A NOVEL OSTEOCHONDRAL COMPOSITE CONSISTING OF A SELF-ASSEMBLING PEPTIDE HYDROGEL AND 3D-PRINTED POLYCAPROLACTONE SCAFFOLD

+*Saatchi, S; *Kisiday, JD; *Wright, JW; *Griffith, LG; *DiMicco, MA; *Grodzinsky, AJ +*Massachusetts Institute of Technology, Cambridge, MA

sanaz@mit.edu INTRODUCTION: The production and long-term function of tissue-engineered osteochondral grafts requires that both the cartilage and bone portions functionally mimic and integrate with the surrounding host tissues. Hydrogels made of agarose, or self-assembling peptides [1,2] provide a 3D environment that maintains chondrocyte morphology and phenotype, and stimulates extracellular matrix (ECM) synthesis and accumulation. Poly- -caprolactone (PCL) is a bioresorbable, biocompatible polymer that enables chondrogenic and osteogenic cell attachment, migration, and development of a tissue-specific ECM [3]. Furthermore, the micro- and macroporosity of PCL scaffolds can be varied in a layer-by-layer fashion during manufacturing, using a rapid prototyping, 3D printing (3DP™) process [4]. To investigate the feasibility of using PCL/hydrogel composites as osteochondral graft materials, the objectives of this study were to determine the effects of the presence of a PCL sublayer, hydrogel composition, and PCL microarchitecture on the behavior of hydrogel-embedded chondrocytes, using biochemical, mechanical, and histological analyses. METHODS: PCL Scaffold Fabrication: PCL scaffolds (9mm diameter x 3mm thick) were fabricated using a Theriform™ 3DP™ machine with an interconnected network of 106-180µm pore size with 70% or 90% porosity. Chondrocyte Isolation, Casting, and Culture: PCL scaffolds were placed into 2.5% agarose molds (9mm inner diameter, 20mm outer diameter, 4mm thick). Bovine chondrocytes were isolated from the knee joint of 2-3 week old calves using sequential pronase and collagenase treatments. Casting solution consisting of chondrocytes resuspended in 10% sucrose, 5mM HEPES, and 0.36% w/v self-assembling peptide solution (sequence: -KLDLKLDLKLDL-, [1]) at 30x106 cells/mL was injected directly into and on top of the PCL scaffolds. Analogous composites were created using a chondrocyte-seeded agarose hydrogel layer (30x106 cells/mL in 2% agarose) atop a PCL scaffold. Chondrocyte-seeded hydrogels cast without an underlying PCL scaffold served as controls. Culture medium (high-glucose DMEM with 1% ITS and 0.2% FBS, [1]) was changed every 2 days. Samples were cultured for 6, 14, and 19 days. Biochemical, Mechanical, and Histological Characterization: 3mm (diam) cylindrical samples were obtained from both the hydrogel and PCL region at each time point. Biochemical characterization was performed via analysis of GAG content (DMMB dye binding), and protein and proteoglycan synthesis (radiolabeled 3H-proline and 35S-sulfate incorporation rates over 20 hours). Measurement of the radially confined equilibrium compressive modulus was performed by compression of the hydrogel portion of the construct (after separation from the PCL) to 5, 10, 12, 14, 16, and 18% compressive strain (30 sec per ramp, 3 min equilibration between ramps). PCL regions designated for histological characterization were fixed in 4% paraformaldehyde overnight, rinsed with a gradient of sucrose solutions, embedded in OCT, cryosectioned to 5µm, and stained with H&E and Toluidine Blue. Peptide regions were treated similarly, except the embedding was in paraffin. Statistical Analysis: The effects of the underlying PCL scaffold porosity and culture duration on chondrocyte behavior were evaluated by 2-way ANOVA. Data are mean SEM. RESULTS: Biochemistry (hydrogel portion): The GAG content of the hydrogel portions of both peptide/PCL and agarose/PCL composites increased with time (Fig. 1A), and the GAG content of composites was higher than that of control (non-composite) hydrogel sheets (p<0.05). In the peptide/PCL composites, there was no effect of porosity on GAG content until day 19, when the peptide atop the 70% porous scaffold contained approximately 40% more GAG per mg hydrogel than the 90% porous sample (p<0.05, Fig 1A). The opposite was true of the agarose/PCL composite, where a slight (p<0.05) increase in GAG content was observed in the 90% porous sample compared with the 70% porous sample. While there was no effect of time on proline incorporation rate in either the peptide/PCL or agarose/PCL scaffold, the presence of PCL stimulated proline incorporation in the agarose/PCL, independent of porosity (p<0.05). Biochemistry (PCL portion): PCL regions of composites displayed similar trends in ECM accumulation, protein synthesis, and proteoglycan synthesis as corresponding hydrogel regions, however values were approximately an order of magnitude less (data not shown). Histology: H&E staining of the peptide portion of a

