Hollow calcium phosphate microcarriers for bone regeneration: In vitro ...

Bio-Medical Materials and Engineering 17 (2007) 277–289 IOS Press

277

Hollow calcium phosphate microcarriers for bone regeneration: In vitro osteoproduction and ex vivo mechanical assessment Brandon G. Santoni a,∗ , G. Elizabeth Pluhar b , Tatiana Motta b and Donna L. Wheeler c a

Department of Mechanical Engineering, Orthopaedic Bioengineering Research Laboratory, Colorado State University, Fort Collins, CO 80523, USA b Department of Veterinary Clinical Sciences, University of Minnesota, St. Paul, MN 55108, USA c BioSolutions Consulting, LLC, Webster, NY 14580, USA Received 7 August 2006 Revised 11 December 2007 Abstract. Synthetic grafting materials, such as calcium phosphates (hydroxyapatite, HA; tricalcium phosphate, TCP), polymers, or composites thereof, can be used as osteoconductive scaffolds and delivery vehicles for osteoinductive growth factors. Carrier materials must be engineered to deliver these factors in a controlled fashion at a rate and dose consistent with the biological need and responsiveness of the system to optimize bone formation and ingrowth. They should also simultaneously provide mechanical support and slowly resorb as new bone is formed. This investigation assessed the elution characteristics of BMP-7 (OP-1) from hollow calcium phosphate spheres of varying chemical composition (HA/β-TCP) and porosity (dense/porous). The pharmacokinetics indicated a bimodal trend of protein release with protein elution peaking between fifteen and thirty minutes in solution (bolus release) and continuing through the eight-week time point (sustained release). Eluted OP-1 bioactivity was characterized over a three-week period using mesenchymal stem cell (MSC) cultures and included assessment of the protein’s differential, proliferative, and calcified nodule forming abilities. Alkaline phosphatase enzyme (ALP) activity in MSCs peaked between 12 and 16 days post-OP-1 exposure. Elutant from the HA dense treatment group induced the highest degree of ALP expression while elutant from the β-TCP treatment groups induced the formation of significantly higher numbers of calcified nodules in culture. The aggregate modulus of a clinically relevant 2 cc dose of carriers was quantified using custom designed testing fixtures to investigate the effects of carrier size, porosity, chemical composition, and the presence of a central hole on mechanical integrity. Significant increases in moduli were noted for carrier size and chemical composition (HA > β-TCP). These preliminary in vitro and ex vivo results indicate the clinical potential of the hollow calcium phosphate carriers as successful load-bearing delivery vehicles for OP-1. Keywords: Biomaterials, biomechanics, drug delivery, bone regeneration, bone morphogenic protein, mesenchymal stem cell

1. Introduction Regeneration of lost bone stock in load-bearing skeletal areas is a serious clinical problem. Most commonly, autograft or allograft is packed into the defect to provide an osteoconductive and/or osteoinductive template for new bone formation. Autograft and allograft have advantages in the aforementioned * Address for correspondence: Brandon G. Santoni, PhD, Musculoskeletal Oncology Research Fellow, Department of Clinical Sciences, Animal Cancer Center, Colorado State University, 300 West Drake Rd., Campus Delivery Code #1620, Fort Collins, CO 80523, USA. Tel.: +1 970 297 4451; Fax: +1 970 297 1254; E-mail: [email protected]

0959-2989/07/$17.00  2007 – IOS Press and the authors. All rights reserved

278

B.G. Santoni et al. / Hollow calcium phosphate microcarriers for bone regeneration

clinical settings, but they also have serious disadvantages [1]. The usefulness of autogenous bone is limited by availability, usually being harvested from the iliac crest, rib, or fibula, and the necessity of a second operation that may result in donor site morbidity following surgery [2]. Allograft bone is an attractive alternative because it is relatively abundant and can be used when large structural restorations are required. Unfortunately, massive allograft bone has minimal osteoinductivity [3] and a limited capacity to incorporate with host bone [4–7]. A biomimetic material providing initial mechanical support, sustained release of growth factors, and biomaterial resorption concurrent with bone ingrowth would characterize an ideal osteoinductive/conductive synthetic scaffold as an alternative to traditional osseous reconstructive materials. Supplementation of bone graft materials with BMP-2, BMP-7 (OP-1), and β-fibroblast growth factor (bFGF) has been shown to improve osseous incorporation of the graft material [8–12]. OP-1 has a proven clinical record as an accelerator of fracture healing [13–15] and bone ingrowth into bone grafts and bone substitutes [14] in clinical trials. While these growth factors demonstrate an osteoinductive effect in vitro, when implanted subcutaneously, they diffuse rapidly and thus require a carrier to retain the growth factor locally and prolong osteogenesis in vivo [16]. Recently, calcium phosphate blocks, particles and spheres have been engineered to enable release of drugs and growth factors [17–20]. Calcium phosphates have numerous advantages over polymer or collagenous carriers. Unlike polymers, naturally occurring calcium phosphates are similar in composition to the mineralized portion of bone and display an innate affinity to bind growth factors. Additionally, their degradation byproducts are biologically friendly in the osteogenic environment. Because of these characteristics, calcium phosphates were used in some of the initial studies to characterize the effects of BMPs on osteogenesis in subcutaneous muscle pockets [21]. They also display superior compressive mechanical strength and modulus compared to that of polymers and collagen carriers making them more appropriate for load-bearing. The overall goal of this research was to evaluate synthetic HA and β-TCP, dense and porous hollow calcium phosphate microcarriers (CaP Biotechnology, Golden, CO, USA) as drug delivery scaffolds for OP-1. Specifically, the mechanical properties of the calcium phosphate carriers were quantified and compared to other marketed bone graft substitutes. Additionally, the pharmacokinetics and bioactivity of eluted OP-1 from the hollow carriers was characterized.

