Unconventional strategies for liver tissue engineering: plant, paper, silk and nanomaterial-based scaffolds

Abstract

The paper highlights how significant characteristics of liver can be modeled in tissue-engineered constructs using unconventional scaffolds. Hepatic lobular organization and metabolic zonation can be mimicked with decellularized plant structures with vasculature resembling a native-hepatic lobule vascular arrangement or silk blend scaffolds meticulously designed for guided cellular arrangement as hepatic patches or metabolic activities. The functionality of hepatocytes can be enhanced and maintained for long periods in naturally fibrous structures paving way for bioartificial liver development. The phase I enzymatic activity in hepatic models can be raised exploiting the microfibrillar structure of paper to allow cellular stacking creating hypoxic conditions to induce in vivo-like xenobiotic metabolism. Lastly, the paper introduces amalgamation of carbon-based nanomaterials into existing scaffolds in liver tissue engineering.

Plain Language Summary

Unconventional scaffolds have the potential to meet the current challenges in liver tissue engineering- loss of hepatic morphology and functions over long-term culture, absence of native-like cell-cell and cell-matrix interactions, organization of hepatocytes into lobular structures exhibiting metabolic variations-which hinder pharmaceutical analysis, regenerative therapies and artificial organ development. Paper with cellulose microfibril network develops cellular aggregates with hypoxic conditions that influence enzymes of xenobiotic metabolism proving to be a better scaffold for hepatotoxicity testing compared with conventional monolayers in tissue culture plates. Decellularized plant stems provide already-built vasculature to be exploited for the development of intricate vessel networks that exist in hepatic lobules aiding in regenerative medicine for hepatic pathologies. Fibrous plant structures are excellent materials for the immobilization of hepatocytes and improve albumin secretion enabling their use in bioartificial liver development. Biomimicry of metabolic zonation in hepatic lobules can be achieved with perfusion culture using silk blend scaffolds with varying proportions of the liver matrix that orchestrate cellular function. The mechanical properties of silk allow the fabrication of structures that resemble liver anatomy to generate native-like hepatic lobules. Nanomaterials have immense potential as a component of composite material development for scaffolds to achieve improved predictive ability in pharmacokinetics. Most of these unconventional scaffolds have the added advantage of being readily available, accessible, affordable and sustainable for liver tissue engineering applications. Conclusively, the shift of attention away from conventional scaffolds poses a promising future in the field of tissue engineering.

Graphical Abstract

Plain language summary

Article highlights.

• Limitations of 2D culture: Current 2D culture practices cannot achieve the complexity of hepatic tissue including metabolic zonation, lobular organization, extensive vasculature, existence of multiple cell types.

• Splenic scaffold: Spleen proves to be the excellent choice of decellularized organ for liver tissue regeneration, however, this cross-organ recellularization procedure is not ideal for clinical translation.

• Plant-based scaffold: Selecting plant material that matches hepatic tissue mechanical properties and anatomical features including the channel arrangements (central vein and portal triad) help hepatic regeneration and pathological modeling. Naturally fibrous plant materials have the potential to be excellent platforms for bioartificial liver development based on high hepatocyte adherence and function.

• Paper-based scaffold: Hepatocyte culture on functionalized filter paper forms a physiologically more relevant model for hepatotoxicity assessment particularly because of higher activity of drug metabolizing enzymes.

• Silk-based scaffold: Hepatic lobular organization that supports metabolic zonation can be mimicked by varying ECM content in silk blend scaffolds to create native-like oxygen and nutrient gradient.

• CNT-based scaffold: Exceptional properties of carbon nanotubes make them ideal to mimic liver ECM allowing formation of hepatic spheroids with in vivo like characteristics- bile canaliculi, polarity and drug clearance.

1. Conventional 2D cultures in liver regeneration & modeling

Liver tissue engineering involves an amalgamation of hepatocyte (HPC) source, biomimetic scaffold and chemical stimuli in the form of growth factors that act synergistically to create a functional system with the goal of mimicking normal organ morphology and physiology that can be used as implantable liver constructs, bioartificial hepatic systems, hepatic specific function study and hepatotoxicity testing models.

Liver is an organ that performs highly diverse functions ranging from regulating carbohydrate homeostasis to being the center of xenobiotic metabolism of the human body. The anatomical organization of liver in the form of hepatic lobules endows liver the ability to perform such diverse metabolic roles. With in this lobular HPC arrangement, a metabolic zonation is observed which is highly challenging to reproduce for such a functional architecture is orchestrated by blood flow that brings in oxygen, nutrients and drugs from the portal vein and takes away the resultant metabolites to central vein. Clearly, the conventionally employed 2D cultures of immortalized hepatic cell lines (e.g., HepG2, HepaRG) or primary HPCs can never give rise to such a continuum of metabolic patterns to conduct hepatic specific functionality studies, rather it is expected from such static monolayer culture systems to produce a similar metabolic phenotype across all the cells present in a monolayer. Therefore, there is a need to integrate these architectural complexities of liver with simultaneous incorporation of the important elements of the localized hemodynamics of the liver regional microenvironments to generate an artificial hepatic tissue capable of performing diverse metabolic functions, like the native liver, which could further be used for transplantation and modeling purposes. As mentioned, conventionally in vitro modeling of hepatic tissue for pharmacological and toxicological research is done using 2D cultures. These monolayer cultures often undergo a shift in genetic program to lose hepatic characteristics as an adaptation to the culture conditions and in absence of the native physiological microenvironment. Such loss of hepatic phenotype and functionality makes 2D culture of HPCs unsuitable for long-term hepatotoxicity studies. Additionally, in HPC monolayers the activity of drug metabolizing enzymes is incomparable with the in vivo context which clearly represents their failure as an optimal model for drug experimentation. Thus, there is a need of a novel approach toward hepatic modeling where the native-like genetic program of HPCs can be maintained for weeks and even months to permit generation of physiologically relevant data, understanding complex mechanisms of hepatotoxicity and development of more accurate computational models. Since cellular responses of individual HPCs in liver are governed through their interactions with other HPCs, cells belonging to non-parenchymatous populations (Kupffer cells, Stellate cells, Liver sinusoidal endothelial cells) and the liver matrix, a platform is required which can mimic the most of these interactions for next generation hepatic modeling.

