Liver tissue engineering and cell sources: issues and challenges


  • Anwar A. Palakkan,

    1. Tissue Injury and Repair Group, University of Edinburgh – MRC Centre for Regenerative Medicine, Edinburgh, UK
    2. Tissue Culture Laboratory, Division of Implant Biology, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Thiruvananthapuram, India
    Search for more papers by this author
  • David C. Hay,

    1. University of Edinburgh – MRC Centre for Regenerative Medicine, Edinburgh, UK
    Search for more papers by this author
  • Anil K. PR,

    1. Tissue Culture Laboratory, Division of Implant Biology, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Thiruvananthapuram, India
    Search for more papers by this author
  • Kumary TV,

    1. Tissue Culture Laboratory, Division of Implant Biology, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Thiruvananthapuram, India
    Search for more papers by this author
  • James A. Ross

    Corresponding author
    1. University of Edinburgh – MRC Centre for Regenerative Medicine, Edinburgh, UK
    • Tissue Injury and Repair Group, University of Edinburgh – MRC Centre for Regenerative Medicine, Edinburgh, UK
    Search for more papers by this author


Professor James A. Ross, Tissue Injury and Repair Group, University of Edinburgh, The Chancellor's Building, 49 Little France Crescent, Edinburgh EH16 4SB, UK

Tel: +44 (0)131-2426354

Fax: +44 (0)131-2426520



Liver diseases are of major concern as they now account for millions of deaths annually. As a result of the increased incidence of liver disease, many patients die on the transplant waiting list, before a donor organ becomes available. To meet the huge demand for donor liver, alternative approaches using liver tissue engineering principles are being actively pursued. Even though adult hepatocytes, the primary cells of the liver are most preferred for tissue engineering of liver, their limited availability, isolation from diseased organs, lack of in vitro propagation and deterioration of function acts as a major drawback to their use. Various approaches have been taken to prevent the functional deterioration of hepatocytes including the provision of an adequate extracellular matrix and co-culture with non-parenchymal cells of liver. Great progress has also been made to differentiate human stem cells to hepatocytes and to use them for liver tissue engineering applications. This review provides an overview of recent challenges, issues and cell sources with regard to liver tissue engineering.


The liver, the largest organ in the body, has a complex architecture and performs a myriad of functions. Even though the liver is highly regenerative, drugs and toxins or viral infections can cause extensive damage to hepatocytes, reducing function and regeneration. In developed countries, drug toxicity and alcoholic liver diseases are of major concern [1, 2], while in developing countries viral hepatitis is a major clinical concern [3, 4]. Globalization and immigration have resulted in the global spread of viral hepatitis and this disease has now become a major world problem [5]. The World Health Organization (WHO) report shows that 2.5% of total deaths in the world are because of liver disease [6], and this will become the 14th most common cause of death by 2030 [7]. In the UK, liver disease is the 5th most common cause of death [5], and continues to rise.

Extensive damage to hepatocytes can lead to liver failure. As a result of liver failure, multisystemic complications arise, leading to neurological impairment, haematological disturbance, renal dysfunction and metabolic abnormalities. Death usually occurs as a result of multiorgan failure. Liver transplantation is the only available and, usually, successful treatment option at this stage, if patient vital signs are favorable. However, the number of liver transplantations is limited because of the lack of availability of donor organs [8]. One strategy devised to cope with the shortage of donor organs is live donor organ transplantation, where partial liver grafts are being employed as an alternative to cadaveric liver transplantation. However, these surgical advancements are still not able to circumvent the requirement for donor livers [9]. Alternatives, such as hepatocyte transplantation [10-12], extracorporeal devices [13-17] and stem cell therapy [18-22], are strategies being pursued throughout the world either to prolong the life of patient until a donor liver is available or to regenerate the normal liver. Whether for liver regeneration therapy, for seeding liver support devices or for in vitro screening, there has been a constant quest for a suitable, reproducible source of hepatocytes. To date it was proved elusive.

Liver tissue engineering

Tissue engineering utilizes biocompatible scaffolds which provide an optimal environment for cell growth and differentiation with the aim of developing constructs that find application either in vitro or in vivo [23]. Liver tissue engineering is mainly concerned with the development of extracorporeal, liver support devices or in vivo liver regeneration using cell transplantation or implantable devices [24]. Recent trends in the field include the construction of micro tissue structures by specifically arranging the cells using of micro fabrication and nanotechnology approaches [25].

