Up-to-date progress in bioprinting of bone tissue

Abstract

The major apparatuses used for three-dimensional (3D) bioprinting include extrusion-based, droplet-based, and laser-based bioprinting. Numerous studies have been proposed to fabricate bioactive 3D bone tissues using different bioprinting techniques. In addition to the development of bioinks and assessment of their printability for corresponding bioprinting processes, in vitro and in vivo success of the bioprinted constructs, such as their mechanical properties, cell viability, differentiation capability, immune responses, and osseointegration, have been explored. In this review, several major considerations, challenges, and potential strategies for bone bioprinting have been deliberated, including bioprinting apparatus, biomaterials, structure design of vascularized bone constructs, cell source, differentiation factors, mechanical properties and reinforcement, hypoxic environment, and dynamic culture. In addition, up-to-date progress in bone bioprinting is summarized in detail, which uncovers the immense potential of bioprinting in re-establishing the 3D dynamic microenvironment of the native bone. This review aims to assist the researchers to gain insights into the reconstruction of clinically relevant bone tissues with appropriate mechanical properties and precisely regulated biological behaviors.

1. Bone: Anatomy and functions

Bone has several crucial functions in the musculoskeletal system, including protecting organs, producing blood cells, storing minerals, and mechanically supporting the human body[1]. According to their macrostructure, the bone can be categorized into cortical (or compact) bone or cancellous (or trabecular) bone. Up to 80% of total bone mass consists of cortical bone, which is dense and has low porosity, contributing to bone’s mechanical strength[2]. Comparatively, the porous structure of trabecular bone results in tissue with a low compressive strength (e.g., one-tenth that of cortical bone)[2]. As a result of a large surface-to-volume ratio, the trabecular bone provides a structure that allows for good contact between bone and blood cells, which is essential for controlling hematopoiesis and homeostasis[3]. As an example, a long bone has three key regions, including a dense cylindrical shaft made of cortical bone (diaphysis) and two sections at the end made of trabecular bone (metaphysis). Microscopically, bone lamella, which is composed of mineralized collagen fibers, exhibits a planar arrangement with a width of 3 – 7 μm[4]. In cortical bone, osteons (or Haversian systems) are formed by concentric layers of lamellae wrapped around a central canal. On the other hand, the arrangement of mineral platelets in the trabecular bone is different, in which the mineral platelets are aligned with the collagen fibers[2].

Furthermore, lamellae, which are considered the basic building blocks of bone, contain mineralized collagen fibrils (~100 nm in diameter). Collagen type I (COL-I) is the primary organic component of the matrix, which is secreted by osteoblasts followed by self-assembling. Apatite crystals distribute discretely along the collagen fibrils. The lamellae and collagen fibers of bone are organized in a way that prevents crack propagation and increases bone toughness[5]. As for the composition of bone, hydroxyapatite (HAp, 60%) and collagen-I (20%) are the major substances. Besides, there are several types of impurities in bone, including sodium (Na⁺), magnesium (Mg²⁺), and potassium (K⁺), as well as proteins, such as osteocalcin (OCN), osteonectin, and sialoprotein[6]. Osteoblasts, osteocytes, osteoclasts, and bone lining cells are the four types of cellular components of bone that are embedded in the extracellular matrix (ECM). Osteoblasts, the bone-forming cells, consist of 4 – 6% of the total number of bone cells. ECM is synthesized by osteoblasts in a two-step process involving the deposition of organic matrix followed by mineralization[7]. During the calcification process, osteoblasts transform into osteocytes, which distribute in the ECM and act as stress sensors in the bone. In some cases, osteoblasts stay on the top of new bone and protect the bone underneath, hence the term lining cells. A bone osteoclast is a very large, multinucleate cell that absorbs bone matrix, in which function is critical in the maintenance, repair, and remodeling of bone.

Bone plays an essential role in providing the body with mechanical support. Calcium phosphate provides mechanical durability and high resistance to compression of bone, whereas collagen is responsible for bone’s elasticity and resistance to tension and stretching. Young’s modulus is approximately GPa for native bone, and tensile and compressive strengths are approximately MPa, which are determined by the location of the bone in the body or by the specific location within the bone. Compared with cancellous bone, cortical bone exhibits much higher Young’s modulus (7 – 30 GPa vs. 50 – 500 MPa), tensile strength (50 – 150 MPa vs. 1.2 – 20 MPa), compressive strength (167 – 193 MPa vs. 1.9 – 10 MPa), and strain to failure (1 – 3% vs. 5 – 7%) in the longitudinal direction[8,9].

2. Consideration of bioprinting a bone

The conventional tissue engineering strategy is to seed cells onto scaffolds, which can then direct cell proliferation and differentiation into three-dimensional (3D) functioning tissues and organs. Although significant success has been achieved in the past decades both in research and clinical applications[10], it is obvious that complex 3D tissues require more precise multicellular structures, which cannot be fulfilled by traditional methods. Bioprinting offers a controllable fabrication process, which allows precise spatial placement of various biomaterials and/or cell types simultaneously mimicking the target natural tissue or organ[11]. Complementing the existing review articles about bioprinting and its applications[12-14], this review focuses on 3D bioprinting of bone, which is rapidly advancing in the field of regenerative medicine. This article discusses the ­up-to-date progress in 3D bioprinting technologies for bone reconstruction by integrating knowledge from in vitro and in vivo studies. Bioprinting approaches and other major considerations are carefully evaluated, such as bioprinting apparatus, biomaterials, bone vascularization, cell source, differentiation factors, mechanical properties and reinforcement, hypoxic environment, and dynamic culture. This review outlines the latest advancement in bioprinting of bone and aims to assist the researchers to gain insights into the reconstruction of clinically relevant bone tissues with appropriate mechanical properties and precisely regulated biological behaviors.

In this review, studies based on conventional 3D printing, in which acellular scaffolds are fabricated followed by cell-seeding, are excluded from the study. Studies on the topic of bone bioprinting, in which cellular components are manipulated together with bioink deposition without subsequent manual cell-seeding process, are carefully selected and reviewed. From the viewpoint of the authors, the selected studies are innovative in certain aspects such as bioink preparation, bioprinting process, or construct design and represent the state-of-the-art progress in this field. When discussing about bone bioprinting, the following questions should be kept in mind (Figure 1):

Figure 1.

Design flow for bioprinting a bone construct.

• (i) What printing processes and structure designs can facilitate micro-/macroscale vascularization of bioprinted bone?

• (ii) What is the optimal bioink formula for derive the greatest benefits from natural and synthetic biomaterials?

• (iii) What are the best combinations of cell types for bone bioprinting? When multiple cell types are loaded in the bioink, how do they interact and promote osteogenesis of stem cells and the formation of vasculature?

• (iv) Besides the well-acknowledged growth factors, are there other biologicsor supplements that can be incorporated in bioprinting and are they conducive to osteogenesis?

• (v) What can be carried out to compensate for the weak mechanical properties of bioprinted bone resulting from the nature of cell-laden bioinks?

• (vi) As bone cells are highly metabolically active, how does the hypoxic culture environment impact the maturation of bioprinted bone?

• (vii) How do bone cells respond to mechanical stresses in the bioprinted constructs in terms of cell morphology, osteogenesis, and mineralization?

