Platelet-derived growth factor production by cells from Dacron grafts implanted in a canine model

Platclct-dcrivcd growth factor production by cells from Dacron grafts implanted in a canine model R o b e r t J. Pitsch, M D , D a v i d J. M i n i o n , M D , M a r g a r e t L. G o m a n , BS, J o h n A. van Aalst, M D , P a u l L. Fox, P h D , and L i n d a M. G r a h a m , M D ,

Cleveland, Ohio Purpose: Previous studies of grafts implanted in dogs documented a time-dependent increase in platelet-derived growth factor (PDGF) production that correlated with inner-capsule thickness. The purpose of this study was to identify the cells in vascular grafts that produce PDGF. Methods: Dacron thoracoabdominal grafts were seeded with autologous endothelial cells (ECs), implanted in 11 beagles, and removed 'after 4 or 20 weeks. ECs and smooth muscle cells (SMCs) were cultured from grafts and adjacent aorta, and PDGF in the conditioned media was measured by radioreceptor assay. The PDGF A-chain mRNA level in freshly harvested cells was assessed using reverse transcriptase, followed by polymerase chain reaction, and expressed as a ratio of glyceraldehyde-3-phosphate dehydrogenase signal. Localization of PDGF A-chain and B-chain protein was also examined with immunohistochemical analysis. Results: Graft and aortic ECs in primary culture did not produce significantly different amounts of PDGY in 72 hours, averaging 368 + 1 6 0 and 340 -+ 81 p g / ~ g DNA, respectively. Graft SMCs in primary culture produced significantly more PDGF than aortic SMCs (584 + 343 and 113 -+ 94 pg/Ixg DNA, respectively; p < 0.01). Graft SMC PDGF secretion remained greater than aortic SMC PDGF secretion through at least six cell passages. PDGF A-chain mRNA levels were not significantly different for aortic or graft ECs. The PDGF A-chain mRNA level was significantly higher for graft SMCs than aortic SMCs (2.44 +- 0.67 and 1.45 -+ 0.57 pg/ixg, respectively; p < 0.03). Immunocytochemical analysis detected PDGF A-chain and B-chain protein in the ECs from both native aorta and graft as well as the subendothelial SMCs in the graft, but not in the SMCs of the native aorta. Conclusions: These results suggest that graft SMCs are functionally altered, producing more PDGF than aortic SMCs. PDGF produced by graft SMCs may contribute to the development of intimal hyperplasia. (J Vasc Surg 1997;26:70-8.) Intimal hyperplasia is a recognized cause o f late graft failure. It is characterized by smooth muscle cell (SMC) accumulation and matrix deposition and is most prominent at the anastomoses o f prosthetic vascular grafts. The exact cause ofintimal hyperplasia is not clear, but many factors, including mechanical injury, altered shear stress, compliance mismatch, anastomotic flow disturbances, and mitogen release From the Department of Surgery, Case Western Reserve University and Department of Veterans Affairs Medical Center; and Department of Cell Biology, Cleveland Clinic Research Institute (Dr. Fox). Supported by grants from the NIH-NHLBI (HL-41178, HL54519) and the Department of Veterans Affairs. Reprint requests: Dr. Linda M. Graham, Surgery Service 112, VA Medical Center, 2215 Fuller Rd., Ann Arbor, MI 48105. 24/1/79341

70

from platclcts, lcukocytcs, and vessel wall cells, may play a role. A role for endothelial cell (EC)-derived mitogens in the development o f graft intimal hyperplasia was suggested by its occurrence at anastomoses, the very areas first covered with endothelium, and by the observation that SMCs proliferated only under ECs, at the anastomoses, and in areas o f transmural ingrowth where SMCs followed ECs migrating into the graft. 1,2 Additional support for the role o f ECderived mitogens in graft intimal hyperplasia was the finding that intimal cells removed from endothelialized grafts in baboons expressed more messenger RNA (mRNA) for platelet-derived growth factor (PDGF) A-chain than native artery. 3 Furthermore, in situ hybridization revealed P D G F A-chain m R N A in graft ECs and SMCs. 4 Previous studies in our labora-

