Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Control Release. Author manuscript; available in PMC 2010 December 3.
Published in final edited form as:
PMCID: PMC2783665

Release Characteristics and Osteogenic Activity of Recombinant Human Bone Morphogenetic Protein-2 grafted to Novel Self-Assembled Poly(lactide-co-glycolide fumarate) Nanoparticles


Functionalized biodegradable nanoparticles (NPs) provide reactive groups and large surface area for grafting Recombinant human bone morphogenetic protein-2 (rhBMP-2) to reduce protein diffusion and maintain sufficient concentration for recruitment and differentiation of osteoprogenitor cells. The objective of this work was to investigate release characteristics and osteogenic activity of rhBMP-2, grafted to biodegradable NPs based on succinimide-terminated poly(lactide fumarate) (PLAF-NHS) and poly(lactide-co-glycolide fumarate) (PLGF-NHS) macromers. The release of rhBMP-2 from the NPs, measured by enzyme-linked immunosorbent assay, was linear with time in the first two weeks, and 24.70±1.30% and 48.7±0.7% of the protein grafted to PLGF-NHS and PLAF-NHS NPs, respectively, was released in the enzymatically active conformation after complete degradation/erosion of the NPs. After 14 days of incubation with bone marrow stromal (BMS) cells, rhBMP-2 grafted to PLAF-NHS and PLGF-NHS NPs was as effective in inducing mineralization as the native rhBMP-2 that was directly added to the cell culture media. At any incubation time, rhBMP-2 grafted to PLAF had the highest expression of osteopontin (OP) and osteocalcin (OC), followed by rhBMP-2 grafted to PLGF and rhBMP-2 directly added to media. Higher OP and OC expression for BMP-gPLAF and BMP-gPLGF groups may be related to other factors in the cascade of osteogenesis, such as differentiation of BMS cells to the vasculogenic lineage and formation of a vascularized/mineralized marix.

1. Introduction

There are approximately 0.5 million fractures in the US annually that require bone graft procedures to ensure rapid skeletal repair [1]. Bone morphogenetic proteins (BMPs) play a major role in initiating the cascade of chemotaxis, differentiation of marrow stromal cells, and bone regeneration [2]. In particular, recombinant human bone morphogenetic protein-2 (rhBMP-2) is a potent differentiation factor that is capable of inducing bone formation following implantation and is used clinically for spinal fusion [3]. BMP signaling is highly regulated in-vivo [4]. Therefore, 4–5 orders of magnitude higher than the amount found endogenously (1 mg/ml for rhBMP-2) have to be loaded in the graft to induce bone formation [5]. High doses, coupled with diffusion of rhBMP-2 away from the intended site of regeneration, cause adverse effects such as bone overgrowth and immunological reactions [6]. Furthermore, the rate at which rhBMP-2 is released from the carrier can affect the efficacy of bone induction [7]. The optimum release profile for rhBMP-2 is not known, but sustained release of rhBMP-2 in-vivo over 4 weeks induced bone formation to a higher extent compared to the same amount released in 3 days [7].

A composite poly(dl-lactic-co-glycolic acid)/calcium phosphate (CaP) cement has been used as a matrix for sustained release of rhBMP-2, but due to interaction of the protein with CaP, a large fraction of rhBMP-2 (50–75%) was not released from the matrix after 28 days [8]. In another study, titanium discs coated with rhBMP-2 incorporated poly(D,L-lactide) (PDLLA) did not induce ectopic bone formation when implanted in sheep muscle, possibly due to the very slow release or deactivation during the coating process [9]. Hydrogel microparticles (MPs) based on dextran, functionalized with carboxylate, sulfate and benzylamide groups, were used as a carrier for rhBMP-2 but >60% of the protein was released from the MPs in the first 24 h followed by a very slow (insufficient) release rate for 4 weeks [10]. Gelatin MPs crosslinked with glutaraldehyde were also used as a carrier for rhBMP-2, but <25% of the encapsulated protein was released after 28 days in collagenase-free media [11]. Glycidyl methacrylated dextran/gelatin MPs can slowly release rhBMP-2 over 4 weeks, but these MPs are limited by their low loading capacity (rhBMP-2 is loaded by swelling of the MPs) [12]. When rhBMP-2 was encapsulated in PLGA NPs, burst release was >50% for low molecular weight (MW) PLGA while <20% of the protein was released after 8 weeks with high MW PLGA [13]. Encapsulation in micro/nano particles and fibers has been used to reduce diffusion of rhBMP-2 away from the application site and to reduce its in-vivo enzymatic degradation [1216]. Although encapsulated rhBMP-2 is shown to enhance mineralization and bone formation [17], a large fraction of the protein is deactivated in the process of emulsification in organic solvents and solidification with isopropanol, and the release profile is not optimal [18, 19]. Therefore, relatively high doses have to be loaded in micro/nano particles which affect the safety profile of rhBMP-2 in clinical applications [6, 20]. Functionalized nanoparticles (NPs) provide large surface area and reactive groups for grafting proteins to the NPs. Grafting reactions can be carried out in aqueous media, thus reducing protein deactivation due exposure to organic solvents.

Our laboratory has developed novel poly(lactide-co-glycolide fumarate) (PLGF), and poly(lactide-co-ethylene oxide fumarate) (PLEOF) unsaturated macromers that self-assemble to form biodegradable NPs [21]. In the process of NPs formation, biodegradable PLEOF macromer acts as a surfactant to stabilize the NPs. NPs ranging 50–500 nm in size can be produced by varying the ratio of PLEOF to PLGF in the blend. The macromers can be functionalized with succinimide groups for grafting proteins to the NPs in aqueous media [22]. The objective of this work was to investigate the release characteristics and osteogenic activity of rhBMP-2, grafted to self-assembled biodegradable NPs. In this work, PLGF macromer is reacted with N,N′-disuccinimidyl carbonate (DSC) to produce succinimide-terminated PLGF-NHS macromer. Next, PLGF-NHS and PLEOF blend macromers are self-assembled by dialysis to produce NPs. Then, rhBMP-2 is grafted to the NPs in aqueous solution by the reaction between amine groups of the protein with PLGF-NHS succinimide groups on the NPs. Release characteristics of rhBMP-2 from the NPs was measured by Enzyme-Linked Immunosorbent Assay (ELISA). Osteogenic activity of the grafted rhBMP-2 was determined by incubation with bone marrow stromal cells.