2.1. Bioprinting apparatus

The common bioprinting processes, including extrusion-based bioprinting (EBB), droplet-based bioprinting (DBB), and laser-based bioprinting (LBB), have been utilized for bone bioprinting, depending on the selective bioink formula. In EBB, the bioink is deposited from a syringe or nozzle onto a build platform based on a computer-aided design of the structure to be printed (Figure 2A). This is accomplished by laying down small cylindrical deposits of the material, either pneumatically, mechanically, or by solenoid-driven deposition. In general, bioprinted bone constructs are primarily fabricated by EBB due to its efficiency in printing large-scale constructs in 3D, and its flexibility to handle a variety of biomaterials to obtain sufficient mechanical strength. Meanwhile, in DBB, bioink with modulated fluid properties (e.g., surface tension and viscosity) is manipulated to form droplets and then constructed using gravity, atmospheric pressure, and fluid mechanics[15] (Figure 2B). The LBB process involves the use of a laser pulse directed through a mirror onto a layer of bioink. In LBB processes based on photopolymerization, an ultraviolet (UV) laser is used to cure hydrogels in a vat that is capable of photocrosslinking (Figure 2C). To build the final construct, this process is repeated several times in a layer-by-layer manner[16,17].

Figure 2.

Bioprinting technologies for bone tissue. (A) Extrusion-based bioprinting with traditional pneumatic/piston extrusion (A1), coaxial nozzle for the printing of core-shell filaments (A2), dual-nozzle setup for bioprinting of multiple bioinks (A3), and post-printing polymerization for photocurable bioink, (B) droplet-based bioprinting, (C) laser-based bioprinting, and (D) aspiration-assisted bioprinting.

Aspiration-assisted bioprinting (AAB), which has recently been developed to manipulate cell spheroids, is a new bioprinting technique that capitalizes on the fact that spheroids can be formed from diverse cell types at high densities, lifted by employing negative air pressure, and bioprinted on a hydrogel[18] (Figure 2D). The above-mentioned bioprinting processes are sometimes integrated to obtain optimal mechanical and biological properties.

2.2. Selection of bioink

Different cellular responses have been observed when natural hydrogels are used for bone bioprinting. For bone bioprinting, alginate has always proven to be a popular biomaterial due to its biocompatibility, low cost, and ease of cross-linking by contact with calcium (Ca²⁺) ions[19]. The low bioactivity of alginate poses a limitation to its use. As compared, the similarities between collagen and native bone make it an ideal material for bone bioprinting. It is, however, difficult to generate collagen hydrogels with high viscosity that have rapid gelation capabilities. Therefore, collagen was only used in a handful of bone bioprinting studies, usually in combination with other biomaterials. Since gelatin is derived from collagen, it is a more economical option and is used in conjunction with other biomaterials to form bioinks. The biological functions of bioprinted bone can also be regulated by other popular biomaterials, such as agarose, chitosan, and hyaluronic acid (HA)[20,21]. Aside from mixing several natural polymers with varying concentrations to tailor bioink properties, polymers can also be modified to have customized properties. For example, methacrylate hydrogels, which are natural components of the ECM modified by methacrylation, are widely used in the field of bioprinting[22]. As one of those hydrogels, gelatin methacrylate (GelMA) is becoming a popular biomaterial for 3D bioprinting[23,24], due to its biocompatibility as well as its ability to cross-link chemically with UV light under physiological conditions[25]. Another example of hydrogel is methacrylated HA (MeHA)[26], which has been combined with GelMA hydrogel for bone bioprinting[27].

These above-mentioned polymers can directly bioprint with cells; however, the application of cell-laden hydrogels in hard tissue regeneration has been restricted by their low mechanical properties[28,29]. Consequently, biomaterials such as ceramics, thermoplastics, or alloys that were traditionally used for the manufacture of bone scaffolds could be incorporated with hydrogels to boost the mechanical strength of the bioprinted bone[28,30-32]. There are multiple options for biomaterials that imitate the main components of bone, including HAp and tricalcium phosphate (TCP)[33-35], as well as polycaprolactone (PCL), polylactic acid (PLA), and polyether ether ketone, which are known for their great mechanical properties[36]. Furthermore, studies have been conducted with 45S5 Bioglass, polyP, and biosilica, which enhance the ability of cells to synthesize mineral deposits and induce the expression of alkaline phosphatase (ALP), bone morphogenetic protein 2 (BMP-2), and COL-I[37,38].

2.3. Vascularized bone structure

Considering that bone is a complex tissue with a hierarchically organized structure, biomimetic design becomes an essential consideration. Cells within the bone are encapsulated by the mineral matrix, which is surrounded by blood vessels[39-42]. It is widely recognized that interconnected pores are necessary for the growth of blood vessels. According to the general viewpoint of tissue engineering, a pore size of 5 microns is optimal for neovascularization, 5 – 15 microns for fibroblast ingrowth, 40 – 100 microns for osteoid ingrowth, and 100 – 400 microns for bone regeneration[43,44]. Other than the use of porous structures, a number of strategies have been developed to promote angiogenesis, including the use of angiogenic growth factors[45,46], dynamic culture[47], and coculture of endothelial cells (ECs) with mesenchymal stem cells (MSCs)[48-50].

A functional blood vessel system at the network level within the implant is required for the effective restoration of bone. It is expected that the endothelium and vessel wall are organized in a networked fashion. In this context, bioprinting is possible to remodel a hierarchically branched vascular network with anatomic similarity. In one case, a geometry consisting of two phases was envisioned, with HAp forming part of the shape (bone tissue) and alginate and gelatin filling the tubular structure (vascular structure)[51]. Moreover, Byambaa et al.[52] designed a pyramidal construct that included a perfusable vascular lumen and was functionalized with gradient concentrations of vascular endothelial growth factor (VEGF). Media perfusion for 5 days was performed to produce a hollow perusable main vessel. For example, Cui et al.[53] fabricated a biphasic structure using a dual bioprinting procedure that combined fused deposition modeling (FDM) and stereolithography (SLA). The honeycomb-pored units in the hard portion of the scaffold mimicked the osteon or Haversian system of bone, while the channels filled with GelMA hydrogel resembled the vascular system. Cell growth and expansion in the hydrogel were expected to render the formation of capillaries.

2.4. Cell source

Bone bioprinting has continued to use similar cell types that are popular in bone tissue engineering. A mouse osteoblast precursor cell line (MC3T3) has been proven to be highly feasible and physiologically relevant for bone bioprinting. Nevertheless, it is important to exercise caution when extrapolating these results to primary cells. In research pertaining to the bone, human osteosarcoma cell line (SaOS-2) that is derived from the primary osteosarcomas is commonly used as osteoblast-like cells[54]. The SaOS-2 cell line has several advantages, including the ability to propagate quickly and to differentiate in a manner similar to that of osteoblasts[55].

The multipotency of stem cells makes them a popular alternative to cell lines and primary bone cells. Numerous studies have been conducted on bone bioprinting using MSCs, such as bone MSCs (BMSCs) and human nasal turbinate-derived MSCs (hTMSCs)[56]. An intriguing finding is that MSCs can be obtained from induced pluripotent stem cells (iPSCs), which can overcome the problem of limited quantities of autologous MSCs[57]. Furthermore, adipose-derived stem cells (ADSCs) are plenty in the human body and are surgically accessible, which makes them another striking source for bioprinting[13]. To support therapeutic revascularization in bioprinted tissue, ECs[26,52,53], endothelial progenitor cells (EPCs)[58], or endothelial colony-forming cells (ECFCs)[59] have been cocultured with the above-mentioned cells[59-61].

