produced by co-cultures of EC and pericytes [32]. Both cell types cultured separately produce latent TGF-j3. In our own studies we have noticed that whereas the response of SMC to ECCM can be quite variable a more consistent stimulation seems to occur when the cells are co-cultures. This suggests that here too the TGF-j3 may be in the active form. The mechanism of activation in vivo is not clear although numerous conditions (heat, acid, alkali [33,34]) and factors (plasmin, cathepsin D [35]) are known to work in vivo. The state of the EC may be an important determinant of both the level of production of latent TGF-P and the degree of activation. It has previously been demonstrated that EC in different states and densities can have different effects on SMC [2,25,26] and the findings presented here, showing that different concentrations of TGF-j3 can have differential effects on GAG types and amounts, suggest that, in vivo, EC in different states may modulate their TGF-/I production. EC in different states may also be able to influence latent TGF-fi activation. It will obviously be important to determine how the level of TGF-P is controlled, how latent TGFand how risk factors for P is activated atherosclerosis which affect endothelium also affect TGF-/3. The state of the SMC may also be important. The SMC used in this study were all in a synthetic state but in recent studies on the effects of ECCM on GAG synthesis by contractile and synthetic phenotypes [9] it was shown that both types of SMC respond in a similar manner to ECCM. The synthetic SMC, however, produced significantly more GAG under both unstimulated and stimulated conditions suggesting that they may be more responsive to TGF-P.

Cell density may also effect the response. SMC express distinct TGF-/3 receptor subtypes at different densities which may allow TGF-j3 to exert opposite effects on SMC growth [36]. Whereas the growth of low density cultures of SMC is inhibited by TGF-fi the growth of confluent cultures is potentiated. Our studies were not designed to investigate the effects on SMC growth and we found no evidence of SMC growth over the 24-h period of exposure to TGF-P. The SMC cultures used in this study were also confluent and it is not clear if low density cultures respond to TGF-/?. In one experiment, however, we did investigate the effect of TGF-fi at a moderately low density of 0.5 x lo6 cells/25 cm2 flask and found a level of GAG stimulation within the range of that seen for confluent SMC. We did notice, however, that the degree of HA stimulation was less than that for the confluent cultures but these low density cultures were already producing a characteristically [37] large proportion of HA relative to total GAG and may not have had the capacity to greatly further increase their production of HA (data not shown). The data presented here indicates that EC-derived TGF-fl can stimulate the synthesis of matrix GAG by SMC. Through this action it could play a role in controlling vessel structure. The effects of TGF-j3 and its subtypes, however, are extraordinarily diverse and its importance to normal morphogenesis of vessels, and to atherogenesis, remains to be elucidated. Nevertheless, it is becoming clear that TGF-P plays an important role in morphogenesis and wound healing in general [21,38-401 and it would be thus reasonable to assume that it is involved in some of the events known to take place as atherosclerotic plaques develop. It will be of interest to determine if TGF-fi is merely contributory to plaque development or whether it initiates alterations to intimal structure.

Acknowledgements

The authors wish to thank Miss Fiona Godfrey for assistance in preparation of the manuscript. This work was supported by a grant from the Medical Research Council of New Zealand.

