Pregnancy-Associated Plasma Protein-A Is Involved in Matrix Mineralization of Human Adult Mesenchymal Stem Cells and Angiogenesis in the Chi
http://www.100md.com
《内分泌学杂志》
Department of Biological Sciences (J.J.), Institute for Complex Engineered Systems (J.J., J.S., P.C.), and Bone Tissue Engineering Center (J.J., D.D, P.C.), Carnegie Mellon University, Pittsburgh, Pennsylvania 15213
Endocrine Research Unit (C.C.), The Mayo Clinic and Foundation, Rochester, Minnesota 55905
Abstract
Pregnancy-associated plasma protein A (PAPP-A) is an IGF binding protein 4 protease that can function to increase local IGF-I bioavailability. Aside from its assumed role during pregnancy, in vitro and in vivo studies have indicated roles for PAPP-A in IGF-I-mediated wound healing, vascular repair, and bone formation. Because bone morphogenetic protein 2 (BMP-2) is known to up-regulate Igf-I gene expression, we hypothesized that PAPP-A may be involved in BMP-2 mechanisms in bone formation. To test this hypothesis, we quantified gene expression of Papp-A in response to BMP-2 treatment and runt-related transcription factor 2, Osterix, and Igf-I in response to PAPP-A protein treatment in human adult mesenchymal stem cells. Our results demonstrate that BMP-2 directly up-regulated Papp-A gene and protein expression. Purified PAPP-A protein directly up-regulated runt-related transcription factor 2 and Igf-I gene expression but not Osterix. When added in combination with recombinant human BMP-2, PAPP-A increased matrix mineralization in the absence of dexamethasone. PAPP-A further demonstrated an angiogenic effect in the chick chorioallontoic membrane, which implicates a critical developmental role and possible therapeutic potential. Our findings suggest that PAPP-A functions in the formation of mineralized tissues through direct up-regulation of key genes. Furthermore, PAPP-A is involved in the formation of new blood vessels, which is essential for proper bone regeneration.
Introduction
PREGNANCY-ASSOCIATED plasma protein-A (PAPP-A) is a zinc-dependent metalloproteinase that was originally isolated in 1974 from normal human pregnancy serum (1). In recent years, circulating PAPP-A levels in maternal serum have been used to screen and diagnose fetal Down’s syndrome with lower levels of PAPP-A correlated with increased Down’s syndrome risk (2). PAPP-A has also been used as a marker for Cornelia de Lange syndrome, which results in mental and growth developmental delay, limb reduction abnormalities, a distinct facial appearance, and congenital heart defects (1). Although PAPP-A appears to have specific and important functions during pregnancy, its function in vivo is not known. PAPP-A has also been localized outside placental tissue in preovulatory granulosa, healing human skin, and atherosclerotic plaques as well as smooth muscle cells, fibroblasts and osteoblasts in culture (3).
PAPP-A has been identified as an IGF binding protein (IGFBP)-4 protease (1). IGFBP-4 binds to and inhibits the actions of IGFs. IGFs are mitogenic peptides that regulate fetal and neonatal growth. It has been postulated that PAPP-A increases bioavailability of IGFs by cleaving IGFBP-4 into fragments with low affinity for IGFs, thus releasing IGFs and permitting their action on local cells. A recent study suggested that increased levels of PAPP-A protein may promote local bone formation (4). Because bone morphogenetic protein (BMP)-2 and IGF-I have important angiogenic properties in normal skeletal and vascular development (5, 6, 7), we reason that PAPP-A could be a component of such mechanisms.
PAPP-A is known to also have pathological functions; elevated levels of PAPP-A are found in atherosclerotic plaques, in which many bone-related extracellular matrix proteins such as osteopontin, osteonectin, and osteocalcin have been localized (8). PAPP-A has been postulated to regulate bone formation via IGF-I signaling and increasing IGF bioavailability. TGF1 was shown to up-regulate PAPP-A message and protein expression (4). Regulation of PAPP-A by TGF1 lends credence to the proposed model of IGF-I and IGFBP-4 as key components of the coupling mechanisms between bone formation and bone resorption (4). Recently Papp-A knockout mice were shown to be significantly smaller in size (60%), compared with their wild-type littermates (9). The phenotype of PAPP-A-null mice is identical with the IGF-II knockout. However, normal levels of Igf-II transcripts are detectable in PAPP-A-null mice. The authors suggest that PAPP-A has an important regulatory role in embryonic growth and postnatal development. They further speculated that PAPP-A is an important modulator of IGF signaling, which helps to determine organ and body size in mammals. The bones of PAPP-A–/– knockout mice were visibly smaller than wild-type mice, thus supporting a role for PAPP-A in bone formation and development.
Thus, it is hypothesized that PAPP-A may have important and pivotal roles in local cellular function related to wound healing, bone remodeling, atherosclerotic plaque development, angiogenesis, and several aspects of human reproduction (10). To our knowledge, there have been no studies to date that investigate bone marker gene expression in vitro in response to administered purified PAPP-A. We hypothesize that PAPP-A has a critical role in bone formation, specifically at the molecular level and possibly within the BMP-2 and IGF-I signaling pathways. BMP-2 and IGF-I are growth factors critical for bone formation and their importance for progression of the osteogenic pathway has been well documented (11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32). Although there appears to be a close connection between BMP-2 and IGF-I (33, 34) and BMP-2 was shown to up-regulate Igf-I gene expression (35), a mechanism of cooperation has not been elucidated. BMP-2 is an inducer of the osteoblast transcription factors runt-related transcription factor (RUNX)2 and osterix (OSX), which are crucial for bone formation (36, 37, 38, 39). Runx2–/– mice display a cartilaginous skeleton, with no bone formation (39). Osx–/– mice do not express osteoblast marker genes such as osteocalcin but exhibit normal Runx2 levels, suggesting that OSX operates downstream of RUNX2 (36).
In the present study, we demonstrate that PAPP-A is involved in BMP-2 signaling and functions in up-regulation of Igf-I and Runx2 gene expression. We further demonstrate that PAPP-A can stimulate matrix mineralization in vitro together with BMP-2 in human adult mesenchymal stem cells (hMSCs). PAPP-A was also shown to have an angiogenic effect in the chick chorioallantoic membrane (CAM) model. The comprehensive data presented here suggest that PAPP-A may be a critical component in bone development and could be widely useful in therapeutic applications.
Materials and Methods
hMSCs were obtained from BioWhittaker, Inc. (Walkersville, MD). hMSCs were certified by the manufacturer as positive for potential toward adipogenic, chondrogenic, and osteogenic pathways. Human mesenchymal stem cell medium, mesenchymal cell growth supplement, L-glutamine, penicillin, and streptomycin were obtained from BioWhittaker and added to the medium according to the manufacturer’s specifications to prepare basal medium. Purified PAPP-A protein was purchased from Advanced Immunochemical (Long Beach, CA). PAPP-A ELISA was obtained from Diagnostic Systems Laboratories (Webster, TX). RNeasy kit and DNase I were obtained from QIAGEN (Valencia, CA). RiboGreen kit was obtained from Molecular Probes (Eugene, OR). All quantitative real-time PCR reagents, primers, and probes were purchased from Applied Biosystems (Foster City, CA). Human fibrinogen was obtained from Aventis (Bridgewater, NJ). Thrombin was obtained from Centeon Pharma (Liederbach, Germany). Quantum Dots were a gift from Quantum Dot Corp. (Hayward, CA). Total protein kits were obtained from Bio-Rad Laboratories (Temecula, CA). All other chemicals were obtained from Sigma Inc. (St. Louis, MO).
Cell culture
hMSCs were seeded in 35-mm dishes and grown in basal media to approximately 70% confluence. hMSCs were then treated with 100 ng/ml recombinant human (rh)BMP-2 for 6, 12, 24, and 48 h. hMSCs were also cultured separately in media containing 200 ng/ml PAPP-A in basal media for 12, 24, 48, and 96 h. Media were renewed every 48 h until RNA extraction.
Analysis of IGFBP-4 proteolytic activity
hMSCs were cultured for 6 d in a 24-well culture dish in media containing 100 ng/ml rhBMP-2. Media was renewed every other day (total volume 500 μl). Conditions were switched to serum free for the final 48 h of culture. Conditioned media were collected and frozen at –80 C until use. Proteolysis of IGFBP-4 was determined by previously established protocols (1). Conditioned media were incubated for 6 and 72 h with 125I-IGFBP-4 and in the presence or absence of IGF-II.
RNA extraction and quantification
Total RNA was extracted using the RNeasy kit according to the manufacturer’s instructions including DNase I treatment to eliminate any residual genomic DNA. RNA was quantified using the RiboGreen kit (Molecular Probes). After addition of the working solution according to the manufacturer’s instructions to RNA samples, fluorescence was quantified using a Tecan (Zurich, Switzerland) Spectrofluor with excitation at 485 and emission at 535 nm. RNA concentrations were calculated based on a standard curve of rRNA of known concentrations.