2.5. Osteogenesis

In principle, bioprinted bone should promote and facilitate the proliferation and osteogenesis of stem cells and osteoprogenitor cells by the release of specific cytokines, such as transforming growth factor (TGF-β), interferons, and interleukins (IL)[62,63]. In addition, bioprinted bone should possess osteoinduction capabilities by depositing bone-related proteins (e.g., BMPs, insulin-like growth factors, and fibroblast growth factor [FGFs])[64-66]. Last but not least, the structure should provide a porous microenvironment for the differentiation of bone cells, facilitating the synthesis of minerals, and collagenous tissue[67].

Some of the above-mentioned studies have shown that stem cell differentiation is preserved in bioprinted bone, and their osteogenic differentiation can be modulated by utilizing the right printing process, biomaterials, design, and bioactive factors[68]. The addition of BMP-2 and TGF-β usually results in greater osteogenic differentiation[69,70]. As to bioink materials, an alginate-based hydrogel supports the viability of MSCs and retains their osteogenic capacity in bioprinted bone[71]. MeHA has also been demonstrated to trigger osteogenic differentiation of hBMSCs in bioprinted constructs without exogenous osteogenic factors[72]. The GelMA scaffold was also demonstrated to induce mineralization of MG63 osteoblasts and primary normal human osteoblasts (NHOst) without requiring any additional osteogenic factors[73]. In addition, the HAp plays an important role in the osteogenic differentiation of preosteoblast cells in vitro[74]. A surge of mineralization in SaOS-2 cell-embedded hydrogel was observed on overlaying polyP·Ca²⁺-complex to the bioprinted alginate/gelatin hydrogel[75]. Furthermore, bone scaffolds that were composed of decellularized porcine bone had significantly enhanced osteogenic gene expression, without the use of an osteogenic medium[76]. Besides, the addition of biologically active ions like strontium (Sr²⁺) has been reported to enhance bone formation[77,78]. In addition, as an alkaloid component extracted from Chinese traditional medicines such as astragalus and coptis, berberine also has an osteogenic effect and antibacterial properties[79]. Stiffer biomaterials also trigger higher MSC osteogenic differentiation as the mechanical properties of the biomaterial affect cellular activity[72]. Moreover, it has been reported that exogenous electrical stimulation can increase mineralization[80]. In short, using appropriate physical and chemical cues, the stem cell activities, particularly the osteogenesis, can be controlled and enhanced.

2.6. Mechanical enhancement

Constructs bioprinted using natural hydrogels have a low modulus of compression (<10 kPa)[74,81] and degrade rapidly, losing most of their structural integrity within a short-time[82]. Despite the design of delicate gradients with different concentrations of GelMA hydrogel, mechanical stability could not be maintained for more than 21 days in culture as a result of the degradation of GelMA[52]. Furthermore, the polyP·Ca²⁺ coated alginate/gelatin construct lost its mechanical stability after a 5-day culture[75]. Young’s modulus (GPa) and tensile/compressive strength (MPa) of native bone are in some orders of magnitude higher than those of hydrogel-based bioinks. It is problematic to use hydrogel-only constructs in clinical applications, and thus, developing constructs with a high and retainable mechanical strength is urgent.

Blending mechanically strong particles within a hydrogel may provide some reinforcement to bioprinted constructs[51,74]. The compression modulus of cell-laden constructs containing alginate, polyvinyl alcohol (PVA), and HAp was 10.3 kPa, but this value dropped to 2.4 kPa after 14 days of culture[83]. An elastic modulus of 0.55 MPa has been measured for the bioprinted TCP-collagen constructs[84], which was less than the elastic modulus of trabecular bone (20 – 52 MPa)[85]. Using silk fibroin/ionic-doped β-TCP, a multilayer structure has been prepared, which provided slightly enhanced mechanical properties (static compression modulus of 0.66 MPa and dynamic mechanical properties of 2.17 – 3.19 MPa)[86]. Furthermore, nanozirconium dioxide powder was blended in PCL, and Young’s modulus and compressive strength increased by ~0.4 and 0.5 times, respectively, as compared with the PCL-only scaffold[87]. In another case, bredigite was mixed with nanosheets of graphene oxide which was reduced by bovine serum albumin, and the addition of reduced graphene oxide enhanced the mechanical properties of scaffolds[88].

For significant reinforcement of the mechanical properties of bioprinted constructs, thermoplastics or ceramic frames are usually 3D printed. The incorporation of thermoplastics not only enhances mechanical properties but also enables the creation of large-scale structures with very high fidelity and a high fiber resolution[89,90]. Using FDM of PLA to mimic the Haversian system of bone and SLA of GelMA to simulate the blood vessels, a scaffold was engineered that exhibited a similar mechanical strength to native bone, as indicated by the compressive modulus of ~0.38 GPa, while the elastic modulus of the vascular region was 10 – 30 kPa, offering an appropriate microenvironment for cell encapsulation[53]. Using PCL or PCL/TCP structures as support for cell-laden hydrogels, a compressive modulus of ~30 – 45 MPa was obtained using various PCL: TCP ratios[91]. In contrast to printing hydrogel and supporting frame separately, core/shell scaffolds were constructed, consisting of calcium-deficient hydroxyapatite (CDHA) core and cellular-laden alginate (shell)[92]. The integration of the CDHA core resulted in a significantly higher compressive modulus (7 MPa) than alginate-only scaffolds and preserved the structural integrity in vitro for 35 days.

2.7. Hypoxic culture

Hyperoxia is another factor that plays a significant role in bone development through the hypoxia-induced transcription factors (HIF), although there is debate about the impact of hypoxia on bone regeneration[93]. Studies have indicated that hypoxia enhances the osteogenesis of BMSCs[94-96], whereas others have suggested that hypoxia inhibits the growth and bone-forming ability of osteoblasts, such as the differentiation of MSCs into osteoblasts[97,98]. Hypoxia and HIF may have multiple roles in osteogenic induction, as suggested by the paradoxical conclusions. In addition, it is well known that hypoxia increases the expression of angiogenic factors in MSCs, such as VEGF[94]. An initial period of hypoxia is present during the process of bone regeneration in the body, which stimulated the deposition of several factors, such as VEGF and IL-6, and promoted vascularization later on[99]. It has been observed that short-term (7 days) hypoxic conditioning did not retard osteogenic differentiation of stromal vascular fraction-derived cells (SVFC) in bioprinted constructs, but it enhanced the vascularization of SVFC as indicated by increased expression of VEGFA and HIF1A[27]. Given the fact that the hypoxic environment is beneficial for capillary formation, but may also inhibit the differentiation of MSCs into osteoblasts[100-102], an ideal design for bone bioprinting is to provide a hypoxic environment with a controlled oxygen diffusion for the embedded human umbilical vein endothelial cells (HUVECs) so that the oxygen supply to the bone region is not compromised[53].

2.8. Dynamic culture

Mechanical stress continuously remodels bone in vivo, and it has been postulated that such stresses are mainly transferred to bone cells through fluid shear stresses[103]. During the loading of a bone, interstitial fluid flows through the pores in the bone, causing a shear stress to be sensed by osteocytes, which are then communicated to osteoblasts and osteoprogenitor cells through their neighboring ECM. During in vivo loading, bone cells experience shear forces ranging from 8 to 30 dyn/cm²[104,105]. Cells can be cultivated in bioreactors with a dynamic environment that mimics the growth conditions of bone, enhancing nutrient transport, exposing cells to fluid shear stresses, and ultimately promoting cell seeding efficiency and differentiation[106,107]. Bioreactors have been shown to benefit bone differentiation and mineralization through mechanical stimulation induced by fluid shear stresses [107-109]. In a study of bone bioprinting, hMSCs were shown to express significantly more COL-I and VEGF when exposed to culture media flowing at a rate of 5 mL/min than when exposed to static culture media[53]. The results of dynamic culture showed superior Ca deposition, an indication that shear stress aided osteogenesis and mineralization. The considerations for bone bioprinting are depicted in Figure 3.

