CELL ORIENTATION INFLUENCES THE MECHANICAL PROPERTIES OF FIBROBLAST POPULATED COLLAGEN VESSELS Jessica E. Wagenseil 1 , Elliot L. Elson 2 and Ruth J. Okamoto 1,3 Departments of 1 Biomedical Engineering, 2 Biochemistry and Biophysics and 3 Mechanical Engineering Washington University St. Louis, MO INTRODUCTION Bio-artificial vessels composed of cells in a collagen matrix are being developed as replacements for damaged small caliber arteries. Most pure collagen-cell constructs lack the required strength to withstand in vivo conditions. The strength and stiffness can be altered by changing the microstructure of the construct such as the cell orientation, degree of collagen crosslinking and amount and/or type of matrix present. If the mechanical results of the microstructural changes could be predicted a priori, replacement vessels could be engineered to match specified design requirements. Zahalak et al (2000) developed a model to correlate micro- structural parameters, such as cell number and orientation, to the mechanical behavior of collagen-cell constructs. The model validation included relaxation tests on constructs with randomly oriented cells. To investigate applications of this model, fibroblast populated collagen vessels (FPCVs) with two different cell orientations have been constructed and subjected to biaxial mechanical tests. METHODS FPCVs were constructed as described in Wakatsuki et al (2000). Primary chick embryo fibroblasts (1E6/mL) were combined with rat tail collagen type I (1 mg/mL) and standard culture media containing DMEM, antibiotics and 10% fetal bovine serum. The mixture was poured into a tubular Teflon mold with a central mandrel and placed in an incubator. After 6 hours each FPCV had contracted around the central mandrel and was removed to a tissue culture dish. L’Heureux et al (1993) showed that cell orientation could be influenced by controlling the contact of a bio-artificial vessel with the inner mandrel. The FPCVs were either allowed to maintain contact with the mandrel during incubation (= no-slip) or were dislodged twice per day using forceps to slip the FPCV along the mandrel (= slip). The FPCVs were removed from the incubator after 4 days and small carbon markers were placed along each central length. The diameter and distance between the markers were measured to define the unloaded length and diameter. The FPCV was then removed from the mandrel and mounted in a biaxial test system. The test system can inflate and/or stretch an FPCV while recording the pressure, longitudinal force, diameter and length (for 1 pair of markers). Each FPCV was first preconditioned by overstretch. This consisted of 10 cycles at 5% over the maximum strain desired for testing. It has been determined that preconditioning by overstretch gives consistent stress- strain results that do not depend on the total number of stretches (Wagenseil et al, 2001). After preconditioning, each FPCV was subjected to a series of unidirectional stretches (i.e. either diameter or length was varied). The FPCV was immersed in HEPES-DMEM with 5% calf serum at pH 7.4 and 37°C throughout the protocols. Stresses and strains were calculated assuming the FPCV was an incompressible, thin-walled cylinder. After mechanical testing, each FPCV was stained with a fluorescent cell membrane dye, fixed and imaged with a confocal microscope. The specimens were prepared so that the horizontal image axis (0°) corresponded with the longitudinal axis of the FPCV and the vertical axis (90°) corresponded with the circumferential axis. Cell orientation was measured in 5 confocal slices equally spaced through the specimen thickness. Round cells and cells that were cut off in the x-y image plane were not included. Cell angles were divided into 9 bins from -90 to 90° (no-slip) or 0 to 180° deg (slip). RESULTS The FPCVs demonstrated anisotropic mechanical properties. The slip FPCVs were stiffer in the circumferential direction, while the no- slip FPCVs were stiffer longitudinally (Fig. 1a and 1b). The slip FPCVs had to be stretched beyond their unloaded length to be inflated at all. If the slip FPCVs were not stretched, they would bend longitudinally instead of inflating. At these large longitudinal strains, the slip FPCVs could not be inflated beyond 0.2 circumferential strain without bursting. The no-slip FPCVs could be inflated at their unloaded length up to 0.4 circumferential strain. The no-slip FPCVs usually could not be stretched beyond their unloaded length without tearing at the ends. Starting page #: 0203