Quantitative real-time PCR
Cells were harvested from the culture treatments at the time points described above. After extraction and quantification of RNA, quantitative real-time PCR (qPCR) analysis was carried out using Taqman one-step RT-PCR master mix. Ten nanograms of total RNA were added per 25-μl reaction with sequence-specific primers (200 nM) and Taqman probes (200 nM). Sequences for all target gene primers and probes are shown in Table 1. 18S primers and probes were designed by and purchased from Applied Biosystems. qPCR assays were carried out in triplicate on an ABI Prism 7000 sequence detection system. Thermocycling conditions were as follows: 48 C for 30 min (reverse transcription) and 95 C for 10 min (initial denaturation) followed by 40 cycles at 95 C for 15 sec (denaturation) and 60 C for 45 sec (annealing and extension). The threshold was set above the nontemplate control background and within the linear phase of target gene amplification to calculate the cycle number at which the transcript was detected [denoted cycle threshold (CT)].
Gene expression values were calculated based on the comparative CT method (separate tubes) as detailed in Applied Biosystems user bulletin 2 (40). For each primers/probe set, validation experiments demonstrated that efficiencies of target and reference gene amplification were approximately equal; the absolute value of the slope of log input amount vs. CT < 0.1 (data not shown). Target genes were normalized to the reference housekeeping gene, 18S. Fold differences were calculated for each treatment group using normalized CT values for the negative control at the appropriate time point as the calibrator.
PAPP-A ELISA
hMSCs were cultured in basal media without serum supplemented with rhBMP-2. Media were collected after 12 and 48 h, 4 and 6 d total in culture and frozen at –80 C until use. Samples were thawed, and 50 μl of conditioned medium were subjected to analysis by ELISA according to the manufacturer’s protocol. PAPP-A concentration (in microinternational units per milliliter) was calculated based on a standard curve of known concentrations.
Inhibition of protein synthesis
hMSCs were seeded in 35-mm dishes as before and grown to approximately 70% confluence. Cells were pretreated with 5 μg/ml cycloheximide for 30 min before treatment with basal media or supplemented with 100 ng/ml rhBMP-2 or 200 ng/ml PAPP-A. After 12 h, total RNA was extracted and quantified as described above. RNA was then analyzed via qPCR for Papp-A, Runx2, or Igf-I gene expression.
Alizarin red staining
hMSCs were cultured in basal media supplemented with 10 mM -glycerophosphate and 100 ng/ml BMP-2, 200 ng/ml PAPP-A, or rhBMP-2 plus PAPP-A. Media were renewed every 4 d. After 28 d in culture, cells were fixed in 70% ice-cold ethanol for 1 h. Cells were rinsed with deionized water and stained with 40 mM alizarin red S (pH 4.2) for 10 min at room temperature with gentle agitation. Fixed cells were rinsed five times with deionized water and then with 1x PBS for 15 min with gentle agitation. Cells were imaged before alizarin red stain extraction with 10% cetylpyridinium chloride in 10 mM sodium phosphate buffer for 20 min at room temperature with gentle agitation. Extracted alizarin red stain in cetylpyridinium chloride was quantified by absorbance reading at 570 nm on a Tecan Spectrofluor. Amount of alizarin red stain was calculated based on a standard curve of known concentrations and normalized to milligrams total protein of the cell lysate.
Angiogenic assay on chick CAMs
Fertilized White Leghorn chicken eggs were incubated at 37 C and 70% humidity for 3 d. On the third day, the eggs were cracked into 100 mm x 20 mm petri dishes containing 4 ml DMEM and placed in a humidified incubator at 37 C for 7 d. To prepare fibrin constructs, a working fibrinogen solution was prepared: 20 mg/ml human fibrinogen in 20 mM HEPES, 100 mM NaCl (pH 7.4) and containing 1 μg/ml aprotinin. PAPP-A was added to the desired concentration (10 pg/ml, 100 pg/ml, and 1 ng/ml) and did not exceed 10% of the total volume. A fibrinogen/BSA construct was prepared to be the negative control. Fifty microliters of the fibrinogen/PAPP-A mix was added to 4 μl of 100 U/ml human thrombin in 100 mM CaCl2. Fifty microliters were immediately transferred to an 8 mm x 1 mm cylindrical mold and allowed to gel for 2–3 min. The fibrin gel constructs were removed and then placed on the chick CAM. After 2 d, 705 nm fluorescent Quantum Dots in 154 mM NaCl were injected into a chick CAM vein, and the blood vessels surrounding the fibrin construct were imaged using a x5 objective on a Stemi SV11 stereomicroscope (Zeiss, Gttingen, Germany) with a Retiga Exi charge-coupled device camera and QImaging software. Excitation was at 450 nm and emission was detected with a 660-nm-long pass filter. No less than five embryos were used for each treatment group.
Statistical analyses
For qPCR assays, the coefficient of variation was calculated from three assay replicates and did not exceed 3% for all treatment groups. Intraday variation did not exceed 5%. Treatment groups within experiments were performed in triplicate and reported as mean ± SEM. Statistical analysis was performed using SYSTAT 9 software (Richmond, CA) to determine significance among treatment groups. ANOVA followed by Tukey’s post hoc test was performed for which time was not an independent variable. ANOVA repeated-measures test followed by least significant differences post hoc test was performed in which time was an independent variable. P < 0.05 was considered significant. All experiments were repeated at least two times and one representative experiment is shown.
Results
BMP-2 directly up-regulates PAPP-A gene expression and protein
hMSCs that had been treated with rhBMP-2 demonstrated a 4.5-fold increase in PAPP-A gene expression after 12 h in culture. After 12 h, PAPP-A gene expression levels declined to basal levels, which were maintained at least until 48 h in culture (Fig. 1A). Treatment with the protein translation inhibitor cycloheximide did not inhibit rhBMP-2-stimulated PAPP-A gene expression, indicating that de novo protein synthesis is not required for BMP-2 to up-regulate PAPP-A (Fig. 1B). Runx2, a known direct target of rhBMP-2, served as a positive control. Thus, PAPP-A is a direct gene target of BMP-2 signaling. rhBMP-2 also stimulated an increase in PAPP-A protein production detected by ELISA after 4 and 6 d in culture (Fig. 1C), and the effect was dose dependent at 4 d (Fig. 1D). Subsequently we determined that treatment of hMSCs with rhBMP-2 did not result in production of a proteolytically active form of PAPP-A (Fig. 1E, lanes 3–8). Furthermore, purified PAPP-A protein used in the gene expression experiments maintained only weak protease activity as 500 ng of PAPP-A, and a 72-h incubation was required to reach substantial cleavage of IGFBP-4 (lanes 15 and 16). We suggest that any residual activity of purified PAPP-A may be extinguished by factors in serum, which was present during gene expression experiments.
PAPP-A directly up-regulates Runx2 gene expression
To determine whether PAPP-A has the capacity to osteoblast marker genes, we treated hMSCs with PAPP-A protein. hMSCs that had been treated with PAPP-A exhibited a 3.2-fold increase in Runx2 gene expression after 12 h in culture (Fig. 2A). After 24 h in culture, Runx2 gene expression continued to increase to 5-fold above untreated control hMSCs. Between 24 and 48 h in culture, Runx2 gene expression declined to basal level. Treatment with cycloheximide did not inhibit the increase in Runx2 gene expression stimulated by PAPP-A (Fig. 2B). Runx2 gene expression in the presence of PAPP-A and cycloheximide was significantly higher than PAPP-A alone. This is due to an accumulation of Runx2 transcript in the absence of protein translation. Runx2 is a direct gene target of PAPP-A; de novo protein synthesis is not required for stimulation of Runx2 gene expression by PAPP-A. Osx is inducible by rhBMP-2 in hMSCs (41, 42). In this study, Osx gene expression was not detected at basal levels or induced by PAPP-A (data not shown due to no fluorescent signal detected).
PAPP-A directly up-regulated Igf-I gene expression
It has been reported in the literature that BMP-2 can up-regulate Igf-I gene expression (35). We demonstrate in the present study that BMP-2 up-regulates Igf-I gene expression indirectly, meaning de novo protein synthesis is required for stimulation of Igf-I gene expression. Treatment of hMSCs with rhBMP-2 resulted in a 6-fold induction of Igf-I gene expression after 12 h in culture (Fig. 3A). Upon treatment with cycloheximide, Igf-I gene expression was equal to that of control (no induction). When hMSCs are treated with PAPP-A, a 2.6-fold increase in Igf-I gene expression was detected after 12 h in culture. In the presence of cycloheximide, Igf-I gene expression induced by PAPP-A was not inhibited. Therefore, PAPP-A does not require de novo protein synthesis to stimulate Igf-I gene expression.
PAPP-A and BMP-2 are sufficient for matrix mineralization in hMSCs
hMSCs do not mineralize in basal media supplemented in -glycerophosphate alone (Fig. 4A). Furthermore, addition of either rhBMP-2 (Fig. 4B) or PAPP-A alone (Fig. 4C) did not result in visible alizarin red staining of hMSC cultures. hMSCs that had been treated with a combination of rhBMP-2 and PAPP-A for 28 d exhibited positive alizarin red staining (Fig. 4D). Figure 4E shows the quantification of alizarin stain for each treatment group. No appreciable staining is observed in basal or rhBMP-2- or PAPP-A-treated wells. The combination of rhBMP-2 and PAPP-A results in an accumulation of significant alizarin red stain.
