Metabolism of Atherogenic Lipoproteinsby Smooth Muscle Cells of Different

Phenotype in CultureJulie H. Campbell, Michael F. Reardon, Gordon R. Campbell, and Paul J. Nestel

Smooth muscle cells of the rabbit aorta, when grown in vitro, express three distin-guishable forms of phenotype (contractile, reversible synthetic, and irreversible syn-thetic). We compared the interactions of these three smooth muscle phenotypes withrabbit very low density lipoprotein (VLDL), low density lipoprotein (LDL), and very lowdensity lipoprotein from cholesterol-fed rabbits (13-VLDL). 13-VLDL showed saturable,high-affinity binding characteristics with each phenotype predominantly through theB/E receptor. The irreversible synthetic cells displayed the greatest binding capacityand the contractile cells, the least. Binding and degradation of normal VLDL was lessthan that of R-VLDL and higher than that of LDL. Only the irreversible synthetic cellsshowed substantial (about threefold) cholesteryl ester formation and cholesterol ac-cumulation, and then only when incubated with 13-VLDL. Substantial stainable lipid,shown chemically to include triglyceride, cholesterol and cholesteryl ester, was alsoobserved only when irreversible synthetic cells were exposed to 13-VLDL. The highcapacity of irreversible synthetic-state, smooth muscle cells to bind and accumulate13-VLDL in contrast to the relative immunity of contractile cells may be relevant to thegenesis of atherosclerosis in the rabbit and possibly also in humans.(Arteriosclerosis 5:318-328, July/August 1985)

In humans, an early stage of atherosclerosis ischaracterized by focal cellular thickenings within

existing diffuse intimal thickenings.1 The cells of thefocal thickenings are predominantly smooth musclecells, but the phenotypic expression of many of thesecells differs from that of medial smooth musclecells.2'3 This has led to their description as "modifiedsmooth muscle," "poorly differentiated smooth mus-cle-like cells," and "fibroblast-like." Morphologically,

From the Baker Medical Research Institute, Commercial Road,Prahran Victoria, and the Department of Anatomy, University ofMelbourne, Parkville Victoria, Australia.

This research was supported by the Life Insurance MedicalResearch Fund of Australia and New Zealand, the National HeartFoundation of Australia, and the National Health and MedicalResearch Council of Australia.

This work was presented in part at the American Heart Associ-ation Meeting, Anaheim, California, November 1983. (ReardonMF, Campbell JH, Nestel PJ, Campbell GR. Metabolism of lipo-proteins by rabbit contractile and synthetic state arterial smoothmuscle cells in culture. Arteriosclerosis 1983;5:504a.)

Michael F. Reardon died on January 8, 1985.Address for reprints: Dr. Julie H. Campbell, Baker Medical Re-

search Institute, Commercial Road, Prahran Victoria 3181, Aus-tralia.

Received October 29,1984; revision accepted March 25,1985.

these cells contain few myofilament bundles and in-creased amounts of organelles involved with synthe-sis such as rough endoplasmic reticulum, free ribo-somes, and Golgi. They closely resemble smoothmuscle cells in development and wound repair.4

Recent studies have shown a differential responseto atherogenic stimuli by smooth muscle cells ac-cording to the phenotypic state of the cells.5"9 Muchof this evidence has come from studies involvingprimary culture of vascular smooth muscle in whichchanges in phenotypic expression can resemble thecells of focal thickenings.56 Cells that have beenenzyme-dispersed from the adult pig or rabbit aorticmedia demonstrate this phenotypic change after 6 to8 days provided they are seeded at less than 106

cells/ml nutrient medium.6 In the first few days ofprimary culture while they are in the "contractile"state closely resembling the cells of the intact aorticmedia, the cells do not incorporate significantamounts of 3H-thymidine into DNA or proliferate inresponse to normolipemic or hyperlipemic whole-blood serum.7 Following phenotypic change to the"synthetic" state, the cells undergo logarithmicgrowth in response to whole blood serum. The cellswill return to the contractile state upon confluency if

318

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from

LIPOPROTEIN METABOLISM OF SMOOTH MUSCLE Campbell et al. 319

they have undergone fewer than five cell doublingsand have remained in the synthetic state for lessthan 3 weeks ("reversible synthetic" state).6 Howev-er, cells that have been subcultured several timesand have undergone multiple cell doublings appearunable to return to the contractile state and are thusin an "irreversible synthetic" state. These cells even-tually exhaust their proliferative capacity and be-come senescent.

In previous reports, we showed that a change inphenotype from the contractile to the reversible syn-thetic state by primary cultures of pig and rabbit aor-tic smooth muscle cells was accompanied by a de-crease in their ability to degrade 125l-human lowdensity lipoprotein (LDL).8'9 Morphologically, revers-ible synthetic-state cells grown in 10% hypercholes-terolemic rabbit serum for 4 days accumulated lipiddroplets, while contractile-state cells under identicalconditions were relatively unaffected. This was re-flected in an increased incorporation of 3H-sodiumoleate into cholesteryl esters in synthetic-state cells.

Since the major lipoprotein of cholesterol-fed rab-bits is a cholesteryl ester-rich, very low density lipo-protein or 13-VLDL,10 we have extended these studiesto comparisons of B-VLDL metabolism by the threephenotypes of rabbit arterial smooth muscle cells.The uptake of B-VLDL appears to be mediatedthrough the interaction of apoprotein E (a major pro-tein of B-VLDL) with cellular receptors that specifical-ly recognize B-VLDL or apoprotein E-rich lipoproteinin general.11 The former has been clearly defined inmacrophages,12 but not in smooth muscle cells. Thesecond receptor, the B/E receptor, interacts specifi-cally also with LDL and VLDL although some cata-bolic changes in VLDL structure have reportedly fa-cilitated specific binding.13'u The capacity of B-VLDLto bind to smooth muscle cells and to induce accu-mulation of lipid has therefore been compared withthat of LDL and of VLDL from normal rabbits.

MethodsMaterials

Sodium 12Sl-iodide (IMS 30) and [9,10(n)-3H]-oleicacid (5.7 Ci/mmol) were from the RadiochemicalCentre, Amersham. Collagenase 4196 CLS wasfrom Worthingion Chemical Company* and elastaseE-0127 was from Sigma Chemical Company.