JOURNAL OF VASCULAR SURGERY Volume 26, Number 1

tory demonstrated higher PDGF protein production in EC-seeded grafts compared with control grafts when removed 4 weeks after implantation, s In addition, PDGF production correlated directly with EC coverage, suggesting that ECs were the sourceofthe PDGF. 5 Interestingly, PDGF production continued to increase when measured in grafts halwested at 20 and 52 weeks; however, it did not correlate with EC coverage but corresponded with the appearance of SMCs in the inner capsule. 6,7 This finding suggested that a significant source of PDGF production in 20and 52-week grafts might be cells other than ECs. The purpose of this work was to identify the graft cells that produced PDGF and were responsible for the time-related increase in mitogen production after graft implantation. PDGF production by graft and aortic ECs and SMCs in vitro and in vivo was assessed by radioreceptor assay and immunohistochemical analysis, respectively, and PDGF gene expression was determined with reverse transcription linked with polymerase chain reaction (RT-PCR). These studies demonstrated that graft SMCs are distinctly different from aortic SMCs with respect to PDGF production, producing high levels of this mitogen. MATERIALS AND M E T H O D S To evaluate PDGF production by ECs and SMCs, adult female beagles underwent placement of knitted Dacron thoracoabdominal grafts seeded with autologous jugular vein ECs. Five grafts were removed 20 weeks after implantation to assess PDGF production by cells in culture. To evaluate mRNA expression for PDGF A-chain, three grafts were removed 4 weeks after implantation and three grafts were removed after 20 weeks. In addition, graft sections were examined using immunohistochemical analysis. Animals ranged in age from 71 to 199 weeks and weighed 7.5 to 12 kg. The mean age of dogs whose grafts were removed at 4 weeks and 20 weeks was not significantly different (165 + 19 weeks and 138 _+ 22 weeks, respectively; p = 0.285). The protocol for animal studies was approved by the institutional committee on animal use. Care complied with the "Principles of Laboratory Animal Care" and

Guide for the Care and Use of Laboratory Animals (National Institutes of Health publication No. 8623, revised 1985). Graft implantation and removal. Cooley Double-Velour knitted Dacron grafts (Meadox Medicals Inc.; Oakland, N.J.), 6 mm internal diameter and 22 cm long, were implanted as thoracoabdominal grafts as described previously, s Grafts were seeded with ECs enzymatically harvested from autologous exter-

Pitsch et al.

71

nal jugular veins and were added to the blood used to preclot the grafts. After 4 or 20 weeks the graft and native vessels were carefully isolated and flushed with medium 199 (M199) tissue culture medium (Sigma Chemical Co.; St. Louis). Tissue on the outer surface of the graft was removed during the harvest. Cell harvest. Cells were harvested from the proximal aorta (ascending aorta and aortic arch) and the distal abdominal aorta below the graft anastomosis. In addition, cells were isolated from four graft sections consisting of proximal anastomosis, proximal midgraft, distal midgraft, and distal anastomosis. The anastomotic graft segments included 5 mm of the adjacent artery. The luminal surface was incubated in collagenase A (630 units/ml) for 10 minutes at 37 ° C in a 5% CO 2 atmosphere, and ECs were removed mechanically using a rounded spatula. SMCs were harvested after removal of the EC layer from aortic and graft segments. The medial layer was removed mechanically, and the media and graft divided into 2 × 2 mm 2 pieces. The pieces were then incubated for 2 hours in an M199-based solution containing 15 units/ml elastase type III (Sigma), 170 units/ml collagenase A, 2 m g / m l crystalline bovine serum albumin (BSA; Boehringer Mannheim Biochemicals), 0.375 m g / m l soybean trypsin inhibitor (Boehringer Mannheim). SMCs were then collected, centrifuged at 300g for 5 minutes, and resuspended in M199. The homogeneity of EC and SMC isolates was confirmed by immunohistochemical analysis with antibody to von Willebrand factor (Chemicon International; Temecula, Calif.) and antibody to SMC ot-actin (Sigma). Cell culture. For analysis of PDGF production by ECs in primary culture, tissue culture plates were coated with fibronectin and ECs were grown to confluence in M199 containing 15 units/ml of heparin sodium, 0.15 m g / m l of endothelial cell growth supplement (ECGS) isolated from bovine hypothalami (Rockland; Gilbertsville, Pa.), 0.05 m g / m l ofgentamicin (Sigma), 2.2 m g / m l of sodium bicarbonate, and 20% fetal bovine serum (FBS; Hyclone, Provo, Utah). After cells reached confluence, medium was changed to M199 with 5% FBS and collected after 72 hours for measurement of PDGF production. SMCs were plated in FBS-coated tissue culture wells and grown to confluence in M199 with 20% FBS. The medium was then changed to M199 with 5% FBS for PDGF collection. In addition, SMCs were passaged after incubation in a 0.05% trypsin solution for 5 minutes at 37 ° C. PDGF assay. The PDGF in conditioned medium was measured by radioreceptor assay as de-

