Influence of scaffold composition on gene expression and cellular organization in tissue-engineered middle phalanx models of human digits
To augment or replace defective, diseased, or impaired human digits, the design and development of tissue-engineered phalanges are important and include a middle phalanx model. This construct consists in part of two square-shaped biodegradable polyglycolic acid (PGA) scaffolds (1 x 1 x 0.2 cm in length, width and thickness, respectively) seeded with cartilage cells (chondrocytes) obtained from young calves. One such seeded scaffold is sutured to each end of a rectangular-shaped scaffold (˜2 x 0.7 x 0.5 cm in length, width and thickness) serving as the midshaft of the model. To examine the biological regenerative capacity of these biomimetic composites, midshafts were left uncovered or wrapped with periosteum, a tissue from calves giving rise to cartilage and bone. Midshafts were composed of poly(L-lactide-ε-caprolactone) [P(LA-CL)] or one of two ceramics, hydroxyapatite (HA) or β-tricalcium phosphate (β-TCP), admixed with P(LA-CL). When engineered middle phalanx models were implanted and grown for up to 20 weeks under dorsal skin flaps of athymic (nude) mice, resulting constructs varied in their midshaft bone and end plate cartilage composition and structure. Harvested from mice at 20 weeks, all constructs (n = 3 for each type) without periosteum developed viable end plate cartilage as determined by Safranin-O staining for chondrocyte-secreted proteoglycans but cells were not organized as in normal growth plate cartilage of human digits. Midshafts remained devoid of cells and mineral. Implanted for the same time 20 week period, constructs of P(LA-CL) (n = 3), HA-P(LA-CL) (n = 3), or β-TCP-P(LA-CL) (n = 3) and enclosed by periosteum each developed viable end plate cartilage whose chondrocytes were organized into columns resembling normal growth plate cartilage of digits. Midshafts mineralized through the normal process of endochondral ossification. While these features were common to all periosteum-wrapped composites, specific differences occurred between them, apparently depending on midshaft copolymer composition. In particular after 10 or 20 weeks of implantation, gene expression of end plate chondrocytes varied in their levels of type II collagen, aggrecan (proteoglycan), or bone sialoprotein, all markers for development of normal cartilage extracellular matrix and mineralization. These results indicate that the composition of midshaft scaffolds comprising middle phalanx models of human digits affects the composition and structure of both midshaft bone and end plate cartilage of constructs. Continuing studies are defining more completely the relationships between structure and composition of bone and cartilage tissues developed and properties of their underlying copolymer scaffolds in these biomineralized models.