Citation
Characterization of Collagen-Bioactive Glass Composites for Treating Osteomyelitis

Material Information

Title:
Characterization of Collagen-Bioactive Glass Composites for Treating Osteomyelitis
Creator:
Goude, Melissa C.
Cooper, Scott
Brennan, Anthony
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Amines ( jstor )
Biomaterials ( jstor )
Bones ( jstor )
Collagens ( jstor )
Composite particles ( jstor )
Crosslinking ( jstor )
Emulsions ( jstor )
Gels ( jstor )
Hydrogels ( jstor )
Ions ( jstor )
Bioactive glasses
Gentamicin-PMMA chains
Osteomyelitis
Genre:
Undergraduate Honors Thesis

Notes

Abstract:
Osteomyelitis is an infection of the bone that is currently treated by gentamycin-impregnated poly(methyl methacrylate) beads. However, the beads’ lack of biodegradability and uncontrolled release of drug is motivation for an improved delivery system. In this study, composites of collagen hydrogel matrix and bioactive glass microspheres (BGMS) were tested for their bioactivity in a stimulated body fluid (SBF) as a possible alternative to the beads. More specifically, the objective was to determine whether the presence of the BGMS acted as a reinforcing phase to the soft collagen matrix by facilitating the mineralization of hydroxyapatite. Four compositions of the composite were prepared: no BGMS, 50wt% BGMS, 50wt% 3-aminopropyltriethoxysilane (APTES)-treated BGMS, and 50wt% 45S5 glass. However, results from mechanical testing showed no significant differences between the compressive moduli at 10-20, 20-30, and 40-50% strain. No significant increase was observed in the stress and strain at failure in the BGMS and 45S5 glass compositions. Heterogeneous distribution of the BGMS in collagen may have impeded the nucleation of hydroxyapatite crystals and prevented the enhancement of the composite moduli. ( en )
General Note:
Melissa C. Goude awarded Bachelor of Science in Materials Science and Engineering; Graduated June 21, 2011 summa cum laude. Major: Materials Science and Engineering
General Note:
College/School: College of Engineering
General Note:
Advisor: Anthony Brennan

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Melissa Goude, Scott Cooper and Anthony Brennan. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

Characterization of Collagen Bioactive Glass Composites for Treating Osteomyelitis Melissa C. Goude 1 Scott Cooper 1 and Anthony Brennan 1 1. Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611, USA Abstract Osteomyelitis is an infection of the bone that is currently treated by gentamycin impregnated poly(methyl methacrylate ) beads. However, the beads lack of biodegradability and uncontrolled release of drug is motivation for an improved delivery system. I n t his study, composites of collagen hydrogel matrix and bioactive glass microspheres (BGMS) were tested for their bioactivity in a stimulated body fluid (SBF) as a possible alternative to the beads More specifically, the objective was to determine whether t he presence of the BGMS acted as a reinforcing phase to the soft collagen matrix by facilitating the mineralization of hydroxyapatite. Four compositions of the composite were prepared: no BGMS, 50wt BGMS, 50wt 3 aminopropyltriethoxysilane ( APTES ) treated BGMS, and 50wt 45S5 glass. However, results from mechanical testing showed no significant differences between the compressive moduli at 10 20, 20 30, and 40 50 strain. No significant increase was observed in the stress and strain at failure in the BGMS and 45S5 glass compositions. Heterogeneous distribution of the BGMS in collagen may have impeded the nucleation of hydroxyapatite crystals and prevented the enhancement of the composite moduli. Introduction Osteomyelitis is the infection of bo ne of ten caused by bacteria such as methicillin resistant Staphylococcus aureus (MRSA). Local colonization of bacteria is promoted in the bone, resulting in the development of a biofilm, which is particularly difficult to treat with systemic antibiotics. Current therapies are costly and involve surgical extraction of necrotic bone and application of local antibiotic treatments. By the current method of treatment, drugs are loaded in poly(methyl methacrylate) ( PMMA ) beads, which are strung and implanted at the site of infection. However, the beads are not biodegradable and require a subsequent surgery for removal [1]. Bioactive glass is a special formulation of si lica glass that is resorbable and easily metabolized by the body. Specifically, bioactive glass has exhibited the ability to bond to bone and stimulate bone growth [2]. Sol gel derived bioactive glass microspheres can be prepared at low er temperature s comp ared to the melt derived particles, allowing incorporation of polymers and drugs. Previous studies by the Brennan group have successfully loaded an antibiotic drug, vancomycin, in bioglass microspheres ( BG MS) through a water in oil emulsion technique

