Bioprinting of tissue engineering scaffolds

Cartilage and bone tissue

To produce a cartilage tissue scaffold that was suitable for load-bearing applications, Xu et al. ¹⁴ inkjet-printed rabbit articular chondrocytes suspended in a fibrin–collagen hydrogel onto an electrospun polycaprolactone (PCL) matrix. The scaffolds were fabricated by alternatively printing a layer of electrospun PCL fibers followed by a layer of chondrocytes/fibrinogen/collagen suspension. The inclusion of the PCL (M_n ~42,500, 10% wt/vol.) fibers within the natural polymer gel improved the mechanical properties. The PCL composite scaffold had a Young’s modulus of 1.76 MPa and an ultimate tensile strength of 1.11 MPa compared to the natural polymer gel alone, which had a Young’s modulus of 0.41 MPa and an ultimate tensile strength of 0.26 MPa. Cell viability of the printed cells, calculated by a comparison of living cells to total number of cells, was 81.58% ± 3.46% 1 week after printing. Scaffolds implanted subcutaneously in immunodeficient mice over an 8-week period developed a dense, well-organized collagen, not present in the control group (unseeded scaffolds). The cartilage tissue formation within the bioprinted scaffolds appeared histologically similar to normal elastic cartilage.

Scaffolds were fabricated by Cui et al. ¹⁵ using human chondrocytes suspended in a poly(ethylene glycol) dimethacrylate (PEGDMA) solution. The PEGDMA solution was polymerized using ultraviolet (UV) light during the printing process. Viability of the human chondrocytes after printing was 84.9% ± 2.2%. Scaffolds were cultured in cell culture medium, supplemented with either fibroblast growth factor-2 (FGF-2) or transforming growth factor-β1 (TGF-β1). TGF-β1 is used to maintain chondrocyte phenotype, while FGF-2 is used for chondrocyte monolayer expansion. The compressive moduli of the hydrogels reduced with the inclusion of cells and were dependent on cell density; however, compressive moduli increased with time spent in cell culture. Culturing with FGF-2/TGF-β1 supplemented medium was shown to be the most beneficial for cartilage development.

Cui et al. ¹⁶ also investigated the possibility for printing in situ. Using the same printing method as before, scaffolds were printed either separately or directly into osteochondral plugs that had been harvested from bovine femoral condyles. The printed scaffolds had a compressive modulus of 395.73 ± 80.40 kPa, similar to natural human collagen. By polymerizing the scaffold as each layer was printed, cellular exposure to UV radiation was reduced by more than 80% compared to manual hydrogel scaffold polymerization that normally requires >10 min UV exposure. Compared to the control of a cast scaffold, cell viability of the printed cells was 26% higher. The inclusion of cells into the polymer matrix made the scaffolds ~20% more flexible, possibly due to the cells absorbing some of the compressive forces. Scaffolds that were printed directly into the bovine femoral condyles defects had enhanced tissue integration compared to non-printed cell-laden hydrogels. The improved tissue integration for scaffolds printed directly into the defect, along with a fabrication timing of 108 s, makes the bioprinting process a viable option for in-situ applications; however, the achieved compressive modulus of the scaffolds is significantly lower than that of human articular cartilage of the lower limbs, which range between 4.5 and 11.8 MPa. ⁴⁹ Therefore, further structural development of the scaffold would need to be considered before human implantation.

Neural tissue

To establish the ability of thermal inkjet printing for depositing mammalian cells, Chinese hamster ovary (CHO) and rat embryonic motoneuron cells were printed, suspended in a phosphate-buffered saline (PBS). ²⁷ After printing, the cells maintained a healthy morphology. The CHO cells proliferated into a confluent layer, while the rat embryonic motoneuron cells proliferated and exhibited a polarized morphology. CHO cell lysis was evaluated immediately after printing and indicated that ~8% of the cells were damaged due to the printing process. Overall, the study demonstrated the potential for inkjet printing to deposit cells into tissue scaffolds.

Further experiments were conducted with rat primary embryonic hippocampal and cortical neurons. ²⁸ Viability of the cells was 74.2% ± 6.3%, which maintained their basic function, phenotype, and electrophysiological properties. The printed cortical cells developed into mature neurons, while both types of cells could produce action potentials.

Lorber et al. ³⁵ were the first group to print cells from the mature adult nervous system. Printing with adult cells is more challenging, as adult neural cells are more prone to damage and are less able to regenerate. Using a high-speed, image-capture setup, it was observed that during droplet ejection, the cells did not undergo major distortion and hence destruction caused by shear forces. There was no difference in cell survival rate and neurite growth between printed and non-printed cells. The glia cells retained their ability to support the growth of the retinal ganglion cells.

Tse et al. ⁵⁰ were the first to print a combination of both neuronal and supportive glial cells. Not only did the printed cells maintain high viabilities, >86% for neuronal cells and >89% for Schwann cells, but more importantly, the printed neuronal cells exhibited earlier and longer neurite growth in comparison with the non-printed cells. By demonstrating the successful printing of both the neuronal and glial cells, the potential for producing a large tissue scaffold with a complete neural network becomes a greater possibility and could have massive potential for the repair of large-area, full-thickness skin burns.

