Delivery of osteogenic factors in 3D-printed scaffolds

Osteoinductive cues are pivotal for bone regeneration. Many growth factors (i.e., proteins and/or peptides that promote cell growth and tissue healing) such as vascular endothelial growth factor (VEGF), fibroblast growth factors, bone morphogenetic proteins (BMPs), parathyroid hormone, and platelet-derived growth factor have been used in bone TE. Delivery of growth factors in 3D-printed bone scaffolds can be enabled during or after the printing process. Several current printing technologies allow the deposition of proteins and peptides during printing.^{Reference Lee, Kang, Lee, Kim, Lee and Cho19}

Alternatively, osteogenic factors can be delivered in previously printed scaffolds, either on the surface or within micropores/channels. Osteogenic factors are typically aimed toward enhanced cell recruitment, proliferation and differentiation, and angiogenesis.^{Reference Fedorovich, Alblas, Hennink, Oner and Dhert17} In a 3D-printed BMP-2-loaded model to repair critical rabbit segmental bone defects, sustained release of BMP-2 accelerated new bone formation without significant inflammatory response, whereas burst released (rapid release of large dose) BMP-2 failed to repair bone defects due to inflammatory response, with burst release defined as rapid unloading of encapsulated factors.^{Reference Shim, Kim, Park, Kundu, Kim, Kang and Cho20,Reference Poldervaart, Wang, van der Stok, Weinans, Leeuwenburgh, Oner, Dhert and Alblas21} Timely vascularization is essential for bone regeneration. Control released (sustained release over time) VEGF in 3D-printed scaffolds led to a significant increase in vessel formation after subcutaneous implantation (underneath the skin) in nude mice for one week, indicating that prolonged release of VEGF promotes vascularization.^{Reference Poldervaart, Gremmels, van Deventer, Fledderus, Oner, Verhaar, Dhert and Alblas22}

Development of osteoinductive 3D scaffolds

Osteoinductive materials are highly desirable for bone regeneration. One of the difficulties in demonstrating osteoinductivity versus osteoconductivity is that these two terms polarize a continuum of osteogenesis. BMP-2 clearly is osteoinductive. However, few biomaterials can achieve the same level of osteoinductivity as BMP-2. Nonetheless, several 3D-printed biomaterials have shown substantial promise for various levels of osteoinductivity. Three-dimensional-printed hydroxyapatite (HA) and tricalcium phosphate (TCP) ceramic scaffolds by projection-based microstereolithography have been shown to improve bone regeneration by accelerated release of calcium ions.^{Reference Seol, Park, Jung, Jang, Girdhari, Kim and Cho23}

The osteoinductive properties of printed scaffolds can also be enhanced by altering printing material ingredients. The combination of SiO₂ and ZnO dopants in 3D-printed β-TCP scaffolds increased the capacity for early bone formation and neovascularization in bicortical (penetrating the opposite bone cortices) femur defects in a murine model for up to 16 weeks. Neovascularization was found to be up to three times higher in combination of SiO₂ and ZnO dopants in TCP compared to TCP alone.^{Reference Fielding and Bose24} Mechanical interlocking between the implant and host tissue was stronger than the force it took to fracture the bone after four weeks.^{Reference Fielding and Bose24} Similarly, the addition of SrO and MgO in 3D-printed TCP scaffolds with multiscale porosity also contributed to early healing through accelerated osteogenesis in a rat distal femoral defect model.^{Reference Tarafder, Davies, Bandyopadhyay and Bose25}

