Synthesis and properties of photocross-linked poly(propylene fumarate) scaffolds

Effect of biomaterial properties on bone healing in a rabbit tooth extraction socket model

In this work we sought to understand the effect of biomaterial properties upon healing bone tissue. We hypothesized that a hydrophilic polymer gel implanted into a bone tissue defect would impede the healing process owing to the biomaterial's prevention of protein adsorption and thus cell adhesion. To test this hypothesis, healing bone was investigated within a rabbit incisor extraction socket, a subcritical size bone defect that resists significant soft tissue invasion by virtue of its conformity. After removal of the incisor teeth, one tooth socket was left as an empty control, one was filled with crosslinked polymer networks formed from the hydrophobic polymer poly(propylene fumarate) (PPF), and one was filled with a hydrogel formed from the hydrophilic oligomer oligo(poly(ethylene glycol) fumarate) (OPF). At five different times (4 days as well as 1, 2, 4, and 8 weeks), jaw bone specimens containing the tooth sockets were removed. We analyzed bone healing by histomorphometrical analysis of hematoxylin and eosin stained sections as well as immunohistochemically stained sections. The proposed hypothesis, that a hydrophilic material would hinder bone healing, was supported by the histomorphometrical results. In addition, the immunohistochemical results reflect molecular signaling indicative of the early invasion of platelets, the vascularization of wound-healing tissue, the differentiation of migrating progenitor cells, and the formation and remodeling of bone tissue. Finally, the results emphasize the need to consider biomaterial properties and their differing effects upon endogenous growth factors, and thus bone healing, during the development of tissue engineering devices.

Effect of physiological temperature on the mechanical properties and network structure of biodegradable poly(propylene fumarate)-based networks

Poly(propylene fumarate) (PPF)-based networks have exhibited increases in mechanical properties during their initial stages of degradation. This study was designed to investigate whether physiological temperatures are the source of this reinforcing behavior by influencing the formation of additional crosslinks within the network. Utilizing a model PPF network formed with the crosslinking agent poly(propylene fumarate)-diacrylate (PPF-DA), cylindrical specimens were stored in an inert environment and conditioned at -20 and 37 degrees C while their mechanical properties and network structure were monitored over a six week period. The PPF/PPF-DA specimens exposed to physiological temperatures showed an increase in compressive modulus from 1674 +/- 88 to 2059 +/- 75 MPa. The double bond conversion improved as well, from 64 +/- 1 to 70 +/- 1%, indicating that crosslinks were being formed in the network. The additional reactivity occurred exclusively with unreacted fumarate bonds. PPF/PPF-DA networks stored at -20 degrees C showed no changes in mechanical properties; however, they increased when subsequently conditioned at 37 degrees C. The results were used to explain that PPF-based networks undergo a biphasic degradation behavior due to the competing hydrolytic degradation and thermal induced crosslinking. In addition, heat treating the networks at higher temperatures can be utilized as a means to further reinforce PPF-based materials.

Quantitative microcomputed tomography analysis of mineralization within three-dimensional scaffoldsin vitro

Synthetic and naturally derived scaffold biomaterials in combination with osteogenic cells or bioactive factors have the potential to serve as bone graft substitutes. Porous poly(l-lactide-co-dl-lactide) (PLDL) scaffolds with mechanical properties comparable to trabecular bone and an oriented, interconnected porosity designed to enhance internal mass transport were recently developed. In this study, PLDL scaffolds were seeded with rat calvarial or rat stromal cells and cultured up to 8 weeks in media containing osteogenic supplements. Cell-seeded human demineralized trabecular bone matrix (DTBM) scaffolds were included for comparison. All constructs were imaged weekly from 4 to 8 weeks using microcomputed tomography (micro-CT) to nondestructively quantify the amount and distribution of mineralized matrix formation. The total mineralized matrix volume increased with time in culture for all construct groups. DTBM constructs contained significantly more mineralized matrix than PLDL constructs. However, an analysis of the acellular DTBM scaffolds exposed to osteogenic media revealed partial remineralization of the demineralized matrix whereas no mineralization was detected in acellular PLDL scaffolds. Differences in mineral distribution were also evident with cell-mediated mineralization found throughout the PLDL constructs but localized to the periphery of the DTBM constructs for both cell types. Expression of bone marker genes indicating osteoblast differentiation was demonstrated in all groups at 8 weeks using a quantitative reverse transcription polymerase chain reaction. Osteocalcin expression was significantly higher for calvarial cell constructs compared to stromal cell constructs, regardless of the type of scaffold. This study demonstrated that micro-CT imaging may be used to nondestructively and quantitatively monitor mineralization within three-dimensional scaffolds in vitro. PLDL scaffolds with an oriented microarchitecture were shown to support cell attachment, differentiation, and cell-mediated mineralization comparable to natural DTBM scaffolds.