Cardiac and vascular tissue

Initial work into printing cardiac tissue was performed using alginate hydrogel and primary feline adult and human H1 cardiomyocytes. ³⁶ Xu et al. printed a half heart shape that had a 1-cm inner diameter and two connected ventricles. In cell culture, the cells remained viable in the large structure due to printed porosities of the scaffold. Using electrical stimulation, the cardiac cells were observed to contract rhythmically which in turn caused the whole structure to beat periodically. Further development of the model should consider structural integrity of the scaffold to prevent the seepage of fluid from within the structure, as well as the capability to withstand physiological fluid pressures.

Endothelial cells have also been successfully printed with a high degree of accuracy and a precise number of cells per drop. ³⁷ Different methods have been explored for the bioprinting of a 3D vascular network. One method is to print a crosslinking agent (calcium chloride) into an alginate/gelatin solution. The calcium chloride crosslinks the alginate/gelatin solution to form a gelled structure. Boland et al. ³⁸ used this method to produce different structures such as tubes, branched tubes, hollow cones, and capillaries. Using a scanning electron microscope (SEM), the internal structure of the scaffolds was evaluated. It was shown that the calcium chloride crosslinked the alginate/gelatin gel on the surface of the droplet to produce hollow shells. This provided a scaffold with a microscopic porosity. Mouse endothelial cells attached to the surface inside the shells, with filopodia leading into the shells and lamellipodia trailing out, suggesting cell migration into pores.

Another method used for inkjet bioprinting vasculature depended on the reaction between thrombin and fibrinogen to form fibrin. This reaction was used by Cui et al. ⁷ to encapsulate human microvascular endothelial cells (HMVECs) in a fibrin hydrogel. HMVECs were suspended in a thrombin calcium chloride solution and printed onto a fibrinogen substrate that reacted to fibrin, thus encapsulating the cells. Using this method, scaffolds were fabricated with micron-sized fibrin channels. The printed hollow fibroin channels were well aligned and had a straight microstructure. The average fiber diameter was 93 µm, had an elastic modulus of 2.9 ± 0.8 MPa, an ultimate tensile strength of 1.7 ± 0.5 MPa, and a burst pressure estimated to be 2955 mm Hg; all properties were comparable to that of extruded tissue-engineered blood vessels. After 21 days in cell culture, the cells had aligned along the printed structures to form confluent linings. A ring-shaped microvasculature had formed that had sealed the fibroin channel with a high level of integrity.

One of the issues associated with inkjet printing is the fabrication of overhanging structures. The printing of overhanging structures is challenging as each layer of printed droplets is supported by either the substrate or previously printed layers. To overcome this issue, Christensen et al. ⁸ used calcium chloride solution for both crosslinking and as liquid support material. Sodium alginate ink, with and without the inclusion of NIH-3T3 mouse fibroblasts, was printed into a calcium chloride solution, which gelled almost instantaneously upon impact. The calcium chloride solution provided buoyancy to the overhanging structures and was easily removed after fabrication, unlike solid support structures that can be difficult to remove post fabrication. The print quality of the rounded overhangs was negatively affected by gravitational forces and droplet impact during printing. The low mechanical integrity of the alginate gel also meant that the designed structure did not maintain its 3D shape during printing. Although the positioning of the walls moved, printing could be compensated to ensure that the structure was completed and the scaffolds exhibited a 90% cell viability after 24 h. The current model would be insufficient for implantation due to the potential for turbulent and vortex blood flows to occur due to poor uniformity of the printed shape. This could induce the development of a thrombosis and eventually block fluid flow. ⁹ Therefore, due to the poor structural integrity encountered during fabrication related to the use of a liquid support structure, it is suggested that a solid support structure may be preferable when printing a tubular hydrogel scaffold.

Skin tissue

To produce a skin graft, Lee et al.^10,11 printed a layered structure with multiple cell types. Primary adult human dermal fibroblasts (hFB) and primary adult human epidermal keratinocytes (hKC) were suspended in cell media and printed in-between layers of a printed collagen (type I and rat tail) gel. ¹⁰ Using a microvalve-based bioprinting system, the number of cells per droplet ranged from 68 ± 13 cells/droplet for hKC to 93 ± 13 cells/droplet for hFB. The inner layers of the scaffold were composed of fibroblasts and the outer layers composed of keratinocytes. There were no morphological differences between the bioprinted cells and manually seeded cells, and no significant difference in cell viabilities. To demonstrate wound-specific scaffold manufacture, scaffolds were printed into polydimethylsiloxane (PDMS) molds with non-planar 3D surface contours. Scaffolds fabricated in the non-planar molds suffered from an inhomogeneous printed layer depth; however, this had no negative influence on the proliferation of the cells.

Scaffolds fabricated in the same way but on a planar surface (shown in Figure 1) were compared to conventional scaffolds that required manual seeding of the cells. ¹¹ After 7 days submerged in cell culture medium, the printed scaffolds maintained their shape and dimensions, as well as produced three densely packed cell layers of hKCs on the surface of the scaffold. In contrast, the conventionally produced skin scaffold continually shrunk and had lost shape after 7 days in culture. When exposed to an air–liquid interface, the printed scaffolds developed a dense epidermal layer, with the number of epidermal layers increasing from 3 to 7 after 14 days of cell culture. The epidermal layers formed tight junctions between neighboring cells, an important feature of the epidermis for providing an effective barrier.

OPEN IN VIEWER

Figure 1.

