An improved method for endothelial cell seeding on polytetrafluoroethylene small caliber vascular grafts _&viva Schneider, MSc, Raphael N. Melmed, M D , Herzl Schwalb, P h D , Matthias Karck, ME), Israel Vlodavsky, P h i ) , and Gideon Uretzky, M D ,
Jerusalem, Israel The creation of nonthrombogenic synthetic surfaces is a major challenge in biomedical research. The feature that clearly distinguishes natural blood vessels from their artificial cotmterparts is the presence of endothelial cell lining that besides being nonthrombogenic is capable of repair and renewal. This study describes a method of coating vascular grafts with a uniform, naturally produced subendothelial extracellular matrix before implantation. This substrate provides a most suitable biolayer for endothelial cell adhesion, growth, and differentiation, as compared with grafts coated with fibronectin or basement membrane extracts. It contains both adhesive glycoproteins (fibronectin, laminin, collagen) and proteoglycans (heparan sulfate) as well as endothelial cell growth factors (basic fibroblast growth factor) that support adhesion and normal growth of suboptimal concentrations of endothelial cells. We suggest that the presence in extracellular matrix of both adhesive macromolecules and potent endothelial cell-growth promoting factors will make the extracellular matrix a promising substrate for vascular grafts, (J VASC SURG 1992;15:649-56.)
The earliest attempts at vascular grafting used veins as replacement material. However, in the 1950s the development of synthetic material, and in particular Dacron and Teflon graft material, gave an enormous :impetus to the field of vascular surgery. With this, it soon became evident that the synthetic graft material was well suited to large artery grafting, marginally suited to medium artery grafting, and not at all suited to small diameter artery grafting. The principal reason for this unsuitability lies in the tendency of the synthetic material to induce thrombus format]on and occlusion of the graft lumen.1 The last decade particularly has seen significant advances made in tTwo parallel lines of biologic research relevant to this field. For one, the endothelial cell
From the Mayer Mitchell Cell BiologyLaboratory,The Joseph Lunenfeld Cardiac Surgery Research Center (Drs. Schneider, Melmed, Schwalb, Karck, and Uretzky), Department of Radiation and Clinical Oncology (Dr. Vlodavsky),Department of Medicine B (Dr. Melmed), Hadassah University Hospital, Jerusalem, ~[srael. Supported in part by the Israel Academy of Sciences and Humanities and by a grant from the Ministryof Science and Technolog)r, Israel, and the GSF, Munich, Germany. Reprint requests: Gideon Uretzky, MD, the Joseph Lunenfeld Cardiac SurgeryResearch Center, HadassahUniversityHospital, P.O. Box 12000, Jerusalem91120, Israel. 24/1/33975
(EC) is now recognized to be active in the production of a growing list of factors important in the regulation of blood fluidity.2 In addition to any physical role, the endothelium may prevent the deposition of fibrin and platelets on subendothelial vascular tissue denuded of ECs. The other line of research concerns the refinement of techniques for in vitro growth of ECs and their seeding on synthetic graft material, a,4 However, it is recognized that with the seeding techniques presently used, approximately 90% of seeded cells are lost from the vascular prosthesis within 24 hours of implantation, s This occurs in spite of prior treatment of graft material with either fibronectin or albumin to increase the plating efficiency. An additional problem with cell seeding is that a large number of donor cells are needed to obtain a reasonable endothelial fining of the graft material. 6 We have been interested in developing a more efficient biosubstrate to try to enhance the efficiency of adhesion of seeded ECs and to reduce the tendency of platelet adhesion to the new graft, which predisposes to thrombus formation. We decided to investigate the use of a naturally produced subendothelial extracellular matrix (ECM) as an adhesion and growth-promoting substrate for ECs on synthetic graft material. This ECM is produced in large amounts by cultured corneal ECs and closely resem649
650 Schneider et aL
Journal of VASCULAR SURGERY L
diameter (6 mm) polytetrafluoroethylene (PTFE; W. L. Gore & Assoc.; Elkton, Md.) cylinders.