Quantitative analysis of interconnectivity of porous biodegradable scaffolds with micro-computed tomography

Pore interconnectivity within scaffolds is an important parameter influencing cell migration and tissue ingrowth needed to promote tissue regeneration. Methods for assessment of interconnectivity are usually qualitative, restricted to two-dimensional images, or are destructive. Microcomputed tomography nondestructively provides three-dimensional (3D) images of intact specimens at high spatial resolutions. We describe an image analysis technique for quantitative assessment of scaffold interconnectivity. Scaffolds were made via a particulate leaching process with 75%, 80%, 85%, and 88% volumetric porogen fractions. Specimens were scanned and resulting 3D, digital images were analyzed with a custom algorithm. A series of virtual, idealized scaffolds were also created for illustration of the algorithm's analysis approach and for its validation. The program calculated accessible void fractions over a range of minimum connection sizes. In real specimens, nearly 100% of the porous volume was connected with outside air for connections greater than or equal to 20 microm in their smallest dimension. In scaffolds made with 75% porogen, the accessible void fraction decreased to 78% if only those connections greater than or equal to 260 microm were considered. The relationship between accessible void fraction and connection size varied as a function of porogen content. The interconnectivity parameter described here may have implications for cell migration and tissue growth into scaffolds.

Evaluation of thermal- and photo-crosslinked biodegradable poly(propylene fumarate)-based networks

Biodegradable networks of poly(propylene fumarate) (PPF) and the crosslinking reagent poly(propylene fumarate)-diacrylate (PPF-DA) were prepared with thermal- and photo-initiator systems. Thermal-crosslinking was performed with benzoyl peroxide (BP), which is accelerated by N,N-dimethyl-p-toluidine (DMT) and enables injection and in situ polymerization. Photo-crosslinking was accomplished with bis(2,4,6-trimethylbenzoyl) phenylphosphine oxide (BAPO), which is activated by long-wavelength UV light and facilitates material processing with rapid manufacturing techniques, such as stereolithography. Networks were evaluated to assess the effects of the initiators and the PPF/PPF-DA double bond ratio on the mechanical properties. Regardless of the initiator system, the compressive properties of the PPF/PPF-DA networks increased as the double bond ratio decreased from 2 to 0.5. BAPO/UV-initiated networks were significantly stronger than those formed with BP/DMT. The compressive modulus of the photo- and thermal-crosslinked PPF/PPF-DA networks ranged from 310 +/- 25 to 1270 +/- 286 MPa and 75 +/- 8 to 332 +/- 89 MPa, respectively. The corresponding fracture strengths varied from 58 +/- 7 to 129 +/- 17 MPa and 31 +/- 13 to 105 +/- 12 MPa. The mechanical properties were not affected by the initiator concentration. Characterization of the network structures indicated that BAPO was a more efficient initiator for the crosslinking of PPF/PPF-DA, achieving a higher double bond conversion and crosslinking density than its BP counterpart. Estimated average molecular weights between crosslinks (Mc) confirmed the effects of the initiators and PPF/PPF-DA double bond ratio on the mechanical properties. This work demonstrates the capability to control the properties of PPF/PPF-DA networks as well as their versatility to be used as an injectable material or a prefabricated implant.

Photoinitiated cross-linking of the biodegradable polyester poly(propylene fumarate). Part I. Determination of network structure

In this work, we investigated the mechanism involved in the photoinitiated cross-linking of the polyester poly(propylene fumarate) (PPF) using the initiator bis(2,4,6-trimethylbenzoyl) phenylphosphine oxide (BAPO). It was hypothesized that BAPO has the ability to cross-link PPF into solid polymer networks, without the use of a cross-linking monomer, because two pairs of radicals, both involving a fast adding phosphinoyl radical, were formed upon UV irradiation of BAPO. Spectroscopic investigation first confirmed the addition of BAPO derived radicals to the PPF olefin. Investigations of fumarate conversion and bulk network properties were then undertaken, using the BAPO initiator and a monoacylphosphine oxide (MAPO) initiator which contains a single photolabile bond. Results show that a single BAPO phosphinoyl radical was primarily responsible for the formation of a highly cross-linked PPF network and the additional radical pair which may be formed does not dramatically alter fumarate conversion or bulk network properties. From these results, the network structure of BAPO initiated, photo-cross-linked PPF may be deduced. Finally, this study demonstrates a method for inferring cross-linked network structures by contrasting properties of bulk materials formed from similar cross-linking initiators.

