The Long-Road to Develop Custom-Built Livers: Current Status of 3D Liver Bioprinting

Abstract

While liver transplantation is the gold-standard therapy for end-stage liver disease, the shortage of suitable organs results in only 25% of waitlisted patients undergoing transplants. 3D bioprinting is an emerging technology and a potential solution for personalized medicine applications. This review highlights existing 3D bioprinting technologies of liver tissues, current anatomical and physiological limitations to 3D bioprinting of a whole liver, and recent progress bringing this innovation closer to clinical use. We reviewed updated literature across multiple facets in 3D bioprinting, comparing laser, inkjet, and extrusion-based printing modalities, scaffolded versus scaffold-free systems, development of an oxygenated bioreactor, and challenges in establishing long-term viability of hepatic parenchyma and incorporating structurally and functionally robust vasculature and biliary systems. Advancements in liver organoid models have also increased their complexity and utility for liver disease modeling, pharmacologic testing, and regenerative medicine. Recent developments in 3D bioprinting techniques have improved the speed, anatomical and physiological accuracy, and viability of 3D bioprinted liver tissues. Optimization focusing on 3D bioprinting of vascular system and bile duct have improved both the structural and functional accuracy of these models, which will be critical in the successful expansion of 3D bioprinted liver tissues towards transplantable organs. With further dedicated research, patients with end-stage liver disease may soon be recipients of customized 3D bioprinted livers, reducing or eliminating the need for immunosuppressive regimens.

INTRODUCTION

Chronic liver disease (CLD) affects 1.5 billion individuals worldwide¹ and represents a substantial economic burden.² End-stage liver disease (ESLD) is the inevitable endpoint of untreated CLD, with unfortunately very few treatment options. Liver transplantation remains the only definitive treatment for ESLD, however, this therapy is severely constrained by the lack of transplantable organs.³ Approximately 75% of patients on the transplant waiting list fail to undergo transplantation as a result of this organ shortage.³ Despite the increased use of marginal livers partly thanks to machine perfusion technology and live donor livers, alternative therapies are still critically needed to overcome this gap. Several approaches have been developed as temporary solutions;⁴ however, no definitive alternative therapy currently exists. There has been extensive research into 3D bioprinting as a potential definitive solution to ESLD.⁵ The following review details the anatomical and physiological constraints to 3D bioprinting of whole livers, the current 3D bioprinting technological modalities, and the recent advances made in the field to bring this promising technology to clinical application.

Anatomical and Physiological Considerations in Liver 3D Bioprinting

The liver is the largest solid organ in the human body and is exceptionally important in maintaining organismal homeostasis. It is responsible for synthesizing crucial compounds necessary for the coagulation cascade, metabolizing nutrients and toxins, regulating blood glucose, and producing bile for emulsifying ingested fats.⁶ This wide array of physiological functions in conjunction with its enormous size makes it a challenging endeavor to bioprint an anatomically and physiologically functional human liver.

Macroscopically, the liver is divided into four anatomical lobes: left, right, caudate, and quadrate.⁷ The vascular supply to the liver is comprised of the portal vein and hepatic arteries, which contribute approximately two-thirds and one-third, respectively, of the nutrient-rich blood to the liver.⁷ Additionally, the liver is responsible for synthesizing and transporting bile through the biliary system, which is macroscopically oriented in parallel with the portal venous and hepatic arterial blood supply. Following metabolic and synthetic processing, blood is returned to the systemic circulation by way of the right, middle, and left hepatic veins which empty into the inferior vena cava.⁷

Functionally, the liver is divided into eight segments, each of which contains branch(es) of the portal vein, hepatic artery, and biliary system.⁸ Each macroscopic segment is further organized in two different ways: 1) hepatic lobules and 2) hepatic acini, as shown in Figure 1. Hepatic lobules consist of a hexagonal orientation of hepatocytes around a central venule, with branches of the portal triad (portal vein, hepatic artery, and bile ductules) located at each corner of the hexagon.⁷ Portal and arterial blood diffuse centripetally around the portal triad through hepatic sinusoids, which are lined by endothelial and Kupffer cells and ultimately deliver nutrients and toxins to hepatocytes for processing via the space of Disse.⁹ The space of Disse is an important microscopic compartment composed of type III collagen and hepatic stellate cells, which regulate sinusoidal blood flow, fibrogenesis, and vitamin A storage.⁹ Processed blood then travels centrally to the central venule, and subsequently drains to the systemic circulation via the inferior vena cava. Hepatic acini, on the other hand, are conceptualized from the perspective of the portal triad, with central venules located at the periphery. This is further arranged into three zones of concentric circles around the portal triad representing decreasing oxygen and nutrient gradients.⁹ Bile duct cells, or cholangiocytes, are oriented in an apicobasal fashion such that bile is exported towards the portal triad, draining into intrahepatic bile ducts and eventually into the extrahepatic right and left hepatic ducts, ultimately converging with the cystic duct to become the common bile duct.⁷

Figure 1. Macroscopic and microscopic anatomy of the liver:

Top: (left) Hepatic arterial blood supply, (middle) portal venous blood supply, (right) venous drainage. (bottom) Macro- and microscopic liver anatomy (image created by biorender.com)

The hierarchical organization of the liver into anatomically and functionally distinct units allows for highly efficient synthetic and metabolic functions. However, its complexity represents a distinct challenge when attempting to recreate the biology in a 3D bioprinted liver. There are three key challenges to address: (i) incorporating all three vascular systems and a complex biliary system into a single 3D bioprinted organ, (ii) producing a physiologically functional micro-macroscopic interface including functional zonation units, and (iii) maintaining high levels of metabolic and synthetic function within 3D bioprinted hepatic parenchymal units. To create an anatomically and physiologically functional 3D bioprinted liver, optimization of existing 3D bioprinting technology is essential.