peptide/PCL composite after 19 days in culture (Fig 2A) shows a homogeneous distribution of chondrocytes throughout the hydrogel, with relatively even proteoglycan distribution indicated by toluidine blue staining (Fig 2B). Staining of the corresponding 70% porous PCL scaffold with H&E (Fig 2C) shows infiltration of the chondrocyte-seeded peptide into pores of PCL (dark spots indicating chondrocytes, light regions are PCL), while toluidine blue staining of the 90% porous PCL scaffold (Fig 2D) shows a gradient of proteoglycan accumulation along cross-section of scaffold. Mechanical properties: The confined compression equilibrium modulus of the hydrogel portion was evaluated after 14 and 19 days of culture. For agarose/PCL constructs, the modulus was ~ 40 kPa, independent of PCL porosity. The peptide/PCL composites exhibited much higher equilibrium moduli, peaking at 80 kPa on day 14. At that time point, the modulus of the 90% porous peptide/PCL constructs was approximately 30% higher than that of the 70% constructs. decreased slightly from day 14 to day 19 of culture (p<0.05), independent of porosity, but there was no significant effect of time or porosity on equilibrium modulus in the peptide/PCL composites. DISCUSSION: Production of a composite construct comprised of a chondrocyte-seeded hydrogel atop a porous PCL scaffold did not impede ECM accumulation or biosynthesis rates of chondrocytes in hydrogel region of composite, highlighting the feasibility of such constructs for osteochondral grafting. Importantly, the application of an agarose or self-assembling peptide suspension, before polymerization, to the rigid PCL substrate allowed cellular infiltration and deposition of cartilage matrix components in the region underlying the bulk hydrogel. Changes in the PCL microarchitecture, which can be controlled during the 3DP™ manufacturing process, may influence the extent of cellular proliferation, and may enable the use of multiple cell populations within a single construct. Future studies may vary the cell source and the mechanical environment during culture to the enhance biomechanical properties of the composite.

Fig 1: Peptide/PCL composite immediately after casting. The peptide portion is stained with Coomassie blue for visualization, and overlays the white PCL scaffold. Surrounding the composite is the agarose mold used for casting.

Fig 2: (A) GAG content, (B) protein synthesis of chondrocyte-seeded peptide or agarose regions of composites.

A B DCAA BB DDCC

Fig 3: (A), (B) Chondrocyte-seeded hydrogel from 90% porous PCL composite, day 19, stained with H&E (A) or toluidine blue (B), both at 20X. (C) 70% porous PCL scaffold stained with H&E, 20X. (D) 90% porous PCL scaffold with gradient of proteoglycan accumulation along cross-section of scaffold, 10X.

REFERENCES: [1] Kisiday+ PNAS 99, 2002. [2] Kisiday+ JB 37, 2004. [3] Cao+ TEng 9, 2003. [4] Sherwood+ Biomaterials 23, 2002. ACKNOWLEDGEMENT: Cambridge-MIT Institute (CMI) and NSF Graduate Research Fellowship. We thank Tom Crowell for his technical assistance.

Peptide ControlPeptide 70% PCLPeptide 90% PCLAgaroseControlAgarose70% PCLAgarose90% PCL

All data presented as N=4-10, Mean ±SEM

Peptide ControlPeptide 70% PCLPeptide 90% PCLAgaroseControlAgarose70% PCLAgarose90% PCL

All data presented as N=4-10, Mean ±SEM

51st Annual Meeting of the Orthopaedic Research Society

Poster No: 1783