Improved in vivo endothelialization of prosthetic grafts by surface modification with fibronectin☆☆☆
Article Outline
Abstract
Endothelial cell growth in vitro is enhanced by coating with fibronectin the surface on which cells grow. Similar coating of prosthetic arterial grafts may promote in vivo graft endothelialization if graft patency is not adversely affected. In each of 15 dogs, two fibronectin-coated polytetrafluoroethylene grafts and two grafts that were not coated were implanted. One graft in each pair was seeded with autologous endothelial cells, so that four different grafts were studied in each animal: a coated, seeded graft; a coated graft that was not seeded; a seeded graft that was not coated; a graft that was neither coated nor seeded. At 2, 4, and 8 weeks, grafts from five animals were examined for patency, surface endothelialization, and indium 111 platelet reactivity. After seeding, surface coverage by endothelium of coated grafts was more complete and more rapid than in uncoated grafts (64% ± 23% vs 31% ± 13% at 4 weeks, p < 0.05). Without seeding, coated grafts also appeared to have increased endothelial cell ingrowth compared with plain grafts (48.8% ± 15.1% vs 37.6% ± 1.5% at 8 weeks). Early (2-week) platelet reactivity of coated grafts was increased (p = 0.06), but patency was not adversely affected. Thus fibronectin coating of prosthetic grafts promotes surface endothelialization in vivo without altering graft patency. (J VASC SURG 1988;8:476-82.)
Prosthetic graft endothelialization does not occur in humans.1 Intimal ingrowth stops after a few centimeters 88despite the presence of endothelial cells capable of migration and multiplication.2 Graft endothelialization can be achieved by the transplantation of autologous endothelial cells onto vascular prostheses (seeding),3 which shows that endothelial cells can grow on the middle of prosthetic grafts. However, large numbers of cells are required for successful seeding,4 which has led to the suggestion that prosthetic graft luminal surfaces may be relatively “inhospitable” to endothelial cell multiplication and migration.5
In vitro, endothelial cell growth is stimulated by several factors, including the adhesive glycoprotein fibronectin.6 Coating prosthetic grafts with fibronectin improves the initial adhesion of seeded endothelial cells to graft luminal surfaces7 and increases retention of seeded cells within grafts after exposure to blood flow.8 In addition, endothelial cell migration on fibronectin-coated prosthetic surfaces is higher than on uncoated surfaces.5 The improvement of these necessary events for endothelial cell growth (adhesion and migration) should promote endothelialization of seeded grafts after graft implantation, but this has yet to be documented.
The purpose of this study was to determine whether coating polytetrafluoroethylene (PTFE) grafts with fibronectin would improve the rapidity of graft endothelialization after endothelial cell seeding. In addition, the effect of fibronectin coating on pannus ingrowth into grafts that were not seeded was studied. Finally, since coating PTFE grafts with fibronectin has previously been shown to increase initial graft platelet accumulation,7 platelet reactivity and patency of coated grafts were investigated and compared with results from plain PTFE grafts.
Material and methods
In each of 15 dogs, two fibronectin-coated, standard PTFE grafts and two plain PTFE grafts were implanted as arterial interposition grafts in both carotid arteries and both femoral arteries. Animal care complied with “Principles of Laboratory Animal Care” and the “Guide for the Care and Use of Laboratory Animals” (National Institutes of Health Publication No. 80-23, revised 1978). One graft in each pair (coated and plain) was seeded with autologous endothelial cells and thus four different experimental settings were studied in each animal: a coated, seeded graft (CS); a coated graft that was not seeded (CUS); a graft that was not coated but was seeded (UCS); a graft that was neither coated nor seeded (UCUS). Each animal was sedated with sodium thiamylal (Surital) and anesthesia was maintained with endotracheal halothane. A 5 cm segment of 4 mm internal diameter (ID) coated or plain PTFE graft was implanted in an end-to-end fashion with standard surgical technique and anticoagulation with heparin (100 U/kg) during graft implantation. The location of coated and plain grafts was alternated between the carotid and femoral arteries. Animals were treated with dipyridamole (Persantine) (50 mg twice daily) for 4 days before graft implantation and aspirin (325 mg twice daily) for one day before surgery. Antiplatelet agents were also continued after graft implantation throughout the period of the study. In addition, each animal was treated with sulfamethoxazole and trimethoprim (Septra DS—one tablet every day) for 2 weeks after surgery to decrease the risk of graft infection.
Fibronectin-coated grafts were prepared the day of graft implantation. Human fibronectin dissolved in medium 199 (50 μg/ml) was added to sterile PTFE grafts and the grafts were incubated for one hour at 37° C, 5% carbon dioxide. Coated grafts were then opened, flushed with sterile medium 199, and immediately either seeded with autologous endothelial cells or brought to the operating field for implantation.
