American Journal of Pathology, Vol. 140, No. 2, February 1992

Copright © American Association of Pathologists

Porcine von Willebrand Disease andAtherosclerosis

Influence of Polymorphism in Apolipoprotein B100Genotype

Timothy C. Nichols,*tt Dwight A. Bellinger,ttsKeir E. Davis,11 Gary G. Koch,tilRobert L. Reddick,tt Marjorie S. Read,tt JanRapacz,# Judith Hasler-Rapacz,#Kenneth M. Brinkhous,tt andThomas R. GriggsttFrom the Departments ofMedicine' and Pathologyt and theCenterfor Thrombosis and Hemostasis* School ofMedicine,the Department of Biostatistics,ll School of Public Health, andthe Division ofLaboratory Animal Medicine,$ University ofNorth Carolina, Chapel Hill, North Carolina; and theDepartments of Genetics and Meat and Animal Sciences, #University of Wisconsin, Madison, Wisconsin

The relationship ofapolipoprotein-B genotype (Lpb)to diet-induced hypercholesterolemia and atheroscle-rosis was studied in von Willebrand disease (vWD)and normal pigs. Von Willebrand and normal pigsdeveloped comparable levels ofhypercholesterolemia(respectively, 757.9 ± 49.4 versus 772.8 ± 47.9 mg/dl, P = 0.95). Pigs with Lpbl/5 andLpb5I8 genotypes,however, developed significantly higher serum cho-lesterol levels than those with other Lpb genotypes(866.1 ± 64.0 mg/dl, P = 0.0343). Coronary andaortic atherosclerosis, measured by computer-assisted automated image analyzer, were not signif-icantly different between vWD and normalpigs. Pigswith an Lpb5 allele developed significantly moreatherosclerosis than those with the Lpb3/8 or Lpb8/8genotypes or the rare Lpbl allele (r - 0.434, P S0.05). Polymorphism in apolipoprotein B100 geno-type, then, significantly influenced the severity ofdiet-induced hypercholesterolemia and atheroscle-rotic plaque formation in vWD and normal swinewithout regard to the vWD genotype. (Am J Pathol1992, 140:403-415)

A role of von Willebrand factor (vWF) in atherogenesishas been suggested by studies in normal and von Wille-brand disease (vWD) pigs.-1 1 In some of these studies,atherosclerosis was induced by high-cholesterol andhigh-fat diets.2-611 These diets have yielded pigs with avariable degree of hypercholesterolemia, and, in somestudies, there was a trend toward higher levels of serumcholesterol in one or the other of the two groups of pigs.Also, within these two groups of atherosclerotic pigs, vari-ation in the extent of coronary and aortic atherosclerosishas been noted.1-11 This variability in diet-induced hy-percholesterolemia and atherosclerosis between normaland vWD pigs suggests that vWF may influence choles-terol metabolism in such a manner as to influence athero-genesis.We have conducted a series of studies on the effect of

vWF on coronary atherogenesis and thrombosis.4i Inthe first study, the amount of coronary atherosclerosiswas related to the degree of hypercholesterolemia thatan individual pig developed and not to the presence ofvWF. In addition, balloon injury of one of the three coro-nary arteries at the start of the experiment did not aug-ment the amount of coronary atherosclerosis an individ-ual pig developed.4 In the subsequent two studies, ath-erosclerosis was produced by diet alone. In the secondstudy, balloon injury of coronary atherosclerotic plaquesin vWD and normal swine at death induced a monolayerof attached platelets similar to what was seen in exposed,nonatherosclerotic subendothelium.5 The presence ofatherosclerosis, however, did not promote platelet-fibrinthrombus formation in either phenotype. The presence ofatherosclerosis, however, did support platelet activation

Supported in part by National Institutes of Health grants HL26309 andHL01 648.

Accepted for publication August 26, 1991.Address reprint requests to Dr. Timothy C. Nichols, Department of

Medicine, University of North Carolina School of Medicine, CB #7075,349 Burnett-Womack Building, Chapel Hill, NC 27514.

403

404 Nichols et alAJP Februaty 1992, Vol. 140, No. 2

on balloon-injured coronary plaques independent of vWFas judged by platelet ultrastructure in the vWD pigs. Inthe third study, it was shown that vWF was required tosupport the development of occlusive thrombosis even inthe presence of coronary atherosclerosis and hypercho-lesterolemia.6 In each of these studies, variations in thedegree of hypercholesterolemia and coronary athero-sclerosis were noted within the two phenotypic groups ofpigs.

Employing immunogenetic and molecular genetictechniques, Rapacz et al12 have documented consider-able variability in porcine apolipoprotein B100 struc-ture.12-14 Apolipoprotein Bi 00 is the major protein com-ponent of plasma low-density lipoprotein (LDL). Low-density lipoprotein is the predominant cholesteroltransport protein in several species, including pigs.12-14Eight Lpb alleles have been identified at the apolipopro-tein B locus in swine.12'13 The codominant expression of16 Lpb allotypes (Lpb 1 to 8 and Lpb 1 1 to 18) allows forthe determination of any of the 36 Lpb genotypes directlyfrom the immunological phenotype.12 13 Feeding high-fatdiets has shown that some Lpb genotypes enhanceatherogenesis.14

In the present study, the fourth in a continuation of thisseries, we described the Lpb polymorphism in the nor-mal and vWD swine from the three earlier studies.4' Thepurpose of this study was to determine if the variability insusceptibility of these vWD and normal swine to diet-induced hypercholesterolemia and atherosclerosis wasrelated to the polymorphism in porcine apolipoproteinBi 00. The results showed that Lpb polymorphism but notvon Willebrand disease status significantly influenced thedevelopment of diet-induced hypercholesterolemia andcoronary and aortic atherosclerosis. These data havebeen reported in part previously.15

Methods

Experimental Animals

All pigs were obtained from the Francis Owen Blood Re-search Laboratory at the University of North Carolina,Chapel Hill, where the vWD pigs have been maintainedas a closed colony since it was established in 1971. Allanimals were treated according to the standards set in'Guide for the Care and Use of Laboratory Animals' (Na-tional Institutes of Health publication no. 85-23). PorcinevWD is characterized by a prolonged bleeding time, de-creased levels of plasma factor VIII coagulant activity,and less than 1% of normal plasma vWF antigen andactivity.1 --19

The following criteria were used to select pigs for thisanalysis: male sex, complete data on vWF levels andgenotype, sufficient stored plasma to allow Lpb genotyp-

ing, ingestion of the atherogenic diet for 16 or 22 to 26weeks for response of serum cholesterol and for quanti-fication of coronary and aortic atherosclerosis by stan-dard methods. Thirty-six pigs (17 vWD and 19 normal)met the criteria for analysis of the effect of the vWD ge-notype on the response of serum cholesterol to theatherogenic diet, 33 (16 vWD and 17 normal) for the ef-fect of Lpb genotype on serum cholesterol, 27 (14 vWDand 13 normal) for the effect of vWD, serum cholesterol,and Lpb genotypes on coronary atherosclerosis, and 21(12 vWD and 9 normal) for aortic atherosclerosis. Thedata on Lpb genotype and aortic atherosclerosis forthese pigs have not previously been reported.

