Skip to main content

Development of culture conditions for the differentiation of hES cells into hepatocytes

Final Report Summary - LIV-ES (Development of culture conditions for the differentiation of hES cells into hepatocytes)

Executive summary:

Chronic liver disease is the fifth cause of death in Europe. Due to the limited availability of donor livers, many people die while on the waiting list for a liver transplant, currently the only effective treatment for life-threatening liver diseases. However, this is not an ideal therapy since post-surgery complications and immunosuppressive drugs pose substantial risks for patients. Thus, development of safer, cost-effective methods for treating chronic liver disease reflects a health priority. The consortium LIVES has pursued the project 'Development of culture conditions for the differentiation of human embryonic stem (hES) cells to hepatocytes' to facilitate cell-based therapy for patients who require regeneration of liver function.

Clinical trials of hepatic cells isolated from donor livers have provided convincing evidence that transplantation of these cells can restore liver function, thereby suggesting this approach as a viable alternative to organ transplantation. However, to implement cell-therapy for chronic liver disease, a renewable source of functional human hepatic cells must be developed. Human embryonic stem cells (hESC) are able to grow indefinitely in vitro and differentiate into all cell types of the human body. Although hESC have been differentiated to hepatic cells by various laboratories, the current approaches are not suitable for clinical applications since they rely on animal-derived components and products with unspecific effects which pose risks to human health. The LIV-ES consortium was created to develop reproducible, animal-free conditions for the efficient differentiation of hESC to hepatic progenitors, with the aim of providing the European community with the necessary knowledge and tools to establish a renewable source of transplantable cells for treating chronic liver disease.

LIV-ES has made tremendous progress towards the development of novel reagents and methods for directing the differentiation of hESC to hepatic cells. One of our basic objectives was to create the technological tools to replace animal products in routine culture of hESC. The LIVES partners have generated biomaterials which support both the growth and differentiation of hESC and the bipotent cell line HepaRG. Moreover, some partners have developed a series of reporter cell lines which are invaluable tools for the identification and purification of cells expressing specific hepatic markers. Other partners have developed a step-wise, chemically-defined protocol for efficiently directing hESC to early-stage progenitors which are committed to the hepatic rather than pancreatic lineage. A renewable source of hepatic cells must expand in vitro: using the reporter tools to identify cells that express specific markers, LIV-ES partners have generated a method to purify hepatic progenitors which can then be expanded in vitro, an important step for future application to cell therapy. The pioneering technology of micro-transcriptome ribonucleic acid (miRNA) has also been employed during the project to identify novel markers of hepatic progenitors. When subjected to further differentiation, these hepatic progenitors yield functional hepatic cells which secrete albumin and express inducible cytochrome P450 3A4, the major detoxifying protein in adult liver and are capable of engrafting in animal models. Finally, the HepaRG cell line has been a critical model for LIV-ES partners to further characterise the culture components and cellular mechanisms which regulate hepatogenesis.

Collectively, the results of LIV-ES provide solid groundwork for developing hepatocyte-based therapy as a standardised procedure of regenerative medicine. The findings of LIV-ES have been published widely in journals of high impact. Thus, the tools and methodology generated by this project are available to both basic and clinical scientists for translation to human liver diseases. The culture conditions developed and validated by LIV-ES will now enable new therapeutic strategies based on stem cells or engineered tissues for restoring liver function.

Project context and objectives:

The field of regenerative medicine offers promising new possibilities for repairing or restoring the function of organs which have been damaged by disease or aging. Defective organs can be regenerated either through activation of endogenous repair mechanisms or by replacement, as in the case of transplants. Regenerative medicine using stem cells has attracted much attention, particularly because these cells are capable of auto-renewal and can differentiate into many types of cells. However, in order to develop regenerative medicine that is useful in a clinical setting, more adequate tools and methods for working with stem cells must be developed.

The liver is associated with many types of diseases, including metabolic disorders and acute liver failure. Chronic liver disease is the fifth cause of death in Europe. Once serious damage occurs to the liver, it loses the ability to repair itself and this is a life-threatening problem because the liver can no longer perform essential physiological functions including removal of toxins from the blood and the production of biochemicals for digestion. The only treatment currently available for chronic liver disorders is an orthotopic liver transplant (OLT). In Europe, inherited metabolic diseases represent 26 % of the indications for OLT. However, the number of patients dying while on liver transplantation waiting lists has increased these last years as a result of the severe shortage of organs (11 %), while it is as high as 80 % in fulminant hepatic failure. There are also substantial mortality and morbidity risks related to post surgical complications (cell rejection, vascular and biliary obstructions) and immunosuppression (infections).

Liver-directed cell therapies, such as engraftment with allogeneic hepatocytes, offer alternatives to orthotopic liver transplantation for the treatment of metabolic disorders. During the last years, some 20 clinical trials have tested hepatocyte transplantation and the results have been encouraging as they have demonstrate that these cells can restore liver function in patients. However, these approaches suffer important limitations: in addition to the paucity in organ donors, adult hepatocytes can divide only once or twice, even in the presence of hepatocyte growth factor and thus cannot be expanded in vitro. They are also difficult to cryopreserve and are highly susceptible to freeze-thaw damage, which reduces their functional capacity.