Extracorporeal temporary liver support devices

Extracorporeal temporary liver support devices are used as life-supporting bridging devices in patients awaiting a donor organ. Earlier efforts to develop artificial or non-biological devices utilized mechanical methods for removing toxic compounds [26, 27] which did not provide statistically significant improvements in patient life span. Biological support devices (bio-artificial liver) contain a cellular component, hepatocytes, combined with the bioreactor component to enhance the performance of the cells. The bioreactors are specifically designed to provide optimal conditions for the performance of the cell and have to ensure the safety of patient. As a result of the varied biochemical activities of hepatocytes, bio-artificial liver (BAL) devices were found to be much more effective than non-biological devices providing synthetic and biochemical functions [28-31]. Other BAL devices using porcine liver slices (HEPAHOPE) are under development.

It is considered that, for an effective clinical therapy, a bio-artificial liver should be able to provide at least 10% of total liver function, which requires about 1010 hepatocytes [32, 33]. A major problem is the procurement of this large number of human hepatocytes as most of these cells were previously isolated from liver tissues unacceptable for transplantation [33]. Unlike in vitro culture, conditions inside a BAL are complicated. Hepatocytes will be in contact with plasma (plasma or plasma fraction determined by the cut-off of the plasmapheresis module) which has different physio-chemical properties and poor oxygen solubility compared with culture medium. Studies have shown that hepatocytes do not perform well in plasma culture [34]. As hepatocytes have a high oxygen uptake rate and are very sensitive to oxygen concentration, a major challenge in BAL design is to provide adequate oxygen to the closely packed hepatocytes. Another major challenge is the shear stress experienced by hepatocytes in BAL devices and is mainly depended on the BAL design. Any shear stress above 5 dyn/cm2 can affect hepatocyte function [33]. As a result BAL can only be perfused very slowly, but at the same time have to provide adequate oxygen to the hepatocyte. Recent methods using modules for oxygenation and computer fluid dynamic modelling of BAL to effectively culture hepatocytes in such a device are promising [35-38]. Another major problem is the limited volume of the bioreactor, as the maximum blood/plasma that can be safely drawn out of a haemodynamically unstable liver failure patient is one litre [39]. It is a challenge to achieve 10% of liver function within a volume of one litre.

Cell transplantation

Cell transplantation has several practical advantages as compared with solid liver transplantation as: (a) several patients can be treated from a single donor organ; (b) cells can be cryopreserved and can be used later at a time of emergency; (c) it is less invasive surgery than organ transplantation; (d) it can be repeated if needed with low morbidity and; (e) the cost of treatment is less compared to whole liver transplantation [40]. Experimental studies are ongoing to assess transplantation of adult hepatocytes, fetal hepatocytes and stem cells. Even though hepatocytes can be transplanted into liver, spleen and peritoneal cavity, the therapeutic effect has been found to vary according to the site of transplantation. Intra-hepatic and splenic sites are recognized as the best sites for cell infusion [41] and the peritoneal cavity is commonly used for transplanting encapsulated hepatocytes [42-47]. There are a few reports on complications of cell infusion such as transient hemodynamic instability during cell infusion through the splenic artery and portal vein [48]. When transplanted hepatocytes do not attach, a large number of cells can reach the pulmonary circulation and can cause pulmonary emboli which can result in cardio-pulmonary compromise. Another major factor limiting the clinical use of hepatocyte transplantation is the lack of a reliable hepatocyte source. Even though cryopreserving can help to prolong the availability of cells, cryopreserved hepatocytes are less efficient in engrafting than freshly isolated hepatocytes [48]. Transplantation of hepatocytes has been found to be effective in prolonging survival in animal models of liver failure and in humans [49]. Even though the rate of engraftment of transplanted hepatocytes is often less than 30%, this procedure has been used successfully in treating inherited metabolic disorders in young patients [50-54], but the long-term function of the transplanted hepatocytes has not been demonstrated [40]. A better understanding of the factors that allow hepatocytes to engraft and integrate into the host tissue is required to avoid repeated cell infusion and possible complications. More optimization of the number of cells required to be infused at any one time is also needed.