Current 3D Bioprinting Technology

Bioprinting is a subset of 3D printing that strategically deposits biomaterial and biochemicals to engineer tissues or fabricate organs to address the increasing demand for the modeling of living constructs with mechanical stability and biological functionality.¹⁰ Since its inception, 3D bioprinting technology has been incorporated into a diverse array of applications, such as tissue engineering and regenerative medicine, toxicology screening, and complex tissue modeling.¹¹ The following section details current 3D bioprinting techniques, the relative advantages and disadvantages in the context of 3D liver bioprinting (Figure 2), the most prevalent constituent cell types used in 3D bioprinted liver constructs, and recent improvements to overcome historical limitations.

Figure 2. 3D Bioprinting Techniques.

ECM = extracellular matrix, dECM = decellularized ECM. (image created by biorender.com)

Inkjet Bioprinting:

Inkjet bioprinting is a technique in which single droplets of cells or biomaterial are patterned onto a substrate through a nozzle using thermal, piezoelectric, acoustic, or electrostatic forces.¹² This method rapidly (~ 200 Hz) fabricates high-resolution structures through modulations in drop size (with diameters < 50μm) with more precision than extrusion-based bioprinting.^13–15 Multiple nozzles may be employed simultaneously to quickly print biomaterials of varying composition in the creation of a large, advanced model.¹⁶ Additionally, these bioprinters are the most cost-efficient, high-speed, and biocompatible – maintaining high cell viability and function after printing.¹⁷ In fact, this technique is currently used in regenerative medicine applications to 3D print skin and cartilage in situ.^18,19 The main limitation of this technique is the necessity for a low-viscosity biomaterial other than cells in order to avoid clogging the nozzle.²⁰ This has major implications for the ultimate stability of the printed structure, as low-viscosity biomaterials tend to lack structural integrity.²⁰ This disadvantage is particularly cumbersome in the context of 3D liver bioprinting, as the liver is a solid organ with inherently robust structural integrity. To mitigate this limitation, several approaches have been implemented. Skardal et al. utilized a 1:1 fibrin/collagen solution mixed with a unique thrombin solution to induce chemical cross-linking, which resulted in a bioink with sufficient structural stability to appropriately model mouse wound-healing.¹⁸ Suntornnond et al. described a two-step modification of gelatin-methacryloyl (GelMA) including saponification and heat treatment to decrease the weight and viscosity of the bioink while retaining its ability to crosslink physically and chemically.²¹ However, this improvement in printability was obtained at the expense of mechanical strength. Additionally, in efforts to create biomimetic articular cartilage, Dufour et al. attempted a novel fabrication method by combining inkjet bioprinting with melt electrowriting (MEW)²², which uses voltage-stabilized jets to place microfibers in pre-determined locations.²³ By doing so, researchers created a more biologically accurate structure that maintained strength, yet increased flexibility. These studies emphasize the capabilities of inkjet bioprinting in current medicine and the different combinations of biofabrication methods that may be combined. With further optimization of bioink composition, inkjet bioprinting may prove to be an effective approach to solid 3D liver bioprinting.

Laser Bioprinting:

Laser-induced forward transfer (LIFT) is a bioprinting technique in which a pulsing laser beam is applied to an energy-absorbing layer (typically a metal), the donor substrate and the desired bioprint, whose droplets eventually accumulate on the receiving substrate placed below.¹² It is a precise, nozzle-free mechanism that includes the printing of bioinks across a wide range of viscosities at high speeds (1–100 kHz) while preserving cell viability.²⁴ These prints provide high resolution at cell concentrations up to 1×10⁸ cells/ml.²⁵ Koch et al successfully printed multiple cell lines in the creation of epidermal tissue, which proved to be resistant to damage from the laser.^26,27 Further, Catros et al demonstrated that optimized laser settings contribute to increased cell viability, such as increased film thickness and bioink viscosity and decreased laser energy.²⁸ This technique has advantages over inkjet bioprinting primarily with respect to biomaterial composition, print resolution, and cell viability. The primary limitations to this technique are that it is slow, highly expensive, and cumbersome.¹² Thus, this technique is ideal for printing features within an organ that demand high resolution, such as fine capillary networks, but is not ideal for printing large swaths of parenchymal tissue as would be needed in the context of 3D liver bioprinting. This technique could conceivably be used in the production of highly detailed liver organoid intermediates, which could subsequently be used as the biomaterial for a 3D printing technique with greater scalability, however, this approach has not been characterized. Incorporating autologous patient cells within 3D bioprints to facilitate tissue immunotolerance remains a key challenge in the field, and the high cell viability achievable with this technique may improve current efforts. LIFT is a relatively new technology that has not fully established its niche in the field of 3D bioprinting,²⁹ thus its application has not been well-characterized in the context of liver organoid modeling. Despite this precocious beginning, LIFT represents a valuable technology that may be used in conjunction with other techniques as a comprehensive approach to producing viable 3D bioprinted livers.