2. Methods and materials

2.1. Materials

L-lactide (LA; >99.5% purity) and glycolide (GL; >99.0% purity) monomers were obtained from Ortec (Easley, SC) and Boehringer Chemicals (Ingelheim, Germany), respectively. LA and GL monomers were dried under vacuum at 40°C for at least 12 h before the reaction. Poly(ethylene glycol) (PEG, nominal molecular weight 3.4 kDa), triethylamine (TEA), tin (II) 2-ethylhexanoate (TOC), and dimethylsulfoxide (DMSO) were purchased from Aldrich (Sigma-Aldrich, St. Louis, MO). PEG was dried by azeotropic distillation from toluene. Fumaryl chloride (FC) was obtained from Aldrich and distilled at 161°C prior to use. Diethylene glycol initiator (DEG; >99% purity) was purchased from Fisher (Pittsburgh, PA). N,N′-disuccinimidyl carbonate (DSC) was obtained from Novabiochem (EMD Biosciences, San Diego, CA). Dichloromethane (DCM), dimethylformamide (DMF), tetrahydrofuran (THF), diethyl ether, and hexane were purchased from Acros (Fairfield, OH). DCM was dried by distillation over calcium hydride. Spectro/Por dialysis tube (molecular weight cut-off 3.5 kDa) was purchased from Spectrum Laboratories (Rancho Dominquez, CA). All solvents (except for DCM) were used as received. Ethylenediaminetetraacetic acid disodium salt (EDTA), penicillin, streptomycin, Alizarin red, and paraformaldehyde were purchased from Sigma (St. Louis, MO). Dulbecco’s phosphate-buffered saline (PBS) and Dulbecco’s Modified Eagle’s Medium (DMEM; 4.5 g/L glucose with L-glutamine and without sodium pyruvate) were obtained from GIBCO BRL (Grand Island, NY). Trypsin and fetal bovine serum (FBS, screened for compatibility with rat bone marrow stromal cells) were obtained from Invitrogen (Carlsbad, CA) and Atlas Biologicals (Fort Collins, CO), respectively. Quant-it PicoGreen dsDNA reagent kit was obtained from Molecular Probes (Eugene, OR). QuantiChrom calcium assay and QuantiChrom alkaline phosphatase assay kits were purchased from Bioassay Systems (Hayward, CA). rhBMP-2 solution (100 μL, 1.5 mg/mL in rhBMP-2 buffer) was generously donated by Medtronic (Minneapolis, MN). Concentration of rhBMP-2 was measured by ELISA using the BMP Quantikine kit (R&D Systems, Minneapolis, MN).

2.2. PLGF and PLEOF Macromer synthesis and succinimide termination

Ultra-low molecular weight poly(lactide) (ULMW-PLA) and poly(lactide-co-glycolide) (ULMW-PLGA) were synthesized by ring-opening polymerization of LA (100%) and LA+GL (50% LA and 50% GL) monomers, respectively, as previously described [23]. Mn¯,Mw¯, and PDI of ULMW-PLA macromer was 1450 Da, 1730 Da, and 1.2, respectively, and those of ULMW-PLGA was 1660 Da, 2150 Da, and 1.3. Next, PLAF or PLGF was synthesized by condensation polymerization of ULMW-PLA or PLGA with FC, as previously described [21, 24]. Similarly, PLEOF was synthesized by reacting ULMW-PLA and PEG with FC [2527]. The molar ratios of FC:(PEG+PLA) and TEA:(PEG+PLA) were 0.9:1.0 and 1.8:1.0, respectively. Succinimide-terminated PLAF-NHS/PLGF-NHS macromers were produced by reacting the hydroxyl end-groups of PLAF/PLGF macromers with DSC, as previously described [28]. Briefly, 800 mg PLAF or PLGF and 26 mg DSC were mixed in 15 ml DMF in a reaction flask. After purging with nitrogen, 40 μL TEA was added while stirring, and the reaction was allowed to continue for 8 h at ambient conditions. The synthesized macromers were characterized by 1H-NMR and GPC [23, 27].

2.3. Characterization of the macromers

The chemical structure of the synthesized macromers was characterized by a Varian Mercury-300 1H-NMR (Palo Alto, CA). The macromer was dissolved in deuterated chloroform (Aldrich, 99.8% deuterated; 50 mg/mL) and 1% v/v trimethylsilane (TMS; Aldrich) was used as the internal standard. The molecular weight distribution of the macromers was measured by GPC [23]. Measurements were carried out with a Waters 717 Plus Autosampler GPC system (Waters, Milford, MA). The columns consisted of a styragel HT guard column (7.8 mm × 300 mm; Waters) in series with a styragel HR 4E column (7.8 mm × 300 mm, Waters) heated to 37°C. The Empower software (Waters) was used for determination of Mn¯,Mw¯, and PDI. The sample (20 μl; 10 mg/mL in THF) was eluted at 1 mL/min. Monodisperse polystyrene standards (PDI<1.1; Waters) were used to construct the calibration curve.

2.4. Nanoparticle formation and rhBMP-2 grafting

Mixtures of PLAF–NHS (or PLGF–NHS) and PLEOF macromers in DMF/DMSO solvent mixture were self-assembled into NPs by dialysis against PBS buffer, as shown in Figure 1. Briefly, 45 mg PLAF–NHS (or PLGF–NHS) and 5 mg 90/10 PLEOF, dissolved in a mixture of 1 mL DMF and 8 mL DMSO, were loaded in the dialysis tube (molecular cutoff: 3.5 kDa) and dialyzed against sterile PBS. The macromer solution was dialyzed for 24 h with change of dialysis buffer every 2–4 h until DMSO and DMF were removed. Next, the NPs suspension was collected from the dialysis tube and freeze-dried to obtain a free flowing powder. For grafting, 2 ml of the suspension (10 mg NPs) was centrifuged at 18,350 rcf (15,000 rpm) for 10 min, the supernatant was decanted, and the precipitated NPs were re-suspended in 0.5 ml PBS by sonication for 1 min with a 3-mm probe Ultrasonic Processor (Cole-Parmer, Vernon Hills, IL). Next, 0.5 ml rhBMP-2 in PBS solution (400 ng/mL) was added to the suspension, and the protein was allowed to react with succinimide terminated NPs under ambient conditions for 12 h (see Figure 1). After reaction, the suspension was dialyzed against PBS to remove the by-product, N-hydroxy succinimide.

Figure 1
Schematic diagram for self-assembly of PLGF-NHS and PLEOF blend to produce NPs.