Cell Culture

For each experiment, the thoracic and abdominalaortic media and intima from seven 9-week-old NewZealand white rabbits were dispersed into singlecells by incubating in 5 ml of 3 mg/ml collagenase for1 hour (discarded), then 5 ml of 1 mg/ml elastase for1.5 hours followed by the addition of 5 ml of 3 mg/mlcollagenase for another hour. All of the tissue dis-persed in the elastase-collagenase mixture, with ayield of 6-7 x 107 viable cells from seven rabbits.

The cells were seeded at 3 x 105 cells/ml into 60 mmor 90 mm plastic culture dishes (Sterilin). The platingefficiency was greater than 80%, and a near-conflu-ent cell layer was produced. For the study of "con-tractile" state cells, the primary cultured cells weretaken on Days 3 to 4. To obtain "reversible synthetic"cells, unused cultures from the previous week weresubcultured and taken on Days 3 to 4 of the firstpassage (i.e., after 10 to 11 days in culture). Suchcells have undergone no more than two cell dou-blings. "Irreversible synthetic" cells were in theirfourth or fifth passage and had been in culture for atotal of 4 to 6 weeks and had undergone six to eightcell doublings. They were used 3 to 4 days after theirfinal subculture.

Rabbit aortic smooth muscle cells have a high re-quirement for glutamine to proliferate. Thus, toachieve two cell doublings (reversible synthetic) orsix to eight cell doublings (irreversible synthetic), thecells were grown in the presence of Medium 199 +5% fetal calf serum (FCS) + 0.2 g/liter glutamine. Tominimize differences in the number of receptors dueto proliferation, however, reversible and irreversiblecells in their final subculture (i.e., at least 3 daysbefore each experiment) were incubated in Medium199 + 5% FCS to which no additional glutamine hadbeen added. Cells in primary culture were also seed-ed in Medium 199 + 5% FCS.

Preparation of Serum and Llpoprotelns

Whole-blood serum from normolipemic (150 mg/dlto 200 mg/dl cholesterol) and hyperlipemic (900mg/dl to 1300 mg/dl cholesterol) rabbits was pre-pared sterile for cell culture. Lipoprotein-deficient se-rum (LDS) from normal rabbits was prepared by ad-justing the salt density to 1.21 g/ml and byultracentrifuging for 48 hours at 104,000 g. The in-franate was dialyzed for 4 days to the osmolality ofplasma.

B-VLDL was obtained from the plasma of rabbitsmade hypercholesterolemic by a diet supplementedwith 2% cholesterol as egg yolks. Plasma was over-laid with an equal volume of 0.15 M NaCI and 0.001M EDTA (pH 7.4) and ultracentrifuged at 110,000 g,at 14° C for 16 hours in a Beckman 50 Ti rotor. The d< 1.006 g/ml supernatant fraction that containedcholesteryl ester-rich VLDL was washed once at thisdensity. Pevikon block electrophoresis15 of this ma-terial showed that more than 90% migrated as Glipoprotein. Normal VLDL was isolated from the plas-ma of normal fasting rabbits by ultracentrifugationunder conditions described for the isolation of B-VLDL. LDL was isolated as a d = 1.025 to 1.055 g/mldensity fraction by a 24-hour ultracentrifugation at104,000 g. Lipoproteins were radioiodinated byusing a modification of the iodine monochloridemethod.16 The ratio of moles of iodine bound permole of protein was routinely kept at less than 1. Thedistribution of radioactivity showed minor dissimilari-ties with B-VLDL and normal VLDL, which reflected

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from

320 ARTERIOSCLEROSIS VOL 5, No 4, JULY/AUGUST 1985

the differences in apolipoprotein composition. Morethan 92% of the radioactivity was trichloracetic acid-precipitable with both preparations. Between 12%and 18% of the radioactivity was extractable withchloroform/methanol (lipid-associated) in the variouspreparations. About 45% of normal VLDL and 58%of (3-VLDL radioactivity were precipitable with isopro-panol and also comigrated with apo B on polyacryla-mide gels. The main difference was in the amountsof label in apo E: about 20% in (3-VLDL and less than10% in VLDL. Conversely, twice as much label mi-grated with apo C in VLDL than in I3-VLDL

Llpoprotein Binding and Degradation

Cultures were preincubated for 17 hours at 37° Cin 3 ml Medium 199 + 5% LDS. At zero time, thismedium was removed and replaced with the givenlipoprotein in Medium 199 + 5% LDS to a final vol-ume of 3 ml. Binding studies were carried out for 3hours at 4° C, and degradation studies were per-formed for 5 hours at 37° C. Parallel studies werecarried out in the presence and absence of tenfoldexcess unlabeled lipoprotein. Additional incubationsof lipoproteins without cells were used to correct fornon-cell-mediated binding and degradation. In someexperiments, the metabolism of increasing amountsof various radioiodinated lipoproteins was mea-sured, whereas in other studies a constant amount(5 /xg/ml) of 125I-B-VLDL was incubated with increas-ing quantities of other unlabeled lipoproteins. Forone study, chloroquine at 25 /xM, 50 /xM, 75/n.M, and100 yM was included in the incubation. Binding ofradioiodinated 125l-lipoproteins was assayed by mea-suring cell-associated radioactivity after the mediumhad been removed and the cells had been washedtwice with 0.2% bovine serum albumen in normalsaline and then four times in Dulbecco's balancedsalt solution. The degradation of lipoproteins wascalculated from the radioactivity in the noniodidetrichloracetic acid-soluble fraction.17

For each lipoprotein concentration, both with cellsand with no-cell controls and in the presence or ab-sence of tenfold unlabeled excess lipoprotein, tripli-cate or quadruplicate dishes were used. To deter-mine the number of cells per dish in the bindingstudies, six dishes of each cell phenotype were treat-ed as for experimental cells then trypsinized andcounted in a ZM coulter counter. For degradationstudies, the cells of each experimental dish werecounted after the degradation products had beenremoved.