JOURNAL OF VASCULAR SURGERY

72 Pitsch et al.

scribed by Fox and DiCorleto, 9 which measures competition with t2SI-labeled PDGF for binding to PDGF receptors on human foreskin fibroblasts. Briefly, PDGF AB (Boehringer Mannheim) was radiolabeled according to the method of Heldin et al. 10 Sparse, quiescent cultures of human foreskin fibroblasts were incubated first with conditioned medium, then medium containing 12SI-labeled PDGF. Cellbound radioactivity was determined in a gammaradiation counter. A standard curve was prepared with known amounts of purified PDGF, and results were fit by nonlinear regression with the logistic equation. The PDGF production by graft and arterial cells was expressed as picograms PDGF per microgram DNA. DNA measurement. DNA in cultured cells was quantitated using the method described by Labarca and Paigen. 11 Briefly, cells were harvested with 0.05% trypsin solution and lysed with a sonicator. The cell lysate (25 Ftl) was then diluted to 750 Ixl with phosphate/NaCl buffered solution (4 m o l / L NaC1) and added to 750 ix1 of bis-benzamide solution. Emission was measured with a fluorimeter with excitation set at 356 nm and emission set at 458 nm. The standard curve was prepared with stock calf thymus DNA (Sigma), and results were fit by nonlinear regression with the logistic equation. Results were expressed as ~g of DNA per well. m R N A isolation. To study gene expression, ECs and SMCs were harvested from grafts and arteries in three dogs at 4 weeks and three dogs at 20 weeks after graft implantation. Isolation of polyA+ RIgA from ECs and SMCs was performed using the Micro-FastTrack mKNA Isolation Kit (Invitrogen Corp.; San Diego). Cells were lysed and the lysate applied to the oligo dT cellulose column. After highand low-salt buffer washes, mRNA was eluted from the oligo dT cellulose column with a salt-ftee buffer, precipitated, and stored at -70 ° C in 100% ethanol. Polymerase chain reaction. The PDGF A-chain and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA from aortic ECs and SMCs were amplified using RT-PCR. Reverse transcriptase, random hexamers, and deoxynucleotides were used to synthesize complementary DNA (cDNA) from isolated mRNA. PDGF A-chain and GAPDH eDNA were selccrively amplified using specific primers, 12,13 excess nucleotides, and a heat-stable DNA polymerase (Taq) for 30 cycles with a denaturing temperature of 92 ° C, an annealing temperature of 58 ° C, and an extension temperature of 72 ° C. The RTPCR products were separated on a 1% agarose gel and transferred to a GeneScreen nylon membrane (Du-

luly1997

pont NEN; Boston)?40ligonucleotide probes, 12,13labeled with [32p] using the T4 polynucleofide ldnase method, Is were established as specific for PDGF Achain and GAPDH in preliminary cxperimcnts, yielding a single, specific amplified product of the correct length. The Southern blots were hybridized overnight at 37 ° C, washed the following day, and exposed to I(odal~ XAR-5 film (Kodak; Rochester, N.Y.) or a phosphorimager screen. Comparison of gene expression was made possible by coamplification of the constitutively expressed gene GAPDH. Quantitative densitometry was performed on autoradiographs using radioactive [I4C] standards (American Radiolabeled Chemicals, Inc.; St. Louis) or with a phosphorimager (Molecular Dynamics; Sunnyvale, Calif.). Results were expressed as a ratio of the PDGF signal to the GAPDH signal. Immunocytochemlcal analysis. Tissue sections were stained for PDGF A-chain and B-chain protein using specific antibodies (Santa Cruz Biotechnology, Inc.; Santa Cruz, Calif.). Fresh tissue was immersionfixed in 2% paraformaldehyde/phosphate-buffered saline solution (PBS) for 5 hours and then was placed in 25% sucrose/PBS overnight. Tissue was then embedded in OCT compound, and 10 txm sections were cut with an AS620 cryotome (Shandon Lipshaw, Inc.; Pittsburgh). Immunoperoxidase staining was performed using the Vectastain Elite ABC Kit (Vector Laboratories, Inc.; Burlingame, Calif.). To quench endogenous peroxidase activity, slides were placed in 0.3% H202 in methanol for 30 minutes. After rinsing with PBS/0.1% BSA, slides were incubated in blocking serum for 30 minutes to lessen nonspecific staining. Primary antibody against PDGF A-chain or B-chain was applied at a 1:100 dilution. The slides were incubated with biotinylated antirabbit antibody and then with Vectastain ABC reagent for 30 minutes. Diaminobenzidine tetrahydrochloride (Polyscienees, Inc.; Warrington, Pa.) was used for the peroxidase enzyme substrate, and produced a reddish brown precipitate. Tissue was counterstained with Harris' hematoxylin stain without acetic acid (Richard-Allen Medical; Richland, Mich.). Controls included omission of the primary antibody and use of antibodies to PDGF A-chain and B-chain that had been preadsorbed with the respective ligand. RESULTS PDGF production by cultured cells. PDGF was mcasured in cells harvested from five dogs 20 weeks after graft implantation. Because of the variability in basal PDGF production bctwcen dogs, PDGF secretion (pg/ptg DNA) was normalized to