2.5. Characterization of NPs

Morphology and size distribution of the NPs was examined using a JSM-5400 scanning electron microscope (JOEL, Japan) at an accelerating voltage of 20 KeV. Freeze-dried NPs were placed on a graphite surface and coated with gold using an Ion Sputter Coater (JOEL, JFC-1100) at 20 mA for 1 min. Size distribution of the NPs was measured by dynamic light scattering with a NICOMP Submicron Particle Sizer (Model 370, NICOMP, Santa Barbara, CA). 500 μl of the diluted suspension was added to a culture tube and placed in the instrument cell holder. The scattered light intensity was inverted to size distribution by inverse Laplace transform using the NICOMP CW370 software.

2.6. Degradation kinetics of NPs

The NPs suspension was centrifuged at 18,350 rcf for 10 min, the supernatant was decanted to remove the unassembled macromers, and the NPs were re-suspended in PBS. For degradation experiments, 50 mg NPs were suspended in 1 ml PBS and the suspensions were incubated at 37°C until complete degradation (no mass remaining or no NPs detected by light scattering). At each time point, size distribution of the NPs was measured by dynamic light scattering. Next, samples were freeze-dried and mass of the dried powder was measured. The fraction of mass remaining was determined by dividing the dried mass at time t by the initial mass.

2.7. Grafting efficiency and release kinetics of rhBMP-2-loaded NPs

Since molecular weight cutoff of the dialysis membrane was lower than rhBMP-2 molecular weight (12 kDa for rhBMP-2 versus 3500 Da membrane cutoff molecular weight; [29]), ungrafted rhBMP-2 was not removed by dialysis. To determine grafting efficiency, the dialyzed suspension was centrifuged at 18,350 rcf for 10 min, and the enzymatically active concentration of rhBMP-2 in the supernatant was measured by ELISA using the BMP Quantikine kit according to manufacturer’s instructions. After the addition of chromogen and assay stop solution, the absorbance was measured with a plate reader (Synergy HT, Bio-Tek) at two wavelengths of 450 and 570 nm. The values at 570 were subtracted from those at 450 nm to correct for background absorbance. The corrected absorbance values were related to concentration using a calibration curve constructed from the absorbance of solutions with known rhBMP-2 concentration. Grafting efficiency was determined by dividing the measured amount of rhBMP-2 by the initial amount in the grafting reaction. For determination of release characteristics, the precipitate was resuspended and incubated in 1 ml PBS at 37°C with orbital shaking. At each time point, suspension was centrifuged at 18350 rcf for 10 min, supernatant was removed and stored in microvials for analysis. The fraction of released rhBMP-2 was determined by dividing the measured amount of rhBMP-2 at each time point to the total amount at time zero (after grafting).

2.8. Bone marrow stromal cell isolation

BMS cells were isolated from the bone marrow of young adult male Wistar rats according to established protocols [27, 30]. Cell isolations were performed under a protocol approved by the Institutional Animal Care and Use Committee of the University of South Carolina. After flushing the bone marrow, cell suspensions were centrifuged at 200×g for 5 min, the cell pellets were re-suspended in primary media (DMEM supplemented with 10% FBS, 100 units/mL penicillin (PEN), 100 μg/mL streptomycin (SP), 50 μg/mL gentamicin sulfate (GS), and 250 ng/mL fungizone (FZ)) and maintained in a humidified 5% CO2 incubator at 37°C. Cultures were replaced with fresh media at 3 and 7 days to remove haematopoietic and other unattached cells. After 10 days, cells were detached from the flasks with 0.05% trypsin-0.53 mM EDTA and used for in-vitro osteogenic experiments.

2.9. In vitro osteogenic activity of BMS cells incubated with rhBMP-2 grafted NPs

The procedure for determination of osteogenic activity of rhBMP-2 grafted to NPs is demonstrated schematically in Figure 2. BMS cells (Figure 2b) isolated from rats (Figure 2a), were seeded in 24-well plates at a density of 5×104 cells/mL in primary media. After 24 h for cell attachment (time zero), media was replaced with standard osteogenic media (primary media supplemented with 100 nM dexamethasone (DEX), 50μg/mL ascorbic acid (AA), 10mM β-glycerophosphate (βGP)) with the addition of time-released rhBMP-2 from a suspension of 200 ng/mL rhBMP-2 grafted NPs (Figure 2c). rhBMP-2 (200 ng) was grafted to the NPs (10 mg) in 1 mL PBS as described in section 2.6. Suspension was centrifuged at 18,350 rcf, supernatant was removed, and the precipitate was resuspended in osteogenic media (Figure 2e) and incubated for 24 h (Figure 2f). After incubation, suspension was centrifuged (Figure 2g), and the supernatant was added to the seeded BMS cells (time zero; Figure 2c). The precipitate (figure 2h) was resuspended in osteogenic media and incubated until the next time point. The time-released rhBMP-2 from the NPs was added to BMS cell cultures at time points 4, 7, 11, 14, and 18 days (the time points for refreshment of the media). NPs were not directly added to cell cultures because refreshment of media every 3–4 days interfered with the release profile of rhBMP-2 from the NPs which would have introduced uncertainty in the total amount of rhBMP-2 released. 200 ng/mL rhBMP-2 directly added to BMS cell cultures at time zero in osteogenic media was used as the positive control, without refreshment at each media change [31]. This group simulated dipping a scaffold in rhBMP-2 solution, followed by implantation in a defect, where rhBMP-2 can not be refreshed after implantation. The same amount of rhBMP-2 grafted to PLGF NPs incubated in osteogenic media was used to simulate dipping a scaffold in rhBMP-2 grafted NPs suspension, assuming that the NPs are retained at the defect by NPs slow migration and larger size (compared to the protein). BMS cells cultured in primary (Control) and osteogenic (OM) media were used as control groups. At each time point (4, 7, 14, and 21 days; Figure 2d)), cultures were washed with PBS, cells were lysed with 0.4 mL lysis buffer (10mM tris, 2% triton) for 1 h. After centrifugation, supernatant was used for determination of DNA content, ALPase activity, and calcium content.

Figure 2
Schematic diagram to demonstrate the procedure for determination of osteogenic activity of rhBMP-2 grafted to NPs.