Scaffolds used for bioprinting can be modified to improve the characteristics of osteoconductivity and osseointegration. Enlargement of surface area via incorporation of nanostructured pores in 3D-printed poly-ε-caprolactone (PCL) scaffolds embedded with nanoporous PCL into a fusion-deposition model processes a thermoplastic filament to a semi-liquid state and extrudes the material across computer-controlled paths to build a given anatomic structure layer-upon-layer. This fusion-deposition model improved its osteogenic potential in repairing calvarial defects (bony defect in skull bones) in pigs.^{Reference Jensen, Rolfing, Svend Le, Kristiansen, Nygaard, Hokland, Bendtsen, Kassem, Lysdahl and Bunger26} Biodegradation of 3D scaffolds also influences in vivo bone formation by cell-material interactions. Rapidly degradable poly(hydroxymethylglycolide-co-ε-caprolactone) scaffolds obtained by fiber deposition (melt-plotting) showed biocompatibility to the extent of tissue-scaffold interactions and vascularization after four and 12 weeks following subcutaneous implantation in mice.^{Reference Seyednejad, Gawlitta, Kuiper, de Bruin, van Nostrum, Vermonden, Dhert and Hennink27}

A scaffold’s internal architecture can influence the outcome of bone regeneration. Porosity characteristics such as pore size, volume fraction, percentage porosity, and pore shape are all important for cell adhesion, viability, and metabolism in scaffolds.^{Reference Sicchieri, Crippa, de Oliveira, Beloti and Rosa28} A major concern regarding porous scaffolds is the compromise in mechanical strength. By introducing microwave sintering, the mechanical strength of 3D-printed β-TCP scaffolds is enhanced to facilitate osteogenesis in repairing femoral defects in rats.^{Reference Tarafder, Balla and Davies29} Higher compressive strength was obtained when calcium phosphate powder was blended with a 3D-printed calcium sulfate (CaSO₄)-based powder.^{Reference Zhou, Buchanan, Mitchell and Dunne30} Critical size murine femoral defects (defects of certain size beyond which self-repair is not possible) were repaired by optimization of material parameters for 3D-printed calcium phosphate scaffolds, and enhanced material properties were obtained by volumetric collagen incorporation via inkjet printing.^{Reference Inzana, Olvera, Fuller, Kelly, Graeve, Schwarz, Kates and Awad31}

Cell delivery in bioprinted bone scaffolds

Currently, the common belief is that critical size bone defects cannot heal without cell delivery.^{Reference Mao, Stosich, Moioli, Lee, Fu, Bastian, Eisig, Zemnick, Ascherman, Wu, Rohde and Ahn32} Anatomically shaped PCL scaffolds with varying internal porosities were used to support the induction of human adipose-derived stem cells to form vascularized bone to repair mandibular and maxillary bone defects.^{Reference Temple, Hutton, Hung, Huri, Cook, Kondragunta, Jia and Grayson33} When 3D-printed highly porous biphasic calcium phosphate bone implants were ectopically implanted (i.e., in a location that is not anatomically the same) in the back of rats, osteogenic cells tended to induce bone formation mainly in the central region of the scaffolds. A combination of osteoblasts and BMP-2 synergistically yielded better bone formation by histomorphometric analysis (volumetric analysis on sections for microscopic structure studies of tissue).^{Reference Strobel, Rath, Maier, Beier, Arkudas, Greil, Horch and Kneser34} Spatially organized, functional osteogenic and endothelial progenitor cells (cells that form blood vessel lining) in printed grafts were observed after porous constructs containing the two cells were printed and subcutaneously implanted in immunodeficient mice.^{Reference Fedorovich, Wijnberg, Dhert and Alblas35} CD117+ hBMSCs (stem cell growth factor receptor and human bone marrow mesenchymal stem cells) were dispersed in different hydrogel mixtures and applied in a 3D-printed β-TCP/PLGA (tricalcium phosphate/poly(lactic-co-glycolic acid) scaffold.^{Reference Weinand, Gupta, Weinberg, Madisch, Neville, Jupiter and Vacanti36} Constructs were implanted subcutaneously into nude mice for six weeks. Biomechanical stiffness, radiological densities, and bone ECM accumulation were significantly enhanced at six weeks.^{Reference Weinand, Gupta, Weinberg, Madisch, Neville, Jupiter and Vacanti36}