264 References 1 Merrilees, M.J. and Scott, L., Interaction of aortic endothelial cells and smooth muscle cells in culture: effect on glycosaminoglycan levels. Atherosclerosis, 39 (1981) 147. 2 Merrilees, M.J. and Scott, L., Stimulation of smooth muscle cell glycosaminoglycan synthesis by cultured endothelial cells is dependent on endothelial cell density. Atherosclerosis, 63 (1987) 145. 3 Merrilees, M.J. and Scott, L., Effect of endothelial cell removal and regeneration on smooth muscle cell glycosaminoglycan synthesis and growth in rat carotid artery in organ culture. Lab. Invest., 52 (1985) 409. 4 Berrou, E., Breton, M., Deudon, E. and Picard, J., Effect of endothelial-cell-conditioned medium on proteoglycan synthesis in cultured smooth muscle cells from pig aorta. J. Cell Physiol., 137 (1988) 430. 5 Alavi, M. and More, S., Glycosaminoglycan composition and biosynthesis in the endothelium-covered neointima of de-endothelialized rabbit aorta. Exp. Mol. Pathol., 42 (1985) 389. 6 Huang, W., Richardson, M., Alavi, M.Z. et al., Proteoglycan distribution in rat aortic wall following indwelling catheter injury. Atherosclerosis, 51 (1984) 59. 7 Richardson, M., Ihnatowycz, I. and Moore S., Glycosaminoglycan distribution in rabbit aortic wall following balloon catheter de-endothelisation. Lab. Invest., 43 (1980) 509. 8 Salisbury, B.G.J., Hajjer, D.P. and Minnick, C.R., Altered glycosaminoglycan metabolism in injured arterial wall. Exp. Mol. Pathol. 42 (1985) 306. 9 Merrilees, M.J., Campbell, J.H., Spanidis, E. and Campbell, G.R., Glycosaminoglycan synthesis by smooth muscle cells of differing phenotypes and their response to endothelial cell conditioned medium, Atherosclerosis, 81 (1990) 245. 10 Hurt, E. and Camejo, G., Effect of arterial proteoglycans on the interaction of LDL with human monocyte-derived macrophages. Atherosclerosis, 67 (1987) 115. 11 Salisbury, B.C.J., Falcone, D.J. and Minnick, C.R., Insoluble low density lipoprotein-proteoglycan complexes enhance cholesteryl ester accumulation in macrophages. Am. J. Pathol., 120 (1985) 6. 12 Vijayagopal, P., Srinivasan, S.R., Jones, K.M. et al., Complexes of low density lipoprotein and arterial proteoglycan aggregates promote cholesteryl ester accumulation in mouse macrophages. Biochim. Biophys. Acta, 837 (1985) 251. 13 Hollander, W., Unified concept on the role of acid mucopolysaccharides and connective tissue proteins in the accumulation of lipids, lipoproteins, and calcium in the atherosclerotic plaque. Exp. Mol. Pathol., 25 (1976) 106. 14 Iverius, P.H., The interaction between human plasma lipoproteins and connective tissue glycosaminoglycans. J. Biol. Chem., 247 (1972) 2607. 15 Iverius, P.H., Possible role of glycosaminoglycans in the genesis of atherosclerosis. In: Atherogenesis: Initiating Factors, Ciba Symposium No. 12, Elsevier, Amsterdam, 1973, p. 185.

16 Derynck, R., Jarrett, J.A., Chen, E.Y. et al., Human transforming growth factor-/3 complementary DNA sequence and expression in normal and transformed cells. Nature, 316 (1985) 701. 17 Rizzino, A., Transforming growth factor-b: multiple effects on cell differentation and extracellular matrixes. Dev. Biol., 130 (1988) 411. 18 Chen, J., Ho&i, H. and McKeehan, W.L., Transforming growth factor-b specifically stimulates synthesis of proteoglycan in human adult arterial smooth muscle cells. Proc. Natl. Acad. Sci. USA, 84 (1987) 5287. 19 Yla-Herttuala, S., Nikkari, T., Hirvonen, J. et al., Biochemical composition of coronary arteries in Finnish children. Arteriosclerosis, 6 (1986) 230. 20 Berenson, G.S., Radhakrishnamurthy, B., Srinivasan, S.R. et al., Chemical changes in arterial wall during regression of atherosclerosis in monkeys. Artery, 9 (1981) 44. 21 Heine, U.I., Munoz, E.F., Flanders, KC. et al., Role of transforming growth factor-p in the development of the mouse embryo. J. Cell Biol., 105 (1987) 2861. 22 Lehnert, S.A. and Akhurst, R.J., Embryonic expression pattern of TGF beta type-l RNA suggests both paracrine and autocrine mechanisms of action. Develop. Biol., 104 (1988) 263. 23 Thompson, N.L., Flanders, K.C., Smith, J.M. et al., Expression of transforming growth factor-b in specific cells and tissues of adult and neonatal mice. J. Cell. Biol., 108 (1989) 661. 24 Herman, I.M. and Castellot, J.J., Regulation of vascular smooth muscle cell growth by endothelial-synthesized extracellular matrices. Arteriosclerosis, 7 (1987) 463. 25 Castellot, J.J., Favreau, L.V., Kamovsky, M.J. and Rosenberg, R.D., Inhibition of vascular smooth muscle growth by endothelial cell-derived heparin. J. Biol. Chem., 257 (1982) 11256. 26 Castellot, J.J., Addonizio, M.K., Rosenberg, R.D. and Kamovsky, M.J:, Cultured endothelial cells produce a heparin-like inhibition of smooth muscle cell growth. J. Cell. Biol., 90 (1981) 372. 27 Willems, CL. et al., Media conditioned by cultured human vascular endothelial cells inhibit the growth of vascular smooth muscle cells. Exp. Cell. Res. 139 (1982) 191. 28 Schwartz, S.M., Molecular biology and the vascular wall. Arterioslerosis, 4 (1984) 647. 29 Schwartz, S.M. and Ross, R., Cellular proliferation in atherosclerosis and hypertension. Prog. Cardiovasc. Dis.,26 (1984) 355. 30 Carr, B.I., Huang, T.H., Itakura, K. et al,, TGF@ gene transcription in normal and nwplastic liver growth. J. Cell B&hem., 39 (1989) 477. 31 Tsunawaki, S., Sporn, M., Ding, A. and Nathan, C., Deactivation of macrophages by transforming growth factor-/3. Nature, 334 (1988) 260. 32 Antonelli-Orhdge, A., Saunders, K.B., Smith, S.R. and D’Amore, P.A., An activated form of transforming growth factor p is produced by co-cultures of endothelial cells and pericytes. Proc. Natl. Acad. Sci. USA, 86 (1989) 4544. 33 Rosa, R., Roberts, A.B., Danielpour, D. et al., Mesoderm