With regard to drug testing, the liver is central to pharmacokinetics and toxicology of xenobiotics, but animal models are often misleading because expression levels and substrate specificity of liver enzymes are different from their human counterparts. As a consequence, hepatic clearance and chemical profile of metabolites do not accurately represent human liver funtion. In fact, unexpected problems of toxicity and pharmacokinetics are responsible for 40 to 50 % of all failures in clinical drug development. Human cell systems, which include human hepatocyte cultures, immortalised cell lines and liver microsomes, have the potential to overcome these limitations but the cell models which are currently available are not amenable to toxicity testing for a variety of reasons. Expression of key liver enzymes, like CYP450, decline rapidly after hepatocyte isolation. Cell lines like HEP-G2 which are derived from tumors display very low expression of transporters and key liver enzymes (CYP450, conjugating enzymes) and do not exhibit the correct cell morphology and polarisation for vectorial drug transport from the plasma to the bile. These limitations for direct therapeutic applications and drug discovery emphasise the need to explore and develop new sources of human hepatic cells, which can be amplified in vitro and subsequently differentiated into functional hepatocytes.

Stem cells represent an ideal source for obtaining renewable cultures of hepatocytes for application to cell-therapy as well as drug testing. Adult livers contain two-types of stem cells that could be potentially exploited to generate hepatocytes:

1. mesenchymal stem cells are cells of extra-hepatic origin with potential therapeutic applications. However, recent reports suggest that these cells may not participate directing in liver regeneration but rather provide trophic support in injured livers by inhibiting endogenous hepatocyte death and stimulating their proliferation. Moreover, in culture, they enter replicative senescence after a limited number of population doubling.
2. liver progenitor and stem cells are present as 'resting facultative stem cells' in a specific niche of the liver. They have been estimated to represent between 0.01 % and 1 % of liver cells in newborn livers but, although they can be expended in vitro, do not display telomerase activity by contrast to fetal liver progenitor cells.

Human embryonic stem (hES) cells are perhaps the most viable option to make new hepatocytes in the laboratory for patients whose liver can no longer regenerate. The embryonic origin of these stem cells confers upon them two important properties: hESCs are able to grow indefinitely in vitro while maintaining pluripotency or the capacity to differentiate into all cell types including hepatocytes. The combination of these two properties suggests that hES cells hold great promise for future clinical application since they could be used to produce large quantities of differentiated cells for cell-based regenerative strategies. However, the generation of fully functional, human hepatocytes from hES cells under clinically-compatible conditions remains a major challenge. In the last few years, several groups have reported the differentiation of hES cells into hepatic cells using diverse culture systems. However, these approaches were all based on culture media containing serum, complex matrices such as matrigel and/ or mouse embryonic fibroblasts as feeders. Animal components pose potential risks for human health and the use of these poorly defined reagents could complicate studies aimed at identifying the molecular mechanisms of human liver development. Moreover, the proliferative capacity of hESC must be carefully controlled for clinical applications since once transplanted, they have been shown in vivo to produce teratomas or adenocarcinomas. Clearly, there are still a number of fundamental questions that must be answered before stem cell-based therapies can be employed in the clinical setting.

Given the urgent need to develop alternative therapeutic strategies for the treatment of advanced liver disease, the consortium LIVES has pursued the project 'Development of culture conditions for the differentiation of hES cells to hepatocytes' to facilitate cell-based regeneration of liver function.

The main objectives of this collaborative project were as follows:

1. development of novel tools which will combine state-of-the-art materials science with cell biology to optimise culture conditions for the differentiation of hES cells: Innovative three dimensional (3D) growth matrices, lentiviral vectors, ES-derived stage-specific cell lines, a miRNA platform devoted to ES-derived endoderm lineages and liver-specific endothelial cells will be developed to generate and expand pure populations of ES-derived hepatic progenitors, as part of work package one (WP1) under animal-free conditions and define new markers.
2. optimisation of culture conditions for the generation of anterior Definitive Endoderm multipotent stem cells: Defined, animal-free conditions will be developed on a reference hES cell line (H9) to efficiently drive differentiation to anterior definitive endoderm (ADE) stem cells; subsequently, proof-of-principle validation of these defined conditions will include testing the protocol with the hES cell lines derived by various Partners and in vitro and in vivo characterisation (WP2).
3. differentiation and expansion of hepatic progenitors from multipotent ADE cells using animal-free conditions: hepatic progenitors obtained by this protocol will be validated by comparison with foetal hepatic progenitors and the HepaRG hepatoma cell line and validated in vivo (WP3). This aim will perform in vivo evaluation of hepatic precursors to provide academic projects with fully characterised ES-derived hepatic progenitors (WP3).
4. differentiation of mature hepatocytes from ES-derived hepatic progenitors using animal-free conditions: Validation of novel hepatocyte lines for expression of detoxifying enzymes to provide the European pharmaceutical industry, small and medium sised entereprises (SMEs) and academic projects with efficient cells to allow drug testing. Products of this protocol will be carefully evaluated in vivo to ensure that these cells do not cause teratomas and that their metabolism and growth resemble functional, human heptatocytes, which will allow future preclinical evaluation for cell therapy (WP4).