JOURNAL OF VASCULAR SURGERY Volume 26, Number 1

PDGF production by the ECs or SMCs harvested from arteries of the same dog. Pure primary EC cultures (n = 4) from graft and native aorta did not produce significantly different amounts of PDGF in 72 hours, the normalized values (mean _+ SEM) being 1.04 _+ 0.18 and 1.06 -+ 0.22, respectively (Table I). Primary cultures of graft SMCs (n = 5) produced significantly more PDGF in 72 hours than aortic SMCs (3.26 _+ 0.84 and 1.06 -+ 0.33, respectively; p < 0.05; Table I). PDGF production by aortic SMCs proximal and distal to the graft was similar to that by aortic SMCs from dogs that did not undergo graft implantation. ~6 To assess the duration of altered mitogen secretion, PDGF production by SMCs was measured during multiple passages of cells harvested from three dogs. The conditioned medium from graft and aortic SMCs in primary culture, as well as passages 2, 4, 6, and 8, was collected and assayed for PDGF secretion. Graft SMC PDGF production remained greater than aortic SMC PDGF production through at least six cell passages (Table II). With increasing numbers of cell passages, graft SMC PDGF production decreased whereas aortic SMC PDGF production increased, reaching nearly equivalent levels by passage 8 (1.43 -+ 0.70 vs 0.59 +_ 0.21, graft and aortic SMCs respectively; p = 0.27). Thus graft SMCs produced significantly more PDGF than aortic SMCs, and this characteristic did not disappear rapidly when both cell types were subjected to the same local environment of cell culture. P D G F A-chain m R N A expression. One possible explanation for the lack of difference in PDGF production by ECs is that they rapidly modulate in cell culture, resulting in loss of unique characteristics possessed in vivo. Therefore, as another method to identify the graft cells responsible for PDGF production and also to determine the level of regulation of PDGF secretion, ECs and SMCs were isolated from Dacron grafts 4 weeks (n = 3) and 20 weeks (n = 3) after implantation. Immunocytochemical analysis using antibodies to yon Willebrand factor and e~-actin was used to verify the identity of the cells, mRNA was then isolated, and PDGF gene expression determined using RT-PCR and represented as the ratio of PDGF A-chain signal to GAPDH signal (Fig. 1). PDGF gene expression was lower in Dacron graft ECs compared with aortic ECs at 4 weeks (0.90 -+ 0.45 and 1.61 _+ 0.67, respectively) and at 20 weeks (4.15 -+ 0.38 and 6.38 _+ 1.36, respectively). On the other hand, graft SMCs expressed more PDGF Achain mRNA than aortic SMCs at both 4 weeks (1.44 -+ 0.38 and 0.45 _+ 0.11, respectively) and 20

Pitsch et al.

73

Table I. Secretion of PDGF during 72 hours in primary culture

Proximal aorta Proximal anastomosis Midgraft Distal anastomosis Distal aorta Iliac artery

E C (n = 4)

S M C (n = 5)

0.86 0.99 1.10 1.54 1.48 0.77

1.34 2.74 4.27 2.01 1.06 1.28

-+ 0.07 -+ 0.17 -+ 0.25 _+ 0.40 -+ 0.57 +_ 0.50

± 0.64 ± 0.44 -+ 1.24 _+ 0.51 ± 0.38 ± 0.57

PDGF (pg/Ixg DNA) expressed as a ratio of mean PDGF secreLion by ECs or SMCs derived from native arteries in that dog.