Incorporation of 3H-Oleate Into Cholesteryl Esterand Trlglycerlde

[9,10(n)-3H]-oleic acid in hexane was evaporatedto dryness and complexed to form sodium oleate in asolution containing 12% bovine albumin in 0.9%NaCI. The final concentration of Na oleate in theexperimental media was 0.2 mM. Cells grown in 60mm dishes were incubated for 17 hours in 3 ml of

Medium 199 + 5% LDS. The medium was replacedwith 3 ml of fresh medium containing unlabeled lipo-protein and 50 /JL\ of 3H-Na oleate-albumin (500,000cpm per 50 /ul), and the cells were incubated at 37° Cfor 5 hours. The cells were washed and harvested.The cellular lipids were extracted with chloroform/methanol (2:1 vol/vol) that contained 14C-cholesterolas an internal recovery standard and cholesteryl pal-mitate as a carrier. The lipids were separated by thin-layer silica gel chromatography (solvent system,hexane/diethyl ether/acetic acid, 150:50:2) and wereassayed for radioactivity. In each experiment, thevalue for 3H-cholesteryl ester and 3H-triglyceride rep-resented the mean of duplicate incubations that usedquadruplicate dishes for each assay.

Measurement of Cholesterol and TrlglycerldeMass

Cells were preincubated for 17 hours in 5% LDSand then for a further 24 hours with unlabeled lipo-protein (mostly 75 M9 protein per ml) in 5% LDS. Thecells were washed and the lipids were extracted withchloroform/methanol (2:1) that contained a mea-sured internal standard, 5a-cholestane). The lipidwas dried, saponified at 72° C for 2 hours with ethan-olic NaOH, and re-extracted in petroleum ether. Thetotal cholesterol mass was quantified as the trimeth-ylsilyl derivative by gas chromatography. A silicacapillary column with OV101 as the liquid phase wasused; standards of cholesterol and 5a-cholestanegave 1:1 ratios; recovery of cholesterol standardswas greater than 95%.

For measurements of cholesteryl esters, the chlor-oform/methanol lipid extract was initially separatedby thin-layer chromatography using 14C-cholesteryloleate as a recovery standard. The isolated choles-teryl ester was saponified and quantified by gaschromatography as free cholesterol.

Cellular triglyceride was measured enzymatically(Boehringer Mannheim).

Fettrot Staining of Cellular Upld

Cultures were grown on coverslips and incubatedfor 17 hours in 5% LDS and then for a further 24hours with 75 /xg/ml of unlabeled lipoprotein in 5%LDS or in 5% LDS alone as control. The cells werefixed in 10% buffered formalin and stained with Fet-trot solution, which renders lipid droplets red, andcounterstained with Meyer's Haemalum (modifiedafter Pease18). The degree of lipid accumulation wasassessed subjectively by light microscopy.

ResultsCell Size and Protein

With phase-contrast microscopy, cells in the irre-versible synthetic state appear considerably largerthan reversible synthetic-state cells, which in turnappear larger than contractile-state cells. This is due

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from

LIPOPROTEIN METABOLISM OF SMOOTH MUSCLE Campbell et al. 321

in part to the progressive spreading of the cells onthe culture substrate. However, trypsinizing the cellsrenders them spherical and their volume can be esti-mated by means of a calibrated ZM coulter countergiving the values ± standard deviation of 211 ± 32/xm3 (contractile), 380 ± 46 /xm3 (reversible synthet-ic), and 857 ± 68 /xm3 (irreversible synthetic). Thisincrease in size is reflected in increases in proteinper cell: 0.316 ± 0.079 ng (contractile), 0.595 ±0.174 ng (reversible synthetic), 1.515 ± 0.369 ng(irreversible synthetic). However, the protein profileof the cell also alters. That is, the cytoplasm of acontractile-state cell contains many myofilaments,while both reversible and irreversible synthetic-statecells contain large amounts of rough endoplasmicreticulum and other synthetic organelles. For thesereasons, metabolic measurements have been ex-pressed on the basis of cell number, that is per 108

cells, rather than per protein mass.

Lipoproteins

The composition of 13-VLDL from hyperlipemic se-rum and of VLDL and LDL from normolipemic serumis shown in Table 1. (Values are means of two sets ofpreparations with similar characteristics). The apoli-poprotein profiles obtained by polyacrylamide gelelectrophoresis19 on 15% and 3.5% gels are shownin Figure 1. The (3-VLDL (90% of which showed betamobility on Pevikon electrophoresis) contains morecholesteryl ester per protein than normal LDL andless triglyceride per protein than normal VLDL. TheB-VLDL is enriched in apolipoprotein E but deficientin apolipoprotein C by comparison with normalVLDL Unlike normal VLDL, I3-VLDL contains sub-stantial apolipoprotein B48 in addition to B100, re-flecting a mixture of chylomicron remnants and liver-derived VLDL20

Llpoproteln Metabolism

The binding characteristics of rabbit B-VLDL andof VLDL and LDL from normal rabbits were tested infour separate sets of experiments. The data, mea-sured as cell-associated radioactivity at 4° C andexpressed as ng lipoprotein protein bound per 106

cells, showed highly consistent differences betweenthe three cell phenotypes. The results of a typicalexperiment are shown in Figure 2. The data from thefour separate experiments have not been meaned,because the cells from different animals and differentbatches of lipoproteins gave quantitatively slightlydifferent results. But with all sets of experiments, thesame similarities and differences were obtained. Thebinding of LDL to the three phenotypes was aboutequally specific (displaceable by excess unlabeledlipoprotein) and nonspecific, suggesting substantialnon-receptor-mediated binding of LDL. By contrast,the binding of both normal VLDL and of (3-VLDL waspredominantly specific and saturable over the con-centrations tested (5-50 /xg protein/ml medium)(Figure 2). There were clear differences in the ca-

Table 1. Composition

Lipoprotein

(3-VLDLVLDLLDL

Pro-tein(%)

7.4

12.5

23.8

of Lipoproteins (% by Weight)

CE(%)

46.34.3

40.7

TG(%)

11.655.8

8.4

FC(%)

8.16.33.8

PL(%)

26.621.123.2

Values are means of two samples. CE = cholesterylester; TG = triglyceride; FC = (ree cholesterol; PL =phospholipid.

pacity (Bmax) of the different phenotypes to bindVLDL and B-VLDL (Table 2); the contractile-statecells showed the lowest binding capacity and theirreversible synthetic-state cells showed the highestcapacity for these lipoproteins. The binding capacityfor B-VLDL by irreversible synthetic cells was invari-ably greater than for normal VLDL. The binding affin-ity by all three cell phenotypes was substantiallyhigher for B-VLDL than for normal VLDL. Approxi-mately half-saturation was achieved at about 5 /ig/mlB-VLDL with the irreversible synthetic-state cells.The Scatchard plots for the binding of B-VLDL andVLDL to both synthetic phenotypes were linear in allcases, suggesting a single binding site. There was agood fit of experimental points to each line.

