Current strategies to generate mature human induced pluripotent stem cells derived cholangiocytes and future applications

ABSTRACT

Stem cell research has significantly evolved over the last few years, allowing the differentiation of pluripotent cells into almost any kind of lineage possible. Studies that focus on the liver have considerably taken a leap into this novel technology, and hepatocyte-like cells are being generated that are close to resembling actual hepatocytes both genotypically and phenotypically. The potential of this extends from disease models to bioengineering, and even also innovative therapies for end-stage liver disease. Nonetheless, too few attention has been given to the non-parenchymal cells which are also fundamental for normal liver function. This includes cholangiocytes, the cells of the biliary epithelium, without whose role in bile modification and metabolism would impair hepatocyte survival. Such can be observed in diseases that target them, so called cholangiopathies, for which there is much yet to study so as to improve therapeutical options. Protocols that describe the induction of human induced pluripotent stem cells into cholangiocytes are scarce, although progress is being achieved in this area as well. In order to give the current view on this emerging research field, and in hopes to motivate further advances, we present here a review on the known differentiation strategies with sight into future applications.

KEYWORDS:

• Cholangiocyte
• cholangiopathy modeling
• induced pluriotent stem cells

INTRODUCTION

Over the last two decades, the advances made in hepatic research have been substantial. Hepatocytes, the main cell line within the liver, have constituted a main target for researchers to study the mechanisms of disease development, regeneration, and recently the developing of functional liver tissue, even also bioengineered organs with the aim of solving many of the current liver transplant issues regarding the lack of donors. Nevertheless, there is still much to explore on the non-parenchymal cells, which might be of benefit considering that their main role concerns the supporting and enhancing of normal hepatic function.¹ Cholangiocytes belong to this group of cells and represent only 3% of the total liver cell population,² nonetheless, their role as the epithelium lining the biliary tree is vital for the processing of bile produced by hepatocytes. Cholangitic diseases not only are responsible for significant morbidity and mortality, but also account for the majority of pediatric liver transplants and a significant percentage of young adult liver transplants; unfortunately, at present there are few models to accurately replicate them and seek new therapeutic goals.³ With the rise of stem cell research, the new ideal cell source to generate cholangiocytes in vitro is being generated both for disease modeling and for potential curative alternatives as well, which involves pluripotent stem cells such as embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) whose continuous proliferation and ability to differentiate into any of the three germ-layers is ideal for this purpose. Due to the lack of limitations that ESCs have shown, including immune rejection after transplantation and ethical concerns, iPSCs have appeared as an alternate and reproducible source of differentiated cells with therapeutic interest, making it even feasible to create patient-matched pluripotent stem cells.⁴ In this review, we will describe what is currently known regarding specification, maturation of iPSCs derived cholangiocytes and future applications, whose field has been actively growing and will surely become a key component for future hepatic research in the basic and clinic area.

DIFFERENTIATION OF HUMAN IPSCS INTO CHOLANGIOCYTE-LIKE CELLS

Biliary system development

The development of the biliary system starts when the diverticulum, derived from the ventral foregut endoderm whose cranial part forms the intrahepatic biliary trees and the caudal part generates the extrahepatic bile ducts, extends into the septum transversum and develops the liver bud. To initiate the formation of the intrahepatic bile ducts, the bi-potential hepatic fate specified endoderm cells in the liver bud, which are also called hepatoblasts, proliferate and migrate into the septum transversum differentiate into both hepatocytes and cholangiocytes.⁵ Subsequently, the biliary precursors interact with adjacent mesodermal tissues and transiently form an asymmetrical structure called the ductal plate which, in mouse embryos, is constituted on the portal side by a single layer of SRY-related HMG box transcription factor 9 (Sox9) and cytokeratin (CK)-19 positive cholangiocyte progenitors, whereas on the parenchymal side hepatocyte nuclear factor 4 (HNF4) positive hepatoblasts line next to them,⁶ thus indicating the unique beginning of tubulogenesis. During the development of the ducts, all cells lining their lumen progressively acquire cholangiocyte morphology and function. Guided by the cell orientation system known as planar cell polarity, the biliary epitheliums uniformly maintain the tubular architecture and proliferate in length following a hilum-periphery axis, surrounded by periportal extracellular matrix and mesenchyme, which leads to postnatal development of hepatic artery branches.⁷ Some ductal plate cells which do not participate in bile duct formation transdifferentiate and become the periportal hepatocytes, thus building the canals of Hering.⁸ The formation of the extrahepatic bile duct system, which involves the hepatic ducts, the cystic duct, the gallbladder, and the common bile duct, initiates earlier and may be regulated by several distinct transcription factors and signaling pathways originating from the intrahepatic biliary trees.⁵ Anatomically, these two ductal systems merge at the level of the hepatic ducts/hilum.