2.10. Measurement of DNA content, ALPase activity, and calcium content

The double stranded DNA (dsDNA) content of the samples was measured using a Quant-it PicoGreen assay according to manufacturer’s instructions. The fluorescence of the solution was measured with a Synergy HT plate reader at emission and excitation wavelength of 485 and 528 nm, respectively. Measured intensities were correlated to cell numbers using a calibration curve constructed with BMS cells of known concentration ranging from zero to 4×104 cells/mL. ALPase activity was assessed using QuantiChrom ALPase assay kit according to manufacture’s instructions. A 10 μL aliquot of the cell lysate was added to 190 μL of reagent solution containing 10 mM p-nitrophenyl phosphate and 5 mM magnesium acetate and absorbance was recorded at time zero and again after 4 min. ALPase activity was calculated using the equation [(At=4−At=0)/(Acalibrator −AddH2O)×808] expressed as IU/L. The absorbance was measured on a plate reader at 405 nm. The measured ALPase activity was normalized to DNA content. Calcium content was measured using QuantiChrom calcium assay kit according to manufacturer’s instructions. The absorbance was measured on a plate reader at 612 nm. Measured intensities were correlated to equivalent amounts of Ca2+ using a calibration curve constructed with calcium chloride solutions of known concentration ranging from zero to 200 μg/mL.

2.11. mRNA analysis

Total cellular RNA was isolated from the cells using TRIzol (Invitrogen, Carlsbad, CA) plus RNeasy Mini-Kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions. The qualitative and quantitative analysis of the RNA samples was performed with NanoDrop 2000 (Thermo Scientific, Waltham, MA). 1μg of the extracted total RNA was subjected to cDNA conversion using the Reverse Transcription System (Promega, Madison, WI). Primers for real-time PCR analysis were designed and selected using the Primer3 web-based software as described [24, 32]. The annealing temperatures were optimized by classical PCR and agarose gel electrophoresis [24, 32]. Real-time PCR (RT-qPCR) was performed to analyze the differential expression of osteopontin, osteocalcin, and osteonectin genes with SYBR green RealMasterMix (Eppendorf, Hamburg, Germany) using Bio-Rad iCycler machine (Bio-Rad, Hercules, CA). The following forward and reverse primers were employed: Osteonectin: forward 5′-ACA AGC TCC ACC TGG ACT ACA and reverse 5′-TCT TCT TCA CAC GCA GTT T; Osteopontin: forward 5′-GAC GGC CGA GGT GAT AGC TT and reverse 5′-CAT GGC TGG TCT TCC CGT TGC; Osteocalcin: forward 5′-AAA GCC CAG CGA CTC T and reverse 5′-CTA AAC GGT GGT GCC ATA GAT; and ARBP: forward 5′-CGA CCT GGA AGT CCA ACT AC and reverse 5′-ATC TGC TGC ATC TGC TTG [24, 32]. Quantification of gene expression was based on the crossing-point threshold (CT) value for each sample [33]. The expression of ARBP (acidic ribosomal phosphoproteins; house-keeping) gene was used as the reference and the fold difference in gene expression was normalized to that at time zero.

2.12. Statistical analysis

Data are expressed as means ± standard deviation. All experiments were done in triplicate. Significant differences between groups were evaluated using a two-tailed student t-test. A value of p<0.05 was considered statistically significant.

3. Results

3.1. Macromer characterization

The assignment of chemical shifts in the NMR spectrum of PLAF was described previously [23, 24]. Incorporation of succinimide group to PLAF chain ends was confirmed by the chemical shift at 2.9 ppm in the NMR spectrum and presence of absorption bands at 2900 and 1780 cm−1 due to methyl and carbonyl vibrations in the FTIR spectra. The average number of succinimide end-groups per chain, obtained from the NMR chemical shifts, was 1.4 for PLAF–NHS and PLGF–NHS macromers. Mn¯,Mw¯, and PDI of the synthesized PLAF were 4.5 kDa, 8.6 kDa, and 1.9, respectively; those of PLGF were 5.2 kDa, 10.9 kDa, and 2.1; and those of PLEOF were 9.7 kDa, 15.4 kDa, and 1.6. Attachment of succinimide end-groups to PLAF macromer increased Mn¯ from 4.5 to 4.7 kDa and slightly increased Mw¯ from 8.61 to 8.63 kDa. PDI decreased slightly from 1.91 to 1.83. Similar results were obtained for PLGF-NHS.

3.2. Particle size distribution and degradation

The change in size distribution of PLAF-NHS and PLGF-NHS NPs with incubation time is shown in Figures 3(a) and 3(b), respectively. NPs had spherical geometry (see insets in Figures 3(a) and (b)). Average size and breadth of the distribution increased with incubation time for PLAF-NHS and PLGF-NHS NPs. The increase in NPs size with incubation could be explained by the swelling of amphiphilic PLEOF macromer of the NPs, as shown in Figure 1.

Figure 3
Change in size distribution of PLAF-NHS (a) and PLGF-NHS (b) NPs with incubation in primary culture media. The inserts are SEM images of the NPs at time zero before incubation (scale bars are 200 nm).

Average size of the PLAF-NHS and PLGF-NHS NPs with incubation is compared with those of PLAF and PLGF NPs (without attachment of succinimide end-group) in Figure 4(a). In general, size of PLGF or PLGF-NHS NPs was significantly less than that of PLAF or PLAF-NHS NPs for all degradation times. Size of PLAF and PLGF NPs was relatively constant with incubation while those of PLAF-NHS and PLGF-NHS increased in the first two weeks followed by a decrease after 3 weeks. These results indicate that NPs self-assembled with succinimide-terminated macromers had higher hydrophilicity compared to those of PLAF and PLGF which resulted in significant swelling of the NPs in the first two weeks.

Figure 4
Change in average diameter (c) and mass loss (d) of the NPs with incubation time in primary culture media before (PLAF and PLGF) and after (PLAF-NHS and PLGF-NHS) succinimide-termination.

Mass loss of PLAF-NHS and PLGF-NHS NPs with incubation time is compared with those of PLAF and PLGF NPs in Figure 4(b). Mass loss was nearly linear in the first four weeks of incubation in which >80% of the mass loss occurred. PLAF and PLGF NPs degraded completely in 7 and 5 weeks, respectively, and 80% and 90% of their mass loss occurred in the first 4 weeks. Succinimide modification of PLGF macromer did not significantly change the mass loss profile of PLGF-NHS NPs, but complete degradation time of PLAF-NHS NPs decreased from 7 to 5 weeks. Succinimide termination slightly increased hydrophilicity of the macromers which reduced degradation time of the more hydrophobic PLAF NPs but had little effect on degradation of PLGF NPs.