Administration of Atherogenic Diet

At initiation of the atherogenic diet, pigs ranged in agefrom 10 to 15 weeks and weighed between 13 and 32 kg.The atherogenic diet was fed to all pigs for 16 or 22 to 26weeks (Tables 1 and 2). At killing, there was no significantdifference in the weights of animals in the normal andvWD groups (normal, 51.79 ± 1.85 kg; and vWD, 52.81± 2.94, P = 0.900, Wilcoxon rank sum test). The athero-genic diet was composed of pig chow supplementedwith 1% or 2% cholesterol, 20% beef tallow, and 0.75%cholate4- (Tables 1 and 2).

Hemostatic Profiles

Plasma was assayed for vWF activity by the tap-tubemacroscopic agglutination test using paraformaldehyde-fixed human platelets and porcine plasma.17 19 Plateletcounts, hematocrits, and leukocyte counts were also de-termined.4i Bleeding times were done by the method ofMertz.17 Except for the tests for vWF, no significant dif-ferences were found between vWD and normal pigs forthese values.'4

Serum Cholesterol Concentration

Serum cholesterol concentration was determined by BioVet Laboratories, Burlington, North Carolina, by themethod of Allain et al,20 as modified by American MonitorCorporation, Indianapolis, Indiana, for automated deter-mination. Values were obtained before feeding theatherogenic diet and every month until killing. The meanlevel of hypercholesterolemia for each pig was definedas the average of monthly values while receiving diet.

Determination of Lpb Genotype

The Lpb phenotypes were determined in the Immunoge-netics Laboratory, University of Wisconsin, Madison, Wis-

von Willebrand Disease, Lipoproteins, and Atherosclerosis 405AJP Febray 1992, Vol. 140, No. 2

Table 1. Lpb Genotype and Increase in Serum Cholesterol Levels in 19 Normal Swine Fed an Atherogenic DietSerum cholesterol (mg/dl)

Pig no. Lpb genotype

174M43N32R34R59R03N56-070-01 43M04N05N85N118Q35R141-01510145N205NI106MI

1/11/11/81/81/81/81/81/81/51/51/51/55/85/85/83/88/8

Dietary cholesterol (%)

22

11

122

12

22

2

* Mean = the average of monthly values while receiving the atherogenic diet.t The cholesterol value was only present at month 5 for this pig.t Mean + SE of the average monthly values.11 Pigs included in the analysis of response to diet-induced hypercholesterolemia and coronary atherosclerosis only (Lpb genotype not

available).

consin, by the double immunodiffusion test using 16 Lpbepitope specific alloimmune sera as described by Ra-pacz et al.12,13 Those testing the samples in the immu-nogenetics laboratory had no knowledge of the von Wille-brand disease genotype nor of the degree of hypercho-lesterolemia and atherosclerosis documented inindividual animals.

Coronary Atherosclerosis

All animals were killed with an overdose of pentobarbitalwhile under halothane anesthesia. The heart and coro-nary arteries were perfusion fixed.' Cross sections ofeach coronary artery were taken perpendicular to the di-rection of blood flow at 1-cm intervals from their origin. In

Table 2. Lpb Genotype and Increase in Serum Cholesterol Levels in 17 von Willebrand Disease Swine Fed anAtherogenic Diet

Serum cholesterol (mg/dl)Pig no. Lpb genotype Dietary cholesterol (%) Weeks on diet Initial Mean*111R 1/1 1 24 145 627112R 1/1 1 25 140 597151-0 1/1 1 16 83 68521S 1/3 1 26 122 49775N 1/3 2 16 112 80860-0 1/3 1 16 104979-0 1/3 1 16 62 76422R 1/8 1 23 108 104416S 1/8 1 25 94 47522N 1/8 2 16 101 61215N 1/8 2 16 126 100525N 1/5 2 16 94 1005154M 5/8 2 16 102 92116N 5/8 2 16 121 89386R 3/8 1 25 96 432199M 3/8 2 16 102 69988Mt 2 16 78 774

105.4 + 5.5 757.9 + 49.4tMean = the average of monthly values while receiving the atherogenic diet.

t Mean ± SE of the average monthly values.t Pig included in the analysis of response to diet-induced hypercholesterolemia and coronary atherosclerosis only (Lpb genotype not

available).

Weeks on diet

16162526241616161616161625221625222416

Initial

86110115101110107

7313012597929410510469

86123

101.6 ± 4.2

Mean*

92164210809857457326578566387627398808841319620528511t479706

772.8 ± 47.9t

406 Nichols et alAJP February 1992, Vol. 140, No. 2

this report, the term 'coronary artery cross section'(CACS) refers to these specimens. The amount of coro-nary atherosclerosis was calculated three ways for eachpig: A) the mean percentage luminal narrowing by intimaobtained by averaging results for all CACS; B) the num-ber of CACS per pig with at least 15% luminal narrowingby intima; and C) the percentage of the total number ofCACS per pig with at least 15% luminal narrowing byintima.4'5

Aortic Atherosclerosis

The aorta was either fixed under systemic pressure in10% buffered formalin or immersion fixed in 1% gluteral-dehyde and 4% formaldehyde when the animal waskilled.4'6 The fixed aorta was divided into 10 to 12 equal-sized segments taken perpendicular to the direction ofblood flow from just below the celiac artery to the origin ofthe iliac arteries. Aortas that were immersion fixed wereopened longitudinally along the ventral aspect and thendivided into segments. Aortas that were fixed under pres-sure were cut into cross sections. Light microscopic sec-tions were prepared from the proximal half of each seg-ment. The extent of aortic atherosclerosis was defined asboth the intimal area in square millimeters and the intimalarea as a percentage of medial area.2122 The summarystatistics are reported as the mean of all values for eachgroup of pigs for the entire abdominal aorta (ie, from theceliac axis to the aortoiliac junction) and for the distal half(ie, sections from the distal five or six of the 10 to 12segments) of the abdominal aorta.