Project results:

HES cells hold great promise for future clinical application since theoretically they can be utilised to produce large quantities of differentiated cells in vitro for cell-based regenerative strategies. Patients with chronic human liver disease would benefit greatly from this advance since liver transplants are expensive and there are simply not enough organ donors to treat all the patients. However, the generation of fully functional, human hepatocytes from hES cells under clinically-compatible conditions remains a major challenge. In the last few years, several groups have reported the differentiation of hES cells into hepatic cells using diverse culture systems. However, these approaches were all based on culture media containing serum, complex matrices such as matrigel and/ or mouse embryonic fibroblasts as feeders. Animal components pose potential safety issues for clinical applications and the use of these poorly defined reagents could complicate studies aimed at identifying the molecular mechanisms of human liver development. Moreover, the proliferative capacity of hESC must be carefully controlled for clinical applications since once transplanted, they have been shown in vivo to produce teratomas or adenocarcinomas. Obviously, there are still a number of fundamental questions that must be answered before cell-based therapies can be applied to treat liver disease.

To facilitate the development of hESC as a potential strategy for regenerating human liver, it is necessary to define the precise conditions for recapitulating liver development in vitro, including the molecular cues which promote the differentiation of stem cells to hepatocytes. The working hypothesis of LIV-ES has been that functional hepatocytes can be obtained by developing the appropriate culture conditions to mimic the sequential stages of development which occur in the human embryo, from pluripotent stem cells to endoderm cells, then bipotent hepatic progenitors to hepatocytes. For translation to clinical procedures, these conditions must be chemically defined to avoid the use of animal products, feeder cells and chromatin modifiers. Thus, the principal objectives of the LIVES consortium have focused on the development of reagents and methods to standardise the culture and differentiation of hESC and bipotent cells to hepatic cells. The strategies for creating these tools were based on the need to purify, expand and/or immortalise hepatic progenitors.

WP1 of the LIV-ES experimental plan has focused on the 'Development of novel tools for generating ES-derived hepatocytes'

During liver development, hepatic progenitors are in contact with other types of cells, namely mesoderm and endothelial cells and these interactions are crucial for inducing hepatic differentiation and specialisation. Studies in mouse models have confirmed that endothelial cells are required for proliferation and maturation of ES-derived hepatic cells. Hepatic cells also interact strongly with proteins of extracellular matrix. In vitro three-dimensional, biomaterial-based matrices may provide the optimal microenvironment for ES cell differentiation. Importantly, these biomaterials for cell culture can be prepared using animal-free materials and obviate the need for feeder cells, thereby enabling the development of much needed good manufacturing practice (GMP) level protocols. Biomaterials are easily expanded to larger scale 3D cell cultures in bioreactors.

Third-generation vectors derived from lentiviruses to express fluorescent marker proteins, i.e. green fluorescent protein (eGFP) and red fluorescent protein (RFP) under the control of different hepatic promoters appear efficient tools to identify and purify differentiating cells and to generate stable reporter cell lines. These cell lines will be useful to monitor the differentiation of ES cells under different conditions. Lentiviral derived vectors carrying immortalising genes should permit to establish cell lines cells at different stages of differentiation.

Moreover, miRNAs are endogenous about 22-nucleotide non-coding RNAs known to regulate gene expression at both the translational and transcriptional levels. Many miRNAs are differentially expressed during the course of development and differentiation and are thought to play important roles in these processes. Thus miRNA profiling appears to be a promising tool that may be used to monitor ES cell differentiation into endodermal cells, hepatic progenitors then hepatocytes and compare the profiles with that obtained with HepaRG cell line at the stage of progenitors and hepatocytes.

The effect of various extracellular matrix (ECM) proteins on hepatic differentiation have been tested. Various 3D commercial biomaterials have been also evaluated in terms of biocompatibility using 3D imaging and cellular viability tests, as well as functionality.

New biomaterials based on different components such as copolymers of hyaluronic acid and polyethyleneglycol were synthesised. The different materials evaluated had different cross-linking densities, porosity and chemical composition.

Evaluation of 3D imaging methods have been set up using first HepG2 cells to study expression and localisation of key proteins of polarised hepatocytes namely efflux proteins, namely multidrug resistance protein one (MDR1), multidrug-resistance associated protein (MRP), breast cancer resistance protein (BCRP) and Bile salt export protein (BSEP).

After extensive comparative studies on HepaRG cells and ES-derived ADE cells, finally a commercial biomaterial (Hydromatrix) and a new biomaterial nanocellulose, which has been developed in P6 lab were selected for further development to optimise the 3D culture condition used for differentiation of hES cell-derived hepatic progenitors into hepatocytes.

Lentiviral based vectors with stage-specific promoters (provided by different partners) driving the expression of two fluorophores: eGFP and dsRED were constructed and produced (P5). The expression and the tissue specificity of these lentivectors, once validated by the different partners, were then provided as highly concentrated and purified batches. Lentivectors with immortalising genes have been also provided. Highly concentrated batches of non-integrative lentiviral based vectors were produced to allow for selection of hepatic progenitors based on specific reporter expression. Finally, EF1-puro-IRES silencing constructs to stably knockdown IRS2 in hES cells have been validated in HepaRG cells. Another version of most of the vectors was constructed in which a gene for selection (neomycine) was inserted and provided to partners.

Efficient protocol to transduce pluripotent hES cells was defined and the protocol shared with partners. Stable ES cell lines transduced with EF1alpha-GFP have been generated as control cell line. Stable lentiviral reporter lines in hES cells have been also generated with APOA-II-GFP, Insulin-GFP and IRS2-GFP using H9 cells as well as the Spanish VAL9 cells. The expression of both IRS2 and APOA-II were shown to coincide with the specification of hepatic progenitors. However, upon passaging APOA-II induction was not observed as well as that of Insulin, which suggested promoter specific silencing in undifferentiated ES cells.

