Vol.
178,
August
No. 15,
BIOCHEMICAL
3, 1991
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS Pages
1991
MODULATION
OF GENE EXPRESSION
LIPOPROTEINS
IN HUMAN
VASCULAR
BY HIGH SMOOTH
1465-1471
AND LOW DENSITY MUSCLE
CELLS
Alfred W.A. Hahn*, Fabrizia Ferracin, Fritz R. Btihler and Alfred Pletscher Department of Research, Base1 University Hospital, Hebelstr. 20, CH- 4031 Basel, Switzerland Received
July
8, 1991
Summary Low density lipoprotein and its oxidized form has been implicated in the process of arteriosclerosis which involves growth- related events in the smooth muscle cells of the arterial wall. The induction of so- called early- growth response genes e.g. c-myc and c-fos can serve as an indicator for these growth- related events. In cultured human vascular smooth muscle cells, both LDL and HDL3 were individually capable of stimulating c-myc and c-fos expression in a concentration dependent manner. However, when they were used in combination, depending on the proportion of HDL3 to LDL, c-fos but not c-myc expression was less pronounced than with the single components. In contrast to HDL3 and LDL alone, a combination of the two lipoproteins also blunted both the expression of autoinduced transforming growth factor 6 transcripts and the transforming growth factor l3- induced increase of c-fos mRNA. It is concluded that a) the inhibition of transforming growth factor 8 autostimulation by HDL3 plus LDL may involve reduced AP-1 activity via a reduction of c-fos expression by the lipoprotein combination and b) the ratio HDL3:LDL might influence the pathogenesis of arteriosclerosis via growth- related events in the arterial wall. 0 1991 *cademic Press, Inc.
B
Plasma lipoproteins are thought to play a role in the pathogenesis of arteriosclerosis (1 -
3). In fact, although elevated concentrations of LDL seem to be a risk factor, HDL appear to exert a protective effect in this condition. LDL has been shown to cause intracellular activation (e.g. stimulation of phosphoinositide metabolism, elevation of [Ca2+]i and enhanced phosphorylation of cellular proteins) in various cell types including vascular smooth muscle, endothelial cells and blood platelets (4.5.6). In addition LDL causes a shape change and has a proaggregatory effect in platelets (578). The action of HDL at the cellular level is less well established. However, HDL3 has been shown to cause a shape change reaction, to antagonise the LDL- induced rise of [Ca2+]i and to counteract
the LDL- mediated
aggregation of human platelets (6.8). HDL3 alone neither caused an increase of intracellular calcium levels, nor induced the aggregation of platelets (6.8) indicating an important interaction between the different lipoproteins.
. . Abbrevratlons: human vascular smooth muscle cells, hVSMC; low- and high density lipoproteins, LDL and HDL, transforming growth factor 6, TGFB. 0006-291X/91 1465
$1.50
Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol.
178,
No.
3,
1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
LDL has also been shown to induce early growth- related events, i.e. transcript expression of c-fos and cmyc in VSMC (9). Such events may be important in the arteriosclerotic process, one characteristic of which is the proliferation of vessel wall constituents. Therefore it is of interest to investigate whether there might also be an interaction between HDL and LDL with respect to these growth- regulating events in cultured hVSMC.
Cell culture and stimulation experimenti Isolation, culture and phenotypic characterization of hVSMC were performed essentially as described before (9). Two different isolates of microarterial hVSMC from different patients undergoing abdominal surgery were used in this study. Cells were used between passage 4- 10. Prior to experimentation, confluent hVSMC were rendered quiescent by culture for 48 hrs (with one medium change after 24 hrs) under serum- free conditions and placed in medium containing 1% plasma- derived serum (mitogen- free) for 12 hrs. Subsequently cells were stimulated with agonist for the times indicated and processed for isolation of total RNA. Northern BJQQQ&& Following stimulation, cultured cells were washed twice with phosphate buffered saline (PBS) and then lysed by the addition of guanidinium isothiocyanate buffer (GT buffer) directly onto the culture dishes. Total RNA was collected on CsCl2- gradients as described before (10). For Northern analysis, 20 pg of total RNA was size- separated on a 1.2% agarose gel containing 2.2 M formaldehyde at a constant voltaae of 5OV for 6-8 hrs in MOPS buffer. The electronhoresed ael was vacublotted onto HYbond N memb;anes (Amersham) with 20x SSC as transfer buffer. Blotted RNA was fixed to the membranes by UV irradiation at 302 nm for 3 mins. Hybridization of blots to random primed cDNA probes (11) was performed according to standard methods (12). The blots were washed at high stringency (0.1x SSC/ 0.1% SDS at 65” C) and exposed to Kodak X-Omat films at -70 C overnight using one intensifying screen. Variabilities of RNA present on the blots were evaluated by tehybridization to a 1.5 kb cDNA probe specific for MHC class I antigens (clone pMF 48) (13) and exposed for 4-6 hrs to X-ray film. Probes used in tbis study were cDNA probes for c-fos (14) and c-myc (15) as well as a 1.2 kb cDNA for TGFB (16). For rehybridizations of blots with different probes, membranes were stripped of their previous signal by incubation for 10 mins. in boiling 10 mM Tris pH 7.5/ 10 mM EDTA. Prior to rehybridization, blots were exposed to Kodak films overnight to ensure complete removal of the previous signals. Signals were densitometrically analyzed by scanning (525 nm) of autoradiographs. The arbitrary optical density (OD) units given in the results were obtained after normalization of signals with respect to the OD values obtained for the MHC class I internal control (highest value within a given series was arbitrarily taken as 100 %). J&.mrotein isolation: LDL (density 1.019 - 1.063) and HDL3 (density 1.125- 1.21) were isolated from the plasma of two different healthy volunteers by ultracentrifugation as described by Have1 et al (17). The lipoprotein fractions were dialysed against isotonic phosphate buffer pH 7.4 containing 1 mM EDTA and 1pM BHT (2.6 di-tert-butyl-p-cresol: Fluka Chemie AG, Buchs, Switzerland) and filtered (0.45 pm). HDL3 was used in the study in preference to total I-IDL. to avoid any possible trace contamination of HDL fractions with LDL. Both, LDL and HDL3 preparations were checked for purity by gel elctrophoresis and HPLC. The protein content was measured (18) and aliquots were stored at 4°C under an atmosphere of nitrogen for not more than a few weeks. During the storage period no change in the activity of the lipoprotein fractions was observed. m
.
