Cell isolation and culture
Human umbilical vein endothelial cells were obtained by the method of Jaffe et al. (20). Briefly, umbilical cords were collected from healthy newborns in ice-cold culture media. Veins were cannulated at both ends, washed with calcium-free HEPES buffer (10 mM HEPES, 119 mM NaCl, 6.7 mM KCl, and 11 mM glucose, at pH 7.35), and a warm collagenase solution (0.160 U/ml in HEPES buffer with 5 mM CaCl2) was injected to rinse and fill the vein. After a 15 min incubation at 37°C, the veins were gently massaged and vigorously perfused with medium. The cell solution obtained was centrifuged, and the cell pellet was resuspended in medium for endothelial cells [medium M199 supplemented with 20% fetal bovine serum (FBS), 2 mM L-glutamine, 50 U/ml of heparin, 25 µg/ml of endothelial cell growth factor supplement (Sigma, St. Louis, Mo.), 100 U/ml of penicillin G, and 25 µg/ml of gentamicin]. ECs were then plated on gelatin-coated tissue culture flasks. The endothelial nature of these cells was confirmed by von Willebrand factor expression and acetylated low density lipoprotein (ac-LDL) incorporation (21). Human vascular SMC were isolated from umbilical veins by the explant method of Ross (22) and cultured in standard medium (Dulbecco's modification of Eagle's medium-Ham's F12 modified medium (3:1), 10% FBS and antibiotics). SMC identity was confirmed by SMC specific -actin immunostaining (23). Human skin fibroblasts were obtained from normal adult skin specimens removed during reductive breast surgery of healthy human subjects (15–37 years old). Small skin fragments were then floated on a 500 µg/ml thermolysin solution in HEPES buffer for 2 h at 37 °C. The dermis was separated from the epidermis with forceps, cut into small pieces, and incubated for 20 h at 37 °C in a solution of collagenase (200 U/ml in Dulbecco's modification of Eagle's medium). After centrifugation, fibroblasts were plated into tissue culture flasks in standard medium. Fibroblasts did not express EC or SMC markers. All cells were plated at a density of 104 cells/cm2, maintained at 37°C in a humidified atmosphere (92% air and 8% CO2), and used between passages 3 and 7. Cells were tested at different passages for mycoplasma infection with Hoechst fluorescent staining for cytoplasmic DNA and were always found to be negative (24).
To induce ECM formation, SMC and fibroblasts were cultured in standard culture medium supplemented with 50 µg/ml of sodium ascorbate in 75 cm2 culture flasks. After approximately 30 days, both cell types formed sheets, comprising cells and ECM, that could be manually peeled off from the culture flask. These sheets could be wrapped around an inert tubular support to produce a cylinder composed of concentric sheet layers. After a maturation period, the layers adhered firmly to one another, forming a cohesive tubular tissue. From this basic technique, we then developed a sequential approach to TEBV construction. The first step was to produce an acellular inner membrane (IM) by dehydrating a tubular tissue made with a fibroblast sheet. The second step was to slip the IM around a perforated tubular mandrel (polytetrafluoroethylene, outside diameter 3.0 mm) and roll a sheet of SMC around it to produce a vascular media. At this stage, the construct was placed in a bioreactor designed to provide both luminal flow of culture medium and mechanical support. After a week of maturation, the third step was to roll a sheet of fibroblasts around the vascular media to provide an adventitia. Finally, after a maturation period of at least 8 wk, the inner tubular mandrel was removed and the TEBV was either tested for mechanical strength or cannulated at both ends for luminal endothelial cell seeding as described before (18). During maturation, tissues were cultured in standard medium with 50 µg/ml of sodium ascorbate except after EC seeding, when EC medium was used. Overall, the production of the graft involves a culture period of 3 months: 3 wk for SMC sheet formation, 1 wk for media maturation, 7 wk for adventitial maturation, and 1 wk for EC growth. This does not include IM production or cell expansion.
Five weeks of culture were found to produce fibroblast sheets with an optimal strength to culture time ratio for a given maturation period (see Results). Consequently, all ad~ventitias or IM were produced from 35-day-old sheets. IM were produced in advance from dehydrated adventitias, stored at -20°C, and rehydrated before use. When adventitias were produced separately for mechanical testing, a 4.6 mm styrene mandrel was used.