PAGE 2

Bon e is a natural composite of organic matrix fortified by an inorganic phase C ollagen fibril s are helical structures composed of three amino acid chains in the following sequence: (Glycine X Y) n for which X and Y can be proline or hydroxyproline. The fibrils are organized into bundle s of banded fibers creating the collagen matrix. The interpenetrating mineral phase, consisting of hydroxyapatite platelets, nucleates and grows at the gaps between the aligned fibrils. The se spheritic clu sters strengthe n bone, while helping it maintain sufficient flexibility. Others have also shown that collagen hydrogels are osteoconductive and function as a drug delivery matrix [3]. Collagen gel can be easily prepared by a number of chemical crosslinking agents listed in Table 1. Specifically, we are interested in the ability of carbodiimides to crosslink collagen, which has been employed in generating an injec table implantable hydrogel [4]. Unlike gluateraldehyde and epoxy carbodiimide does not introduce cytotoxicit y to the system. In addition, the carbodiimide does not become incorporated into the crosslinked matrix during the catalysis of bonding between carboxyl and amine groups so they can continuously facilitate the reaction Our goal is to engineer a composite of bioactive glass microspheres in collagen hydrogel. The delivery system will allow for controlled release of vancomycin and foster growth of new bone tissue without the need for a second surgery. To act as an orthoped ic scaffold, the composite need s to exhibit an elastic modulus similar to that of bone ( 0. 1 G Pa 4.5 GPa [12] ). We are interested in investigating the extent that bioactive glass microspheres act as a reinforcing phase in the collagen gel which has a modulus documented as low as 4 6kPa [5] depending on the crosslinking mechanism. If mineralization of hydroxyapatite occurs when the composite is immersed in a simulated body fluid (SBF), additional crosslinking of the gel may further enhance its elastic modulus. Table 1 Collagen Crosslinking Mechanisms Chemical Mechanism Example Structure Cytotoxicity Modulus Tensile Strength lysyl oxidase [4] enzyme mediated oxidation of lysine group to form inter and intra fibril covalent crosslinks protein lysine 6 oxidase naturally occurs in the body ~ 75kPa [4] n/a aldehydes (gluateraldehyde, formaldehyde) [5, 6] reaction between lysine and asparagine or glutamine to form an imine, which crosslinks with tyrosine potentially toxic residue 4 6kPa [5] 7.9MPa [11] hexamethylene diisocyanate [2, 10] reaction with e amino group (lysine) to form urea type bond toxicity from degradation products 2 26MPa [10] 16.5MPa [10]