Figure 3.

Considerations for bioprinting a bone construct.

3. Up-to-date progress of bone bioprinting

3.1. Extrusion-based bioprinting of bone

3.1.1. Alginate-based composite bioinks

As a popular biomaterial for bone regeneration, alginate-based composite bioinks have been extensively explored for bone bioprinting. Fedorovich et al.[71] conducted an early study in 2008 to examine the bioprinting of Lutrol^® F127, agarose, alginate, and methylcellulose hydrogels with an EBB bioprinter and observed that the applied extrusion conditions did not reduce the survival and differentiation capacity of BMSCs. This process was capable of developing bone constructs containing multiple cell types indicated by bioprinting of two fluorescently labeled cell populations within a single scaffold. The authors subsequently created porous constructs which encapsulated two types of cells, namely, EPCs and MSCs[58]. A rectangular 10-layer scaffold, which consisted of two parts (EPC-loaded Matrigel and MSC-loaded Matrigel with the addition of biphasic calcium phosphate [BCP]), demonstrated that cell distribution could be maintained after 2 weeks of culture. Furthermore, the MSC/BCP-laden Matrigel part demonstrated apparent bone formation in immune-deficient mice after 6 weeks of implantation, which demonstrated by Goldner’s trichrome and COL-I staining, while cartilage formation was evident in the MSC/Matrigel part as determined by Safranin-O and COL-II staining. In another work of the same group, different fiber spacings or angles of fiber deposition were used to fabricate heterogeneous constructs with adjustable porosity (35 – 66%) and elastic modulus (4.7 – 6.6 kPa)[81]. Osteochondral constructs were bioprinted using alginate containing chondrocytes (cartilage compartment) and alginate containing MSCs and osteoinductive BCP particles (bone compartment). As a result, the cell viability of alginate constructs containing human chondrocytes and osteogenic progenitors remained high (90%) during the printing process. Furthermore, chondrogenic and osteogenic differentiation has been demonstrated in different parts of the construct in vitro and in vivo (subcutaneously in immune-deficient mice).

In a study using a pneumatic EBB system, Loozen et al.[110] developed porous or solid constructs containing alginate hydrogel combined with MSCs, BCP particles, and plasmid-DNA-encoding bone morphogenetic protein-2. As a result of plasmid DNA transfection, osteogenic differentiation of cells was observed by enhanced BMP-2 and ALP production, and porous constructs displayed superior BMP-2 production to solid constructs. Bendtsen et al.[83] extensively explored the effect of varying materials with different concentrations on their printability, and thus effectively bioprinted mouse calvaria 3T3-E1 (MC3T3) cells within alginate-PVA-HAp bioink, which supported 96% cell viability, indicating that the suspension optimized the printability and improved the cellular activity. Cunniffe et al.[111] generated a bioink by combining an arginylglycylaspartic acid (RGD)-γ-irradiated alginate and nano-HAp blended with plasmid DNA (pDNA). The bioprinted constructs containing pDNA showed a higher level of mineralization than those without pDNA. The MSC-encapsulated constructs that were implanted subcutaneously into nude mice demonstrated greater levels of mineral deposition and vascularization compared to cell-free groups.

Cidonio et al.[112] developed a bioink composed of synthetic nanoclay (Laponite, LPN) and GelMA, which showed good fidelity and interlinked porosity during EBB. The hBMSCs maintained good viability in the bioprinted construct (86 ± 10% at 21 days of culture), and cell-laden constructs cultured without dexamethasone showed areas of mineralization, indicating the formation of osteogenic tissue. Subsequently, the same team used human BMSCs encapsulated in a Laponite^Ò-alginate-methylcellulose hydrogel to create bioprint scaffolds, showing 90% cell viability on day 7 of culture with decreased ALP expression and increased calcium mineral deposition over 21 days of culture[113]. MSC-laden scaffolds were implanted in athymic BALB/c mice, where extensive mineralization was observed after 4 weeks of implantation, while bone volume and bone density increased significantly from 2 to 8 weeks. Furthermore, subcutaneous implantation of BMP-2-containing bioprinted scaffolds in mice exhibited increased level of glycosaminoglycans (GAGs) deposition and mineralization.

Alginate is readily processable using cross-linking mechanisms such as ionic interaction, cell cross-linking, photopolymerization, and Schiff-base reaction[114]. The prevalent method for alginate gelation is to combine the alginate with divalent cations, such as Ca²⁺, Mg²⁺, Ferrous (Fe²⁺), Barium (Ba²⁺), or Sr²⁺. In the cell cross-linking mechanism, the ligands (e.g., RGD) were grafted onto alginate for cell adhesion[111]. As cells were loaded into the RGD-modified alginate, the receptors on the cell surface can bind to ligands of the modified alginate. In addition, some researchers exploited photopolymerization to solidify methacrylated alginate by employing UV irradiation[115]. Alginate was also prepared in the form of microspheres to deliver growth factors, proteins, and drugs in tissue engineering[114]. Inspired by the advantage of the alginate microsphere, Wu et al. embedded the alginate microspheres within the cell aggregates to generate porous tissue strands with high cell density[116]. The incorporation of alginate microspheres facilitated the permeation of oxygen and nutrition, promoting cell viability within the cell aggregates, which offers new insight into scaffold-free biofabrication.

3.1.2. Gelatin-based composite bioinks

As mentioned in section 2.2., bioinks whose main composite is gelatin have also been reported by several studies. Das et al.[56] used silk in a bioprinted bone study, where silk fibroin-gelatin (SF-G) bioink was embedded with hTMSC, and in situ cytocompatible gelation occurred (enzymatic cross-linking using mushroom tyrosinase and physical cross-linking using sonication). SF-G bioink caused no harm to the cells, and a higher chondrogenic and adipogenic potential was observed in the tyrosinase cross-linked groups, whereas a higher osteogenic differentiation was observed in the sonication cross-linked groups. Using six hydrogel blends of fibrin, gelatin, HA, and glycerol (F/G/H/Gl), Wehrle et al.[117] identified the best hydrogel blend which showed high cell viability (>96%) and cell proliferation. The osteogenic differentiation process was carried out in the hydrogels with immortalized hTERT-overexpressing MSCs, fetal MSCs from the umbilical cord MSCs, BMSCs, and ADSCs, among which ADSCs showed the greatest osteogenic differentiation potential indicated by the ALP assay, which were then chosen for bioprinting with the F/G/H/Gl hydrogel. Chiesa et al.[118] fabricated a vascularized bone model using gelatin-nanohydroxyapatite (Gel-nHAp), hMSCs, and HUVECs. Gel-nHAp was first extruded followed by hMSCs being seeded on scaffolds and osteogenically differentiated for 2 weeks. Next, lentiviral-GFP transfected HUVECs were placed into the macropores of 3D bioprinted scaffolds. The assembly of a complex capillary-like network was observed, and vascular lumen formation and osteogenic differentiation were confirmed by immunostaining and gene expression.