Reporter HepaRG cell lines have also been established with the same APOA-II-GFP and IRS2-GFP vectors. HepaRG cells were also transduced with the CYP3A4-GFP leentivector to analyse the influence of the microenvironment on the phenotype and the differentiation potential of HepaRG progenitors, to test the effect of DMSO and inducibility of GFP expression by rifampicin, a known inducer of CYP3A4 expression. Stable HepaRG cell line with IRS2 knockdown was also established to assess the differentiation and cell cycle markers (P4, P1C).

Many co-culture experiments performed between HepaRG and circulating progenitor endothelial cells and ES-derived endothelial cells failed to induce these cells to acquire a liver phenotype. Therefore conditions to isolate and purified foetal liver endothelial cells from mouse and human foetal livers have been set up. Immortalised cell lines have been generated and characterised (P1B).

Amongst the miRNA population, miRNA 122 has been identified as an essential and predominant miRNA expressed during liver development. This miRNA is liver specific and expressed at high level in the adult liver where it represents about 70 % of all miRNAs. To better understand the role of this miRNA the molecular mechanism controlling its expression have been investigated. It was found that miR-122 primary transcript consists of 3 exons and that miR122 is generated from the 3rd exon.

Characterisation of the miR-122 promoter showed that this promoter was expressed only in cells (human hepatoma cells) that express miR-122 and that there was a correlation between the level of expression of miR-122 and miR-122 promoter activity: Huh7 which express a high level of miR-122 have the highest promoter activity (compared to HepG2 and PLC/PRF5).

The miR-122 promoter was also found to have an enhancer activity for the liver specific human alpha1 antitrypsin promoter (hAAT) and not on control EF1alpha promoter.

Dissection of the promoter was made by serial deletions in front of the luciferase gene and it was shown that the site for the CCAAT/enhancer-binding protein-alpha (CEBP) was required for maximal activity.

Other results suggest that other factors such as HNF3B are required for optimal miR-122 promoter activity (P3).

These results lead to the identification of miR-122 promoter which was sent to LIV-ES partners in particular P5 for cloning in lentiviral vectors and use for monitoring hepatic progenitor differentiation toward more mature hepatocytes.

WP2 of LIVES focused on 'Differentiation of hES cells into multipotent endoderm stem cells'.

Generating functional, differentiated hepatocytes from hES cells using culture conditions that are compatible with clinical applications remains a major challenge. Development of a protocol which reproduces in vitro the early steps of human embryonic development may provide the best approach for generating differentiated cells with functional properties. The first event of differentiation during mammalian development occurs at the gastrulation stage with the specification of the primary germ layer ectoderm, endoderm and mesoderm from which all the cell types of the adult body are derived. The endoderm germ layer is quickly patterned during gastrulation to give rise to the ADE from which the hepatic bud will be later specified. Therefore, definite endoderm (DE) and then ADE cells represent the earliest progenitors of liver cells during development.

The effect of various ECM-proteins were tested on hepatocyte differentiation on hESC-derived DE cells such as Col 1, Col III, Col IV, fibronectin, gelatin and growth factor reduced matrigel. Based on immunostaining and Alb secretion it was concluded that fibronection may support ADE cell differentiation into hepatocytes and could be used for 3D cultures.

For 3D cultures various commercial matrixes were tested and amongst them, three that were chosen for more detailed studies: HydroMatrix (Sigma), Hystem-C (Sigma) and ExtraCel (TebiBio). Studies were performed on HepG2 cells then on HepaRG cells (P6A/B).

A protocol driving differentiation of hESCs into homogenous population of endoderm cells in chemically defined medium which is devoid of animal products and of source of undefined factor such as Matrigel or chromatin modifiers was developed (P2). These cells have been extensively characterised by expression of specific markers (Hex, CXCR4, Cerberus, Otx1, HNF1beta, goosecoid, MixL1, HNF3beta) by Q-RTPCR, immunostaining and micro-array gene expression profiling. Cells were also validated for the absence of expression of visceral endoderm markers. Importantly the protocol was found to be also effective on Finnish (FE22, 29) hES cells as well as on Spanish VAL9 hES cells.

For ADE cells to be considered as authentic endoderm progenitor, they must be competent to generate multiple endoderm cell types (liver, pancreatic cells and even lung).

Available methods to generate pancreatic progenitors often contain undefined animal products such as feeders or foetal bovine serum (FBS). Starting from our ADE cells, we screened defined culture conditions to differentiate human DE into a near homogenous population of pancreatic and liver specified endoderm from multiple hPSC lines. The result of this screening shows that RA has an essential function in promoting pancreatic specification while BMP signalling blocks the expression of pancreatic markers. In addition, we observed that inhibition of FGF signalling decreases cells survival of pancreatic progenitors, thus justifying the use of FGFs in our protocol. More importantly, our analyses also revealed that Activin/TGFb controls DE cell fate choice toward the pancreas lineage by inhibiting dorsal foregut (DF) specification while promoting the hepatic lineage.

These results have important practical significance since protocols currently available to generate pancreatic cells from hPSCs often rely on feeders, Matrigel? and serum all which represent potential source of TGFb signalling with the capacity to compromise pancreatic specification.