Extrusion-based Bioprinting:

Extrusion-based bioprinting employs temperature-controlled pneumatic pressure or a mechanical piston to dispense biomaterial as a continuous filament, rather than droplets as described above.¹² This layering process can be applied to a multi-nozzle system in which 3D heterogeneous tissue-like formations may be constructed, as seen in the liver model.¹¹ A defining feature of the extrusion method is the ability to use a broad range of bioinks of varying densities, viscosities, and cross-linking capacities.^12,30 With more adjustable properties, a specific microenvironment may be mimicked with higher precision. Unfortunately, extrusion-based bioprinting has several drawbacks, including limitations in printing resolution and inconsistent cell viability (40–86%).³¹ Currently, this bioprinting method is used for the generation of aortic valves and in vitro-pharmacokinetic models.³² Because this method possesses great versatility potential, advances surrounding 3D hepatic models are well explored in the literature.^33,34

In efforts to create a more authentic model for extrusion bioprints, researchers have aimed to create the most physiologically accurate conditions for functional evaluation. Khati et al. describe a decellularization process of porcine livers to retain the maximum extracellular matrix material created at 37 °C, 7–7.4 pH.³⁴ A permeable 3D construct from the decellularized liver matrix (dLM)-G-PEG (G=gelatin) allowed HepG2 cells to conform to a structural framework while crosslinking with tyrosinase immersion to prevent deformation. This multi-material structure improved rheology, extrudability, and stability while retaining cell viability. Given the xenogeneic nature of this decellularized matrix, the potential for recipient immunogenic responses cannot be excluded. Future evaluation of the immunogenicity of this construct is warranted. Extrusion-based bioprinting represents an effective approach to efficiently 3D bioprinting tissues with greater speed, which is crucial to maximizing the viability and functionality and viability of the printed organ.

Constituent Cells

The majority of cells used in current 3D liver bioprinting are derived from human embryonic stem cells (hESCs), immortalized human cell lines, or human induced pluripotent stem cells (iPSCs). hESCs represent the most physiologically similar constituent cell type to use for 3D bioprinting, however, the ethical challenges surrounding the use of these cells essentially preclude any serious consideration for their use on a large scale. Immortalized human cell lines are commonly used for 3D bioprinting organ models. One of the most commonly used cell lines for 3D bioprinted hepatic parenchymal models is the commercially available HepaRG, an immortalized human hepatic progenitor stem cell line.^30,35 These cells demonstrate high proliferative capacity, robust physiologic function, and favorable viability, however the primary limitation is the immortalization process which precludes them from long-term use in clinically functional 3D bioprinted organs. Recent work has focused on the use of iPSCs for use in 3D bioprinted constructs.³⁶ The key advantages to using iPSCs are the relatively robust cell viability, replicative capacity, and ability to be re-differentiated into the desired human parenchymal cell type.³⁷ These cells are highly versatile and may be particularly suited to printing a genetically homogeneous patient-derived organ given the ability to de-differentiate a single human cell line from the patient and re-differentiate it into multiple differentiated cell subtypes (hepatocytes, Kupffer cells, cholangiocytes, etc.) via sequential stages of treatment with differentiation media.³⁷ Recent work has demonstrated the incorporation of iPSCs into spheroids has allowed for superior cell survival and hepatic function as compared to single cell culture, in addition to reduced epithelial-mesenchymal transition.³⁸ Human iPSCs were incorporated into a heart ventricle model with a perfusable vascular network and promising biologic functionality³⁹, further demonstrating the promise of this cell type in eventual solid organ 3D bioprinting. The primary limitation to using each of these cell types is the amount of time required to culture enough cells to print an entire organ. Additionally, iPSCs demonstrate significant potential for oncologic transformation, particularly with respect to the formation of teratomas in the long-term.⁴⁰ This risk is also a concern in the use of hESCs, however it is substantially greater in iPSCs as a result of the de-differentiation and re-differentiation processes and the resulting cellular stress that increases the frequency of DNA alterations.⁴⁰ Given the number of iPSCs that would be required to print an entire organ, the risk of teratoma formation is non-negligible and further risk characterization will be required as the technology progresses. Although a consensus statement on the ideal cell type to use in bioprinting would be premature at this point in time, it is preferable to include all 5 major liver cells, such as hepatocytes, hepatic stellate cells, liver endothelial cells, cholangiocytes, and Kupffer cells to include a complete liver microenvironment.

Given the uniform, hierarchical anatomical organization of the liver, studies of engineering this intricate structure have focused on building from micro- to macroscale. Kang et al. successfully created a highly vascularized liver lobule with a central lumen from hepatocyte-laden bioinks, endothelial cell-laden bioinks, and cell-free bioinks as sacrificial materials.³⁰ This technique allowed heterogeneous pattern spatial arrangements of cells, promoting cellular organization, structural integrity, and improved cellular function. Further, it was shown that these complex functional units can be printed in series, enabling large-scale organ engineering. With this knowledge of patterned co-cultures through extrusion bioprinting, Taymour et al. applied core-shell bioprinting through coaxial nozzles to establish a hepatocyte core (human HepG2) encapsulated in the fibroblast core (murine NIH 3T3).⁴¹ Hepatic function was shown to improve in co-culture with fibroblasts. The addition of fibrin stimulated cell interconnection between fibroblasts and furthered cell spreading. Additionally, this method produces macropores and channels which may be perfused to supply nutrition to the cells, and thus mimic the micro-architecture of the liver sinusoid.