PAPP-A stimulates new blood vessel formation in the chick CAM
Crucial to bone development and remodeling is the presence of a blood supply. To determine whether PAPP-A has angiogenic properties in vivo, we exploited the chick CAM model system. The CAM is widely used to study both angiogenic and antiangiogenic agents. Several reviews have discussed its capabilities and advantages as well as its considerations for experimental design (43, 44, 45). Here we demonstrate that a BSA control does not induce new blood vessel formation in the CAM (Fig. 5A), whereas PAPP-A protein loaded within fibrin constructs induced an angiogenic response in the CAM (Fig. 5, B–D). Angiogenesis in this CAM model is evidenced by the spoke-wheel appearance of blood vessels around the fibrin construct. The lowest dose of PAPP-A tested, 10 pg/ml, showed a slight response despite minor degradation of the fibrin construct (Fig. 5B). Positive responses were evident for both the 100 pg/ml and 1 ng/ml doses of PAPP-A (Fig. 5, C and D).
Discussion
The results of this study implicate PAPP-A as a key molecule in osteoblast differentiation. Treatment of hMSCs with rhBMP-2 resulted in direct up-regulation of Papp-A gene expression. TGF was previously shown to up-regulate Papp-A gene expression in human osteoblasts (4). As a member of the TGF superfamily, it is reasonable that BMP-2 has a similar effect. In addition, studies by others have shown that BMP-2 can up-regulate Igf-I gene expression (35). In our laboratory, we observed that this up-regulation is indirect and a protein mediator is required for stimulation of Igf-I gene expression by BMP-2. We also demonstrated that uncommitted hMSCs when treated with PAPP-A initiate changes in gene expression including up-regulation of the transcription factor Runx2. Figure 6 depicts our proposed model of the cascade of events that could contribute to Runx2 gene expression and the relationship among BMP-2, IGF-I, and PAPP-A. Although PAPP-A did not induce Osx gene expression when added in combination with BMP-2, PAPP-A can stimulate matrix mineralization in the absence of other osteogenic supplements such as dexamethasone. This suggests alternate signaling pathways for BMP-2 and PAPP-A that are each essential for mineralization. However, the common involvement of Runx2 suggests a potential point of convergence between the pathways. The cooperation between PAPP-A and BMP-2 in matrix mineralization could involve Osx but remains unclear at this time. Our data further suggest that PAPP-A, which we demonstrated to be regulated by BMP-2, is also involved in angiogenesis because new blood vessel formation was observed in the chick CAM model after PAPP-A treatment.
PAPP-A is an IGFBP-4 protease and potentially has a role in bone formation by regulating IGF bioavailability. We and others (15) demonstrated the effects of IGF-I on osteoblast marker gene expression, specifically in the early commitment of a progenitor cell to the osteogenic lineage (41). In the present study, we demonstrate that PAPP-A can up-regulate Runx2 gene expression directly, in the absence of de novo protein synthesis. Our first thought was that PAPP-A functions in up-regulation of Runx2 via an increase in IGF-I bioavailability to the cell; however, our recent data suggest a nonproteolytic role. PAPP-A used in this study was only weakly active as protease, as was the PAPP-A induced by BMP-2. Therefore, it is unlikely that PAPP-A induction of Runx2 is dependent on IGF-I bioavailability and suggests a growth factor-like signaling mechanism for PAPP-A.
We therefore suggest that PAPP-A may elicit effects on gene expression via nonproteolytic mechanism(s). As such, the stimulation of hMSCs toward osteoblast maturation by PAPP-A does not rely solely on IGFBP-4 proteolytic cleavage from IGFs. Recently Sivanandam et al. (46) demonstrated that protein kinase C (PKC) is involved in the conversion of PAPP-A470 to lower-molecular-weight PAPP-A400, resulting in a 4-fold reduction in protease activity. Because BMP-2 is known to stimulate PKC activity (23), it is possible that PAPP-A stimulated by BMP-2 presented here may also involve PKC activity and thus induce the less active form of PAPP-A400. PKC stimulation did not affect proMBP levels; therefore, a less active PAPP-A400 is not likely due to complex formation with its inhibitor proMBP (proform of eosinophil major basic protein). The authors speculate that this conversion by PKC is unlikely caused by proteolytic cleavage of PAPP-A because reduced monomers of PAPP-A470 comigrated with PAPP-A400. A similar phenomenon could occur in our system where BMP-2 might dictate PAPP-A activity and perhaps its functions.
The combination of PAPP-A and rhBMP-2 was sufficient for matrix mineralization of hMSCs, which is indicative of end-stage osteoblastic differentiation. Significant calcium deposition was not detected in the presence of either PAPP-A or rhBMP-2 alone. However, it is interesting to note that PAPP-A plus BMP-2-induced mineral deposition in the absence of any other traditional osteogenic supplements such as dexamethasone. Recently we observed an identical outcome of hMSC matrix mineralization after adenoviral transduction with Igf-I and treatment with rhBMP-2 (data not shown). Our findings have important implication for the design of new cocktails that promote proper bone formation with reduced side effects. These data agree with our recent report as well as studies by others that support a synergistic role between IGF-I and BMP-2 (33, 41). Perhaps PAPP-A, either through IGF-I or independently, permits BMP-2 to signal optimally and thus enhance bioactivity and improve outcome (matrix mineralization). Whereas neither factor can act alone to sufficiently achieve mineralization, their combined actions result in the cells’ ability to differentiate toward the osteogenic lineage and form a mineralized extracellular matrix. The present study also agrees with respect to Osx gene expression. Osx was not induced by adenoviral delivery of Igf-I or treatment with PAPP-A. Osx is required for ossification and as stimulated by BMP-2 seems insufficient to induce mineralization in hMSCs alone. However, in the presence of either IGF-I or PAPP-A, BMP-2-treated cells exhibit a mineralized extracellular matrix. Perhaps the increase in IGF-I due to viral gene delivery or PAPP-A stimulation modifies BMP-2 signaling such that mineralization is possible. It is unclear whether BMP-2 signaling modulated by IGF-I involves Osx.
PAPP-A’s angiogenic capacity alone reinforces the theory that PAPP-A itself is a powerful signal in key developmental, repair, and regenerative processes. In the literature, BMP-2 and IGF-I have been reported to be involved in angiogenesis (5, 7, 47). It is logical that PAPP-A, as a downstream direct target of BMP-2 and potential intermediate between BMP-2 and IGF-I, could have an angiogenic role. Reports from others (9) have strongly implicated PAPP-A in a developmental and healing context. As shown in the present study, the induction of new blood vessels formation by PAPP-A in the chick CAM model makes PAPP-A an attractive candidate for therapeutic applications.
In conclusion, our data suggest that PAPP-A plays a key regulatory role in bone formation. PAPP-A could function in osteogenic lineage progression via proteolytic and/or nonproteolytic mechanisms. At this time, it is unclear what dictates one function over another or the advantages of each when considering the desirable end result of matrix mineralization, angiogenesis, and tissue development. Given that PAPP-A is regulated by many factors (4, 48), it could have multiple roles. Our future work will focus on elucidating the signaling mechanisms of PAPP-A in bone formation, particularly addressing PAPP-A as a potential mediator of BMP-2 and IGF-I.
Acknowledgments
We especially thank Laurie Bale of Conover Lab for performing the protease activity assay. The authors thank Dr. Jeffrey Hollinger (director of the Bone Tissue Engineering Center) for providing the rhBMP-2 and partial financial support for this work.
Footnotes
This work was supported by Pennsylvania Infrastructure Technology Alliance and Bone Tissue Engineering Center, Carnegie Mellon University.
Abbreviations: BMP, Bone morphogenetic protein; CAM, chick chorioallantoic membrane; CT, cycle threshold; hMSC, human adult mesenchymal stem cell; IGFBP, IGF binding protein; OSX, osterix; PAPP-A, pregnancy-associated plasma protein A; PKC, protein kinase C; qPCR, quantitative real-time PCR; rh, recombinant human; RUNX, runt-related transcription factor.
