OTA 1997 Posters - Scientific Basis for Fracture Care

Re-Induction of the Hedgehog Signaling Pathway during Fracture Healing

Anh X. Le, MD, Motoki Iwasaki, MD, PhD, David S. Bradford, MD, Theodore Miclau, MD, Jill A. Helms, DDS, PhD

Orthopaedic Molecular Biology Laboratory, San Francisco, California, USA

Purpose: An estimated 5.6 million fractures occur annually, and approximately 10% of them develop into delayed unions or nonunions. This clinical situation has generated an interest in developing biologically based treatments to stimulate fracture healing. Exogenous application of growth promoting proteins, such as bone morphogenetic proteins, to the fracture site appear clinically promising. However, the basic molecular mechanisms mediating cartilage and bone formation are largely unknown. Unlike other tissues, bone heals by generating new bone instead of by forming a scar. Histomorphologically, the mesenchymal stem cells that populate the wound site are similar to those recruited during embryonic bone formation. This has lead to the proposition that fracture healing recapitulates stages of fetal bone formation. The purpose of our study was to determine whether genes which regulate cartilage and bone formation during embryonic skeletogenesis were re-induced during fracture repair. In addition, we also investigated the effects of fracture stabilization on the temporal and spatial expression patterns of these genes.

The genes we focused on were members of the Hedgehog signaling cascade and included indian hedgehog (ihh), bone morphogenetic protein 6 (bmp6) and gli. Based on work in our laboratory and others, ihh is expressed in mature and hypertrophic cells in the cartilage anlage of the limb. Ihh induces a number of target genes, among them are bmp6, which is expressed by the hypertrophic chondrocytes, and gli, a transcription factor expressed in the perichondrium and in resting chondrocytes. Together, these signaling molecules participate in a negative feed-back loop that regulates the rate of chondrogenic differentiation. These genes continue to be expressed in the neonatal growth plate in the same relative expression domains. Paradoxically, these genes are no longer expressed in the physes of 2-week-old mice.

We tested the hypothesis that fracture healing recapitulates embryonic bone formation by determining whether ihh, bmp6 and gli were re-induced in a fracture callus. If these genes are involved in determining the rate at which bone heals, then we hypothesized that we could influence their temporal expression by stabilizing the fracture site.

Methods: Adult mice (age 10-12 weeks) underwent a closed tibia fracture with and without internal fixation. The contralateral tibia served as control. Animals were sacrificed after 3, 5, 7, 14 and 28 days post-fracture. All tissues were collected, fixed in 4% paraformaldehyde, decalcified in buffered EDTA for 10 days at 4°C, dehydrated in ethanol and embedded in paraffin wax. Adjacent sections were processed for histologic analysis and for in situ hybridization. For in situ analyses, we used (35 ) S-radiolabeled antisense probes corresponding to type II collagen (col 2), type X collagen (col 10), ihh, bmp6 and gli, and sense riboprobes were used as negative controls. Both the fracture site and the proximal tibial growth plate were included on each slide to serve as an internal control for probe sensitivity.

Results: All mice (n=40) ambulated within five hours after the operation and all healed without signs of infection. In the unstabilized fracture, mesenchymal cells expressed collagen type II (col 2) by day 3 postfracture, suggesting that they were committed to differentiate along a chondrocytic lineage. At day 5 post-fracture, ihh, bmp6 and col 10 mRNA transcripts were first detected in the cartilage islands within the callus. By day 7, ihh, gli, bmp6 and col 10 were all expressed within the cartilage callus. The expression domain of ihh, which overlapped with that of col 10, was localized to a subset of mature and hypertrophic chondrocytes. Bmp6 was expressed by a group of hypertrophic chondrocytes adjacent to ihh-expressing cells. Gli transcripts were detected in the neoperiosteum and in a population of chondrocytes at the periphery of the callus. By day 14, newly formed woven bone could be distinguished at the fracture site, and ihh, gli and bmp6 were expressed in the same relative domains in the partially ossified callus. By 28 days post-fracture, the expression of ihh, bmp6 and gli were no longer detectable in the completely ossified callus.

In the stabilized group, the expression of ihh, bmp6 and gli was not detected until day 7 post-fracture. Ossification of the callus occurred by day 14, at which time ihh, bmp6 and gli expression patterns were no longer detected. However, the chondrocytes in these cartilage tissues still expressed col 10, indicating that the process of hypertrophy persisted.

Discussion and Conclusion: Several conclusions can be derived from these data. First, we demonstrated that the key molecules that mediate fetal bone development are re-induced in an adult fracture callus. The expression of ihh, bmp6 and gli in a fracture callus suggests a possible regulatory role for the Hedgehog signaling cascade during bone repair. Second, we provide direct evidence that immobilization of a fracture accelerates callus maturation at a molecular level. Our results indicated that the cartilage callus in the stabilized model ceased to express ihh, bmp6 and gli at an earlier time point than in the unstabilized model. This was expected since ihh appears to regulate the rate of chondrogenic differentiation, and stabilization favors intramembranous bone formation. Collectively, our results demonstrate that fracture healing appears to recapitulate fetal bone development. These data represent an important step towards understanding the molecular processes governing fracture healing. Our long-term goal is to use this information to develop biologically-based therapies for difficult clinical entities, such as delayed unions, non-unions and pseudarthroses.