2. The unconventional approaches in liver tissue engineering

The complexity of living systems that exist naturally can only be achieved by mother nature. Therefore, using whole organs in the field of organ regeneration heightens the field of tissue engineering. Decellularized livers can be the best scaffold for liver tissue engineering for it provides nearly all the mechanical, architectural, physiological and chemical cues necessary for functional liver generation [1]. However, a novel unconventional cross-organ recellularization technology is emerging where a decellularized organ is used as a scaffold to regenerate another organ. In the context of liver tissue engineering, the use of a splenic scaffold is preferred as spleen is obtained more extensively and bears a microenvironment, interstitial matrix and blood sinuses, in particular, that resembles the liver. This helps combat the limited availability of donor's livers for decellularization procedures and potential risk of introducing zoonotic diseases or immunological rejection in case of xenograft scaffoldings. Xiang et al. [2] successfully recellularized rat splenic scaffold obtained by freeze-thaw induced cell lysis and following trypsin Triton X-100 perfusion which shared similarity with decellularized liver in terms of extracellular matrix (ECM) components after seeding with bone marrow mesenchymal stem cells (MSCs) exhibiting high retention rate (>90%). To assess histocompatibility, the decellularized spleen matrix was subcutaneously implanted on rats and resulted in milder inflammatory reaction, better wound healing and higher survival rate. Later, the recellularization of the decellularized spleen with HPCs was achieved through the directed differentiation of implanted rat bone marrow MSCs [2]. Gao et al. [3] even observed albumin (ALB) production and urea secretion in rat decellularized spleen matrix that retained the natural vessel network produced by sequential detergent (sodium dodecyl sulphate and Triton X-100) perfusion within 10 days of seeding primary HPCs. The study even demonstrated that there exists no significant difference between the biological functions of HPCs cultured on splenic or hepatic scaffolds. Therefore, use of spleen acts as an extensively available scaffold for liver regeneration bearing necessary microvasculature, mechanical strength, ECM components, histocompatibility and functionality supportiveness [4–6]. In vivo, recellularization of the spleen with HPCs can be achieved upon injection of a tumor extract that develops an immunosuppressive and pro-regenerative microenvironment followed by implantation of liver cells into splenic tissue 6. Such transformation of the spleen into hepatic tissue can rescue the host mice from severe liver damage including 90% hepatectomy demonstrating the immense clinical potential of this novel approach [6]. However, the clinical translation of splenic scaffolds for in vitro liver regeneration is hindered by the thrombogenic properties of the scaffold which could be dealt with on a short-term basis only through anticoagulant coating [7]. Achieving complete endothelization of the engineered construct or formation of a perfusable microvasculature in vitro remains a major challenge [8].

Therefore, the major limiting factor that obstructs the developments in liver tissue engineering practices is the absence of an ideal biocompatible scaffold with the ability to sustain genetic program required for hepatic characteristics in the cells, create a metabolic gradient within the HPCs, mimic lobular architecture of the native liver, induce in vivo-like drug metabolism in HPCs, provide growth and proliferation stimulus to HPCs and hepatic differentiation cue to stem cells, facilitates microvascular network development to support gaseous exchange, nutrient uptake and metabolite excretion. Although biodegradable polymers rule the regime of tissue engineering as ideal scaffolds, however in the field of liver tissue engineering the intricate requirements are largely unmet, therefore, there is a need to look for alternative scaffolding materials. Not only these, the need for easily available, accessible, flexible, sustainable and economical material for bringing the prospect of tissue engineering-based personalized medicine to realization has shifted the focus away from the traditional synthetic polymer scaffolds to repurpose readily available materials around us as scaffolds [9]. The application of unusual and ubiquitous materials that are atypical in regenerative medicine including plants, paper, textile fibers as 3D matrices forms the basis of unconventional scaffolds in liver tissue engineering. Each of these materials possess one or more characteristics valuable for hepatic regeneration and/or modeling. For example- pre-vascularized scaffolds obtained from plants can be adapted to facilitate small molecule movements. The naturally occurring spectacular vascular system of plants responsible for the efficient distribution of movement of water, minerals and organics (including food and hormones) throughout the body can be easily exploited to benefit the lobular vasculature requirement of tissue-engineered hepatic constructs. Since paper scaffolds are demonstrated to be useful for raising the longevity of hepatocyte cultures while maintaining functionality, recent studies have evaluated the differentiation status of cells in paper-based hepatocyte cultures and their usefulness in performing drug toxicity and metabolism studies. The use of silk in scaffolding material is appreciated because of the mechanical properties of the constituent proteins, which help mimic liver microenvironment to an extent to pave the way toward achieving fully biomimetic artificial liver constructs. With the modern era taking biotechnology to the nanoscale, the evaluation of nanomaterial influences on HPCs is being explored by researchers to bring unseen achievements into liver tissue engineering.

The aim of this review is to highlight the advancements in scaffold technology underscoring such unconventional approaches to meet the challenges of current liver tissue engineering approaches. Similar papers have been published in the past focusing on cardiac and skeletal regeneration, in vitro cultured meat production or tissue engineering in general, however, liver tissue engineering has never been the focus [9–12].

3. Potential of decellularized plant scaffolds in liver tissue engineering

Plants serve as a natural source of polymers, particularly hydrogels like cellulose, chitosan and chitin, to prepare scaffolds. Chitosan, for example, is known for cell adherence and supporting cellular proliferation. Having antibacterial activity and porous scaffold-forming ability, it has been used to prepare microfibers for bio-artificial liver chip and scaffold for hepatic tissue engineering [13,14]. Both studies indicate chitosan supports HPC assembly in a manner similar to in vivo liver structure and hepatic-specific functions. Cellulose nanofibrils retain bulk properties of hydrophilicity, amenability to chemical modification and mechanical strength which are coupled with unique properties arising at nanoscale, hence, are attractive substrates for biomedical engineering purposes. The use of nanofibrillar cellulose for the growth of HepaRG cells and their progenitors induces the spheroid formation and enhanced differentiation along with the establishment of correct cell polarity and functional bile canaliculi-like structures respectively [15,16]. However, the process of hydrogel extraction can be expensive and currently, they do not seem to satisfy clinical translation criteria [17,18]. Now the focus of tissue engineers has shifted to directly utilizing the natural plant structures for mammalian cell growth. Such a scaffold preparation strategy, at first, requires pre-treatment for the removal of plant cells (decellularization) followed by seeding of desired cells and their proliferation (recellularization) as depicted in Figure 1. The most commonly employed decellularization techniques employ detergents (such as Triton-X) followed by exposure to bleaching agents (such as hydrogen peroxide) to disrupt the interactions between lipids and proteins and solubilize the plasma membrane while other emerging technologies focus on milder chemical treatments through a detergent-free or enzymatic approach and application of physical forces. Details on plant-material decellularization are reviewed by Rabbani et al. [19]. The detergent-free protocol clears plant cellular material with heated bleach and salt bath as demonstrated on Ficus hispida leaves, it may, however, cause scaffold damage due to prolonged heating step [20]. Cellular breakdown in plant tissues can also be achieved with the lyophilization process, where temperature shock induces mechanical permeabilization of the cell wall followed by DNase treatment to get rid of plant genetic material [21]. Another emerging decellularization strategy for plant tissues, the supercritical fluid technology, completely eradicates the need of chemical treatment taking advantage of the gas-like transport properties and liquid-like density of carbon dioxide under the conditions of high pressure and low temperature i.e. the supercritical phase of carbon dioxide (ssCO₂) which allow the gas to penetrate deep into dense material [22]. Effective decellularization is achieved when ssCO₂ is employed with an entrainer, such as peracetic acid, to remove residual plant content as seen in the study by Harris et al. on baby spinach leaves [22]. As already mentioned, decellularized scaffolds help to achieve the greatest level of complexity, limit possible immunological rejection of allografts, reduce spreading of zoonotic diseases, decellularized plant scaffolds have an added advantage of eliminating the need of donor organ which speeds up the tissue regeneration process as there's no need to look for a suitable donor. This advantage can be life-saving as in numerous cases patients have to face death due to the inability to find the organ donor within a short period of time. Table 1. summarizes works done on plant material decellularization for tissue engineering applications in the past 5 years. Many of the plant structures that have the potential to be a scaffold are produced as wastes or by-products during food production and processing. The strategies to use agricultural wastes in neural tissue engineering and culture of bovine muscular tissue for the production of lab-grown meat are discussed by Chai & Chen [23] and Perreault et al. [24]. Therefore, the valorization of such agro-food wastes into products of therapeutic value is an innovative strategy toward building a sustainable circular economy.

