Adele L. Boskey

Targeted disruption of the biglycan gene leads to an osteoporosis-like phenotype in mice

The resilience and strength of bone is due to the orderly mineralization of a specialized extracellular matrix (ECM) composed of type I collagen (90%) and a host of non-collagenous proteins that are, in general, also found in other tissues. Biglycan (encoded by the gene Bgn) is an ECM proteoglycan that is enriched in bone and other non-skeletal connective tissues. In vitro studies indicate that Bgn may function in connective tissue metabolism by binding to collagen fibrils and TGF-ß (Refs 5,6), and may promote neuronal survival. To study the role of Bgn in vivo, we generated Bgn-deficient mice. Although apparently normal at birth, these mice display a phenotype characterized by a reduced growth rate and decreased bone mass due to the absence of Bgn. To our knowledge, this is the first report in which deficiency of a non-collagenous ECM protein leads to a skeletal phenotype that is marked by low bone mass that becomes more obvious with age. These mice may serve as an animal model to study the role of ECM proteins in osteoporosis.

Extracellular Matrix Mineralization and Osteoblast Gene Expression by Human Adipose Tissue–Derived Stromal Cells

Human adipose tissue represents an abundant reservoir of stromal cells with potential utility for tissue engineering. The current study demonstrates the ability of human adipose tissue-derived stromal cells to display some of the hallmarks of osteoblast differentiation in vitro. Following treatment with ascorbate, beta-glycerophosphate, dexamethasone, and 1,25 dihydroxy vitamin D(3), adipose tissue-derived stromal cells mineralize their extracellular matrix based on detection of calcium phosphate deposits using Alizarin Red and von Kossa histochemical stains. Fourier transform infrared analysis demonstrates the apatitic nature of these crystals. Mineralization is accompanied by increased expression or activity of the osteoblast-associated proteins osteocalcin and alkaline phosphatase. These and other osteoblast-associated gene markers are detected based on polymerase chain reaction. In contrast, the adipocyte gene markers--leptin, lipoprotein lipase, and peroxisome proliferator activated receptor gamma2--are reduced under mineralization conditions, consistent with the reciprocal relationship postulated to exist between adipocytes and osteoblasts. The current work supports the presence of a multipotent stromal cell population within human extramedullary adipose tissue. These findings have potential implications for human bone tissue bioengineering.

Osteopontin-hydroxyapatite interactions in vitro: inhibition of hydroxyapatite formation and growth in a gelatin-gel.

Osteopontin is a phosphorylated bone matrix sialoprotein, postulated to play a regulatory role in biomineralization. The effects of a crude preparation of rat bone osteopontin and a more highly purified bovine bone osteopontin were evaluated using a gel diffusion system to measure effects of 0.1-100 micrograms/ml of this matrix protein on hydroxyapatite formation and crystal proliferation. Bovine osteopontin at concentrations greater than 25 micrograms/ml inhibited both hydroxyapatite formation and growth in a dose-dependent manner. Osteopontin at concentrations lower than 25 micrograms/ml had no detectable effect on the amount of mineral accumulated in experiments with and without pre-formed hydroxyapatite seed crystals either when initial mineral deposition was assessed at 3.5 days, or when mineral formation and growth were assessed at 5 days. There was a statistically significant dose-dependent decrease in crystal length at all concentrations tested. The rat osteopontin preparation had similar inhibitory abilities. Partial dephosphorylation of bovine osteopontin with alkaline phosphatase removed its inhibitory ability, and reduced its ability to bind calcium. The affinity of bovine osteopontin for hydroxyapatite was determined based on a Langmuir adsorption isotherm, with values of K (binding affinity) and N (number of binding sites) being 0.026 ml/microgram and 1084 micrograms/m2, respectively. The data suggest that, in this system, osteopontin is an effective inhibitor of hydroxyapatite formation and growth due to its affinity for the hydroxyapatite crystals. In this system, osteopontin, distinct from other phosphoproteins which both promote and inhibit hydroxyapatite deposition, did not enhance mineral formation at any concentration tested.