For EC staining, living tissues were labeled with CMFDA (5-chloromethylfluorescein-diacetate; Molecular Probes, Eugene, Oreg.) for 2 h (10 µM) and with DiI-ac-LDL (DiI=1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine chlorate; BTI, Stoughton, Mass.) for 3 h (7.5 µg/ml) in standard culture conditions. Cells were then fixed in 4% formaldehyde, rinsed in phosphate-buffered saline, permeabilized with 0.1% saponin, and incubated with a mouse anti-human von Willebrand factor monoclonal antibody (Chemicon, El Segundo, Calif.) for 45 min, rinsed, and incubated with a Cascade blue conjugated goat anti-mouse antibody (Molecular Probes) for 30 min. Green, red, and blue stains were taken as separate color micrographs, using appropriate filters on a Nikon epifluoroscopic microscope, and digital images were overlaid as true optical colors. For medial staining, frozen cross sections were fixed in acetone and incubated with mouse anti-desmin monoclonal antibody (Sigma) and Texas red conjugated goat anti-mouse antibody (Molecular Probes). Nuclei were stained blue with Hoechst 33258. For adventitial staining, frozen cross sections were fixed in cold acetone and double stained with rabbit anti-human elastin antibody (A. Grimaud, Institut Pasteur, Lyon, France), mouse anti-vimentin monoclonal antibody (N. Marceau, Hôtel-Dieu, Québec, Canada), FITC conjugated goat anti-rabbit (Cederlane, Hornby, Canada), and Texas red conjugated goat anti-mouse antibody.
Adventitias (inside diameter 4.6 mm) were cannulated on a specially designed system and pressurized with phosphate-buffered saline (137 mM NaCl, 2.7 mM KCl, 8.1 mM Na2HPO4 and 1.5 mM KH2PO4). Hydrostatic pressure was increased by 5 mmHg steps until vessel failure. Fresh human saphenous veins (HSV) were diameter matched with TEBV (inside diameter 3 mm) and free of collaterals. HSV were unused segments of carefully dissected autografts from patients undergoing distal vascular reconstruction.
Gelatinase activity was determined using 10% sodium dodecyl sulfate-polyacrilamide gels copolymerized with 3.5% gelatin as described before (25). Briefly, conditioned culture medium samples were centrifuged at 10000 g and immediately frozen at -20°C until use. After running, gels were rinsed for 15 min in washing buffer (50 mM Tris, pH 7.4) and then for 30 min in buffer containing 2.5% Triton X-100. They were rinsed once more in washing buffer for 15 min and put overnight at 37°C in digestion buffer (50 mM Tris, pH 7.4, containing 10 mM CaCl2 and 100 mM NaCl) under slow agitation. Gels were fixed with a 10% methanol solution containing 10% acetic acid, stained with 0.05% Coomassie blue prepared in the fixative, and scanned. All mediums were conditioned for 48 h and sample volumes were adjusted for total culture medium volumes.
Cell density was obtained by counting endothelial cell nuclei per surface area of Hoechst 33258 stained en face prepara~tions of endothelialized IM using a Nikon epifluoroscopic microscope equipped with appropriate UV filters. Five specimens were used and five randomly chosen fields were counted per specimen. Cell coverage of the luminal surface was calculated from morphometric analysis of hematoxylin stained en face preparation of endothelialized IM. Six specimens were used and three to five randomly chosen fields were counted per specimen. In both cases cells were fixed with 10% formalin for 30 min prior to staining. Prostacyclin (PGI2) synthesis was measured by detection of its stable degradation product 6-keto-prostaglandin F1α with a radioimmonoassay kit (Amersham, U.K.) in culture medium conditioned by confluent ECs on the IM. EC medium was perfused at a rate of 2.6 ml/h for 12 h with or without thrombin (2 U/ml). For platelet adhesion studies, fresh heparinized human blood from healthy male volunteers (14 U/ml) was slowly perfused (0.2 ml/min) in the lumen of an IM or an endothelialized IM with an eight-roller peristaltic pump at 37°C for 30 min. Lumens were gently flushed with culture medium to remove unattached platelets, fixed in 2% glutaraldehyde, and processed for scanning electron microscopy (SEM). Three separate experiments were performed with three endothelialized and unendothelialized vessels. Each vessel was cut into five specimens. Platelets were counted in three random fields from each specimen.
ECs were omitted to avoid acute rejection. Consequently, dogs were anticoagulated with warfarin (days -1 to 7) to raise their prothrombin time to twice the normal value. Dogs also received acetylsalicylic acid (325 mg/day on days -2 to 7) and heparin (100 U/kg on days 0 and 1). Immunosuppression was obtained with cyclosporin-A at a blood concentration of 500 ng/ml (days -2 to 7). Anastomosis was performed using a continuous running suture of 6–0 polypropylene. Graft patency was monitored daily by Doppler signaling techniques. Animal experiments were approved by the Ethics Committee of Laval University in accordance with the guidelines of the Canadian Council on Animal Care.