One week before graft implantation, endothelial cells from one jugular vein from each animal were harvested with collagenase as previously described.7 Harvested cells were then grown to confluence in fibronectin-coated polystyrene culture flasks with medium 199 supplemented with 20% fetal calf serum, penicillin (100 U/ml), amphotericin B (Fungizone, 2 μg/ml), streptomycin (100 ng/ml), transferrin (5 μg/ml), seleneum (5 ng/ml), and insulin (5 μg/ml) at 37° C, 5% carbon dioxide. At the time of seeding, cells were removed from culture flasks by incubation with trypsin (0.05%) and EDTA (0.02%) in phosphate-buffered saline solution. An aliquot of these cells was examined for viability by Evans blue exclusion and counted by means of a hemocytometer to determine cell density.
Five hundred thousand cells (3.4 × 104 cells/cm2 of graft surface area) were seeded into previously prepared fibronectin-coated grafts and plain grafts of similar length (5 cm, 4 mm ID). Seeded grafts were sealed and incubated for 1 hour at 37° C, 5% carbon dioxide, rotating grafts 90 degrees every 15 minutes to improve cell distribution. The grafts were then opened, flushed with sterile culture medium, and immediately implanted into the animal from which the cells had been harvested.
Two, four, and eight weeks after graft implantation, grafts from five animals were examined for patency, surface endothelialization, and 8 indium 111 platelet reactivity. Twenty-four hours before graft harvest, autologous platelet labeling with indium 111 oxine was accomplished with a modification of the method of Heaton et al.,9 and the labeled platelets were reinjected. At the time of graft harvest, the artery proximal and distal to each graft was exposed, taking care not to disturb the graft, and cannulated with a 16-gauge catheter. Grafts were then perfused in situ at 100 mm Hg with Sorensen's phosphate buffer followed by 1% glutaraldehyde in phosphate buffer. After fixation had been accomplished, grafts were dissected free from the surrounding tissue and carefully washed. Each excised graft was then divided into five equal sections to include the two anastomotic sites and the proximal, middle, and distal graft segments. The weight of each segment was determined, segment radioactivity measured in a gamma well counter, and platelet specific activity expressed as counts per minute per milligram of graft tissue. Graft sections were then stored in Trumps' fixative for 2 weeks to allow decay of indium 111 radioactivity before further processing.
Samples were processed for scanning electron microscopy by postfixation in 1% osmium tetroxide and dehydration in progressive concentrations of ethanol. Samples in 100% ethanol were critical-point dried with a Denton vacuum and coated with gold palladium with a Tecknics Hummer IV. Representative specimens of each of the segments (accounting for approximately half of the circumference of the segment) were then mounted and examined with a JEOL JSM-35C scanning electron microscope. An overall picture of each specimen was obtained with magnification ×20. In addition, three standard locations along the specimen were photographed with magnification ×200 to allow identification of surface morphology.
With the morphologic criteria of Jaffe10 to identify endothelium ultrastructurally, all photographs were graded in a blinded fashion for degree of surface endothelialization, first from the low power magnification and second from the three selected higher power fields. These two estimates of surface endothelialization were then compared and the specimen was reexamined if the estimates varied by more than 5%. Overall graft endothelialization was determined by assuming that each anastomotic section accounted for 10% of the overall graft surface area (because of the tapered anastomosis) while each middle section of the graft accounted for 26.3% of the graft surface.
Results are expressed as mean ± standard error of the mean. Differences in surface endothelialization, graft platelet reactivity, and overall patency within the experimental groups were analyzed with analysis of variance or chi-square analysis for overall comparisons and Student's t test for individual pair comparison. Differences were only considered significant at the 95% confidence level.
Results
Endothelialization of fibronectin-coated grafts was more rapid than cellular coverage of plain grafts after seeding (Table I).
Table I. Percentage of surface coverage by endothelium (mean + SEM)
| 2 wk | 4 wk | 8 wk | |
|---|---|---|---|
| CS | 35.2 + 12.2 | 64 ± 10.4*† | 68.4 ± 7.5‡ |
| UCS | 26 ± 9.7 | 30.8 ± 6.5* | 73.2 ± 11.1‡ |
| CUS | 16.6 ± 4.3 | 29.0 ± 4.1 | 48.8 ± 15.1 |
| UCUS | 15.3 ± 3.8 | 43.3 ± 2.0† | 37.6 ± 1.5 |
| *CS vs UCS, p < 0.05. †CS vs UCUS, p = 0.06. ‡p < 0.01, vs UCUS. | |||
Fig. 1.