Figure 1.

Chemical decellularization of a plant structure (here leaf) by perfusion of detergents through its vasculature after continuous washing (not shown) to remove the cuticle. Detergents bring out cell lysis by acting upon plasma membrane components (proteins and lipids) resulting in its rupture and expulsion of intracellular components. This leaves behind a skeleton composed of relatively unaffected cell wall components (cellulose and hemicellulose) usually with lesser tensile strength than the original structure. Further recellularization is done by seeding with desired cells after biofunctionalization of the obtained cellulosic skeleton (not shown). Created with BioRender.com.

Table 1.

Decellularized plant scaffolds for tissue engineering applications.

Plant Material	Decellularization procedure	Scaffold modification	Engineered structure	Cell line	Ref.
Celery stem	Immersed in 10% SDS for 5 days then incubation with 1% NaOCl containing 0.1% Triton X-100 for 48 h	–	Hepatic lobules	hiPSC-derived HPCs	[25]
Tomato thorny leaves	Immersed in 10% SDS for 5 days further incubation in 10% NaOCl for 3 h	–	Hepatocellular carcinoma model	HepG2 cells	[26]
Celery stem	Immersed in 10% SDS for one day under gentle agitation then incubation in 0.1% Triton-X-100 with 1% NaOCl for 3–4 h followed by ultrasonic bath thrice at 40 kHz for 5 min	Nanoamyloid and nanohydroxyapatite coating	In vivo trabeculae regeneration	N/A	[71]
Apple	Added into 0.5% SDS for 2 days and shaken at 180 rpm	–	Bone-like tissue elements	hiPSCs	[72]
Green plant leaf Onion leaf	Incubated in 10% SDS for one day and then with 10% NaOCl for 12 h Incubated in 10% SDS buffer for one day	PLGA-based rapamycin nanoparticle loading	Vascular patch	N/A	[73]
Bamboo stem	Treatment with 1% NaOCl for 3 days	Chemical oxidation with sodium periodate	Osteogenic differentiation	MSCs	[74]
Asparagus stalks	Incubated in 0.1% SDS or 2 days with shaking at 180 rpm, then in PBS for 12 h	–	In vivo spinal cord regeneration	N/A	[75]
Green onion	Submerged in 1% SDS and shaken at 70 rpm for 3 weeks	–	Aligned skeletal muscle	C2C12 cells	[81]
Celery stalk	Immersed in 0.1% SDS then agitated 120 rpm for 3 days	–	Myotubes	C2C12 cells	[82]
Golden apple Carrot Celery	Immersed in 0.1% SDS for 2 days under continuous shaking at 180 rpm with intervening sonication for 5 min at 40°C after 24 h	–	Adipogenic differentiation Osteogenic differentiation Anisotropic tissue regeneration	3T3-L1 pre-adipocytes MC3T3-E1 pre-osteoblasts L929 cells	[83]
Spinach leaf	Incubated in 10X SDS for 5 days with shaking then incubation in 0.1% Triton-X100 in 10% NaOCl for 48 h	–	Osseous tissue	Bone marrow MSCs	[84]
Grass blades	Soaked and agitated in 1% SDS, 1% Tween-20 and 10% bleach for 1–2 days	–	Aligned muscle bundles	C2C12 myoblasts	[85]
Leatherleaf plant leaves	Treatment with SDS, egtazic acid and/or Tergitol, followed by a bleach and Triton X-100 solution	Gelatin crosslinking	Small-caliber vascular grafts	Endothelial cells	[86]
Sorghum leaves	Treatment with 5% NaOH for 24 h then 50% bleach solution containing 4.13% NaOCl for 1 h	poly(PEGMEMA-r-VDM-r-GMA) copolymer	Aligned myotubes	Myogenic progenitor cells	[87]
Cabbage leaves	Incubated in 10% SDS for 5 days under stirring then transferred in 0.1% Triton-X100 in 10% NaOCl for 48 h	–	Osteogenic differentiation	Bone marrow MSCs	[88]
Lotus petiole	Treatment with 10% SDS for 3 days, then in 10% NaOCl bleach for 2 days	Christmas tree microfluidics integration	Multifunctional neuron-on-a-chip system	PC12 cells	[89]
Cabbage leaf	Serial perfusion of 1X PBS and 10% SDS for ten days then with 1% Triton-X-100 for 48 h	–	Vasculature for large-sized grafts	Endothelial cells	[90]
Spinach leaves	Incubated in 10% SDS for 5 days and in 1% Triton-X100 solution for 2 days on orbital shaker at 100 rpm then submerged into 10% NaOCl for 1 h	2-deoxy-D-ribose loading	Proangiogenic wound dressing	N/A	[91]
Malabar Nightshade leaves	Incubated in 10% SDS for 5 days then treated with 1% Tween-20 in 10% NaOCl for 48 h	–	Cardiac tissue mass	Cardiac fibroblasts	[92]

°C: Degree celsius; h: hour; hiPSC: Human induced pluripotent stem cells; HPC: Hepatocyte; kHz: Kilo hertz, min: Minute; MSC: Mesenchymal stem cell; N/A: Not Applicable; PBS: Phosphate buffer saline; PEGMEMA-r-VDM-r-GMA: Poly(ethylene glycol) methyl ether methacrylate-ran-vinyl dimethyl azlactone-ran-glycidyl methacrylate; PLGA: Poly(lactic-co-glycolic acid); rpm: Revolution per minute; SDS: Sodium dodecyl sulfate.