References
Lawrence JB, Oxvig C, Overgaard MT, Sottrup-Jensen L, Gleich GJ, Hayes LG, Yates 3rd JR, Covover CA 1999 The insulin-like growth factor (IGF)-dependent IGF binding protein-4 protease secreted by human fibroblasts is pregnancy-associated plasma protein-A. Proc Natl Acad Sci USA 96:3149–3153
Brambati B, Macintosh MC, Teisner B, Maguiness S, Shrimanker K, Lanzani A, Bonacchi I, Tului L, Chard T, Grudzinskas JG 1993 Low maternal serum levels of pregnancy associated plasma protein A (PAPP-A) in the first trimester in association with abnormal fetal karyotype. Br J Obstet Gynaecol 100:324–326
Chen BK, Leiferman KM, Pittelkow MR, Overgaard MT, Oxvig C, Conover CA 2003 Localization and regulation of pregnancy-associated plasma protein a expression in healing human skin. J Clin Endocrinol Metab 88:4465–4471
Ortiz CO, Chen BK, Bale LK, Overgaard MT, Oxvig C, Conover CA 2003 Transforming growth factor- regulation of the insulin-like growth factor binding protein-4 protease system in cultured human osteoblasts. J Bone Miner Res 18:1066–1072
Rabinovsky ED, Draghia-Akli R 2004 Insulin-like growth factor I plasmid therapy promotes in vivo angiogenesis. Mol Ther 9:46–55
Dunn SE 2000 Insulin-like growth factor I stimulates angiogenesis and the production of vascular endothelial growth factor. Growth Horm IGF Res 10(Suppl A):S41–S42
Shigematsu S, Yamauchi K, Nakajima K, Iijima S, Aizawa T, Hashizume K 1999 IGF-1 regulates migration and angiogenesis of human endothelial cells. Endocr J 46(Suppl):S59–S62
Bayes-Genis A, Conover CA, Overgaard MT, Bailey KR, Christiansen M, Holmes Jr DR, Virmani R, Oxvig C, Schwartz RS 2001 Pregnancy-associated plasma protein A as a marker of acute coronary syndromes. N Engl J Med 345:1022–1029
Conover CA, Bale LK, Overgaard MT, Johnstone EW, Laursen UH, Fuchtbauer EM, Oxvig C, van Deursen J 2004 Metalloproteinase pregnancy-associated plasma protein A is a critical growth regulatory factor during fetal development. Development 131:1187–1194
Qin QP, Laitinen P, Majamaa-Voltti K, Eriksson S, Kumpula EK, Pettersson K 2002 Release patterns of pregnancy associated plasma protein A (PAPP-A) in patients with acute coronary syndromes. Scand Cardiovasc J 36:358–361
Alden TD, Beres EJ, Laurent JS, Engh JA, Das S, London SD, Jane Jr JA, Hudson SB, Helm GA 2000 The use of bone morphogenetic protein gene therapy in craniofacial bone repair. J Craniofac Surg 11:24–30
Campisi P, Hamdy RC, Lauzier D, Amako M, Rauch F, Lessard ML 2003 Expression of bone morphogenetic proteins during mandibular distraction osteogenesis. Plast Reconstr Surg 111:201–208; discussion 209–210
Lee MH, Kwon TG, Park HS, Wozney JM, Ryoo HM 2003 BMP-2-induced osterix expression is mediated by Dlx5 but is independent of Runx2. Biochem Biophys Res Commun 309:689–694
Diefenderfer DL, Osyczka AM, Reilly GC, Leboy PS 2003 BMP responsiveness in human mesenchymal stem cells. Connect Tissue Res 44(Suppl 1):305–311
Walsh S, Jefferiss CM, Stewart K, Beresford JN 2003 IGF-I does not affect the proliferation or early osteogenic differentiation of human marrow stromal cells. Bone 33:80–89
Rubin J, Ackert-Bicknell CL, Zhu L, Fan X, Murphy TC, Nanes MS, Marcus R, Holloway L, Beamer WG, Rosen CJ 2002 IGF-I regulates osteoprotegerin (OPG) and receptor activator of nuclear factor-B ligand in vitro and OPG in vivo. J Clin Endocrinol Metab 87:4273–4279
Nohe A, Hassel S, Ehrlich M, Neubauer F, Sebald W, Henis YI, Knaus P 2002 The mode of bone morphogenetic protein (BMP) receptor oligomerization determines different BMP-2 signaling pathways. J Biol Chem 277:5330–5338
Fiedler J, Roderer G, Gunther KP, Brenner RE 2002 BMP-2, BMP-4, and PDGF-bb stimulate chemotactic migration of primary human mesenchymal progenitor cells. J Cell Biochem 87:305–312
Zhang M, Xuan S, Bouxsein ML, von Stechow D, Akeno N, Faugere MC, Malluche H, Zhao G, Rosen CJ, Efstratiadis A, Clemons TL 2002 Osteoblast-specific knockout of the insulin-like growth factor (IGF) receptor gene reveals an essential role of IGF signaling in bone matrix mineralization. J Biol Chem 277:44005–44012
Deckers MM, van Bezooijen RL, van der Horst G, Hoogendam J, van Der Bent C, Papapoulos SE, Lowik CW 2002 Bone morphogenetic proteins stimulate angiogenesis through osteoblast-derived vascular endothelial growth factor A. Endocrinology 143:1545–1553
Tanaka H, Wakisaka A, Ogasa H, Kawai S, Liang CT 2002 Effect of IGF-I and PDGF administered in vivo on the expression of osteoblast-related genes in old rats. J Endocrinol 174:63–70
Boden SD, Kang J, Sandhu H, Heller JG 2002 Use of recombinant human bone morphogenetic protein-2 to achieve posterolateral lumbar spine fusion in humans: a prospective, randomized clinical pilot trial: 2002 Volvo Award in Clinical Studies. Spine 27:2662–2673
Hay E, Lemonnier J, Fromigue O, Marie PJ 2001 Bone morphogenetic protein-2 promotes osteoblast apoptosis through a Smad-independent, protein kinase C-dependent signaling pathway. J Biol Chem 276:29028–29036
Lou J, Tu Y, Li S, Manske PR 2000 Involvement of ERK in BMP-2 induced osteoblastic differentiation of mesenchymal progenitor cell line C3H10T1/2. Biochem Biophys Res Commun 268:757–762
Boden SD, Zdeblick TA, Sandhu HS, Heim SE 2000 The use of rhBMP-2 in interbody fusion cages. Definitive evidence of osteoinduction in humans: a preliminary report. Spine 25:376–381
Thomas T, Gori F, Spelsberg TC, Khosla S, Riggs BL, Conover CA 1999 Response of bipotential human marrow stromal cells to insulin-like growth factors: effect on binding protein production, proliferation, and commitment to osteoblasts and adipocytes. Endocrinology 140:5036–5044
Higuchi T, Kinoshita A, Takahashi K, Oda S, Ishikawa I 1999 Bone regeneration by recombinant human bone morphogenetic protein-2 in rat mandibular defects. An experimental model of defect filling. J Periodontol 70:1026–1031
Boden SD, Martin Jr GJ, Horton WC, Truss TL, Sandhu HS 1998 Laparoscopic anterior spinal arthrodesis with rhBMP-2 in a titanium interbody threaded cage. J Spinal Disord 11:95–101
Han VK, Bassett N, Walton J, Challis JR 1996 The expression of insulin-like growth factor (IGF) and IGF-binding protein (IGFBP) genes in the human placenta and membranes: evidence for IGF-IGFBP interactions at the feto-maternal interface. J Clin Endocrinol Metab 81:2680–2693
Mohan S, Baylink DJ 1996 Insulin-like growth factor system components and the coupling of bone formation to resorption. Horm Res 45(Suppl 1):59–62
Muschler GF, Hyodo A, Manning T, Kambic H, Easley K 1994 Evaluation of human bone morphogenetic protein 2 in a canine spinal fusion model. Clin Orthop 229–240
Ahrens M, Ankenbauer T, Schroder D, Hollnagel A, Mayer H, Gross G 1993 Expression of human bone morphogenetic proteins-2 or -4 in murine mesenchymal progenitor C3H10T1/2 cells induces differentiation into distinct mesenchymal cell lineages. DNA Cell Biol 12:871–880
Raiche AT, Puleo DA 2004 In vitro effects of combined and sequential delivery of two bone growth factors. Biomaterials 25:677–685
Raiche AT, Puleo DA 2004 Cell responses to BMP-2 and IGF-I released with different time-dependent profiles. J Biomed Mater Res 69A:342–350
Canalis E, Gabbitas B 1994 Bone morphogenetic protein 2 increases insulin-like growth factor I and II transcripts and polypeptide levels in bone cell cultures. J Bone Miner Res 9:1999–2005
Nakashima K, Zhou X, Kunkel G, Zhang Z, Deng JM, Behringer RR, de Crombrugghe B 2002 The novel zinc finger-containing transcription factor osterix is required for osteoblast differentiation and bone formation. Cell 108:17–29
Ducy P, Starbuck M, Priemel M, Shen J, Pinero G, Geoffroy V, Amling M, Karsenty G 1999 A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev 13:1025–1036
Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G 1997 Osf2/Cbfa1: a transcriptional activator of osteoblast differentiation. Cell 89:747–754
Otto F, Thornell AP, Crompton T, Denzel A, Gilmour KC, Rosewell IR, Stamp GW, Beddington RS, Mundlos S, Olsen BR, Selby PB, Owen MJ 1997 Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 89:765–771
Applied Biosystems 2001 User bulletin #2: ABI Prism 7700 Sequence Detection System. Foster City, CA: Applied Biosystems
Celil AB, Hollinger JO, Campbell PG 2005 Osx transcriptional regulation is mediated by additional pathways to BMP2/Smad signaling. J Cell Biochem 95:518–528
Jadlowiec J, Koch H, Zhang X, Campbell PG, Seyedain M, Sfeir C 2004 Phosphophoryn regulates the gene expression and differentiation of NIH3T3, MC3T3–E1, and human mesenchymal stem cells via the integrin/MAPK signaling pathway. J Biol Chem 279:53323–53330
Ribatti D, Presta M 2002 The role of fibroblast growth factor-2 in the vascularization of the chick embryo chorioallantoic membrane. J Cell Mol Med 6:439–446
Ribatti D, Nico B, Vacca A, Roncali L, Burri PH, Djonov V 2001 Chorioallantoic membrane capillary bed: a useful target for studying angiogenesis and anti-angiogenesis in vivo. Anat Rec 264:317–324
Ribatti D, Vacca A, Roncali L, Dammacco F 1996 The chick embryo chorioallantoic membrane as a model for in vivo research on angiogenesis. Int J Dev Biol 40:1189–1197
Sivanandam AS, Mohan S, Kita H, Kapur S, Chen ST, Linkhart TA, Bagi G, Baylink DJ, Qin X 2004 Studies on regulation of IGF binding protein (IGFBP)-4 proteolysis by pregnancy-associated plasma protein-A (PAPP-A) in cells treated with phorbol ester. Biochem J 379:57–64
Marrony S, Bassilana F, Seuwen K, Keller H 2003 Bone morphogenetic protein 2 induces placental growth factor in mesenchymal stem cells. Bone 33:426–433
Conover CA, Chen BK, Resch ZT 2004 Regulation of pregnancy-associated plasma protein-A expression in cultured human osteoblasts. Bone 34:297–302(Julie Jadlowiec, Diana Do)
Endocrine Research Unit (C.C.), The Mayo Clinic and Foundation, Rochester, Minnesota 55905
Abstract
Pregnancy-associated plasma protein A (PAPP-A) is an IGF binding protein 4 protease that can function to increase local IGF-I bioavailability. Aside from its assumed role during pregnancy, in vitro and in vivo studies have indicated roles for PAPP-A in IGF-I-mediated wound healing, vascular repair, and bone formation. Because bone morphogenetic protein 2 (BMP-2) is known to up-regulate Igf-I gene expression, we hypothesized that PAPP-A may be involved in BMP-2 mechanisms in bone formation. To test this hypothesis, we quantified gene expression of Papp-A in response to BMP-2 treatment and runt-related transcription factor 2, Osterix, and Igf-I in response to PAPP-A protein treatment in human adult mesenchymal stem cells. Our results demonstrate that BMP-2 directly up-regulated Papp-A gene and protein expression. Purified PAPP-A protein directly up-regulated runt-related transcription factor 2 and Igf-I gene expression but not Osterix. When added in combination with recombinant human BMP-2, PAPP-A increased matrix mineralization in the absence of dexamethasone. PAPP-A further demonstrated an angiogenic effect in the chick chorioallontoic membrane, which implicates a critical developmental role and possible therapeutic potential. Our findings suggest that PAPP-A functions in the formation of mineralized tissues through direct up-regulation of key genes. Furthermore, PAPP-A is involved in the formation of new blood vessels, which is essential for proper bone regeneration.