Implantable devices

To enhance the survival and functional stability of transplanted hepatocytes, biomaterials have been developed and incorporated into implantable devices. Implantable devices are either in the form of microcarriers [55], biodegradable scaffolds [56, 57] or as encapsulation vehicles [58-60]. Since cells in microcarriers and scaffolds are open to the systemic circulation, there is the possibility of immunoreactions. Encapsulation can avoid immunological problems, but there is a requirement for a biomaterial that can provide adequate mass transfer of nutrients and the oxygen which is essential for the cells. Promising results have been shown using collagen-coated microcarriers [61, 62] and alginate encapsulation [42, 47, 63]. However, a major issue during long-term implantation is the formation of a fibrotic capsule around encapsulated structures [64-66]. These capsules can prevent the exchange of solutes between the cell and outside environment and frequently lead to cell death.

Cell sources for liver tissue engineering

Primary human and porcine hepatocytes, immortalized hepatocytes, human hepatic cell lines and stem cells have been utilized for liver tissue engineering (Table 1).

Table 1. Cell choices for liver tissue engineering
Human hepatocytesDiscarded human liver tissue

Bio-artificial liver.

Implantable devices.

Cell transplantation.

High host compatibility.

Reduced availability.

Poor in vitro proliferation.

Porcine hepatocytesPorcine liver

Bio-artificial liver.

Implantable devices.

Enhanced availability compared with human hepatocytes.

Potential zoonotic disease transmission.

Protein–protein incompatibility.

Possible immune response.

Tumour derived human hepatocyte cell linesCell lines derived from human hepatic carcinoma.

Bio-artificial liver.

Implantable devices.

Easy storage, maintenance and in vitro proliferation.


Reduced functional performance.

Immortalized human hepatocyte cell linesHuman hepatocytes immortalized by gene transfection.

Bio-artificial liver.

Implantable devices.

Easy in vitro proliferation, storage and maintenance.

Reduced tumorigenicity compared with tumour derived cell lines.

Potential tumorigenicity.

Long-term safety concerns.

Reduced functional performance.

Adult stem cells (MSCs)Adult tissues Cell transplantation. Implantable devices.

Unlimited availability.

Less safety concerns.

Transdifferentiation of MSCs to myofibroblasts.

Lower efficiency of the available hepatic differentiation protocols.

Longer duration of hepatic differentiation.

Oval cellsHuman fetal liverCell transplantation.

Native liver stem cells.

Can differentiate to both hepatocyte and biliary cells.

Cell isolation is difficult.

Limited availability.

Tumorigenic potential.

Human fetal hepatocytesHuman fetal liver

Bio-artificial liver.

Implantable devices. Cell transplantation.

Easy cell isolation.

Can undergo few cell divisions in vitro.

Lower functional efficiency.

Reduced availability.

Ethical concerns.

Possible tumorigenicity.

HepatoblastHuman fetal liver at early stage of gestationCell transplantation.Extensive in vitro proliferation.

Ethical issues.

Limited availability.

Embryonic/ induced pluripotent stem cellsPluripotent cell lines derived from discarded human embryo (embryonic) or genetically modified adult cells (induced pluripotent)

Bio-artificial liver.

Implantable devices.

Unlimited availability and indefinite proliferation.

Hepatic differentiation capability.

Ethical issues.

Possible teratoma formation.

Induced hepatocyte like cells (iHep)Genetically modified adult cells to express hepatic genes. Easy method for generating hepatocytes within a limited time period.Functional stability and safety.

Porcine hepatocytes

As the availability of human hepatocytes is limited and pigs have a physiology similar to humans, porcine hepatocytes were used in BAL and implantable devices [67]. Most of the BALs that entered clinical trials have utilized porcine hepatocytes [68, 69]. Development of improved cryopreservation techniques made it more convenient to use cryopreserved cells in BALs [70-74]. The main limitation of primary hepatocyte culture is the deterioration of hepatocyte functions and lack of long-term functional stability under in vitro culture conditions [75]. Various efforts have been made to maintain the functional performance of hepatocytes by providing proper extracellular matrix for attachment either as scaffolds [56, 76-81], microcarriers [82, 83] or cell entrapment/encapsulation devices [84-86]. The main concerns in using xenogenic porcine hepatocytes is the transfer of zoonotic diseases [87], protein–protein incompatibility and the possible immune responses generated during treatment [88-90]. There is therefore a compelling need to seek alternative cell sources.