Schematic diagram of multilayered skin tissue with two layers of keratinocytes and three layers of fibroblasts embedded in collagen, fabricated with inkjet bioprinting (taken from a study performed by Lee et al. ¹¹ ).

An alternative PEG-based skin model has been developed by Rimann et al. ⁵¹ Printing at a temperature of 20°C, alternate layers of PEG-based bioink and primary human dermal fibroblasts were deposited to produce a scaffold. Each layer of the PEG-based bioink was immediately polymerized using UV light. The top layers of cells were encased within an additional seven layers of the bioink. DNA damage was measured by the formation of cyclobutane pyrimidine dimer (CPD), a product of UV damage to DNA. UV exposure during printing was shown not to damage cell DNA when compared to unprinted cells. Cell viability, proliferation, and morphology of the printed cells were similar to that of the unprinted cells. After 2 days in cell culture, the fibroblasts were shown to adhere to the printed structure with a typical morphology and spread within the printed matrix. Over a 3-week period, cells continued to proliferate and produce their own extra cellular matrix. Primary human dermal keratinocytes seeded on top of the printed scaffolds did not achieve a fully stratified epidermis when exposed to the air–liquid interface for 14 days. By not printing the keratinocytes onto the scaffold, cell distribution and positioning were affected. Achieving a more uniform layer of keratinocytes during the seeding process could have helped to achieve a stratified epidermis, as demonstrated by Lee et al. ¹¹

Yanez et al. ⁵² compared a bioprinted scaffold to a commercial graft (Apligraf®) for treating full-thickness skin wounds in mice. Human dermal microvascular endothelial cells (HMVECs) were suspended in a PBS containing thrombin and calcium chloride. The HMVECs were printed onto a gel consisting of neonatal human dermal fibroblast (NHDF) cells and collagen. After printing and a brief period of incubation, neonatal human epidermal keratinocytes (NHEK) mixed in collagen were applied to the top of the scaffold. Full-thickness wounds were created on the backs of mice, which were then treated with the bioprinted scaffolds, Apligraf®, or treated without a graft. After 14 days, the bioprinted scaffolds had adhered to the wound while the Apligraf had dried out and detached from the implant site, although no infections were recorded for any of the test groups. By Day 21, complete closure of the wound had occurred for the bioprinted scaffold, followed by the Apligraf-treated wounds after 28 days. Wound contraction at all time points was improved in the bioprinted group. By Week 6, the bioprinted group had a wound contraction of 62.5% ± 9% compared to the Apligraf with 74% ± 5%, and the negative control with 79% ± 2%. All groups developed epidermal and dermal layers; however, the printed skin graft was shown to have the most histologically similar appearance to native skin. The printed scaffold also showed evidence of vascularization, as the printed cells had integrated with the native tissue and helped to develop a microvessel network.

An in-situ printing method has been tested by Skardal et al. ⁵³ Amniotic fluid–derived stem (AFS) cells and bone marrow–derived mesenchymal stem cells (MSCs) were separately suspended in fibrinogen–collagen solutions. The cell suspensions were printed onto full-thickness skin wounds on the backs of mice and crosslinked by printing a separate thrombin solution between layers of the cell suspensions. Only one cell type was used for the treatment of each wound and was compared to scaffolds produced in the same way without the inclusion of printed cells. Printing produced a 100% coverage of the wound with a tight seal between the skin and graft. Wound contraction and re-epithelialization were at all time points significantly higher in the groups with printed cells. After 1 week, wound contraction was 83% for scaffolds with printed cells, compared to 73% for the gel alone. Printed scaffolds exhibited re-epithelialization of up to 89% which had organized, well-defined layers, while the control only exhibited a 51% re-epithelialization with a poor structural quality. Unlike the study of Yanez et al., the printed cells did not integrate with the tissue and had almost disappeared after 7 days. Therefore, in the study conducted by Skardal et al., the benefit of using printed cells was the secretion of trophic factors. The concentration of secreted factors was highest for the AFS cells, which coincided with better regeneration results. Overall, it was determined that AFS cells bioprinted onto a wound site could be beneficial for burn treatments. This study demonstrates an immediate and practical application of bioprinting. Not only can the graft be printed in situ, the AFS cells could come from a cell bank, enabling the possibility of “off-the-shelf” treatments.

Stem cells

Printing of stem cells presents the potential issue of changing their phenotype, which can be influenced by experimental conditions such as induced shear stresses. ⁵⁴ The effects of shear stresses on stem cell viability, phenotype, and function were analyzed by Blaeser et al. ⁵⁵ Over a range of shear stresses, printed primary human mesenchymal stem cells (hMSCs) were shown to have a reduced viability immediately after printing at high shear rates. However, after 7 days in cell culture, no significant difference between cells printed at low and high shear rates was observed. Cell proliferation rate was stimulated by a moderate level of shear stress, yet, above a certain threshold, there was an observed negative effect. Cell phenotype was unaffected by shear forces during printing. The study demonstrated that below a certain shear stress, hMSCs can be printed without side effects.