PAGE 3

poly(epoxy) compound [ 6 ] reaction with lysine group, slower than gluteraldehyde but more flexible gel moderate cytotoxicity 2 9MPa [11] 9.4MPa [11] c arbodiimides [3, 7] catalyzes bonding between carboxyl and amine groups without becoming incorporated themselves urea product n/a 100 200kPa [7] glycation (ribose, glucose, fructose) [8] nonenzymatic crosslinking of amine groups by reduction of sugars ribose non cytotoxic and implant friendly n/a 25kPa [8] T ransglutaminase [9] reaction between glutamine and lysine forms intermolecular or intramolecular amide crosslink calcium dependent enzyme found in the body [ 9 ] n/a 4 9x10 2 M P a [9] Materials and Methods SBF Preparation A solution of the following salts, KCl, K 2 HPO 4 3H 2 O, MgCl 2 6H 2 O, CaCl 2 Na 2 SO NaCl, NaHCO 3 and KCl was prepared near 37C using a protocol modified from Kokubo et al. [ 13, 14 ]. The solution was also buffered to pH=7.4 with tris(hydroxymethyl)aminomethane (TRIS). The resulting SBF contains ions in the concentration shown below in Table 2 The collagen gel composites were submerged in SBF for 0, 7, and 14 days. Table 2 Concentration of i ons in the SBF s olution Ion Na + K + Mg 2+ Ca 2+ Cl HCO 3 HPO 4 2 SO 4 2 Concentration (mM) 143 5 1.5 2.6 187.8 4.2 1 0.5 Microsphere Synthesis T he sol gel derived BGMS of composition 80 SiO 2 16 CaO 4 P 2 O 5 (77S) were prepared by Scott Cooper using a water in oil (w/o) emulsion technique from the pro tocol described by Park, et al. [15]. The oil phase, which is composed of octanol with 3wt Span 80 and 1.4 wt hydroxypropylcellulose, was prepared by adding the hydroxypropylcellulose around 80C and stirred for four hours Then, Span poly(epoxy) compound [ 6 ] reaction with lysine group, slower than gluteraldehyde but more flexible gel moderate cytotoxicity 2 9MPa [11] 9.4MPa [11] c arbodiimides [3, 7] catalyzes bonding between carboxyl and amine groups without becoming incorporated themselves urea product n/a 100 200kPa [7] glycation (ribose, glucose, fructose) [8] nonenzymatic crosslinking of amine groups by reduction of sugars ribose non cytotoxic and implant friendly n/a 25kPa [8] T ransglutaminase [9] reaction between glutamine and lysine forms intermolecular or intramolecular amide crosslink calcium dependent enzyme found in the body [ 9 ] n/a 4 9x10 2 M P a [9] Materials and Methods SBF Preparation A solution of the following salts, KCl, K 2 HPO 4 3H 2 O, MgCl 2 6H 2 O, CaCl 2 Na 2 SO NaCl, NaHCO 3 and KCl was prepared near 37C using a protocol modified from Kokubo et al. [ 13, 14 ]. The solution was also buffered to pH=7.4 with tris(hydroxymethyl)aminomethane (TRIS). The resulting SBF contains ions in the concentration shown below in Table 2 The collagen gel composites were submerged in SBF for 0, 7, and 14 days. Table 2 Concentration of i ons in the SBF s olution Ion Na + K + Mg 2+ Ca 2+ Cl HCO 3 HPO 4 2 SO 4 2 Concentration (mM) 143 5 1.5 2.6 187.8 4.2 1 0.5 Microsphere Synthesis T he sol gel derived BGMS of composition 80 SiO 2 16 CaO 4 P 2 O 5 (77S) were prepared by Scott Cooper using a water in oil (w/o) emulsion technique from the pro tocol described by Park, et al. [15]. The oil phase, which is composed of octanol with 3wt Span 80 and 1.4 wt hydroxypropylcellulose, was prepared by adding the hydroxypropylcellulose around 80C and stirred for four hours Then, Span poly(epoxy) compound [ 6 ] reaction with lysine group, slower than gluteraldehyde but more flexible gel moderate cytotoxicity 2 9MPa [11] 9.4MPa [11] c arbodiimides [3, 7] catalyzes bonding between carboxyl and amine groups without becoming incorporated themselves urea product n/a 100 200kPa [7] glycation (ribose, glucose, fructose) [8] nonenzymatic crosslinking of amine groups by reduction of sugars ribose non cytotoxic and implant friendly n/a 25kPa [8] T ransglutaminase [9] reaction between glutamine and lysine forms intermolecular or intramolecular amide crosslink calcium dependent enzyme found in the body [ 9 ] n/a 4 9x10 2 M P a [9] Materials and Methods SBF Preparation A solution of the following salts, KCl, K 2 HPO 4 3H 2 O, MgCl 2 6H 2 O, CaCl 2 Na 2 SO NaCl, NaHCO 3 and KCl was prepared near 37C using a protocol modified from Kokubo et al. [ 13, 14 ]. The solution was also buffered to pH=7.4 with tris(hydroxymethyl)aminomethane (TRIS). The resulting SBF contains ions in the concentration shown below in Table 2 The collagen gel composites were submerged in SBF for 0, 7, and 14 days. Table 2 Concentration of i ons in the SBF s olution Ion Na + K + Mg 2+ Ca 2+ Cl HCO 3 HPO 4 2 SO 4 2 Concentration (mM) 143 5 1.5 2.6 187.8 4.2 1 0.5 Microsphere Synthesis T he sol gel derived BGMS of composition 80 SiO 2 16 CaO 4 P 2 O 5 (77S) were prepared by Scott Cooper using a water in oil (w/o) emulsion technique from the pro tocol described by Park, et al. [15]. The oil phase, which is composed of octanol with 3wt Span 80 and 1.4 wt hydroxypropylcellulose, was prepared by adding the hydroxypropylcellulose around 80C and stirred for four hours Then, Span

