Aim

Implant-associated infection remains one of the biggest challenges facing orthopaedics and there is an urgent clinical need to develop new prophylactic strategies. We have previously shown that CSA-90, a broad-spectrum antimicrobial, prevented infection in an infected open fracture model. In this study we developed a novel model of implant-associated infection, in which to further test the potential of CSA-90 as a prophylactic agent.

Introduction

Bone tissue engineering approaches are an emerging strategy to treat bone defects, and commonly involve the delivery of osteogenic cells and/or drugs via a porous scaffold. We have been exploring an alternative injectable approach for drug delivery that would obviate the need for invasive surgery.

CPT is a uniquely difficult condition, often associated with Neurofibromatosis (NF1), where bone healing is compromised. Although rare, the severity of this condition and the multiple procedures often entailed in treating it, warrant research attention. As study material is limited, animal models of the disorder are desirable for testing new treatments.

We sought to create a model of CPT where both copies of the NF1 gene were ablated at the fracture site, as has been found in some clinical specimens. NF1 floxed mice had fracture surgery; both closed fracture and open osteotomy were performed. Either a Cre- or control GFP-adenovirus was injected into the fracture site at day zero. Recombination was confirmed in ZAP reporter mice. Additionally, cell culture studies were used to examine the possible responses of NF1+/+ (wild type) NF1+/− or NF1−/− to drugs which may rescue the dysregulated Ras/MAPK pathway in NF1.

In closed fractures, radiographic bridging was 100% in NF1+/+ calluses and <40% in NF1−/− calluses (P<0.05). In open fractures, radiographic bridging was 75% in NF1+/+ calluses and <30% in NF1−/− calluses (P<0.05). In both fracture repair models the NF1−/− state was associated with a significant up to 15-fold increase in fibrotic tissue invading the callus by week 3. In NF1−/− fractures, large numbers of TRAP+ cells were observed histologically in the fibrotic tissue. Closed fractures also showed a significant increase in BRDU labelled proliferating cells in the callus. In cell culture models of NF1 deficient osteogenesis, NF1−/− progenitors were found to be significantly impaired in their capacity to form a calcified matrix as measured by Alizarin Red S staining and osteogenic markers (Runx2, Osteocalcin, Alp expression). However, when differentiated calvarial NF1 floxed osteoblasts were treated with Cre adenovirus, mineralization was not affected, suggesting that NF1 impacts on osteogenic differentiation rather than mature cell function. Treatment with MEK inhibitor PD0325901 was found to rescue the NF1−/− progenitor differentiation phenotype and permit robust mineralization. Treatment with the JNK inhibitor SP600125 was also able to improve ALP activity and mineralization in NF1+/− osteoprogenitors compared to control cells.

This model of NF1 −/− induction at a fracture or osteotomy site closely replicates the clinical condition of CPT, with lack of bone healing and fibrous tissue invasion. Underlying defects in bone cell differentiation in NF1 deficiencies can be at least partially rescued by JNK and MEK inhibitors.

A Ruys, School of Aerospace, Mechanical and Mechatronic Engineering, University of Sydney, Sydney

The effects of bone anabolics can be maximised by systemic co-treatment with an anti-catabolic. Local treatment may reduce the total drug required and produce superior outcomes, although high dose local bisphosphonate has been reported to impair bone formation. We have explored local co-delivery of anabolic/anti- catabolic bone drugs at different doses.

We manufactured biodegradable poly-D,L-lactic acid (PDLLA) polymer pellets containing 25g BMP-7 as an anabolic with or without 0.002mg-2mg Pamidronate (PAM) as an anti-catabolic. Polymer pellets were surgically implanted into the hind limb muscle of female C57BL6 mice. Animals were sacrificed at three weeks post- implantation and bone formation was assessed by radiography, microcomputed tomography (microCT) and histology. Histological staining on five Âm paraffin sections included haematoxylin/eosin, alcian blue/picrosirius red, and tartrate- resistant acid phosphatase (TRAP).

Radiographic and microCT data confirmed that 0.02mg and 0.2mg local PAM doses significantly augmented BMP-7 induced bone formation. In contrast, 2mg local PAM dramatically reduced the amount of bone present. This dose was comparable to that used by Choi et al who also reported impaired bone formation in a skull defect model.2 three-dimensional microCT and histological analyses of the ectopic bone and surrounding muscle showed a cortical shell covering the polymer pellet, which had not completely resorbed.

Histological analysis at the pellet/bone interface showed tissue granulation and no inflammation, suggesting a high biocompatibility of the PDLLA polymer. The presence of bisphosphonate also decreased the amount of fatty marrow tissue seen within between the cortical shell and the unresorbed polymer.

For the first time we can demonstrate synergy with local BMP/bisphosphonate. This study confirms that high local PAM doses can have negative effects, indicating a need to avoid overdosing. The lack of implant degradation suggests a need to optimise polymer degradation for bone tissue engineering application.

Bone morphogenetic proteins (BMPs) are able to induce osteogenic differentiation in many cells, including muscle cells. However, the actual contribution of muscle cells to bone formation and repair is unclear. Our objective was to examine the capacity of myogenic cells to contribute to BMP-induced ectopic bone formation and fracture repair.