Introduction
PREGNANCY-ASSOCIATED plasma protein-A (PAPP-A) is a zinc-dependent metalloproteinase that was originally isolated in 1974 from normal human pregnancy serum (1). In recent years, circulating PAPP-A levels in maternal serum have been used to screen and diagnose fetal Down’s syndrome with lower levels of PAPP-A correlated with increased Down’s syndrome risk (2). PAPP-A has also been used as a marker for Cornelia de Lange syndrome, which results in mental and growth developmental delay, limb reduction abnormalities, a distinct facial appearance, and congenital heart defects (1). Although PAPP-A appears to have specific and important functions during pregnancy, its function in vivo is not known. PAPP-A has also been localized outside placental tissue in preovulatory granulosa, healing human skin, and atherosclerotic plaques as well as smooth muscle cells, fibroblasts and osteoblasts in culture (3).
PAPP-A has been identified as an IGF binding protein (IGFBP)-4 protease (1). IGFBP-4 binds to and inhibits the actions of IGFs. IGFs are mitogenic peptides that regulate fetal and neonatal growth. It has been postulated that PAPP-A increases bioavailability of IGFs by cleaving IGFBP-4 into fragments with low affinity for IGFs, thus releasing IGFs and permitting their action on local cells. A recent study suggested that increased levels of PAPP-A protein may promote local bone formation (4). Because bone morphogenetic protein (BMP)-2 and IGF-I have important angiogenic properties in normal skeletal and vascular development (5, 6, 7), we reason that PAPP-A could be a component of such mechanisms.
PAPP-A is known to also have pathological functions; elevated levels of PAPP-A are found in atherosclerotic plaques, in which many bone-related extracellular matrix proteins such as osteopontin, osteonectin, and osteocalcin have been localized (8). PAPP-A has been postulated to regulate bone formation via IGF-I signaling and increasing IGF bioavailability. TGF1 was shown to up-regulate PAPP-A message and protein expression (4). Regulation of PAPP-A by TGF1 lends credence to the proposed model of IGF-I and IGFBP-4 as key components of the coupling mechanisms between bone formation and bone resorption (4). Recently Papp-A knockout mice were shown to be significantly smaller in size (60%), compared with their wild-type littermates (9). The phenotype of PAPP-A-null mice is identical with the IGF-II knockout. However, normal levels of Igf-II transcripts are detectable in PAPP-A-null mice. The authors suggest that PAPP-A has an important regulatory role in embryonic growth and postnatal development. They further speculated that PAPP-A is an important modulator of IGF signaling, which helps to determine organ and body size in mammals. The bones of PAPP-A–/– knockout mice were visibly smaller than wild-type mice, thus supporting a role for PAPP-A in bone formation and development.
Thus, it is hypothesized that PAPP-A may have important and pivotal roles in local cellular function related to wound healing, bone remodeling, atherosclerotic plaque development, angiogenesis, and several aspects of human reproduction (10). To our knowledge, there have been no studies to date that investigate bone marker gene expression in vitro in response to administered purified PAPP-A. We hypothesize that PAPP-A has a critical role in bone formation, specifically at the molecular level and possibly within the BMP-2 and IGF-I signaling pathways. BMP-2 and IGF-I are growth factors critical for bone formation and their importance for progression of the osteogenic pathway has been well documented (11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32). Although there appears to be a close connection between BMP-2 and IGF-I (33, 34) and BMP-2 was shown to up-regulate Igf-I gene expression (35), a mechanism of cooperation has not been elucidated. BMP-2 is an inducer of the osteoblast transcription factors runt-related transcription factor (RUNX)2 and osterix (OSX), which are crucial for bone formation (36, 37, 38, 39). Runx2–/– mice display a cartilaginous skeleton, with no bone formation (39). Osx–/– mice do not express osteoblast marker genes such as osteocalcin but exhibit normal Runx2 levels, suggesting that OSX operates downstream of RUNX2 (36).
In the present study, we demonstrate that PAPP-A is involved in BMP-2 signaling and functions in up-regulation of Igf-I and Runx2 gene expression. We further demonstrate that PAPP-A can stimulate matrix mineralization in vitro together with BMP-2 in human adult mesenchymal stem cells (hMSCs). PAPP-A was also shown to have an angiogenic effect in the chick chorioallantoic membrane (CAM) model. The comprehensive data presented here suggest that PAPP-A may be a critical component in bone development and could be widely useful in therapeutic applications.
Materials and Methods
hMSCs were obtained from BioWhittaker, Inc. (Walkersville, MD). hMSCs were certified by the manufacturer as positive for potential toward adipogenic, chondrogenic, and osteogenic pathways. Human mesenchymal stem cell medium, mesenchymal cell growth supplement, L-glutamine, penicillin, and streptomycin were obtained from BioWhittaker and added to the medium according to the manufacturer’s specifications to prepare basal medium. Purified PAPP-A protein was purchased from Advanced Immunochemical (Long Beach, CA). PAPP-A ELISA was obtained from Diagnostic Systems Laboratories (Webster, TX). RNeasy kit and DNase I were obtained from QIAGEN (Valencia, CA). RiboGreen kit was obtained from Molecular Probes (Eugene, OR). All quantitative real-time PCR reagents, primers, and probes were purchased from Applied Biosystems (Foster City, CA). Human fibrinogen was obtained from Aventis (Bridgewater, NJ). Thrombin was obtained from Centeon Pharma (Liederbach, Germany). Quantum Dots were a gift from Quantum Dot Corp. (Hayward, CA). Total protein kits were obtained from Bio-Rad Laboratories (Temecula, CA). All other chemicals were obtained from Sigma Inc. (St. Louis, MO).
Cell culture
hMSCs were seeded in 35-mm dishes and grown in basal media to approximately 70% confluence. hMSCs were then treated with 100 ng/ml recombinant human (rh)BMP-2 for 6, 12, 24, and 48 h. hMSCs were also cultured separately in media containing 200 ng/ml PAPP-A in basal media for 12, 24, 48, and 96 h. Media were renewed every 48 h until RNA extraction.
Analysis of IGFBP-4 proteolytic activity
hMSCs were cultured for 6 d in a 24-well culture dish in media containing 100 ng/ml rhBMP-2. Media was renewed every other day (total volume 500 μl). Conditions were switched to serum free for the final 48 h of culture. Conditioned media were collected and frozen at –80 C until use. Proteolysis of IGFBP-4 was determined by previously established protocols (1). Conditioned media were incubated for 6 and 72 h with 125I-IGFBP-4 and in the presence or absence of IGF-II.
RNA extraction and quantification
Total RNA was extracted using the RNeasy kit according to the manufacturer’s instructions including DNase I treatment to eliminate any residual genomic DNA. RNA was quantified using the RiboGreen kit (Molecular Probes). After addition of the working solution according to the manufacturer’s instructions to RNA samples, fluorescence was quantified using a Tecan (Zurich, Switzerland) Spectrofluor with excitation at 485 and emission at 535 nm. RNA concentrations were calculated based on a standard curve of rRNA of known concentrations.
Quantitative real-time PCR
Cells were harvested from the culture treatments at the time points described above. After extraction and quantification of RNA, quantitative real-time PCR (qPCR) analysis was carried out using Taqman one-step RT-PCR master mix. Ten nanograms of total RNA were added per 25-μl reaction with sequence-specific primers (200 nM) and Taqman probes (200 nM). Sequences for all target gene primers and probes are shown in Table 1. 18S primers and probes were designed by and purchased from Applied Biosystems. qPCR assays were carried out in triplicate on an ABI Prism 7000 sequence detection system. Thermocycling conditions were as follows: 48 C for 30 min (reverse transcription) and 95 C for 10 min (initial denaturation) followed by 40 cycles at 95 C for 15 sec (denaturation) and 60 C for 45 sec (annealing and extension). The threshold was set above the nontemplate control background and within the linear phase of target gene amplification to calculate the cycle number at which the transcript was detected [denoted cycle threshold (CT)].