B

E B100

B48

A B C D

Figure 1. Sodium dodecyl sulfate (SDS)-polyacryla-mide gel electrophoresis of rabbit lipoproteins. Gels fromleft to right indicate the differences in apoprotein E contentwhen 100 ^tg lipoprotein protein of N-VLDL (A) and B-VLDL (B) were applied to 15% (wt/vol) gels, and the rela-tive differences in apoprotein B100 and B48 content when100 ^g lipoprotein protein of N-VLDL (C) and B-VLDL (D)were applied to 3.5% (wt/vol) gels.

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from

322 ARTERIOSCLEROSIS VOL 5,

BINDING: B-VLDL

50-s

,S 40-

| 30-

I 20-

10-

cortractUe synthetic

10 25

BINDING: N-VLDL

0 10 25 50 10 25 50

ug 1 2 5 1 - B-VLDL per ml medium

« 60-

"b 40-

| 30-

I 20-

9 10-

contract!* synthetic IrreversWesynthetic

10 25 60 10 25 50 10 25

ug 125r- N-VLDL per ml medium

BINDING: LDL

50

B 50-o

\ 40-

| 30-

I 20'a 10

contractile synthetic irrevefslbtesynthetic

10 25 SO 10 25 50 10 25

uo 126I-LOL per ml meoium50

Figure 2. Specific (• •) and nonspecific ( • — • )binding of 125l-labeled lipoproteins to rabbit aortic smoothmuscle cells in culture.

No 4, JULY/AUGUST 1985

DEGRADATION: B-VLDL

1C> 25 50 ID 2S 50 10 25 50ug '"l-B-VLDL per ml mecHum

DEGRADATION: N-VLDL

- 1600'

u •

"fe 1200.

i

400-

contractile •yrrthetlc irreversibleayntrietlc

10 25 50 10 26 50 10 2-6 50ug 1ZBI-N-VLDL per ml mecHum

DEGRADATION: LDL

• 1600-

• ynthetlc IrreversAjle•ynthetic

10 26 60 10 25 50 10 25 50

Figure 3. Specific (• •) and nonspecific (• •)degradation of 125l-labeled lipoproteins by rabbit aorticsmooth muscle cells in culture.

Table 2. Binding and Dissociation ConstantsObtained from Scatchard Analyses

Cellphenotype

Contractile

Reversiblesynthetic

Irreversiblesynthetic

Upo-proteinligand

S-VLDLN-VLDL

B-VLDLN-VLDL

B-VLDLN-VLDL

B™x(ngprotein/

106 cells)

10.914.7

22.632.0

54.446.4

kd

2.131.36

3.052.66

1.721.10

*(M)

x 10~9

x 10"8

x 10-9

x 10-8

x 10"9

x 10"8

The same original data were used in this table and inFigure 2.

'Assumes molecular weights of 3.225 x 106 for B-VLDLprotein and 1.118 x 10s for N-VLDL protein.

Degradation characteristics of the lipoproteins re-sembled those for binding with the exception of LDL(Figure 3). Degradation of normal VLDL and of (3-VLDL was highly saturable and showed high-affinity,high-capacity characteristics with the irreversiblesynthetic cells. The capacity of contractile state cellsto degrade these lipoproteins was small and that ofthe reversible synthetic-state cells was intermediate.The degradation of I3-VLDL was greater than of nor-mal VLDL in both synthetic phenotypes. However,the degradation of LDL appeared to be at least asgreat with contractile-state cells (which we had ob-served previously in incubations with human LDL9).

The capacity of the various lipoproteins to interactat common binding sites was examined by definingthe degree of displacement of 12SI-I3-VLDL (5 / l )

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from

LIPOPROTEIN METABOLISM OF SMOOTH MUSCLE Campbell et al. 323

100

BO •

80 -

40 •

20 - \ \N-VLDL

B-VUX * " - - . / * * " • - — - ^• 4

100

8 0 -

8 0 -

4 0 -

2 0 -

10 50 66

ug unlabeled llpoprotefn per ml

6 60

ug unlabolod B-VUX per ml

B

Figure 4. A. Degradation of 125l-labeled (3-VLDL (5^g/ml) in the presence of increasing concentrations of un-labeled B-VLDL, normal VLDL, or LDL in irreversible syn-thetic state cells. B. Displacement of 125l-labeled B-VLDL(5 figJm\) by unlabeled B-VLDL in irreversible syntheticcells in which B/E receptor activity was suppressed (S) orunsuppressed (U).

by unlabeled B-VLDL, normal VLDL, and LDL fromirreversible synthetic state cells (Figure 4 A). Degra-dation was half-suppressed at concentrations of 5/xg/ml of B-VLDL and of 10 /xg/ml of normal VLDL.Initial supression by LDL also occurred at about 5 figprotein/ml, but only about 60% of B-VLDL degrada-tion was suppressible with excess LDL, in contrast toalmost complete suppression with normal VLDL andB-VLDL.

These findings suggested that at least some of theinteraction between B-VLDL and the smooth musclecells occurred through the B/E (LDL) receptor. A fur-ther study was therefore carried out in which the B/Ereceptor had been repressed by preincubating irre-versible synthetic cells with 20% FCS for 17 hours.Then a saturating concentration of rabbit LDL (75 figprotein/ml) was added to the medium +5% LDS.During this latter part of the experiment, 12SI-B-VLDL(5 yu,g/ml) and increasing amounts of unlabeled B-VLDL were added during a 2-hour incubation at 37°

C. As shown in Figure 4 B, the unlabled B-VLDLreduced the uptake (cell-associated radioactivity) oflabeled lipoprotein very much less in the cells inwhich B/E receptor activity had been suppressedthan in cells that had been preincubated in LDS andnot exposed to excess LDL.

Degradation of lipoproteins was not mediatedthrough nonspecific surface proteases, since mini-mal degradation occurred in the absence of cellseven in the presence of medium that had been pre-viously incubated with cells. On the other hand, deg-radation was inhibited by chloroquine, with almosttotal suppression at a concentration of 100 /xM (datanot shown). This was associated with marked accu-mulation of cell associated lipoprotein radioactivity.