265

34

35

36

37

induction in amphibians: the role of TGF-/I, like factors. Science, 239 (1988) 783. Lawrence, D.A., P&her, R. and Julien, P., Conversion of a high molecular weight latent /3-TGF from chicken embryo fibroblasts into a low molecular weight active P-TGF under acidic conditions. Biochem. Biophys. Res. Commun., 133 (1985) 1026. Lyons, R.J., Keski-Oja, J. and Moses, H.L., Proteolytic activation of latent transforming growth factor-p from fibroblast-conditioned medium. J. Cell. Biol., 106 (1988) 1659. Goodman, L.V. and Majack, R.A., Vascular smooth muscle cells express distinct transforming growth factor-p receptor phenotypes as a function of cell density in culture. J. Biol. Chem., 264 (1989) 5241. Merrilees, M.J., Synthesis of glycosaminoglycans and proteoglycans. In: J.H. Campbell and G.R. Campbell (Eds.), Vascular Smooth Muscle in Culture, Vol. 1, CRC Press, Boca Raton, FL, 1987, p. 133.

38 Mustoe, R.A., Pierce, G.F., Thomason, A., et al., Accelerated healing of incisional wounds in rats induced by transforming growth factor-p. Science, 237 (1987) 1333. 39 Montesano, R. and Orci, L., Transforming growth factor p stimulates collagen-matrix contraction by fibroblasts: Implications for wound healing. Proc. Natl. Acad. Sci. USA, 85 (1988) 4894. 40 Smiddy, W.E., Glaser, B.M., Green, R.W. et al.. Transforming growth factor beta: A biologic chorioretinal glue. Arch. Ophthalmol., 107 (1989) 577. 41 Merrilees, M.J., Sodek, J. and Aubin, J.E., Effect of cells of epithelial rests of Malassez and endothelial cells on synthesis of glycosaminoglycans by periodontal ligament fibroblasts in vitro. Dev. Biol., 97 (1983) 146. 42 Danielpour, D., Dart, L.L., Flanders, K.C., Roberts, A.B. and Spom, M.B., Immunodetection and quantification of the two forms of transforming growth factor-beta (TGF-B, and TGF-B,) secreted by cells in culture. J. Cell. Physiol., 138 (1989) 79.

Endothelial cell stimulation of smooth muscle glycosaminoglycan synthesis can be accounted for by transforming growth factor beta activity.

Endothelial cell conditioned medium (ECCM) contains a factor which markedly stimulates smooth muscle cell (SMC) glycosaminoglycan (GAG) synthesis. We ...
1014KB Sizes 0 Downloads 0 Views