3.3. Release characteristics of rhBMP-2

Average size of PLAF-NHS NPs increased from 242±67 to 251±78 nm after grafting with rhBMP-2 while that of PLGF-NHS NPs increased from 195±42 to 199±72 nm. This was consistent with size of the protein in the native conformation [34]. Grafting efficiency was relatively high at 97.0±0.5% for PLAF-NHS and 97.5±0.4% for PLGF-NHS NPs, which was consistent with high grafting efficiencies reported with heparin-modified particles [12]. The release characteristics of rhBMP-2 from PLAF-NHS and PLGF-NHS NPs are shown in Figure 5. Protein stability with time was tested by incubating 100 μg/mL rhBMP-2 in PBS with 5 wt% PLEOF. The relative enzymatic activity after incubation for 1, 15, and 30 days was 100±8, 102±8, and 103±8%, respectively, which demonstrated that incubation time did not affect protein stability. Release of active rhBMP-2 from PLGF-NHS NPs was linear in the first 5 days followed by a slower release rate period from day 5 to 15. PLAF-NHS NPs displayed linear release of rhBMP-2 in the first 15 days followed by a slower release rate period from 15 to 28 days. Nearly 25% (24.70±1.30%) and 50% (48.7±0.7%) of the grafted rhBMP-2 was released in enzymatically active conformation after complete degradation of PLGF-NHS and PLAF-NHS NPs, respectively.

Figure 5
Release characteristics of rhBMP-2 from PLAF-NHS and PLGF-NHS NPs with incubation time in primary culture media, measured by Enzyme-Linked Immunosorbent Assay.

3.4. Osteogenic activity of released rhBMP-2

Experimental groups included BMS cells cultured in osteogenic media supplemented with time-released rhBMP-2 from grafted PLAF-NHS (BMP-gPLAF) and PLGF-NHS (BMP-gPLGF) NPs; rhBMP-2 directly added to BMS cell cultures in osteogenic media (BMP); without rhBMP-2 in osteogenic media (OM); and without rhBMP-2 in primary media (Control). The DNA content of BMS cells with incubation time for the five groups is shown in Figure 6(a). The DNA content of Control group (BMS cells cultured in primary media in the absence of osteogenic factors) showed a significant increase from day 4 to 7 as the cells proliferated, followed by a decrease from day 7 to 14, with no further change after 21 days. Relatively high seeding density (5×104 cells/cm2) was used because cell-cell contact plays a critical role in differentiation of adherent BMS cells [35]. Consequently, the DNA content of Control group decreased from day 7 to 14 as the seeded cells became highly confluent, ceased to proliferate, and eventually a fraction of the cells underwent apoptosis. At day 4, there was no statistical difference between the DNA content of the four groups; however, the DNA content of control and OM groups was significantly higher than BMP, BMP-gPLAF, and BMP-gPLGF groups after 7 and 14 days. At day 14, DNA content of BMP group was statistically lower (173±12 ng/mL) than those of BMP-gPLAF (353±32 ng/mL) and BMP-gPLGF (363±46 ng/mL) groups. At day 21, DNA content of BMP, BMP-gPLAF, and BMP-gPLGF groups was similar to that of OM but significantly lower (indicated by a star) than that of Control group. This is consistent with previous results showing that the addition of osteogenic factors and rhBMP-2 to culture media reduces proliferation of progenitor BMS cells and enhances their differentiation to the osteogenic lineage. For example, Marolt and collaborators observed that DNA content of silk scaffolds, initially seeded with equal cell numbers, was 0.0015±0.0003% of the scaffold weight after incubation for 36 days in osteogenic media versus 0.008±0.002% in primary media [36].

Figure 6
DNA content (a), ALPase activity (b), calcium content (c), mRNA expression levels (as fold difference) of osteopontin (d; OP), osteocalcin (e, OC), and osteonectin (f, ON) genes of BMS cells as a function of incubation time cultured in osteogenic media ...

ALPase activity of BMS cells with incubation time for the five groups is shown in Figure 6(b). ALPase activity of OM group statistically increased in days 7 and 14 while that of Control did not increase significantly for any of the time points. ALPase activity of BMP-gPLAF and BMP-gPLGF groups statistically increased (indicated by one star with respect to Control) from day 4 to 7 and then returned to baseline level after 14 and 21 days, while that of BMP group increased statistically from day 4 to 7, remained at the elevated level at day 14, and then returned to baseline level after 21 days. At day 7, ALPase activity of BMP-gPLAF and BMP groups was significantly higher than those of OM and control groups while ALPase activity of BMP-gPLGF group was significantly higher than Control but not OM. Overall, rhBMP-2 grafted to PLAF NPs showed higher ALPase activity compared to PLGF NPs.

Calcium content of BMS cells with incubation time for the five groups is shown in Figure 6(c). At day 4, there was statistically significant increase in calcium content of BMP-gPLAF (3.1±0.5 mg/dl), BMP-gPLGF (2.4±0.3 mg/dl), and BMP (8.2±0.5 mg/dl) groups with respect to Control (1.3±0.3 mg/dl) and OM (0.7±0.3 mg/dl) groups (indicated by one and two stars, respectively), and calcium content of BMP group was significantly higher than those of BMP-gPLAF and BMP-gPLGF (indicated by three stars). Calcium content of BMP-gPLAF, BMP-gPLGF, and BMP groups showed a statistically significant 6.9-, 2.5-, and 19.5-fold increase, respectively, from day 4 to 7 and 6.6-, 7.8-, and 2.7-fold increase from day 7 to 14. The group with BMP directly added to the culture media showed early increase in calcium content in day 7 (19.5-fold increase from day 4 to 7), but the three BMP groups ultimately showed similar levels of mineralization after 14 days. After 21 days, the three BMP groups and OM group had similar levels of calcium content. It is interesting to note that the OM group did not show increase in calcium content at days 7 and 14 but its calcium content increased to the level of BMP groups after 21 days. The calcium contents demonstrate that rhBMP-2 directly added or grafted to NPs, accelerates (as opposed to increase) mineralization, which is consistent with in-vivo results that rhBMP-2 soaked collagen sponge implanted in lumbar spine produces signs of fusion as early as eight weeks [37].