Middle sections from fibronectin-coated (A) and plain (B) PTFE grafts 4 weeks after seeding and implantation. Increased coverage by cells that resemble endothelial cells is seen on the coated graft (original magnification ×200).
At 8 weeks, improved surface coverage by endothelium was also found in unseeded, fibronectin-coated grafts compared with unseeded, plain grafts (Fig. 2).
Fig. 2.
Middle sections from fibronectin-coated (A) and plain (B) PTFE grafts that were not seeded 8 weeks after graft implantation. Almost complete surface coverage by cells resembling endothelial cells can be seen on the coated graft. No identifiable cells seen on the plain graft (original magnification ×200).
When individual values of graft surface coverage by endothelium are considered rather than group means, the impact of fibronectin coating on graft endothelialization is even more evident. By 4 weeks, coverage of the graft surface by endothelium exceeded 50% in four of five coated grafts that were seeded whereas none of the grafts in the other experimental groups were found to have that degree of surface coverage (Table II).
Table II. Grafts with greater than 50% endothelialization
| 2 wk | 4 wk | 8 wk | Overall | |
|---|---|---|---|---|
| CS | 20% | 80% | 80% | 80%* |
| UCS | 0 | 0 | 80% | 44% |
| CUS | 0 | 0 | 60% | 33%† |
| UCUS | 0 | 0 | 0 | 0 |
| *CS vs UCS, p < 0.05. †CUS vs UCUS, p = 0.06. | ||||
Two weeks after graft implantation, platelet accumulation over a 24-hour period was higher on coated grafts than on uncoated grafts (Table III).
Table III. Graft platelet accumulation (indium 111 activity/mg/min)
| 2 wk | 4 wk | 8 wk | |
|---|---|---|---|
| CS | 2393 ± 1930* | 532 ± 260 | 543 ± 255 |
| UCS | 91 ± 27 | 292 ± 64 | 543 ± 283 |
| CUS | 1720 ± 12971 | 758 ± 284 | 758 ± 284 |
| UCUS | 178 ± 77 | 841 ± 195 | 526 ± 173 |
| *Coated grafts vs uncoated grafts, p = 0.06. | |||
Fig. 3.
Middle sections from fibronectin-coated (A) and plain (B) PTFE grafts that were not seeded, 2 weeks after graft implantation. Surface coverage by thrombus (by high power consistent with fibrin, platelet, and red cell coagulum) on the coated graft is significantly increased compared with the plain graft.
Table IV. Graft patency
| 2 wk (%) | 4 wk (%) | 8 wk (%) | Overall (%) | |
|---|---|---|---|---|
| CS | 100 | 100 | 100 | 100 |
| UCS | 80 | 80 | 100 | 87 |
| CUS | 100 | 80 | 100 | 93 |
| UCUS | 80 | 80 | 100 | 87 |
Discussion
The reason that the ingrowth of endothelial cells into prosthetic grafts stops after a short distance is unknown. Wounding of endothelial cells in pannus ingrowth results in thymidine uptake and cell regrowth, which would be impossible if cells were senescent.2 In addition, Watkins et al.11 have reported that endothelial cells derived from adult saphenous vein undergo up to 58 population doublings in vitro, sufficient to cover the surface area of most commonly used prosthetic grafts. In light of these observations, it is reasonable to assume that surface properties and surface chemistry of the prosthesis must be important determinants of the arrest of endothelial cell ingrowth.
Prosthetic graft endothelialization should be enhanced by modifying prosthetic graft luminal surfaces to promote endothelial cell attachment, migration, and multiplication. The results from this study support that concept. Both endothelialization after seeding and endothelial cell ingrowth into unseeded grafts were improved by coating the luminal surface of PTFE grafts with fibronectin. The morphologic observations used in this study to quantify cellular coverage cannot conclusively determine the type of cells seen on the luminal surface of grafts and distinguish endothelial cells from myointimal cells. However, comparisons of the degree of cellular coverage in coated and plain grafts in the same animal demonstrate the ability of fibronectin to stimulate the growth of cells that resemble endothelium on prosthetic surfaces in vivo.
Fibronectin is an important endothelial cell attachment protein and attachment is necessary for endothelial cell growth.6 Seeding density that is improved by fibronectin coating also influences the rapidity of endothelial cell growth as shown by the report of McCall et al.12 that increasing seeding density significantly decreases the time required for growth to confluence. Finally, fibronectin stimulates endothelial cells to synthesize and secrete basement membrane components13; Madri and Stenn14 have shown that endothelial cell migration is preceded by the secretion of types IV and V collagens and laminin. Any one or a combination of these factors is probably important in the increased migration and growth of endothelial cells on fibronectin-coated grafts.