The effect of printing on other stem cells has also been investigated. Faulkner-Jones et al. ⁵⁶ investigated the influence of printing on human embryonic stem cells (hESCs) and human-induced pluripotent stem cells (hiPSCs). The use of hiPSCs for tissue engineering applications could be more beneficial than using hESCs, as the hiPSCs could be harvested from the patient and therefore immunosuppressive drugs would not be required for their implantation. The printing process did not change the differentiation of the cells nor change their lineage when compared to unprinted cells. The expression of pluripotency markers was slightly lower than the non-printed cells, however still within an acceptable range. When printing large circular structures, cell viability dropped to 55.5% 1 h after printing; however, the size of the scaffold did not influence the differentiation and maturation process of the cells. A low cell viability of the large scaffolds could be linked to a structural design restricting the availability of nutrients. Another possibility could be extended fabrication time, during which the cells are subjected to non-physiological conditions.

By printing stem cells into a scaffold, it is possible to develop a complex tissue. Gurkan et al. ⁵⁷ printed MSCs for the development of a multiphase tissue scaffold. Multiphase tissue scaffolds could be useful for repairing soft–hard tissue interfaces such as between tendon and bone. Scaffolds were produced by printing inks composed of methacrylated gelatin (Me-Gel) and hMSC cells, supplemented with bone morphogenic protein-2 (BMP-2), TGF-β1, or a combination of both. Scaffolds were printed with one side containing only BMP-2 and the other only TGF-β1, with a multiphase section in the middle. A smooth transition was produced between each phase. Osteogenesis- and chondrogenesis-related genes were upregulated within the respective BMP-2 and TGF-β1 regions, as well as in the multiphase section after 36 days in culture. Due to the complexity of tissue junctions and the difficulty of producing the intricate organizational structure, there are yet to be any translational studies to test current tissue-engineered scaffolds in a clinical setting. ⁵⁸ Therefore, the success of bioprinting to differentiate stem cells into different lineages within the same scaffold could be the solution for future development of these complex structures.

Bone tissue

Fedorovich et al. ¹⁹ printed intricate porous constructs containing endothelial progenitor cells (EPCs) and multipotent stromal cells (MSCs). Tissue development was studied for scaffolds composed of either Matrigel, or Matrigel and alginate. For the composite scaffold, EPCs were encapsulated in the Matrigel and MSCs in alginate. Scaffolds were implanted into subcutaneous dorsal pockets of mice. The Matrigel scaffolds promoted better cell migration in comparison with the Matrigel–alginate constructs due to fewer interactive properties of alginate. It was found that bone-like tissue had developed in the Matrigel scaffolds at 6 weeks after implantation. ¹⁹ However, due to the lower mechanical properties of Matrigel, load-bearing applications would require the addition of another material construct to provide extra support.

Hung et al. ²⁰ fabricated scaffolds out of PCL and decellularized bone matrix laden with human adipose–derived stem cells (hASCs) for craniofacial regeneration in mice. Compared to a pure PCL scaffold, the composite scaffold had an increased surface roughness. As would be expected with addition of hydroxyapatite, a bioactive material, enhanced bone regeneration was observed with increased cell adhesion in the composite scaffolds compared to pure PCL at 1 and 3 months of post-implantation. ²⁰

Due to the limited processability of natural polymers, synthetic polymers are often used for bioprinting applications. However, synthetic polymers generally lack similar cellular interactions experienced by natural polymers. In order to attain a more biologically functional 3D-printed scaffold, printed scaffolds comprising PCL, poly(lactic-co-glycolic acid) (PLGA), and β-tricalcium phosphate (β-TCP) were subsequently seeded with human nasal inferior turbinate tissue–derived mesenchymal stromal cells (hTMSCs) to lay down a layer of ECM. ⁶¹ After which, the scaffolds were decellularized and prepared for subcutaneous implantation in rats. Compared to printed scaffolds without a layer of ECM, the ECM/PCL/PLGA/TCP scaffolds demonstrated greater osteoinductive potential and an enhanced density of mineralized tissue throughout the defect. Overall, this work demonstrates a potential alternative to xenogeneic or allogeneic bone grafts.

Chondral and osteochondral defects are difficult to treat or reconstruct due to their complicated structure. To date, no clinical methodologies are able to reproduce hyaline cartilage.^21,62 To repair chondral defects, O’Connell et al. ⁴³ introduced a new approach using an “in-situ biopen,” which can reconstruct chondral, maxilla, and mandible defects. The biopen comprised two ink chambers and an extruder nozzle with a UV source. The biopen enables the deposition of living cells and biomaterials in a manual direct-write fashion. Using the biopen, a human infrapatellar fat pad (IPFP) adipose stem cell–laden gelatin-methacrylamide (GelMA)/hyaluronic acid-methacrylate (HAMa) hydrogel scaffold was fabricated for in-vitro test. It was observed that the viability of biopen-printed cells was 97% after 7 days. ⁴³ The biopen may be limited to be used for simple applications, where a specific structure is not required. For complex tissue structures that rely on the precise positioning of cells and material, using a bioprinting method with a fixed set of axis could prove to be more beneficial.

Cartilage tissue

To meet the increasing demand for cartilage tissue regeneration, Schuurman et al. ⁴⁴ fabricated a scaffold using GelMA embedded with chondrocytes and PCL to provide structural support. GelMA was also combined with hyaluronic acid to provide an anabolic effect on extracellular matrix synthesis and may be beneficial for cartilage tissue engineering. ²⁹ After printing, cell viability was significantly higher in GelMA/hyaluronic acid hydrogels (82%) compared to GelMA hydrogels (73%). The scaffolds were then implanted in the subcutaneous pockets of nude mice and they observed cartilaginous tissue formation after 4 weeks in both materials.