Finally, a chemically defined 2D platform was also developed for the generation of multipotent endoderm progenitors which can produce lung, liver and pancreatic progenitor cells. A methodology to expand this foregut endoderm population to produce homogeneous population of cells was also developped. All data has been replicated in multiple hESC and hIPSC lines demonstrating this is not a cell line or cell type specific observation (P2).

WP3 focussed on the 'Differentiation of endoderm stem cells into hepatic progenitors'.

During embryogenesis, distinct mesodermal signals, either secreted or membrane-bound, are required for the specification of the liver bud and proliferation of foetal hepatic progenitors. The conditions developed to convert ADE cells to hepatic progenitors in vitro were based on our understanding of the processes that occur during normal liver organogenesis. The use of hepatic-specific primary endothelial cells was one critical approach as these cells are known to play a major role in promoting the proliferation of hepatic progenitors in situ. Human foetal hepatic progenitors (isolated and characterised as defined by Partner 1A) as well as HepaRG cells (developed and patented by Partner 1C) served as positive controls for the analysis of hepatic progenitors produced within this WP.

The inductive effect of various growth factors involved in liver bud organogenesis; such as FGFs, BMPs and inhibitors of signalling pathways (PI3Kinase) were evaluated on ADE cells generated from hESCs. The resulting culture system was devoid of animal products, serum or chromatin modifyers and was refined in order to achieve successful differentiation of ES-derived ADE cells to hepatic progenitors and fetal hepatocytes using fully defined culture conditions (P1A/P2).

The ability of various cytokines to enhance differentiation towards endodermal lineages was analysed. These studies showed that TGFbeta/activin favoured hepatic specification. Therefore, a novel cell culture system was developed by P2 to generate fetal hepatocytes in which hepatic specification was obtained by replacing the CDM/PVA medium + factors by RPMI+B27+ Activin 50ng/ml. Further differentiation was then achieved by culturing the progenitors in CRML medium+HGF 10ng/ml +oncostatin 20ng/ml from day 10 to day 25 (P2).

Hepatic progenitors derived by both methods were characterised by Q-RT-PCR, RT-PCR, immunostaining and FACS analysis for cell surface markers. Functional studies were also used to characterise and validate the cells generated. These included; urea excretion assays, albumin synthesis, indocyanin green uptake and excretion and glycogen storage.

An example of hepatic markers expressed by hepatic progenitors expressed between d12 and D16 of differentiation. AFP and HNF4alpha are markers of the hepatocyte lineage whereas CK19 is a marker of cholangiocyte lineage.

When ES cells were transduced with APOA-II-GFP vector and differentiated into hepatic progenitors expression of GFP was observed in most cells.

Importantly, the hepatic progenitor cells derived in vitro from hESCs using the first protocol, were able to graft and maintain their differentiated characteristics in vivo following transplantation into an immunodeficient (Rag2gammac-/-x uPa) mouse model. These studies demonstrated the presence of clusters of grafted cells expressing human hepatocyte markers AA1T and albumin within the liver parenchyma of the transplanted animals.

Cell culture conditions were developed and defined during the study that would allow the passage and expansion of progenitor cells generated by the differentiation protocol. The hepatic progenitors derived from hESCs from days 13 to 15 of the protocol could be amplified on collagen IV and Collagen I coated plates, in chemically define -hepatic progenitor culture medium (HPM). Amplification of hepatic progenitors also enabled us to define conditions to then purify these cells. The approach used was based on a combination of cell sorting, a method for producing purified integrating (ILV) and integration-defective lentivectors (IDLV) yielding high-titer pure preparations of viral particles (P5) and the use of a hepatic promoter and of a specific integrase inhibitor.

The human apolipoprotein A-II (APOA-II) promoter was used to drive GFP expression in hepatocyte progenitor cells following infection with either integrating lentivirus or with Integrase-deficient lentivirus. The specificity of APOA-II-GFP lentivirus was validated in various cell lines and in differentiating hES cells prior to being used to isolate pure cultures of progenitor cells from differentiating hESCs.

Conditions for cell sorting of the hepatic progenitor population yielded a population of 99 % GFP positive cells 89 % of which co-expressed hepatoblast markers such as AFP and CK19.

In order to circumvent the potential drawbacks of using integrating viral reporters to label precursor cells, one of the major objectives of the study was to develop an approach to temporarily label precursor cells with the APOA-II reporter so that they could be isolated without permanent genetic alteration. We therefore developed the use of Integrase-defective lentivirus and a viral integrase inhibitor (raltegravir) to express the selectable reporter in a fully reversible manner. Hepatocyte precursor cells generated using these methods were shown to be integration free and could be differentiated further into more mature hepatocytes expressing albumin (P1A).

Whether proliferating hepatoblasts behaved as true bipotential progenitors and were able to also undergo differentiation to the cholangiocyte lineage has been investigated. Conditions have been defined to generate fetal cholangiocytes, co-expressing HNF6 and CK19 and which are able to adopt tubular forms, characteristic of cholangiocytes.

HepaRG cells at the stage of progenitors were also shown to be able to differentiate into cholangiocyte like cells expressing specific markers such as GGT1 or KRT14 (P1A/C).