Ramalanjaona et al.8 estimated that fibronectin coating of PTFE grafts might reduce the time interval for graft endothelialization after seeding from between 5 and 7 weeks to between 3 and 4 weeks. This agrees well with the results seen in this study; to obtain approximately 70% surface coverage after seeding required 4 weeks in coated grafts compared with 8 weeks in uncoated grafts. This decrease in the “period of vulnerability” for thrombosis after seeding was associated with an increased platelet reactivity of fibronectin-coated grafts during the period of graft endothelialization despite pretreatment with antiplatelet agents. This finding may appear to contradict the report by Kempczinski et al.15 that antiplatelet agents reduced platelet accumulation on fibronectin-coated grafts to a level comparable to that on plain grafts. However, in that study platelet accumulation on fibronectin-coated grafts was determined in animals treated with antiplatelet agents whereas platelet accumulation on plain grafts was measured in untreated animals.
Increased graft platelet reactivity of fibronectin-coated grafts in the present study was only seen 2 weeks after graft implantation and graft patency was unaffected in this canine model where the animals were treated with aspirin and dipyridamole. Because antiplatelet agents may not be as effective in preventing graft thrombosis in humans,16 the effect of fibronectin coating on early graft patency in that setting remains to be determined.
Recent reports have demonstrated the possibility of obtaining endothelialization in prosthetic grafts without the complexities of seeding. Zacharias et al.17 have shown that porous PTFE grafts can heal by transinterstitial ingrowth after implantation in baboons. Malone et al.18 have also demonstrated endothelialization of detergent-extracted biologic grafts implanted in dogs that tend to heal arterial grafts poorly. These findings suggest that given proper graft configuration or luminal surface characteristics, “spontaneous” endothelialization is possible. The finding that fibronectin coating of arterial grafts improves endothelialization both with and without seeding suggests an additional factor that might be useful in promoting graft surface endothelialization, without endothelial cell transplantation.
References
- . Interspecies healing of porous arterial prosthesis: observations (1960-1974). Arch Surg. 1974;109:698–705
- . Endothelial regeneration. V. Inhibition of endothelial regrowth in arteries of rat and rabbit. Lab Invest. 1983;49:569–575
- . Endothelial seeding of enzymatically derived and cultured cells on prosthetic grafts. In: Stanley JC, Burkel WE, Lindnauer SN, Bartlett RH, Turcotte JG editor. Biologic and synthetic vascular prostheses. New York: Grune & Stratton, Inc; 1982;p. 631–652
- . Seeding endothelium on canine arterial prosthesis—the size of the inoculum. J Surg Res. 1980;28:35–38
- . Migration of adult human vascular endothelial cells: effect of extracellular matrix proteins. Surgery. 1986;100:384–390
- . Fibronectins: multi-functional modular glycoproteins. J Cell Biol. 1982;95:369–377
- . Improved endothelial cell seeding with cultured cells in fibronectin-coated grafts. J Surg Res. 1985;38:641–647
- . The effect of fibronectin coating on endothelial cell kinetics in polytetrafluoroethylene grafts. J Vasc Surg. 1986;3:264–272
- Indium: a new radionuclide label for studying human platelet kinetics. Br J Haematol. 1979;42:613–622
- . Culture of human endothelial cells. XII Transplant Proc. 1980;3(Suppl 1):49–53
- Adult human saphenous vein endothelial cells. Assessment of their reproductive capacity for use in endothelial seeding of vascular prostheses. J Surg Res. 1984;36:588–596
- . The culture of vascular endothelial cells to confluence on microporous membranes. Thromb Res. 1981;24:417–431
- . Basal lamina formation by cultured microvascular endothelial cells. J Cell Biol. 1984;99:692–698
- . Aortic endothelial cell migration. I. Matrix requirements and composition. Am J Pathol. 1982;106:180–186
- . Thrombogenicity of fibronectin-coated, experimental polytetrafluoroethylene graft. Surgery. 1987;101:439–444
- The effect of aspirin and dipyridamole on the patency of lower extremity bypass grafts. Surgery. 1984;96:462–466
- . Mechanisms of healing in synthetic grafts. J Vasc Surg. 1987;6:429–436
- . Detergent-extracted small-diameter vascular prostheses. J Vasc Surg. 1984;1:181–191
☆ Supported by funds from the Veterans Administration Research Service.
☆☆ Reprint requests: James M. Seeger, MD, VA Medical Center, Surgical Service (112), Gainesville, FL 32602.
PII: 0741-5214(88)90114-0
doi:10.1067/mva.1988.avs0080476
© 1988 Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. Published by Elsevier Inc. All rights reserved.