Photoinitiated cross-linking of the biodegradable polyester poly(propylene fumarate). Part II. In vitro degradation

This study investigated the in vitro degradation of both solid PPF networks and porous PPF scaffolds formed by photoinitiated cross-linking of PPF polymer chains. Three formulations of scaffolds of differing porosity and pore size were constructed by varying porogen size and content. The effects of pore size and pore volume on scaffold mass, geometry, porosity, mechanical properties, and water absorption were then examined. Throughout the study, the solid networks and porous scaffolds exhibited continual mass loss and slight change in length. Porogen content appeared to have the greatest effect upon physical degradation. For example, scaffolds initially fabricated with 80 wt % porogen content lost approximately 30% of their initial PPF content after 32 weeks of degradation, whereas scaffolds fabricated with 70 wt % porogen content lost approximately 18% after 32 weeks of degradation. For all scaffold formulations, water absorption capacity, porosity, and compressive modulus were maintained at constant values following porogen leaching. These results indicate the potential of photo-cross-linked PPF scaffolds in tissue engineering applications which require maintenance of scaffold structure, strength, and porosity during the initial stages of degradation.

Mechanical evaluation of a porous bone graft substitute based on poly(propylene glycol-co-fumaric acid)

A porous, resorbable polymer composite based on poly(propylene glycol-co-fumaric acid) (PPF) was mechanically evaluated in vitro for use as a bone graft substitute and fracture fixative. The test material created a dynamic system capable of initially providing mechanical integrity to bony voids and a degradative mechanism for ingrowth by native bone. The unsaturated polymer, PPF, was crosslinked in the presence of effervescent agents to yield a porous microstructure upon curing. An in vitro degradation study first assessed the temporal mechanical properties of the test material. This research was followed by an ex vivo study using a long-bone osteotomy model to characterize the mechanics of fixation. Results showed the initial compressive strength of the cross-linked PPF system was comparable to cancellous bone. The rate of strength loss was commensurate with the predicted mechanical recovery of healing bone with analogous results in a composite that comprised also 25% (by weight) autograft. Mechanical testing in the long-bone model demonstrated that PPF-based bone-graft substitute increased the flexural strength of K-wire stabilized osteotomies. These results suggest that this type of bone graft substitute may have clinical utility in the stabilization of complex tubular bone fractures.

Poly(propylene fumarate) and poly(DL-lactic-co-glycolic acid) as scaffold materials for solid and foam-coated composite tissue-engineered constructs for cranial reconstruction

This pilot study investigates the osseointegration of four types of critical-size (1.5-cm diameter) rabbit cranial defect (n = 35) bone graft scaffolds. The first is a solid poly(propylene fumarate)/beta-tricalcium phosphate(PPF/beta-TCP) disk; the three remaining constructs contain a PPF/beta-TCP core coated with a 1-mm resorptive porous foam layer of PPF or PLGA [poly(DL-lactic-co-glycolic acid)], and bone marrow. Animals were killed at 6, 12, and 20 weeks. There was no evidence of a foreign body inflammatory response at any time during the study. Histomorphometric analyses of new bone formation sorted lineal and areal measures of new bone into three cranial layers (i.e., external, middle, and internal). Statistical analyses revealed significantly more bone in the PLGA foam-coated constructs than in the PPF foam-coated constructs (p < 0.03). No implant fixation was used; there is no strength at time 0. Twenty percent of all explants were tested for incorporation strength with a one-point "push-in" test, and failure ranged from 8.3 to 34.7 lb. The results of this study support the use of PPF as a biocompatible material that provides both a structural and osteogenic substrate for the repair of cranial defects.