weeks (3.33 + 0.74 and 2.34 + 0.86, respectively). When results for both 4- and 20-week Dacron grafts were combined, PDGF A-chain gene expression was significantly higher for graft SMCs than aortic SMCs (2.44 + 0.67 and 1.45 + 0.57, respectively; p < 0.05 by Wilcoxon signed rank). As seen in Fig. 2, the pattern of gene expression was consistent between dogs, with the mean ratio of graft to aortic PDGF A-chain/GAPDH RT-PCR signal being 0.62 + 0.08 for ECs and 2.34 + 0.36 for SMCs. Immunocytochemical analysis. To determine whether these changes in mRNA expression correlated with protein production in vivo, immunohistochemical analyses for PDGF A-chain and B-chain was performed. ECs from both graft and aortic segments stained positively for PDGF A-chain and B-chain proteins in a diffuse and comparable manner (Figs. 3 and 4). Both PDGF A-chain and B-chain were detected in SMCs from the graft segments, but in contrast, there was essentially no staining for PDGF A-chain or B-chain in SMCs of the native aorta (Figs. 3 and 4). The graft SMCs that stained for PDGF were often located just beneath the endothelium. No intracellular staining was observed in negative controls. These results suggest that differences found in PDGF production in vitro reflect changes seen in vivo, with graft SMCs producing more PDGF than aortic SMCs.

DISCUSSION Previous studies suggested that PDGF production by endothelialized prosthetic grafts was higher than native artery or control grafts, but the actual cellular source of the PDGF was not identified.4,s Results of the current study indicate that PDGF secretion by graft and aortic ECs is not significantly different and that PDGF production by ECs does not account for the higher levels of PDGF secreted from grafts segments compared with normal aorta. Although the similarity in PDGF secretion by graft and

JOURNAL OF VASCULAR SURGERY

74

July 1997

Pitsch et al.

S m o o t h M u s c l e Cells

E n d o t h e l i a l Cells A-Chain

GAPDH 0.45 0.79 1.33 2.42 0.68

o~

,
ox-

3.56 0.58 1.84 2.92 2.33

o" .e,,o~ , . ~ " .e,.o~

A-Chain GAPDH

~,.0"-

Fig. 1. Southern blots of SMC and EC PDGF A-chain and GAPDH mRNA amplified by RT-PCR. Aortic and graft ECs and SMCs are from a representative dog. Images are computerenhanced. A-chain/GAPDH numbers represent the ratios of signal intensities before enhancement.

Table I I . SMC Secretion o f P D G F (pg/Ixg DNA) during multiple passages

Proximal aorta Proximal anastomosis Midgraft Distal anastomosis Distal aorta Iliac artery

PO

P2

P4

P6

P8

1.36 _+0.91 3.14 ± 0.24 6.07 ± 1.01 2.78 _+0.29 1.56 _+0.36 0.51 _+0.51

1.02 _+0.07 4.67 _+.2.84 17.18 -+ 5.22 7.12 +_2.31 0.37 ± 0.19 1.83 ± 0.56

0.74 ± 0.20 1.77 ± 0.40 6.79 ± 1.54 4.43 ± 0.99 1.08 -+ 0.14 0.74 ± 0.01

0.63 _+0.22 2.80 _+ 1.68 7.39 ± 2.06 2.52 _+0.79 0.66 -+ 0.10 0.63 _+0.31

0.42 _+0.34 0.85 _+0.53 2.15 _+ 1.93 0.96 -+ 0.72 0.77 +- 0.37 0.44 ± 0.21

PDGF (pg/~g DNA) expressed as a ratio of mean PDGF secretion by SMCs derived from native arteries in that dog (n = 3). aortic ECs cultured from 20-week animals might be a result o f rapid EC modulation in primary passage causing cells to lose unique characteristics possessed in vivo, results o f immunohistochemical analysis and R T - P C R suggest that protein production and gene expression are similar in aortic and graft ECs in vivo. Development o f substantial subendothelial tissue, especially 20 weeks after graft implantation, may minimize the "foreign" environment o f the prosthetic graft, resulting in similar levels o f mitogen production by both graft and aortic ECs. Although ECs on grafts produce PDGF, our findings suggest that ECs are not responsible for increased PDGF secretion found in 20-week grafts compared with 4-week grafts. SMCs from 20-week grafts produce significantly more PDGF than aortic SMCs in primary cell culture. This characteristic persists through multiple passages, which suggests that it is not a transient effect of the local environment in the graft but that

graft SMCs have altered synthetic machinery compared with aortic SMCs. Furthermore, findings with immunohistochemical analysis and R T - P C R indicate that graft SMCs in vivo produce more PDGF and express higher levels m R N A for PDGF A-chain than aortic SMCs. These findings arc consistent with those o f Golden et al. 4 that documented higher levels o f PDGF A-chain m R N A in graft neointima than aorta, but the cells expressing the PDGF gene were not identified in that study. Increased PDGF secretion by graft SMCs compared with aortic SMCs, combined with previous findings o f a time-related increase in SMC number in the prosthetic graft neointima, suggest that the graft SMCs are responsible for the increased PDGF production by prosthetic grafts over time. 6,7 Furthermore, the R T - P C R data demonstrate that the increased PDGF production in graft SMCs is regulated at the level o f gene expression (i.e., either increased gene transcription or m R N A stability).