Although gelatin is easily accessible and has good biocompatibility, its low viscosity, low yield stress, and relatively long cross-linking time during or after bioprinting lead to poor shape retention properties, resulting in difficulty in creating 3D reliable structures with an interconnected pore network[118]. To overcome this obstacle, non-modified gelatin can be bioprinted below the melting temperature (usually below 4°C) to increase the viscosity for extrusion[119] or with the aid of a sacrificial frame to ensure sufficient time for cross-linking with cross-linkers, such as genipin and transglutaminase[118].

3.1.3. Alginate/Gelatin-based composite bioinks

To integrate the benefits of alginate (fast cross-linking) and gelatin (good biocompatibility), alginate/gelatin-based composite bioinks have drawn a lot of attention. As reported by Neufurth et al.[75], a polyphosphate Ca salt overlay (polyP·Ca²⁺-complex) was applied to the bioprinted alginate/gelatine/SaOS-2 cell scaffold to modulate the biological result of the construct. PolyP·Ca²⁺-complex was surprisingly found to significantly increase the ability of cells to proliferate in the underlining hydrogel. The hardness and mineralization of cell-laden alginate/gelatin hydrogels significantly increased when the overlaid polymer was present. A follow-up study by Wang et al.[120] investigated the effect of bioglasses including polyP·Ca²⁺-complex, silica, and biosilica on the mineralization of SaOS-2 cells embedded in a gelatine/alginate hydrogel. Bioglass particles significantly enhanced the mineralization ability of the entrapped cells, as evidenced by staining with alizarin red S, while element analysis of mineral nodules formed by the SaOS-2 cells indicated a gathering of minerals. Wüst et al.[51] took advantage of the thermal gelation of gelatin and the irreversible cross-linking of alginate to develop a two-step process. Cell viability was 85% after 3 days of culture after bioprinting when MSCs were mixed into the hydrogel precursor. Furthermore, a two-phased structure was produced with a HAp-containing phase and a tubular structure filled with alginate/gelatin to represent bone tissue and vascular structure, respectively.

Different from the blending of alginate and gelatin, alginate/gelatin composite hydrogel has been bioprinted using an in situ cross-linking mechanism, which is attributed to the Schiff-base reaction between aldehyde groups of oxidized alginate and amino groups of gelatin[114]. When gelatin was extruded into a support bath that contained oxidized alginate, freeform fabrication of complex 3D structures can be obtained[121].

3.1.4. Photocurable bioinks

Modification of biomaterials (such as gelatin, alginate, and HA) with methacrylate enables chemically cross-linking using UV radiation[122], which overcomes the above-mentioned limitations, such as low viscosity and slow gelation. Photopolymerization has also been integrated into EBB processes to enhance the stability and complexity of the bioprinted bone. In the work by Poldervaart et al.[72], MeHA has been investigated for bone bioprinting, where UV light exposure enhanced the storage modulus and elastic modulus of the MeHA, and hBMSCs residing in MeHA hydrogels retained 64% viability after a 21-day culture. Despite the absence of additional osteogenic stimuli, Ca deposition in hydrogels with higher MeHA concentrations (2.5 – 3%) was significantly greater than that observed in hydrogels with lower MeHA concentrations (1.5%). Byambaa et al.[52] used GelMA hydrogels containing different concentrations of VEGF to produce a gradient structure, and a gel rod in the center was bioprinted with GelMA, forming a perfusable vascular lumen with an endothelial lining. Cocultured HUVECs and hMSCs reached 93% cell viability after 7 days, and MSCs in the inner fibers differentiated into smooth muscle cells, which promoted vascular vessel formation. In the three outer layers, the MSCs and bioactive silicate nanoplatelets were embedded in GelMA-VEGF bioink, which induced osteogenic differentiation in vitro. The Alizarin Red S staining, immunostaining, and RT-qPCR results demonstrated that a mature bone with angiogenesis was generated after 21 days of culture.

3.1.5. Bioinks with particle reinforcement

As biomaterials such as silicate and calcum phosphates are well-known for their roles in increasing mechanical properties and mineralization of bone scaffolds, such materials in the form of particles are supplemented in bioinks. A recent study by Liu et al.[123] encapsulated rat BMSCs in a bioink containing nanosilicate (nSi), gelatin, and alginate, in which the addition of nSi resulted in low hydrogel swelling and improved shear-thinning character, storage and loss moduli, and compressive modulus of hydrogels. Hydrogels containing 2% nSi were used for bioprinting and cells could retain their viability and proliferation, accompanied by increased ALP activity and increased gene expression of Runt-related transcription factor 2 (RUNX2), osterix, ALP, collagen type I alpha 1 (COL1A1), OCN, and osteopontin (OPN). Critical-sized cranial defects of Sprague Dawley (SD) rats were also transplanted with scaffolds, and micro-CT, Van Gieson, histochemistry, and Masson’s trichrome staining results demonstrated that the MSCs-loaded 2% nSi scaffold showed better outcomes than the control groups (without scaffolds or nSi). Using a nanoengineered bioink consisting of GelMA, kappa-carrageenan, and nSi, which formed an ionic-covalent entanglement network, Chimene et al.[124] succeeded in bioprinting a hollow cylinder with 150 layers (3 cm in height). The presence of nSi resulted in the increased compression modulus and the significant deposition of GAGs, proteoglycans, Ca, and phosphate in human MSCs post-bioprinting, demonstrating that the bioink could induce endochondral differentiation of MSCs.

In a study by Demirtaş et al.[74], chitosan, chitosan-HAp composite, and MC3T3-E1 cells were printed using an EBB, in which alginate and alginate-HAp hydrogels were used as comparisons. With the incorporation of HAp, both chitosan and alginate hydrogels had 3 – 6 folds greater elastic modulus. Spherical morphology of cells was seen in alginate and chitosan hydrogels on day 7 of culture, whereas spread cells were observed in alginate-HAp and chitosan-HAp hydrogels, indicating a positive effect of HAp on cellular morphology. Cells within chitosan-HAp hydrogel expressed higher levels of osteogenic proteins and Ca deposition as compared to other groups. Using BMSCs, RAW264.7 macrophages, gelatin, GelMA, polyethylene glycol (PEG), and BMP-4-loaded mesoporous silica nanoparticles (MSNs) as bioinks, Sun et al.[125] demonstrated that MSNs improved the mechanical and shear-thinning properties of the hydrogel. In vitro and in vivo (implantation in calvarial defects of DM rats) tests were conducted using RAW/BMP-4 scaffolds, where BMP-4-loaded MSNs induced macrophage polarization and inhibited the inflammatory response. Moreover, the BMSC/RAW/BMP-4 group was tested in vitro, which indicated that the expression of RUNX2, OPN, and ALP was up-regulated. The in vivo results, further, highlighted the superiority of the BMSC/RAW/BMP-4 group over others regarding new bone volume and degree of neovascularization.

In light of the fact that amorphous magnesium phosphates (AMPs) stimulate rapid differentiation and mineralization of pre-osteogenic cells, a shear-thinning and printable ECM-based bioink containing 2% octapeptide FEFEFKFK with AMP particles (ECM/AMP bioink) were developed[126]. Dental pulp stem cells were added to ECM/AMP bioink and maintained 90% viability after 5 days in bioprinted constructs. In comparison with the AMP-free control group, ECM/AMP constructs showed greater mineralization and increased RUNX2, OPN, and COL1A1 mRNA expression at 21 days without the presence of growth factors. Eight weeks after implanting ECM/AMP constructs in rat cranial defects, a remarkable increase in bone density as well as new bone formation were observed in the defect. Using a microparticulate bioink composed of poly(lactic-co-glycolic acid) (PLGA), PEG, and carboxymethyl cellulose, Sawkins et al.[127] achieved a bioprinted construct with mechanical properties comparable to those of human cancellous bone (Young’s modulus: 57.3 MPa) and the pore sizes of 65 – 77 μm. Furthermore, the bioprinted constructs released lysozyme for 15 days, and a high level of protein activity was observed for 9 days.

