Objective: To examine the feasibility of using human dermal fibroblasts (DFbs) and polyglycolic acids (PGA) to engineer tendon in vitro.
Methods: Human dermal fibroblasts (DFbs) were isolated from the foreskin tissues of children obtained during operation with collagenase and cultured in vitro. Human tendon was obtained from a patient undergoing amputation during operation to isolate tenocytes. The DFbs of second passage were seeded on PGA fibers to form cell-scaffold constructs in shape of tendons. Those constructs were divided into 4 groups: experimental group (n = 15) with the DFbs inoculated on PGA scaffold under constant tension generated by a U-shaped spring, control group 1 (n = 15) with the DFbs inoculated on PGA scaffold without tension, control group 2 (n = 3), i. e., cell-free pure PGA scaffolds under tension, and control group 3 (n = 5), i. e., tenocyte-scaffold constructs under tension that was harvested only at the ninth week. Samples were harvested 2, 5, 9, 14, and 18 weeks later to undergo histological examination and biomechanical test.
Results: Two weeks later histological examination showed that the constructs were mainly composed of PGA fibers in both the experimental group and the group without tension. Transmission electron microscopy showed fine cell attachment and stretching on the scaffold. By the 5th week, a neo-tendon was formed in all groups except for the cell-free group, and histology revealed the formation of collagen fibers. At the 9th week, the PGA fibers of the cell-free group were broken and partially degraded, the neo-tendon's diameter of the experimental group was (1.18 +/- 0.25) mm, significantly thinner than that of the group without tension[ (2.43 +/- 0.49) mm, P = 0.017]. The gross morphology of tendons of the experimental group and tenocyte group were similar to each other except for more cells in the experimental group. In experimental group, immunohistochemistry revealed the production of fibers of collagen type I & III that were aligned longitudinally along the force axis like the normal tendon pattern. An irregular collagen pattern was observed in the group without tension. The maximum tensile stress of the experimental group was (2.75 +/- 0.59) MPa, similar to that of the tenocyte group [(3.08 +/- 0.30) MPa, P = 0.439], and significantly greater than that of the group without tension [(0.82 +/- 0.21) MPa, P = 0.006]. At the 14th week the PGA fibers of the cell-free group were mostly degraded. In addition, more dead cells and tissue atrophy were observed in the experimental group, and the tensile stress was higher than that of the same group by the 9th week. In the 18th week the number of hollow fiber of the experimental group was more obvious, the number of dead cells increased, and the tensile stress was lower, however, there was no significant difference in other characteristics compared with those in the 14th week.
Conclusions: DFbs can be used for in vitro tendon engineering as tenocytes. Mechanical stimulation by statistic strain is beneficial for tissue formation, but the effect may not be optimal if the tension is applied for too long.