Ectopic versus orthotopic environment

Many bone regeneration studies have been conducted in an ectopic environment by subcutaneous implantation in the dorsal region. Whereas ectopic implantation is appropriate for initial screening, an orthotopic environment is preferable or is required to ascertain bone repair. Femoral and calvarial defects are common models for bone repair studies using bioprinting strategies.^{Reference Tarafder, Davies, Bandyopadhyay and Bose25,Reference Inzana, Olvera, Fuller, Kelly, Graeve, Schwarz, Kates and Awad31} Bioprinting technology was attempted for repairing bone defects in mouse calvaria. Nano-HA was printed to repair bone defects using a 4 mm diameter trephine.^{Reference Keriquel, Guillemot, Arnault, Guillotin, Miraux, Amedee, Fricain and Catros37} Bone regeneration of critical-sized segmental diaphyseal defects in rabbits was allowed by long-term delivery of BMP-2 in collagen solution in printed PCL/PLGA scaffolds.^{Reference Shim, Kim, Park, Kundu, Kim, Kang and Cho20} Porcine calvaria defects were repaired by 3D-printed PCL scaffolds embedded with nanoporous PCL.^{Reference Jensen, Rolfing, Svend Le, Kristiansen, Nygaard, Hokland, Bendtsen, Kassem, Lysdahl and Bunger26}

Biomechanical properties of 3D-printed cartilage

Three-dimensional bioprinting technology is applicable to cartilage regeneration not only due to its ability to reproduce both patient-specific grafts and complex 3D-zonal ECM architecture, but also due to its integration of properties of various biomaterials. Recently, a hybrid inkjet printing system using off-the-shelf components was reported to produce cartilage constructs with improved biological and mechanical properties. Electrospun poly(caprolactone) fibers were alternated with inkjet-printed rabbit elastic chondrocytes (mature cells) suspended in a fibrin-collagen hydrogel. The fabricated constructs formed ectopic cartilage-like tissue subcutaneously in mice with deposition of Type II collagen, which is specific to cartilage, and glycosaminoglycans, and improvement of mechanical properties over alginate or fibrin-collagen gels alone.^{Reference Xu, Binder, Albanna, Dice, Zhao, Yoo and Atala40} The feasibility of fabricating anatomic cartilage structures was shown by delivering chondrocytes and poly(ethylene glycol) dimethacrylate to precise target locations in 3D for mimicking different zones of articular cartilage. The interface failure strength was enhanced as assessed by push-out testing.^{Reference Cui, Breitenkamp, Finn, Lotz and D’Lima41} However, integration of cartilage grafts with surrounding native articular cartilage in long-term preclinical or clinical studies remains a challenge, regardless of the use of 3D-printed scaffolds or non-3D-printed scaffolds.

Cell delivery in 3D-printed scaffolds for cartilage regeneration

Ectopic tissue formation in anatomically-shaped human tibial condyle scaffolds were seeded with human bone marrow mesenchymal stem/stromal cells.^{Reference Cui, Breitenkamp, Finn, Lotz and D’Lima41} A composite of poly-ε-caprolactone and HA was fabricated and seeded with human bone marrow stem cells (hMSCs) with or without osteogenic differentiation. A 1 mm overlaying layer of poly(ethylene glycol)-based hydrogel encapsulating hMSCs or hMSC-derived chondrocytes was molded into anatomic shape. After six weeks of subcutaneous implantation in athymic rats, hMSCs generated substantially more angiogenesis, whereas the hMSC-derived progenies yielded more mineralized tissue in microchannels and glycosaminoglycan matrix in the cartilage-like layer.^{Reference Lee, Marion, Hollister and Mao42} Recently, Fedorovich et al. fabricated cell-laden, heterogeneous alginate hydrogel scaffolds of varying porosity and elastic modulus by the 3D fiber deposition technique, and encapsulated human chondrocytes and osteogenic progenitors in alginate hydrogel to yield centimeter-scale intricate hydrogel scaffolds with high cell viability with different parts.^{Reference Fedorovich, Schuurman, Wijnberg, Prins, van Weeren, Malda, Alblas and Dhert43} The structured tissues formed distinctive osteochondral regions indicated by collagen Types I, II, and VI after subcutaneous implantation in immunodeficient mice for six weeks.^{Reference Fedorovich, Schuurman, Wijnberg, Prins, van Weeren, Malda, Alblas and Dhert43}