.
Both LDL and HDL3 were found to induce c-fos and c-myc transcripts in
quiescent hVSMC (Fig. 1). The stimulation kinetics for c-fos were rapid and transient (maximal at 30 mins. poststimulatory, return to control levels within 1 hr). Relative to c-fos, stimulation of c-myc was also transient but delayed with maximal induction occuring only -3-5 hrs poststimulation (Fig. !A) . The stimulatory effects of the lipoproteins on protooncogene expression were concentration dependent. LDL and HDL3 exhibited similar potencies and maximal effects ( EC50 -100 l&ml; 1466
saturation at
BIOCHEMICAL
Vol. 178, No. 3, 1991
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
hrs 0 .5 3 5 500
0
1000
concentration
1500
(pghl)
. Don in hVSMC -3u. hVSMC were stimulated by either LDL or HDL3 (each at 200 pg/ml) for the times indicated. Northern Mats hybridized to cDNA probes for the protconcogenes were exposed overnight at Fieure
1. &m&ion Panel A: Quiescent
of c-mvc and c-fos mRNA
70°C with one intensifying screen: blots hybridised to the MHC class I probe were exposed under identical conditions for 4-6 hrs only. Comparable results were obtained with two different isolates of cells. Panel B displays a representative concentration response curve for c-fos expression promoted by
different amountsof LDL (open circles) or HDL3 (full circles). The normalized arbitrary optical density (OD) units were obtained asdescribedin the methodssection. -3OOpg/ml)(Fig.
IB) for both c-fos (Fig.lB) and c-myc (not shown) transcript induction. The results
confirm the earlier findings with LDL (9) and in addition demonstrate induction of the protooncogenes by HDL3. This study also demonstrates that combinations of the two lipoproteins do not have additive effects on cfos expression and depending on the HDL3:LDL ratio, c-fos transcript levels were lower than with either HDL3 or LDL alone (Fig. 2A and B). c-fos transcript induction was reduced by 50-605 when the I-ILIL3 concentration was at least equal to or greater than LDL. On the other hand, no significant reduction was
2:I
3:l
4:l
HDL : LDL
Fipure
-of
.
I:1 ratio
III hVSMC bv different
were sumutated for 30 mitts. with different amounts of LDL and HDL3 (l= 200 pdml) as indicated (Panel A). Panel B gives a gtaphic representation of the nonnahsed arbitrary optical density (OD) units of a representative Northern blot for each combination of HDL3 : LDL. Qualitatively similar results were obtained on two separateoccasions with two different isolates of cells. Quiescent
hVSMC
1467
Vol.
178,
No.
3,
1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
evident for those combinations in which the LDL concentration was greater than HDL3. It is unlikely that the effects observed on protooncogene expression with lipoproteins and their combinations are due to trace- contaminants sometimes contained in lipoprotein preparations like platelet- activating factor, platelet-derived growth factor or lipopolysaccharides. These factors, at reasonable concentrations (at least lo-lo M) have also been shown to induce early response genes, however, in an additive and/or synergistic fashion. Therefore, the reduction of c-fos expression by LDL/HDL combinations most likely represents a lipoprotein- specific effect. In addition, neutralising antibodies to platelet- derived growth factor and the platelet- activating factor- antagonist (Ro 19-3704, Hofmann LaRoche Ltd. Base], Switzerland) did not influence the lipoprotein- induced expression of the pmtooncogenes. Individual lipoprotein induction of c-fos might be assumed to reflect receptor- specific responses at different sites (or pathways), with activation of either one leading to stimulation of c-fos expression. However, the combination studies suggest that HDL3 and LDL might influence c-fos expression through a common site/pathway, but via mechanisms which are distinct and mutually exclusive in effect. Mutual inhibition of c-fos induction by HDL3 and LDL could explain why the stimulation of c-fos expression by some combinations of HDL3/LDL
was inferior to that by either lipoprotein alone. However. the
inhibitory influence of HDL3 on LDL is apparently greater than that of LDL on HDL3 since blunting of c-fos expression was maximal when the concentration of HDL3 was 2 LDL, but negligible when their concentration of LDL was > HDL3. HDL3 and LDL have previously been shown to exert mutually inhibitory effects in other cellular systems. In human platelets, HDL3 (which induced only shape change but no aggregation) inhibited LDL- induced aggregation (8) and LDL- induced elevations in [Ca2+]i (6) and displaced LDL from its binding site (19). Conversely, in hVSMC LDL counteracted the HDL3- induced stimulation
of
phosphoinositide metabolism (Dr. T.J.Resink, pers. commun.). Whether interference of HDL3 and LDL with the biophysical properties of the cell membrane (20) or other mechanisms, such as intracellular feedback control or competition at the receptor level, are involved in the inhibitory effects of the lipoproteins is a subject for further studies. In contrast to the reduced stimulation of c-fos expression, c-myc induction was not affected by the various HDL3 : LDL combinations (not shown). This difference may reflect different coupling of the signalling pathways for stimulation of c-fos and c-myc protooncogene expression. The oncoprotein c-myc is known to co- regulate the cell cycle in an Al’-1 independent manner at late Gl/early S-phase (2122 ). Since c-myc expression induced by the combination of HDL3 plus LDL was not inferior to that induced
1468
Vol.