Statistical Methods

Primary Evaluation and Correlations

The data for serum cholesterol and degree of coro-nary and aortic atherosclerosis were described for Lpbgenotypes with means, standard errors, and minimumand maximum values. The association of the respectiveLpb genotypes with serum cholesterol and degrees ofatherosclerosis was primarily evaluated with Spearmanrank correlation coefficients. Unless otherwise stated, rand P values are from this correlation analysis. Also,comparisons across groups based on Lpb genotypeswere made with the Kruskal-Wallis test.

Groups of Lpb Genotypes

Multiple linear regression methods for the analysis ofvariance (ANOVA) were used to investigate further therelative roles of different Lpb genotypes, von Willebrand

disease, and serum cholesterol in explaining variation indegree of coronary and aortic atherosclerosis. Thesemethods also were applied to relationships between se-rum cholesterol and different Lpb genotypes and vonWillebrand disease. These analyses supported groupingof the seven Lpb genotypes as follows: Group 1 withLpbl/1, Lpbl/3, and Lpbl/8 genotypes (ie, the Lpbl al-lele but not the Lpb5 allele) (n = 19); Group 2 with theLpbl/5 and Lpb5/8 genotypes (ie, genotypes containingthe Lpb5 allele) (n = 10); and Group 3 with Lpb3/8 orLpb8/8 genotypes (ie, neither the Lpbl nor the Lpb5 al-lele) (n = 4).

Results

vWD and Lpb Genotypes

The results of the Lpb phenotypic classification of theindividual swine are shown in Tables 1 and 2. Table 1 liststhe normal pigs that had bleeding times of 2.38 + 0.2minutes and vWF activity of 94.3 ± 8.1%. Table 2 lists thevWD pigs, which had bleeding times of more than 10minutes and less than 1% vWF activity. Four Lpb alleleswere identified in the normal and vWD atheroscleroticpigs: Lpb 1, 3,5, and 8. The Lpbl and Lpb8 alleles werepresent in homozygous and heterozygous genotypeswhereas the Lpb3 and Lpb5 alleles were present only asheterozygotes. No mutant Lpb5 alleles for spontaneoushypercholesterolemia were found.23

Serum Cholesterol: Effect of vWD and LpbGenotypes

The response of serum cholesterol to the atherogenicdiet, percentage dietary cholesterol, and length of feed-ing diet according to genotypes are tabulated in Tables 1and 2. No significant difference was found between theaverage of mean values in cholesterol levels betweennormal and vWD pigs (respectively in mg/dl, 772.8 ±47.9 versus 757.9 + 49.4, P = 0.95, Wilcoxon rank sumtest). No significant differences were found in the degreeof hypercholesterolemia related to Lpb genotype in pigsthat were fed either a 1% or 2% diet (mean cholesterol forpigs receiving 1% cholesterol was 769.0 ± 58.5 mg/dl;and for those receiving 2% cholesterol diet, 762.6 ± 36.4mg/dl, P = 0.776, Kruskal-Wallis). In addition, when theindividual cholesterol levels were compared at eachmonth of receiving the diet for 16 weeks or 22 to 26 weeksand the average of monthly values, no significant differ-ences were found in the levels achieved (data notshown).

Mean cholesterol levels for the pigs are listed in Table

von Willebrand Disease, Lipoproteins, and Atherosclerosis 407AJP Febnay 1992, Vol. 140, No. 2

3 according to the groups of Lpb genotypes and therespective homozygous and heterozygous Lpb geno-types. The degree of diet-induced hypercholesterolemiawas not significantly associated with the presence of anyparticular allele, although the correlation with the Lpb5allele (r = 0.263, P = 0.1) was somewhat stronger thanfor the other alleles. A comparison of mean cholesterollevels for each of the three groups of Lpb genotypes,however, showed that pigs in group 2 (ie, pigs with theLpb5 in their genotypes) developed a significantly highermean cholesterol than those in groups 1 or 3 (P =

0.0343, Kruskal-Wallis test) (Table 3). An overall compar-ison of mean cholesterol for each of six respective Lpbgenotypes (ie, with Lpb3/8 and Lpb8/8 pooled becauseof small sample size) showed the highest mean choles-terol value in Lpb5/8 genotype (927.4 ± 112.0, n = 5, P= 0.128, Kruskal-Wallis test) and the lowest mean cho-lesterol in pigs with Lpb3/8 and Lpb8/8 genotypes (542.5± 56.2, n = 4).

Morphology of Atherosclerosis

Coronary Arteries

The morphology of the coronary atherosclerosis foundin these normal and vWD pigs has been described indetail.46 In general, examination of the serial cross sec-tions of all three coronary arteries showed that certainareas were uninvolved, others had accumulations offoam cells and smooth muscle cell proliferation with afibrous cap, and in some sections calcification, necrosis,and hemorrhage were also evident.

Aortic Lesions

Two types of aortic lesions were evident on gross

specimens in both normal and vWD pigs. The aortic archand thoracic aorta contained flat, translucent lesions asdescribed in previous studies.23 In contrast, the lesionsfound in the abdominal aorta not only contained flat,translucent lesions but also ones that were raised, palegray, and fibrous.23

By light microscopy, some regions of the aorta were

uninvolved. The flat lesions in the aorta contained accu-mulations of foam cells that ranged from a few cells toseveral cells thick. Sections from the raised plaques inthe abdominal aorta contained foam cells, smooth mus-cle cells, and a fibrous cap (Figures 1-3). Sections werechosen for illustration based on two criteria. First, each isfrom one of three distal sections taken from the abdom-inal aorta (ie, section 9, 10, or 1 1 of the 10 to 12 from eachaorta as described in Materials and Methods). Second,the amount of atherosclerosis was equivalent to the mean± standard error values for the normal or vWD pigs con-tained within the three groups of Lpb Genotypes (data infigure legends).

Coronary Atherosclerosis: Effect of vWDand Lpb Genotypes and Serum Cholesterol

There were no significant differences in the three criteriaused to describe coronary atherosclerosis between nor-mal and vWD pigs as previously noted when pigs werereported separately (P - 0.628, Wilcoxon rank sum test)(Table 4).4 For the three groups of Lpb genotypes, thepresence of the Lpb5 allele correlated significantly withthe development of the greatest amount of coronary ath-erosclerosis. This finding was significant for all three cri-teria (Table 5).