Cholesterol and Trlglycerlde Accumulation

The mass of total cholesterol in cells of the threephenotypes following 24-hour incubations with 5%LDS alone or with 5% LDS plus LDL, normal VLDL,or B-VLDL (at 75 fig protein/ml) are shown in Table3. The larger cell size and the greater content ofmembranous organelles were reflected in the great-er cholesterol mass in the synthetic, compared withthe contractile, phenotype (see the lipoprotein-freeincubations). The addition of 120 fig/m\ cholesterolof LDL failed to increase the cellular cholesterol overthat in lipoprotein-free incubations. Incubations with400 /i.g/ml cholesterol of B-VLDL doubled the choles-terol mass of the reversible-synthetic cells and morethan tripled that in the irreversible-synthetic cells.The amount of triglyceride rose in parallel with cho-lesterol, with the greatest levels in irreversible-syn-thetic state cells incubated with B-VLDL (56 ± 2/j.g/106 cells) compared with 28 ± 5 pig/108 cells inlipoprotein-free incubations.

In further experiments, the incubations containedamounts of LDL or B-VLDL that were equivalent interms of cholesterol content. However, tripling theLDL cholesterol concentration in the medium failedto raise the cholesterol mass in the irreversible-syn-thetic cells beyond 21 ̂ g/106 cells, whereas B-VLDLinduced a three- to fourfold increase. Whereas cho-lesteryl ester contributed only about 5% to the cho-lesterol mass in irreversible-synthetic cells incubat-ed without lipoprotein, this rose to 30% of the totalfollowing exposure to B-VLDL. The rise in the total

Table 3. Mass of Total Cholesterol following 24-Hour Incubations in 5% LIpoprotein-DeflclentSerum Alone or with 75 /tg/ml Lipoprotein

Cell phenotypeLDS LDL N-VLDL

(tig total cholesterol per 10* cells)(3-VLDL

ContractileReversible syntheticIrreversible synthetic

6±114±422±2

8±124±623±3

NDND

28±4

13±533±574 + 8

Each value represents the mean ± standard deviation of between three and five separate determina-tions. ND = not determined; LDS = lipoprotein-deficient serum.

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from

324 ARTERIOSCLEROSIS VOL 5, No 4, JULY/AUGUST 1985

Table 4. Incorporation of 3H-Sodium Oleate Into Cellular Cholesteryl Ester

Cell phenotype

ContractileReversible synthetic

Irreversible synthetic

LDS

2.0±0.2

4.1 ±0.64.0 ±0.1

LDL(nmole oleate

2.2±0.4

4.3 ±0.4

4.1+0.1

N-VLDLper 106 cells)

2.0±0.14.1 ±0.4

4.5±0.5

B-VLDL

2.7 ±0.4

6.0±0.7*t5.8±0.6*t

Each value represents the mean ± standard deviation of three separate determinations. LDS =lipoprotein-deficient serum.

'Significantly different from the same cell phenotype incubated in LDS by unpaired nest (p < 0.001).tSignificantiy different from the contractile state cell in B-VLDL by unpaired f test (p < 0.001).

cholesterol content of the cells was therefore due toboth free cholesterol and esterified cholesterol, al-though the percent increment was much greater forthe latter.

Incorporation of Na Oleate Into Cell Llplds

The incorporation of 3H-Na oleate into cellular cho-lesteryl ester was influenced by lipoproteins and cellphenotype, and the changes resembled those ob-served for cholesterol mass. The means for threeexperiments are shown in Table 4. The baseline es-terification rates with LDS were twice as great withthe two synthetic-state cells as with the contractile-state cells. Incubations with LDL, VLDL, or B-VLDL

led to minimal changes with the contractile cells.Whereas neither LDL nor VLDL stimulated incorpor-ation of 3H-oleate into cholesteryl esters in incuba-tions with either form of synthetic cell, incorporationrose by about 50% in the presence of (3-VLDL. Thefindings for 3H-oleate incorporation into cellular tri-glyceride were almost identical; the only clearchanges related to an approximate 50% increasewhen B-VLDL was incubated with reversible and irre-versible synthetic-state cells (data not shown). Itshould be noted that the triglyceride content of B-VLDL was much less than that of VLDL, yet the latterfailed to stimulate oleate incorporation into triglycer-ide above that seen with LDS alone.

B

Figure 5. Contractile (A), reversible-synthetic (B), and irreversible-synthetic (C) state cells stained with Fettrot solutionfor lipid droplets and counterstained with Meyer's Haemalum following a 17-hour incubation in 5% lipoprotein-deficientserum (LDS), then a 24-hour incubation with 75 ^g/ml B-VLDL in 5% LDS. x 600.

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from

LIPOPROTEIN METABOLISM OF SMOOTH MUSCLE Campbell et al. 325

Morphologic Studies

There was a marked variation in the accumulationof stainable lipid when the three smooth muscle cellphenotypes were incubated with the three classes oflipoprotein. Incubations of all three smooth musclephenotypes with 5% LDS or with LDL (120 /xg/mlcholesterol) led to minimal Fettrot staining. The con-tractile-state cells also failed to show more than mi-nor lipid droplets when exposed to VLDL or B-VLDL;however, both synthetic phenotypes showed in-creased lipid staining when incubated with VLDL orwith B-VLDL; by far the greatest lipid staining was inirreversible synthetic cells incubated in 3-VLDL (Fig-ures 5 and 6).

In the presence of 10% unfractionated hyperlipe-mic rabbit serum, marked lipid accumulation oc-curred in the irreversible synthetic cells, which was ofthe order seen with B-VLDL incubations.

Discussion

The main finding was related to the marked het-erogeneity among the three smooth muscle cellphenotypes with respect to lipoprotein metabolism.There were, in addition, significant differences in themetabolism of the three classes of rabbit lipopro-teins. Arterial smooth muscle cells, tested early inprimary culture while they still retained the morpho-logical features of contractile cells, showed a verylow capacity to bind and to degrade VLDL and LDLfrom normal rabbits and B-VLDL from cholesterol-

fed rabbits. This was associated with minimal accu-mulation of cellular cholesterol, low esterification ofcholesterol, and morphological evidence of no in-crease in lipid.