The expression level of osteogenic markers OP, OC, and ON with incubation time is shown in Figures 6(d), 6(e), and 6(f), respectively. OP and OC expression levels increased while that of ON decreased with incubation time. At each time point, BMP-gPLAF group had significantly higher expression of OP and OC, followed by BMP-gPLGF and BMP groups. At each time point, OP and OC expression of BMP-gPLAF was significantly higher than that of BMP-gPLGF, consistent with higher cumulative release of rhBMP-2 from BMP-gPLAF NPs, as shown in Figure 5. After 21 days, OP expression of Control, OM, BMP, BMP-gPLGF, and BMP-gPLAF was 2, 6.5, 6.8, 13, and 29, respectively, and OC expression was 12, 68, 67, 83, and 150. ON expression decreased with incubation as shown in Figure 6(f), and BMP-gPLAF had the lowest ON expression (0.01 fold), followed by BMP-gPLGF (0.09 fold) after 21 days. The increase in OP and OC expression for BMP groups is consistent with the findings of Hu and collaborators [38].

4. Discussion

The unique property of these NPs is that the grafted rhBMP-2 is released concurrent with degradation/erosion of the matrix (compare mass loss in Figure 4(b) with rhBMP-2 release in Figure 5). While the release of model drugs from high molecular weight (40–60 kDa) PLA and PLGA systems is by diffusion through a porous matrix [3941], release from PLAF-NHS and PLGF-NHS NPs is dominated by matrix erosion [22]. NPs self-assembled from the relatively short PLGF-NHS chains (compared to PLGA) degrade primarily by swelling and erosion. The amphiphilic PLEOF chains provide the driving force for NPs swelling in aqueous media, followed by reduction in chain overlap density and finally erosion of the swollen layer. As the swollen layer erodes, a new layer is exposed, and the process of swelling and erosion is repeated until the NP is completely degraded. This process results in linear mass loss of the NPs and linear rhBMP-2 release with incubation time in the first two weeks. Furthermore, the high density of chain ends in low molecular weight PLGF-NHS macromers increases the extent of rhBMP-2 grafting to the NPs. For example, grafting density of heparin to PLGA NPs increased by 29-fold when PLGA molecular weight was reduced from 53 to 15 kDa [42]. Cell viability of BMS cells (4×104 cells/cm2) incubated with 150 mg/mL PLAF NPs (15 times higher than the concentration used for in-vitro osteogenic experiments) was 95±9, 94±2, and 95±3% after 1, 2, and 3 days of incubation, respectively. Furthermore, PLGA and PEG (the building blocks of PLGF and PLEOF) are FDA approved for certain clinical applications, and fumaric acid occurs naturally in the Kreb’s cycle. PLGA degrades to lactic acid and glycolic acid and PEG with molecular weights <4 kDa is excreted by the kidneys. Therefore, PLGF NPs are a viable carrier for sustained in-vivo delivery of rhBMP-2.

Theoretical models and molecular dynamic simulations predict that stability of proteins tethered/grafted to a substrate has enthalpic as well as entropic contributions [43, 44]. While the reduction in entropy of unfolding in the tethered state has a stabilizing effect, the energetic interaction of tethered protein with the substrate can adversely affect stability. rhBMP-2 grafted to the less polar PLAF-NHS has less energetic interaction with the substrate, resulting in higher stability and higher fraction of the released rhBMP-2 in the active conformation. On the other hand, rhBMP-2 grafted to relatively (compared to PLAF-NHS) more polar PLGF-NHS interacts more strongly with the substrate, resulting in lower stability and significantly lower fraction of the released rhBMP-2 in the active conformation. Chung and collaborators conjugated rhBMP-2 to heparin-functionalized PLGA (90 kDa MW) NPs in a fibrin gel [16]. When rhBMP-2 was directly added to fibrin gel, >50% active rhBMP-2 (measured by ELISA) was released from the gel after 2 weeks, but only 22% rhBMP-2 was released when conjugated to PLGA NPs. Our results demonstrate that 37% rhBMP-2 can be released in enzymatically active conformation by grafting to NPs produced from succinimide-terminated low molecular weight PLAF.

The higher OP and OC expression for BMP-gPLAF and BMP-gPLGF groups with sustained release of rhBMP-2 may be related to other factors in the cascade of osteogenesis and mineralization. For example, we have previously shown that differentiation of BMS cells seeded in collagen type-I tubes to vasculogenic lineage and formation of capillary-like structures is related to higher expression of osteopontin [32]. It has also been shown that murine aortic endothelial (MAE) cells transfected with fibroblast growth factor-2 (FGF-2) gene overexpress OP compared to parental cells, and the cross-talk between OP and FGF-2 triggers vasculogenesis [45]. Hirama and collaborators reported that OP-overexpressing neuroblastoma cells contribute to vasculogenesis and tumor growth and the tumorogenic activity correlated with the amount of secreted OP [46]. Furthermore, Hamada and collaborators have demonstrated that a peptide from osteopontin (SVVYGLR) has as potent vasculogenic activity on endothelial cells seeded on collagen gels as vascular endothelial growth factor (VEGF) [47].

Since rhBMP-2 is a potent factor for osteogenic differentiation of progenitor BMS cells, diffusion of rhBMP-2 away from the intended site can cause bone overgrowth [5]. In a practical situation like spine fusion or segmental fractures, rhBMP-2 grafted NPs will be suspended in a natural or synthetic hydrogel precursor solution, and a porous scaffold (the hard phase with long-term degradation time to provide structural stability to the regenerating region) will be dipped in the precursor solution and allowed to crosslink to suspend the NPs in the pore volume of the scaffold. The hydrogel phase immobilizes the NPs and localizes its delivery to the scaffold while the NPs provide a sustained dose of rhBMP-2 with time. rhBMP-2 grafted to degradable PLAF-NHS/PLGF-NHS NPs not only provides a sustained delivery system for recruitment and differentiation of osteoprogenitor cells, but it has the potential to reduce diffusion of the protein away from the regeneration site.

5. Conclusions

Release of active rhBMP-2 from PLGF-NHS and PLAF-NHS NPs was linear in the first 5 and 15 days, respectively, followed by a slower release rate period. Nearly 25% (24.70±1.30%) and 50% (48.7±0.7%) of the grafted rhBMP-2 was released in enzymatically active conformation after complete degradation of PLGF-NHS and PLAF-NHS NPs, respectively. The extent of mineralization of BMS cells as a function of time for rhBMP-2 grafted to PLAF-NHS, PLGF-NHS NPs, and rhBMP-2 directly added to the culture media showed a statistically significant 6.9-, 2.5-, and 19.5-fold increase, respectively, from day 4 to 7 compared to the control group and 6.6-, 7.8-, and 2.7-fold increase from day 7 to 14. For any of the time points, BMP-gPLAF group had significantly higher expression of OP and OC, followed by BMP-gPLGF and BMP groups.