Multiple regression methods for ANOVA also were

used to analyze relationships between degree of coro-nary atherosclerosis and combinations of Lpb alleles andgenotypes as well as presence or absence of von Wille-brand disease. Descriptive statistics for the three groupsof Lpb genotypes are given in Table 5. A significantlygreater degree of coronary atherosclerosis was presentin the coronary arteries of pigs with the Lpb5 allele (ie,group 2) when compared with pigs in group 1 (0.026 -

P < 0.107, ANOVA) or group 3 (0.002 - P < 0.007)(Table 5). The presence of the Lpbl allele in the absenceof the Lpb5 allele (ie, group 1) was associated with anintermediate degree of coronary atherosclerosis. Theseresults also suggested that the Lpb3/8 or Lpb8/8 geno-

types (ie, group 3) were associated with somewhat lowerdegrees of coronary atherosclerosis. The data on pigswith these genotypes should be viewed cautiously be-

Table 3. Lpb Genotype and Diet-induced HypercholesterolemiaLpb genotypes Serum cholesterol (mg/dl)*

Group Genotypes n Mean + SE Minimum Maximum1. Lpbl"/'113,1/8 19 777.8 + 43.8 475 10802. Lpbl15.5/8 10 866.1 ± 64.0t 620 13193. Lpb3/8,8/8 4 542.5 ± 56.2 432 699

* Average of mean of monthly values for the groups while receiving the atherogenic diet.t P = 0.0343, Kruskal-Wallis test for cholesterol across the three Groups of Lpb Genotypes.

408 Nichols et alAJP February 1992, Vol. 140, No. 2

IA

lB

Figure 1. Diet-induced atherosclerosis ofabdominal aorta in a von Willebrand disease pig uith Lpb'8genotype (pig number 165, Table 2,dietfedfor 25 weeks). Morphometric anallsis indicated the atheromatous plaque area was 0.938 mmu?' or 6.6% of medial area. The meanvalues of these measurements for the z'WD pigs in group 1 on Table 7 considered separately was 0.811 + 0.201 mm2 and 8.2 + 2.3% (n= 8). For comparison, the mean 'alues of these measurements for normal pigs in group 1 on Table 7 considered separately uwas 1.854 +0.723 mm' and 20.5 + 8.4% (n = 4). Pigs with the Lpb' allele developed an intermediate amount of coronary and aortic atherosclerosisrelative to pigs with other alleles (group 1 on Tables 5 and 7); (A) is a whole mount; (B) is a x4.6 magnification which includes lesion. Aortaopened and immersion fixed Verhoeff-van Gieson. Bar = 1 mm in both (A) and (B).

cause group 3 only contained four pigs. These differ-ences in coronary atherosclerosis between pigs ingroups 1, 2, and 3 were similarly evident in these regres-sion analyses when serum cholesterol was taken into ac-

count. Also with Lpb genotype taken into account, neitherserum cholesterol nor presence or absence of vWD wassignificantly associated with severity of coronary athero-sclerosis. When analyzed without adjustment for Lpb ge-

von Willebrand Disease, Lipoproteins, and Atherosclerosis 409AJP February 1992, Vol. 140, No. 2

e..i_L.

>,...i;--

L ..

r

% ,.

.SL' 8

;; vt:YX,

g g. B ,.

_ E __ E nr M.; E :S.#..MS.

- ,P4fff[,:

I B i"7'e' _* __;.S W _

_ | -t"" >s" .* ss' [_ .;.4 . ,_ r| ,,!::ss .§ t§ t-KM3* 8 |

|--l- -i | - | .. . #.i Y.l7W R _- - l | l N Bg | i !N l i lNf " % ^ re£.f#, .,.st *' 4|4;ht-

[ . tS ,, " w ; 3 *

Figure 2. Diet-induced atherosclerosis ofabdominal aorta in a ion Willebrand disease pig with Lph5 8 genotype (pig number 1 54M, Table2, dietfedfor 16 weeks). The atheromatous plaque contains lipid-ladet foam cells and afibrous cap encroaching on the lumen ofthe vesselMorphometric analysis indicated intinmal plaque area uwas 4.445 mm-' or 34% of medial area. The mean values of these measurementsforthe vWD pigs in group 2 on Table 7 considered separately u'as 1.886 ± 0.6738 mm12 and 21.6 ± 9.4% (n = 3). For comparison, the meanvalues of these measurenmentsfor the normalpigs in group 2 on Table 7 considered separatelly uas 1.478 0.445 mm2 and 15.5 -+- 3.1%(n = 4); Pigs with the Lpbs allele developed the most severe atherosclerosis (group 2 on Tables 5 and 7); (A) is a whole mount of the crosssection; (B) is x3 magnffication uhich includes lesion. Aorta pressure perfused. Methblene blue and basicfuscin, bar = 1 nim in both (A)and (B).

notype, serum cholesterol was significantly correlatedwith the three criteria used to describe coronary athero-sclerosis, as has been previously reported4: for A), r =

0.39, P = 0.045; for B), r = 0.31, P = 0.091; for C), r =

0.39, P = 0.035.

Analyses were performed to determine if there was an

effect of 1% versus 2% cholesterol diet and the durationof diet feeding (16 versus 22 to 26 weeks) on the severityof coronary atherosclerosis. Analyses that adjusted forthese variables supported the principal conclusions con-

2A

410 Nichols et alAJP February 1992, Vol. 140, No. 2

'J!J

ijf

3B

Figure 3. Diet-induced atherosclerosis ofabdominal aorta in a von Willebrand diseasepig uith Lpt' genotype (pig number 86R, Table 2,diet fed for 25 uweeks). Morphometric analysis indicated lesion area uwas 0.0672 mm or 1% of medial area. The mean value of thesemeasurementsfor the vWD pig in group 2 on Table 7 considered separately was 0.121 mm and 1.1% (n = 1). For comparison, the meanvalue of these measurementsfor the normal pig in group 3 on Table 7 considered separately was 0.1-35 mm2 and 2.1% (n = 1); (A) is auhole mount; (B) is a x 4 magnification which includes lesion. Pigs in this group ofLpb Genotvpes developed the least severe coronary andaortic atherosclerosis (group .3 on Tables 5 and 7). Aorta opened and immersion fixed. Verhoeff-van Gieson, bar = 1 mm in both (A) and(B).

cerning Lpb genotypes, as did analyses that did not ad-just for these variables (data not shown).

To determine the effect of including diet and ballooninjury methods of inducing coronary atherosclerosis,analyses were performed that excluded the subset ofcoronary arteries that were ballooned. There was nochange in the association of severity of coronary athero-sclerosis with Lpb genotype when the ballooned coro-nary artery was excluded (data not shown).