JOURNAL OF VASCULAR SURGERY Volume 26, Number 1

"~ 10

A

"E

--

<

IJ_

(.9 D

a.

Pitsch et al.

I~ ----B--- 4 week

----B---- 4 week

75

6

""

0

0

Aorta

Graft

Endothelial Cells

Aorta

Graft

Smooth Muscle Cells

Fig. 2. Graph of RT-PCR results shows PDGF A-chain/GAPDH signal intensity. Graft ECs express lower levels of PDGF A-chain mRNA than aortic ECs, whereas graft SMCs consistently express higher levels than aortic SMCs. The source of the SMCs on the graft is uncertain. They are probably not derived from the initially seeded cells, because jugular vein ECs maintained in culture are rarely contaminated with SMCs. SMCs may grow onto the graft from the adjacent artery or by transinterstitial growth, perhaps stimulated by chemotactic factors produced by ECs. These two sources of SMCs may explain differences seen in PDGF production by midgraft and anastomotic SMCs, with anastomotic SMCs producing levels of PDGF midway between midgraft and aortic SMCs. The elevated PDGF production by graft SMCs suggests that they are modulated to a secretory state, similar to SMCs found early in development or in pathologic states, such as atherosclerosis or injur y ) 7-21 Neonatal SMCs secrete high levels of PDGF, but as they mature they change in phenotype and stop producing measurable quantities of PDGF. 22 In addition, the type of PDGF produced by SMCs appears to be distinct for different stages of normal development, with rat pup SMCs expressing mRNA for A-chain and B-chain whereas uninjured adult rat SMCs express only mRNA for PDGF A-chain. 17 In pathologic states such as athcrosclerosis or intimal hyperplasia, adult SMCs appear to be capable of reverting to a "fetal" phenotype, with altered PDGF gene expression and mitogen production. A1-

though SMCs from normal arteries do not produce measurable quantities of PDGF, SMCs cultured from human atherosclerotic plaques produce PDGF and express mRNA for PDGF A-chain. 19 In situ hybridization studies demonstrate PDGF A-chain mRNA in the mesenchymal cells of atherosclerotic plaque, but not in normal artery. 2° Arterial injury in animal models also induces SMCs to revert to a juvenile state after injury, secreting PDGF and expressing mRNA for PDGF A-chain and B-chain) 7,18 These studies suggest that mature SMCs retain the ability to modulate their phenotype after injury or in disease states to produce increased amounts of PDGF. Relatively little is known about changes in SMC phenotype and PDGF production or genc expression after prosthetic graft implantation. Birinyi et al. 21 found that SMCs culturedfrom human graft hyperplastic lesions express mRNA for PDGF A-chain and secrete a PDGF-like molecule. By immunohistochemical analysis, Kraiss et al. 23 demonstrated PDGF A-chain in the SMC layers immediately subjacent to the endothelial surface of grafts removed from baboons. These studies and our findings of significant differences in graft and arterial SMC production of PDGF suggest that SMCs on grafts are modulated to a phenotype similar to that found after injury or in atherosclerosis.

JOURNAL OF VASCULAR SURGERY

76

luly 1997

Htsch et al.

Fig. 3. Immunohistochemical analysis using antibody to PDGF A-chain (original magnification, ×400). Aortic section (A) has less staining of SMCs than graft section (B).

Fig. 4. Immunohistochemical analysis using antibody to PDGF B-chain (original magnification, ×400). Aortic section (A) has less staining of SMCs than graft section (B).

One possible explanation for the unique phenotype o f graft SMCs compared with aortic SMCs is that SMCs that migrate onto the graft represent a small subpopulation of cells in the normal aorta that retain their proliferative and secretory capacity. In the bovine neonatal pulmonary artery, a distinct subpopulation o f SMCs proliferate in response to hypoxia. 24 Similarly, in balloon-injured rat carotid arteries, a subpopulation o f intimal cells express mRNA for PDGF B-chain. 2s These cells are occasionally

arranged in clusters, suggesting a c o m m o n precursor. Perhaps after prosthetic graft implantation, a subpopulation of normal arterial SMCs migrate onto the graft and proliferate there, and these cells are capable of secreting PDGF. If this is the case, however, one would expect homogeneous staining for PDGF on immunohistochemical analysis, but instead there is a definite increase in the subendothelial SMCs, similar to that found by Kraiss et al. 23 Perhaps a more plausible explanation is that the SMCs that