3.1.6. Hybrid constructs with mechanical reinforcement

As mentioned in section 2.6., to obtain a clinically relevant mechanical strength, hybrid constructs are usually fabricated by EBB, such as using a dual-nozzle setup to print mechanical supporting frames along with cell-laden bioinks and coaxial nozzle to print core-shell filaments (Figure 2A). Lee et al.[44] demonstrated a hybrid scaffold containing PCL and cell-embedded alginate fibers, where the alginate fibers provided biological functionality to the construct, while the PCL fibers regulated the mechanical properties. Such a design resulted in a notable increase in Young’s modulus and ultimate tensile strength compared to alginate-only scaffolds. Furthermore, an integrated tissue-organ printer able to handle several biomaterials (including gelatin, fibrinogen, HA, and PCL) was established by Kang et al.[91] to construct human-scale and mechanically-stable tissues. In the process of reconstructing the mandible bone, authors bioprinted an amniotic fluid-derived stem cell-laden hydrogel, a mixture of PCL and TCP, as well as Pluronic F127. Furthermore, rat calvarial bone constructs were generated in a circular shape using human amniotic fluid-derived stem cells (hAFSCs), revealing restored vascularized bone without necrosis at all implant regions, while the control group treated with scaffold only had negligible bone tissue formation. Kuss et al.[27] adopted SVFCs, which maintained the characteristics of ECs, in creating bone constructs with PCL/HAp and SVFC-laden MeHA/GelMA bioinks. In a short-term (<21 days) hypoxic culture in vitro, osteogenic differentiation of SVFCs was not affected, which, in turn, promoted vascular gene expression of VEGFA and HIF1A. In athymic mice, short-term hypoxia (7-day hypoxia and 14-day normoxia) promoted microvessel formation in vivo at 4 weeks and integration with the host vascular network but did not affect osteogenic differentiation of SVFCs, representing the beneficial effects of short-term hypoxia in bone regeneration.

With coaxial EBB, Raja et al.[92] created 3D scaffolds that used CDHA (core) and MC3T3-E1 cell-laden alginate (shell), thereby avoiding conventional sintering of ceramics after simultaneous printing of CDHA and the cell-laden hydrogel. Compared to alginate-only scaffolds (0.3 MPa), core/shell scaffolds have a higher compressive modulus (7 MPa), and CDHA-only scaffolds disintegrated after compression, showing that core/shell scaffolds balance the poor mechanical properties of hydrogel with the brittleness of ceramic. Kim et al.[84] fabricated an α-TCP/collagen cell-laden scaffold with MC3T3-E1 cells, where a layer of porous α-TCP/collagen fibers without cells was extruded for mechanical stability, followed by deposition of collagen bioinks onto the porous layer, and this process was repeated until a 3D scaffold was obtained. On the bioprinted scaffolds, more viable cells and a more homogenous distribution of cells were found compared with the controls, namely, the cell-laden collagen-only scaffold and α-TCP/collagen scaffold with dipping-loaded cells. Ahlfeld et al.[128] prepared biphasic scaffolds by alternately extruding two materials, namely, the cell-laden alginate-methylcellulose blend (ALG/MC) and calcium phosphate cement (CPC). The compression modulus of the biphasic scaffold containing 50% CPC and ALG/MC was much higher than that of the monophasic scaffolds of ALG/MC (31 ± 9 MPa vs. 37 ± 5 kPa). Osteochondral scaffolds were generated with calcified cartilage between CPC and ALG/MC zones, which resembled articular cartilage and subchondral bone, respectively.

Zhai et al.[129] developed a biodegradable material (bioink A) composed of poly(ethylene glycol) diacrylate (PEGDA) and laponite nanoclay (PEG-Clay). A structure was created using a two-channel 3D bioprinting method together with another composite (bioink B) composed of rat osteoblasts (ROBs) loaded within HA. Bioink A improved the mechanical properties and cell adhesion, and the release of magnesium and silicon bioactive ions was conducive to the osteogenic differentiation of cells. The bioink B ensured the uniform distribution of cells in the scaffold and a high survival rate (e.g., >95% after 1 day). In vitro experiments showed that the ALP activity of ROBs within the PEG4K-Clay scaffolds was remarkable. In vivo tibia repair showed that the regenerated bone size using ROB-laden PEG4K-Clay scaffolds was significantly larger than that of other groups.

Table 1 summarizes the studies on extrusion-based bone bioprinting and the properties of the bioprinted bone constructs.

Table 1.

Summary of the studies on extrusion-based bone bioprinting and properties of the constructs