This work was supported by research grants to E. Jabbari from the National Science Foundation (under Grant No. CBET-0756394), The National Institutes of Health (under grant number R03-DE19180), and the National Football League Charities. The authors also thank Dr. Erin L. Connolly and Dr. Zhihuan Gao (Department of Biological Sciences at USC) for the use and for assistance with RT-qPCR machine.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


1. Yaszemski MJ, Oldham JB, Lu L, Currier BL. Clinical needs for bone tissue engineering technology. Em Squared; Toronto: 2000.
2. Wozney JM. Overview of bone morphogenetic proteins. Spine. 2002;27(16):S2–S8. [PubMed]
3. Robinson Y, Heyde CE, Tschoke SK, Mont MA, Seyler TM, Ulrich SD. Evidence supporting the use of bone morphogenetic proteins for spinal fusion surgery. Exp Rev Med Dev. 2008;5(1):75–84. [PubMed]
4. Hillger F, Herr G, Rudolph R, Schwarz E. Biophysical comparison of BMP-2, ProBMP-2, and the free pro-peptide reveals stabilization of the pro-peptide by the mature growth factor. J Biol Chem. 2005;280(15):14974–14980. [PubMed]
5. McKay B, Sandhu HS. Use of recombinant human bone morphogenetic protein-2 in spinal fusion applications. Spine. 2002;27(16):S66–S85. [PubMed]
6. Shields LBE, Raque GH, Glassman SD, Campbell M, Vitaz T, Harpring J, Shields CB. Adverse effects associated with high-dose recombinant human bone morphogenetic protein-2 use in anterior cervical spine fusion. Spine. 2006;31(5):542–547. [PubMed]
7. Jeon O, Song SJ, Yang HS, Bhang SH, Kang SW, Sung MA, Lee JH, Kim BS. Long-term delivery enhances in vivo osteogenic efficacy of bone morphogenetic protein-2 compared to short-term delivery. Biochem Biophys Res Comm. 2008;369(2):774–780. [PubMed]
8. Ruhe PQ, Boerman OC, Russel FGM, Spauwen PHM, Mikos AG, Jansen JA. Controlled release of rhBMP-2 loaded poly(DL-lactic-co-glycolic acid)/calcium phosphate cement composites in vivo. J Contr Rel. 2005;106(1–2):162–171. [PubMed]
9. Wildemann B, Kandziora F, Krummrey G, Palasdies N, Haas NP, Raschke M, Schmidmaier G. Local and controlled release of growth factors (combination of IGF-I and TGF-beta I, and BMP-2 alone) from a polylactide coating of titanium implants does not lead to ectopic bone formation in sheep muscle. J Contr Rel. 2004;95(2):249–256. [PubMed]
10. Maire M, Chaubet F, Mary P, Blanchat C, Meunier A, Logeart-Avramoglou D. Bovine BMP osteoinductive potential enhanced by functionalized dextran-derived hydrogels. Biomaterials. 2005;26(24):5085–5092. [PubMed]
11. Patel ZS, Yamamoto M, Ueda H, Tabata Y, Mikos AG. Biodegradable gelatin microparticles as delivery systems for the controlled release of bone morphogenetic protein-2. Acta Biomater. 2008;4(5):1126–1138. [PMC free article] [PubMed]
12. Chen FM, Zhao YM, Sun HH, Jin T, Wang QT, Zhou W, Wu ZF, Jin Y. Novel glycidyl methacrylated dextran (Dex-GMA)/gelatin hydrogel scaffolds containing microspheres loaded with bone morphogenetic proteins: Formulation and characteristics. J Contr Rel. 2007;118(1):65–77. [PubMed]
13. Wei G, Jin Q, Giannobile WV, Ma PX. The enhancement of osteogenesis by nano-fibrous scaffolds incorporating rhBMP-7 nanospheres. Biomaterials. 2007;28:2087–2096. [PMC free article] [PubMed]
14. Lin H, Zhao Y, Sun W, Chen B, Zhang J, Zhao W, Xiao Z, Dai J. The effect of crosslinking heparin to demineralized bone matrix on mechanical strength and specific binding to human bone morphogenetic protein-2. Biomaterials. 2008;29:1189–1197. [PubMed]
15. Fu YC, Nie H, Ho ML, Wang CK, Wang CH. Optimized bone regeneration based on sustained release from three-dimensional fibrous PLGA/HAp composite scaffolds loaded with BMP-2. Biotechnol Bioeng. 2008;99(4):996–1006. [PubMed]
16. Chung YI, Ahn KM, Jeon SH, Lee SY, Lee JH, Tae G. Enhanced bone regeneration with BMP-2 loaded functional nanoparticle-hydrogel complex. J Contr Rel. 2007;121(1–2):91–99. [PubMed]
17. Schrier JA, Fink BF, Rodgers JB, Vasconez HC, DeLuca PP. Effect of a freeze-dried CMC/PLGA microsphere matrix of rhBMP-2 on bone healing. AAPS PharmSciTech. 2001;2(3):E18. [PMC free article] [PubMed]
18. Jabbari E, Florschutz AV, Petersen LG, Liu N, Lu L, Currier BL, Yaszemski MJ. Release characteristics of recombinant human bone morphogenic protein-2 from PLGA microspheres embedded in a poly(propylene fumarate) porous scaffold. Transactions of 7th World Biomaterials Congress; Sidney, Australia. 2004. p. 512.
19. Wei Q, Wei W, Lai B, Wang LY, Wang YX, Su ZG, Ma GH. Uniform-sized PLA nanoparticles: Preparation by premix membrane emulsification. Int J Pharmaceut. 2008;359(1–2):294–297. [PubMed]
20. Meikle MC. On the transplantation, regeneration and induction of bone: The path to bone morphogenetic proteins and other skeletal growth factors. Surgeon - Journal of the Royal College of Surgeons of Edinburgh and Ireland. 2007;5(4):232–243. [PubMed]
21. He X, Ma J, Mercado AE, Xu W, Jabbari E. Cytotoxicity of Paclitaxel in biodegradable self-assembled core-shell poly(lactide-co-glycolide ethylene oxide fumarate) nanoparticles. Pharm Res. 2008;25(7):1552–1562. [PubMed]