Endothelial cells isolated from dissociated liver were sorted according to expression of CD146 and CD31 and the resulting cells were demonstrated to be Lyve-1 positive (a marker of liver sinusoidal endothelial cells). Following isolation the cells were immortalised by lentiviral infection with EF1 driven hTert and E6-E7. Using this method different immortalised human liver endothelial cell lines were generated and characterised. All lines tested expressed endothelial markers (CD31, CD144, KDR, CD146) and were able to form tubules in matrigel, incorporate acetylated LDL and were activated by TNF-alpha (induction of ICAM-1 and VCAM-1). Coculture with hESC derived endodermal cells is a work in progress (P1B).

Little is known regarding the complex expression patterns of both mRNAs and miRNAs during the early stages of human liver development and the role of miRNAs in the regulation of this process has not been studied. HepaRG cells were used because they reproduce the process of differentiation of bipotent hepatic progenitors and are capable of differentiating into both mature hepatocytes and biliary cells. Analyses were performed comparing the gene expression profiles of HepaRG progenitors at D3 and differentiated HepaRG at D30 without DMSO. They revealed profound changes between transcriptome of progenitors and differentiated HepaRG. More than 2 000 genes whose expressions are changed more than 1.5 fold and 1 500 whose expression are changed more than 2 fold between the two conditions. Interestingly, addition of DMSO between D15 and D30 known to potentialise the differentiation process induced also profound transcriptome changes with a strong increase of several drug metabolising enzymes expression, coagulation protein expression and a decrease of cell cycle protein and cell communication protein.

Comparison of the gene expression profiles of progenitors on Day 3 and differentiated HepaRG Day 30 without DMSO with gene expression profiles obtained from mouse livers at different stage of development was then performed. Hierarchical clustering analysis also allowed P1C to identify two major clusters which organisation was driven by the culture condition of HepaRG cells. The first cluster corresponds to embryonic liver and includes HepaRG progenitors. The second cluster corresponding to post-natal liver includes differentiated HepaRG. Interestingly, the HepaRG signature in cluster 1 fit with early liver development stage corresponding to liver bud growth and liver hematopoiesis which demonstrates that HepaRG progenitor is a hepatoblast.

These transcriptomic profiles of HepaRG cells at the stage of progenitors and differentiated cells were then used to predict miRNA expression by using MiRABELLE software developed by P3. Three microRNAs (miR-210, miR-488 and miR-380-5p) were predicted to have a significant increased activity in progenitor HepaRG compared to differentiated HepaRG.

In parallel, another approach consisting to hybridise RNA from HepaRG on Agilent chip (G4471A Unrestricted Amadid miRNA 8x15k) was performed. Eight miRNA highly expressed in HepaRG progenitors compared to differentiated HepaRG (fold change more than two, p less than 0.05) were identified: miR-17, miR19a, miR-221, miR-301a, miR-31, miR-34c-5p, miR-424 and miR-99p.

Interestingly, among these miRs, miR-17 and 19a were shown to be expressed in liver at early stages of development (E11.5) and to decrease during liver organogenesis (Jevnaker et al., J Cell Physiol. 2011;226:2257-66). Using a mouse model of liver cancer, it was also shown that miR-221 overexpression stimulates growth of tumorigenic murine hepatic progenitor cells (P1C).

Global gene and miRNA expression profiled from adult and 9-12w human embryonic livers were obtained (high-density microarrays and quantitative RT-PCR) by P3. Comparison of the expression of the most regulated miRNAs to that of their putative targets using a novel algorithm revealed a significant anti-correlation for several miRNAs and identified the most active miRNAs in embryonic and in adult liver. Furthermore, our algorithm facilitated the identification of TGF?-R1 as a novel target gene of let-7.

miRNA expression profiled from undifferentiated and NaButyrate-induced differentiated hESC of two lines were also performed using microarray and quantitative RT-PCR. The miRNA profiling revealed expression of three novel miRNAs in undifferentiated and differentiated hESC. Upon NaButyrate induction, two of the most upregulated miRNAs common to both cell lines were miR-24 and miR-10a, which target genes shown to inhibit endodermal differentiation. In parallel, induction of several liver-enriched miRNAs, including miR-122 and miR-192, was observed to induction of endodermal gene expression. Moreover, ectopic expression of mir122 in hESC cells promotes expression of HHEX and CXCR4, transcription factors required for generation of endoderm-derived progenitors. Therefore, mir122 was identified as a specific marker of the endoderm lineage that is also expressed early in liver development (P3).

It has been shown that insulin signalling via IRS2 is required for the proliferation and differentiation of HepaRG precursor cells. Knockdown of endogenous IRS2 prevented the cells from differentiating to hepatocytes, whilst omitting insulin from the differentiation media resulted in a blocked state of differentiation in which IRS2 was highly expressed. Insulin/IRS2 signalling was found to promote the proliferation of precursor cells and downregulation of endogenous IRS2 was necessary for proper maturation of HepaRG cells to hepatocytes.

Using pIRS2-GFP and APOA-II-GFP or dsRED lentivirus P4 made reporter cell lines and showed that distinct populations of cells in mature cultures expressed IRS2 or APOA-II, the latter corresponding to the more differentiated cells. All the data suggest that elevated expression of IRS2 marks a transitional state between precursor and mature hepatocyte and that insulin and IRS2 drive the early differentiation of bipotent progenitor cells. This hypothesis has recently been confirmed by double labelling of cells with pIRS2-GFP and APO-II-dsRED showing that APO-II-dsRED cells arise from IRS2 positive cells in the early stages of differentiation (P4).