Aortic Atherosclerosis: Effect of vWD andLpb Genotypes and Serum Cholesterol

The extent of aortic atherosclerosis, expressed as boththe intimal area in square millimeters and the mean inti-mal area as a percentage of media, tended to be lower inthe vWD pigs when compared with normal pigs (Table6). The differences were not significant considering either

von Willebrand Disease, Lipoproteins, and Atherosclerosis 411AJP Febn-y 1992, Vol. 140, No. 2

Table 4. Coronary Atherosclerosis Morphometry in Normal and vWD Swine

Normal and vWDCoronary atherosclerosis morphometry

swine genotypes n Percent luminal narrowing (A) Number CACS 15% (B) Percent CACS ,15% (C)

Normal 14 14.44 ± 2.68% 4.14 ± 1.01 34.05 ± 7.91%vWD 13 13.97 ± 3.22% 3.92 ± 0.93 33.90 ± 8.95%P valuet 0.628 0.752 0.950

* Summary statistics of the coronary atherosclerosis morphometry are reported for the normal and vWD pigs as mean ± SE. The severityof coronary atherosclerosis is reported by three criteria: A) mean percent luminal narrowing by intima by genotype; B) the mean number ofcoronary artery cross sections (CACS) per genotype with a15% luminal narrowing by intima; and C) the mean percent of the total number ofcoronary artery cross sections per genotype with 2 15% luminal narrowing by intima. The individual values were calculated for each pig andwere averaged, and then the result was averaged with those for all other pigs with normal or vWD genotypes. Coronary atherosclerosismorphometry data and calculations are from the formulas reported in references 4 and 5.

t P values are from Wilcoxon rank sum test.

the entire abdominal aorta or the distal half (P - 0.201,Wilcoxon rank sum test, Table 6).

For the Lpb genotypes, the correlation of the Lpb5allele with severity of atherosclerosis approached signif-icance for the entire abdominal aorta (for mean intimalarea in square millimeters, r = 0.35, P = 0.120; for intimalarea as a percentage of medial area, r = 0.38, P =

0.086). Because most of the raised fibrous plaques ap-peared in the distal aorta, as has been noted before,23the distal half of the abdominal aorta was evaluated sep-arately. The Lpb5 allele was significantly correlated withthe development of atherosclerosis in the distal half (formean intimal area in square millimeters, r = 0.450, P =

0.041; and for intimal area as a percent medial area: r =

0.434, P = 0.0495) (Table 7, Figure 2).The degree of diet-induced hypercholesterolemia

was significantly correlated with aortic atherosclerosisexpressed as intimal area as a percentage of mediafor both the entire abdominal aorta and the distalhalf (respectively, r = 0.42, P = 0.058; and r =

0.44, P = 0.045). The same trend was suggestedwhen aortic atherosclerosis was considered as intimalarea in square millimeters (for the entire abdominal aorta,r = 0.31, P = 0.165; for the distal half, r = 0.39, P =

0.08).

Multiple regression analyses for analysis of variance(ANOVA) were used also for both measures of abdomi-nal aortic atherosclerosis to analyze relationships be-tween severity of atherosclerosis and combinations ofLpb alleles and genotypes as well as presence or ab-sence of von Willebrand disease. In addition, these anal-yses were made with and without adjustment for choles-terol. The correlation of the Lpb5 allele or serum choles-terol with the amount of aortic atherosclerosis in theseanalyses tended to be significant when considered sep-arately (P < 0.05, Wilcoxon rank sum test) or at leastsuggested trends. With adjustment for serum cholesterol,however, differences in the degree of atherosclerosis be-tween Lpb genotypes were no longer significant (P 30.15, ANOVA); these differences were also not significantwhen there was adjustment for 1% versus 2% cholesteroldiet and duration of diet feeding (16 versus 22 to 26weeks) (data not shown). Similarly serum cholesterol wasno longer significant with adjustment for Lpb status.These latter findings could be mainly due to the smallsample size for analysis of aortic atherosclerosis. Impor-tantly, comparisons of severity of aortic atherosclerosisbased on the absence or presence of von Willebranddisease were clearly not significant (P 3 0.15, Wilcoxonrank sum test).

Table 5. Lpb Genotype and Coronary Atherosclerosis MorphometryLpb genotypes Coronary atherosclerosis morphometry*

Group Genotypes n Percent luminal narrowing (A) Number CACS >15% (B) Percent CACS a15% (C)

it Lpbl/1"1/3,1/8 14 12.9 + 2.4% 3.5 t 0.8 31.5 ± 8.1%(0.1, 29.4) (0,8) (0,100)2t Lpbl/5t5t./8 9 21.6 t 3.3% 6.4 t 1.1 50.8 t 8.5%(9.0, 39.7) (2, 13) (13.3, 92.9)

3t Lpb3/8a8/8 4 2.3 ± 1.2% 0.3 ± 0.3 2.1 t 2.1%(0, 4.8) (0, 1) (0, 8.3)

Summary statistics of the severity of coronary atherosclerosis are reported for the three groups of Lpb genotypes as mean ± SE with minand max values listed below in parenthesis. Coronary atherosclerosis is reported for each group with three criteria as defined in Table 4 andin Materials and Methods. The individual values were calculated for each pig and were averaged, and then the result was averaged with thosefor all other pigs in a given group. Calculations are from the formulas reported in references 4 and 5.

t 0.002 - P < 0.007 for the three criteria for coronary atherosclerosis for group 2 versus group 3 and 0.026 - P - 0.107 for group 2 versusgroup 1 in analysis of variance.

t The Lpb5 allele correlates with all three criteria for coronary atherosclerosis: for A) r = 0.52, P = 0.005; for B) r = 0.48, P = 0.01 1; andfor C) r = 0.46, P s 0.015.

412 Nichols et alAJP Febrary 1992, Vol. 140, No. 2

Table 6. Abdominal Aortic Atherosclerosis Morphomety in Normal and vWD Swine

Abdominal aortic atherosclerosis*

Entire Distal half

Normal and vWD Intimal area Intimal areaswine genotypes n mm2 % of media mm2 % of media

Normal 9 1.495 ± 0.391 16.24 ± 4.16% 2.358 ± 0.569 25.09 ± 6.46vWD 12 1.022 ± 0.260 10.99 ± 3.16% 1.510 ± 0.436 16.14 ± 4.98P valuet 0.320 0.227 0.201 0.256

* Aortic atherosclerosis was defined as the mean intimal area expressed as percent of media and the mean intimal area in mm2 for 10-12sequential sections per pig for both the entire abdominal aorta and the distal half (see Materials and Methods, "Aortic Atherosclerosis"). Thevalues were calculated for each pig and were averaged, and then the result was averaged with those for all other pigs in a given group. Thesummary statistics are listed as mean ± SE.

t P values are from Wilcoxon rank sum test.