178,
No
3,1991
BIOCHEMICAL
mc Cl.IIB I
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
B 81
MHC class I Figure 3. m of TGFB -ion and c-fos mRNA e-o s of HDL and Quiescent hVSMC were. stimulatedwith TGFE (5 &ml) either atone or in comkition wii HDL3 (6OOpg/mt)and/or LDL (200 p@ml). The MHC specific hybridization servesas the internal control for RNA variabilities on the Northern blots. Exposure times were as indicated for tigure 1 with the exceptionof the experiments with HDW plus LDL, where exposure was 36 hrs for the TGFB specific hybridisation and 16 hrs for MHC . Panel A shows representative Northern blots; in Panel B the normalized (for MHC- and exposure time) optical densities (OD) for each individual stimulation experimentare depicted. [3= TGFB A= TGFB+HDL3 0 =TGFB+LDL .= TGF!Z+HDLj+LDL Panel C: c-fos transcript induction by TGFD either alone (5ng/ml) or in combination with LDL and/or HDL3 as describedfor panel A.
by singular lipoprotein
stimulation, interaction between the lipoproteins probably do not directly
influence the cycling of cells. The interactive effect of HDL3 and LDL on c-fos mRNA expression is of interest because the c-fos oncoprotein is involved in the activity of the AP-1 multiprotein complex which controls transcriptional
activity of a variety of genes (23.24). Thus, alteration of c-fos transcript
expression by lipoprotein
interaction may also lead to modulation of the activity of the AP-1
transcription complex. To investigate whether lipoproteins
can indeed influence AP-1 regulated genes, we studied the
autostimulation of TGFB, a process dependent on AP-1 activity (25,26) and inhibited by actinomycin D. a blocker of transcription. Simultaneous addition of HDL3 and LDL to hVSMC was also observed to affect the autostimulation of TGFB transcripts in these cells (Fig.3). Stimulation of quiescent hVSMC by Sng/ml of TGFB autoinduced TGFB mRNA levels with a maximum at 5-7 hrs poststimulation (Fig.3 A and B). In addition to a 2.5 Kb transcript (encoding TGFBI), a low level of a -3.5 Kb isoform (encoding TGFl32) was expressed. When TGFB was added to hVSMC in combination with either LDL or HDL3,
the autostimulatory
effect of TGFD was not influenced, as indicated by the unaltered
autostimulated TGFD mRNA levels. Analogous, the lipoprotein combination without
TGFD addition did
not result in expression of TGFE mRNA (not shown). However, exposure of hVSMC to TGFB together with 600 pg/ml HDL3 plus 200 pg/ml LDL resulted in a markedly blunted autostimulation of TGFB mRNA (Fig.3 A and B). The experiments presented in figure 3C demonstrate c-fos mRNA induction by TGFR either alone or in combination with the lipoproteins as described for the A and B panels of the 1469
Vol.
178, No.3,
1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
figure. TGFB alone, at the concentration of Sng/ml, caused a maximal stimulation of c-fos mRNA which was not influenced by either HDL3 or LDL alone. Jn hVSMC exposed to a simultaneous combination of LDL, HDL3 and TGFB, levels of c-fos transcripts were considerably reduced compared with those in hVSMC exposed either to TGFB alone or to a combination of TGFg and either one of the lipoproteins (Fig. 30 Given the associations between c-fos expression and AP-1 activity (23,24) and between AP-1 activity and TGFl3 autostimulation (25,26), our demonstration of inhibition of both TGFB- induced c-fos expression and TGFB autostimulation by the combination of HDL3 and LDL (Fig. 3) suggests that interactions between these lipoproteins
may indeed influence AP-1 activity. However, it is noteworthy that
transcription of genes which also involve AP-1 complexes e.g. of endothelin, was not affected by lipoproteins and their combinations (not shown). Therefore, the inhibition of TGFB autostimulation
is
not indicative of a “simple” relationship between AP-1 activity and transcription, but requires further investigation. The present findings have possible physiological and pathophysiological implications. In the arterial wall, lipoproteins may thus not only play the role of cholesterol- carriers but also act as co- mitogens and modulators of (autocrine) growth factors such as TGFO. The latter has been demonstrated to regulate VSMC differentiation (27). to modulate
VSMC synthesis of extracellular matrix components (28.29 ),
and to chemoattract neutrophils and monocyte/macrophages (30,31). The present results indicate that the balance
between lipoprotein fractions (e.g. HDL3 and LDL) may be important for maintaining integrity
of the arterial wall in conditions where endothelial cell layers are damaged and VSMC are exposed to serum lipoproteins. Pre-existing imbalances of lipoprotein ratios may then be involved in the progression of arteriosclerotic lesions e.g. extensive proliferation of VSMC and alterations of extracellular matrix (2,3). Interestingly, under physiological conditions the relation between HDL3 and LDL is of the order of 2: 1 (pg protein/ml) whereas a preponderance of LDL over HDL3 (i.e. a ratio below 1) is thought to be a risk factor for the development of arteriosclerosis.