2. Materials and methods 2.1. Mechanical testing A two-cc volume of each microcarrier treatment group (n = 16) consisting of various combinations of carrier composition (HA or β-TCP), density (dense or porous), size (medium or large) and central hole (presence or absence) was soaked in PBS for 30 minutes and placed between two porous steel platens (Mott Corporation, Farmington, CT, USA) in a custom designed acrylic confined compression chamber. The samples were placed into a materials testing machine (MTS, Eden Prairie, MN, USA) and exposed to stepwise stress relaxation tests in strain increments of 0.25, 0.5, 1.0, 1.5, 2.0 and 5.0% strain at a loading rate of 0.4 mm/min [22]. Between each stepwise increment in strain, samples were allowed to reach equilibrium levels of stress. Each treatment group was tested twice and each value of equilibrium stress and its corresponding level of strain were then plotted. The slope of this line was taken to be the aggregate modulus, HA [23]. The same technique was used to obtain HA values for


279

Pro Osteon 500R™ (Interpore Orthopaedics, Irvine, CA, USA), Perioglass® (USBiomaterials, Alachua, FL, USA), BioOss® (Geistlich Biomaterials, Switzerland), Ceros® TCP (Mathy Dental, Switzerland) as well as ovine cortical and cancellous autograft. Biomaterial HA values were compared using a 1-way ANOVA (α = 0.05) using SAS (Statistical Analysis Software, Cary, NC, USA). 2.2. Pharmacokinetics 2.2.1. OP-1 reconstitution and elution from microcarriers Lyophilized OP-1 (Stryker Biotech, Hopkington, MA, USA) was reconstituted in a 47.5%EtOH/ 0.01%trifloroacetic acid (TFA) solution. Aliquots containing 760 µg of OP-1 were transferred to 15 ml conical vials containing 1 ml of HA dense, TCP dense, HA porous, or TCP porous microcarriers prefabricated with a central hole from exterior to interior (Fig. 1) to maximize surface area for protein adsorption. The experimental design and treatment designations are given in Table 1. Protein adsorption was accomplished using a vacuum desiccation technique pulling approximately 25 psi of vacuum pressure for 10 min at 37◦ C. The cycle of vacuum application and release was repeated several times. The microcarriers were then transferred to separate 5 ml cryovials for the duration of the experiment. One ml of PBS at pH 7.0 was added to the microcarriers and the cryovials were maintained at 37◦ C under constant agitation in a shaking water bath. The elutant was collected and more PBS added at the

Fig. 1. Microscopic examination of variations in diameter, hole size, and overall shape for the TCP medium sized porous carrier (A) and the HA medium sized porous carrier (B). The diameter of the medium sized carriers is approximately 750 µm. Density is a qualitative parameter determined through high-resolution SEM images. Table 1 Pharmacokinetics experimental design Carrier HA-Dense HA-Dense HA-Porous HA-Porous TCP-Dense TCP-Dense TCP-Porous TCP-Porous

Protein 0.76 mg OP-1 none 0.76 mg OP-1 none 0.76 mg OP-1 none 0.76 mg OP-1 none

Code HSO HSX HPO HPX TSO TSX TPO TPX

1 ml samples n=6 n=6 n=6 n=6 n=6 n=6 n=6 n=6

280

B.G. Santoni et al. / Hollow calcium phosphate microcarriers for bone regeneration Table 2 Sampling times and pH adjustments Time VIAL, 0, 15, 30, 45, 60 min; 2, 4, 8 h; 1, 1. 5, 3 day 5 day 1 week 2 week 3 week 4 week 6, 8 weeks