PAGE 4

80 wa s added to the solution around 40C. Tetrae thoxysilane (TEOS) was introduced to the oil phase a nd hydrolyze d at the interface between oil and water. The aqueous phase was prepared by adding a catalyst, HNO 3 to 2.8M CaO in order to achieve pH=1 and p olyvinyl pyrrolidone (PVP) (40kg/mol) to produce a transparent, homogenous mixture. About 6.1mL of the aqueous phase wa s added dropwise to 46g of the oil phase to form the emulsion, which wa s stirred by a Caframo mechanical mixer at 500RPM for 30 minutes TEOS and triethlphosphate were added to the emulsion as well to attain a H 2 O:TEOS molar ratio of 4. The emulsion continued to stir at 20C for 24 hours. The BGMS were washed by centrifugation at 2000RPM, rinsed with 20mL of 100 ethanol and re dispersed with a vortex mixer three times. The BGMS were dried in a vacuum oven at 45 C fo r three hours until all ethanol was visibly removed. APTES Treatment 3 aminopropyltriethoxysilane (APTES), a silane coupling agent, was used to functionalize the BGMS with a primary amine. A mixture of 0.1g BGMS, 0.8mL of 95 ethanol and 0.2mL of APTES was shaken by hand at room temperature for 3 minutes and centrifuged for 5 minutes at 1500 RPM. The pellet was rinsed with 95 ethanol and suspended with a vortex mixer. Centrifugation was rep eated for another 5 minutes and dried in a convection oven at 100C for 45 minutes. The APTES treated BGMS were stored in a desiccators until use. 45S5 Glass Particle Preparation The 45S5 or standard, melt derived bioactive glass particles (approximately made by US Biomaterials and provided by the Mecholsky lab. Collagen Gel Preparation B ovine Type I collagen solution (Advanced Biomatrix PureCol 3mg/mL) was first c hilled and 6mL was c rosslinked physically using 0.1M NaOH neutralization solution in a cylindrical glass vial After incubation at 37C for at least two hours the hydro gels are then chemically cros slinked by a solution containing 1 (3 Dimethylaminopropyl) 3 ethylcarbodiimide hydrochloride (EDC) and N hydroxysuccinimide (NHS) at a 5:1 molar ratio The hydrogels were then incubated at 37C for 24 hours. Upon removal, the samples were rinsed with a 0.1M Na 2 HPO 4 neutralizing solution to remove any residual crosslinking solution. The prepared samples were s tored in phosphate buffered saline (PBS) at 37C before mechanical testing or characterization Four compositions of composites were prepared with the collagen gels: no BG MS, BGMS equal to 50wt of the dry collagen gel 50wt APTES treated BG MS, and 50wt 45S5 glass as seen in Figure 1 The BGMS and 45S4 glass particles were suspended in the 0.1M NaOH solution prior to crosslinking the collagen solution through sonication and physically drawing the solution up and down the micropipette All samples were in cubated at 37C and 5 CO 2

PAGE 5

Figure 2 Sample set up on the Texture Analyzer. Mechanical Testing Using the T exture A nalyzer (Stable Micro Systems Ltd. UK) compression testing was performed on the composite samples which w ere affixed with superglue to the stage as seen in Figure 2 Prior to testing, the samples were immersed in PBS at 37 C. The probe was programmed to compress samples at a strain rate of 0.1mm/sec The typical diameter of composite samples is about 12mm and height of 5mm A macro was developed with the Exponent sof tware (Stable Micro Systems Ltd., UK) to evaluate the compressive modulus of each sample at 10 20, 20 30, 40 50 strain and the strength and strain at rupture. Collagen (control) Collagen + 50wt BGMS Collagen + 50wt APTES Treated BGMS Collagen + 50wt 45S5 Glass Figure 1 Four compositions of the composite samples.