Scaffolding Approaches:

Scaffold-based:

In bioprinting whole organs and tissues, limitations exist in replicating the complex spatial arrangement of cells while maintaining cellular functionality and structural integrity.⁴² In a scaffold-based approach, cells are printed within an exogenous biomaterial matrix, such as hydrogels to provide structural and cellular support – promoting growth, cellular proliferation, and cellular interactions on a 3D scale.¹² Scaffolds may be naturally occurring or synthetically created to mimic the extracellular matrix (ECM) microenvironment. Ozbolat describes the importance of scaffold flexibility comparable to that of natural tissue. It is imperative that the hydrogel is shear thinning, the property of the viscosity decreasing as shear stress increases. This trait is considered one of the most impactful properties of printability, including resolution, pattern fidelity, and clogging prevention.¹⁰ Ozbolat explained the importance of the hydrogel possessing a quick gelation rate or chemical cross-linking to generate stable anatomically accurate structures. Lowered shear also reduces shear-induced damage of cells within the bioink during passage through the nozzle lumen. Overall, scaffolds allow for the ease and practicality of tissue deposition. Further, Park et al. described how microcarriers, reservoirs of bioactive molecules that may regulate cellular behaviors, can be implemented into scaffolds upon which cells can be cultured to promote quick proliferation.⁴³ Levato et al. demonstrated that cells cultured on microcarriers interact and aggregate more efficiently than cells in hydrogels alone.⁴⁴

Scaffold-free:

Unlike scaffold-based bioprinting, the scaffold-free approach does not use exogenous biomaterials (e.g., hydrogel, Matrigel) to hold cells in a 3D matrix. Instead, a bolus of cells is confined to a specifically shaped mold, allowing the cells to aggregate and secrete their own ECM components to provide support.⁴² This method allows for quick fusion and maturation of cells that have been used to fabricate a variety of tissues, including cardiac patches, blood vessels, and nervous tissue, as well as the liver.^42,45,46 Tissue spheroids are the most well-characterized scaffold-free bioprint to be fabricated. In one embodiment of this technique, pre-fabricated tissue spheroids are transferred from a mold using aspiration from a nozzle controlled using computer vision. The spheroids are placed, one-by-one, on to an array of needles called a Kenzan, which keeps the spheroids in contact with one another, allowing them to fuse to one another, forming a larger tissue. Spheroid placement (patterning) can be used to change the spheroid spacing, which regulates access to nutrients⁴⁶ (Figure 3E). However, this model is limited by the difficulty in creating large-scale tissues compatible with organ function. This is because scaffold-free bioprinting requires a large number of cells to create sufficient spheroid or other aggregates and expanding this number of cells is labor-intensive and costly.⁴² Additionally, the parenchymal cells do not produce the ECM, so formed aggregates are weak and do not maintain structural integrity.⁴² Thus, as an alternative approach, ECM-secreting cells (e.g., hepatic stellate cells) should be included.⁴⁵ In fact, in the setting of capillary formation, Liu et al. co-cultured fibroblasts and endothelial cells, induced them to secrete ECM, and then seeded the epithelial cells into the scaffold-free spheroid.⁴⁷ This method provided structural integrity to the bioprint while ensuring all the necessary factors to promote blood vessel formation and skin repair were available for optimal capillary network replication.

Figure 3. Bioreactor System and Scaffold-free 3D Bioprinting.

(A) FABRICA Bioreactor system used to culture and perfuse 3D bioprinted constructs in an aseptic environment. Details show the inlet and outlet ports, how the camera mount is installed onto the port, and a tissue construct loaded onto Kenzan microneedles within the bioreactor chamber. (B) Interior blueprint demonstrating flow and perfusion of constructs within the FABRICA. (C) FABRICA camera image of tissue construct on the Kenzan microneedles immediately after bioprinting. (D) The FABRICA is loaded inside of the Regenova, and the camera attached to the front sensor port of the FABRICA was used to capture images of spheroid placement during printing. (E) Microscope image of the 3D-bioprinted tissue construct immediately after printing on day 1. (F) Photograph of the perfused 3D-bioprinted pig liver model after a week of continuous pulsatile perfusion and removal from Kenzan microneedles (images obtained from Smith et al³⁸ with permission).

Multi-Component Bioinks:

In the creation of complex structures, such as the liver, single-component bioink suffers limitations of bio-incompatibility and printability.⁴⁸ Multi-component bioinks utilize two or more biomaterials to replicate the in vivo environment more accurately while maintaining structural stability.⁴⁹ These materials may be interconnected using 4 major methods. First, homogenous networks are created by covalently cross-linking materials with each other.^48,50 Second, through interpenetrating networks, independently cross-linked materials are intertwisted without covalent bonds.⁵¹ Nanocomposite bioinks include nanoparticles that increase the stiffness of the structure and enhance the shear-thinning behavior while increasing bioactivity.¹¹ Lastly, supramolecular bioinks utilize a bond that breaks and reforms between the biomaterial according to the stress.⁵² Knowledge of the interconnections between biomaterials can be used to optimally utilize them in the bioprinting of different types of tissue. In this section, the methods are outlined with emphasis on those with greater potential in liver tissue bioprinting.

Chitosan-Based:

A derivative of chitin, chitosan demonstrates biocompatibility, biodegradability, and antibacterial characteristics while possessing components of the ECM, including structures like glycosaminoglycan.⁵³ Though Li et al. have reported that various hepatocyte functions improve when cultured on chitosan, its insolubility in water and incompatibility with physiologically accurate conditions have made it incongruous for this use.⁵⁴ As a result of these limitations, chitosan has not been well-characterized in the context of 3D liver bioprinting.