Despite economic, environmental and humane advantages to date just two reports of using decellularized plant structures- celery (Apium graveolens) stem and tomato (Solanum lycopersicum L.) thorny leaves as scaffolds for liver tissue engineering have been published [25,26]. Both studies adopted the chemical method to decellularize respective plant structures taken as substrate as it is highly efficient. These studies immersed fresh substrates in the most common or universal compounds for decellularization procedures- sodium dodecyl sulphate and sodium hypochlorite. Progress of the decellularization procedure can be observed with the naked eye as the plant material turns white and sub-transparent with the removal of green plant cells. An essential requirement for the decellularization procedure is to ensure the safety and biocompatibility of the scaffold for biomedical applications which depends upon the amount of deoxyribonucleic acid (DNA) and proteins left in the decellularized tissue. In both of the studies, leftover DNA meets the prescribed limits of decellularized tissue (<50 ng DNA/mg tissue). The high water-holding capacity and porosity of both these scaffolds indicate their ability to efficiently absorb culture media and nutrients providing liver cells an ambient environment to proliferate. Natural structures tend to degrade out soon but decellularized plant tissues exhibit favorable stability for culturing cells due to the low degradation rate of cellulose. Both J. Wang et al. [25] and Ahmadian et al. [26] have reported no practical degradation in their respective scaffolds till 2–3 weeks after culture initiation.

To assess hepatic compatibility of decellularized celery stem, J. Wang et al. [25] cultured Matrigel embedded human-induced pluripotent stem cell-derived HPCs into the scaffold and Ahmadian et al. [26] made use of HepG2 cells for cellularization of decellularized tomato thorny leaves. The use of celery stem results in a spheroid formation having a high density of individual cells and cellular interaction mediated by tight junctions while both the scaffolds support the large expansion of HPCs. Celery stems were chosen to regenerate liver tissue for transplantation because it has a tightly arranged parallel hollow microstructure that closely resembles the liver lobule structure [25]. In detail, celery stems are extended thick and fleshy leaf stalks with a collection of many thin running tubes with a centrally-located larger tube that corresponds with the central vein of the hepatic lobule and portal triads (bile duct, hepatic artery, hepatic portal vein) located at the periphery of the hepatic lobule. It is worth mentioning here that resemblance between liver microstructure and natural plant structure is of no use if it is lost during the decellularization procedure, fortunately, J. Wang et al. [25] reported that the integrated porous morphology of celery stem is conserved in the decellularized skeleton. However, under in vitro conditions, transcriptional levels of cytochrome P450 enzymes (CYPs) were found to be lower than tissue culture plates making the celery system unsuitable for drug screening platforms.

Ahmadian et al. [26] successfully employed tomato thorny leaves to mimic the human hepatocellular carcinoma microenvironment based on their stiffness and surface trichomes evident from higher expression of angiogenic growth factor, membrane and plasma proteins correlated and/or associated with hepatocellular carcinoma in HPCs cultured in decellularized leaf scaffold compared with control. Additionally, the expression levels were found to be increasing with the length of culture highlighting the enhancement of malignant phenotype. To assess the applicability of decellularized tomato thorny leaves in drug screening, the effects of prilocaine (3.6 mM) on HepG2 cells were analyzed [26]. Cytotoxicity to cancerous HPCs was found to be significantly increased in the presence of the decellularized scaffold indicating the impact of the tumor microenvironment on drug response. Also, prilocaine treatment reduced malignancy in the culture indicated by a significant decrease in all three cancer-indicating proteins as well as a several-fold increase in the expression of pro-apoptotic factors that prevent proliferation of malignant cells. Conclusively, it can be said that decellularized tomato thorny leaves provide an adequate scaffold to develop a hepatocellular carcinoma model that could be used for efficient drug screening purposes.

Apart from these recently developed decellularization techniques, early attempts to incorporate plant material into hepatic tissue engineering were based on the use of whole plant structure, loofa sponge (LS) which is obtained from the fibrous interior of dried fruits of Luffa cylindrica (The sponge gourd), as a scaffold in a packed-bed bioreactor for perfusion culture of HPCs with the goal of producing a bioartificial liver. LS has been utilized as an excellent tool to act as a matrix for immobilization purposes of microalgal, fungal, bacterial, yeast, plant cells and HPCs from humans and rats attributed to its high porosity, low density, large pore volume, structure stability, the possibility of repeated sterilization [27]. LS as a tissue engineering scaffold has the advantage of being a naturally fabricated macroporous chitinous structure with a netting-like fibrous vascular system. The first-ever report evaluating the tissue engineering potential of LS was given by J. Chen et al. [28]. Higher HPC density and up-regulation of hepato-specific function of HepG2 cells were observed on LS compared with polyurethane scaffold probably due to surface indentations that provide greater surface area for cell attachment and support aggregation on LS fibers. Operation of HPC metabolism at a high rate in LS scaffolds evident from ammonia production and α-fetoprotein secretion rates and high cell immobilization characteristics imply the suitability of LS as a bioartificial liver substrate [28]. One of the most important factors affecting the generation of successful bioartificial organs is the maintenance of in vivo functionality in the engineered construct. Next to functionality, cell density determines the utility of bioartificial tissue-engineered constructs, because even if cells are functional and the numbers are low, they will fail to achieve the desired level of biological activity. Since in HPC perfusion culture, the LS scaffold demonstrates much higher ALB synthesis rates compared with other scaffolds and the density of HPCs in LS exceeds most biodegradable polymers and natural beads, LS is a classic substrate to develop bioartificial liver [28]. Another study by J. Chen et al. [29] described a simple, effective and scalable model for bioartificial liver development using LS as a scaffold with dynamic seeding and perfusion cultures of HPCs. Since the biofunctionalization of scaffold material can help mimic cell-ECM interactions, improvement in HPC loading and metabolic performance can be achieved by incorporating galactose moieties into the fiber surface of LS through the covalent binding of lactobionic acid [29]. This galactosylation of LS mimics cell-ECM interactions in vitro through galactose recognition by HPC surface present lectins to promote long-term maintenance of liver functions.