The capacity of differentiation of hepatic progenitors toward the pancreatic pathway, was evaluated by growing these cells for 12 days in culture conditions inductive for pancreatic progenitors. Cells expressing PDX1/Sox9 could not be generated.

Use of H9 cells reporters for human insulin and Alx-3, markers which facilitate identification of pancreatic progenitor cells confirmed that hepatic progenitors were incapable of differentiating to cells of the pancreatic lineage.

These results from both P2 and P4 suggest that the differentiation protocol developed by the LIV-ES consortium efficiently induces specification of the hepatic lineage since our ADE-derived progenitors do not yield pancreatic cells when subjected to differentiation conditions that generate precursors of the pancreatic lineage. These hepatic progenitors have been fully characterised in vitro and validated in vivo.

WP4 focussed on 'Differentiation of hepatic progenitors into mature hepatocytes'.

Animal-free culture conditions (CDM+factors) have been defined for further differentiation of the population of fetal hepatocytes into more mature hepatocytes. These cells have been characterised for hepatocyte functions (P1A).

Lentiviral vectors expressing GFP under the control of CYP3A4 were also used for hepatocyte characterisation.

Expressions and activities of metabolic enzymes such as cytochrome P450 have been then studied in comparison with HepaRG-derived hepatocytes known to express high levels of drug metabolism enzymes (Aninat et al., 2006). Expression of CYP3A4 was detected and inducible by rifampicin, although no CYP3A4 activity was yet detected. In ES-derived hepatocytes CYP1A2 expression was also inducible by 3Methylcholanthrene (P1C).

Transfection of HepaRG cells at the progenitor stage with a plasmid encoding human CYP2E1 under a CMV promoter was performed to obtain high CYP2E1 expression levels. We showed that transfection allowed enforced expression of CYP2E1 in both progenitors and differentiated HepaRG cells and that theinsulin-dependent mechanism of CYP2E1 regulation occurs only during the process of hepatocyte differentiation (P1C).This opens new means for drug metabolism and toxicity studies.

ES-derived hepatocytes have been partially validated in vivo (one experiment).

HepaRG were also to engraft and clusters expressing human AA1T and cells expressing CK19 were detected.

The 3D culture conditions used for differentiation of hES cell-derived hepatic progenitors into hepatocytes were optimised. HydroMatrix and nanocellulose promote the expression of ALB and AAT, relevant markers for hepatocyte phenotype.

P6B has demonstrated using HepaRG cells that the 3D culture in nanofibrillar cellulose (NFC) improved the cell polarity and the localisation of MRP2 and also MDR1 ABC transporters at the apical membrane of the spheroids, testifying the achieved apical-basolateral cell polarity. The cells cultured in NFC were successfully stained and imaged with confocal microscopy. Nuclei, filamentous actin and some MRP2 and also MDR1 ABC transporters were successfully imaged.

The cell cycle progression is controlled by the sequential induction and/or activation of cell cycle regulating proteins including the Cyclin dependent kinase (Cdk) and their regulatory binding partners the cyclins.

The search for Cdk11 target genes led to the identification of several genes involved in the regulation of the progression in the G1 phase of the cell cycle. To determine whether the Cdk11-cyclin L complexes could have a role in the entry into and progression through the G1 phase in mature hepatocytes and hepatic progenitors HepaRG cells were used as model.

Protocols to synchronise the progression of progenitor and differentiated HepaRG cells throughout the cell cycle were set up. We demonstrated that while Cdk11 is expressed at a constant level throughout the cell cycle, cyclin L1alpha was barely detectable in quiescent cells but rapidly induced in early G1 phase with a maximal expression level in mid-G1. These data provide the first evidence for a cell cycle regulation of the cyclin L proteins with an early induction in G1 phase.

Major conclusions:

1. we have generated different tools to characterise, purify expand and differentiate hepatic progenitors and hepatocytes including reporter lentivectors and cell lines and miRNA platforms.
2. novel 3D cell culture methods using matrices, such as nanofibrillar cellulose and peptide nanofibers, have been established. The studies confirmed that the biomaterials maintain the apical and basolateral cell polarity of mature hepatocytes in 3D culture.
3. immortalised lines of mouse and human fetal liver endothelial cells have been established
4. we have developed a chemically defined 2D-platform for the generation of multipotent, endoderm progenitors from several ES cell lines. Under appropriate conditions, these ADE cells can generate liver, pancreas and lung progenitors
5. we have defined conditions for progenitor expansion and cell maturation and to purify hepatic progenitors. Chemically-defined conditions were also defined to generate fetal cholangiocytes from ES-derived progenitors and from HepaRG progenitors.
6. we have defined conditions to obtain more functionally mature hepatocytes
7. insulin/IRS2 signalling has been identified as a potential target for promoting the proliferation of hepatic precursor cells

All our data pave the way to generate GMP-hepatocytes for clinical application for the treatment of life-threatening liver diseases. They also will provide conditions to generate expanded hepatocytes either at a fetal stage or at a differentiated stage for toxicology studies.