Spectroscopic characterization of collagen cross-links in bone

Collagen is the most abundant protein of the organic matrix in mineralizing tissues. One of its most critical properties is its cross‐linking pattern. The intermolecular cross‐linking provides the fibrillar matrices with mechanical properties such as tensile strength and viscoelasticity. In this study, Fourier transform infrared (FTIR) spectroscopy and FTIR imaging (FTIRI) analyses were performed in a series of biochemically characterized samples including purified collagen cross‐linked peptides, demineralized bovine bone collagen from animals of different ages, collagen from vitamin B6‐deficient chick homogenized bone and their age‐ and sex‐matched controls, and histologically stained thin sections from normal human iliac crest biopsy specimens. One region of the FTIR spectrum of particular interest (the amide I spectral region) was resolved into its underlying components. Of these components, the relative percent area ratio of two subbands at ∼1660 cm−1 and ∼1690 cm−1 was related to collagen cross‐links that are abundant in mineralized tissues (i.e., pyridinoline [Pyr] and dehydrodihydroxylysinonorleucine [deH‐DHLNL]). This study shows that it is feasible to monitor Pyr and DHLNL collagen cross‐links spatial distribution in mineralized tissues. The spectroscopic parameter established in this study may be used in FTIRI analyses, thus enabling the calculation of relative Pyr/DHLNL amounts in thin (∼5 μm) calcified tissue sections with a spatial resolution of ∼7 μm.

Adhesion to Vitronectin and Collagen I Promotes Osteogenic Differentiation of Human Mesenchymal Stem Cells.

The mechanisms controlling human mesenchymal stem cells (hMSC) differentiation are not entirely understood. We hypothesized that the contact with extracellular matrix (ECM) proteins normally found in bone marrow would promote osteogenic differentiation of hMSC in vitro. To test this hypothesis, we cultured hMSC on purified ECM proteins in the presence or absence of soluble osteogenic supplements, and assayed for the presence of well-established differentiation markers (production of mineralized matrix, osteopontin, osteocalcin, collagen I, and alkaline phosphatase expression) over a 16-day time course. We found that hMSC adhere to ECM proteins with varying affinity (fibronectin>collagen I≥collagen IV≥vitronectin>laminin-1) and through distinct integrin receptors. Importantly, the greatest osteogenic differentiation occurred in cells plated on vitronectin and collagen I and almost no differentiation took place on fibronectin or uncoated plates. We conclude that the contact with vitronectin and collagen I promotes the osteogenic differentiation of hMSC, and that ECM contact alone may be sufficient to induce differentiation in these cells.

FTIR microspectroscopic analysis of human osteonal bone

Fourier Transform Infrared Microspectroscopy (FTIRM) has been used to study the changes in mineral and matrix content and composition in replicate biopsies of non-osteoporotic human osteonal bone. Spectral maps in four orthogonal directions (in 10 μm steps) from the centers towards the peripheries of individual osteons were obtained from iliac crest biopsies of two necropsy cases. Mineral to matrix ratios, calculated from the ratio of integrated areas of the phosphate v1,v3 band at 900–1200 cm-1 to the amide I band at 1585–1725 cm-1, increased from the center to the periphery of the osteon. The total carbonate (based on the v2 band at ≈850–900 cm-1) to phosphate v1,v3 ratio decreased as the mineral to matrix ratio increased. Analysis of the v2 CO32- band with a combination of second-derivative spectroscopy and curve fitting revealed a decrease in “labile” carbonate, a slight decrease in Type A and a slight increase in Type B carbonate from the center to the periphery of the osteon. Similar analysis of the components of the v1,v3 phosphate band with a combination of second-derivative spectroscopy and curve fitting revealed the presence of 11 major underlying moieties. These components were assigned by comparison with published frequencies for apatite and acid-phosphate containing calcium phosphates. The most consistent variations were alterations in the relative percent areas of bands at ≈1020 and ≈1030 cm-1, which had previously been assigned to nonstoichiometric and stoichiometric apatites, respectively. This ratio was used as an index of variation in crystal perfection throughout the osteon. This ratio decreased as the mineral to matrix ratio increased. The reproducibility of these parameters at multiple sites in multiple biopsies suggests their applicability for the analysis of mineral changes in disease.