The network and pathways of the growth factors that control early biliary development, from cell fate determination to tubulogenesis, is not fully uncovered. Several essential proteins and signaling pathways which may play important roles in cholangitic specification and bile duct morphogenesis have been demonstrated. Li, et al.⁹ suggests that by inhibiting the nuclear factor-kappa B-dependent pathway, which is specifically activated by interleukin (IL) −6, can terminate bile duct development through forkhead box A1 (FoxA1) and FoxA2 transcriptional systems in a mice model. The Notch signaling pathway, constituted by four known Notch receptors (Notch 1–4) and five ligands belonging to the Jagged (Jagged1 and 2) and Delta-like (Delta-like ligand [Dll]1, 3, and 4) family, is well known for its power to determine biliary cell fates and guide proper morphogenesis of the developing biliary tree.¹⁰ Some previous studies about Alagille syndrome carried out in animal models showed that Jagged1/Notch2 interactions regulated potentially during tubulogenesis of the biliary tree after the formation of the ductal plate and initial lineage commitment of hepatoblasts to biliary cells.^11-14 Transforming growth factor beta (TGF-β) can also support cholangiocyte differentiation. TGF-β can stimulate cholangiocytes maturation around portal vein, where there is more expression of the TGF-β receptor type II, without HNF6 and OC2 control.¹⁵ It was shown as well that canonical wingless (Wnt)/beta-catenin and its ligands, such as Wnt3a, induce biliary differentiation, since beta-catenin deficient mice develop impaired maturation, expansion, and survival of hepatoblasts.¹⁶ Another important morphogenetic cue comes from Hedgehog signaling (Hh), by temporally promoting early hepatoblast proliferation; interestingly, mature cholangiocytes do not respond to these signals.¹⁷

During later gestation and postnatally, owing to the maternal feeds, liver metabolism undergoes numerous modifications which may start to act as the main stimuli guiding biliary system maturation until reaching eventual full physiological capacity.¹⁸ Due to both increasing levels in the neonate of bilirubin derived from senescent red blood cells and increasing expression levels of the hepatic enzyme uridine diphosphate glucuronyl transferase (UDPGT), from 1% of adult values at 30 to 40 weeks of gestation to adult levels during the first few weeks of postnatal life,¹⁹ the level of conjugated bilirubin in the canaliculus and bile is elevated. It is transported via the biliary tract to the small intestine and either excreted, or deconjugated and reabsorbed into blood as part of the bilirubin enterohepatic circulation^20,21 all of which may promote maturation of the secretory function of cholangiocytes. The different ratio of the primary bile acid, cholic acid, and chenodeoxycholic acid, as well as the chemical position conjugated demonstrate unique changes of the bile acid synthetic pathway between fetal and adult livers.^22-24 Another aspect to consider is the fact that the bile acid pool size goes through a significant expansion at the end of gestation.²⁵ In infants, it becomes comparable to that of adults at the age of 7 weeks after correcting for body surface area,²⁶ and can be explained by the increased activity of enzymes involved in bile acid synthesis among hepatocytes just before birth.²⁷ During the first week of life, the primary bile acids circulating in the serum reach a significantly higher concentration compared to healthy children and adults, which may accelerate the maturation of channels and receptors in cholangiocytes.²⁸ With regard to the neonatal gallbladder, its complete development is necessary for efficient bile flow into the duodenum. Additionally, the sodium bile acid co-transport activity of the terminal ileum also attains full function postnatally, normally by the time of weaning.²⁹

Generating cholangiocytes from human iPSCs derived hepatoblasts

Given that hepatic specification initiates from hepatoblasts which originate both hepatocytes and cholangiocytes, the derivation of hepatoblasts from pluripotent stem cells has been sought as a main objective by researchers.³⁰ According to the formation of hepatic tissue in vivo, several of these well know stepwise protocols (developed in mouse and human ESCs) employ the formation of embryonic bodies (EBs) and short exposure to TGF-beta superfamily members activin A and fibroblast growth factor 2 (FGF2) to generate definitive endoderm (DE) cells, followed by hepatoblast specification which requires both 1% dimethyl sulfoxide (DMSO) and hepatocyte growth factor (HGF).^31,32 After being proven that such differentiation is also possible in mouse iPSCs,³³ Hay, et al.³⁴ devised a non-EB method to generate hepatic-like cells from human ESCs that only requires DMSO for hepatoblast formation,³⁴ and proved the important role of Wnt3 during the DE commitment stage; the utility of this method has been proven and applied in several human iPSCs studies.^35,36,37 Song, et al.,³⁸ who were the first to report the hepatic differentiation of human iPSCs, reported hepatoblast induction, as a result of the combined stimuli of HGF and keratinocyte growth factor. Si-Tayeb, et al.³⁹ developed an iPSCs hepatic differentiation protocol by using bone morphogenic protein 4 (BMP4) and FGF2. Two different teams which recently established their own differentiation procedures to derive cholangiocyte-like cells used this protocol as hepatoblast induction.^40,41,42