Gene expression values were calculated based on the comparative CT method (separate tubes) as detailed in Applied Biosystems user bulletin 2 (40). For each primers/probe set, validation experiments demonstrated that efficiencies of target and reference gene amplification were approximately equal; the absolute value of the slope of log input amount vs. CT < 0.1 (data not shown). Target genes were normalized to the reference housekeeping gene, 18S. Fold differences were calculated for each treatment group using normalized CT values for the negative control at the appropriate time point as the calibrator.
PAPP-A ELISA
hMSCs were cultured in basal media without serum supplemented with rhBMP-2. Media were collected after 12 and 48 h, 4 and 6 d total in culture and frozen at –80 C until use. Samples were thawed, and 50 μl of conditioned medium were subjected to analysis by ELISA according to the manufacturer’s protocol. PAPP-A concentration (in microinternational units per milliliter) was calculated based on a standard curve of known concentrations.
Inhibition of protein synthesis
hMSCs were seeded in 35-mm dishes as before and grown to approximately 70% confluence. Cells were pretreated with 5 μg/ml cycloheximide for 30 min before treatment with basal media or supplemented with 100 ng/ml rhBMP-2 or 200 ng/ml PAPP-A. After 12 h, total RNA was extracted and quantified as described above. RNA was then analyzed via qPCR for Papp-A, Runx2, or Igf-I gene expression.
Alizarin red staining
hMSCs were cultured in basal media supplemented with 10 mM -glycerophosphate and 100 ng/ml BMP-2, 200 ng/ml PAPP-A, or rhBMP-2 plus PAPP-A. Media were renewed every 4 d. After 28 d in culture, cells were fixed in 70% ice-cold ethanol for 1 h. Cells were rinsed with deionized water and stained with 40 mM alizarin red S (pH 4.2) for 10 min at room temperature with gentle agitation. Fixed cells were rinsed five times with deionized water and then with 1x PBS for 15 min with gentle agitation. Cells were imaged before alizarin red stain extraction with 10% cetylpyridinium chloride in 10 mM sodium phosphate buffer for 20 min at room temperature with gentle agitation. Extracted alizarin red stain in cetylpyridinium chloride was quantified by absorbance reading at 570 nm on a Tecan Spectrofluor. Amount of alizarin red stain was calculated based on a standard curve of known concentrations and normalized to milligrams total protein of the cell lysate.
Angiogenic assay on chick CAMs
Fertilized White Leghorn chicken eggs were incubated at 37 C and 70% humidity for 3 d. On the third day, the eggs were cracked into 100 mm x 20 mm petri dishes containing 4 ml DMEM and placed in a humidified incubator at 37 C for 7 d. To prepare fibrin constructs, a working fibrinogen solution was prepared: 20 mg/ml human fibrinogen in 20 mM HEPES, 100 mM NaCl (pH 7.4) and containing 1 μg/ml aprotinin. PAPP-A was added to the desired concentration (10 pg/ml, 100 pg/ml, and 1 ng/ml) and did not exceed 10% of the total volume. A fibrinogen/BSA construct was prepared to be the negative control. Fifty microliters of the fibrinogen/PAPP-A mix was added to 4 μl of 100 U/ml human thrombin in 100 mM CaCl2. Fifty microliters were immediately transferred to an 8 mm x 1 mm cylindrical mold and allowed to gel for 2–3 min. The fibrin gel constructs were removed and then placed on the chick CAM. After 2 d, 705 nm fluorescent Quantum Dots in 154 mM NaCl were injected into a chick CAM vein, and the blood vessels surrounding the fibrin construct were imaged using a x5 objective on a Stemi SV11 stereomicroscope (Zeiss, Gttingen, Germany) with a Retiga Exi charge-coupled device camera and QImaging software. Excitation was at 450 nm and emission was detected with a 660-nm-long pass filter. No less than five embryos were used for each treatment group.
Statistical analyses
For qPCR assays, the coefficient of variation was calculated from three assay replicates and did not exceed 3% for all treatment groups. Intraday variation did not exceed 5%. Treatment groups within experiments were performed in triplicate and reported as mean ± SEM. Statistical analysis was performed using SYSTAT 9 software (Richmond, CA) to determine significance among treatment groups. ANOVA followed by Tukey’s post hoc test was performed for which time was not an independent variable. ANOVA repeated-measures test followed by least significant differences post hoc test was performed in which time was an independent variable. P < 0.05 was considered significant. All experiments were repeated at least two times and one representative experiment is shown.
Results
BMP-2 directly up-regulates PAPP-A gene expression and protein
hMSCs that had been treated with rhBMP-2 demonstrated a 4.5-fold increase in PAPP-A gene expression after 12 h in culture. After 12 h, PAPP-A gene expression levels declined to basal levels, which were maintained at least until 48 h in culture (Fig. 1A). Treatment with the protein translation inhibitor cycloheximide did not inhibit rhBMP-2-stimulated PAPP-A gene expression, indicating that de novo protein synthesis is not required for BMP-2 to up-regulate PAPP-A (Fig. 1B). Runx2, a known direct target of rhBMP-2, served as a positive control. Thus, PAPP-A is a direct gene target of BMP-2 signaling. rhBMP-2 also stimulated an increase in PAPP-A protein production detected by ELISA after 4 and 6 d in culture (Fig. 1C), and the effect was dose dependent at 4 d (Fig. 1D). Subsequently we determined that treatment of hMSCs with rhBMP-2 did not result in production of a proteolytically active form of PAPP-A (Fig. 1E, lanes 3–8). Furthermore, purified PAPP-A protein used in the gene expression experiments maintained only weak protease activity as 500 ng of PAPP-A, and a 72-h incubation was required to reach substantial cleavage of IGFBP-4 (lanes 15 and 16). We suggest that any residual activity of purified PAPP-A may be extinguished by factors in serum, which was present during gene expression experiments.
PAPP-A directly up-regulates Runx2 gene expression
To determine whether PAPP-A has the capacity to osteoblast marker genes, we treated hMSCs with PAPP-A protein. hMSCs that had been treated with PAPP-A exhibited a 3.2-fold increase in Runx2 gene expression after 12 h in culture (Fig. 2A). After 24 h in culture, Runx2 gene expression continued to increase to 5-fold above untreated control hMSCs. Between 24 and 48 h in culture, Runx2 gene expression declined to basal level. Treatment with cycloheximide did not inhibit the increase in Runx2 gene expression stimulated by PAPP-A (Fig. 2B). Runx2 gene expression in the presence of PAPP-A and cycloheximide was significantly higher than PAPP-A alone. This is due to an accumulation of Runx2 transcript in the absence of protein translation. Runx2 is a direct gene target of PAPP-A; de novo protein synthesis is not required for stimulation of Runx2 gene expression by PAPP-A. Osx is inducible by rhBMP-2 in hMSCs (41, 42). In this study, Osx gene expression was not detected at basal levels or induced by PAPP-A (data not shown due to no fluorescent signal detected).
PAPP-A directly up-regulated Igf-I gene expression
It has been reported in the literature that BMP-2 can up-regulate Igf-I gene expression (35). We demonstrate in the present study that BMP-2 up-regulates Igf-I gene expression indirectly, meaning de novo protein synthesis is required for stimulation of Igf-I gene expression. Treatment of hMSCs with rhBMP-2 resulted in a 6-fold induction of Igf-I gene expression after 12 h in culture (Fig. 3A). Upon treatment with cycloheximide, Igf-I gene expression was equal to that of control (no induction). When hMSCs are treated with PAPP-A, a 2.6-fold increase in Igf-I gene expression was detected after 12 h in culture. In the presence of cycloheximide, Igf-I gene expression induced by PAPP-A was not inhibited. Therefore, PAPP-A does not require de novo protein synthesis to stimulate Igf-I gene expression.
PAPP-A and BMP-2 are sufficient for matrix mineralization in hMSCs
hMSCs do not mineralize in basal media supplemented in -glycerophosphate alone (Fig. 4A). Furthermore, addition of either rhBMP-2 (Fig. 4B) or PAPP-A alone (Fig. 4C) did not result in visible alizarin red staining of hMSC cultures. hMSCs that had been treated with a combination of rhBMP-2 and PAPP-A for 28 d exhibited positive alizarin red staining (Fig. 4D). Figure 4E shows the quantification of alizarin stain for each treatment group. No appreciable staining is observed in basal or rhBMP-2- or PAPP-A-treated wells. The combination of rhBMP-2 and PAPP-A results in an accumulation of significant alizarin red stain.
PAPP-A stimulates new blood vessel formation in the chick CAM
Crucial to bone development and remodeling is the presence of a blood supply. To determine whether PAPP-A has angiogenic properties in vivo, we exploited the chick CAM model system. The CAM is widely used to study both angiogenic and antiangiogenic agents. Several reviews have discussed its capabilities and advantages as well as its considerations for experimental design (43, 44, 45). Here we demonstrate that a BSA control does not induce new blood vessel formation in the CAM (Fig. 5A), whereas PAPP-A protein loaded within fibrin constructs induced an angiogenic response in the CAM (Fig. 5, B–D). Angiogenesis in this CAM model is evidenced by the spoke-wheel appearance of blood vessels around the fibrin construct. The lowest dose of PAPP-A tested, 10 pg/ml, showed a slight response despite minor degradation of the fibrin construct (Fig. 5B). Positive responses were evident for both the 100 pg/ml and 1 ng/ml doses of PAPP-A (Fig. 5, C and D).