pH 7 7 7 7.1 7.1 7.4 7.3 7.2 7.4

collection time points and pH adjustments noted in Table 2. The pH adjustments were made to mimic pH changes in the local wound environment during osseous healing [24,25], which may effect protein pharmacokinetics [26]. The collected elutant from all time points was assayed in duplicate to quantify protein concentration using a colorimetric assay (BioRad Laboratories, Hercules, CA, USA) and an automated plate reader (Phenix GENics) at a wavelength of 595 nm. 2.3. Bioactivity/osteoproduction 2.3.1. Mesenchymal stem cell culture MSC cultures were prepared from bone marrow harvested from 60 day old F344 rats (Charles River Lab, Wilmington, MA) using a previously described technique [27–29]. Briefly, the rats were euthanized by CO2 asphyxiation and the femora, tibiae, and humeri were disarticulated and dissected free of soft tissue. Epiphyses were removed using a bone ronguer. A 10 ml syringe filled with supplemented media (sMedia) composed of Dulbecco’s modified Eagle’s medium containing 500 mg/l glucose supplemented with 10% fetal bovine serum (FBS) and antibiotics (penicillin 100 µg/ml, streptomycin 100 µg/ml) and an 18-gauge needle were used to flush the bone marrow from the diaphysis into a sterile culture dish. The resulting cell suspension was centrifuged, the supernatant aspirated, and cells resuspended in sMedia. Following primary culture in 56 cm2 tissue culture dishes, the cells were seeded at a density of 2.0 × 104 cells/well in 24-well plates and incubated. A total of 180 2-cm2 wells were seeded to study the effect of OP-1 on cell differentiation, proliferation, and calcified nodule forming ability. 2.3.2. Exposure to growth factor Twenty-four hours after seeding in the 2-cm2 wells the sMedia was supplemented with 100 ng of OP-1 collected either from the pooled elutant of the 7–28 days time points of all four treatment groups (HPO, HSO, TPO, TSO) or from freshly reconstituted OP-1 to serve as the positive control. In separate wells, equal volumes of elutant from the four treatment groups without OP-1 (HPX, HSX, TPX, TSX) were added to the MSCs in culture. For the negative control, sterile PBS was used. All control wells were supplemented with the same volume of sample as the treatment OP-1 wells. All treatments were run with three replicates (n = 3) per treatment group per time point. Following initial exposure, the sMedia plus protein was replaced every 96 hours. At 4, 8, 12, 16 and 20 days, the cells were evaluated for differentiation and proliferation with the alkaline phosphatase and crystal violet assays, respectively. At day 20 only, Von Kossa staining for calcified nodules was performed.


281

2.3.3. Cellular differentiation Alkaline phosphatase (ALP) has been reported as an early, distinctive marker of mesenchymal cell differentiation into osteoblasts [30,31]. ALP activity levels were quantified using the methods described by Cassiede et al. [29]. Briefly, all wells were washed twice in Tyrode’s salt solution and filled with 300 µl of alkaline phosphatase buffer containing p-nitrophenyl phosphate (Sigma-Aldrich Corp., St. Louis, MO, USA), which when exposed to ALP is converted to p-nitrophenol (PNP). After ten minutes, the buffer was removed and transferred to a separate 24-well plate containing 300 µl of 1 N NaOH to stop the color reaction. The absorbance of the wells was read at a wavelength 405 nm with an automated plate reader (Phenix GENics) and values were compared to a standard curve created using concentrations of PNP (Sigma-Aldrich Corp., St. Louis, MO, USA) ranging from 0.0045 to 0.045 µmol/ml. Alkaline phosphatase activity was expressed as µmol p-nitrophenol liberated per 10 minutes of exposure to the p-nitrophenyl phosphate substrate. Each sample was read in triplicate and the mean enzyme activity was compared statistically to activity levels of the positive control after twelve days in culture using a oneway ANOVA [32]. A Tukey’s Studentized range post-hoc test was performed to determine differences between treatment groups. Effects of carrier density and composition on eluted OP-1 bioactivity over the twenty-day experimental period were investigated with a two-way ANOVA. 2.3.4. Cellular proliferation Immediately following the ALP assay, the cells were assayed for proliferation using a modification of the method proposed by Westergren-Thorsson et al. [33]. Briefly, the wells were rinsed and fixed in 1% gluteraldehyde in Tyrode’s salt solution (v/v) for 15 min, rinsed twice with distilled, deionized milli-q H2 O (mqH2 O) and air-dried. Cultures were then stained with 1 ml 0.1% crystal violet in water (w/v) for 30 min and rinsed. The crystal violet solution was removed from the wells and discarded, and 2 ml of 1% Triton X-100 in water (v/v) was added to each well and agitated overnight on a rotary shaking table to extract the stain taken up by the viable cells. The following day, the absorbance of the solution in each well was measured with the automated plate reader (Phenix GENics) at a wavelength of 595 nm. The amount of crystal violet extracted per well is directly proportional to the number of viable cells [27]. To assess the relationship between cell number and bound dye, first passage cells were seeded in 2 cm2 wells in triplicate at densities ranging from 6.25 × 103 to 2.0 × 105 cells/well. Twenty-four hours after plating the cells were fixed and stained with crystal violet. The plot of absorbance versus cell number was used as the standard curve for determination of cell number for all subsequent crystal violet assays. With these results, the effect of OP-1 as a proliferative agent was investigated. In addition, alkaline phosphatase activity levels were normalized to cell number. These data were analyzed in a fashion identical to that described above for cellular differentiation. 2.3.5. Bone nodule formation Calcified matrix production within the culture wells was measured at day 20 using Von Kossa staining as described previously [29]. Briefly, the wells were rinsed and fixed for 30 min with 10% glutaraldehyde. One ml of a 2% silver nitrate in mqH2 O (w/v) solution was added to each well and the plates were left in darkness for 30 min. The wells were rinsed and filled with distilled water and exposed to bright light for 15 min at which time the reaction was stopped by rinsing with mqH2 O. The mineralized nodules were counted using an image analysis system (Image Pro Plus, Silver Springs, MD, USA). A one-way ANOVA was performed to examine the effect of treatment on bone nodule formation.