Study	Materials	Cell source	Animal model	Pore size (porosity)	Mechanical reinforcement	Compressive modulus	Viability	Zonal structure
Fedorovich et al.[71]	Alginate, Lutrol F127, PEO-PPO-PEO block copolymer, Matrigel, agarose, methylcellulose	BMSCs	---	300 μm	---	---	~70 – 95%	---
Fedorovich et al.[58]	Matrigel, alginate, BCP, Lutrol® F127	EPC, MSCs	Female nude mice	---	---	---	---	EPC-laden Matrigel part on the left and MSC-laden Matrigel part with BCP on the right
Fedorovich et al.[81]	Alginate, BCP	MSCs, chondrocytes	Female nude mice	0.8, 1.5, 2.0, and 2.5 mm (0, 35%, 48%, and 66%)	---	6.4 – 14.8 kPa	89%	MSC-laden alginate+BCP on the left and chondrocyte-laden alginate on the right
Loozen et al.[110]	Alginate, calcium phosphate, plasmid DNA encoding BMP-2	MSCs	Mice	1 mm (40%)	---	Porous: 5.6 kPa Solid: 7.3 kPa	80 – 90%	---
Bendtsen et al.[83]	Alginate, PVA, HAp	MC3T3-E2	---	---	---	8.6 – 10.3 kPa	~78 – 96%	---
Cunniffe et al.[111]	Alginate (RGD-γ-irradiated), nHAp, PCL, pDNA, BMP-2, TGF-β3	MSCs	Nude mice	---	---	---	~100%	PCL supporting mesh
Cidonio et al.[112]	Nanoclay LPN, GelMA	hBMSCs	---	---	---	---	86 ± 10% at 21 days	---
Cidonio et al.[113]	LAP, alginate, methylcellulose BMP-2, VEGF	hBMSCs, HUVECs	Male athymic immunodeficient mice	---	---	---	~90%	---
Das et al.[56]	Silk fibroin, gelatin	hTMSCs	---	1 mm	---	---	>86%	---
Wehrle et al.[117]	Fibrin, gelatin, alginate agarose, collagen, HA, glycerol Hap	iMSCs, ucMSCs, BMSCs, ADSCs	---	---	---	---	~97%	--
Chiesa et al.[118]	Type A gelatin, nHAp, genipin	hMSCs, HUVECs	--	--	--	36.4 ± 9.6 kPa	--	--
Neufurth et al.[75]	Alginate, gelatin, agarose, polyphosphate+CaCl2 Complex	SaOS-2	---	---		---	92%	---
Wang et al.[120]	Alginate, gelatin, polyphosphate+CaCl2 complex, silica, biosilica	SaOS-2	---		---	---	---	---
Wüst et al.[51]	Alginate, gelatin	MSCs	---	---	HAp	With 8% HAp: 36 kPa; With 4% HAp: 32 kPa; Without HAp: 29 kPa.	85%	---
Poldervaart et al.[72]	MeHA	hBMSCs	---	1 mm	---	1.3 – 10.6 kPa	~60 – 75%	---
Byambaa et al.[52]	GelMA, VEGF, silicate nanoplatelets	hMSCs and HUVECs	---	---	---	5.7 – 6.5 MPa	~50 – 93%	Bone 3D architecture containing osteogenic and vasculogenic niches
Liu et al.[123]	nSi, gelatin and alginate	rBMSCs	SD rats	---	nSi	0%nSi: 46 kPa; 1%nSi: 58 kPa; 2%nSi: 101 kPa; 3%nSi: 214 kPa	---	---
Chimene et al.[124]	GelMA, kCA, nSi	hMCS	---	---	nSi	103 kPa	---	---
Demirtaş et al.[74]	Chitosan, alginate, HAp	MC3T3-E1	---	100 – 200 μm	HAp particles	Without HAp: 3.5 – 4.6 kPa; With HAp: 15 – 19 kPa	>90%	---
Sun et al.[125]	Gelatin, GelMA, PEG, MSNs, BMP-4	BMSCs, RAW264.7 macrophages	Male SD rats	---	MSNs	0% MSNs: 121 kPa 0.4% MSNs: 195 kPa 0.8% MSNs: 233 kPa	>90%	---
Dubey et al.[126]	Octapeptide FEFEFKFK, AMP	DPSCs	Male rats	---	---	---	~90%	---
Sawkins et al.[127]	PLGA, PEG, CMC, Pluoronic F-127	MSCs	---	65 – 76 μm, (10.8 – 12.4%)	---	54.4 – 57.3 MPa	50 – 87%	---
Lee et al.[44]	PCL, alginate	MC3T3-E1	---	300 – 500 μm	PCL struts	8.3 – 15.4 MPa (tensile)	84%	---
Kang et al.[91]	Gel composite (gelatin, fibrinogen, HA and glycerol), PCL, TCP, Pluronic F127	hAFSCs, rabbit ear chondrocytes, Mouse C2C12 myoblasts	Rat and athymic mice	70%	PCL, PCL/TCP structures	~30 – 45 MPa	>91%	---
Kuss et al.[27]	MeHA, GelMA and HA, PCL/Hap	SVF derived cells	Mice	400 μm	PCL/HAp frame	~55 MPa	---	---
Raja et al.[92]	TCP, alginate	MC3T3-E1	---	1.7 mm	TCP strust in the core	Core/shell scaffold: 7 MPa, Alginate scaffold: 0.3 MPa	---	---
Kim et al.[84]	TCP, COL-I	MC3T3-E1	---	~500 μm	TCP/collagen struts	Collagen: 0.04 MPa; Construct: 0.55 MPa	>91%	---
Ahlfeld et al.[128]	CPC, ALG/MC	hMCSs	--	--	--	30.7 ± 9.1 MPa	--	--
Zhai et al.[129]	PEGDA, Laponite XLG nanoclay, HA	ROBs	Male rats	--	Laponite XLG nanoclay	20% PEG4K-7% Clay: 0.47 MPa 20% PEG10K-7% Clay: 0.98 MPa	>90% at 1 day	--

PEO: Polyethylene glycol, PPO: Poly (phenylene oxide), BCP: Biphasic calcium phosphate, DNA: Deoxyribonucleic acid, BMP-2: Bone morphogenetic protein 2, HAp: Hydroxyapatite, PLGA: Poly (lactic-co-glycolic acid), PEG: Poly (ethylene oxide), CMC: Carboxymethyl cellulose, COL-I: Collagen type I, HA: Hyaluronic acid, PCL: Polycaprolactone, TCP: Tricalcium phosphate, PLA: Polylactic acid, GelMA: Gelatin methacrylate, VEGF: Vascular endothelial growth factor, MeHA: Methacrylated hyaluronic acid, PVA: Polyvinyl alcohol, RGD: Arginylglycylaspartic acid, nHAp: Nano-hydroxyapatite, pDNA: Plasmid DNA, TGF-β3: Transforming growth factor-beta 3, nSi: Nanosilicate, kCA: Kappa-carrageenan, AMP: Amorphous magnesium phosphate, LAP: Laponite, CPC: Calcium phosphate cement, PEGDA: Poly ethylene glycol diacrylate, ALG/MC: Alginate-methylcellulose blend, LPN: Laponite, MSNs : Mesoporous silica nanoparticles, BMP-4: Bone morphogenetic protein 4, MSCs: Mesenchymal stem cells, HUVECs: Human umbilical vein endothelial cells, hTMSCs: Human nasal inferior turbinate tissue-derived MSCs, hAFSCs: Human amniotic fluid-derived stem cells, SVF: Stromal vascular fraction, NHOst: Normal human osteoblasts, iMSCs: Immortalized hTERT-overexpressing MSCs, ucMSCs: Fetal MSCs from umbilical cord, BMSCs: Bone marrow stem cells, ADSCs: Adipose-derived stem cells, rBMSCs: Rat BMSCs, hMSCs: Human MSCs, DPSCs: Dental pulp stem cells, SD: Sprague Dawley, ROBs: Rat osteoblasts

3.2. Droplet-based bioprinting of bone

Campos et al.[130] applied DBB to bone reconstruction, and thermo-responsive agarose hydrogels with COL-I were inkjet-printed at different concentrations. By increasing the agarose content in agarose/collagen composites, more precise contours could be captured in the bioprinted tissue, since the viscosity and mechanical stiffness of the bioinks were improved, therefore facilitating their printability. However, softer hydrogels resulted in the greater elongation of MSCs after osteogenic differentiation. Cells survived in the bioprinting and preserved the mesenchymal phenotype, as confirmed by high cell viability (>98%), immunostaining of vimentin+ and CD34−, ALP activity, and bone-related gene expression.

3.3. Laser-based bioprinting of bone

Using the digital light processing-based 3D bioprinting technique, Anada et al.[131] generated a construct that consists of two rings, which are an outer ring of octacalcium phosphate (OCP) and GelMA imitating the cortical shell as well as an inner ring of HUVEC spheroids and GelMA imitating the bone marrow. On increasing the concentration of OCP, MSC differentiation was enhanced, resulting in increased ALP activity, and the vascularization was observed in hydrogels, which contained HUVEC spheroids in combination with GelMA.

Bernal et al.[132] developed an optical-tomography-inspired volumetric printing (VBP) apparatus, which enables structures of complex shapes to be bioprinted within seconds, and used in bone remodeling. This method overcomes the limitations of bioprinting such as a long printing time and difficulty of printing complex structures. It was used to bioprint an anatomical bone trabecular model containing MSCs, in which the smallest feature was 145 μm, and endothelial colony-forming cells (ECFCs) and MSCs were further added into the pores to form a heterocellular structure, where the vascular formation was seen after a 3-day culture.