Discussion

The relationship between vWD genotype, polymorphismin Lpb genotype, and diet-induced atherosclerosis andhypercholesterolemia was studied in normal and vWDpigs. Four different Lpb alleles were identified in the thesepigs: Lpbs 1, 3, 5, and 8. The presence of the Lpb5 allelecorrelated significantly with development of greatestamounts of coronary and aortic atherosclerosis. TheLpb5 allele occurs frequently in populations of pigs ofvarious breeds and has been associated with atheroscle-rosis.12-14 In contrast, the Lpbl allele is rare and unt-ested in studies of atherosclerosis.14 In the normal andvWD pigs, the presence of the Lpbl allele correlated withan intermediate amount of disease relative to the otheralleles. Pigs with Lpb3/8 or Lpb8/8 genotypes developedthe least amount of atherosclerosis. Although pigs withthe Lpb5 allele tended to have higher cholesterol levels,this relationship was not as strong as that for atheroscle-rosis. Importantly, vWD genotype did not influence thedevelopment of either hypercholesterolemia or coronaryor aortic atherosclerosis. These results suggest that apo-lipoprotein B100 genotype is a major factor in determin-ing the severity of coronary and aortic atherosclerosisthat develops in this porcine model.

For simplicity, the atherosclerosis and cholesteroldata in the seven different genotypes found in the normaland vWD pigs were presented in three groups. It is im-portant to note, however, that the pigs with the Lpb5 allelehad the most amount of coronary and aortic atheroscle-rosis when analyzed before genotype grouping. In addi-tion, these pigs developed higher cholesterol levels, par-

ticularly ones with the Lpb5/8 genotypes, which devel-oped the highest cholesterol levels and the greatestamount of coronary atherosclerosis. This analysis thenidentified pigs with Lpb3/8 and Lpb8/8 genotypes as

having the least amount of atherosclerosis and lowestserum cholesterol levels. All other pigs were in betweenthese two extremes and shared the Lpbl allele. Althoughother groupings may be possible, the three Lpb groups

were identified by having mild, moderate, or severe dis-ease when each pig was considered individually.

Previous studies with varying protocols have foundthat the Lpb5 allele is associated with diet-induced hy-percholesterolemia and aortic atherosclerosis.14 Pigswith the Lpb5 allele fed a cholesterol-free, high-fat dietdeveloped both higher mean serum cholesterol levelsand aortic intimal lipidosis than pigs with the Lpb8 allele.In a separate study, Thompson miniature swine with theLpb5 allele developed hypercholesterolemia when fed a

Table 7. Lpb Genotype and Abdominal Aortic Atherosclerosis MorphometyAbdominal aortic atherosclerosis*

Entire Distal half

Lpb genotypes Intimal area Intimal area

Group genotypes n mm2 % media mm2 % media

1. Lpbl/11/3l1/8 12 1.159 ± .0294 12.3 ± 3.4% 1.589 ± 0.438 17.9 ± 5.4%(0.049, 3.771) (0.8, 41.9) (0.085, 5.616) (1.4, 65.9)

2. Lpbl/5t.5t/8 7 1.65 ± 0.375 18.1 ± 4.1% 2.826 ± 0.574t 28.5 t 6.4%t(0.442, 2.877) (4.8, 37.3) (0.706, 4.768) (6.5, 58.0)

3. Lpb3/8 2 0.128 ± 0.007 1.6 + 0.5% 0.195 t 0.47 2.6 t 0.3%(0.121, 0.135) (1.1, 2.1) (0.147, 0.242) (2.3, 2.9)

Aortic atherosclerosis is reported as defined in the legend to Table 6. The summary statistics are listed as mean + SE with min and maxvalues listed below in parenthesis.

t The Lpb5 allele correlates with the degree of aortic atherosclerosis in the distal half of the abdominal aorta (for mean intimal area in mm2,r = 0.450, P = 0.041; and for intimal area as a percent medial area: r = 0.434, P = 0.0495).

von Willebrand Disease, Lipoproteins, and Atherosclerosis 413AJP Febnuary 1992, Vol. 140, No. 2

2% cholesterol, 25% fat diet, whereas pigs with theLpb3/3 genotype were nonresponders.12 No data on thedegree of hypercholesterolemia or atherosclerosis foundin these animals were reported.

The present study confirms and extends these asso-ciations of Lpb genotypes and degree of atherosclerosisto normal and vWD male pigs fed a 1% to 2% cholesterol,high-fat diet. Pigs with the Lpb5 allele not only developedthe most severe aortic atherosclerosis and highest cho-lesterol levels but also the most severe coronary athero-sclerosis. In addition, the normal and vWD pigs with theLpb3/8 and Lpb8/8 genotypes developed the smallestrise in serum cholesterol and the least amount of coro-nary and aortic atherosclerosis. It should be noted thatthis group contained only four pigs: one received 2%cholesterol in the diet for 16 weeks and three received1% or 2% cholesterol for 22 to 25 weeks. The influence ofLpb genotype on diet-induced atherosclerosis is furthersupported by the additional new finding of an intermedi-ate amount of disease in pigs with the rare Lpb1 allele.These findings have identified groups of normal and vWDpigs with specific Lpb genotypes suited for further studyof graded susceptibility to diet-induced atherosclerosis.

How different Lpb genotypes might influence lipidmetabolism and atherogenesis is understood in part.One strain of pigs with a mutant Lpb5 allele phenotypi-cally exhibits spontaneous hypercholesterolemia and ac-celerated atherosclerosis.2325 These Lpb5 mutant pigshave normal LDL receptor binding activity but have ab-normal low density lipoproteins and defective LDL catab-olism, resulting in hypercholesterolemia even while re-ceiving a cholesterol-free diet.23 24 Also a direct relation-ship exists between the mutant Lpb5 genotype and theconcentration and composition of LDL subfractions.26 Al-though there appears to be considerable homology be-tween human and porcine LDL molecules at the pro-posed LDL receptor binding sites, LDL from pigs withLpb2 and Lpb4 alleles has a higher affinity than humanLDL for both the pig and human receptor.27 Interestinglythe LDL receptor binding regions of alleles Lpb2,3,7 andmutant Lpb5 do not have major differences by DNA se-quence analysis.28 The differences in the phenotypic ex-pression of these alleles, then, may result from differ-ences outside of the LDL receptor binding region. Thesefindings are also consistent with the observation in thispresent study that the correlation of Lpb5 allele withamount of atherosclerosis was greater than that found fordegree of hypercholesterolemia. These studies indicate,then, that the polymorphism at the apo-B locus exerts astrong influence on lipid metabolism and atherogenesisin pigs.

It is possible that the various Lpb genotypes produceimportant differences in cholesterol distribution amongdifferent lipoproteins, lipid particle size, or lipoprotein

clearance, similar to what has been found in the sponta-neously hypercholesterolemic mutant Lpb5 pig.242629Such differences would not have been detected by mea-surement of total cholesterol as reported in this study.Further study is needed to document whether or not theLpb 1, 3, 5, and 8 alleles alter specific aspects of lipopro-tein metabolism and atherogenesis in this model of diet-induced atherosclerosis.

