INFLUENCE OF MECHANICAL STRAIN ENVIRONMENT ON PATTERNS OF GENE EXPRESSION DURING SECONDARY FRACTURE HEALING

Participants: E.A. Smith, S.K. Volkman, S.A. Goldstein, M.R. Moalli, J.A. Goulet, B.T. Nolan, K.A. Sweet, D.C. Kayner, M.W. Stock

Keywords: fracture healing, growth factors, angiogenesis

Introduction

Fracture healing is a complex process that involves progression through specific stages of tissue formation, including: inflammation, hematoma, and granulation tissue; chondrogenesis; cartilage calcification; and finally osteogenesis. Cellular proliferation, differentiation, and extracellular matrix formation in each stage is under the influence of specific growth factors, as well as the hormonal, vascular, and mechanical environment. Alterations in the mechanical environment of a healing fracture has been shown to effect the rate and success of the healing process, although specific relationships between mechanical environment and the cellular process that occur during fracture healing have yet to be elucidated. The goal of this study is, therefore, to correlate patterns of mechanical strain with patterns of gene expression and angiogenesis in a rat model of fracture healing with mechanical stimulation.

Materials and Methods

A custom hinged, external fixator with a locking plate is applied to both femora of male, Sprague Dawley rats, and a 2mm transverse osteotomy is made in the mid-diaphysis. From day 7 until day 21, the animal is anesthetized, and the fracture in one leg is loaded in bending 3 times per week at 0.5 Hz for 17 minutes, for a total of 450 load cycles. Mechanical loading occurs by a custom loading apparatus, which consists of a rotating motor, torque cell, and an eccentric cam that attaches to the external fixator via a four-bar linkage. Input torque is then sampled for 2 minutes at the onset of each loading session. Rats are sacrificed at 1, 2, 3, 4, and 6 weeks and both femora were harvested. The femora are then snap frozen and stored at -80C.

Digital image-based finite element models will be constructed from µMRI images of a subset of femurs using MnSO₄ as a contrast agent. Areas of constituent materials will be thresholded from the µMRI images, and then be directly converted into finite elements. Scanning acoustic microscopy will be performed on three orthogonal slices to determine material properties of the resident materials in three dimensions. Linear, 3-D finite element analysis will be run to estimate the local mechanical strain environment.

A further subset of femora will be longitudinally frozen and stained using toluidine blue and safranin-O/fast green. Expression domains of growth and angiogenic factors bFGF, VEGF, and transferrin, and differentiation factors Ihh, PTHrP, and BMP-6 will be determined by immunolocalization. Blood vessel invasion will be visualized using immunohistochemistry for endothelial cell maker PECAM-1 and type IV collagen. The mechanical strain environment at each time point will then be related to the

Progress

The loading apparatus and data acquisition system have been constructed and used to load a total of 27 femora. These animals were then sacrificed at 1 week (n=3), 2 weeks (n=5), 3 weeks (n=6), 4 weeks (n=7) or 6 weeks (n=6) and microradiographs were taken. Immunohistochemistry protocols are currently being refined, and new techniques for cryosectioning are being developed. One µMRI image has been acquired and is being used for development of image processing software for finite element model construction.

Results

Preliminary results from microradiographs indicate that at weeks 1 and 2, there is no apparent difference between mechanically stimulated and unstimulated femora. At 3 weeks (i.e. after 2 weeks of mechanical stimulation), the radiographic appearance of the unstimulated leg seems to indicate more advanced healing than the stimulated leg. However, by 4 weeks, healing in the stimulated leg appears to be more advanced than the unstimulated leg, and distinct spatial patterns of tissue formation are emerging. On the compressed cortex, the amount of mineralized callus is increased when compared to the tensile cortex. These patterns persist until 6 week post-operative, when the mechanically stimulated fracture appears to be bridged with mineralized tissue. We hypothesize that the mechanical signals generated during stimulation are delaying callus mineralizing and allowing proliferation and accumulation of cartilage. Then, when stimulated has ceased, mineralization proceeds, ultimately leading to increased mineralized tissue and earlier healing. This hypothesis will continue to be tested.