282


3. Results 3.1. Mechanical testing Confined compression stress relaxation testing (Fig. 2) indicated that aggregates of hollow carriers composed of HA were significantly stiffer than those of β-TCP (p < 0.05). Aggregate moduli ranged from a minimum of 18.5 ± 6.82 MPa (β-TCP, 750 µm, porous; TMP in Table 3) to a maximum value of 109.7 ± 14.6 MPa (HA, 750 µm, dense; HMD in Table 3). The presence of a central hole significantly reduced HA . In general, carrier porosity had no significant effect on aggregate modulus (Fig. 3). Summary mechanical data for all permutations of the hollow calcium phosphate microcarriers and other commercially available products are presented in Table 3.

Fig. 2. Stress relaxation curve for the TCP, large, dense (A) treatment group; Provided in (B) is the linear regression line determined from (A) calculated from a total of ten data points; Data points (•) in (B) represent average levels of stress calculated from the four-second interval immediately proceeding the ensuing ramp to a particular level of strain; HA was calculated as the slope of the line in (B). Table 3 Comparison of confined compression stress relaxation results for the hollow calcium phosphate microcarriers versus common bone graft substitutes. Results expressed as mean ± SEM Grafting material a

HMD Perioglass Cortical Bone Chips TMPb

HA (MPa) 109.7 ± 14.6∗ 29.5 ± 4.30 23.2 ± 2.32 18.5 ± 6.82

BioOss

6.61 ± 0.85∗

Ceros TCP

5.49 ± 2.08∗

Pro Osteon 500R

5.02 ± 1.47∗

Cancellous Bone

1.72 ± 0.49∗

a,b

+

− − − −

Signify calcium phosphate maximum (HMD) and minimum (TMP), respectively.

*+/−

Significantly greater+ or less− than TMP, p < 0.05; HMD > Perioglass, p < 0.0001.


283

Fig. 3. Aggregate modulus (HA ) comparisons examining the effects of density (A and C) and composition (B and D). Presented here are moduli from the large (A and B) and medium (C and D) carriers without the central hole. As an example, the bar graphs in A indicate that a 2 cc volume of the large HA, dense microcarriers (2) exhibits a significantly higher aggregate modulus than the same volume of the large, HA, porous carriers (1, p = 0.003) while the effect of porosity for the large TCP carriers was not significant (p = 0.172). In general, HA carriers displayed higher aggregate moduli than β-TCP carriers while the effects of density on mechanical performance were negligible.

3.2. Elution kinetics Osteogenic protein-1 was released from each of the four carrier types in a bimodal fashion (Fig. 4). An initial bolus of OP-1 elution peaked between fifteen and thirty minutes for all treatment groups, though the rate of OP-1 elution from the HA carriers was significantly greater than that of the β-TCP carriers (p < 0.05) during this period. Furthermore, at the fifteen minutes time point, an effect of density of OP-1 elution was noted for the HA carriers (Fig. 4). The early bolus release was followed by a sustained release phase that lasted the duration of the eight-week evaluation period. At 28 days, the TCP dense and the HA porous carriers were eluting OP-1 at a rate of 20.5 and 23.4 ng OP-1/h, respectively. 3.3. Bioactivity/osteoproduction The effect of OP-1 eluted from the microcarrier architecture on alkaline phosphatase expression in MSCs resulted in a very consistent temporal pattern of activity for all treatment groups (Fig. 5A, 5B),

284


Fig. 4. Pharmacokinetic data normalized to time (i.e. µg/h) illustrates a bi-modal trend of OP-1 release from carriers with a central hole (Fig. 1) with an early bolus period followed by a longer period of sustained release; the rate of OP-1 elution from all carriers peaked at 15 minutes with HA carriers eluting protein at a significantly higher rate than TCP carriers (*, p < 0.05); furthermore HA, porous carriers eluted OP-1 at a rate twice that of the HA dense carriers (ζ, p < 0.05); all carriers continued to elute OP-1 at concentrations effective in standard bioactivity assays and at therapeutic doses throughout the course of the experiment (inset).