By contrast, smooth muscle cells that had under-gone phenotypic change to the synthetic state inculture had a higher capacity (expressed per unitcell) to metabolize lipoproteins, specifically B-VLDLand normal VLDL. The greatest accumulation of cel-lular cholesterol, which was reflected in marked lipidstaining, occurred when B-VLDL was incubated withcells that were irreversibly in the synthetic state. Withall cells, a much higher proportion of VLDL and B-VLDL was bound through a saturable process thanwas the case for LDL.

Paradoxically, contractile-state cells appeared todegrade at least as much LDL as the synthetic-statecells, which is consistent with our earlier finding ofmore degradation of human LDL by contractile- thanby synthetic-state cells.9 Van Lenten et al.21 havealso recently reported a sharp reduction in the degra-dation of human LDL by human monocytes after thefirst 2 days in culture.

A notable finding was the relative immunity of thecontractile-state cells to cholesterol enrichment,even when they were incubated with B-VLDL; thisreflects their low capacity to bind the classes of lipo-proteins tested. Even the reversible synthetic-statecells, which can revert to the contractile state underappropriate conditions,9 showed only a modest in-crease in cholesterol content with LDL and B-VLDL.

» • * • • • . - ?

B

Figure 6. Irreversible-synthetic state cells stained with Fettrot solution for lipid droplets and counterstained with Meyer'sHaemalum following a 17-hour incubation in 5% LDS, then a 24-hour incubation in (A) 5% LDS alone (control), (B) 5% LDSand 75 /^g/ml LDL, (C) 5% LDS and 75 ng/ml normal VLDL, and (D) 5% LDS and 75 ^m\ B-VLDL. x 600.

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from

326 ARTERIOSCLEROSIS VOL 5, No 4, JULY/AUGUST 1985

Yet the reversible synthetic cells did show a highcapacity for binding R-VLDL, which was only a littleless than that shown by the irreversible syntheticcells; this demonstrated equivalent incorporation ofoleate into cholesteryl esters.

In this study we chose to express the data in termsof cell number rather than cell protein. The mainreason was that the mass and type of cellular orga-nelles (which may have little bearing on lipoproteinuptake) change markedly with change in phenotype.However, since the size of the cell increases withphenotypic change, the surface area of the synthet-ic-state cells is larger than that of the contractile cell.If the number of available receptors rises proportion-ately, then some of our results may be explicable inthose terms. That does not, of course, detract fromthe important consequence that synthetic-state cellsare therefore more susceptible to lipid overload. Nordoes it negate the significance of the greater uptakeof R-VLDL than of the other lipoproteins.

The accumulation of cholesteryl ester by smoothmuscle cells so that they resemble the classic "foamcell" has been difficult to demonstrate in culture.22

This is presumably due to the fact that the LDL or apoB/E receptor, which mediates the uptake of nativelipoproteins by smooth muscle cells, is down-regu-lated as the amount of cellular cholesterol in-creases.23 Thus, incubation of subcultured smoothmuscle cells in high concentrations of LDL or R-VLDL results in about a twofold increase in choles-terol content.24"27 In each of these studies, the per-cent increase for cholesteryl ester was far greaterthan for free cholesterol. The accumulation of cho-lesterol can be increased by using chloroquine28 andby incubating the cells with chemically modifiedLDL,29 which is not subject to regulation by the LDLreceptor. Moderate increases in cholesteryl ester incultured smooth muscle cells have also beenachieved with unfractionated hyperlipemic serumproduced by cholesterol feeding.30"33 This may bedue to the presence in hyperlipemic serum of R-VLDL, the only naturally occurring lipoprotein thatwill cause a massive (20- to 160-fold) increase incholesteryl esters in cultured macrophages.34 Atten-tion has thus focused on I3-VLDL as a potentiallyimportant atherogenic lipoprotein.

In our studies, (3-VLDL increased the cholesterolcontent of irreversible synthetic cells to a muchgreater extent than did LDL, at equivalent lipoproteincholesterol concentrations. This was also true for theincorporation of 3H-oleate into cholesteryl esters;only R-VLDL, not VLDL nor LDL, stimulated oleateincorporation beyond that observed in the absenceof lipoprotein. However, other studies have shownthat LDL obtained from cholesterol-fed monkeysdoes stimulate cholesteryl oleate formation.30

Irreversible synthetic arterial smooth muscle cellsbound R-VLDL by a high-affinity, highly saturableprocess. Since normal VLDL displaced the binding ofR-VLDL almost as effectively as did 6-VLDL itself,and since the initial displacement by LDL was equal-

ly effective, binding of B-VLDL presumably occurs atthe B/E receptor. This was confirmed by suppressionof the receptor, which led to much less displacementof labeled (3-VLDL by unlabeled lipoprotein. Thegreater capacity for (3-VLDL (and to a lesser extentfor VLDL) than for LDL presumably reflects the pro-portionate differences in apo E content and shows amuch higher affinity for the B/E receptor than LDLitself.35 These results do not exclude the recognitionof VLDL and of R-VLDL through an additional proc-ess, although the linearity of the Scatchard plots forthe binding of R-VLDL and VLDL to both syntheticphenotypes favors a single binding site.

We have recently observed that human plasmatriglyceride-rich lipoproteins, which are also rich inapo E, bind as effectively to macrophages as R-VLDL and they also promote esterification of choles-terol.36 The present studies also suggest that lipopro-teins related to triglyceride metabolism (VLDL andR-VLDL) bind more effectively to arterial smoothmuscle cells and influence cellular cholesterol me-tabolism more profoundly than does LDL itself. Theimportance of remnants of triglyceride metabolism inatherogenesis is being recognized through the foam-cell transformation of macrophages by triglyceride-rich lipoproteins37 and from recent epidemiologicaldata.38

In the present experiments, substantial lipid accu-mulation was observed in irreversible synthetic cellsthat had been incubated with R-VLDL and even withnormal VLDL. Considerable lipid staining was seenalso with unfractionated hyperlipemic serum. Fettrotstain will stain both cholesterol and triglyceride, andwe have not distinguished between these possibili-ties. At least with the incubations with R-VLDL, thetriglyceride concentration in the irreversible synthet-ic-state cells doubled. There was also greater incor-poration of 3H-oleate into cellular triglyceride in bothsynthetic phenotypes in the presence of R-VLDLthan with VLDL, LDL, or LDS. Chen and Fischer-Dzoga26 and Pearson39 have reported an increase inboth cellular cholesterol and triglyceride after incu-bations of rabbit aortic smooth muscle cells with un-fractionated hyperlipemic serum.