Alginate-Based:

Alginate hydrogel is a biopolymer obtained from seaweed or algae that is widely used in tissue bioengineering, including the heart, bone, cartilage, and liver.⁵² Its extensive use in various forms of tissue constructs is due to its biocompatibility and applicability to several bioprinting techniques.^55,56 However, alginate has several notable shortcomings, including its overall bioinert characteristics and low viscosity.⁵⁷ To address these limitations, recent work has focused on combining alginate with other compounds. Wu et al. created a multicomponent bioink, incorporating cellulose nanocrystals and GelMA to improve printability and biological performance.¹⁰ Additionally, the polymerization of GelMA and the electrostatic interaction of alginate resulted in an independent interpenetrating network with high structural stability.¹⁰ The stiffness afforded by this approach allowed for cell growth and promoted cellular behaviors, such as cell adhesion, spreading, and proliferation. Taymour et al. combined alginate with methylcellulose which formed a cytocompatible environment that could be employed for creating volumetric structures with high precision.⁴¹ By using a core-shell extrusion bioprinting technique as described above, the structure maintained structural stability without the need for chemical cross-linking. To further replicate the hepatocyte microenvironment, the addition of Matrigel improved cell viability and proliferation, seen with increased numbers of clusters of HepG2 cells with a prominent actin structure. These studies highlight the potential for alginate-based multicomponent bioinks in the future of liver bioprinting.

Collagen-Based:

Because collagen is a significant component of the in vivo ECM, its use in bioprinting has been widely characterized. It is known to structurally support embedded cells and regulate cellular functions/behaviors, including metabolism, proliferation, cell adhesion, and migration.^51,58 However, as a sole component within a bioink, collagen exhibits low mechanical stability and slow gelation rate at physiological temperatures (37°C), limiting its ability to hold its shape after extrusion.⁵⁸ Despite these limitations, collagen has favorable biological properties, thus efforts have been directed at creating collagen-based multicomponent bioinks to overcome its relatively weak structural integrity. Mazzocchi et al. conducted experiments to optimize the ratio of methacrylated type I collagen and thiolated hyaluronic acid using UV cross-linking to improve bioprinting and maintain cellular viability and function.⁵⁹ At a 3:1 ratio respectively, human hepatocytes were able to maintain urea and albumin production and responded to acetaminophen appropriately over 2 weeks.

Gelatin-Based:

Often used as an alternative to collagen, gelatin, a collagen derivative, has been extensively studied as a bioink for 3D printing applications due to its similarity to the ECM microenvironment.⁶⁰ Gelatin has molecular bonds that are easily broken down by heat and is able to form hydrogels once cooled to low temperatures, making it a suitable bioink for controlled printing at different viscosities and stacking.⁶¹ Additionally, because gelatin is biocompatible and holds many side chains allowing for cross-linking and modification, its use for tissue printing is more widely used than collagen.^62,63 However, the use of pure gelatin is met with many limitations. It possesses poor thermal stability and limited mechanical use at low viscosities.^10,51 This issue is compounded by the fact that gelatin liquifies at physiological temperatures.¹⁰ Additionally, besides its ability to bind cells, gelatin does not possess any bioactive properties, limiting its use in tissues.⁶⁴ Therefore, to enhance the bioactivity and printing mechanisms, multicomponent bioinks with gelatin base have been investigated.^64–66

Gelatin may be conjugated with a photopolymerizable functional group⁵¹ or cross-linked with hydrogels.⁵⁷ Gelatin is often used as its derivative, GelMA, through a methacrylic anhydride reaction – used in bone, cartilage, skin, and vascular networks.⁶² Cui et al. found that with 20% methacrylic anhydride, the GelMA underwent rapid cross-linking with high cell proliferation and high structural fidelity, which was maintained for 10 days and provided an appropriate cellular microenvironment.¹⁹ When applied to match the stiffness of healthy liver tissue, Ma et al. found that bioink supported hepatocytes in cell migration and proliferation.⁶⁷ Further, a combination of glycidyl methacrylate hyaluronic acid (GMHA) and GelMA has been characterized to enhance the microenvironment for endothelial cells and promote further vascularization.⁶⁸ These gelatin multicomponent bioinks show great promise for use as hydrogels in future 3D bioprinting approaches.

dECM-Based:

More recently, decellularized ECM (dECM) hydrogels are largely considered within the future of organ bioprinting. A dECM undergoes a physical, chemical, enzymatic, or a combination thereof decellularization processes dependent on the tissue/organ that is being emulated.^69,70 This offers a method to replicate tissue-specific characteristics, including cell density, matrix density, tissue thickness, and lipid content, while removing components that impose risk of immune rejection.⁶⁹ Additionally, because dECM is derived from native tissue, it can provide the intrinsic cues for cellular function from native ECM.⁷¹ To further enable bioprinting, dECM hydrogels are abundant in collagen which easily undergoes gelation at physiological temperatures.¹¹ However, shortcomings exist due to the insufficient mechanical properties of dECM-based bioinks and slow gelation process, which impedes the creation of large complex structures.^71,72 Additionally, because restricted cell proliferation and cell colonization are noted, dECM scaffold models may hinder new tissue growth.⁴² Therefore, efforts have been directed to improve printability while using these bioinks. These include cross-linking dECM bioinks with other hydrogels,⁵⁷ combination with methacrylate,¹¹ and photopolymerization using digital light processing (DLP).⁷³ Using dECM as a scaffolding approach provides a new avenue for exploration in 3D bioprinting.⁴²

In a mixture of vascular tissue dECM and alginate, Gao et al. demonstrated precise printing through structural support due to rapid alginate crosslinking.⁵⁷ Similarly, Ma et al. mixed liver dECM and GelMA in the formation of a photo-crosslinked liver-derived dECM bioink.⁷³ The structure overcame mechanical instability issues, shown by its high fidelity and ability to remain stable and viable for 7 days. Further, liver-specific functional markers were expressed at higher levels within the multi-component bioink than those of only dECM composition, suggesting a more physiologically consistent liver microenvironment.⁷³ Kim et al. evaluated the optimal ratio of liver dECM and gelatin mixture (gelatin, hyaluronic acid, and fibrinogen) bioink for an extruded bioprint. At 2% w/v liver dECM, higher shape fidelity, increased cell viability, and enhanced cell function were noted.¹¹ By incorporating dECM with other compounds to produce multicomponent bioinks, higher resolution, fidelity, and viability were established while maintaining the natural framework from original tissue, showing potential for optimization in the bioprinting of a liver. Overall, extensive work has been done to create a multicomponent bioink that optimizes collagen’s biological properties while promoting stronger structural stability and printing mechanics.