Lpb genotypes as determinants of the severity of cor-onary and aortic atherosclerosis may provide insight intothe wide range of differences in the degree of atheroscle-rosis noted in the earlier studies.1'- In this study, vWDpigs with severe hypercholesterolemia, male gender,and especially the Lpb5 allele developed the most se-vere coronary and aortic atherosclerosis, which was notsignificantly different from normal pigs that also had theLpb5 allele. With different Lpb genotypes, both normaland vWD pigs still developed comparable degrees ofcoronary and aortic atherosclerosis but at graded levelsof severity. It is clear that polymorphism in Lpb genotypescould have affected the results of the atherogenesis ex-periments in the vWD stock, confusing the interpretationof the results either to cause apparent enhancement ofthe perceived effect of vWF or to mask a more importanteffect of vWF. It seems likely that any effect of vWF on thedevelopment of atherosclerosis is part of a complex arrayof variables affecting atherogenesis, including the impor-tant contribution of lipids and lipoproteins.

A mechanism linking vWF with atherogenesis wouldmost likely involve support by vWF of platelet adhesionand activation on arterial walls. Von Willebrand factorfunctions as a ligand binding platelets to exposed suben-dothelium, particularly under high shear conditions3033and in the development of occlusive arterial thrombo-sis.6.3.3 Platelets thus attached to the vessel wall andactivated could release platelet-derived growth factor(PDGF) and thereby promote atherogenesis.3536 It isprobable that PDGF can be delivered to the vessel wallby a vWF-independent fashion or that other in vivo factorsplay a significant role in the proliferative response ofsmooth muscle cells to injury.37-42 Data from our labora-tory have shown that both pseudopod formation andspread of platelets adhering to injured arterial walls isimpaired in normocholesterolemic, vWD pigs.543 Inter-estingly the number of adherent platelets in these injuredcoronary arteries is not significantly influenced by thepresence or absence of vWF.543 Reddick et al, however,have recently found that platelets adhering to ulceratedor damaged coronary artery plaques in vWD pigs fed ahigh-cholesterol diet are spread or activated to 73% ofthe degree found in pigs with normal levels of vWF.5 Thisraises the possibility that the Lpb5 allele leads to a greateruptake of lipid by the vessel wall and that the lipid-richvessel wall enhances platelet activation by a mechanism

414 Nichols et alAJP February 1992, Vol. 140, No. 2

that bypasses the need for vWF as demonstrated byReddick et al.5 Such a series of interactions would ex-plain to some degree the loss of protection from athero-sclerosis in vWD pigs when Lpb5 and hypercholesterol-emia are present.

Acknowledgments

The authors thank Mr. Jeff Sigman for excellent technical assis-tance and Ms. Robin Raymer, manager of the Francis OwenBlood Research Laboratory.

References

1. Brinkhous KM, Reddick RL, Griggs TR: Arterial thrombosis,atherosclerosis, and the factor VlIl/vWF complex, The Roleof Factor VIII and von Willebrand Factor. Edited by TS Zim-merman, ZN Ruggeri. New York, Marcel Dekker, 1989, pp283-303

2. Griggs TR, Reddick RL, Sultzer D, Brinkhous KM: Suscep-tibility to atherosclerosis in aortas and coronary arteries ofswine with von Willebrand's disease. Am J Pathol 1981,102:137-145

3. Sultzer DL, Brinkhous KM, Reddick RL, Griggs TR: Effects ofcarbon monoxide on atherogenesis in normal pigs and pigswith von Willebrand's disease. Atherosclerosis 1982,43:303-319

4. Griggs TR, Bauman RW, Reddick RL, Read MS, Koch GG,Lamb MA: Development of coronary atherosclerosis inswine with severe hypercholesterolemia. Lack of influenceof von Willebrand factor or acute intimal injury. Arterioscle-rosis 1986, 6:155-165

5. Reddick RL, Read MS, Brinkhous KM, Bellinger DA, NicholsTC, Griggs TR: Coronary atherosclerosis in the pig: In-duced-plaque injury and platelet response. Arteriosclerosis1990,10:541-550

6. Nichols TC, Bellinger DA, Tate DA, Reddick RL, Read MS,Koch GG, Brinkhous KM, Griggs TR: von Willebrand factorand occlusive arterial thrombosis: A study in normal and vonWillebrand's disease pigs with diet-induced hypercholester-olemia and atherosclerosis. Arteriosclerosis 1990, 10:449-461

7. Bowie EJW, Fuster V, Owen CA Jr, Brown AL: Resistance tothe development of spontaneous atherosclerosis in pigswith von Willebrand's disease. Thromb Diathes Haemorrh1975, 34:599

8. Fuster V, Bowie EJW, Lewis JC, Fass DN, Owen Jr CA,Brown AL: Resistance to arteriosclerosis in pigs with vonWillebrand's disease. J Clin Invest 1978, 61:722-730

9. FusterV, Fass DN, Kaye MP, Josa M, Zinsmeister AR, BowieEJW: Arteriosclerosis in normal and von Willebrand pigs:Long-term prospective study and aortic transplantationstudy. Circ Res 1982, 51:587-593

10. Badimon L, Steele P, Badimon JJ, Bowie EJW, Fuster V:Aortic atherosclerosis in pigs with heterozygous and vonWillebrand disease. Comparison with homozygous von

Willebrand and normal pigs. Arteriosclerosis 1985, 5:366-370

11. Fuster V, Lie JT, Badimon L, Rosemark JA, Badimon JJ,Bowie EJW: Spontaneous and diet-induced coronary ath-erosclerosis in normal swine and swine with von Willebranddisease. Arteriosclerosis 1985, 5:67-73

12. Rapacz J: Lipoprotein immunogenetics and atherosclero-sis. Am J Med Genet 1978, 1:377-405

13. Rapacz J, Hasler-Rapacz J, Kuo WH: Immunogenetic poly-morphism of lipoproteins in swine. Immunogenetics 1978,6:405-424

14. Rapacz J, Elson CE, Lalich JJ: Correlation of an immunoge-netically defined lipoprotein type with aortic intimal lipidosisin swine. Exp Mol Pathol 1977, 27:249-261

15. Nichols TC, Bellinger DA, Koch GG, Davis K, Reddick RL,Read MS, Rapacz J, Hasler-Rapacz J, Brinkhous KM,Griggs TR: Porcine von Willebrand disease and atheroscle-rosis: Influence of polymorphism in apolipoprotein Bi 00 ge-notype in 35 male pigs. Arteriosclerosis 1990, 10(5):858a