Our findings are significant for interpretating thevast amount of data on lipoprotein metabolism incultured smooth muscle cells, since in most reportsthe phenotypic characteristics of the cells studiedhave not been described. Indeed, cells from the"second-to-tenth passages" have often been used.In some experiments these would be "reversible syn-thetic" cells, while in other experiments in the sameseries they would be "irreversible synthetic" cells ormixtures of the two.

In addition, the findings of this study may be rel-evant to the etiology of atherosclerosis. During twoor three decades in humans, diffuse intimal thicken-ings form from smooth muscle cells that migrate fromthe media and proliferate.1 Many of the cells of dif-fuse intimal thickenings, and the cells of the athero-sclerotic plaques that develop within them, are mor-

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from

LIPOPROTEIN METABOLISM OF SMOOTH MUSCLE Campbell et al. 327

phologically in a synthetic state and are thus differentfrom the contractile-state cells of the media.3 If themorphologically identical synthetic-state cells in vivoare functionally equivalent to synthetic-state cells inculture, then at least some are likely to be in theirreversible synthetic-state due to the large numberof divisions that they must have undergone to formthe thickening or plaque. Certainly senescent cells(that is, those that have exhausted their proliferativecapacity) have been demonstrated in humanplaques.40 The increased susceptibility of these cells(compared to the contractile-state cells of the media)to accumulate lipid in the presence of R-VLDL maycontribute to the predilection of diffuse intimal thick-enings for atherosclerotic plaque formation.

AcknowledgmentsThe excellent technical assistance of Lucy Popadynec, Trish

Nugent, and Bill Kaegi is gratefully acknowledged.

References1. Vellcan C, Vellcan D. The precursors of coronary athero-

sclerotic plaques in subjects up to 40 years old. Atherosclero-sis 1980:37:33-46

2. Haust MD. Atherosclerosis and smooth muscle cells. In: Ste-phens NL, ed. Biochemistry of smooth muscle, vol II. BocaRaton: CRC Press, 1983:189-250

3. Campbell GR, Campbell JH. Smooth muscle cells. In: Rob-ert L, ed. The biology and clinical science of atherosclerosis.Section D. Cellular and molecular biology of the vessel wall,Edinburgh: Churchill Livingstone (in press)

4. Campbell GR, Chamley-Campbell JH, Burnstock G. Dif-ferentiation and phenorypic modulation of arterial smoothmuscle. In: Schwartz CJ, Werthessen NT, Wolf S, eds. Struc-ture and function of the circulation, vol 3. New York: Plenum,1981:357-399

5. Campbell GR, Campbell JH. Smooth muscle phenotypicchanges in arterial wall homeostasis: implications for thepathogenesis of atherosclerosis. Exp Mol Pathol 1985;42:(inpress)

6. Chamley-Campbell JH, Campbell GR. What controlssmooth muscle phenotype? Atherosclerosis 1981 ;40:347-357

7. Chamley-Campbell JH, Campbell GR, Ross R. Phenotype-dependent response of cultured aortic smooth muscle to se-rum mitogens. J Cell Biol 1981;89:379-383

8. Chamley-Campbell JH, Nestel P, Campbell GR. Smoothmuscle metabolic reactivity in atherogensis: LDL metabolismand response to serum mitogens differ according to pheno-type. ln:BornGRV, CatapanoAL, PaolettiRP, eds. Factors information and regression of the atherosclerotic plaque. NatoAdvanced Study Institute Series A, Vol 51. New York: Ple-num, 1982:115-124

9. Campbell JH, Popadynec L, Nestel PJ, Campbell GR. Lip-id accumulation in arterial smooth muscle cells. Influence ofphenotype. Atherosclerosis 1983,47:279-295

10. Mahley RW. Alterations in plasma lipoproteins induced bycholesterol feeding in animals including man. In: Dietschy JM,Gotto AM Jr, Ontko JA, eds. Disturbances in lipid and lipopro-tein metabolism. Bethesda: American Physiological Society,1978:181-197

11. Hul DY, Innerarlty TL, Mahley RW. Defective hepatic lipo-protein receptor binding of B-very low density lipoproteinsfrom Type III hyperiipoproteinemic patients. Importance of

apolipoprotein E. J Biol Chem 1984;259:860-86912. Brown MS, Goldstein JL. Lipoprotein metabolism in the ma-

crophage: Implications for cholesterol deposition in athero-sclerosis. Annu Rev Biochem 1983:52:223-261

13. Glanturco SH, Gotto AM Jr, Hwang S-LC, et al. Apolipopro-tein E mediates uptake of S,100-400 hypertriglyceridemicvery low density lipoproteins by the low density lipoproteinreceptor pathway in normal human fibroblasts. J Biol Chem1983;258:4526-4533

14. Schonfeld G, Patsch W, Ptleger B, Wrtztum J, WeldmanJL. Lipolysis produces changes in the immunoreactlvity andcell reactivity of very low density lipoproteins. J Clin Invest1979:64:1288-1297

15. Mahley RW, Welsgraber KH, Innerarlty TL. Canine lipopro-teins and atherosclerosis II. Characterization of the plasmalipoproteins associated with atherogenic and non-athero-genic hyperiipidemia. Circ Res 1974:35:722-733

16. Fldge NH, Poulls P. Studies on the radio-iodination of verylow density lipoprotein obtained from different mammalianspecies. Clin Chim Acta 1974;52:15-26

17. Blerman EL, Stein O, Stein Y. Lipoprotein uptake and me-tabolism by rat aortic smooth muscle cells in tissue culture.Circ Res 1974:35:136-150

18. Pearse AGE. Histochemistry. Theoretical and applied. Vol I.London: J & A Churchill, 1968

19. Weber K, Osborn M. The reliability of molecular weight de-terminations by dodecyl sulphate-polyacrylamide gel electro-phoresis. J Biol Chem 1969:244:4406-4412

20. Falnaru M, Mahley RW, Hamilton RL, Innerarrty TL. Struc-tural and metabolic heterogeneity of B-very low density lipo-proteins from cholesterol-fed dogs and from humans withType III hyperiipoproteinemia. J Upid Res 1982:23:702-714