An initial investigation of photocurable three-dimensional lactic acid based scaffolds in a critical-sized cranial defect

Degradable polymer networks formed by the photoinitiated polymerization of multifunctional monomers have great potential as in situ forming materials, especially for bone tissue engineering. In this study, one specific chemistry was analyzed with respect to bone formation in a critical-sized defect model with and without adsorbed osteoinductive growth factors present. The scaffolds degraded in approximately 8 months and possessed an elastic modulus similar to that of trabecular bone. A porous scaffold fabricated with approximately 80% porosity and pore diameters ranging from 45 to 150 mm was implanted in a critical-sized cranial defect in rats. When implanted alone, the scaffolds were filled primarily with fibrous tissue after 9 weeks with only mild inflammation at the defect site. When the scaffolds released osteoinductive growth factors, statistically more bone filled the scaffold. For instance, 65.8+/-9.4% (n=5) of the defects were filled with radiopaque tissue in the osteoinductive releasing scaffolds, whereas only 24.2+/-7.4% (n=5) of the defects were filled in the untreated defects 9 weeks after implantation. These results illustrate not only the benefits of delivering osteoinductive factors when developing synthetic bone grafts, but the potential of these materials for supporting the infiltration and development of bone in large defects.

Controlled release of an osteogenic peptide from injectable biodegradable polymeric composites

Poly(D,L-lactic-co-glycolic acid)/poly(ethylene glycol) (PLGA/PEG) blend microparticles loaded with the osteogenic peptide TP508 were added to a mixture of poly(propylene fumarate) (PPF), poly(propylene fumarate)-diacrylate (PPF-DA), and sodium chloride (NaCl) for the fabrication of PPF composite scaffolds that could allow for tissue ingrowth as well as for the controlled release of TP508 when implanted in an orthopedic defect site. In this study, PPF composites were fabricated and the in vitro release kinetics of TP508 were determined. TP508 loading within the PLGA/PEG microparticles, PEG content within the PLGA/PEG microparticles, the microparticle content of the PPF composite polymer component, and the leachable porogen initial mass percent of the PPF composites were varied according to a fractional factorial design and the effect of each variable on the release kinetics was determined for up to 28 days. Each composite formulation released TP508 with a unique release profile. The initial release (release through day 1) of the PLGA/PEG microparticles was reduced upon inclusion in the PPF composite formulations. Day 1 normalized cumulative mass release from PPF composites ranged from 0.14+/-0.01 to 0.41+/-0.01, whereas the release from PLGA/PEG microparticles ranged from 0.31+/-0.02 to 0.58+/-0.01. After 28 days, PPF composites released 53+/-4% to 86+/-2% of the entrapped peptide resulting in cumulative mass releases ranging from 0.14+/-0.01 microg TP508/mm(3) scaffold to 2.46+/-0.05 microg TP508/mm(3) scaffold. The results presented here demonstrate that PPF composites can be used for the controlled release of TP508 and that alterations in the composite's composition can lead to modulation of the TP508 release kinetics. These composites can be used to explore the effects varied release kinetics and dosages on the formation of bone in vivo.

Photocrosslinking characteristics and mechanical properties of diethyl fumarate/poly(propylene fumarate) biomaterials

The development of tissue engineered materials for the treatment of large bone defects would provide attractive alternatives to the autografts, allografts, non-degradable polymers, ceramics, and metals that are currently used in clinical settings. To this end, poly(propylene fumarate) (PPF), a viscous polyester synthesized from diethyl fumarate (DEF), has been studied for use as an engineered bone graft. We have investigated the photocrosslinking of PPF dissolved in its precursor, DEF, using the photoinitiator bis(2,4,6-trimethylbenzoyl) phenylphosphine oxide (BAPO) and low levels of ultraviolet light exposure. A three factor, 2 x 2 x 4 factorial design was developed, studying the effects of PPF number average molecular weight, BAPO initiator content, and DEF content upon photocrosslinking characteristics and mechanical properties. Uncured DEF/PPF solution viscosity fell over three orders of magnitude as DEF content was increased from 0% to 75%. The exothermic photocrosslinking reaction released low levels of heat, with no more than 160J/g released from any formulation tested. As a result, the maximum photocrosslinking temperature remained below 47 degrees C for all samples. Both sol fraction and swelling degree generally increased with increasing DEF content. Compressive mechanical properties were within the range of trabecular bone, with the strongest samples possessing an elastic modulus of 195.3 +/- 17.5 MPa and a fracture strength of 68.8 +/- 9.4MPa. Finally, the results indicate that PPF crosslinking was facilitated at low DEF precursor concentrations, but hindered at higher precursor concentrations. These novel DEF/PPF solutions may be preferred over pure PPF as the basis for an engineered bone graft because they (1) exhibit reduced viscosity and thus are easily handled, (2) form polymer networks with compressive strength at fracture suitable for consideration for trabecular bone replacement, and (3) may be readily fabricated into solids with a wide range of structures.