Acknowledgments: We gratefully acknowledge the assistance given in the preparation of the manuscript to Drs. TJ Resink and H Langemann. This study was funded by the Swiss National Fund grant no. 31-29275.90
1) Cameo G, Ponce E, Lopez F, Starosta R, Hurt E, Roman0 M, (1983) Atherosclerosis 49: 241- 254 2) Steinberg D, Parthasarathys S, Carew TE, Khoo JC, Wibtum JL, (1989) New Engl.J.M& 320: 915 924 3) Haberland ME, Fong D, Cheng L, (1988) Science 241: 215-218 1470
Vol.
178,
No.
3, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
4) Block LH, Knorr M, Vogt E, Lecher R, Vetter W, Aroscurth P, Qiao BY, Pometta D, James R, Regenass M, Pletscher A, (1988) Proc. Natl. AcadSci. (USA) 85: 885 899 5) Andrews HE, Aitken JW, Hassall DG, Skinner OVG, Bruckdorfer KR. (1987) B&hem J. 242: 559- 564 6) Tkachuk V, Bochkov V, Voyno- Yasenetskaya TA, Avdonin PV, Bobik A, J.Biol.Chem (in press) 7) Knorr M, Lecher R, Vogt E, Vetter W, Block LH, Ferracin F, Lefkovits H, Pletscher A, (1988) Eur.J.Biochem 172: 753- 759 8) Pletscher A, Ferracin F, Block LH, (199 1) Thrombosis Res. 60: 43 l- 432 9) Scott-Burden T, Resink TJ, Hahn AWA, Baur U. Box RJ, Btihler FR, (1989) J.Biol.Chem 264: 12582- 12589 10) Ullrich A, Shine J, Chirgwin J, Pichet R, T&her E, Rutter WJ, Goodman HM, (1977) Science 196: 1313- 1315 11) Feinberg AP, Vogelstein B, (1983) Anal. B&hem. 132: 6- 13 12) Church GM, Gilbert W., (1984) Proc.Natt.Acad.Sci. 81: 1991- 1995 13) Pohla H, Kuon W, Tabaczewski P, DUmer C, Weiss HE, (1989) Immunogenetics 29: 297-307 14) Co&on BH, Zullo J, Verma ZM, Stiles CD, (1984) Science 226: 1080- 1082 15) Miyamoto R, Chizzonite R, Crow1 R, Rupprecht K, Kramer R, Schaber M, Kumar G, Poonian M, Ju G, (1985) Proc.Natl.Acad.Sci. 82: 7232- 7236 16) Derynck R, Jarret JA, Chen EY, Eaton DH, Bell Jr, Assoian RK, Geddel DV, (1985) Nature 316: 701- 705 17) Have1 RJ, Eder HA. Bradgen JH, (1955) J.Clin.Invest. 34: 1345- 1353 18) Lowry OH, Rosebruogh NJ, Farr AL, Randall RT, (1951) J.Biol.Chem 193: 265- 275 19) Koller E, (1986) FEBS Lett. 200: 97- 102 20) Epand RM, Lester DS, (1990) Trends Pharm. Sci. 11: 317- 320 21) Blanchard JM, Piechaczyk R, Dani C, (1985) Nature 317: 443- 445 22) Heikkila R, Schwab G. Wickstrom E, Loke SL, Pluznik DH, Watt R, Neckers LM, (1987) Nature 328: 445- 449 23) Rollins DJ, Stiles CD, (1988) In Virto Cell & Devel. Biol. 24: 81- 83 24) Curran T, Franza BR, (1988) Cell 55: 395 397 25) Obberghen- Schilling E, Roche NS, Flanders KC, Spom MB, Roberts AB, (1988) J.Biol.Chem 263: 7741- 7746 26) Seong-Jin K, Angel P, Lafyatis R, Hattori K, Kyung YK, Spom MB, Karin M, Roberts AB, (1990) Mol.Cell.Biol. 10: 1492- 1497 27) Goodman LV, Majack RA, (1989) J.Biol.Chem. 264: 5241- 52 44 28) Spom MB, Roberts AB, Wakefield LM, Assoian RK, (1986) Science 260: 532- 534 29) Spom MB, Roberts AB, Wakefield LM, Crombrugghe 8, (1987) J.Cell.Biol. 105: 1039- 1045 30) Gamble JR, Vadas MA, (1988) Science 242: 97- 99 31) Wahl SM, Hunt DA, Wakefield LM, McCartney- Francis N, Wahl LM, Roberts AS, Spom MB, (1987) Proc.Natl.Acad.Sci. 84: 5788- 5792
1471
178,
August
No. 15,
BIOCHEMICAL
3, 1991
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS Pages
1991
MODULATION
OF GENE EXPRESSION
LIPOPROTEINS
IN HUMAN
VASCULAR
BY HIGH SMOOTH
1465-1471
AND LOW DENSITY MUSCLE
CELLS
Alfred W.A. Hahn*, Fabrizia Ferracin, Fritz R. Btihler and Alfred Pletscher Department of Research, Base1 University Hospital, Hebelstr. 20, CH- 4031 Basel, Switzerland Received
July
8, 1991
Summary Low density lipoprotein and its oxidized form has been implicated in the process of arteriosclerosis which involves growth- related events in the smooth muscle cells of the arterial wall. The induction of so- called early- growth response genes e.g. c-myc and c-fos can serve as an indicator for these growth- related events. In cultured human vascular smooth muscle cells, both LDL and HDL3 were individually capable of stimulating c-myc and c-fos expression in a concentration dependent manner. However, when they were used in combination, depending on the proportion of HDL3 to LDL, c-fos but not c-myc expression was less pronounced than with the single components. In contrast to HDL3 and LDL alone, a combination of the two lipoproteins also blunted both the expression of autoinduced transforming growth factor 6 transcripts and the transforming growth factor l3- induced increase of c-fos mRNA. It is concluded that a) the inhibition of transforming growth factor 8 autostimulation by HDL3 plus LDL may involve reduced AP-1 activity via a reduction of c-fos expression by the lipoprotein combination and b) the ratio HDL3:LDL might influence the pathogenesis of arteriosclerosis via growth- related events in the arterial wall. 0 1991 *cademic Press, Inc.
B
Plasma lipoproteins are thought to play a role in the pathogenesis of arteriosclerosis (1 -