3.4. Aspiration-assisted bioprinting of bone

Heo et al.[133] used AAB technology to control the precise printing of osteogenically differentiated hMSC/HUVEC spheroids. The method allowed the printed spheroids to tightly self-assemble and retain their original shape without disintegration. The survival rate of cells in hMSC/HUVEC spheroids was significantly higher than that in hMSC-only spheroids. The addition of HUVECs also promoted the osteogenic differentiation of cells. The hMSC/HUVEC spheroids showed a higher expression of COL1, ALP, and bone sialoprotein (BSP) compared to control groups (2D cultured hMSC and hMSC-only spheroids). This strategy provided a new possibility for printing bone tissue with anatomically-relevant cell density. Table 2 summarizes the above-mentioned studies on bone bioprinting using DBB, LBB, and AAB and the properties of the bioprinted bone constructs.

Table 2.

Summary of the studies on bone bioprinting using DBB, LBB, and AAB

Study	Technology	Materials	Cell source	Animal model	Pore size (porosity)	Mechanical reinforcement	Compressive modulus	Viability	Zonal structure
Duarte Campos et al.[130]	Inkjet	COL-I, agarose	MSCs	---	---	---	18.1, 53.1 and 89.1 kPa	98% at 21 days	---
Anada et al.[131]	SLA	OCP, GelMA	HUVECs	---	---	---	---	---	Peripheral OCP-containing GelMA ring + central GelMA ring containing HUVEC spheroids
Bernal et al.[132]	VBP	GelMA	ACPCs, MSCs, ECFCs	---	---	---	266 kPa	>85%	---
Heo et al [133]	AAB	Cell spheroid	hMSCs, HUVECs	---	---	---	---	>85%	---

SLA: Stereolithography, VBP: Volumetric bioprinting, AAB: Aspiration-assisted bioprinting, COL-I: Collagen type I, GelMA: Gelatin methacrylate, OCP: Octacalcium phosphate, MSCs: Mesenchymal stem cells, HUVECs: Human umbilical vein endothelial cells, ACPCs: Articular Cartilage-resident chondroprogenitor cells, ECFCs: Endothelial colony forming cells, hMSCs: Human MSCs

3.5. Hybrid bioprinting of bone

To replicate a bone construct with interior vascular networks, hybrid bioprinting processes which integrated EBB and LBB or DBB have been reported. Cui et al.[53] used a hybrid bioprinting platform composed of FDM and SLA. Using this technique, they deposited polylactide (PLA) fibers and GelMA hydrogel to reproduce a vascularized Haversian system of bone tissue. In the manufacturing process, the polydopamine (pDA)-coated PLA scaffold (bone region) was immobilized with BMP-2 peptides, and then, VEGF peptides were conjugated to GelMA chains (vascular region). PLA scaffolds were pre-seeded with hMSCs, whereas hMSCs and HUVECs were embedded at a 1:1 ratio in the GelMA hydrogel, where the pDA- and BMP-2-modification was found to promote cell growth and spread on the scaffold and the BMP-2 benefited osteogenesis. In a dynamic culture of perfusion, notable osteogenesis and the formation of vascular networks can be observed, as indicated by the higher expression of COL-I, Ca deposition, and VEGF than in static culture.

Combined bioprinting of ADSCs and HUVECs has also been reported to promote vascularization in bioprinted bone tissue[134]. ADSCs underwent osteogenic differentiation for 7 days before being added into the hydrogel composite (containing fibrinogen, gelatin, HA, glycerol, HAp, VEGF, basic fibroblast growth factor [bFGF], and aprotinin), which was subsequently extruded. An array of droplets was deposited through drop-on-demand bioprinting of HUVECs-loaded hydrogel composite (including fibrinogen, VEGF, bFGF, and aprotinin) after every second layer of ADSC-laden bioink was printed. The bioprinted construct was enclosed by a PCL frame to ease transportation and was, then, implanted subcutaneously into mice. On 12-day implantation, large volumes of calcified ECM, as well as blood vessels with lumens, were observed. Table 3 summarizes the above-mentioned studies on bone bioprinting using hybrid processes and the properties of the bioprinted bone constructs.

Table 3.

Summary of the studies on bone bioprinting using hybrid processes

Study	Technology	Materials	Cell source	Animal model	Pore size (porosity)	Mechanical reinforcement	Compressive modulus	Zonal structure
Cui et al.[53]	Extrusion+ SLA	PLA, GelMA, BMP-2 and VEGF peptides	hMSCs and HUVECs	---	260 μm (20%)	PLA fibers	Construct: 0.38 GPa; Hydrogel: 10 – 30 kPa	Vascularized construct with capillary networks
Rukavina et al.[134]	Extrusion + DoD	Fibrinogen, Gelatin, HA, Glycerol, VEGF, bFGF, HAp, PCL	HUVECs ADSCs	SCID mice	---	---	---	PCL frame

SLA: Stereolithography, DoD: Drop-on-demand, PLA: Polylactic acid, GelMA: Gelatin methacrylate, BMP-2: Bone morphogenetic protein 2, VEGF: Vascular endothelial growth factor, HA: Hyaluronic acid, bFGF: basic fibroblast growth factor, HAp: Hydroxyapatite, PCL: Polycaprolactone, hMSCs: Human mesenchymal stem cells, HUVECs: Human umbilical vein endothelial cells, ADSCs: Adipose-derived stem cells

4. Conclusion and future prospects

The advancement of bone bioprinting with the use of various hydrogels, cell types, and other osteoconductive components has recently been remarkable, with several studies presenting mineralized structures with the vascularized network. It takes a considerable period (until the healing and remodeling process is complete) to create a durable structure with appropriate mechanical properties and physiological functions, which requires a balance between bioink degradation and tissue regeneration[135]. It is also challenging to develop large-scale vascularized bone grafts that meet the requirements of the clinical setting and easy integration with host tissue[136]. As discussed herein, researchers are striving to provide solutions to tackle the limitations in bone bioprinting, and it is recommended that constant efforts are made to integrate acellular scaffolds with cellular bioinks to modify the mechanical properties of bioprinted bone[13]. It is anticipated that more advanced processes and bioprinters will be developed that can deposit both cellular and non-cellular biomaterials with high efficiency, biocompatibility, and control over printing conditions (e.g., temperature, oxygen tension, and humidity). The piezoelectric nature of bone aids in promoting bone adaptation and remodeling by electromechanical mechanisms[137]. Hence, the addition of electromagnetic materials as a means of controlling the microenvironment of bioprinted bone might prove to be an effective avenue in the future when it comes to discovering new bioinks for bone bioprinting. For instance, currently available bioinks can be incorporated with conduction polymers, such as carbon nanotubes and graphene, to provide structural stability, guide cell growth, as well as stimulate bone formation[138]. A large number of biomaterials are expected to be developed in the future to achieve functional 3D bone bioprinting, which can be achieved with the integration of multidisciplinary knowledge and continuous financial support.

Funding

This work was funded by the National Natural Science Foundation of China (grant number 52205305), the GuangDong Basic and Applied Basic Research Foundation (grant number 2020A1515110724), and the Open Foundation of the State Key Laboratory of Fluid Power and Mechatronic Systems (grant number: GZKF-202004).

Conflict of interest

The authors declare no potential conflicts of interest concerning the research, authorship, and/or publication of this article.

Author contributions

Conceptualization: Yang Wu

Funding acquisition: Yang Wu

Visualization: Yang Wu, Ming Li

Writing - original draft: Yang Wu, Ming Li, Hao Su

Writing - review & editing: Yang Wu, Ming Li, Hao Su, Huaying Chen, Yonggang Zhu