including the positive control (Fig. 5C). ANOVA indicated no significant osteoinductive effect of ceramic alone on MSC differentiation (p = 0.8832). At 12 days, the ALP activity of OP-1 eluted from the HA dense carrier was greater than the activity of the positive control, yet not significant (Fig. 5D, p = 0.1817). Activity levels for the HA porous and TCP porous treatment groups were significantly lower than the positive control at day 12 (p = 0.0033 and p = 0.0078, respectively). Normalized ALP activity revealed trends of increased ALP expression per cell number at 12 and 16 days post OP-1 exposure. OP-1 eluted from the TCP dense and porous carriers and the positive control groups induced the formation of significantly more nodules than the HA dense and porous carriers (Fig. 6, p < 0.05). In general, elutant from the TCP carriers induced significant increases in nodule production compared to the HA treatment groups (p = 0.0022). 4. Discussion In this investigation, synthetically produced hydroxyapatite (Ca10 (PO4 )6 (OH)2 ) and β-tricalcium phosphate (Ca3 (PO4 )2 ) were evaluated as delivery vehicles for OP-1 by in vitro pharmacokinetics and cell culture conditions and ex vivo mechanical testing. The osteoinductive characteristics of OP-1 have been well documented through various in vitro cell culture experiments [30,32], and in treating and healing critical-sized defects in a variety of mammalian models [12,14,20,34,35]. Pharmacokinetic data indicated bimodal release of OP-1 from the ceramic carrier with the eluted protein stimulating peak alkaline phosphatase expression in mesenchymal stem cells at 12 and 16 days post-exposure. The stepwise


285

Fig. 5. Alkaline phosphatase (ALP) results: (A) Illustration of TPO treatment induced activity levels (2) on the MSCs versus the treatment control (TPX, 1) through twenty days and (B) the resulting effect of OP-1 eluted from TCP porous carriers (i.e. 2–1 for each time point); (C) bioactivity of carrier released OP-1 (2–1 for each time point for each carrier treatment group) versus the positive control for all time points; (D) day twelve bioactivity comparison; a > b, p < 0.05; b > c, p < 0.05.

stress relaxation experiments performed on the 2 cc carrier aggregates in confined compression resulted in a very wide range of aggregate moduli ranging from 18 to 110 MPa. The carriers performed as well as, or better than, other synthetic and natural bone graft substitutes under mechanical compression. Hydroxyapatite and β-TCP have distinguishable advantages over the natural and synthetic polymers and osseous grafting materials as protein vehicles [1]. Their application as bone substitutes has received considerable attention due to a close chemical and crystal resemblance to natural bone mineral [36] and very desirable biocompatibility and bioactivity traits [18,36,37]. Confined compression tests successfully determined the aggregate modulus of these and other bone grafting materials. In general, the HA carriers displayed significantly higher moduli than the β-TCP carriers, while density had no significant effect on aggregate modulus. The maximum HA for the hollow calcium phosphate microcarriers was 110 MPa, five times greater than morsellized cortical allograft and twenty times greater than the coralline Pro Osteon 500R™. Aggregate moduli results for the carrier permutations indicated a wide range in mechanical properties that may be suitable for different skeletal

286


Fig. 6. Bone nodule formation expressed as number of nodules formed per treatment group relative to the control. OP-1 eluted from all treatment carriers, including the positive control, induced the formation of significantly more nodules per well than did the treatment controls (p < 0.05). (*) indicates significantly enhanced nodule formation compared to the HPO and HSO treatment groups. Results are expressed as mean ± SEM.

structural demands and clinical applications. Despite favorable comparisons to synthetic and natural bone substitutes, microcarrier moduli values were approximately two orders of magnitude less than fresh human femoral trabecular bone (7 GPa) [38] and cortical bone (17–20 GPa) [39], though the method of quantifying cortical bone modulus was different than that used in this study. Nevertheless, the moduli for these microcarriers are equal to or greater than other published synthetically produced calcium phosphate biomaterials [40,41]. Because aggregates of calcium phosphate ceramics display pseudo-viscoelastic properties in physiological solutions [42], the stepwise stress relaxation environment accommodated an aggregate characteristic and proved a successful means of modulus testing. Since these materials are brittle in nature, it is assumed that the materials themselves are not biphasic but rather, under confined compression settings, the graft particles slide past one another and re-position themselves within the compression chamber under applied load. Such a phenomenon may likely be experienced in the clinical situation during the grafting procedure. Pharmacokinetics indicated a bimodal protein release with an early bolus phase that peaked within thirty minutes, followed by a period of sustained release at lower, though therapeutically relavent [10], concentrations lasting the experimental duration. This mechanism of protein release has been hypothesized as optimal for eliciting the cascade of events that leads to osteogenesis [42]. Our findings indicate that the adsorption process effectively binds the OP-1 deep within ceramic architecture, allowing the slow release of residual protein over time. Unfortunately, the cumulative amount of protein quantified after eight-weeks in solution was significantly less than the 760 µg initially added in solution to the microcarriers. We hypothesized that some OP-1 adhered to the conical vial surface instead the ceramic. Indeed, additional experiments confirmed this hypothesis as we observed that increasing the number of


