Cloning of the Minimal Functional Domain of Human Lim Mineralization Protein-3 Able To Induce Bone Mineralization: In Vitro and In Vivo Study

Human LIM mineralization protein (LMP)-3 is one of the three splice variants of LMP recently identified. LMPs are involved in the osteoblast differentiation program and structurally characterized by the two conserved LIM and PDZ domains. Human LMP-1 (hLMP-1) shows one N-terminal PDZ domain and three C-term LIM domains connected by a non-conserved Unique region, deleted in the hLMP-2. hLMP-3 misses almost completely the LIM domains along with part of the unique region, due to a frame shift mutation. The three isoforms are expressed almost ubiquitously but show quantitative differences, hLMP-3 being the less expressed in all the analyzed tissues. Both hLMP-1 and hLMP-3 has been demonstrated to induce bone formation in vitro and ectopic bone formation in vivo, while hLMP-2 is not osteoinductive, suggesting that LIM domains are not essential for this function. Thus it has been hypothesized that the osteoinductive domain could reside in the Unique region that is partially conserved in hLMP-3. To examine the osteoinductive properties of this minimal domain we have cloned three different length of the Unique region of the hLMP-3 gene, corresponding to 120, 90 and 60 bp, fused to the enhanced green fluorescent protein (eGFP) and named L40-eGFP, L30-eGFP and L20-eGFP respectively. Thus we tested the ability of these constructs to induce bone specific gene expression and bone mineralization in vitro and ectopic bone formation in vivo in comparison to the full-length gene hLMP-3. Here we demonstrate that adenoviral-mediated gene transfer of all the 3 domains induces expression of certain bone-specific genes in a mouse fibroblasts cell line. The up-regulation of osteo-specific genes was assessed in mouse fibroblasts also by means of biolistic transfection using a plasmid containing a L20-eGFP fusion gene. In addition, we demonstrate that all the domains are able to induce mineralization in fibroblast and mesenchymal stem cells. An experiment to evaluate if direct gene transfer of the three constructs into murine skeletal muscle results in ectopic bone formation as efficiently as using LMP-3 is being performed. Finally in order to propose these new constructs as an effective approach to induce bone formation in vivo for clinical applications, we have synthesized a peptide of 20 aminoacid, corresponding to the fragment of 60 bp of the Unique region (named PTD-OD-1). The peptide enter the cells by a protein transduction domain (PTD-5) and its ability to induce in vitro expression of bone-specific genes and bone mineralization both in fibroblast and in human mesenchymal stem cells will be evaluated. PTD-OD-1 could represent a safe and powerful tool for clinical applications, and merit several analysis to evaluate its ability.