3). In fact, although elevated concentrations of LDL seem to be a risk factor, HDL appear to exert a protective effect in this condition. LDL has been shown to cause intracellular activation (e.g. stimulation of phosphoinositide metabolism, elevation of [Ca2+]i and enhanced phosphorylation of cellular proteins) in various cell types including vascular smooth muscle, endothelial cells and blood platelets (4.5.6). In addition LDL causes a shape change and has a proaggregatory effect in platelets (578). The action of HDL at the cellular level is less well established. However, HDL3 has been shown to cause a shape change reaction, to antagonise the LDL- induced rise of [Ca2+]i and to counteract
the LDL- mediated
aggregation of human platelets (6.8). HDL3 alone neither caused an increase of intracellular calcium levels, nor induced the aggregation of platelets (6.8) indicating an important interaction between the different lipoproteins.
. . Abbrevratlons: human vascular smooth muscle cells, hVSMC; low- and high density lipoproteins, LDL and HDL, transforming growth factor 6, TGFB. 0006-291X/91 1465
$1.50
Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol.
178,
No.
3,
1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
LDL has also been shown to induce early growth- related events, i.e. transcript expression of c-fos and cmyc in VSMC (9). Such events may be important in the arteriosclerotic process, one characteristic of which is the proliferation of vessel wall constituents. Therefore it is of interest to investigate whether there might also be an interaction between HDL and LDL with respect to these growth- regulating events in cultured hVSMC.
Cell culture and stimulation experimenti Isolation, culture and phenotypic characterization of hVSMC were performed essentially as described before (9). Two different isolates of microarterial hVSMC from different patients undergoing abdominal surgery were used in this study. Cells were used between passage 4- 10. Prior to experimentation, confluent hVSMC were rendered quiescent by culture for 48 hrs (with one medium change after 24 hrs) under serum- free conditions and placed in medium containing 1% plasma- derived serum (mitogen- free) for 12 hrs. Subsequently cells were stimulated with agonist for the times indicated and processed for isolation of total RNA. Northern BJQQQ&& Following stimulation, cultured cells were washed twice with phosphate buffered saline (PBS) and then lysed by the addition of guanidinium isothiocyanate buffer (GT buffer) directly onto the culture dishes. Total RNA was collected on CsCl2- gradients as described before (10). For Northern analysis, 20 pg of total RNA was size- separated on a 1.2% agarose gel containing 2.2 M formaldehyde at a constant voltaae of 5OV for 6-8 hrs in MOPS buffer. The electronhoresed ael was vacublotted onto HYbond N memb;anes (Amersham) with 20x SSC as transfer buffer. Blotted RNA was fixed to the membranes by UV irradiation at 302 nm for 3 mins. Hybridization of blots to random primed cDNA probes (11) was performed according to standard methods (12). The blots were washed at high stringency (0.1x SSC/ 0.1% SDS at 65” C) and exposed to Kodak X-Omat films at -70 C overnight using one intensifying screen. Variabilities of RNA present on the blots were evaluated by tehybridization to a 1.5 kb cDNA probe specific for MHC class I antigens (clone pMF 48) (13) and exposed for 4-6 hrs to X-ray film. Probes used in tbis study were cDNA probes for c-fos (14) and c-myc (15) as well as a 1.2 kb cDNA for TGFB (16). For rehybridizations of blots with different probes, membranes were stripped of their previous signal by incubation for 10 mins. in boiling 10 mM Tris pH 7.5/ 10 mM EDTA. Prior to rehybridization, blots were exposed to Kodak films overnight to ensure complete removal of the previous signals. Signals were densitometrically analyzed by scanning (525 nm) of autoradiographs. The arbitrary optical density (OD) units given in the results were obtained after normalization of signals with respect to the OD values obtained for the MHC class I internal control (highest value within a given series was arbitrarily taken as 100 %). J&.mrotein isolation: LDL (density 1.019 - 1.063) and HDL3 (density 1.125- 1.21) were isolated from the plasma of two different healthy volunteers by ultracentrifugation as described by Have1 et al (17). The lipoprotein fractions were dialysed against isotonic phosphate buffer pH 7.4 containing 1 mM EDTA and 1pM BHT (2.6 di-tert-butyl-p-cresol: Fluka Chemie AG, Buchs, Switzerland) and filtered (0.45 pm). HDL3 was used in the study in preference to total I-IDL. to avoid any possible trace contamination of HDL fractions with LDL. Both, LDL and HDL3 preparations were checked for purity by gel elctrophoresis and HPLC. The protein content was measured (18) and aliquots were stored at 4°C under an atmosphere of nitrogen for not more than a few weeks. During the storage period no change in the activity of the lipoprotein fractions was observed. m
.