287

cycles of vacuum desiccation resulted in decreased OP-1 absorption within the carriers concomitant with increased OP-1 absorption to the inside of the conical vials. Certainly this finding represents a limitation of the study and future studies should identify a method that optimizes protein adsorption only within the carrier architecture. It is unknown if OP-1 release kinetics demonstrated here would be adequate for initiating osteogenesis in vivo. The existence of an optimal pharmacokinetic profile for osteoinduction is unknown. The extremes of release are clearly not beneficial to bone induction (i.e. initial bolus with no prolonged sustained release or vice versa), however as illustrated by Uludag et al. [43] using a collagen carrier and rhBMP-2, it has been shown that intermediate profiles can induce new bone formation. The ultimate test for this calcium phosphate material would be a detailed assessment of the dynamic in vivo responses to this drug delivery vehicle. It was observed during the crystal violet assays that a number of the cells were released from tissue culture dishes during the fixation step. Though this is a procedural shortcoming, if this loss occurs similarly in each culture dish, normalization of ALP activity relative to cell number is justifiable. In all treatment groups, normalized ALP activity peaked at 16 days followed by a decrease at 20 days, indicating that increased enzyme levels were the result of increased expression in individual cells. At day twelve, the elutant from the HA dense treatment group elicited the highest level of enzyme activity. This seems to correspond to previous findings reported by Alam et al. of enhanced in vivo bone induction determined by ALP quantification using an HA-rhBMP-2 delivery system [9]. This enhanced ALP activity could be attributed to a synergistic effect of this particular ceramic with OP-1, although the actual mechanism of such an effect remains uncharacterized. Eluted OP-1 activity peaked for all carriers at day 16, whereas peak activity for the positive control occurred at day 12, a finding that corroborates other studies using osteoblast-like cells [32]. This delay in peak activity of OP-1 eluted from the carrier architecture may be the result of ceramic microparticulate in the elutant, though microparticulate presence alone was found to have an insignificant osteoinductive effect. Our findings are in agreement with earlier reports that calcium phosphates are only osteoconductive [18]. Furthermore, OP-1 elutant from the TCP ceramic produced significantly more nodules than the HA composition and positive control. This finding may be attributed to accelerated TCP degradation relative to HA. This ceramic particulate may have served as an osteoconductive scaffold to which cells could adhere and lay down extracellular calcium matrix. In conclusion, bimodal release of OP-1 from the hollow microcarriers was demonstrated and protein bioactivity was maintained after direct exposure to the ceramic architecture. This preliminary study indicates that these hollow HA and β-TCP microcarriers would provide an effective load-bearing delivery vehicle for sustained release of OP-1 and possibly other proteins or drugs. Future studies may aim to quantify the pharmacokinetics, bioactivity, and mechanical performance of composite aggregates combining the effects of the osteoinductive HA with the osteoconductive β-TCP. Once the in vitro and ex vivo characteristics of this graft material are fully characterized, in vivo studies may aim to fully characterize the clinical viability and applications of this novel grafting material. Acknowledgements The authors would like to thank the Colorado Commission of Higher Education and the NSF for financial support of this project, CaP Biotechnology for providing the graft material, Stryker Biotech for providing the OP-1, and Steve Jackinsky, Elena Young and Dr. Chad Lewis for their contributions to the mechanical testing.