PCL–alginate gel scaffolds containing chondrocytes were successfully fabricated by Kundu et al. using a multihead deposition system (MHDS). The MHDS had four dispensing heads enabling different biomaterials to be dispensed at the same time. ³¹ The scaffolds were fabricated using layer-by-layer deposition of PCL followed by a chondrocytes alginate hydrogel bioink. Each layer of the scaffold comprised PCL lines with bioink printed in the gaps between each line. The scaffolds were implanted in the dorsal subcutaneous space of mice. The results indicated the formation of cartilaginous tissue in the scaffold after 4 weeks with 85% cell viability. ³¹

Periodontal tissue

Periodontal ligament stem cells (PDLSCs) have great potential for periodontal tissue regeneration. ¹² The behavior of PDLSCs after printing was investigated by Ma et al. ⁶³ using GelMA and poly(ethylene glycol) (PEG) dimethacrylate and PDLSCs. Cell viability test showed high viability of 94% after 24 h post printing.

Kim et al. ³³ demonstrated a successful reconstruction of maxillary bone in a 12-year-old dog using a printed PCL/β-TCP scaffold. The scaffold was fabricated using micro-extrusion-based 3D bioprinting and demonstrated successful reconstruction 8 months after surgery. This case is believed to be the first case of reconstruction using a 3D-printed scaffold in the maxillary bone defect. ³³

Neural tissue

Autologous grafts are the gold standard for repairing damaged neural tissue. However, such procedures involve high risk due to donor site morbidity, multiple surgery sites, and an increased threat of infection. Hsieh et al. printed neural stem cells (NSCs) embedded in the polyurethane (PU) using fused deposition. The printed scaffolds were studied in a zebra fish traumatic brain injury model. ³⁹ To formulate the bioink, NSCs suspended in culture medium were mixed with PU nanoparticles. The NSC/PU hydrogels promoted the repair of damaged central nervous system in zebra fish. After the implantation with the NSC-laden constructs, nerve function was recovered for the adult zebra fish. ³⁹ This research supports the use of PU for neural grafts have been observed with conventional tissue engineering fabrication techniques. ⁴⁰

Owens et al. ⁶⁴ developed a technology to biofabricate purely cellular nerve grafts using multicellular cylindrical units composed of mouse bone marrow stem cells (BMSCs) and Schwann cells (SC). ⁵⁹ Scaffolds were fabricated using a multichannel extrusion-based bioprinter (NovoGen MMX; Organovo, San Diego) enabling cellular suspensions to be printed alongside an array of agarose rods that provide a temporary support structure (demonstrated in Figure 2). The agarose rod support structure was removed after a maturation period of 7 days. The bioprinted nerve grafts were implanted to bridge a 1-cm gap of the sciatic nerve in rats. The rats recovered both motor and sensory functions at 40 weeks. This study is an excellent representation of the use of bioprinting to aid with scaffold self-assembly for implantation. After a period of 7 days, the cells had produced enough extra-cellular matrix to support themselves and withstand the mechanical forces inflicted during surgery. By implanting only cellular material and cells, the scaffolds more closely represent an autologous graft, which are the gold standard for nerve reconstructive surgeries and currently offer the best option for nerve rehabilitation.

OPEN IN VIEWER

Figure 2.

Schematic diagram of a nerve graft fabrication, in which the outer layer is made of bioink containing BMSCs supported by agarose rods. Structures were fabricated with an extrusion-based technique (taken from a study performed by Owens et al. ⁶⁴ ).

Skin tissue

Lee et al. ¹⁰ fabricated a multilayered skin-like tissue structure with the inclusion of fibroblasts and keratinocytes using both extrusion and inkjet printing processes. The cells were suspended in culture medium and printed using an inkjet printhead, while a support structure of poly(dimethylsiloxane) (PDMS) was printed under pneumatic pressure-based extrusion bioprinting. A layer of hydrogel was printed to provide structural integrity for the printed cells. Droplets of cell suspension were dispensed between layers of hydrogel. Each layer of hydrogel was treated with NaHCO₃ to induce crosslinking. After printing, the fibroblasts had a viability of 95%, whereas the keratinocytes had a viability of 83.9%. After 8 days of culture, printed fibroblasts showed a higher cell density compared to the keratinocytes. The tissue exhibited better shape retention compared to solid-state scaffold material during in-vitro studies. ¹⁰

Kim et al. ⁶⁵ demonstrated a 3D bioprinting strategy using an extrusion-based, integrated composite tissue/organ building system (ICBS) bioprinting module to fabricate PCL and collagen-based construct. This printing system provides nine different dispensing modules in which nine different types of biomaterials could be used. Five of the dispensing modules used an extrusion-based printhead, whereas the other four modules used a droplet-based printhead. PCL was extruded as a base mesh construct for an inkjet-printed, collagen-based bioink that contained either human primary dermal fibroblasts (HDFs) or epidermal keratinocytes (HEKs). The samples were cultured in a cell culture medium for 7 days. The results revealed that the PCL construct with collagen showed biological characteristics such as a fibroblast-stretched dermis and epidermal layers after 14 days, similar to a native human epidermal structure. ⁶⁵

Corneal tissue

Corneal transplantation or penetrating keratoplasty (PK) has been a gold standard for the treatment of corneal diseases over the past decades, since Dr Eduard Zirnm performed first cornea transplantation in a human eye in 1905. ⁶⁶ Limitations of the currently used procedure include the quality of visual recovery due to an increased early endothelial cell loss, detachment of the posterior lamellar grafts, and vascular in-growth into the lamellar plane. Bioprinting technology can replace the limitations of corneal transplantation.