PAGE 6

Characterization Composite samples were freeze dried for characterization using x ray diffraction ( XRD ) and sputtered coated with Au/Pd for scanning electron microscopy ( SEM ) for 60 seconds at 38 mA by Scott Cooper The SEM was done with a 15 keV accelerating voltage 15mm working distance, and Condensor Lens setting = 11 in SE mode. Results The compressive moduli for composite samples were obtained for three strain s as shown in Figure 3A As seen in Figure 3B n o significant differences were observed in the moduli of day 0, 7, 14 samples for each composition at 10 20 strain (see Ap p endix A for ANOVA test results) Overall, the compressive modulus of samples at low strain ranged from 2kPa to 7kPa. Figure 3 All mechanical testing was performed at 0.1mm/sec strain rate. Day 0 (n=3), day 7 (n=3), day 14 (n=3), and day 14 repeat (n=4). A) Typical stress and strain curve from a composite sample. The compressive moduli are calculated at the three highlighted regi ons of strain. B) Compressive moduli for all composites at 10 20 strain. C) Compressive moduli for all composites at 20 30 strain. D) Compressive moduli for all composites at 40 50 strain.

PAGE 7

At 20 30 strain the compressive moduli for composites without MS and composites with 50 wt APTES treated BGMS did not change significantly (Appendix A) with extended immersion in SBF as shown in Figure 3C Increase in moduli was detected from day 0 to day 7 in samples containing 50 wt BGMS and 50 wt 45S5 glass, but not at day 14. The moduli at this strain ranged from 2kPa to16kPa. At 40 50 strain the compressive moduli of composite samples did not vary significantly between number of days in SBF or compositions. The moduli ranged from 10kPa to 35kPa. As displayed in Figure 4 no significant variations were observed in the strain and stress at failure for all composite samples. Figure 4 All mechanical testing was performed at 0.1mm/sec strain rate. Day 0 (n=3), day 7 (n=3), day 14 (n=3), and day 14 repeat (n=4). A) Typical stress and strain curve from a composite sample. The stress and strain corresponding to the highlighted point on the curve are taken as the point of rupture. B) Stress at rupture for all composite samples. C) Strain at rupture for all composite samples. A fibrillar morphology was observed in the control samples with SEM as shown in Figure 5 indicating that the collagen had successfully undergone physical gelation In the day 0 samples, patches of amorphous collagen are present near the clusters of BGMS and glass particles. However, the distributions of the BGMS and particles were non uniform and the cau liflower like morphology that is associated with mineralization of HA was absent in all samples. Also, dimples were seen on the surfaces of the BGMS.

PAGE 8

Figure 5 SEM images of day 0 composite samples taken at 15 keV accelerating voltage 15mm working distance, and Condensor Lens setting = 11 in SE mode. All s amples were sputter coated with Au/Pd for 60 seconds at 38 mA A) no BGMS. B) with 50wt BGMS C) 50wt APTES treated BGMS D) 45S5 glass particle The XRD spectra of day 14 samples shown in Figure 6 did not display peaks indicative of HA crystals The large peak in Figure 6A was identified as a Silicon peak (see Appendix B ). No precipitation of hydroxyapatite were detected visually in Figure 7 The fibrillar collagen network remained intact with sporadic integration of BGMS and particle aggregates Amorphous collagen formed near the BGMS and glass particles as well. Figure 6 (Cu A) The XRD spectra of day 14 samples (n=3). B) The XRD spectra of day 14 repeat samples (n=4). *

PAGE 9

Figure 7 SEM images of day 14 composite samples taken at 15 keV accelerating voltage 15mm working distance, and Condensor Lens setting = 11 in SE mode. All s amples were sputter coated with Au/Pd for 60 seconds at 38 mA A) no BGMS. B) 50wt BGMS C) 50wt APTES tr eated BGMS D) 45S5 glass particle Discussion Overall, the presence of BGMS APTES treated BGMS and 45S5 bioglass particles did not appear to reinforce the collagen gel sc affold significantly. Some webbing of collagen gel was observed in the vicinity of BGMS clusters, suggesting possible crosslinking Similar webbing morphology was found with APTES treated BGMS in sporadic clusters throughout the composite However, the additional crosslinking sites on the functionalized surfa ce produced by the APTES treatment did not appear to enhance crosslinking Samples that were immersed in the SBF for 14 days were expected to result in the highest extent of mineralization. However, no HA precipitates were observed visually or indirectly via mechanical testing in any sample, unlike a similar study from Eglin et al. HA were found to have nucleated on the surface of BGMS within three days of immersion in the SBF and no obstruct ion by the collagen network was reported [16]. Although the protocol included son ication and mechanical mixing to ensure a homogeneous suspension of BGMS and particles in the collagen solution before incubation the particles may have settled out before crosslinking is complete. Consequently, t he uneven distribution of particl es or BGMS in the collagen gel may have explained the large variance of mechanical properties in the samples especially at sites where amorphous collagen morphology were detected It may have also obstructed the ease with which the ions from the SBF could approach the site of the particles and nucleate into hydroxyapatite crystals