Novel infrared spectroscopic method for the determination of crystallinity of hydroxyapatite minerals

Biologically important apatite analogues have been examined by Fourier Transform Infrared Spectroscopy (FT-IR), and a method developed to quantitatively assess their crystalinity. Changes in the phosphate v1 and v3 regions, 900-1,200 cm-1, for a series of synthetic (containing hydroxide, fluoride, or carbonate ion) and biological apatites with crystal sizes of 100-200 A were analyzed with curve-fitting and second derivative spectroscopy. The v1,v3 contour was composed of three main subbands. Correlations were noted between two spectral parameters and crystal size as determined by x-ray diffraction. The percentage area of a component near 1,060 cm-1 decreased as the length of the c-axis of the hydroxyapatite (HA) compounds increased, while the frequency of a band near 1,020 cm-1 increased with increasing length of the apatite c-axis. These parameters are thus proposed as indices of crystallinity for biological (poorly crystalline) HA. The FT-IR spectra of highly crystalline apatitic compounds were also analyzed. For crystal sizes of 200-450 A, the percentage area of the phosphate v1 band (near 960 cm-1) decreased with increasing HA crystal size. IR indices of crystallinity have thus been developed for both well crystallized and poorly crystallized HA derivatives. The molecular origins of the various contributions to the v1,v3 contour are discussed, and a preliminary application of the method to a microscopic biological sample (rat epiphyseal growth plate) is illustrated.

Dkk2 has a role in terminal osteoblast differentiation and mineralized matrix formation.

Human and mouse genetic and in vitro evidence has shown that canonical Wnt signaling promotes bone formation, but we found that mice lacking the canonical Wnt antagonist Dickkopf2 (Dkk2) were osteopenic. We reaffirmed the finding that canonical Wnt signaling stimulates osteogenesis, including the differentiation from preosteoblasts to osteoblasts, in cultured osteoblast differentiation models, but we also found that canonical Wnts upregulated the expression of Dkk2 in osteoblasts. Although exogenous overexpression of Dkk before the expression of endogenous canonical Wnt (Wnt7b) suppressed osteogenesis in cultures, its expression after peak Wnt7b expression induced a phenotype resembling terminal osteoblast differentiation leading to mineralization. In addition, osteoblasts from Dkk2-null mice were poorly mineralized upon osteogenic induction in cultures, and Dkk2 deficiency led to attenuation of the expression of osteogenic markers, which could be partially reversed by exogenous expression of Dkk2. Taken together with the finding that Dkk2-null mice have increased numbers of osteoids, these data indicate that Dkk2 has a role in late stages of osteoblast differentiation into mineralized matrices. Because expression of another Wnt antagonist, FRP3, differs from Dkk2 expression in rescuing Dkk2 deficiency and regulating osteoblast differentiation, the effects of Dkk2 on terminal osteoblast differentiation may not be entirely mediated by its Wnt signaling antagonistic activity.

Von Kossa Staining Alone Is Not Sufficient to Confirm that Mineralization In Vitro Represents Bone Formation