Isaacson et al. ⁶⁷ fabricated a collagen-based cornea using pneumatic extrusion bioprinting. The bioinks were prepared using sodium alginate and methacrylated type I collagen, mixed with corneal keratinocytes. A plastic support material was printed to help mold the scaffold shape. The support material was printed with a concave hollowed-out structure at the top. The hollowed-out section of the support material was filled with a gelatin slurry in order to support the collagen and alginate bioinks. Cell viability of corneal keratocytes remained was 90% at Day 1 after printing, and 83% after 7 days. Although cell viabilities remained high, a 7% decrease after 7 days could indicate that either the printing process or scaffold design affected the cell survivability. A longer study would be required to evaluate if this was a continuing trend.

Cardiac tissue

Zhang et al. ⁶⁸ demonstrated a hybrid methodology to fabricate endothelialized myocardium. The construct was fabricated using extrusion bioprinting (NovoGen MMX; Organovo) with multihead dispensing nozzles. Induced pluripotent stem cells (iPSCs) were used to fabricate endothelialized human myocardium. The iPSCs are capable of differentiating into multiple cell lineages including cardiomyocytes. Microfibrous scaffolds were fabricated using GelMA and iPSCs, and seeded with cardiomyocytes to induce the formation of myocardium. It was observed that the printed endothelialized microfibrous scaffold was capable of spontaneous and synchronous contraction.

Duan et al. ⁶⁹ developed a 3D bioprinted hybrid hydrogel based on methacrylated hyaluronic acid (Me-HA) and Me-Gel to encapsulate human aortic valvular interstitial cells (HAVICs). Cell viability determined using a live/dead assay gave a 90% cell viability. In addition, cells easily adhered and formed a monolayer on the surface of the bioprinted structure, and the encapsulated cells below the surface started to remodel the hydrogel after 3 days in cell culture. Remodeling of the hydrogel is a positive indication for fast graft integration with the host tissue.

EPCs are known to be a promising cell source for the treatment of ischemic diseases. Gao et al. ⁷⁰ fabricated a bio-blood-vessel (BBV) using vascular tissue–derived decellularized extracellular matrix (VdECM) and atorvastatin-loaded PLGA microspheres (APMS). Atorvastatin is a proangiogenic drug that promotes endothelial cell function. The BBV was implanted in a nude mouse hind limb ischemia model. Comparing different combinations of BBV, EPCs, and APMSs, it was discovered that BBVs laden with both EPCs and AMPSs significantly increased capillary and arteriole densities.

Jang et al. ⁷¹ fabricated a pre-vascularized stem cell patch using a bioink consisting of suspended decellularized porcine heart (dECM), and human c-kit + cardiac progenitor cells (hCPCs), MSCs, or a combination of both cell types. For in vivo testing, PCL was printed as a support for the bioink. Patches were either implanted subcutaneously into mice to test for neovascularization or implanted in a rat myocardial infarction model. The patches improved epicardial activation capability with enhanced angiogenesis. The patterned stem cells promoted strong vascularization and tissue matrix formation with enhanced cardiac functions. Therefore, bioprinted stem cell patches are demonstrated to be a promising therapeutic method for ischemic heart diseases.

Muscular tissue

Miri et al. ⁷² created hierarchical cell-laden structures to mimic multicellular tissues. Pneumatic extrusion was used to dispense multiple bioinks onto microfluidic chips that were then crosslinked using UV light. In-vitro studies using poly(ethylene glycol) diacrylate (PEGDA) and GelMA loaded with NIH-3T3 fibroblasts, hMSCs, osteoblasts, and C2C12 skeletal muscle cells were printed into structures resembling musculoskeletal junctions, muscle strips, and tumor angiogenesis. The prints retained interfaces and adequate proliferation rates after 3, 5, and 7 days in cell culture. For instance, a microfluidic chip with two different bioinks was fabricated to mimic bone–tendon junctions (Figure 3). The bone side of the chip contained a suspension of osteoblasts and hMSCs in GelMA. The tendon side contained a suspension of fibroblasts in GelMA. To fabricate a tumor angiogenesis resembling model, a microvasculature-like mask was used during UV curing of PEGDA, followed by washing with PBS and extruding endothelial cell-laden bioink into the microchannels. The interfaces remained explicit and the print patterns were well-defined. For in vivo studies, PEGDA-framed chips that had a concentration gradient of Me-Gel (GelMA) ranging 5%–15 % were implanted subcutaneously in rats. After 30 days, enhanced cell proliferation occurred for 10% GelMA. Having an optimum concentration of 10% GelMA is supported by previous studies performed by Zhou et al.^22,73 and Zhu et al. ⁷⁴ which will be further discussed later in the review.

OPEN IN VIEWER

Figure 3.

Schematic diagram of a microfluidic chip that mimics bone–tendon junctions, fabricated with pneumatic-based extrusion (taken from a study performed by Miri et al. ⁷² ).

Bone and cartilage tissue

The first instance of stereolithographic-fabricated, cell-laden scaffolds were produced by Dhariwala et al. ⁴¹ using poly(ethylene oxide) (PEO) and CHO cells. Using SLA, they were able to achieve a print resolution of 250 µm. Cells maintained a good viability and distribution between 1 and 2 days after printing. The scaffolds had insufficient mechanical properties; however, future work could incorporate a second polymer for greater structural strength.