16. Hogan AG, Muhrer ME, Bogart R: A hemophilia-like diseasein swine. Proc Soc Exp Biol Med 1941, 48:217-219

17. Mertz ET. The anomaly of a normal Duke's and very pro-longed bleeding time in swine suffering from an inheritedbleeding disease. Am J Physiol 1942,136:360-362

18. Griggs TR, Webster WP, Copper HA, Wagner RH, Brink-hous KM: Von Willebrand factor: Gene dosage relationshipsand transfusion response in bleeder swine-A new bioas-say. Proc NatI Acad Sci USA 1974, 71:2087-2090

19. Allain JP, Cooper HA, Wagner RH, Brinkhous KM: Plateletsfixed with paraformaldehyde: A new reagent for assay ofvon Willebrand factor and platelet aggregating factor. J LabClin Med 1975, 85:318-328

20. Allain CC, Poon LS, Chan CSG, Richmond W, Fu PC: En-zymatic determination of total serum cholesterol. Clin Chem1974, 20:470-475

21. Clarkson TB, Alexander NJ, Morgan TM: Atherosclerosis ofcynomolgus monkeys hyper- and hyporesponsive to dietarycholesterol: Lack of effect of vasectomy. Arteriosclerosis1988, 8:488-498

22. Parks JS, Kaduck-Sawyer J, Bullock BC, Rudel LL: Effect ofdietary fish oil on coronary artery and aortic atherosclerosisin African green monkeys. Arteriosclerosis 1990, 10:1102-1112

23. Rapacz J, Hasler-Rapacz J, Taylor KM, Checovich WJ, AttieAD: Lipoprotein mutations in pigs are associated with ele-vated plasma cholesterol and atherosclerosis. Science1986, 234:1573-1577

24. Checovich WJ, Fitch WL, Krauss RM, Smith MP, Rapacz J,Smith CL, Attie AD: Defective catabolism and abnormalcomposition of low density lipoprotein from mutant pigs withhypercholesterolemia. Biochemistry 1988, 27:1934-1941

25. Prescott MF, McBride CH, Hasler-Rapacz J, Von Linden J,Rapacz J: Development of complex atherosclerotic lesionsin pigs with inherited hyper-LDL cholesterolemia (IHLC)bearing mutant alleles for apolipoprotein B. Am J Pathol1991, 139:139-147

26. Lee DM, MokT, Hasler-Rapacz J, Rapacz J: Concentrationsand compositions of plasma lipoprotein subfractions of

von Willebrand Disease, Lipoproteins, and Atherosclerosis 415AJP Februaiy 1992, Vol. 140, No. 2

Lpb5-Lpu1 homozygous and heterozygous swine with hy-percholesterolemia. J Lipid Res 1990, 31:839-847

27. Ebert DL, Maeda N, Lowe SW, Hasler-Rapacz J, Rapacz J,Attie AD: Primary structure comparison of the proposed lowdensity lipoprotein (LDL) receptor binding domain of humanand pig apolipoprotein B: Implications for LDL-Receptor In-teractions. J Lipid Res 1988, 29:1501-1509

28. Maeda N, Ebert DL, Doers TM, Newman M, Hasler-RapaczJ, Attie AD, Rapacz J, Smithies 0: Molecular genetics of theapolipoprotein B gene in pigs in relation to atherosclerosis.Gene 1988, 70:213-229

29. Checovich WJ, Aiello RJ, Attie AD: Overproduction of abuoyant low density lipoprotein subspecies in spontane-ously hypercholesterolemic mutant pigs. Arteriosclerosis1991, 11:351-361

30. Tschoop T, Weiss HJ, Baumgartner HR: Decreased adhe-sion of platelets to subendothelium in von Willebrand's dis-ease. J Lab Clin Med 1974, 83:296-300

31. Sakariassen KS, Bolhuis PA, Sixma JJ: Human blood plate-let adhesion to artery subendothelium is mediated by factorVIII/von Willebrand factor bound to the subendothelium. Na-ture 1979, 279:636-638

32. Turitto VT, Weiss HJ, Baumgartner HR: Decreased plateletadhesion on vessel segments in von Willebrand's disease:A defect in initial platelet attachment. J Lab Clin Med 1983,101:551-564

33. Nichols TC, Bellinger DA, Johnson AJ, Lamb MA, GriggsTR: Von Willebrand's disease prevents occlusive thrombo-sis in stenosed and injured porcine coronary arteries. CircRes 1986, 59:15-26

34. Bellinger DA, Nichols TC, Read MS, Reddick RL, Lamb MA,Brinkhous KM, Evatt BL, Griggs TR: Prevention of occlusivecoronary artery thrombosis by a murine monoclonal anti-

body to porcine von Willebrand factor. Proc Natl Acad SciUSA 1987, 84:8100-8104

35. Ross R, Glomset J, Kariya B, Harker L: A platelet-dependentserum factor that stimulates the proliferation of arterialsmooth muscle cells in vitro. Proc Natl Acad Sci USA 1974,71:1207-1210

36. Ross R: The pathogenesis of atherosclerosis-an update. NEngl J Med 1986, 314:488-500

37. Friedman RJ, Burns ER: Role of platelets in the proliferativeresponse of the injured artery. Prog Hemost Thromb 1978,4:249-278

38. Ingerman-Wojenski CM, Silver MJ: Model system to studyinteraction of platelets with damaged arterial wall. Exp MolPathol 1988, 48:116-134

39. Schwartz SM, Reidy MA: Common mechanisms of prolifer-ation of smooth muscle in atherosclerosis and hypertension.Hum Pathol 1987,18:240-247

40. Barrett TB, Benditt EP: Platelet-derived growth factor geneexpression in human atherosclerotic plaques and normalartery wall. Proc NatI Acad Sci U S A 1988, 85:2810-2814

41. McCusker RH, Campion DR, Jones WK, Clemmons DR:The insulin-like growth-factor binding proteins of porcine se-rum: Endocrine and nutritional regulation. Endocrinology1989,125:501-509

42. Antoniades HN, Galanopoulos T, Neville-Golden J, KiritsyCP, Lynch SE: Injury induces in vivo expression of platelet-derived growth factor (PDGF) and PDGF receptor mRNAsin skin epithelial cells and PDGF mRNA in connective tissuefibroblasts. Proc Natl Acad Sci USA 1991, 88:565569

43. Reddick RL, Griggs TR, Lamb MA, Brinkhous KM: Plateletadhesion to damaged coronary arteries: Comparison of nor-mal and von Willebrand diseased swine. Proc Natl Acad SciU S A 1982, 79:5076-5079