Samuel S. Murray

Secreted phosphoprotein 24 kDa (spp24) is a bone matrix protein. It contains a TGF‐β receptor II homology 1 (TRH1) domain. A cyclic, synthetic 19 amino acid peptide (bone morphogenetic protein binding peptide or BBP) based on the sequence of the TRH1 domain enhances BMP‐2 induced osteogenesis. Many observations suggest that different size forms of this protein have very different effects (inhibiting or enhancing) on BMP‐2 induced osteogenesis. Using the stable recombinant Met(His)₆‐tagged secretory form of full‐length (fl) bovine spp24 [Met(His)₆‐spp24 (residues 24–203)] and transgenic (TG) mice expressing fl bovine spp24 (residues 1–203), we have demonstrated that spp24 inhibits BMP‐2 induced bone formation. The effects of Met(His)₆‐spp24 (24–203) were determined in the ectopic bone‐forming bioassay in male mice. Implantation of 5 µg of BMP‐2 stimulated bone formation, assessed densitometrically as bone area and mineral content. When Met(His)₆‐spp24 (24–203) was implanted with BMP‐2, it elicited a dose‐dependent decrease in BMP‐2‐medicated ectopic bone formation. When added at a 50‐fold excess (w/w), Met(His)₆‐spp24 (24–203) completely ablated the effects of BMP‐2, while addition of a 10‐fold excess had no effect. Constitutive expression of fl bovine spp24 (1–203) under the control of the osteocalcin promoter in TG female mice reduced femoral and vertebral bone mineral density at 3 months of age and reduced femoral BMD at 8 months of age, but had no effects in male mice, which can exhibit less osteocalcin‐promoter driven gene transcription than females. Histomorphometric analysis demonstrated that bone volume and trabecular thickness were lower in TG female mice at 3 months of age than in sex‐ and age‐matched wild type (WT) controls. Thus, fl spp24 and its secretory isoform (Met(His)₆‐spp24 [24–203]), which contain a BMP‐binding or TRH1 motif, inhibit ectopic bone formation in male mice and adversely affects BMD and histological parameters related to bone mass and formation in female mice expressing the human transgene. Under these conditions, fl spp24 acts as a BMP antagonist in vivo. © 2008 Orthopaedic Society. Published by Wiley Periodicals, Inc. J Orthop Res 26:753–758, 2008

Demineralised bone matrix (DBM) products are complex mixtures of proteins known to influence bone growth, turnover, and repair. They are used extensively in orthopaedic surgery, and are bioassayed in vivo prior to being used in clinical applications. Many factors contribute to the osteogenic potency of DBM, but the relative contributions of these factors, as well as the possibility of interactive effects, are not completely defined. The “gold standard” measure of the therapeutic value of DBM, the in vivo assay for ectopic bone formation, is costly, time-consuming, and involves the use of numerous animal subjects. We have measured the levels of five growth factors released by the collagenase digestion of DBM, and statistically related these levels with osteogenic potency as determined by a standard in vivo model, in order to determine which value or combination of values of growth factors best predict osteogenic activity. We conclude that the level of BMP-2 is the best single predictor of osteogenic potency, and that adding the values of other growth factors only minimally increases the predictive power of the BMP-2 measurement. A small, but significant, interactive effect between BMP-2 and BMP-7 was demonstrated. We present a statistical model based on growth factor (e.g. BMP-2) analysis that best predicts the in vivo assay score for DBM. This model allows the investigator to predict which lots of DBM are likely to exhibit in vivo bioactivity and which are not, thus reducing the need to conduct in vivo testing of insufficiently active lots of DBM. This model uses cut-point analysis to allow the user to assign an estimate of acceptable uncertainty with respect to the “gold standard” test. This procedure will significantly reduce the number of animal subjects used to test DBM products.