4. Paper-based scaffolds for hepatotoxicity testing

The compatibility of paper as a scaffold for HPCs has been reported as early as 2000 in a study aimed at developing an effective hybrid bioartificial liver device. In this study, Mizuguchi et al. [30] reported that the existence of intersecting cellulose fibrils in paper-form microcavities which allows the stacking up of HPCs in three to four layers without loss in the natural morphological appearance and further formation of bile canalicular structures. Clearly, the paper supports hepatocyte growth in a manner more closely related to their in vivo organization providing an edge over conventional monolayer cultures. Being highly affordable and easily available, resistant to degradation due to the absence of degradative enzymes in mammalian cell cultures, paper leaves behind the costlier polymeric scaffolds that support the 3D growth of cells or even hydrogels that match the stiffness of liver tissue. However, due to the absence of any motif that supports cell adhesion in cellulose, paper has to be biofunctionalized prior to cell culture applications. Cell-adhesive moieties- cell adhesive proteins (laminin, fibronectin, collagen, etc.) and functionalized nanoparticles or organosilanes bearing terminal groups that enhance protein adsorption can either be physically or covalently immobilized on the paper. Physical immobilization is a result of molecular trapping within cellulosic fibers or electrostatic tethering of positively charged regions or ionic species to the slight negative charge on the cellulose surface. Such physical immobilization allows patterning of the paper surface via contact and non-contact printing techniques which are highly employed for microfluidic applications [31]. Otherwise, direct coating by wetting the paper with solution of the desired molecule is carried out. Once the molecular adsorption is complete, excess of the solution is washed out. Direct coating is extensively used to create a cell adhesive protein layer, especially collagen, on the paper matrix for both general cell culture assays and hepatic applications in particular [32–34]. Since higher surface energy facilitates cellular attachment and spreading, surface treatment by exposure to corona discharge, laser, plasma improves interaction of paper with cells. To achieve long-term attachment and reduce loss by desorption, chemical modifications could be done on the paper to induce covalent binding of cell-adhesive moieties. Although, cellulose is naturally abundant in hydroxyl groups but they are not reactive enough at normal conditions to form covalent bonds with proteins. Thus, they are used to generate activated groups through treatment with respective chemical agents, example- glutaraldehyde and sodium periodate to generate aldehyde, which can directly bind to surface-exposed amino groups on the cell adhesive proteins or to functionalized nanoparticles [35]. Derivatization with organosilanes carrying desired terminal groups is possible even without the intervening activation step, since a direct reaction occurs between cellulosic hydroxyl groups and alkoxy group of organosilanes [36]. Such chemical functionalization of paper with carefully chosen nanoparticles and organosilanes are proves useful in modifying cellular responses and improving physical parameters of the surface. For example- Agarwal et al. [35] reported that HepG2 cells adopt a spreaded morphology on poly(amidoamine)-derivatized filter paper (grade 4) contrasting to their spherical shape on naked paper which is because of the non-adhesive nature of the paper matrix. Biofunctionalization of paper with poly(amidoamine) using glutaraldehyde as a cross-linking agent increases surface roughness, improves diffusion parameters and water uptake and enhances protein adsorption without affecting original network fibrillar structure which is crucial for cell culture applications. Clearly, the biofunctionalization of paper not only supports cell adhesion but also positively influences a variety of other properties that enhance its potential as a tissue scaffold. A major challenge in paper-based cell studies is the selection of a suitable paper type as substrate, followed by defining the desired design (if necessary) and modulation of its properties [37].

More recently, Y. Wang et al. [33] showed that using a simple collagen (100 μg/ml)-coated paper-based scaffold prepared from filter paper (Hangzhou), human-induced HPCs in co-culture with umbilical vein endothelial cells undergo aggregation and maintain the hepatic specific functions of ALB and urea production for up to two months. Decreasing the diameter of filter paper (grade 4 or 6) disc coated with Caprine liver-derived ECM (200 μg/ml) was shown to increase the cell density which facilitates the establishment of a higher degree of cell-cell interaction leading to better stimulation of the biological activity of HepG2 cells, ALB secretion [38]. These studies provide a firm basis to strike out classical limitations in liver-centric pharmaceutical studies imposed by the functional deterioration of the HPCs/induced HPCs in vitro and the lack of 3D histotypic morphology through co-culture techniques on paper scaffolds. Any toxicological study is based on precise detection methods with ideally no background interference to yield results closest to the actual drug behavior. To fulfil these requirements, paper scaffolds used to culture liver cells are amenable to immunocytochemical and ELISA analyzes without any nonspecific matrix-antibody interaction [38].

Both these studies provide a bioassay model for the assessment of drug-induced hepatotoxicity using common clinical hepatotoxins namely Acetaminophen and Valproate respectively and report enhanced sensitivity of HPCs to these toxins in the 3D state on biofunctionalized paper devices compared with 2D monolayer cultures [33,38]. These findings align with data obtained in preclinical and clinical stages, justifying the use of paper-based scaffolds over conventional monolayer systems to provide physiologically relevant results in preliminary hepatotoxicity assessment. The increased sensitivity to drugs is attributed to increased expression of phase I metabolism CYP enzymes, which convert Acetaminophen and Valproate into oxidative and mitochondrial stress-inducing metabolites that exert hepatotoxicity [33,38].

Multi-layered paper-based scaffolds are reported to produce native physiological-like oxygen gradients in disease models which are extremely significant in hepatic-functional studies, particularly drug-based studies, as hypoxic conditions positively modulate CYP activity in HPCs [39]. Operation of this effect in paper-based scaffolds was recently studied by Sitte et al. [40]. Also, Agarwal et al. [37] reported that with increasing filter paper (grade 4 or 6) disc diameter, cell proliferation enhanced leading to high cell numbers that create local hypoxic conditions in the culture resulting in high sensitivity to Acetaminophen through CYP activity stimulation. However, Diprospero et al. [41] demonstrated that the use of collagen (2 mg/mL)-coated paper scaffold prepared from sheets of Whatman 105 lens paper decreased drug metabolizing ability of HepaRG cells compared with monolayer cultures over collagen layer due to an overall fall in CYP activity, not expression and external reduction in oxygen partial pressures to create representative periportal or perivenous environment increases the activity of drug-metabolizing enzymes. This is attributed to the adoption of a cholangiocyte-like phenotype evident from a significant increase in expression of differentiation markers for HPCs contributing to the formation of bile canaliculi resulting from ten-times more adhesion protein on a substrate compared with other studies [41]. This is in accordance with results obtained by Janani & Mandal [42] who reported for the first time the impact of ECM proportion on inducing HPC metabolism and found that high liver ECM content supported cholangiocyte-like functions. Therefore, the concentration of ECM/adhesion protein is chiefly important in determining HPC metabolic fate which itself is crucial to pharmaceutical studies in hepatic systems.

Because of the amenability of cell migration assays, Transwell assays for example, to high-throughput screening they are commonly used to study cell movement, however, they are limited by the loss of in vivo histotypic structure of the cells. On the other hand, cell movement studies through cell invasion assays are currently complicated. Since the network arrangement of cellulose fibrils in paper generates a porous structure allowing cellular migration to the interior of the paper matrix, a simple paper-based invasion assay can be performed. Easily prepared gradients through layering of paper scaffolds help screen out chemicals that drive cellular movement. Agarwal et al. [37] have explored the application of paper-based devices to understand the migration of MSCs in Transwell format. These cells were chosen for their liver regenerative potential and the results showed that the presence of HPC secretome significantly increases migration [37].

5. Carbon-based nanomaterial application in drug metabolism assessment

Carbon-based nanomaterials have the potential to be extraordinary substrates for the preparation of 3D tissue engineering scaffolds. However, only a couple of these nanomaterials out of the vast array have been explored for potential in liver tissue engineering with carbon nanotubes (CNTs) being the most used. G. Bai et al. [43] have critically reviewed the application of graphene and its derivatives- graphene oxide, reduced graphene oxide, graphene quantum dots as 3D scaffolds in tissue engineering highlighting their scope in the arena of liver regeneration. Briefly, it has been described that despite having high specific surface area and porosity, two factors excellent for scaffold fabrication, graphene and its derivatives were lesser examined for liver tissue engineering purposes because of cytotoxicity concerns. Later in vitro studies have proven the non-toxic and biocompatible and even bioactive (influence on cell growth, proliferation and differentiation) nature of graphene materials and interestingly some specific methods of preparation were found to produce biocompatible graphene nanomaterials. According to G. Bai et al. [43] graphene-based materials are an excellent scaffold material for liver tissue engineering purposes because of hydrophilicity, anti-inflammatory properties, ability to direct stem cell differentiation and degrade in a biocompatible manner.