Current Obstacles and Recent Progress in 3D Liver Bioprinting

Despite the array of anatomical, physiological, and mechanical challenges inherent to liver 3D bioprinting, ongoing efforts to optimize the technology have delivered promising results. Recent technological advancements in the context of 3D bioprinting functional 1) hepatic parenchymal units, 2) vascular systems, and 3) biliary systems are described here and summarized in Table 1.

Table 1:

Recent Advancements in Liver 3D Bioprinting.

Author	Year	Model	Method	Hydrogel	Cells	Strengths	Limitations
Yang et al [35]	2021	Organoid	Extrusion-based	Gelatin Alginate	HepaRG	Physiologic functionality High viability Perfusion kinetics Neovascularization	Immortalized cells Poor internal architecture No internal vasculature
Kang et al [30]	2020	Organoid	Preset extrusion-based	Collagen Alginate	HepaRG	Physiologic functionality High viability High resolution (10 micrometer) Lobule microarchitecture Internal vasculature	Immortalized cells
Son et al [76]	2021	Vasculature	Extrusion-based	HA Gelatin Alginate Fibrinogen	NHDFs, HUVECs	Physiologic microvasculature Customizable Host-induced neovascularization Lobule-like architecture with improved hepatocyte function	Non-physiologic cell lines Slow production time
Mazari-Arrighi et al [78]	2021	Biliary	DLS	MA-collagen MA-HA Fibrinogen LAP	NRCs	Self-organization into branched arrays High cell viability Physiologic functionality	Non-human cells
Zhang et al [86]	2022	Organoid	uFCP	None	HepG2 HSC	Micrometer resolution High cell viability Uniform microarchitecture Rapid production Primary HSCs	Immortalized hepatocytes
Bernal et al [87]	2022	Organoid	VBP	MA-gelatin LAP	EpCAM+ adult stem cells	Physiologic functionality High cell viability Extremely rapid production High resolution Sterile perfusion chamber Patient-derived cells	No internal vasculature

DLS = digital-light stereolithography, EpCAM = epithelial cell adhesion molecule, HA = hyaluronic acid, HSC = hepatic stellate cell, HUVEC = human umbilical vein endothelial cell, LAP = lithium phenyl-2,4,6-trimethylbenzoylphosphinate, MA = methacrylated, NHDF = normal human dermal fibroblast, NRC = normal rat cholangiocyte, uFCP = microfluidic flow cytometric printing, VBP = volumetric bioprinting.

Hepatic Parenchyma

In 2021, Yang et al successfully 3D-bioprinted HepaRG-based hepatorganoids which demonstrated physiologically appropriate synthetic and metabolic function when transplanted into liver-injury mouse models.³⁵ In addition to acquiring synthetic and metabolic functionality, these hepatorganoids formed functional microvasculature, and more importantly, prolonged survival in the recipients by almost 2-fold.³⁵ These hepatorganoids with robust function show promise as hepatic parenchymal building blocks for use in 3D bioprinted livers. Additionally, a recent study demonstrated the use of a preset extrusion bioprinting technique in order to generate hepatic lobules consisting of hepatocytes and endothelial cells around a central lumen.³⁰ These hepatic lobules significantly increased the synthesis of albumin and urea, and demonstrated superior cytochrome P450 enzyme activity compared to hepatocyte/endothelial cell controls lacking organized lobule architecture.³⁰ Adapting these physiologically robust and anatomically accurate constructs and incorporating them into a stable, viable bioink may prove effective in 3D bioprinting hepatic parenchyma. A key limitation to developing a hepatic parenchymal bioink for 3D liver printing is the necessity of maintaining tissue perfusion during the printing process. Printing a human-sized liver takes a considerable amount of time and subjects the printed hepatic parenchyma to considerable warm ischemia times. Further optimization of perfusion kinetics within the bioink in addition to establishing a robust perfusion system that can be maintained throughout the printing process is crucial. The FABRICA bioreactor is a 3D-printed bioreactor that can culture and perfuse 3D printed tissues in an aseptic environment in real-time.⁴⁵ Chen et al. demonstrated the FABRICA bioreactor is capable of maintaining physiologic oxygen saturation levels in the water, human blood, and culture media.⁷⁴ Given this promising feature, the system may be adapted to provide a means of maintaining perfusion throughout 3D organ printing, though this has been only characterized in the case of 3D bioprinting of genetically-engineered pig liver model.⁷⁴ Further optimization of current bioink composition and perfusion kinetics will be necessary to ensure viable 3D bioprinting of hepatic parenchyma.