PAGE 10

The lack of mineralization may have also resulted from the premature removal of calcium ions before nu cleation. PVP, a reagent used to prepare the BGMS, is water so luble Howe ver, in the process of soaking the BGMS in the aqueous neutralizing solution may have led to the dissolution of PVP from the BGMS along and the porosity of the silica gel allow the ca lcium ions to leach out Once the collagen has crosslinked, the hydrogel composites were rinsed with 0.1M Na 2 HPO 4 solution. This procedure could potentially eliminate the presence of any residual calcium ions and prevented precipitation from occurring as well The dissolution of PVP can be seen from the dimples on the BGMS surfaces In addition, the concentration of ions in the SBF may not have been high enough to induce mineralization at a detectable level Scaffolds of crosslinked collagen were achieved in this study as evident in the fibrillar microstructure. However, scaffolds containing BGMS, APTES treated BGMS or 45S5 glass particles did not demonstrate an increase of mechanical property as hypothesized M ineralization of hydro xyapatite was not observed in the composites after immersion in SBF which may be explained by events that prevent ed calcium ion from fully nucleating Acknowledgement The author would like to thank Dr. Anthony Brennan for his mentorship. Much gratitude also goes out to Scott Cooper for his guidance and assistance throughout the project Assistance from the Brennan Lab is gratefully acknowledged. Projec t funding is contributed by the Margaret A. Ross Professorship References [1] Xie Z., Liu, X., Jia, W., Zhang, C., Huang, W., & Wang, Q. (2009). Treatment of osteomyelitis and repair of bone defect by degradable bioactive borate glass releasing vancomycin. Journal of Controlled Release 139 118 126. [2] Freiss W. (1998). Collagen biomaterial for drug delivery. European Journal of Pharmaceutics and Biopharmaceutics 45 113 136. [3] Park, S.N., Park, J.C., Kim, H.O., Song, M.J., & Suh, H. (2002). Characterization of porouscollagen/hyaluronic acid scaffold modi fied by 1 ethyl 3 (3 dimethylaminopropyl) carbodiimide cross linking. Biomaterials 23 1205 1212. [4] Elbjeirami, W.M., Yonter, E.O., Starcher, B.C., & West, J.L. (2003). Enhancing mechanical properties of tissue engineered constructs via lysyl oxidase cr osslinking activity. Journal of Biomedical Materials Research 66A 513 521. [5] Sheu, M.T., Huang, J.C., Yeh, G.C., & Ho, H.O. (2001) Characterization of collagen gel solutions and collagen matrices for cell culture. Biomaterials 22 1713 1719. [6] Miyat a, T., Taira, T., Noishiki, Y. (1992). Collagen engineering for biomaterial use. Clinical Materials 9 139 148. [7] Angele, P., Abke, J., Kujat, R., Faltermeier, H., Schumannm, D., Nerlich, M. et al. (2004) Influence of different collagen species on physi co chemical properties of crosslinked collagen matrices. Biomaterials 25 2831 2841.