3D-graphene foam (GF) is a new foam-like material made as a defect-free graphene architecture that has been applied in cardiac, nerve, bone and skin tissue regeneration [44]. For the first time, it was evaluated as a scaffold after oxygen-plasma-treatment for a hepatic system using HepG2 cells by Loeblein et al. [45]. As described, GF is electrically conducting, highly porous and extremely light in weight with a high specific surface area giving an edge over insulating and bulky polymer 3D scaffolds. The intrinsic structure of GF supports longer periods of growth, formation of cellular aggregates and growth in the form of a layer without impacting ALB production and its extensive porosity maintains aerobic conditions in the culture [45]. Despite its biocompatibility and biodegradability, GF fails to enhance the liver-specific function of ALB production due to the lack of cell-cell interaction owing to a large surface area beyond the conventional tissue culture plate levels. Additionally, extensive oxygenation of culture due to high porosity may exert a negative impact on drug metabolism through the inactivity of phase I enzymes making it unsuitable for toxicological analysis.

Surface topography of ECM is a crucial factor in determining cell behavior, therefore efforts have been made to mimic this microlevel guidance to cells through tunable substrates. Since CNTs have controlled nanoscale topography, they are an excellent option for ECM replacements. CNTs are actually rolled-up graphene sheets producing single- or multi-walled structures. Similar to GFs, CNTs are mechanically robust, electrically conductive, biologically inert and can be easily functionalized in addition to providing nanoscale topographic surface, therefore have immense potential as a tissue engineering scaffold. The electroactive aligned structures generated by CNTs are extensively employed for nerve and cardiac regeneration [46–48]. The first-ever use of multiwalled CNT-based substrates for applications to liver tissue engineering was reported by Che Abdullah et al. [49]. In the study, primary rat HPCs displayed normal morphology and polarity on multi-walled CNT structures; and a significant increase in ALB secretion, however, no significant change in the activity of phase I metabolism enzymes was observed compared with glass coverslip control. Preliminary findings by Che Abdullah et al. [49] conclude that multi-walled CNT yarns support in vivo-like HPC growth. Such yarns with high mechanical strength, the biocompatibility of CNTs and the ability to be molded into any 3D structure, may prove to be attractive scaffolds for HPC growth for transplantation if tackled with their hydrophobicity. Also, they may not be the perfect scaffold for drug discovery processes since phase I transformations do not reach beyond conventional cultures. Moreover, CNT-functionalized surface cannot only maintain hepatic characteristics of HPCs but also have the potential to be exploited for the expansion of HPCs from the source cell population as demonstrated by Zhao et al. [50]. In the study, human liver stiffness was established using a polyacrylamide gel coated with polyethylene glycol linked multi-walled CNTs to develop a platform for differentiating HPC-like cells from human amniotic epithelial cells. This polyacrylamide-CNT composite improved mature HPC character in differentiated cells both at transcriptional and translational levels, unlike other reports where hepatic differentiated cells from human amniotic epithelial cells retain fetal nature. The HPC-like cells thus obtained demonstrated higher molecular uptake and release compared with HepG2 cells, although the basal CYP3A4 activity and ALB secretion was found to be lower. The hydrophobic nature of CNTs may prove to be a blessing in disguise as it results in weaker HPC-substrate interactions facilitating cellular aggregation and therefore spheroid formation [51]. Koga et al. [51] in the earliest study evaluating CNT influence on HPC morphology and functionality, reported that presence of CNT reduces cellular extension, promotes cell-cell communication by higher expression of connexin which establishes gap junctions and enhances liver-specific functions of ammonia removal and ALB secretion in primary rat HPCs under serum-free conditions compared with collagen-coated substrates. Another study by Wei et al. [52] reported presence of CNTs necessary for self-assembly of rat primary HPCs into spheroids on the fabricated scaffolds. Here, however, the adhesive cues to HPCs are provided by the electrospun scaffold grafted with galactose triggering a spheroid morphology with formation of in vivo-like bile canaliculi, apico-basal polarity and long-term maintenance of ALB and urea production [52]. The clearance rate of CYP2C9 metabolic probes-tolbutamide and S-warfarin, midazolam and testosterone by CYP3A11, and acetaminophen by phase II enzymes in HPC spheroids was significantly higher than those for conventional tissue culture plates, almost approaching the in vivo values. Furthermore, in vitro–in vivo extrapolation of metabolism of the five model molecules by linear regression analyzes was accurate for CNT containing scaffolds and showed lowest decline over a period of 15 days. Clearly, CNT introduction improves circulatory and xenometabolic function of HPCs, therefore, CNTs can be employed to study elimination and release kinetics of compounds with higher predictive ability making them a remarkable scaffolding component for hepatic tissue modeling to conduct drug studies. With more attention, CNTs may find a crucial role in hepatic drug metabolism studies. Notably, the studies by Koga et al. [51] and Wei et al. [52] are in contrast with spreaded cellular morphology observed by Che Abdullah et al. [49] this could be attributed to the 3D arrangement of CNTs in these studies which can better mimic the natural ECM microenvironment opposed to CNT monolayer sheets used by Che Abdullah et al. [49].

6. Biomimetic artificial liver matrix development with silk-based composites

Mimicking the cell microenvironment in tissue engineering is challenging because it incorporates several signals that regulate several cell activities. Among the materials used for this purpose, plant-derived polymers are the most common choice, however, they may face a lack of desired mechanical strength which led to a shift in focus on other natural polymers of animal origin, the silk proteins. However, silk can be immunogenic which raises biocompatibility issues probably because of its sericin component [53]. To overcome this challenge, silk has to be degummed (removal of sericin) to regenerate silk fibroin (SF) which now has been established as a chief component of scaffold material for tissue engineering applications. SF alone cannot act as an ideal scaffold material for tissue regeneration, since it has a slower degradation rate due to its extensively H-bonded structure and its hydrophobic nature is not suitable for cell adherence [54]. On the other hand, after implantation aborted macrophage-mediated proteolysis may result in encapsulation and granuloma formation [54]. Therefore, a variety of SF composites with natural and synthetic polymers have. been developed and utilized in a variety of tissue engineering approaches [55].

However, recent work in hepatic tissue engineering employs pure SF or SF blends for scaffold fabrication. Janani et al. [56] created an in vitro hepatic microenvironment using a 3D scaffold made from an SF blend of mulberry and RGD motif-rich non-mulberry silks which was characterized using rat HPCs and HepG2 cells. The SF blend scaffold supported a high degree of cellular interactions with neighboring cells and the surrounding matrix, while HPC growth on individual scaffolds made from mulberry SF and non-mulberry SF occurred in the form of clusters with a high level of cell-cell interaction and along the walls in a uniform manner due to better cell-matrix interactions respectively. Further, using gene expression and functional analysis it was demonstrated that the SF-blend scaffold supported long-term maintenance of hepatic-specific functions of HPCs i.e. ALB synthesis, urea synthesis and CYP activity [56]. In vivo compatibility of scaffolds was examined by subcutaneous implantation which showed surplus infiltration of cells in the case of SF blend and non-mulberry SF scaffold leading to accelerated tissue ingrowth and absence of immune cells at later stages of growth [56].