Osteogenic gene expression was measured by quantitative PCR in osteoprogenitors, myoblasts, and fibroblasts following BMP-2 treatment. The MyoD-Cre x ROSA26R and MyoD-Cre x Z/AP mouse strains were used to track the fate of MyoD+ cells in vivo. In these double-transgenic mice, MyoD+ progenitors undergo a permanent recombination event to induce reporter gene expression. Ectopic bone was produced by the intramuscular implantation of BMP-7. Closed tibial fractures and open tibial fractures with periosteal stripping were also performed. Cellular contribution was tracked at one, two and three week time points by histological staining.

Osteoprogenitors and myoblasts exhibited comparable expression of early and late bone markers; in contrast bone marker expression was considerably less in fibroblasts. The sensitivity of cells to BMP-2 correlated with the expression of BMP receptor-1a (Bmpr1a). Pilot experiments using the MyoD-Cre x Rosa26R mice identified a contribution by MyoD expressing cells in BMP-induced ectopic bone formation. However, false positive LacZ staining in osteoclasts led us to seek alternative systems such as the MyoD-cre x Z/AP mice that have negligible background staining. Initially, a minor contribution from MyoD expressing cells was noted in the ectopic bones in the MyoD-cre x Z/AP mice, but without false positive osteoclast staining. Soft tissue trauma usually precedes the formation of ectopic bone. Hence, to mimic the clinical condition more precisely, physical injury to the muscle was performed. Traumatising the muscle two days prior to BMP-7 implantation: (1) induced MyoD expression in quiescent satellite cells; (2) increased ectopic bone formation; and (3) greatly enhanced the number of MyoD positive cells in the ectopic bone. In open tibial fractures the majority of the initial callus was MyoD+ indicating a significant contribution by myogenic cells. In contrast, closed fractures with the periosteum intact had a negligible myogenic contribution.

Myoblasts but not fibroblasts were highly responsive to BMP stimulation and this was associated with BMP receptor expression. Our transgenic mouse models demonstrate for the first time that muscle progenitors can significantly contribute to ectopic bone formation and fracture repair. This may have translational applications for clinical orthopaedic therapies.

Congenital pseudarthrosis of the tibia is an uncommon manifestation of neurofibromatosis type 1 (NF1), but one that remains difficult to treat due to anabolic deficiency and catabolic excess. Bone grafting and more recently recombinant human bone morphogenetic proteins (rhBMPs) have been identified as pro-anabolic stimuli with the potential to improve the outcome after surgery. As an additional pharmaceutical intervention, we describe the combined use of rhBMP-2 and the bisphosphonate zoledronic acid in a mouse model of NF1-deficient fracture repair.

Fractures were generated in the distal tibiae of neurofibromatosis type 1-deficient (Nf1^+/−) mice and control mice. Fractures were open and featured periosteal stripping. All mice received 10 μg rhBMP-2 delivered in a carboxymethylcellulose carrier around the fracture as an anabolic stimulus. Bisphosphonate-treated mice also received five doses of 0.02 mg/kg zoledronic acid given by intraperitoneal injection.

When only rhBMP but no zoledronic acid was used to promote repair, 75% of fractures in Nf1^+/− mice remained ununited at three weeks compared with 7% of controls (p < 0.001). Systemic post-operative administration of zoledronic acid halved the rate of ununited fractures to 37.5% (p < 0.07).

These data support the concept that preventing bone loss in combination with anabolic stimulation may improve the outcome following surgical treatment for children with congenital pseudarthoris of the tibia and NF1.

Background: Recombinant bone morphogenic proteins (BMPs) are potent bone anabolic agents suggested for the treatment of orthopaedic complications associated with neurofibromatosis type 1 (NF1), in particular, congenital pseudarthrosis of the tibia. We have explored the effect of Nf1 haploinsufficiency on ex vivo and in vivo models of BMP-induced bone formation in Nf1+/− mice.

Methods: Using an Nf1+/− knockout mouse model, we expanded primary cell cultures from calvarial and long bone osteoblasts and measured osteogenic markers, such as alkaline phosphatase and mineralization using Alizarin Red staining, and the responses of these markers to BMP-2 treatment. We also developed an in vivo muscle pouch heterotopic ossification model to assess the ability of BMP-2 to form bone.

Results: Primary osteoblast cultures from Nf1+/− mice showed reduced ALP staining, ALP activity and mineralization, denoting an anabolic deficiency. Nf1+/− osteoblasts responded to BMP-2 treatment, although osteogenic markers were reduced compared to BMP-2 treated Nf1+/+osteoblasts. Heterotopic bone was induced in both genotypes by surgically implanting BMP-2, however less bone was formed in Nf1+/− mice than Nf1+/+ controls.

Conclusion: These data indicate that BMP therapies have potential utility in treating orthopaedic defects in children with NF1, but that dosing may need to be optimized for this patient subgroup or that catabolism may need to be also controlled.

The literature on fracture repair has been reviewed. The traditional concepts of delayed and nonunion have been examined in terms of the phased and balanced anabolic and catabolic responses in bone repair. The role of medical manipulation of these inter-related responses in the fracture healing have been considered.