288


References [1] V.M. Goldberg and S. Stevenson, Natural history of autografts and allografts, Clin. Orthop. Relat. Res. 225 (1987), 7–16. [2] B.N. Summers and S.M. Eisenstein, Donor site pain from the ilium. A complication of lumbar spine fusion, J. Bone Joint Surg. Br. 71 (1989), 677–680. [3] C.J. Damien and J.R. Parsons, Bone graft and bone graft substitutes: A review of current technology and applications, J. Appl. Biomater. 2 (1991), 187–208. [4] R.R. Pelker, G.E. Friedlaender and T.C. Markham, Biomechanical properties of bone allografts, Clin. Orthop. Relat. Res. 174 (1983), 54–57. [5] K.G. Heiple, S.W. Chase and C.H. Herndon, A comparative study of the healing process following different types of bone transplantation, J. Bone Joint Surg. Am. 45 (1963), 1593–1616. [6] C. Delloye, M. Verhelpen, J. d’Hemricourt, B. Govaerts and R. Bourgois, Morphometric and physical investigations of segmental cortical bone autografts and allografts in canine ulnar defects, Clin. Orthop. Relat. Res. 282 (1992), 273–292. [7] W.F. Enneking, D.A. Campanacci, Retrieved human allografts: A clinicopathological study, J. Bone Joint Surg. Am. 83A(7) (2001), 971–986. [8] M. Kanayama, T. Hashimoto, K. Shigenobu, S. Yamane, T.W. Bauer and D. Togawa, A prospective randomized study of posterolateral lumbar fusion using osteogenic protein-1 (OP-1) versus local autograft with ceramic bone substitute: Emphasis of surgical exploration and histologic assessment, Spine 31 (2006), 1067–1074. [9] M.I. Alam, I. Asahina, K. Ohmamiuda, K. Takahashi, S. Yokota and S. Enomoto, Evaluation of ceramics composed of different hydroxyapatite to tricalcium phosphate ratios as carriers for rhBMP-2, Biomaterials 22 (2001), 1643–1651. [10] S.D. Cook and D.C. Rueger, Osteogenic protein-1: Biology and applications, Clin. Orthop. Relat. Res. 324 (1996), 29–38. [11] K. Hanada, J.E. Dennis and A.I. Caplan, Stimulatory effects of basic fibroblast growth factor and bone morphogenetic protein-2 on osteogenic differentiation of rat bone marrow-derived mesenchymal stem cells, J. Bone Miner. Res. 12 (1997), 1606–1614. [12] G.E. Pluhar, A.S. Turner, A.R. Pierce, C.A. Toth and D.L. Wheeler, A comparison of two biomaterial carriers for osteogenic protein-1 (BMP-7) in an ovine critical defect model, J. Bone Joint Surg. Br. 88 (2006), 960–966. [13] G.E. Friedlaender, Osteogenic protein-1 in treatment of tibial nonunions: Current status, Surg. Technol. Int. 13 (2004), 249–252. [14] M. Pecina, L.R. Giltaij and S. Vukicevic, Orthopaedic applications of osteogenic protein-1 (BMP-7), Int. Orthop. 25 (2001), 203–208. [15] R. Dimitriou, Z. Dahabreh, E. Katsoulis, S.J. Matthews, T. Branfoot and P.V. Giannoudis, Application of recombinant BMP-7 on persistent upper and lower limb non-unions, Injury 36 (Suppl. 4) (2005), S51–S59. [16] J.M. Wozney, Bone morphogenetic proteins, Prog. Growth Factor Res. 1 (1989), 267–280. [17] K. Kurashina, H. Kurita, Q. Wu, A. Ohtsuka and H. Kobayashi, Ectopic osteogenesis with biphasic ceramics of hydroxyapatite and tricalcium phosphate in rabbits, Biomaterials 23 (2002), 407–412. [18] K. Takaoka, H. Nakahara, H. Yoshikawa, K. Masuhara, T. Tsuda and K. Ono, Ectopic bone induction on and in porous hydroxyapatite combined with collagen and bone morphogenetic protein, Clin. Orthop. Relat. Res. 234 (1988), 250–254. [19] M. Itokazu, T. Sugiyama, T. Ohno, E. Wada and Y. Katagiri, Development of porous apatite ceramic for local delivery of chemotherapeutic agents, J. Biomed. Mater. Res. 39 (1998), 536–538. [20] P. Habibovic, H. Yuan, M. van den Doel, T.M. Sees, C.A. van Blitterswijk and K. de Groot, Relevance of osteoinductive biomaterials in critical-sized orthotopic defect, J. Orthop. Res. 24 (2006), 867–876. [21] M.R. Urist, Y.K. Huo, A.G. Brownell, W.M. Hohl, J. Buyske, A. Lietze, P. Tempst, M. Hunkapiller and R.J. DeLange, Purification of bovine bone morphogenetic protein by hydroxyapatite chromatography, Proc. Natl. Acad. Sci. USA 81 (1984), 371–375. [22] Y. Zhang and M. Zhang, Three-dimensional macroporous calcium phosphate bioceramics with nested chitosan sponges for load-bearing bone implants, J. Biomed. Mater. Res. 61 (2002), 1–8. [23] R.K. Korhonen, M.S. Laasanen, J. Toyras, J. Rieppo, J. Hirvonen, H.J. Helminen and J.S. Jurvelin, Comparison of the equilibrium response of articular cartilage in unconfined compression, confined compression and indentation, J. Biomech. 35 (2002), 903–909. [24] A.J. Schoutens, J. Artlet and J.W.M. Gardeniers, Haemodynamics of bone healing in a model stable fracture, in: Bone Circulation and Vascularization in Normal and Pathological Conditions, S.P.F. Hughes, ed., Plenum Press, New York, 1993, pp. 121–128. [25] M. Brookes and W.J. Revell, Blood Supply of Bone: Scientific Aspects, Butterworth & Co., London, 1971. [26] R.S. Taichman and S.G. Emerson, Human osteosarcoma cell lines MG-63 and SaOS-2 produce G-CSF and GM-CSF: identification and partial characterization of cell-associated isoforms, Exp. Hematol. 24 (1996), 509–517. [27] D.P. Lennon, S.E. Haynesworth, R.G. Young, J.E. Dennis and A.I. Caplan, A chemically defined medium supports in vitro proliferation and maintains the osteochondral potential of rat marrow-derived mesenchymal stem cells, Exp. Cell Res. 219 (1995), 211–222.