22. Mercado AE, He X, Xu W, Jabbari E. The release characteristics of a model protein from self-assembled succinimide-terminated poly(lactide-co-glycolide ethylene oxide fumarate) nanoparticles. Nanotechnology. 2008;19(32):325609. [PubMed]
23. Jabbari E, He X. Synthesis and characterization of bioresorbable in situ crosslinkable ultra low molecular weight poly(lactide) macromer. J Mater Sci Mater Med. 2008;19:311–318. [PubMed]
24. Jabbari E, He X, Valarmathi M, Sarvestani AS, Xu W. Material properties and bone marrow stromal cells response to in situ crosslinkable rgd-functionlized lactide-co-glycolide scaffolds. J Biomed Mater Res. 2009;89A(1):124–137. [PubMed]
25. Sarvestani A, He X, Jabbari E. Viscoelastic characterization and modeling of gelation kinetics of injectable in situ cross-linkable poly(lactide-co-ethylene oxide-co-fumarate) hydrogels. Biomacromolecules. 2007;8:406–415. [PubMed]
26. Sarvestani A, Xu W, He X, Jabbari E. Gelation and degradation characteristics of in situ photo-crosslinked poly(L-lactid-co-ethylene oxide-co-fumarate) hydrogels. Polymer. 2007;48:7113–7120.
27. He X, Jabbari E. Material properties and cytocompatibility of injectable MMP degradable poly(lactide ethylene oxide fumarate) hydrogel as a carrier for marrow stromal cells. Biomacromolecules. 2007;8:780–792. [PubMed]
28. Morpurgo M, Bayer EA, Wilchek M. N-hydroxysuccinimide carbonates and carbamates are useful reactive reagents for coupling ligands to lysines on proteins. J Biochem Biophys Methods. 1999;38:17–28. [PubMed]
29. Lindholm S. Advances in skeletal reconstruction using bone morphogenetic proteins. University of Oulu; Finland: 2002.
30. He X, Ma J, Jabbari E. Effect of Grafting RGD and BMP-2 Protein-Derived Peptides to a Hydrogel Substrate on Osteogenic Differentiation of Marrow Stromal Cells. Langmuir. 2008;24(21):12508–12516. [PubMed]
31. Vehof JWM, De Ruijter AE, Spauwen PHM, Jansen JA. Influence of rhBMP-2 on rat bone marrow stromal cells cultured on titanium fiber mesh. Tissue Eng. 2001;7(4):373–383. [PubMed]
32. Henderson JA, He XZ, Jabbari E. Concurrent differentiation of marrow stromal cells to osteogenic and vasculogenic lineages. Macromol Biosci. 2008;8(6):499–507. [PubMed]
33. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001;29(9):e45. [PMC free article] [PubMed]
34. Scheufler C, Sebald W, Hülsmeyer M. Crystal Structure of Human Bone Morphogenetic Protein-2 at 2.7 Å Resolution. J Mol Biol. 1999;287:103–115. [PubMed]
35. Huss R, Hoy CA, Deeg HJ. Contact- and growth factor-dependent survival in a canine marrow-derived stromal cell line. Blood. 1995;85(9):2414–2421. [PubMed]
36. Marolt D, Augst A, Freed LE, Vepari C, Fajardo R, Patel N, Gray M, Farley M, Kaplan D, Vunjak-Novakovic G. Bone and cartilage tissue constructs grown using human bone marrow stromal cells, silk scaffolds and rotating bioreactors. Biomaterials. 2006;27(36):6138–6149. [PubMed]
37. Hecht BP, Fischgrund JS, Herkowitz HN, Penman L, Toth JM, Shirkhoda A. The use of recombinant human bone morphogenetic protein 2 (rhBMP-2) to promote spinal fusion in a nonhuman primate anterior interbody fusion model. Spine. 1999;24(7):629–636. [PubMed]
38. Hu ZM, Peel SAF, Ho SKC, Sandor GKB, Clokie CML. Role of bovine bone morphogenetic proteins in bone matrix protein and osteoblast-related gene expression during rat bone marrow stromal cell differentiation. J Craniofac Surg. 2005;16(6):1006–1014. [PubMed]
39. Klose D, Siepmann F, Elkharraz K, Krenzlin S, Siepmann J. How porosity and size affect the drug release mechanisms from PLIGA-based microparticles. Int J Pharmaceut. 2006;314(2):198–206. [PubMed]
40. Messaritaki A, Black SJ, van der Walle CF, Rigby SP. NMR and confocal microscopy studies of the mechanisms of burst drug release from PLGA microspheres. J Contr Rel. 2005;108(2–3):271–281. [PubMed]
41. Siepmann J, Elkharraz K, Siepmann F, Klose D. How autocatalysis accelerates drug release from PLGA-based microparticles: A quantitative treatment. Biomacromolecules. 2005;6(4):2312–2319. [PubMed]
42. Jeon O, Kang SW, Lim HW, Chung JH, Kim BS. Long-term and zero-order release of basic fibroblast growth factor from heparin-conjugated poly(L-lactide-co-glycolide) nanospheres and fibrin gel. Biomaterials. 2006;27(8):1598–1607. [PubMed]
43. Knotts TA, Rathore N, de Pabloz JJ. An entropic perspective of protein stability on surfaces. Biophys J. 2008;94(11):4473–4483. [PubMed]
44. Friedel M, Baumketner A, Shea JE. Stability of a protein tethered to a surface. J Chem Phys. 2007;126(9):1–12. [PubMed]
45. Leali D, Dell’Era P, Stabile H, Sennino B, Chambers AF, Naldini A, Sozzani S, Nico B, Ribatti D, Presta M. Osteopontin (Eta-1) and fibroblast growth factor-2 cross-talk in angiogenesis. J Immunol. 2003;171(2):1085–1093. [PubMed]
46. Hirama M, Takahashi F, Takahashi K, Akutagawa S, Shimizu K, Soma S, Shimanuki Y, Nishio K, Fukuchi Y. Osteopontin overproduced by tumor cells acts as a potent angiogenic factor contributing to tumor growth. Cancer Lett. 2003;198(1):107–117. [PubMed]
47. Hamada Y, Egusa H, Kaneda Y, Hirata I, Kawaguchi N, Hirao T, Matsumoto T, Yao M, Daito K, Suzuki M. Synthetic osteopontin-derived peptide SVVYGLR can induce neovascularization in artificial bone marrow scaffold bioniaterials. Dental Mater J. 2007;26(4):487–492. [PubMed]