Lu et al. ³² crosslinked PEGDA hydrogels laden with stem cells derived from mouse bone marrow stromal cells (OP-9 cells). The cells remained viable after 24 h. Osteogenic differentiation of mouse mesenchymal stem cells (mMSCs) seeded on to the scaffold showed intensive mineralization after 2 and 4 weeks.

To create biomimetic bone matrices to study metastasis of breast cancer in bone tissue, Zhou et al. ²² employed a table-top commercial stereolithographic printer. The aim was to create a cell-laden 3D structure and seed it with breast cancer cells (BrCas) and observe the interactions in vitro. Currently, most in-vitro cancer studies are performed on monolayered cell cultures which do not accurately mimic the 3D structure of bodily tissues. Osteoblasts and hMSCs were suspended in hydrogels consisting of different concentrations of GelMA and nanocrystalline hydroxyapatite (nHA). Cell viability of the scaffolds was negatively affected by UV irradiation; however, cell densities increased with time in cell culture. It was found that scaffolds with lower GelMA concentration of 10% exhibited a higher cell viability compared to 15% concentration, suggesting that higher concentrations of GelMA created higher encapsulation stresses and hindered nutrient transport. The osteoblast- and MSC-embedded scaffolds were seeded with BrCa cells and cultured for 2 weeks. After 5 days of culture on 10% and 15% GelMA scaffolds, osteoblast proliferation decreased by 46.9% and 34.7%, respectively, while BrCas proliferation increased by 16.5% and 23.8%, respectively. After 5 days of culture on 10% and 15% GelMA scaffolds, MSC proliferation had also decreased by 37.6% and 30%, respectively, while BrCas increased by 32.5% and 40%, respectively. The decrease in osteoblast and MSC proliferation, and a higher BrCa proliferation, suggests that osteoblasts and MSCs secrete macromolecules that promote BrCas growth which corresponds to other research.^79,80 Vascular endothelial growth factor (VEGF) secretion increased within the 2 weeks, while alkaline phosphates (ALP) activity decreased. As bioprinting techniques develop further, it may be possible to produce more complex biological system and enable the studying of disease interactions not just on a cellular level but also on the surrounding tissue. Building complex tissue models could enable a more accurate study of diseases and even provide a method for the testing of treatments to determine the most effective solutions.

Oxygen inhibition presents a limitation to SLA printing. When SLA is carried out in the presence of air, free radicals from the initiators, used during crosslinking, are seized/scavenged by O₂ molecules. This hinders the polymerization process, impacting print fidelity and ultimately can cause a physical collapse of the scaffold.^81,82 The effect of oxygen inhibition can be reduced by using a high-radiation intensity, increasing photoinitiator concentration, or by performing SLA under inert conditions such as in a nitrogen and carbon dioxide gas atmosphere. However, all of these approaches will negatively impact cell viability.

Lim et al. ¹³ conducted a study to compare UV and visible light (VIS) crosslinkage of cell-laden hydrogels to minimize the effect of oxygen inhibition. GelMA and bovine collagen I were mixed along with Irgacure 2959 and ruthenium (Ru) + sodium persulfate (SPS) for crosslinking with UV and VIS, respectively. Breast adenocarcinoma cells (MCF-7) were encapsulated in the hydrogels with different concentrations of the photoinitiators to study cytotoxicity effects. It was found that an increase in Irgacure 2959 will affect the cell viability drastically, down to <50% with a concentration of 0.5 wt%, while the different concentrations of Ru/SPS had similar viability rates. To determine the light intensity’s effect of UV and VIS on cell viability, a concentration of 0.05 wt% Irgacure 2959 with UV light and 0.2/2 Ru/SPS with visible light were used. The higher the UV intensity, the more the cell viability decreased; however, no significant difference was noted with VIS. Oxygen inhibition was measured by the change in diameter of the printed struts. Printing with VIS-Ru/SPS, the difference in strut diameter was much less than that when using UV-Irgacure 2959, <10%. Further studies were conducted using human articular chondrocytes (HAC) and human bone morrow–derived mesenchymal stem cells (hMSCs), suspended in the GelMA/collagen hydrogels and crosslinked with UV and visible light. Visible light and Ru/SPS exhibited higher cell viability than UV and Irgacure 2959 after 1 day of culture, >85% and <59%, respectively. Irgacure 2959 is a commonly used photoinitiator due to its lower cytotoxicity in comparison with other photoinitiators. ²⁶ As the comparison between cell viabilities of UV and visible light approaches is so stark, the use of visible light may provide the greatest potential for stereolithographic-based bioprinting.

Because of its avascular nature and the presence of low cell densities, cartilage defects will not completely self-regenerate. ⁸³ Zhou et al. included trophic factors into cell-laden cartilage scaffolds to improve cartilage regeneration. hMSCs were combined with 10% GelMA, and different concentrations of PEGDA and PLGA nanospheres containing TGF-β1. ⁷³ Lower concentrations of PEGDA improved cell proliferation after 1, 3, and 5 days. Release studies showed that the nanospheres were able to maintain a sustained release over a 21-day period. After 1, 2, and 3 weeks in cell culture, the expression of chondrogenic genes SOX-9 and Aggrecan were reported to be higher for MSCs in scaffolds containing TGF-β1, suggesting chondrogenic differentiation.