PAGE 11

[8] Seliktar, D., Black, R.A., Vito, R.P., & Nerem, R.M. (2000) Dynamic mechanical conditioning of collagen gel blood vessel constructs induces remodeling in vitro. Annal s of Biomedical Engineering 28 351 362. [9] Chen, R.N., Ho, H.O., & Sheu, M.T. (2005) Characterization of collagen matrices crosslinked using microbial transglutaminase. Biomaterials 26 4229 4235. [10] Damink, L.H., Dijkstra, P.J., Van Luyn, J.A., & Van, P.B. (1995). Crosslinking of dermal sheep collagen using hexamethylene diisocyanate. Journal of Materials Science: Materials in Medicine 6 429 434. [11] Zeeman, R. (1998). Crosslinking of collagen based materials. Ph.D. dissertation, University of Twente, Enschede, The Netherlands. [12] Turner, C., Cowin, J., Rho, R., Ashman, J., & Rice, J. (1990). The fabric dependence of the orthotropic elastic constants of cancellous bone. J. Biomech. 23 549 561, 1990. [1 3 ] Kokubo, T., Takadama, H. (2006). How useful is SBF in predicting in vivo bone bioactivity? Biomaterials 27 (15), 2907 2915. [1 4 ] Kokubo, T., Kushitani, H., Sakka, S., Kitsugi, T., Yamamuro, T. (1990). Solutions Able to Reproduce Invivo Surface Structure Changes in Bioactive Glass Ceramic A W. Journal of Biomedical Materials Research 24 (6), 721 734. [15] Park J H, Oh C, Shin S I, Moon S K, Oh S G. (2003). Preparation of hollow silica microspheres in W/O emulsions with polymers Journal of Colloid and Interface Science 266(1):107 114. [16] Eglin D Maalheem S Livage J Coradin T (2006). In vitro apatite forming ability of type I collagen hydrogels containing bioactive glass and silica sol gel particles. J Mater Sci Mater Med 17(2):161 7.

PAGE 12

Appendix A. The 2 factor ANOVA test results indicating lack of statistical significance between composite moduli, failure stresses and strains for all compositions and SBF treatments. ANOVA (Strain to Failure) Source of Variation SS df MS F P value F crit Sample 110.8845 2 55.44227 0.22262 0.802054 3.402826 Columns 1192.786 3 397.5954 1.59648 0.216342 3.008787 Interaction 2728.888 6 454.8147 1.826235 0.136257 2.508189 Within 5977.079 24 249.045 Total 10009.64 35 ANOVA (Stress at Failure) Source of Variation SS df MS F P value F crit Sample 6.22E+09 2 3.11E+09 1.230487 0.309931 3.402826 Columns 2.96E+10 3 9.88E+09 3.907517 0.020964 3.008787 Interaction 5.44E+10 6 9.07E+09 3.587085 0.011117 2.508189 Within 6.07E+10 24 2.53E+09 Total 1.51E+11 35 ANOVA (10 20 Strain) Source of Variation SS df MS F P value F crit Sample 15901262 2 7950631 1.232941 0.309242 3.402826 Columns 21927549 3 7309183 1.133469 0.355443 3.008787 Interaction 42161940 6 7026990 1.089708 0.39663 2.508189 Within 1.55E+08 24 6448507 Total 2.35E+08 35 ANOVA (20 30 Strain) Source of Variation SS df MS F P value F crit Sample 73676346 2 36838173 2.957797 0.071083 3.402826 Columns 1.27E+08 3 42412041 3.405333 0.033856 3.008787 Interaction 1.79E+08 6 29793013 2.39213 0.059184 2.508189 Within 2.99E+08 24 12454596 Total 6.79E+08 35

PAGE 13

ANOVA (40 50 Strain) Source of Variation SS df MS F P value F crit Sample 2.77E+08 2 1.39E+08 1.465643 0.250871 3.402826 Columns 1.12E+09 3 3.73E+08 3.94883 0.020168 3.008787 Interaction 6.1E+08 6 1.02E+08 1.075375 0.404518 2.508189 Within 2.27E+09 24 94521812 Total 4.28E+09 35

PAGE 14

Appendix B Crystallographic data from the 1996 JCPDS International Centre for Diffraction Data Silicon (Si) 01 0787 Cubic, nm Int h k l 28.514 100 1 1 1 47.086 80 2 2 0 56.078 75 3 1 1 69.062 25 4 0 0 76.156 45 3 3 1 87.978 50 4 2 2 64.481 40 5 1 1 106.843 20 4 4 0 113.850 30 5 3 1 127.383 25 6 2 0 136.503 10 5 3 3 154.764 5 4 4 4 Hydroxyapatite Ca 10 (PO 4 ) 6 (OH) 2 Hexagonal, nm Int 25.900 40 28.705 8 32.081 100 34.224 8 39.707 16 42.438 4 43.954 4 46.829 20 49.540 20 53.594 16 64.237 8 72.100 8 76.882 4 87.978 8