Discussion
The results of this study implicate PAPP-A as a key molecule in osteoblast differentiation. Treatment of hMSCs with rhBMP-2 resulted in direct up-regulation of Papp-A gene expression. TGF was previously shown to up-regulate Papp-A gene expression in human osteoblasts (4). As a member of the TGF superfamily, it is reasonable that BMP-2 has a similar effect. In addition, studies by others have shown that BMP-2 can up-regulate Igf-I gene expression (35). In our laboratory, we observed that this up-regulation is indirect and a protein mediator is required for stimulation of Igf-I gene expression by BMP-2. We also demonstrated that uncommitted hMSCs when treated with PAPP-A initiate changes in gene expression including up-regulation of the transcription factor Runx2. Figure 6 depicts our proposed model of the cascade of events that could contribute to Runx2 gene expression and the relationship among BMP-2, IGF-I, and PAPP-A. Although PAPP-A did not induce Osx gene expression when added in combination with BMP-2, PAPP-A can stimulate matrix mineralization in the absence of other osteogenic supplements such as dexamethasone. This suggests alternate signaling pathways for BMP-2 and PAPP-A that are each essential for mineralization. However, the common involvement of Runx2 suggests a potential point of convergence between the pathways. The cooperation between PAPP-A and BMP-2 in matrix mineralization could involve Osx but remains unclear at this time. Our data further suggest that PAPP-A, which we demonstrated to be regulated by BMP-2, is also involved in angiogenesis because new blood vessel formation was observed in the chick CAM model after PAPP-A treatment.
PAPP-A is an IGFBP-4 protease and potentially has a role in bone formation by regulating IGF bioavailability. We and others (15) demonstrated the effects of IGF-I on osteoblast marker gene expression, specifically in the early commitment of a progenitor cell to the osteogenic lineage (41). In the present study, we demonstrate that PAPP-A can up-regulate Runx2 gene expression directly, in the absence of de novo protein synthesis. Our first thought was that PAPP-A functions in up-regulation of Runx2 via an increase in IGF-I bioavailability to the cell; however, our recent data suggest a nonproteolytic role. PAPP-A used in this study was only weakly active as protease, as was the PAPP-A induced by BMP-2. Therefore, it is unlikely that PAPP-A induction of Runx2 is dependent on IGF-I bioavailability and suggests a growth factor-like signaling mechanism for PAPP-A.
We therefore suggest that PAPP-A may elicit effects on gene expression via nonproteolytic mechanism(s). As such, the stimulation of hMSCs toward osteoblast maturation by PAPP-A does not rely solely on IGFBP-4 proteolytic cleavage from IGFs. Recently Sivanandam et al. (46) demonstrated that protein kinase C (PKC) is involved in the conversion of PAPP-A470 to lower-molecular-weight PAPP-A400, resulting in a 4-fold reduction in protease activity. Because BMP-2 is known to stimulate PKC activity (23), it is possible that PAPP-A stimulated by BMP-2 presented here may also involve PKC activity and thus induce the less active form of PAPP-A400. PKC stimulation did not affect proMBP levels; therefore, a less active PAPP-A400 is not likely due to complex formation with its inhibitor proMBP (proform of eosinophil major basic protein). The authors speculate that this conversion by PKC is unlikely caused by proteolytic cleavage of PAPP-A because reduced monomers of PAPP-A470 comigrated with PAPP-A400. A similar phenomenon could occur in our system where BMP-2 might dictate PAPP-A activity and perhaps its functions.
The combination of PAPP-A and rhBMP-2 was sufficient for matrix mineralization of hMSCs, which is indicative of end-stage osteoblastic differentiation. Significant calcium deposition was not detected in the presence of either PAPP-A or rhBMP-2 alone. However, it is interesting to note that PAPP-A plus BMP-2-induced mineral deposition in the absence of any other traditional osteogenic supplements such as dexamethasone. Recently we observed an identical outcome of hMSC matrix mineralization after adenoviral transduction with Igf-I and treatment with rhBMP-2 (data not shown). Our findings have important implication for the design of new cocktails that promote proper bone formation with reduced side effects. These data agree with our recent report as well as studies by others that support a synergistic role between IGF-I and BMP-2 (33, 41). Perhaps PAPP-A, either through IGF-I or independently, permits BMP-2 to signal optimally and thus enhance bioactivity and improve outcome (matrix mineralization). Whereas neither factor can act alone to sufficiently achieve mineralization, their combined actions result in the cells’ ability to differentiate toward the osteogenic lineage and form a mineralized extracellular matrix. The present study also agrees with respect to Osx gene expression. Osx was not induced by adenoviral delivery of Igf-I or treatment with PAPP-A. Osx is required for ossification and as stimulated by BMP-2 seems insufficient to induce mineralization in hMSCs alone. However, in the presence of either IGF-I or PAPP-A, BMP-2-treated cells exhibit a mineralized extracellular matrix. Perhaps the increase in IGF-I due to viral gene delivery or PAPP-A stimulation modifies BMP-2 signaling such that mineralization is possible. It is unclear whether BMP-2 signaling modulated by IGF-I involves Osx.
PAPP-A’s angiogenic capacity alone reinforces the theory that PAPP-A itself is a powerful signal in key developmental, repair, and regenerative processes. In the literature, BMP-2 and IGF-I have been reported to be involved in angiogenesis (5, 7, 47). It is logical that PAPP-A, as a downstream direct target of BMP-2 and potential intermediate between BMP-2 and IGF-I, could have an angiogenic role. Reports from others (9) have strongly implicated PAPP-A in a developmental and healing context. As shown in the present study, the induction of new blood vessels formation by PAPP-A in the chick CAM model makes PAPP-A an attractive candidate for therapeutic applications.
In conclusion, our data suggest that PAPP-A plays a key regulatory role in bone formation. PAPP-A could function in osteogenic lineage progression via proteolytic and/or nonproteolytic mechanisms. At this time, it is unclear what dictates one function over another or the advantages of each when considering the desirable end result of matrix mineralization, angiogenesis, and tissue development. Given that PAPP-A is regulated by many factors (4, 48), it could have multiple roles. Our future work will focus on elucidating the signaling mechanisms of PAPP-A in bone formation, particularly addressing PAPP-A as a potential mediator of BMP-2 and IGF-I.
Acknowledgments
We especially thank Laurie Bale of Conover Lab for performing the protease activity assay. The authors thank Dr. Jeffrey Hollinger (director of the Bone Tissue Engineering Center) for providing the rhBMP-2 and partial financial support for this work.
Footnotes
This work was supported by Pennsylvania Infrastructure Technology Alliance and Bone Tissue Engineering Center, Carnegie Mellon University.
Abbreviations: BMP, Bone morphogenetic protein; CAM, chick chorioallantoic membrane; CT, cycle threshold; hMSC, human adult mesenchymal stem cell; IGFBP, IGF binding protein; OSX, osterix; PAPP-A, pregnancy-associated plasma protein A; PKC, protein kinase C; qPCR, quantitative real-time PCR; rh, recombinant human; RUNX, runt-related transcription factor.