Primary human hepatocytes

Primary human hepatocytes are the most preferred cell source for BAL sand implantable devices. However, the availability of primary human hepatocytes is limited, as most of them are isolated from discarded cadaveric liver samples that are not suitable for transplantation. Even though improved cryopreservation techniques have been developed to preserve these cells in an efficient way, a major concern is the deterioration of differentiated functions under in vitro culture conditions [91, 92]. In vitro models such as sandwich and spheroid culture have been developed to preserve the differentiated functions of hepatocytes [93, 94]. However, most of the in vitro cell models are not able to express all differentiated functions and often lack many of the enzymes, transporters, cytochrome P450 enzymes and cell polarization characteristics required [24].

Human hepatocyte cell lines

In an attempt to enhance cell availability various efforts have been made to produce new or to utilize existing human hepatocyte cell lines. Both tumour-derived hepatocyte cell lines and immortalized cells have been investigated.

Human tumour-derived cell lines

BALs such as Extracorporeal Liver Assist Device (ELAD) utilized the C3A cell line [14, 95]. C3A is a clone, derived from the human hepatoma cell line HepG2, which has improved hepatocyte functions compared with the parent line [14]. Studies show that C3A has an improved synthetic (albumin and alpha foetoprotein) capacity and nitrogen-metabolizing ability. However, factors limiting their usage are the low levels of ammonia removal (because of lack of complete urea enzymes), amino acid metabolism, cytochrome P450s and lack of many drug-metabolizing functions (mainly because of the reduced number of mitochondria)[96-98]. Another major concern in using cancer cell lines is the possibility of transmission of tumorogenic products or agents and possible complications arising from this [99, 100].

Immortalized human hepatocyte cell lines

Many efforts have been made to overcome the poor in vitro proliferation of primary hepatocytes by immortalizing these cells [101].Transfection of simian virus 40 Tantigen (SV40 Tantigen) has been used to develop many immortalized cell lines [102, 103]. Human hepatocytes have also been immortalized by lipofectamine-mediated co-transfection of albumin-promoter-regulated antisense constructs and genes coding for the cellular transcription factor E2F and D1 cyclin [104]. However, the risks of potential tumorigenic effects were not addressed in such studies. Recent efforts were made to develop systems which tightly control cell growth using reversible immortalization. In reversible immortalization, cells are immortalized by retrovirus-mediated transfer of oncogenes and subsequently these oncogenes are specifically excised out using the principles of Cre/loxP recombination. As an additional safety feature, to protect the patient from possible migration of the cells from BALs, some reversible immortalized cells also have a suicide gene such as HSVTK incorporated which makes the cell sensitive to anti-viral drugs such as ganciclovir [105]. The human immortalized hepatocyte cell line NKNT-3 was developed by using this advanced technology [106, 107]. Non-parenchymal cell lines such as endothelial cells (HNNT-2) [108], stellate cells (TWNT-1) [109] and cholangiocytes (MMNK-1) were also developed by this technique and may be useful for hepatocyte co-culture studies. However, possible inhibition of suicide gene transcription because of promoter methylation and DNA damage because of Cre/lox recombination are of concern. Moreover, the long-term safety and differentiated functionality of these cells have not been explored in much detail.

Stem cells

Adult, embryonic and induced pluripotent stem cells have been utilized for liver tissue engineering. The major advantage of stem cells is their proliferation capacity, which allows scalable populations of cells to be grown. In addition, pluripotent stem cells (human embryonic or induced pluripotent stem cells) can be expanded while retaining the ability to form any cell type in the human body, representing an inexhaustible cell source. Thus, stem cells are a promising cell choice for the production of hepatocytes as they can meet the great demand for these cells. The prospects for the use of different stem cell types in liver regeneration has been reviewed elsewhere [110, 111].