Muscular tissue

Cell patterning and encapsulation were implemented using dielectrophoresis (DEP) and SLA by Bajaj et al. ⁸⁴ Different cell types, mouse embryonic stem cells (mESCs), and mouse myoblast cell line (C2C12) were patterned and encapsulated in PEGDA hydrogels. Viability rates were similar before and after DEP. After 5 days, live/dead staining showed a slight decrease in cell viability, perhaps because of the non-adhesive property of PEGDA. ²⁵

Neural tissue

Zhu et al. ⁷⁴ studied the differentiation potential of human neural stem cells (hNSCs) into neurons after printing using SLA apparatus. Over a range of concentrations, it was determined that low GelMA concentrations yielded higher proliferation rates. 10% GelMA hydrogels were combined with hNSCs and graphene nanoplatelets. Graphene is known to stimulate hNSCs into neural differentiation. ³⁴ Glial fibrillary acidic protein (GFAP) expression is associated with the development of ependymal cells, and β-tubulin III is associated exclusively with neurons and testis cells. After 14 days, β-tubulin III expression increased while GFAP expression declined. Changes in the expression of GFAP and β-tubulin III, along with neurites elongation, suggest that neural differentiation had occurred.

Bone tissue

Biopapers, the biological substrate receivers, are produced to enhance cell viability of LAB prints. Such examples are gelatin, fibrin, alginate, nanohydroxyapatite, PCL, and PLGA hydrogels. Galbraith et al. ⁸⁹ developed a self-assembled osseous biopaper using osteogenically differentiated hASCs. Osseous biopapers exhibited higher ALP activity than stromal biopapers after 5, 15, and 21 days. Alizarin red staining displayed higher matrix mineralization within the osseous biopapers. This result is concurrent with current research that indicates that hASCs may be better suited for certain applications in comparison with stromal cells, and that the more easily isolated hASCs may be better suited for LAB-based techniques. ⁹²

Rapid development of an adequate vasculature within the implanted scaffold is important to avoid hypoxia leading to a necrotic core. Pre-vascularization can be used to improve scaffold regeneration and can be achieved through in-vitro cell seeding and/or in-situ flap techniques. Kawecki et al. ⁹³ performed laser-assisted positioning of human umbilical vein endothelial cells (HUVECs) within the biopapers, as a pre-vascularization step. The endothelial cells were viable immediately after printing. Orientation of the HUVECs showed better alignment for the osseous biopapers. Preservation of cell alignment was favored in the locations where the HUVECs were aligned with cells of the biopapers. After 7 days of printing, tubule-like endothelial cells were observed.

Catros et al. ⁹⁴ evaluated the effect of 3D positioning of cells on PCL biopapers. The scaffolds consisted of consecutive layers of electrospun sheets and monolayers of osteosarcoma cells (MG63) (demonstrated in Figure 4). For a positive control, the MG63 cells were only printed on top of the scaffolds. Cells were shown to maintain viability after printing. In vitro, cell proliferation rates did not significantly differ between the two groups after 1 and 7 days. However, after 14 days, the scaffolds with the cells printed within the structure exhibited a significantly higher proliferation rate. The scaffolds were implanted in calvarial defects in mice. After 2 months, a significant increase in bone tissue in-growth was observed for scaffolds with organized layers, when compared to the positive control. Histological analysis showed thick fibrous tissue growth in the location of the sequentially layered scaffolds.

OPEN IN VIEWER

Figure 4.

Schematic diagram of consecutive layers of electrospun PCL sheets and monolayers of osteosarcoma cells. Cell placement was controlled using laser-assisted bioprinting (taken from a study performed by Catros et al. ⁹⁴ ).

Skin tissue

Koch et al. ⁸⁸ studied the LAB of skin cells; fibroblasts and keratinocytes, and hMSCs. Cell viability immediately after printing was 98% ± 1% and 90 ± 10% for the skin cells, both keratinocytes and fibroblasts, and hMSCs, respectively. Compared to non-printed cells, there was no significant difference in apoptotic activity and proliferation rate between each cell type. hMSCs maintained their phenotype over the course of the study, implying that the transfer process did not induce any up-regulation of surface markers, and hence differentiation.

Corneal tissue

The limbus is an anatomical site that borders the cornea and provides it with limbal epithelial stem cells (LESCs) for regeneration. ⁹⁵ Destruction of the limbus by trauma or diseases can deprive the cornea from LESCs and, as a result, cause a limbal stem cell deficiency. Using LAB, Sorkio et al. ⁹⁶ printed human embryonic stem cell–derived limbal epithelial stem cells (hESC-LESCs) and hASCs for repair of the limbus. Three scaffolds were developed: a stratified corneal epithelium (with hESC-LESCs), a lamellar corneal stroma (with alternating layers of hASCs and acellular bioink), and a scaffold containing both corneal epithelium and stroma. The receiver-slide of the LAB setup contained human-sourced collagen I and recombinant laminin. After printing, the number of cells in each structure tripled after 7 days. Ex vivo assessment of the structures was performed on porcine-excised corneal tissue and compared to acellular Matriderm® sheets as a control. The multicell-laden structures exhibited strong adhesion with the host tissue and hASCs migration to it, while control groups did not exhibit such activities.