Numerous techniques are currently used to characterize biological mineralization in intact tissues and cell cultures; the von Kossa staining method, electron microscopic analysis (EM), X-ray diffraction, and Fourier transform infrared spectroscopy (FTIR) are among the most common. In this study, we utilized three of these methods to compare the mineralization of cultured fetal rat calvarial cells (FRC) and the osteoblast cell lines 2T3 and MC3T3-E1 with the in vivo mineral of rat calvarial bone. The cells were cultured with or without ascorbic acid (100 µg/ml) and β-glycerophosphate (2.5, 5, or 10 mM βGP), and harvested between 16 and 21 days (FRC cells and 2T3 cells) or at 30 days of culture (MC3T3-E1 cells). In the FRC cultures, maximal von Kossa staining was observed with 2.5 and 5 mM βGP in the presence of 100 µg/ml ascorbate. FRC cells also showed some von Kossa staining when cultured with βGP alone. In contrast, maximal von Kossa staining for MC3T3-E1 cells was observed with 10 mM βGP. Only the cultures of MC3T3-E1 cells that received both ascorbate and βGP produced von Kossa positive structures. The 2T3 cultures produced von Kossa positive staining only upon treatment with ascorbic acid and βGP, which was greatly accelerated by bone morphogenic protein-2 (BMP-2). FTIR was performed on the mineral and matrix generated in FRC, MC3T3, and 2T3 cultures, and the results were compared with spectra derived from 16-day-old rat calvaria. The mineral-to-matrix ratios calculated from FTIR spectra for rat calvaria ranged from 2.97 to 7.44. FRC cells made a bonelike, poorly crystalline apatite, and, with increasing βGP, there was a statistically significant (P ≤ 0.02) dose-dependent increase in the mineral-to-matrix ratio (0.56 ± 0.16, 1.00 ± 0.32, and 2.46 ± 0.76, for 2.5, 5, and 10 mM βGP, respectively). The mean carbonate-to-phosphate ratios of the FRC cultures were 0.015, 0.012, and 0.008, in order of increasing βGP concentration, compared with rat calvaria values of 0.009–0.017. The 2T3 cells treated with BMP-2 also made bonelike crystals, similar to those observed in FRC cultures. In contrast, the cultures of von Kossa positive MC3T3-E1 cells did not display a significant amount of mineral (maximum mineral-to-matrix ratio was 0.4). Thus, although the von Kossa stainings of FRC, 2T3, and MC3T3-E1 were very similar, FTIR analysis indicated that calcium phosphate mineral was not present in the MC3T3 cultures. By EM, the mineral in FRC cell cultures and 2T3 cultures was generally associated with collagen, whereas rare or sparse dystrophic mineralization of unknown chemical origin was evident in the MC3T3-E1 cultures. These studies demonstrate that von Kossa staining alone is not appropriate for the identification and quantitation of bonelike mineral and, hence, other techniques such as X-ray diffraction, EM, or FTIR should be utilized to verify the presence and quality of calcium phosphate phases.

Osteopontin Deficiency Increases Mineral Content and Mineral Crystallinity in Mouse Bone

Fourier transform infrared microspectroscopy (FTIRM) and infrared imaging (FTIRI) were used to characterize the mineral in bones of two different lines of Opn-deficient (Opn-/-) mice and their background-matched wild-type controls (Opn+/+). Sections of tibia and femur from 12-week-old and 16-week-old mice were evaluated with a spatial resolution between 10 microm (FTIRM) and 7 microm (FTIRI). FTIRI was used to examine 400 microm x 400 microm areas in cortical bone and trabecular bone and FTIRM examined selected 20 microm x 20 microm areas at sites within these anatomically defined areas. Despite the absence of an obvious phenotype in Opn-deficient mice, being undetectable by radiographic and histological methods, FTIRM analyses revealed that the relative amount of mineral in the more mature areas of the bone (central cortical bone) of Opn-knockout mice was significantly increased. Moreover, mineral maturity (mineral crystal size and perfection) throughout all anatomic regions of the Opn-deficient bone was significantly increased. The 2-dimensional, color-coded data (images) produced by FTIRI showed similar increases in mineral maturity in the Opn-/- bone, however, the crystallinity parameters were less sensitive, and significance was not achieved in all areas analyzed. Nonetheless, the findings of increased mineral content and increased crystal size/perfection in both lines of Opn-deficient mice at both ages are consistent with in vitro data indicating that Opn is a potent inhibitor of mineral formation and mineral crystal growth and proliferation, and also support a role for Opn in osteoclast recruitment and function.