Potential impact:

Chronic liver disease has a significant impact on the well-being of patients and contributes to the significant morbidity associated with these conditions in the EU. Regenerative medicine comprises the development and application of innovative medical therapies aiming to cure life-threatening and/or chronic diseases or support the regeneration of injured cells, tissues or entire organs. The goal is to avoid organ transplantations, pure technical solutions, or permanent pharmacotherapy. This will therefore lead to a complete paradigm shift in medical treatment and not merely new technologies and products. As a result of significant progress in biotechnology, regenerative medicine is evolving as a new discipline that could significantly change therapeutic practices.

Medicine, stem cell biology, tissue engineering, nanotechnology, genomic research, material science and system- and process-engineering are coming together to answer fundamental problems of human disease, including trauma and aging and to enable new therapeutic possibilities. The long-term goal is to create new individualised cell-based therapies that will enable autologous regeneration. One area of potential impact for regenerative medicine is the requirement for alternatives to liver transplantation. In particular, novel regenerative cell-based strategies represent some of the most attractive prospects for new therapies of liver failure.

The LIV-ES consortium has established clear objectives and a delivery plan for translating our increased biological understanding of human embryonic stem cells and hepatogenesis into clinical impacts that will benefit both patients, researchers and perhaps even the economy of the European community. The base of knowledge and methods developed by LIV-ES provide a platform for moving hESC and iPS models towards product development and delivery of cell therapies to the clinic.

Clinical trials of hepatic cells isolated from donor livers have provided convincing evidence that transplantation of these cells can restore liver function, thereby suggesting this approach as a viable alternative to organ transplantation. However, to implement cell-therapy for chronic liver disease, a renewable source of functional human hepatic cells must be developed. hESC are able to grow indefinitely in vitro and differentiate into all cell types of the human body. Although hESC have been differentiated to hepatic cells by various laboratories, the current approaches are not suitable for clinical applications since they rely on animal-derived components which pose risks to human health. The LIV-ES consortium was created to develop reproducible, animal-free conditions for the efficient differentiation of hESC to hepatic progenitors, with the aim of providing the European community with the necessary knowledge and tools to establish a renewable source of transplantable cells for treating chronic liver disease.

LIV-ES has made tremendous progress towards the development of novel reagents and methods for directing the differentiation of hESC to hepatic cells. One of our basic objectives was to create the technological tools to replace animal products in routine culture of hESC. The LIV-ES partners have generated biomaterials which support both the growth and differentiation of hESC and of hepatic progenitor cells. Moreover, we have developed a series of reporter cell lines which are invaluable tools for the identification and purification of cells expressing specific hepatic markers. Other partners have developed a step-wise, chemically-defined protocol for efficiently directing hESC to early-stage progenitors which are committed to the hepatic rather than pancreatic lineage. A renewable source of hepatic cells must expand in vitro: using the reporter tools to identify cells that express specific markers, LIV-ES partners have generated a method to purify hepatic progenitors which can then be expanded in vitro, an important step for future application to cell therapy. The pioneering technology of miRNA has also been employed during the project to identify novel markers of hepatic progenitors. When subjected to further differentiation, these hepatic progenitors yield functional hepatic cells, which secrete albumin and display cytochrome P450 activity and are capable of engrafting in animal models. Finally, the bipotent HepaRG cell line has been a critical model for LIVES partners to further characterise the culture components and cellular mechanisms which regulate hepatogenesis.

Collectively, the results of LIV-ES provide a means for developing hepatocyte-based therapy as a standardised procedure of regenerative medicine. The findings of LIV-ES have been published widely in journals of high impact; thereby furthering the general knowledge of hESC. Thus, the tools and methodology generated by this project are available to both basic and clinical scientists for translation to human liver diseases. The culture conditions developed and validated by LIV-ES will now enable new therapeutic strategies based on stem cells or engineered tissues for restoring liver function.

In particular our data can be transposed to induced pluripotent stem cell models which opens up not only the approach of cell/gene therapy for the treatment of monogenic disorders or as alternatives to ES cells but also in vitro liver disease modelling, as already demonstrated by partners of this project and the wide market of pharmacotoxicology.

At an economic level, the deliverables provided by LIV-ES provide new opportunities for European biotech companies to develop and market the tools for working in animal-free conditions with pluripotent cells. The methodology developed by LIV-ES will enable EU companies, which have the appropriate facilities to move these procedures to GMP conditions for enabling cell therapy.

List of websites:

http://www.liv-es.eu

Partners

Institut National de la Santé et de la Recherche Médicale (Inserm), France; Anne Weber-Benarous/Georges Uzan/ Christiane Guillouzo/Anne Corlu

The Chancellor, Masters and Scholars of the University of Cambridge (UCAM), United Kingdom; Ludovic Vallier

The Hadassah Medical Organisation (HMO) Israel; Eithan Galun

Fundacion de la Comunidad Valenciana Centro de Investigacion Principe Felipe (CIPF), Spain; Deborah Burks

Vectalys SAS (Vectalys), France; Pascale Bouillé

Helsingin Yliopisto (HU), Finland; Arto Urtti/Timo Otonkoski

Inserm Transfert SA (IT), France; Christiane Dascher-Nadel

Coordinator's contact details:

Anne Weber-Benarous, Inserm U972, Hôpital Kremlin-Bicêtre, Bâtiment Gregory Pincus, 80, rue du Général Leclerc, 94276 Le Kremlin-Bicêtre Cedex, France, anne.weber@inserm.fr

Related documents