289

[28] J.E. Dennis, S.E. Haynesworth, R.G. Young and A.I. Caplan, Osteogenesis in marrow-derived mesenchymal cell porous ceramic composites transplanted subcutaneously: Effect of fibronectin and laminin on cell retention and rate of osteogenic expression, Cell Transplant. 1 (1992), 23–32. [29] P. Cassiede, J.E. Dennis, F. Ma and A.I. Caplan, Osteochondrogenic potential of marrow mesenchymal progenitor cells exposed to TGF-beta 1 or PDGF-BB as assayed in vivo and in vitro, J. Bone Miner. Res. 11 (1996), 1264–1273. [30] T.K. Sampath, J.C. Maliakal, P.V. Hauschka, W.K. Jones, H. Sasak, R.F. Tucker, K.H. White, J.E. Coughlin, M.M. Tucker, R.H. Pang et al., Recombinant human osteogenic protein-1 (hOP-1) induces new bone formation in vivo with a specific activity comparable with natural bovine osteogenic protein and stimulates osteoblast proliferation and differentiation in vitro, J. Biol. Chem. 267 (1992), 20352–20362. [31] C. Maniatopoulos, J. Sodek and A.H. Melcher, Bone formation in vitro by stromal cells obtained from bone marrow of young adult rats, Cell & Tissue Res. 254 (1988), 317–330. [32] I. Asahina, T.K. Sampath and P.V. Hauschka, Human osteogenic protein-1 induces chondroblastic, osteoblastic, and/or adipocytic differentiation of clonal murine target cells, Exp. Cell. Res. 222 (1996), 38–47. [33] G. Westergren-Thorsson, P.O. Onnervik, L.A. Fransson and A. Malmstrom, Proliferation of cultured fibroblasts is inhibited by L-iduronate-containing glycosaminoglycans, J. Cell Physiol. 147 (1991), 523–530. [34] S.L. Salkeld, L.P. Patron, R.L. Barrack and S.D. Cook, The effect of osteogenic protein-1 on the healing of segmental bone defects treated with autograft or allograft bone, J. Bone Joint Surg. Am. 83-A (2001), 803–816. [35] T.J. Blokhuis, F.C. den Boer, J.A. Bramer, J.M. Jenner, F.C. Bakker, P. Patka and H.J. Haarman, Biomechanical and histological aspects of fracture healing, stimulated with osteogenic protein-1, Biomaterials 22 (2001), 725–730. [36] I. Manjubala, M. Sivakumar, R.V. Sureshkumar and T.P. Sastry, Bioactivity and osseointegration study of calcium phosphate ceramic of different chemical composition, J. Biomed. Mater. Res. 63 (2002), 200–208. [37] M.R. Urist, A. Lietze and E. Dawson, Beta-tricalcium phosphate delivery system for bone morphogenetic protein, Clin. Orthop. Relat. Res. 187 (1984), 277–280. [38] P.K. Zysset, X.E. Guo, C.E. Hoffler, K.E. Moore and S.A. Goldstein, Elastic modulus and hardness of cortical and trabecular bone lamellae measured by nanoindentation in the human femur, J. Biomech. 32 (1999), 1005–1012. [39] S.C. Cowin, Bone Mechanics Handbook, CRC Press, New York, 2001. [40] T.M. Chu, D.G. Orton, S.J. Hollister, S.E. Feinberg and J.W. Halloran, Mechanical and in vivo performance of hydroxyapatite implants with controlled architectures, Biomaterials 23 (2002), 1283–1293. [41] N. Verdonschot, C.T. van Hal, B.W. Schreurs, P. Buma, R. Huiskes and T.J. Slooff, Time-dependent mechanical properties of HA/TCP particles in relation to morsellized bone grafts for use in impaction grafting, J. Biomed. Mater. Res. 58 (2001), 599–604. [42] R.H. Li and J.M. Wozney, Delivering on the promise of bone morphogenetic proteins, Trends Biotechnol. 19 (2001), 255–265. [43] H. Uludag, D. D’Augusta, R. Palmer, G. Timony and J. Wozney, Characterization of rhBMP-2 pharmacokinetics implanted with biomaterial carriers in the rat ectopic model, J. Biomed. Mater. Res. 46 (1999), 193–202.