In vitro degradation and fracture toughness of multilayered porous poly(propylene fumarate)/beta-tricalcium phosphate scaffolds

This study investigated the in vitro degradation of poly(propylene fumarate)/beta-tricalcium phosphate (PPF/beta-TCP) scaffolds in pH 7.4 phosphate-buffered saline at 37 degrees C. Scaffold design consisted of three layers: two solid layers about a central layer of porous PPF foam. Solid PPF with molecular weights of 810 and 1450 Da was crosslinked under UV light. PPF foam was prepared by a photocrosslinking, porogen-leaching method with an initial porogen content of 80 wt % and two sizes, 150-300 and 300-500 microm. Comparison of initial and residual weights demonstrated a 14.3 +/- 2.0% loss of mass at 3 weeks and a 16.6 +/- 1.8% loss of mass at 6 weeks. Observed pH values for all constructs remained stable (7.15-7.40) throughout the 3 to 6 weeks. Scanning electron micrographs of these scaffolds revealed some loss of foam material between 3 and 6 weeks; however, foam microarchitecture was intact. Solid PPF fracture toughness was tested for high and low molecular weight PPF, 0.376 +/- 0.004 and 0.134 +/- 0.015 MPa(m)1/2, respectively. These values are roughly one magnitude less than human cortical bone.

Bone formation in transforming growth factor beta-1-coated porous poly(propylene fumarate) scaffolds

This study determined the bone growth into pretreated poly(propylene fumarate) (PPF) scaffolds implanted into a subcritical size, rabbit cranial defect. PPF scaffolds were constructed by using a photocrosslinking-porogen leaching technique. These scaffolds were then either prewetted (PPF-Pw), treated with RF glow-discharge (PPF-Gd), coated with fibronectin (PPF-Fn), or coated with rhTGF-beta1 (PPF-TGF-beta1). One of each scaffold type was then placed into the cranium of nine rabbits. The rabbits were sacrificed after 8 weeks, and the scaffolds were retrieved for histological analysis. The most bone formation was present in the PPF-TGF-beta1 implants; the newly formed bone had a trabecular appearance together with bone marrow-like tissue. Little or no bone formation was observed in implants without rhTGF-beta1. These histological findings were confirmed by image analysis. Bone surface area, bone area percentage, pore fill percentage, and pore area percentage were significantly higher in the rhTGF-beta1-coated implants than in the noncoated implants. No statistical difference was seen between the PPF-Fn, PPF-Pw, or PPF-Gd scaffolds for these parameters. Quadruple fluorochrome labeling showed that in PPF-TGF-beta1 implants bone formation mainly started in the interior of a pore and proceeded toward the scaffold. We conclude that (a) PPF-TGF-beta1 scaffolds can indeed adequately induce bone formation in porous PPF, and (b) PPF scaffolds prepared by the photocrosslinking-porogen leaching technique are good candidates for the creation of bone graft substitutes.

Soft and hard tissue response to photocrosslinked poly(propylene fumarate) scaffolds in a rabbit model

The treatment of large cranial defects may be greatly improved by the development of precisely formed bone tissue engineering scaffolds. Such scaffolds could be constructed by using UV laser stereolithography to photocrosslink a linear, biodegradable polymer into a three-dimensional implant. We have previously presented a method to photocrosslink the biodegradable polyester, poly(propylene fumarate) (PPF). To ensure the safety and effectiveness of this technique, the soft and hard tissue response to photocrosslinked PPF scaffolds of different pore morphologies was investigated. Four classes of photocrosslinked PPF scaffolds, constructed with differing porosities (57-75%) and pore sizes (300-500 or 600-800 microm), were implanted both subcutaneously and in 6.3-mm-diameter cranial defects in a rabbit model. The rabbits were sacrificed at 2 and 8 weeks, and the implants were analyzed by light microscopy, histological scoring analysis, and histomorphometric analysis. Results showed the PPF scaffolds elicit a mild tissue response in both soft and hard tissues. Inflammatory cells, vascularization, and connective tissue were observed at 2 weeks; a decrease in inflammatory cell density and a more organized connective tissue were observed at 8 weeks. Scaffold porosity and scaffold pore size were not found to significantly affect the observed tissue response. Evidence of scaffold surface degradation was noted both by histology and histomorphometric analysis. Bone ingrowth in PPF scaffolds implanted into cranial defects was <3% of the defect area. The results indicate that photocrosslinked PPF scaffolds are biocompatible in both soft and hard tissues and thus may be an attractive platform for bone tissue engineering.