\

JOURNAL OF VASCULARSURGERY Volume 26, Number 1

migrate onto the graft and proliferate there are altered by the environment to express a secretory phenotype, and this does not quickly revert to a contractile phenotype. The production of PDGF may be important in the development of graft intimal hyperplasia for several reasons. If graft SMCs have the ability to respond to the PDGF they produce, continued proliferation may be stimulated in an autocrine fashion. Kraiss et al.2s demonstrated PDGF receptor [3-subunit mRNA and protein, and lower levels of the (x-subunit mP,NA, in the neointima of grafts in baboons. Although the response of graft SMCs to PDGF was not studied, the expression of mRNA for PDGF receptor suggests that the SMCs are capable of responding to PDGF. On the other hand, if graft SMCs do not have the ability to respond to PDGF, production of this mitogen may contribute to ongoing SMC migration onto the graft, especially at the anastomoses. Work by others suggests that PDGF contributes to intimal hyperplasia after mechanical injury. Infusion of antibody to PDGF reduced the area ofneointimal hyperplasia by 40.9% after balloon catheter injury of rat carotid arteries by interfering with migration of medial SMCs into the neointima. 26 Furthermore, increased amounts of PDGF, either by systemic infusion or gene transfer, 27,28 results in increased intimal thickening. Although these studies suggest that PDGF plays a role in SMC accumulation in the intima, the pathophysiologic significance of PDGF in the production ofintimal hyperplasia or the mechanism by which altered SMC production of PDGF might contribute to intimal hyperplasia is uncertain. Our studies suggest that SMCs that accumulate on prosthetic grafts are altered and have higher PDGF production and PDGF mRNA expression than aortic SMCs. Although graft healing in animal models is not identical to that in human beings, this altered phenotype in graft SMCs is similar to that seen in SMCs from human atherosclerotic lesions. If graft SMCs are also responsive to the effects of PDGF, autocrine growth stimulation may occur and contribute to the development of anastomotic intireal hyperplasia in grafts. Understanding of the mechanism of increased PDGF production would provide insight into the cause of intimal hyperplasia and promote the development of strategies to prolong graft patency. We gratefully acknowledge Meadox Medicals Inc., Oakland, N.J., for their donation of 6 mm internal diameter Cooley double-velour knitted Dacron grafts.

Pitsch et al.

77

REFERENCES 1. Clowes AW, Gown AM, Hanson SR, Reidy MA. Mechanisms of arterial graft failure. 1. Role of cellular proliferation in early healing of PTFE prostheses. Am J Pathol 1985;118:43-54. 2. Clowes AW, Kirkman TR, Reidy/VIA. Mechanisms of arterial graft healing: rapid transmural capillary ingrowth provides a source of intimal endothelium and smooth muscle in porous PTFE prostheses. Am J Pathoi 1986;123:220-30. 3. Golden MA, Au YPT, Kenagy RD, Clowes AW. Growth factor gene expression by intimal cells in healing polytetrafluoroethylene grafts. J Vasc Surg 1990;11:580-5. 4. Golden MA, Au YPT, Kirkman TR, Wilcox JN, Raines EW, Ross R, et al. Platelet-derived growth factor activity and mRNA expression in healing vascular grafts in baboons: association in vivo of platelet-derived growth factor mRNA and protein with cellular proliferation. J Clin Invest 1991;87:40614. 5. Kaufman BR, Fox PL, Graham LM. Platelet-derived growth factor production by canine aortic grafts seeded with endotheliai cells. J Vasc Surg 1992;15:699-707. 6. Kaufman BR, DeLuca DJ, Folsom DL, Mansell SL, Gorman ML, Fox PL, Graham LM. Elevated platelet-derived growth factor production by aortic grafts implanted on a long-term basis in a canine model. J Vasc Surg 1992;15:806q6. 7. Margolin DA, Kaufman BR, DeLuca DJ, Fox PL, Graham LM. Increased platelet-derived growth factor production and intimal thickening during healing of Dacron grafts in a canine model. J Vasc Surg 1993;17:858-67. 8. Graham LM, Vinter DW, Ford JW, ILahn RH, Burkel WE, Stanley JC. Endothelial ceil seeding of prosthetic vascular grafts: early experimental studies with cultured autologous canine endothelium. Arch Surg 1980;115:929-33. 9. Fox PL, DiCorleto PE. Regulation of production of a platelet-derived growth factor-like protein by cultured bovine aortic endothelial cells. J Cell Physiol 1984;121:298-308. 10. Heldin C-H, Westermark B, Wasteson A. Platelet-derived growth factor: purification and partial characterization. Proc Nail Acad Sci U S A 1979;76:3722-6. 11. Labarca C, Paigen K. A simple, rapid, and sensitive DNA assay procedure. Anal Biochem 1980;102:344-52. 12. Wang AM, Doyle MV, Mark DF. Quantitation of mRNA by the polymerase chain reaction. Proc Natl Acad Sci U S A 1989;86:9717-21. 13. Diamond SL, Sharefldn JB, Dieffenbach C, Frasier-Scott K, McIntire LV, Eskin SG. Tissue plasminogen activator messenger RNA levels increase in cultured human endothelial cells exposed to laminar shear stress. J Cell Physiol 1990;143: 364-71. 14. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a laboratory manual. Cold Spring Harbor: Cold Spring Harbor Laboratory Press, 1989;9.34-9.35. 15. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a laboratory manual. Cold Spring Harbor: Cold Spring Harbor Laboratory Press, 1989; 10.60-10.61. 16. Madura JAII, Kaufman BR, Margolin DA, Spencer DM, Fox PL, Graham LM. Regional differences in platelet-derived growth factor production by the canine aorta. J Vasc Res 1996;33:53-61. 17. Majesky MW, Giachelli CM, Reidy MA, Schwartz SM. Rat carotid neointimal smooth muscle cells reexpress a developmentally regulated mRNA phenotype during repair of arterial injury. Circ Res 1992;71:759-68. 18. Walker LN, Bowen-Pope DF, Ross R, Reidy MA. Production