.
Both LDL and HDL3 were found to induce c-fos and c-myc transcripts in
quiescent hVSMC (Fig. 1). The stimulation kinetics for c-fos were rapid and transient (maximal at 30 mins. poststimulatory, return to control levels within 1 hr). Relative to c-fos, stimulation of c-myc was also transient but delayed with maximal induction occuring only -3-5 hrs poststimulation (Fig. !A) . The stimulatory effects of the lipoproteins on protooncogene expression were concentration dependent. LDL and HDL3 exhibited similar potencies and maximal effects ( EC50 -100 l&ml; 1466
saturation at
BIOCHEMICAL
Vol. 178, No. 3, 1991
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
hrs 0 .5 3 5 500
0
1000
concentration
1500
(pghl)
. Don in hVSMC -3u. hVSMC were stimulated by either LDL or HDL3 (each at 200 pg/ml) for the times indicated. Northern Mats hybridized to cDNA probes for the protconcogenes were exposed overnight at Fieure
1. &m&ion Panel A: Quiescent
of c-mvc and c-fos mRNA
70°C with one intensifying screen: blots hybridised to the MHC class I probe were exposed under identical conditions for 4-6 hrs only. Comparable results were obtained with two different isolates of cells. Panel B displays a representative concentration response curve for c-fos expression promoted by
different amountsof LDL (open circles) or HDL3 (full circles). The normalized arbitrary optical density (OD) units were obtained asdescribedin the methodssection. -3OOpg/ml)(Fig.
IB) for both c-fos (Fig.lB) and c-myc (not shown) transcript induction. The results
confirm the earlier findings with LDL (9) and in addition demonstrate induction of the protooncogenes by HDL3. This study also demonstrates that combinations of the two lipoproteins do not have additive effects on cfos expression and depending on the HDL3:LDL ratio, c-fos transcript levels were lower than with either HDL3 or LDL alone (Fig. 2A and B). c-fos transcript induction was reduced by 50-605 when the I-ILIL3 concentration was at least equal to or greater than LDL. On the other hand, no significant reduction was
2:I
3:l
4:l
HDL : LDL
Fipure
-of
.
I:1 ratio
III hVSMC bv different
were sumutated for 30 mitts. with different amounts of LDL and HDL3 (l= 200 pdml) as indicated (Panel A). Panel B gives a gtaphic representation of the nonnahsed arbitrary optical density (OD) units of a representative Northern blot for each combination of HDL3 : LDL. Qualitatively similar results were obtained on two separateoccasions with two different isolates of cells. Quiescent
hVSMC
1467
Vol.
178,
No.
3,
1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
evident for those combinations in which the LDL concentration was greater than HDL3. It is unlikely that the effects observed on protooncogene expression with lipoproteins and their combinations are due to trace- contaminants sometimes contained in lipoprotein preparations like platelet- activating factor, platelet-derived growth factor or lipopolysaccharides. These factors, at reasonable concentrations (at least lo-lo M) have also been shown to induce early response genes, however, in an additive and/or synergistic fashion. Therefore, the reduction of c-fos expression by LDL/HDL combinations most likely represents a lipoprotein- specific effect. In addition, neutralising antibodies to platelet- derived growth factor and the platelet- activating factor- antagonist (Ro 19-3704, Hofmann LaRoche Ltd. Base], Switzerland) did not influence the lipoprotein- induced expression of the pmtooncogenes. Individual lipoprotein induction of c-fos might be assumed to reflect receptor- specific responses at different sites (or pathways), with activation of either one leading to stimulation of c-fos expression. However, the combination studies suggest that HDL3 and LDL might influence c-fos expression through a common site/pathway, but via mechanisms which are distinct and mutually exclusive in effect. Mutual inhibition of c-fos induction by HDL3 and LDL could explain why the stimulation of c-fos expression by some combinations of HDL3/LDL
was inferior to that by either lipoprotein alone. However. the
inhibitory influence of HDL3 on LDL is apparently greater than that of LDL on HDL3 since blunting of c-fos expression was maximal when the concentration of HDL3 was 2 LDL, but negligible when their concentration of LDL was > HDL3. HDL3 and LDL have previously been shown to exert mutually inhibitory effects in other cellular systems. In human platelets, HDL3 (which induced only shape change but no aggregation) inhibited LDL- induced aggregation (8) and LDL- induced elevations in [Ca2+]i (6) and displaced LDL from its binding site (19). Conversely, in hVSMC LDL counteracted the HDL3- induced stimulation
of
phosphoinositide metabolism (Dr. T.J.Resink, pers. commun.). Whether interference of HDL3 and LDL with the biophysical properties of the cell membrane (20) or other mechanisms, such as intracellular feedback control or competition at the receptor level, are involved in the inhibitory effects of the lipoproteins is a subject for further studies. In contrast to the reduced stimulation of c-fos expression, c-myc induction was not affected by the various HDL3 : LDL combinations (not shown). This difference may reflect different coupling of the signalling pathways for stimulation of c-fos and c-myc protooncogene expression. The oncoprotein c-myc is known to co- regulate the cell cycle in an Al’-1 independent manner at late Gl/early S-phase (2122 ). Since c-myc expression induced by the combination of HDL3 plus LDL was not inferior to that induced
1468
Vol.