Blood flow through hepatic lobules generates a chemical gradient across its length resulting in distinct metabolic profiles of cells across various zones. Cells in the periportal zone lie close to the portal triad and are therefore in an oxygen-rich microenvironment and perform oxidative metabolism, progression toward the pericentral zone where oxygen availability is the minimum increase in xenobiotic metabolism is observed [57]. Other factors affecting the metabolic plasticity of HPCs include ECM composition, nutrient availability, hormone distribution and influence from non-parenchymatous HPCs. Earlier attempts to model such native-like metabolic heterogeneity through co-culture of HPCs with non-parenchymal cells in perfusion bioreactor [58]. Recently, Janani & Mandal [42] combined the mechanical strength of SF and the native environment provided by liver ECM to achieve hepatic zonation without any co-culture. To recreate this metabolic zonation, a chemically cross-linked liver−ECM SF blend scaffold was fabricated with varying ECM component degrees to mimic native-like oxygen gradient and nutrient levels by stacking of seeded scaffolds in a perfusion bioreactor and successful maturation of a whole in vitro liver construct was achieved. The root of such a stacked system lies in the observation of different functional properties of HPCs in scaffolds of varying ECM components under static culture conditions. Static culture revealed that long-term ALB and urea secretions are high in ECM-rich (3 parts of ECM out of 4) scaffolds due to proteoglycans and glycosaminoglycans with low activity of CYP reductase probably because of lower mechanical strength, and high expression of cholangiocyte genes stimulated by cues from liver ECM, all of which correspond to HPCs of periportal zone in native liver [42]. Cells on scaffolds with 2 and 1 parts of ECM had high expression of phase I and II metabolism genes along with moderate to high CYP reductase activity [42]. Therefore, low ECM content in scaffolds supports zone 3-specific pericentral cell-like functions of detoxification. Conclusively, physiological hepatic functionality was restored by stacking up scaffolds in decreasing order of ECM content with media flowing vertically upwards in a perfusion bioreactor culture [42].

A strategy to reproduce the lobular organization of hepatic tissue in vitro involves the use of radially aligned SF sheets as scaffolds for the co-culture of HPCs and primary endothelial cells [59]. The scaffold featuring radial lamellar sheets alternating with lamellar channels and a central cavity was prepared through rapid freezing of degummed silk solution in a directional manner, from the periphery to the central region, followed by lyophilization and precoated with growth factor reduced-Matrigel before seeding. Imaging revealed a unique distribution of cells on the radial scaffolds-endothelial cells occupying extended morphology attached to the sheets and HPCs organizing into a compact hepatic pad between the spaces-mimicking the native-like hepatic parenchymal−mesenchymal arrangement [59]. Furthermore, after 3 weeks of cultivation endothelial cells differentiated and organized themselves into a network that invaded the hepatic pad forming a lobule-like structure in vitro. Functional assays reveal improved hepatic-specific functions due to 3D architecture and the presence of endothelial cells indicating their significance in maintaining liver structure and physiology. This is possible because of the resemblance of the scaffold structure to the anatomic form of the native lobule and the sequential seeding strategy (inoculation of endothelial cell suspension at first, followed by an HPC-collagen mixture adopted for tissue construction).

7. Conclusion

The plant, paper, silk and carbon-nanomaterial based unconventional scaffolds have the potential to address the challenges of 2D culture in liver regeneration and modeling. However, none of these scaffold materials can be regarded as ideal which is suitable for all kinds of tissue engineering applications. Plants with anatomical and architectural features resembling the hepatic lobules can be used for liver regeneration and modeling of hepatic pathologies, however, their long-term term effects are still unknown. The potential of decellularized plant scaffolds in regeneration of large tissue constructs is unexplored and demands for an efficient criterion to numerically realize to the goal of decellularization, the elimination of all immunogens. However, large scale bioartificial liver systems can be successfully developed using the fibrous interiors of dried sponge gourd. The microfibril network within paper significantly upregulates activity of drug metabolizing enzymes by creating hypoxia-like conditions in cellular aggregates proving their utility as a substrate for assay development in hepatotoxicity testing. Among the carbon-based nanomaterials, CNTs promote spheroid formation with native liver-like circulatory and drug metabolizing function, but the cytotoxicity of graphene nanomaterials remains questioned. Devoid of immunogenic compound, SF proves to be the most successful in mimicking the normal physiological unit of liver, the hepatic lobule with its classical metabolic zonation among cells. Owing to mechanical strength, SF based scaffolds sustain perfusion that helps create the in vivo hemodynamics. Although these unconventional materials definitely help move away from the complex fabrication process of polymer scaffolds, they themselves need processing- decellularization, degumming, derivatization, etc. Being readily available, sustainable and economical plant, paper and silk-based materials can increase access to liver tissue engineering on a world-wide level after significant improvements to the field as discussed in the next section.

8. Future perspective

The introduction of novel scaffold materials into the realm of liver tissue engineering has raised high hopes for the availability of new-generation regenerative therapeutics or artificial organs. However, there are voids to fill in to accelerate this process as discussed below.

The metabolic zonation of liver cells which is crucial for normal organ physiology can be mimicked with the help of decellularized horizontal sections of higher plants with vascular arrangement sharing features of liver lobules. Since celery's biocompatibility and resemblance with hepatic lobules have been established, the formation of chemical gradients that give rise to metabolic heterogeneity must be examined. The decellularization of plant material in both of the reported studies for hepatic tissue engineering is achieved by chemical means which involve harsh detergents and bleaches that may leave environmentally toxic residues. Additionally, the decellularization process is time-consuming since long and intense washing steps post-decellularization are needed as a precautionary measure against the potential toxicity of the chemicals involved [60]. Among the other softer and sustainable methods available for plant decellularization, supercritical fluid technology can be applied to plant substrates for hepatic tissue engineering (celery stem and tomato thorny leaves) since it has been reported to successfully decellularize herbaceous plants' leaves and stems [60]. The use of scCO₂ in decellularization is a green, easily-available, low-cost technique with biologically compatible physical conditions which comes with an added advantage of being a time-efficient approach where tissues processing is achieved within a few hours and simultaneous sterilization is done [22]. Additionally, the use of decellularized plant scaffolds can also be extended to impart mechanical strength to weak scaffolds without exerting any cytotoxic effects or affecting scaffold biocompatibility, unlike chemical crosslinkers. Arin et al. [61] showed that through cross-linking of decellularized liver-derived ECM with decellularized kelp mechanical strength of the scaffold can be improved. Furthermore, decellularized plant scaffolds have an inherent elastic nature resembling the mammalian system to which cultured cells respond by reorienting nuclei, raising cytosolic calcium levels and increasing collagen secretion [62]. For the first time, Grilli et al. [63] embedded decellularized plant leaves in a 3D hydrogel scaffold to mimic the characteristic vasculature of regenerated adipose tissue. Since fennel and dill leaves bear macroscopic structural similarity to human micro-circulation, they can be incorporated into liver tissue regeneration approaches as well to mimic fluid flow through the portal triads and the central vein.