References
Lawrence JB, Oxvig C, Overgaard MT, Sottrup-Jensen L, Gleich GJ, Hayes LG, Yates 3rd JR, Covover CA 1999 The insulin-like growth factor (IGF)-dependent IGF binding protein-4 protease secreted by human fibroblasts is pregnancy-associated plasma protein-A. Proc Natl Acad Sci USA 96:3149–3153
Brambati B, Macintosh MC, Teisner B, Maguiness S, Shrimanker K, Lanzani A, Bonacchi I, Tului L, Chard T, Grudzinskas JG 1993 Low maternal serum levels of pregnancy associated plasma protein A (PAPP-A) in the first trimester in association with abnormal fetal karyotype. Br J Obstet Gynaecol 100:324–326
Chen BK, Leiferman KM, Pittelkow MR, Overgaard MT, Oxvig C, Conover CA 2003 Localization and regulation of pregnancy-associated plasma protein a expression in healing human skin. J Clin Endocrinol Metab 88:4465–4471
Ortiz CO, Chen BK, Bale LK, Overgaard MT, Oxvig C, Conover CA 2003 Transforming growth factor- regulation of the insulin-like growth factor binding protein-4 protease system in cultured human osteoblasts. J Bone Miner Res 18:1066–1072
Rabinovsky ED, Draghia-Akli R 2004 Insulin-like growth factor I plasmid therapy promotes in vivo angiogenesis. Mol Ther 9:46–55
Dunn SE 2000 Insulin-like growth factor I stimulates angiogenesis and the production of vascular endothelial growth factor. Growth Horm IGF Res 10(Suppl A):S41–S42
Shigematsu S, Yamauchi K, Nakajima K, Iijima S, Aizawa T, Hashizume K 1999 IGF-1 regulates migration and angiogenesis of human endothelial cells. Endocr J 46(Suppl):S59–S62
Bayes-Genis A, Conover CA, Overgaard MT, Bailey KR, Christiansen M, Holmes Jr DR, Virmani R, Oxvig C, Schwartz RS 2001 Pregnancy-associated plasma protein A as a marker of acute coronary syndromes. N Engl J Med 345:1022–1029
Conover CA, Bale LK, Overgaard MT, Johnstone EW, Laursen UH, Fuchtbauer EM, Oxvig C, van Deursen J 2004 Metalloproteinase pregnancy-associated plasma protein A is a critical growth regulatory factor during fetal development. Development 131:1187–1194
Qin QP, Laitinen P, Majamaa-Voltti K, Eriksson S, Kumpula EK, Pettersson K 2002 Release patterns of pregnancy associated plasma protein A (PAPP-A) in patients with acute coronary syndromes. Scand Cardiovasc J 36:358–361
Alden TD, Beres EJ, Laurent JS, Engh JA, Das S, London SD, Jane Jr JA, Hudson SB, Helm GA 2000 The use of bone morphogenetic protein gene therapy in craniofacial bone repair. J Craniofac Surg 11:24–30
Campisi P, Hamdy RC, Lauzier D, Amako M, Rauch F, Lessard ML 2003 Expression of bone morphogenetic proteins during mandibular distraction osteogenesis. Plast Reconstr Surg 111:201–208; discussion 209–210
Lee MH, Kwon TG, Park HS, Wozney JM, Ryoo HM 2003 BMP-2-induced osterix expression is mediated by Dlx5 but is independent of Runx2. Biochem Biophys Res Commun 309:689–694
Diefenderfer DL, Osyczka AM, Reilly GC, Leboy PS 2003 BMP responsiveness in human mesenchymal stem cells. Connect Tissue Res 44(Suppl 1):305–311
Walsh S, Jefferiss CM, Stewart K, Beresford JN 2003 IGF-I does not affect the proliferation or early osteogenic differentiation of human marrow stromal cells. Bone 33:80–89
Rubin J, Ackert-Bicknell CL, Zhu L, Fan X, Murphy TC, Nanes MS, Marcus R, Holloway L, Beamer WG, Rosen CJ 2002 IGF-I regulates osteoprotegerin (OPG) and receptor activator of nuclear factor-B ligand in vitro and OPG in vivo. J Clin Endocrinol Metab 87:4273–4279
Nohe A, Hassel S, Ehrlich M, Neubauer F, Sebald W, Henis YI, Knaus P 2002 The mode of bone morphogenetic protein (BMP) receptor oligomerization determines different BMP-2 signaling pathways. J Biol Chem 277:5330–5338
Fiedler J, Roderer G, Gunther KP, Brenner RE 2002 BMP-2, BMP-4, and PDGF-bb stimulate chemotactic migration of primary human mesenchymal progenitor cells. J Cell Biochem 87:305–312
Zhang M, Xuan S, Bouxsein ML, von Stechow D, Akeno N, Faugere MC, Malluche H, Zhao G, Rosen CJ, Efstratiadis A, Clemons TL 2002 Osteoblast-specific knockout of the insulin-like growth factor (IGF) receptor gene reveals an essential role of IGF signaling in bone matrix mineralization. J Biol Chem 277:44005–44012
Deckers MM, van Bezooijen RL, van der Horst G, Hoogendam J, van Der Bent C, Papapoulos SE, Lowik CW 2002 Bone morphogenetic proteins stimulate angiogenesis through osteoblast-derived vascular endothelial growth factor A. Endocrinology 143:1545–1553
Tanaka H, Wakisaka A, Ogasa H, Kawai S, Liang CT 2002 Effect of IGF-I and PDGF administered in vivo on the expression of osteoblast-related genes in old rats. J Endocrinol 174:63–70
Boden SD, Kang J, Sandhu H, Heller JG 2002 Use of recombinant human bone morphogenetic protein-2 to achieve posterolateral lumbar spine fusion in humans: a prospective, randomized clinical pilot trial: 2002 Volvo Award in Clinical Studies. Spine 27:2662–2673
Hay E, Lemonnier J, Fromigue O, Marie PJ 2001 Bone morphogenetic protein-2 promotes osteoblast apoptosis through a Smad-independent, protein kinase C-dependent signaling pathway. J Biol Chem 276:29028–29036
Lou J, Tu Y, Li S, Manske PR 2000 Involvement of ERK in BMP-2 induced osteoblastic differentiation of mesenchymal progenitor cell line C3H10T1/2. Biochem Biophys Res Commun 268:757–762
Boden SD, Zdeblick TA, Sandhu HS, Heim SE 2000 The use of rhBMP-2 in interbody fusion cages. Definitive evidence of osteoinduction in humans: a preliminary report. Spine 25:376–381
Thomas T, Gori F, Spelsberg TC, Khosla S, Riggs BL, Conover CA 1999 Response of bipotential human marrow stromal cells to insulin-like growth factors: effect on binding protein production, proliferation, and commitment to osteoblasts and adipocytes. Endocrinology 140:5036–5044
Higuchi T, Kinoshita A, Takahashi K, Oda S, Ishikawa I 1999 Bone regeneration by recombinant human bone morphogenetic protein-2 in rat mandibular defects. An experimental model of defect filling. J Periodontol 70:1026–1031
Boden SD, Martin Jr GJ, Horton WC, Truss TL, Sandhu HS 1998 Laparoscopic anterior spinal arthrodesis with rhBMP-2 in a titanium interbody threaded cage. J Spinal Disord 11:95–101
Han VK, Bassett N, Walton J, Challis JR 1996 The expression of insulin-like growth factor (IGF) and IGF-binding protein (IGFBP) genes in the human placenta and membranes: evidence for IGF-IGFBP interactions at the feto-maternal interface. J Clin Endocrinol Metab 81:2680–2693
Mohan S, Baylink DJ 1996 Insulin-like growth factor system components and the coupling of bone formation to resorption. Horm Res 45(Suppl 1):59–62
Muschler GF, Hyodo A, Manning T, Kambic H, Easley K 1994 Evaluation of human bone morphogenetic protein 2 in a canine spinal fusion model. Clin Orthop 229–240
Ahrens M, Ankenbauer T, Schroder D, Hollnagel A, Mayer H, Gross G 1993 Expression of human bone morphogenetic proteins-2 or -4 in murine mesenchymal progenitor C3H10T1/2 cells induces differentiation into distinct mesenchymal cell lineages. DNA Cell Biol 12:871–880
Raiche AT, Puleo DA 2004 In vitro effects of combined and sequential delivery of two bone growth factors. Biomaterials 25:677–685
Raiche AT, Puleo DA 2004 Cell responses to BMP-2 and IGF-I released with different time-dependent profiles. J Biomed Mater Res 69A:342–350
Canalis E, Gabbitas B 1994 Bone morphogenetic protein 2 increases insulin-like growth factor I and II transcripts and polypeptide levels in bone cell cultures. J Bone Miner Res 9:1999–2005
Nakashima K, Zhou X, Kunkel G, Zhang Z, Deng JM, Behringer RR, de Crombrugghe B 2002 The novel zinc finger-containing transcription factor osterix is required for osteoblast differentiation and bone formation. Cell 108:17–29
Ducy P, Starbuck M, Priemel M, Shen J, Pinero G, Geoffroy V, Amling M, Karsenty G 1999 A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev 13:1025–1036
Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G 1997 Osf2/Cbfa1: a transcriptional activator of osteoblast differentiation. Cell 89:747–754
Otto F, Thornell AP, Crompton T, Denzel A, Gilmour KC, Rosewell IR, Stamp GW, Beddington RS, Mundlos S, Olsen BR, Selby PB, Owen MJ 1997 Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 89:765–771
Applied Biosystems 2001 User bulletin #2: ABI Prism 7700 Sequence Detection System. Foster City, CA: Applied Biosystems
Celil AB, Hollinger JO, Campbell PG 2005 Osx transcriptional regulation is mediated by additional pathways to BMP2/Smad signaling. J Cell Biochem 95:518–528
Jadlowiec J, Koch H, Zhang X, Campbell PG, Seyedain M, Sfeir C 2004 Phosphophoryn regulates the gene expression and differentiation of NIH3T3, MC3T3–E1, and human mesenchymal stem cells via the integrin/MAPK signaling pathway. J Biol Chem 279:53323–53330
Ribatti D, Presta M 2002 The role of fibroblast growth factor-2 in the vascularization of the chick embryo chorioallantoic membrane. J Cell Mol Med 6:439–446
Ribatti D, Nico B, Vacca A, Roncali L, Burri PH, Djonov V 2001 Chorioallantoic membrane capillary bed: a useful target for studying angiogenesis and anti-angiogenesis in vivo. Anat Rec 264:317–324
Ribatti D, Vacca A, Roncali L, Dammacco F 1996 The chick embryo chorioallantoic membrane as a model for in vivo research on angiogenesis. Int J Dev Biol 40:1189–1197
Sivanandam AS, Mohan S, Kita H, Kapur S, Chen ST, Linkhart TA, Bagi G, Baylink DJ, Qin X 2004 Studies on regulation of IGF binding protein (IGFBP)-4 proteolysis by pregnancy-associated plasma protein-A (PAPP-A) in cells treated with phorbol ester. Biochem J 379:57–64
Marrony S, Bassilana F, Seuwen K, Keller H 2003 Bone morphogenetic protein 2 induces placental growth factor in mesenchymal stem cells. Bone 33:426–433
Conover CA, Chen BK, Resch ZT 2004 Regulation of pregnancy-associated plasma protein-A expression in cultured human osteoblasts. Bone 34:297–302(Julie Jadlowiec, Diana Do)