Adult stem cells

Adult stem cells are often the preferred cells for transplantation and implantable devices. Adult stem cells have no risk of teratoma formation and their use avoids many ethical issues compared with embryonic stem cells. Many in vitro studies have shown that mesenchymal stem cells (MSC) harvested from various sources have the capacity to transdifferentiate to hepatocyte-like cells [112, 113]. Animal studies have shown that direct injection of MSC into damaged liver can generate hepatocyte-like cells [114, 115]. However, the trans differentiation of MSC to myofibroblasts at the site of injury is of major concern [116, 117]. Therefore, recent efforts have been directed at transplantation of in vitro trans differentiated MSC to avoid the possible formation of myofibroblasts [118]. However, the differentiation of stem cells to mature adult hepatocytes in the quantities required for clinical therapy is still a challenge. None of the available differentiation protocols are able to differentiate stem cells to fully functional hepatocytes. Moreover, the process of differentiation requires complex growth factors and supplements that increase the effective costs of providing therapy.

Oval cells

Oval cells are quiescent stem cells residing in adult liver that proliferate extensively during prolonged liver injury, giving rise to both hepatocyte and cholangiocytes. Analysis of markers suggest that oval cells consist of a heterogeneous population of cells that differ in their differentiation capacity and stage of differentiation [119]. The low number of oval cells in liver limits their isolation and purification. Moreover, concern about possible tumour formation caused by oval cell predominance in hepatocellular carcinoma limits their application [120].

Human fetal hepatocytes

Unlike adult hepatocytes, fetal hepatocytes can undergo multiple cell divisions in vitro [121]. To ensure unlimited availability, human fetal hepatocytes have been immortalized by SV40 large T antigen [122]or hTERT transfection [123]. However, fetal hepatocytes show only low levels of ammonia elimination (49%) and urea production (1.1%) when compared with primary hepatocytes, limiting their application in a BAL [124]. Functional comparison of fetal hepatocytes with mature hepatocytes in AMC-BAL shows that fetal hepatocytes are not functionally mature enough to employ in clinical treatment regimes [125].

Even though fetal liver transplantation studies are showing promising results [126, 127], their availability, the ethical concerns, possible tumorigenicity and their incomplete differentiation limits their use in clinical application [128].


Hepatoblasts are progenitor cells that can be isolated from human fetal livers at early stages of gestation [129]. Unlike fetal hepatocytes, hepatoblasts can proliferate extensively under in vitro conditions and can give rise to both hepatocyte and cholangiocytes in vivo. Transplantation studies in animal models show that hepatoblast can integrate and proliferate in injured liver [130-132]. However, ethical issues and availability limit their application for regenerative therapy.

Embryonic and induced pluripotent stem cells

The main advantages of embryonic stem cells are their pluripotency, availability and their in definite proliferation in vitro [133]. Even though feeder-free cultures using matrigel have been developed, enabling ES cell culture in BAL devices, the risk of introducing murine viruses from matrigel is of concern. There are many reports on the in vitro differentiation of embryonic stem cells to hepatocyte-like cells [134-138]. However, none of the available protocols have been delivered at scale and have thereby limit the application of ES-derived hepatocytes in a BAL [139]. Major concerns for the use of embryonic stem cells for cell based therapy are the ethics, immunocompatiblty and possible teratoma formation from non-differentiated cells [140].

The development of induced pluripotent stem (iPS) cell technology has, to some extent, circumvented the ethical and immune issues. Recent efforts aimed at differentiating iPS cells to hepatocytes are promising [141, 142]. In spite of these developments, the concerns over use of viral vectors, changes in cell cycle regulators and teratoma formation are still limiting the application of iPS cells for liver regenerative therapy. However, the differentiation of embryonic stem cells and induced pluripotent stem cells to hepatocytes and other lineages will find application in drug and toxicity screening studies in the shorter term and may in the future have a role to play in cell based therapies.

Induced hepatocyte-like cells (iHep)

Recent innovative studies show that hepatocyte-like cells can be generated from adult cells by over-expressing hepatocyte lineage specific transcription factors [143, 144]. iHep cells were able to express both hepatic genes and functions and were able to engraft and reconstitute hepatic tissue in murine models. This novel method is a great advantage for liver tissue engineering as it can bypass all the complicated steps involved in generating iPS cells and their differentiation to functional hepatocytes. However, more studies are needed to further develop this technology. Moreover, the functional stability and safety of using iHep cells for liver tissue engineering applications has to be investigated.