Vasculature

Functional integration of the macro- and microvasculature remains one of the major barriers to successful 3D bioprinting of anatomically and physiologically functional solid organs. In contrast to the hepatic parenchyma which may withstand lower-resolution bioprinting while maintaining gross functionality, hepatic vasculature is highly intricate and requires high-resolution printing in order to appropriately establish functional perfusion kinetics.⁷⁵ Recent work has sought to evaluate potential solutions. Son et al. developed a technique to produce tissue-specific multiscale vasculature adaptable to hepatic lobule-like architecture and capable of integrating into host vasculature, maintaining appropriate hepatic parenchymal function.⁷⁶ This multicellular construct consisted of human dermal fibroblasts, human umbilical vein endothelial cells, and three different bioinks. In order to induce capillary network growth, perfusable endothelialized constructs were 3D printed utilizing endothelial cell-laden bioink. This was achieved by printing a bridge between the perfusable vascular constructs, with an angiogenic-factor secreting pattern interspersed between the perfusable vascular constructs. The interspersed placement of angiogenic-factor secreting cells allowed for an angiogenic gradient upon which physiologically functional capillary networks grew between the perfusable endothelialized vascular constructs. The constructs were then evaluated in a hepatic lobule-like array and demonstrated a drastic increase in albumin and urea secretion when compared to non-vascularized hepatocyte control lobules,⁷⁶ demonstrating functional superiority. Most notably, the hepatic lobule-like vascular constructs were explanted into a chorioallantoic membrane and demonstrated a substantial amount of host-induced neovasculature which subsequently penetrated the implant.⁷⁶ These findings provide evidence that the 3D bioprinted vasculature, when transplanted into patients, may be able to integrate into the existing human host vasculature. This technology has promising implications for the transplantation of functional hepatocyte constructs into patients with ESLD. Given the three separate vascular systems present in the liver, each of which requires both high-resolution 3D printing and a slightly different bioink composition to produce, there remains a significant time obstacle to 3D printing the entire organ. Future attempts may require producing a scaffold in which the vasculature is first printed, upon which the lower-resolution, higher-speed bioprinting of hepatic parenchyma can be placed. Further evaluation of how to incorporate 3D bioprinting of a high-resolution vascular system into the overall time-sensitive process of 3D bioprinting a liver remains a crucial obstacle to overcome.

Biliary System

The accurate recreation of a functional biliary system within the context of liver 3D bioprinting remains a challenging technological barrier to clinically viable 3D bioprinted livers. The biliary system consists of a complex array of branched intrahepatic bile ducts oriented in an apicobasal pattern that collect and transport bile to the hepatic ducts, ultimately joining with the cystic duct to form the common bile duct. Similarly to the challenge of integrating the macro- and microvasculature using 3D printing, ensuring that bile produced by hepatocytes is appropriately taken up by cholangiocytes and transported out of the liver into the common bile duct is a challenging feat to achieve with 3D bioprinting. Previous work using immortalized murine cholangiocytes demonstrated that cholangiocytes grown in scaffolds lacking components of the basal lamina failed to effectively branch.⁷⁷ Mazari-Arrighi et al. recently expanded upon this work and evaluated the use of digital light stereolithography in 3D bioprinting cholangiocytes.⁷⁸ This study utilized rat cholangiocytes in conjunction with a 3D printed ECM scaffold consisting of type I collagen, hyaluronic acid, and fibrinogen necessary to maintain the apicobasal polarity and subsequent structural integrity of the bile duct structures. The group subsequently identified the incorporation of Matrigel into the culture medium to be necessary in order to maintain long-term viability and structural integrity of the 3D printed biliary system.⁷⁸ These results provide valuable insight into the optimal parameters necessary to produce a viable, functional, and mechanically sound 3D bioprinted biliary system. Sampaziotis et al used single-cell derived extrahepatic cholangiocyte organoids to reconstruct mouse extrahepatic bile tree.^79,80 However, the biliary system presents an even greater barrier to 3D liver bioprinting than the vasculature as it requires high-resolution printing while also being particularly susceptible to ischemic injury,⁸¹ precluding it from being pre-printed into a scaffold without sufficient perfusion support. This unique challenge remains a key barrier and optimizing current bioprinting technology will likely be necessary to overcome the current limitations.

3D Bioprinting Applications and Technological Advancements

3D bioprinting holds great promise as a source of clinically transplantable organs, however the technology has implications beyond regenerative medicine. In recent years 3D bioprinting of liver organoids has been incorporated into existing organoid technology with promising results.⁸² Liver organoids have wide-reaching applications including cancer⁸³ and liver disease modeling,⁸⁴ drug development,⁸⁵ and regenerative medicine. Additionally, bioreactor technology has made impressive technological advancements,⁴⁵ and the integration of advanced 3D-printed liver organoid models in conjunction with high-fidelity bioreactors has the potential to revolutionize current liver models (Figure 3). A review of the most recent advancements in non-regenerative liver 3D bioprinting applications is discussed here and summarized in Table 1.

Microfluidic flow cytometric printing

Zhang et al. recently developed an approach to synthesize homogeneous spheroids with superior functional and anatomical uniformity using microfluidic flow cytometric printing (uFCP).⁸⁶ In this system, individual cells are directly printed into a specific configuration based on a customizable computer-based template. Similarly to fluorescence-activated cell-sorting (FACS), each cell is tagged with a fluorescent probe which is detected by the flow cytometer and subsequently delivered to the appropriate channel. The difference in the microfluidic technique is the spatial resolution; individual cells can be delivered to a particular point in space on the scale of a few microns, whereas in FACS the variability can fluctuate on a centimeter scale. This high resolution is achieved via a high-fidelity dielectrophoretic droplet deflection mechanism. Additionally, this technology allows for the exclusion of non-viable cells from the spheroid construct, which is achieved by the use of a viability dye parameter. The group showcases the superior functional uniformity using a 2-cell hepatocyte/hepatic stellate cell liver fibrosis spheroid model, in which measurement variability in type I collagen synthesis (a surrogate for fibrosis) decreased from 40% to 15%.⁸⁶ Cell viability was significantly higher in the uFCP-printed spheroids as compared to the random compartmentalization spheroids.⁸⁶ Given the capacity of current flow cytometers to accurately accommodate between 10–20 fluorescent markers, this technology has substantial implications in the context of generating physiologically optimized liver spheroids with multiple different cell types. These high-resolution spheroids may also be amenable to incorporation into a bioink for whole-liver bioprinting.