Cell homing approach for cartilage regeneration

Recruitment of endogenous cells by delivery of growth factor is an alternative to tissue regeneration by cell transplantation.^{Reference Mao, Stosich, Moioli, Lee, Fu, Bastian, Eisig, Zemnick, Ascherman, Wu, Rohde and Ahn32} We fabricated an anatomically correct bioscaffold by layer-by-layer bioprinting to replace the humeral head of the rabbit forelimb joint, an implant with spatial delivery of transforming growth factor beta 3 (TGFβ3). At four months, the articular surface of TGFβ3-infused bioscaffolds was fully covered with hyaline (i.e., translucent) cartilage over the scaffold surface and integrated with regenerated subchondral bone that had well-defined blood vessels, and with mechanical properties similar to that of native articular cartilage. In contrast, TGFβ3-free bioscaffolds had only isolated areas of cartilage formation, whereas no cartilage formation occurred in defect-only rabbits (injury induced, but no treatment). Our findings suggest that regeneration of cartilage may benefit from recruitment of endogenous cells from spatial delivery of biological cues (without cell transplantation) in an anatomically correct scaffold.^{Reference Lee, Cook, Mendelson, Moioli, Yao and Mao38} Subsequently, we showed that TGFβ3 and stromal cell-derived factor 1 (SDF-1; a signaling protein) not only chemotactically recruited adipose stem/progenitor cells, bone marrow MSCs, and synovium stem cells, but also induced chondrogenesis of the recruited cells, suggesting that homing of multiple stem/progenitor cell populations may potentially serve as an alternative or adjunctive approach to cell transplantation for cartilage in clinical applications.^{Reference Mendelson, Frank, Allred, Jones, Chen, Zhao and Mao44}

Regeneration of tooth root

Three-dimensional layer-by-layer printing was used to fabricate anatomically correct rat molar and incisor scaffolds from a hybrid of PCL and HA. These were implanted orthotopically (i.e., in a natural position) following mandibular incisor extraction and ectopically (i.e., in an abnormal position) into the dorsum with delivery of SDF-1 and bone morphogenetic protein 7 (BMP-7). Regeneration of tooth-like structures and periodontal integration by cell homing was observed after nine weeks.^{Reference Kim, Lee, Kim and Mao45} In another study, dental-pulp tissue was regenerated by recruitment of cell homing upon in vivo implantation by delivery of a series of cues including FGF, VEGF, PDGF, nerve growth factor, and BMP-7.^{Reference Kim, Xin, Moioli, Chung, Lee, Chen, Fu, Koch and Mao46}

Regeneration of tooth-supporting complex by 3D bioprinting

Tooth-supporting periodontium consists of multiple tissues, including cementum, periodontal ligament (PDL), and alveolar bone. Park et al. followed a protocol to facilitate the regeneration of tooth dentin-ligament-bone complexes in vivo by controlling the multicompartmental scaffold architecture by rapid prototyping, manufacturing by 3D printing, surgical implanting into the periodontal defect site, and evaluating dual-architecture scaffolds as well as newly regenerated tissue physiology for clinical implementation. They observed newly formed cementum-like tissue, PDL, and bone structures.^{Reference Park, Rios, Jin, Bland, Flanagan, Hollister and Giannobile47,Reference Park, Rios, Taut, Padial-Molina, Flanagan, Pilipchuk, Hollister and Giannobile48}

Multiphase periodontium tissue was recently developed by spatiotemporal delivery of multiple proteins and multiphase microstructure of poly(carprolactione)-hydroxylapatite. The scaffolds were fabricated using 3D printing seamlessly in three phases with amelogenin, connective tissue growth factor, and BMP-2 spatially delivered, and time-released in different phases. The complex structure formed PDL-like collagen I-rich fibers, mineralized tissues, and dentin/cementum-like tissues both in in vitro culture and in vivo implantation. This indicated that the optimal combination of microstructure and bioactive cues with target cell types is beneficial for periodontal tissue regeneration.^{Reference Lee, Hajibandeh, Suzuki, Fan, Shang and Mao49}