LS has proved to be an attractive substrate for bioartificial liver engineering. However, no further development in the translation of the perfusion technology for HPC culture has been reported since then, on the other hand, the role of LS in the immobilization of microbial cells for industrial production has expanded pretty well [27]. Hence the current generation of tissue engineers working toward building bioartificial liver systems should make efforts to incorporate LS and LS-like natural scaffolds. Similar to LS, another porous natural scaffold composed of collagen fibers that can be applied to bioartificial liver engineering through perfusion culture is a marine sponge. Currently, decellularized marine sponges are being explored as scaffolds for skeletal tissue engineering as its constituent material, collagen, forms a major component of human connective tissue ECM [64]. Since, the fibrillar network generated by cellulose in paper allows cell invasion assay to be performed, it can further be coupled with fluorescence microscopy for real-time monitoring [65]. Such real-time visualization may be of chief importance in liver tissue engineering as it can reveal crucial information about cultured HPCs which include patterns of cell death induced by drug molecules, oxygen uptake and consequent metabolic state of cells, organellar-specific impacts of drugs, especially changes in mitochondrial membrane potential and dynamics [66–68].

As described earlier, GF is not a good scaffold for liver tissue modeling and drug-based studies, however, easy clinical manipulation, extended periods of cultures and extremely large cell numbers reached, along with great mechanical properties evident from withstanding rotary shaking throughout the incubation throw light toward the use of GF in building composites to alleviate brittleness and yield high desired cell numbers in scaffolds for transplants. This is supported by the development of biocompatible a 3D GF/polylactic acid–poly-ε-caprolactone copolymer hybrid with significantly enhanced mechanical strength and improved ductility used for proliferation and differentiation of MSCs toward chondrogenic phenotype [69]. In vitro modeling of liver metabolic zonation is crucial to examine hepatotoxicity, investigate hepatic pathologies and understand normal liver physiology. Acetaminophen perfusion in co-culture modelled hepatic zonation mimics in vivo cell death pattern [58]. Therefore, the evaluation of hepatotoxic drugs on metabolic zones in the stacked cross-linked liver−ECM SF blend scaffold should be done to validate the potential of this liver model for studying drug metabolism and toxicology applications. Though radial scaffolds are successful in generating an in vivo-like hepatic lobular organization with an endothelial network incorporating into compact HPC mass, hepatic-specific metabolic zonation remains unknown [59]. Dynamic culture could be employed to mimic oxygen gradient across the radius of scaffolds that induce the development of distinct metabolic zones in hepatic lobules.

An important consideration while dealing with any scaffold of transplantation is its in vivo compatibility, degradation and integration with surrounding tissue. In case of biocompatibility of graphene nanomaterials, contrasting in vitro and in vivo studies are seen raising a serious concern on the usage of these materials for tissue engineering applications [43]. A study by Walaa et al. [70] indicated that mice on exposure to graphene oxide in form of nanosheets via intraperitoneal route developed DNA fragmentation in hepatic tissue indicating genotoxicity, higher levels of reactive oxygen species and lower activity of hepatic antioxidative enzymes along with various histopathological alternations in the liver tissue. Implantation of decellularized plant material scaffolds shows promising results, generally there is no observation of any pain behavior in the animal, large-scale scaffold degradation, massive immune or inflammatory reaction, thrombosis, with spontaneous endothelization or vasculature development and impressive tissue integration [25,71–75]. Li et al. [71] observed successful trabeculae regeneration in rat femur-implanted decellularized plant scaffolds and significant improvement in motor function of rats with traumatic spinal cord injury was observed upon implantation of decellularized vascular bundles by Modulevsky et al. [75]. However, such realization of successful liver regeneration using decellularized plant scaffolds is yet to be achieved. There is a need of studies to explore the suitability of such scaffolds in generating large tissue structures, till date there exists just single study by J. Wang et al. [25] evaluating in vivo performance of cell-laden decellularized plant scaffold in liver regeneration. Although no inflammatory response, vascularization and significantly higher ALB expressions were observed in nude mice after two weeks of implantation, a long-term observation is rightly needed to assess scaffold degradation and hepatic functionality maintenance. It is worth mentioning that the poor immunogenicity of plant-based scaffolds is based on efficient decellularization protocol that eliminates all the plant cellular material which may potentially trigger a strong immune response. No matter how non-immunogenic a material is, it still stays foreign to the host body and therefore an acute immune reaction is expected following scaffold implantation. Such an acute foreign body reaction to plant material is described by Modulevsky et al. [76] after subcutaneous implantation of cellulose scaffold prepared by decellularization of apple hypanthium tissue in immunocompetent mice. The efficiency of a decellularization process is measured in terms of the quantity of immunogenic material left in the tissue. Currently, a competent decellularization process has been defined as the one which leaves no visible nuclear material under histological and cytological stains and the residual genetic material does not exceed a concentration of 50 ng/mg tissue with the length of remnants lesser than 200 base pairs [77]. However, mere presence of DNA within decellularized scaffold does not evoke an immune response without any adjuvants, and therefore the immunological tolerance of some commercially available biomaterials in the host body which do not meet the above listed criteria can be understood [78]. Focus on DNA elimination shifts the attention away from removal of cellular proteins which can provoke a high immunogenic response in vivo as demonstrated by Böer et al. [79]. Use of harsh conditions for decellularization may alter ECM making it immunogenic. Therefore, there exists no suitable criterion definitive for an efficient decellularization protocol. This challenge to tackle immunogenicity of decellularized scaffolds is critically reviewed by Kasravi et al. [80] from the perspective of animal tissues where the impact of decellularization procedure on immunogenicity of scaffold obtained has been discussed along with the presentation of solutions to reduce the scaffold's immunogenic nature. Moreover, clinical translation of this technology requires critical analysis of the cell type to be used for recellularization purposes. When stem cells, particularly pluripotent and multipotent, are employed, the inevitable risk of teratoma formation or malignancy must be considered. Therefore, cell-free decellularized plant scaffolds should be evaluated in pre-clinical models to achieve the goal of hepatic regeneration via the growth, proliferation and differentiation of resident HPCs and hepatic stem cell populations.

Author contributions

Ideation, conceptualization, literature search, manuscript preparation, final revision by Sanyam Jain. Critical review by Jai Gopal Sharma.

Financial disclosure

The authors have no financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Competing interests disclosure

The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, stock ownership or options and expert testimony.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.