Fig. 1. Rotation device for lining small caliber cylinders of Gore-Tex (PTFE) prosthesis with ECM and endothelium. The device consists of four axes rotating at a velocity of four rotations/hour, to guarantee even distribution of adhering ECs. PTFE cylinders (6 cm length x 0.6 cm diameter) were used for lining experiments. Bovine corneal cells in DMEM containing 5% FCS, 10% CS, 4% dextran, and 10 ng/ml FGF were seeded at a concentration of 5 × 105 cells/cm2 on segments first coated with fibronectin. bles the subendothelium in vivo in its organization and characteristic constituentsl It contains primarily collagens (mostly type Ili and IV, with smaller amounts of type I and V), proteoglycans (mostly heparan sulfate-proteoglycans and dermatan sulfateproteoglycans), laminin, nidogen, fibronectin, and elastin. 7 The corneal cells are easy to maintain in culture, and their ECM, unlike that produced by fibroblasts and smooth muscle cells, is deposited in a polar fashion, exclusively underneath the EC monolayer. Because it is firmly attached to the entire area of the culture substrate, the EC layer can be readily removed, leaving the underlying ECM intact and free of nuclei, cytoskeletal elements, and cellular debris. 7,~ Moreover, this ECM has recently been shown to contain basic fibroblast growth factor (bFGF) that is stable and firmly bound to heparan sulfate. 8 The presence of both adhesive macromolecules and potent EC growth-promoting factors in ECM makes it a promising substrate for vascular grafts. This article describes a technique we have devised to provide seeded vascular ECs with a natural ECM by the prior growth of corneal ECs on the graft material. The cells that produce ECM are then removed, leaving the underlying ECM firmly attached to the synthetic substrate. A monolayer of vascular ECs can then be grown to confluence on the matrix (provided by the corneal endothelium) in small
MATERIAL A N D M E T H O D S Dulbecco's modified Eagle medium (DMEM containing 1 gm and 4.5 gm glucose/L), medium199, bovine calf serum (CS), fetal calf serum (FCS), glutamine, phosphate-buffered saline (PBS) containing 0.05% trypsin and 0.02% ethylenediamine tetraacetic acid (STV), penicillin, and streptomycin, were obtained from Gibco (Grand Island, N.Y.). Brain FGF was from Biomedical Technologies (Stoughton, Mass.). Dextran (T-40) was from Pharmacia Fine Chemicals (Uppsala, Sweden). Tissue culture dishes were obtained from Falcon Labware Division, Becton Dickinson (Oxnard, Calif.), and four-well tissue culture plates were from Nm(Roskilde, Denmark). Matrigel extract of basement membrane-like matrix produced by EngelbrethHolm-Swarm (EHS) tumor was obtained from Collaborative Research Inc. (Bedford Mass.). Fibronectin, and all other chemicals of reagent grade were purchased from Sigma Chemicals (St. Louis, Mo.). Gore-Tex PTFE vascular graft material was purchased from W. L. Gore & Assoc. (Munich, Germany). Cell culture Bovine aortic ECs. Clonal populations of adult bovine aortic ECs were obtained as described. 9,1° Cells were routinely cultured in DMEM (1 gm glucose/L), supplemented with 10% bovine CS glutamine, and antibiotics at 37 ° C in an 8% CO 2 atmosphere humidified incubator. Cells were dissociated with STV and passaged weekly at a split ratio of i : 6, and bFGF (10 ng/ml) was added every other day until the cells were nearly confluent. Confluent cultures composed of flattened, closely apposed, and nonoverlapping cells were kept without further addition of FGF. Cells were grown in 90 mm dishes (Falcon). Bovine corneal ECs. A primary cell culture was established from steers' eyes as described, l~ and stock cultures maintained as the vascular ECs but with the addition of 5% FCS and 4% dextran (T-40) to the growth medium. H u m a n colon carcinoma cells. A cell line (HS703T) 7 was cultured in DMEM (4.5 gm glucose/L) supplemented with 10% FCS, insulin (5 >g/ml), penicillin (50 units/ml) and streptomycin (50 ~g/ml). The cells were loosely attached to the tissue culture dish and were detached by repeated pipetting.
Volume 15 Number 4 April 1992
Coating of PTFE w i t h fibronectin, collagen, or Matrigel Small pieces o f PTFE (15 m m diameter) were inserted into Nunc four-well plates (15 m m well diameter) and immobilized by narrow Teflon rings. These were coated with either PBS or fibronectin (3 ixg/cm2) in PBS by incubation for 1/2 hour at 37 ° C, followed by three washes with PBS. r This step preceeded the seeding o f the corneal ECs. A similar effect on cell adhesion was obtained when the graft material was preincubated with higher concentrations o f fibronectin, up to 15 >g/cm 2. T o coat the PTFE with collagen, 0.25 ml collagen solution, prepared as described, 12'13was allowed to solidify on the PTFE disks at 37 ° C for 30 minutes. The excess of the gel was removed by suction, leaving a thin )llagen film on the PTFE. In studies with disks coated with both collagen and fibronectin, collagen was first coated onto the disk as described above. Matrigel (10 mg/ml), an extract o f basement membranes produced by the E H S tumor, was allowed to solidify on PTFE disks at 37 ° C for 30 minutes. Excess material was removed by suction, leaving a thin film on the PTFE.