21. Van Lenten BJ, Fogelman AM, Hokom MM, Benson L,Haberiand ME, Edwards PA. Regulation of the uptake anddegradation of B-very low density lipoprotein in human mono-cyte macrophages. J Biol Chem 1983:258:5151-5157

22. Stein O, Vanderhoek J, Stein Y. Cholesterol ester accumu-lation in cultured aortic smooth muscle cells. Induction ofcholesterol ester retention by chloroquine and low densitylipoprotein and its reversion by mixtures of high densityapolipoprotein and sphingomyelin. Atherosclerosis 1977;26:465-482

23. Goldstein JL, Brown MS. The low density lipoprotein path-way and its relation to atherosclerosis. Annu Rev Biochem1977:46:897-930

24. Alam SS, Solen KA, Layman DL, Riddle MC, Connor WE.The mass uptake of cholesterol ester from low density lipo-proteins by cultured smooth muscle cells and adventitial cellsof human aortas. Circ Res 1980;47:374-383

25. St. Clalr RW, Lelght MA. Differential effects of isolated lipo-proteins from normal and hypercholesterolemic Rhesus mon-keys on cholesterol esterification and accumulation in arterialsmooth muscle cells in culture. Biochim Biophys Acta 1978;530:279-291

26. Chen RM, Flscher-Dzoga K. Effect of hyperiipemic serumlipoproteins on the lipid accumulation and cholesterol flux ofrabbit aortic medial cells. Atherosclerosis 1977:28:339-353

27. Mahley RW, Innerarrty TL, Welsgraber KH, Fry DL. Caninehyperiipoproteinemia and atherosclerosis: accumulation oflipid by aortic medial cells in vivo and in vitro. Am J Pathol1977:87:205-225

28. Leake DS, Heald B, Peters TJ. Subcellular fractionation ofarterial smooth muscle cells laden with lipid following incuba-tion with low density lipoproteins and chloroquine. Exp MolPathol 1983;38:82-99

29. Basu SK, Goldstein JL, Anderson RGW, Brown MS. Deg-radation of cationized low density lipoprotein and regulationof cholesterol metabolism in homozygous familial hypercho-lesterolemia fibroblasts. Proc Natl Acad Sci USA 1976;73:3178-3182

30. Bates SR, Jett CM, Miller JE. Prevention of the hyperlipide-mic serum or LDL-induced cellular cholesterol ester accumu-lation by 22-hydroxycholesterol and its analogue. BiochimBiophys Acta 1983:753:281-293

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from

328 ARTERIOSCLEROSIS VOL 5, No 4, JULY/AUGUST 1985

31 . Fischer-Dzoga K, Chen R, Wissler RW. Effects of serumlipoproteins on the morphology, growth and metabolism ofarterial smooth muscle cells. In: Wagner WD, Clarkson TBeds. Arterial mesenchyme and arteriosclerosis (Adv Exp MolBiol No 43). New York: Plenum 1974:299-311

32. Smith BP, St. Clair RW, Lewis JC. Cholesterol esterificationand cholesteryl ester accumulation in cultured pigeon andmonkey arterial smooth muscle cells. Exp Mol Pathol 1979;30:190-208

33. Bates SR, Wissler RW. Effect of hyperlipemic serum oncholesterol accumulation in monkey aortic medial cells. Bio-chim Biophys Acta 1976;450:78-88

34. Mahley RW, Innerarity TL, Brown MS, Ho YK, GoldsteinJL. Cholesteryl ester synthesis in macrophages: stimulationby B-very low density lipoproteins from cholesterol-fed ani-mals of several species. J Lipid Res 1980;21:970-980

35. Pitas RE, Innerarity TL, Arnold KS, Mahley RW. Rate andequilibrium constants for binding of apo-E HDLC (a cholester-ol-induced lipoprotein) and low density lipoproteins to humanfibroblasts: Evidence for multiple receptor binding of apo-EHDLC. Proc Natl Acad Sci USA 1979;76:2311-2315

36. Nestel PJ, Billington T, Bazelmans J. Plasma chylomicronuptake in macrophages via R-VLDL receptors; stimulation ofcholesteryl ester formation [abstr]. Arteriosclerosis 1984;4:539a

37. Bates SR, Murphy PL, Feng Z, Kanazawa T, Getz GS. Verylow density lipoproteins promote triglyceride accumulation inmacrophages. Arteriosclerosis 1984;4:103-114

38. Reardon MF, Craig IH, Harper RW, Nestel PJ. Coronarydisease severity related to triglyceride-rich lipoproteins inwomen and cholesterol-rich lipoproteins in men [abstr]. AustNZJMed 1983;13:408

39. Pearson JD. Lipid metabolism in cultured aortic smooth mus-cle cells and comparison with other cell types. Part 1. Compo-sition of cells grown in hyperlipemic serum. Atherosclerosis1976;24:233-242

40. Ross R, Wight TN, Strandness E, Thiele B. Human athero-sclerosis I. Cell constitution and characteristics of advancedlesions of the superficial femoral artery. Am J Pathol1984;114:79-93

41. Scatchard G. The attractions of proteins for small moleculesand ions. Ann NY Acad Sci 1949;51:660-672

Index Terms: smooth muscle culture • lipoprotein metabolismsmooth muscle phenotype • atherosclerosis

I3-VLDL VLDL LDL

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from

J H Campbell, M F Reardon, G R Campbell and P J Nestelculture.

Metabolism of atherogenic lipoproteins by smooth muscle cells of different phenotype in

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 1985 American Heart Association, Inc. All rights reserved.

Avenue, Dallas, TX 75231is published by the American Heart Association, 7272 GreenvilleArteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/01.ATV.5.4.3181985;5:318-328Arterioscler Thromb Vasc Biol. 

http://atvb.ahajournals.org/content/5/4/318World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

  http://www.lww.com/reprints

Information about reprints can be found online at: Reprints: 

document.Permissions and Rights Question and AnswerFurther information about this process is available in theis being requested is located, click Request Permissions in the middle column of the Web page under Services.Clearance Center, not the Editorial Office. Once the online version of the published article for which permission

can be obtained via RightsLink, a service of the CopyrightArteriosclerosis, Thrombosis, and Vascular Biology Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on July 6, 2018http://atvb.ahajournals.org/

Dow

nloaded from