178,
No
3,1991
BIOCHEMICAL
mc Cl.IIB I
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
B 81
MHC class I Figure 3. m of TGFB -ion and c-fos mRNA e-o s of HDL and Quiescent hVSMC were. stimulatedwith TGFE (5 &ml) either atone or in comkition wii HDL3 (6OOpg/mt)and/or LDL (200 p@ml). The MHC specific hybridization servesas the internal control for RNA variabilities on the Northern blots. Exposure times were as indicated for tigure 1 with the exceptionof the experiments with HDW plus LDL, where exposure was 36 hrs for the TGFB specific hybridisation and 16 hrs for MHC . Panel A shows representative Northern blots; in Panel B the normalized (for MHC- and exposure time) optical densities (OD) for each individual stimulation experimentare depicted. [3= TGFB A= TGFB+HDL3 0 =TGFB+LDL .= TGF!Z+HDLj+LDL Panel C: c-fos transcript induction by TGFD either alone (5ng/ml) or in combination with LDL and/or HDL3 as describedfor panel A.
by singular lipoprotein
stimulation, interaction between the lipoproteins probably do not directly
influence the cycling of cells. The interactive effect of HDL3 and LDL on c-fos mRNA expression is of interest because the c-fos oncoprotein is involved in the activity of the AP-1 multiprotein complex which controls transcriptional
activity of a variety of genes (23.24). Thus, alteration of c-fos transcript
expression by lipoprotein
interaction may also lead to modulation of the activity of the AP-1
transcription complex. To investigate whether lipoproteins
can indeed influence AP-1 regulated genes, we studied the
autostimulation of TGFB, a process dependent on AP-1 activity (25,26) and inhibited by actinomycin D. a blocker of transcription. Simultaneous addition of HDL3 and LDL to hVSMC was also observed to affect the autostimulation of TGFB transcripts in these cells (Fig.3). Stimulation of quiescent hVSMC by Sng/ml of TGFB autoinduced TGFB mRNA levels with a maximum at 5-7 hrs poststimulation (Fig.3 A and B). In addition to a 2.5 Kb transcript (encoding TGFBI), a low level of a -3.5 Kb isoform (encoding TGFl32) was expressed. When TGFB was added to hVSMC in combination with either LDL or HDL3,
the autostimulatory
effect of TGFD was not influenced, as indicated by the unaltered
autostimulated TGFD mRNA levels. Analogous, the lipoprotein combination without
TGFD addition did
not result in expression of TGFE mRNA (not shown). However, exposure of hVSMC to TGFB together with 600 pg/ml HDL3 plus 200 pg/ml LDL resulted in a markedly blunted autostimulation of TGFB mRNA (Fig.3 A and B). The experiments presented in figure 3C demonstrate c-fos mRNA induction by TGFR either alone or in combination with the lipoproteins as described for the A and B panels of the 1469
Vol.
178, No.3,
1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
figure. TGFB alone, at the concentration of Sng/ml, caused a maximal stimulation of c-fos mRNA which was not influenced by either HDL3 or LDL alone. Jn hVSMC exposed to a simultaneous combination of LDL, HDL3 and TGFB, levels of c-fos transcripts were considerably reduced compared with those in hVSMC exposed either to TGFB alone or to a combination of TGFg and either one of the lipoproteins (Fig. 30 Given the associations between c-fos expression and AP-1 activity (23,24) and between AP-1 activity and TGFl3 autostimulation (25,26), our demonstration of inhibition of both TGFB- induced c-fos expression and TGFB autostimulation by the combination of HDL3 and LDL (Fig. 3) suggests that interactions between these lipoproteins
may indeed influence AP-1 activity. However, it is noteworthy that
transcription of genes which also involve AP-1 complexes e.g. of endothelin, was not affected by lipoproteins and their combinations (not shown). Therefore, the inhibition of TGFB autostimulation
is
not indicative of a “simple” relationship between AP-1 activity and transcription, but requires further investigation. The present findings have possible physiological and pathophysiological implications. In the arterial wall, lipoproteins may thus not only play the role of cholesterol- carriers but also act as co- mitogens and modulators of (autocrine) growth factors such as TGFO. The latter has been demonstrated to regulate VSMC differentiation (27). to modulate
VSMC synthesis of extracellular matrix components (28.29 ),
and to chemoattract neutrophils and monocyte/macrophages (30,31). The present results indicate that the balance
between lipoprotein fractions (e.g. HDL3 and LDL) may be important for maintaining integrity
of the arterial wall in conditions where endothelial cell layers are damaged and VSMC are exposed to serum lipoproteins. Pre-existing imbalances of lipoprotein ratios may then be involved in the progression of arteriosclerotic lesions e.g. extensive proliferation of VSMC and alterations of extracellular matrix (2,3). Interestingly, under physiological conditions the relation between HDL3 and LDL is of the order of 2: 1 (pg protein/ml) whereas a preponderance of LDL over HDL3 (i.e. a ratio below 1) is thought to be a risk factor for the development of arteriosclerosis.