78

19.

20.

21.

22.

23.

Pitsch ¢t al.

of platelet-derived growth factor-like molecules by cultured arterial smooth muscle cells accompanies proliferation after arterial injury. Proc N a t Acad Sci U S A 1986;83:7311-5. Libby P, Warner SJC, Salomon RN, Birinyi LK. Production of platelet-derived growth factor-like mitogen by smooth muscle cells from human atheroma. N Engl J Med 1988;318: 1493-8. Wilcox JN, Smith KM, Williams LT, Schwartz SM, Gordon D. Platelet-derived growth factor mRNA detection in human atherosclerotic plaques by in situ hybridization. J Clin Invest 1988;82:1134-43. Birinyi LK, Warner SJC, Salomon RN, Callow AD, Libby P. Observations on human smooth muscle cell cultures from hyperplastic lesions of prosthetic bypass grafts: production of a platelet-derived growth factor-like mitogen and expression of a gene for a platelet-derived growth factor receptor--a preliminary study. J Vase Surg 1989;10:157-65. Seifert RA, Schwartz SM, Bowen-Pope DF. Developmentally regulated production of platelet-derived growth factor-like molecules. Nature 1984;311:669-71. Kralss LW, Raincs EW, Wilcox JN, Seifcrt RA, Barrett TB, Kirkman TR, et al. Regional expression of the platelet-derived growth factor and its receptors in a primate graft model of vessel wall assembly. J Clin Invest 1993;92:338-48.

JOURNAL OF VASCULARSURGERY ~uly i997

24. Wohrley JD, Frid MG, Moiseeva EP, Orton EC, Belknap JK, Stenmark KR. Hypoxia selectively induces proliferation in a specific subpopulation of smooth muscle cells in the bovine neonatal pulmonary arterial media. J Clin Invest 1995;96: 273-81. 25. Lindner V, Giachelli CM, Schwartz SM, Reidy MA. A subpopulation of smooth muscle cells in injured rat arteries expresses platelet-derived growth factor-B chain mRNA. Circ Res 1995;76:951-7. 26. Ferns GAA, Raines EW, Sprugel KH, Motani AS, Reidy MA, Ross R. Inhibition of neointimal smooth muscle accumulation after angioplasty by an antibody to PDGF. Science 1991; 253:1129-32. 27. Nabel EG, Yang Z, Liptay S, San H, Gordon D, Haudenschild CC, et al. Recombinant platelet-derived growth factor B gene expression in porcine arteries induces intimal hyperplasia in vivo. J Clin Invest 1993;91:1822-9. 28. Jawien A, Bowen-Pope DF, Lindner V, Schwartz SM, Clowes AW. Platelet-derived growth factor promotes smooth muscle migration and intimal thickening in a rat model of balloon angioplasty. J Clin Invest 1992;89:507-11.

Submitted Sep. 10, 1996; accepted Nov. 20, 1997.