Volumetric bioprinting

An alternative volumetric bioprinting approach has recently been characterized by Bernal et al.⁸⁷ This approach involves applying visible light from multiple angles onto a photoresponsive hydrogel in order to “activate” the hydrogel in a pre-specified pattern. The result is the ability to rapidly construct a 3D pattern from an existing hydrogel. Given the rapidity of the construction and the use of visible light in place of mechanically harsh extrusion-based printing, the resulting liver organoid models created using this technique demonstrated superior viability and functionality when compared to extrusion-printed and cast controls.⁸⁷ Since this approach does not involve dynamic manipulation of the biomaterial and generates less shear stress, it is particularly useful for 3D bioprinting in vitro culture models particularly susceptible to shear stress. Incorporation of this approach into current existing methodologies may revolutionize rapid pharmaceutical testing in physiologically robust organoid models.

Induced Pluripotent Stem Cells (iPSCs)

Optimizing the composition of 3D printed liver organoids is necessary to develop a highly reliable in vitro model of in vivo liver physiology. Currently, hepatocytes used in these systems can come from several different sources, including iPSCs, re-programmed liver cells, and primary liver cells.⁸⁸ Ramli et al. demonstrated the successful production of a 2-cell liver organoid model with functional bile canaliculi by differentiating iPSCs into hepatoblast spheroids using a step-wise differentiation process.³⁷ The organoids demonstrated ductular reaction and cholestasis when treated with free fatty acids, consistent with a physiologically accurate model of non-alcoholic fatty liver disease in vitro. Jiang et al. described the use of a micropatterned agarose scaffold to generate uniform human liver organoids from human pluripotent stem cells.⁸⁹ This approach allowed for precise control of the organoid microenvironment superior to the conventional Matrigel approach.⁸⁹ These recent advancements in optimized organoid composition improve the accuracy of these models as well as the speed at which they can be produced, allowing for high-fidelity, rapid testing crucial to pharmaceutical testing and disease modeling, with the potential for adaptation into regenerative medicine technology.

Remaining Challenges

Despite several advances, there remain several challenges to successfully 3D bioprinting livers. The sheer size of the liver represents a key obstacle, as bioprinting necessitates the use of living tissue which inherently sustains ischemic damage in the absence of blood supply. One approach to this problem is to utilize mathematical models to evaluate the amount of hepatic parenchyma necessary for the intended recipient to carry out a sufficient hepatic function. This would allow 3D printed livers to be substantially smaller than typical human livers. However, the small-for-size syndrome should be tested in a large animal model. Another approach would be to 3D print functionally independent liver segments which could be perfused on a pump while subsequent segments are printed. This approach would necessitate complex surgical reconstruction during transplantation but may also allow for minimized warm ischemia time for each hepatic segment and thus improved overall functionality post-operatively.

In addition to size constraints, the incorporation of three separate vascular systems and a fully functional biliary system adds substantial tissue engineering and time obstacles to the successful 3D bioprinting of a liver. The vasculature could conceivably be printed first such that its time-consuming construction does not add to overall warm ischemia time. The biliary system likely must be printed in conjunction with the hepatic parenchyma, introducing a complicated simultaneous integration of the micro- and macroscopic biliary architecture which current 3D bioprinting technology is not yet capable of facilitating in a sufficiently time-efficient manner to maintain organ viability. Adapting current volumetric bioprinting technology to cholangiocytes may allow for rapid production of the biliary system after vasculature printing but before parenchymal printing to minimize warm ischemia time while maximizing high-resolution printing. Further research into the incorporation of multiple bioprinting modalities will likely be necessary before 3D liver bioprinting reaches clinical application.^79,80

CONCLUSIONS

Recent advances in 3D bioprinting techniques have improved the speed, accuracy, and viability of current 3D-printed liver tissues. Dedicated research into optimizing vascular and biliary 3D printing has substantially improved both the functionality and anatomical accuracy of these systems, and incorporation of these improvements into larger-scale functional 3D liver models will be crucial for regenerative medicine applications. Depending on the inclusion of liver cell types (iPSCs, hESCs, primary cells, and multiple liver cells, such as hepatocytes, hepatic stellate cells, liver endothelial cells, cholangiocytes, and Kupffer cells), the immunogenetic nature and potential immune responses to the 3D bioprinted liver would change. Although obstacles remain between current technology and clinical implementation, the solutions are within reach and patients may be able to receive their own 3D bioprinted organs in the future.

Abbreviations:

CLD

chronic liver disease

dECM

decellularized extracellular matrix

dLM-G-PEG

decellularized liver matrix-gelatin-polyethylene glycol

ECM

extracellular matrix

ESLD

end-stage liver disease

FACS

fluorescence-activated cell sorting

GelMA

gelatin-methacryloyl

GMHA

glycidyl methacrylate hyaluronic acid

iPSC

induced pluripotent stem cells

LIFT

laser-induced forward transfer

MEW

melt electrowriting

uFCP

microfluidic flow cytometric printing