Preserving hepatocyte function under in vitro culture

The major challenge in the use of primary isolated hepatocyte culture, or, indeed, hepatocytes differentiated from either ES or iPS cells, is the functional deterioration of the mature cells under in vitro conditions within a short period of culture. For a more complete hepatocyte function-appropriate microenvironment, cell–cell interactions, cell–matrix interactions and chemical gradients are required which mimic the hepatic plate of the in vivo liver [23, 145-148]. Mimicking the in vivo conditions by utilizing extracellular matrix and the co-culture of hepatocytes with non-parenchymal cells are key areas of investigation aimed at extending the life of the mature hepatocyte and preserving mature liver functions. Extracellular matrix has a profound influence on the structure and function of liver cells. Studies show that extracellular matrix resembling that of the perisinusoidal space is critical in maintaining normal liver functions [149, 150]. In vitro cultures in dishes coated with extracellular matrix such as matrigel, collagen, fibronectin and mixtures of collagen and fibronectin were only able to preserve hepatocyte functions for a short time [151]. However, collagen-sandwich systems were found to be able to maintain cellular morphology and albumin secretion for extended periods [152]. Microarray gene expression analysis demonstrates that collagen-sandwich cultures outperformed their counterpart monolayer culture in lipid, amino acid carbohydrate and alcohol metabolism, synthesis of bile acids and expression of cytochrome P450 enzymes [153, 154]. It has been shown that collagen-sandwich-cultured hepatocytes can develop well-polarized bile canaliculi and express many transport proteins [155-158].

Heterotypic interaction between hepatocytes and non-parenchymal cells has a profound role in liver physiology. Many studies have been carried out on the co-culture of hepatocytes with liver-derived cells such as biliary epithelium, stellate, endothelial and Kupffer cells. Hepatocytes have also been co-cultured with non-liver-derived cells such as human fibroblasts, aortic endothelium, lung epithelium and dermal fibroblasts. Co-culture of hepatocytes with non-parenchymal cells and the influence of these cells on liver-specific functions has been reviewed elsewhere [159]. Although the exact mechanisms by which cell–cell interactions modulate hepatocyte phenotype and function is not clear, co-culture has been found to have a profound influence on hepatocyte phenotype [158], stable long-term retention of mature liver functions such as albumin secretion [161], cytochrome P450 enzyme activity [162], phase II enzymes [163] and expression of junctional complex proteins [161, 164]. Despite the evidence of improved function using co-culture, there are no bioreactors with a co-culture component which have entered clinical trials although some are at the research stage [165]. Engineering constraints in designing such a large-scale complex bioreactor system is one of the main problems in developing this approach.

The use of biological subcellular components in BAL and implantable devices are not particularly useful because of the possible immune reactions, transfer of infections and lack of consistency. Identifying non-biological materials that can preserve hepatocyte functions will be extremely beneficial for liver tissue engineering applications. Recent approaches aimed at screening such synthetic sub-cellular matrix replacements which can support the hepatic functions of differentiated human pluripotent stem cells are promising [166]. Identification of such polymers will help to develop scalable non-biological subcellular components that can be used in BAL and implantable devices.

Conclusion and perspective

The major issue in liver support and regeneration therapy is the availability of suitable cell types. Approaches using cutting-edge biotechnology techniques, such as reversibly immortalizing human hepatocytes or non-viral methods of creating iPS cells are promising. The uses of stem cells are still in their infancy, and the need for a robust, completely efficient protocol for the differentiation of stem cells into functionally mature hepatocytes is still essential. A better understanding of the cellular signalling involved in the differentiation of stem cells into functional hepatocytes and maintenance of mature hepatocyte function would be extremely beneficial. Techniques using polymer library screening, microfluidics and micropatterning have helped to better understand cell–cell interactions, cell–matrix interactions and the microenvironment [166, 167]. This knowledge will not only help to improve the design of BALs for clinical use but will provide substantial benefits for physiologically relevant and high-throughput screening of drugs and new medicines.


This study was supported by the Commonwealth Scholarship Commission.