In vitro bioprinted skin grafts

Lee et al. developed a skin biofabrication method using a 3D bioprinter with different dispensers independently operated by electromechanical valves and mounted on three axis robotic stages. Varied volume of the dispensed droplet, the liquid material, was controlled by differing the gate opening phase of the valves and the air pressure, so that it was possible to construct biomimetic skin by printing keratinocytes and fibroblasts in a layer-by-layer manner.^{Reference Lee, Debasitis, Lee, Lee, Fischer, Edminster, Park and Yoo51} The constructs were printed into a planar tissue culture dish and in a poly(dimethylsiloxane) mold containing 3D surface contours as a wound model. Three-dimensional bioprinted skin showed similar morphology to native skin layers prior to in vivo implantation.^{Reference Lee, Singh, Trasatti, Bjornsson, Xu, Tran, Yoo, Dai and Karande52} Koch et al. fabricated 3D skin grafts through deposition of fibroblasts and keratinocytes embedded in collagen, using a laser-assisted bioprinting technique, with formation of the basement membrane and intercellular junctions.^{Reference Koch, Deiwick, Schlie, Michael, Gruene, Coger, Zychlinski, Schambach, Reimers, Vogt and Chichkov53}

Skin bioprinting for wound healing

Skin grafts by 3D printing may promote skin wound healing and restore the barrier function.^{Reference Orgill54–Reference MacNeil56} Three-dimensional printing studies for skin regeneration are summarized in Table IV. One study used Matriderm, a clinically approved 3D bovine collagen-elastin matrix, microinjected with cells inside the matrix.^{Reference Killat, Reimers, Choi, Jahn, Vogt and Radtke57} This dermal substitute was applied in a rodent full-thickness wound model. Cells appear to migrate from the transplantation site into deeper, distant layers.^{Reference Skardal, Mack, Kapetanovic, Atala, Jackson, Yoo and Soker58} Amniotic fluid-derived stem cells showed greater microvessel density and capillary diameters than bone marrow MSCs and showed some efficacy in wounds and burns.^{Reference Skardal, Mack, Kapetanovic, Atala, Jackson, Yoo and Soker58} Michael et al. showed the successful 3D printing of a cell construct and the subsequent tissue formation in full-thickness skin wounds in vivo with printed keratinocytes forming a multilayered epidermis.^{Reference Michael, Sorg, Peck, Koch, Deiwick, Chichkov, Vogt and Reimers59} A portable inkjet delivery system for the in situ printing of skin cells into the lesion site was developed with capacity to print human keratinocytes and fibroblasts into full-thickness skin lesions of mice.^{Reference Xu, Binder, Aboushwareb, Dice, Atala and Yoo60} Results show complete wound closure in three weeks.

Table IV.

Three-dimensional printing for skin regeneration: in vivo studies.

Note: BMSCs, bone marrow mesenchymal stem cells.

A remaining substantial challenge in 3D skin printing is whether 3D-printed skin grafts, frequently printed with cells, necessarily have an advantage over non-3D-printed skin grafts including decellularized matrices and/or biodegradable materials. Non-3D-printed skin grafts have shown some efficacy in dermal wounds including dermis deep burn wounds^{Reference Burke, Yannas, Quinby, Bondoc and Jung61,Reference Yannas, Burke, Orgill and Skrabut62} and have the distinct advantage of the absence of the need for cell seeding. Can 3D-printed cell-laden skin scaffolds or 3D-printed cell-free scaffolds outperform non-3D-printed native matrices or biomaterial scaffolds in preclinical and clinical studies of skin wounds? Unless an affirmative answer emerges from multiple well-designed studies, 3D-printed cell-laden scaffolds remain an exciting scientific exercise but without demonstrated efficacy (and safety) in dermal wound healing.