Preparation o f ECM coated PTFE Bovine ,corneal endothelial cells were suspended in D M E M containing 10% CS, 5% FCS, and 4% dextran-T40. Cells were seeded at a concentration o f 1 × l 0 s cells or 2 × 10 ~ cells/15 m m wells. The PTFE in the dishes was previously coated with fibronectin or collagen. The cells were cultured in 8% CO 2 incubator for 14 days. Fibroblast growth factor (10 ng/ml) was added every other day until the cells reached confluency. Fourteen days a~er seeding, the attached corneal cells were removed by addition of 0.5% Triton X-100 (Sigma) in PBS for 10 minutes; at room temperature, leaving the underlying ECM intact and firmly attached to the entire P T F E disk. Remaining nuclei and cytoskeletons were removed by 2 to 3 minutes of exposure to 0.025 rnol/L N H 4 O H followed by four washes with PBS.
Attachment assay o f colon carcinoma cells Colon carcinoma cells suspended in D M E M at a concentration o f 2.5 × 10 s cells/ml were added to uncoated PTFE or PTFE coated with ECM. After 40 minutes ofiincubation at 37 ° C in 8% CO 2 incubator, unattached cells o f triplicate cultures were removed by gently pipetting and rinsing in PBS. The remaining firmly attached cells were dissociated with STV
Seeding of endothelial cells on a vascular prosthesis
Fig. 2. Proliferation of bovine corneal ECs seeded on tissue culture plastic or PTFE. Bovine ECs were seeded (2.5 x 104 cells/well) on tissue culture wells or PTFE segments (precoated with various substrates). Seven days after seeding, the cells were dissociated with STV and counted. A, Nontreated; B, Fibronectin; C, Collagen; D, Fibronectin + collagen.
and counted in a Coulter counter (model ZM) (Coulter Electronics, Luton, Beds, England).
Attachment and proliferation o f bovine aortic or corneal ECs Bovine aortic or corneal ECs were seeded at concentrations o f 2 x 103 or 2.5 x 104 cells/ml in complete growth medium on coated or uncoated PTFE in 15 m m four-well plates. At various time intervals, the cells o f triplicate cultures were dissociated by STV and counted in a Coulter counter. Lining o f small caliber cylinders of PTFE prosthesis with ECM and endothelittm Polytetrafluoroethylene cylinders (6 cm length × 0.6 cm diameter) were used for lining experianents. Bovine corneal ECs in D M E M containing 5% FCS, 10% CS, and 4% dcxtran were seeded at a concentration of 5 × 10 5 cells/cm 2 on PTFE segments coated with fibronectin. The rotation device (Fig. 1), developed in our laboratory, consisted o f four rotation axes rotating at a velocity o f four rotations per hour, to guarantee an even distribution o f adhering ECs. The PTFE cylinders were put in sterile Falcon tubes on the rotation device and were incubated/rotated for 48 hours at 37 ° C in 8% CO 2 incubator. At the end o f the rotation period the incubation o f the cylinders continued for another 5 days (the expected time for ECs to reach confluence). At this point the growth o f the cells was evaluated by cell dissociation with STV and counting in a Coulter counter, and visualization o f cells on the
Journal of VASCULAR SURGERY
652 Schneider et al.
8' 5' ........
Fig. 3. Attachment and proliferation of bovine aortic ECs on Gore-Tex (PTFE) disks coated with different biosubstrates. A, Adult bovine ECs were seeded (2 × 103cells/well) on untreated PTFE, PTFE coated with ECM, or PTFE that was first coated with fibronectin and then by ECM. After 7 days in culture cells were dissociated with STV and counted. B, Adult bovine ECs seeded (2 × 103 cells/well) on untreated PTFE disks (control) or on PTFE disks coated with fibronecfin, fibronecfin + ECM, or Matrigel. After 1 day ("attachment") and 8 days ("proliferation") in culture, cells were dissociated with STV and counted. No FGF was added during this experiment.
Extracellular matrix staining
Table I. Attachment of human colon carcinoma cells to PTFE coated with ECM PercenF
No. of attached cells PTFE coated with fibronectin and then ECM PTFE coated with ECM alone Untreated PTFE PTFE coated with fibronectin
of cell attachment
6.06 × 104
1.90 × 104 < 1 × 10 3 < 1 × 10 3