Acknowledgments: We gratefully acknowledge the assistance given in the preparation of the manuscript to Drs. TJ Resink and H Langemann. This study was funded by the Swiss National Fund grant no. 31-29275.90
1) Cameo G, Ponce E, Lopez F, Starosta R, Hurt E, Roman0 M, (1983) Atherosclerosis 49: 241- 254 2) Steinberg D, Parthasarathys S, Carew TE, Khoo JC, Wibtum JL, (1989) New Engl.J.M& 320: 915 924 3) Haberland ME, Fong D, Cheng L, (1988) Science 241: 215-218 1470
Vol.
178,
No.
3, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
4) Block LH, Knorr M, Vogt E, Lecher R, Vetter W, Aroscurth P, Qiao BY, Pometta D, James R, Regenass M, Pletscher A, (1988) Proc. Natl. AcadSci. (USA) 85: 885 899 5) Andrews HE, Aitken JW, Hassall DG, Skinner OVG, Bruckdorfer KR. (1987) B&hem J. 242: 559- 564 6) Tkachuk V, Bochkov V, Voyno- Yasenetskaya TA, Avdonin PV, Bobik A, J.Biol.Chem (in press) 7) Knorr M, Lecher R, Vogt E, Vetter W, Block LH, Ferracin F, Lefkovits H, Pletscher A, (1988) Eur.J.Biochem 172: 753- 759 8) Pletscher A, Ferracin F, Block LH, (199 1) Thrombosis Res. 60: 43 l- 432 9) Scott-Burden T, Resink TJ, Hahn AWA, Baur U. Box RJ, Btihler FR, (1989) J.Biol.Chem 264: 12582- 12589 10) Ullrich A, Shine J, Chirgwin J, Pichet R, T&her E, Rutter WJ, Goodman HM, (1977) Science 196: 1313- 1315 11) Feinberg AP, Vogelstein B, (1983) Anal. B&hem. 132: 6- 13 12) Church GM, Gilbert W., (1984) Proc.Natt.Acad.Sci. 81: 1991- 1995 13) Pohla H, Kuon W, Tabaczewski P, DUmer C, Weiss HE, (1989) Immunogenetics 29: 297-307 14) Co&on BH, Zullo J, Verma ZM, Stiles CD, (1984) Science 226: 1080- 1082 15) Miyamoto R, Chizzonite R, Crow1 R, Rupprecht K, Kramer R, Schaber M, Kumar G, Poonian M, Ju G, (1985) Proc.Natl.Acad.Sci. 82: 7232- 7236 16) Derynck R, Jarret JA, Chen EY, Eaton DH, Bell Jr, Assoian RK, Geddel DV, (1985) Nature 316: 701- 705 17) Have1 RJ, Eder HA. Bradgen JH, (1955) J.Clin.Invest. 34: 1345- 1353 18) Lowry OH, Rosebruogh NJ, Farr AL, Randall RT, (1951) J.Biol.Chem 193: 265- 275 19) Koller E, (1986) FEBS Lett. 200: 97- 102 20) Epand RM, Lester DS, (1990) Trends Pharm. Sci. 11: 317- 320 21) Blanchard JM, Piechaczyk R, Dani C, (1985) Nature 317: 443- 445 22) Heikkila R, Schwab G. Wickstrom E, Loke SL, Pluznik DH, Watt R, Neckers LM, (1987) Nature 328: 445- 449 23) Rollins DJ, Stiles CD, (1988) In Virto Cell & Devel. Biol. 24: 81- 83 24) Curran T, Franza BR, (1988) Cell 55: 395 397 25) Obberghen- Schilling E, Roche NS, Flanders KC, Spom MB, Roberts AB, (1988) J.Biol.Chem 263: 7741- 7746 26) Seong-Jin K, Angel P, Lafyatis R, Hattori K, Kyung YK, Spom MB, Karin M, Roberts AB, (1990) Mol.Cell.Biol. 10: 1492- 1497 27) Goodman LV, Majack RA, (1989) J.Biol.Chem. 264: 5241- 52 44 28) Spom MB, Roberts AB, Wakefield LM, Assoian RK, (1986) Science 260: 532- 534 29) Spom MB, Roberts AB, Wakefield LM, Crombrugghe 8, (1987) J.Cell.Biol. 105: 1039- 1045 30) Gamble JR, Vadas MA, (1988) Science 242: 97- 99 31) Wahl SM, Hunt DA, Wakefield LM, McCartney- Francis N, Wahl LM, Roberts AS, Spom MB, (1987) Proc.Natl.Acad.Sci. 84: 5788- 5792
1471