Dianat, N. et al. settled for the first time robust conditions to drive differentiation of human pluripotent stem cells (ESCs, iPSCs) derived hepatoblasts into functional cholangiocyte-like cells using feeder-free and defined culture settings, including the use of growth hormone, epidermal growth factor (EGF), IL-6, and sodium taurocholate. The cholangiocytic lineage obtained was found to express cell-specific markers including CK-7 and osteopontin, the transcription factors SOX9 and hepatocyte HNF6 and functional proteins including cystic fibrosis transmembrane conductance regulator (CFTR), secretin receptor (SR), and nuclear receptors; primary cilia and response to hormonal stimulation were observed as well. In a three-dimensional (3D) matrix, those cells developed epithelial/apicobasal polarity and formed functional cysts and biliary ducts.⁴³ Afterwards, two systematic studies were published applying different approaches to generate specific cholangiocyte-like cells: Ogawa, et al. described a protocol that achieved efficient differentiation of cholangiocytes from iPSCs by means of a 3D co-culture of hepatoblasts with OP9 stromal cells and the combination of different growth factors including HGF, EGF and TGF-β,⁴⁴ while Sampaziotis, et al. emphasized on the pre-cholangiocyte stage, which can be called cholangiocyte progenitor specification, using FGF10, Activin A, and retinoic acid before the 3D condition.⁴⁵ Both studies accomplished creating functional cells that acquire epithelial functions such as rhodamine efflux and CFTR-mediated fluid secretion, and that yield cystic and/or ductal structures that express mature biliary markers, including apical sodium-dependent bile acid transporter, SR, cilia and CFTR. Ogawa et al. demonstrated the importance of NOTCH signaling in cholangiocytes differentiation, whereas Sampaziotis et al pointed out the functional characteristics of cholangiocytes, which comprises bile acids transfer, alkaline phosphatase activity, gamma-glutamyl-transpeptidase activity and physiological responses to secretin, somatostatin, and vascular endothelial growth factor. Another novel stepwise cholangiocyte differentiation strategy was designed by De Assuncao, et al., who fully characterized the cells in vitro and in vivo, by exposing them to high doses of Jagged1 and TGF-β. In vivo, the cells engrafted within mouse liver following retrograde intrabiliary infusion.⁴⁶ The most recent study which applies a similar differentiation protocol was carried by Dianat, N. et al., who proved that laminin could promote gene expression levels of the cholangiocyte markers, AQP1, SOX9, CFTR, G protein-coupled bile acid receptor 1 (GPBAR1), JAG1, SCTR, and gama-glutamyl transferase (GGT1).⁴⁷

Current and potential strategies for maturation of human iPSCs derived cholangiocyte-like cells

The evaluation and control of the transcriptional network and markers for maturation

Defining and controling a unique transcriptome that allows hepatoblasts to enter the biliary lineage and maturate into cholangiocytes instead of hepatocytes, should be the first strategy for optimizing this type of iPSC differentiation process. Such an approach has been already considerably taken advantage of in phenotype characterization. The cholangitic transcriptional network has been found to be closely linked in part to factors that are crucial to define further growth of the endoderm into either the pancreatic or hepatic lineage, since the suppression of the marker genes of the definitive endoderm Sox17 and FoxA2 (needed to specify foregut endoderm and subsequent derivation), results in cholangiocyte hyperplasia associated to an excessive expression of IL-6.^48,49 During the hepatoblast stage, T-box transcription factor (Tbx)-3 not only takes part in hepatoblast migration into the septum transversum along with other factors such as prospero-related homeobox (Prox)-1, but is fundamental too in increasing the expression of HNF 6 and HNF1β and decreasing the level of HNF4α. This is an essential process, as HNF4α together with alpha-fetoprotein (AFP) are key markers of hepatocyte fate commitment; consequently, their expression should be repressed in order to facilitate biliary differentiation.^50,51 When progenitor cells reach the cholangiocyte stage, an important mediator is HNF-6 whose expression, and that of another related factor, onecut (OC)-2, is much higher in cholangiocytes than in hepatoblasts or hepatocytes, and is related to the modulation of TGF-β gradients.⁴⁹ Furthermore, the relevance of the liver enriched factors HNF6 and HNF1β concerns not only their embryological role, but their function in normal biliary physiology as well, as both influence bile acid metabolism besides participating in other hepatic pathways. Interestingly, Prox-1 is repressed by the factor liver receptor homolog (LRH)-1, which also contributes to cholangiocytes physiologically.⁵² Another factor with a significant role is SOX9 which can be observed first supporting normal TGF-β activity in the cells that will constitute the ductal plate that, as mentioned previously, maintains cholangiocyte differentiation. Seemingly, it is downstream of HNF6 and upstream of CAAT/enhancer-binding protein (C/EBP)-α, a hepatocyte enriched transcription factor known previously as HNF2. The latter must be inhibited by SOX9 in order to allow normal biliary-favoring signaling of HNF6 and HNF-1β, even though low C/EBPα expression is also associated with low HNF-6 levels which have been found to induce a hybrid phenotype.⁵³ Cytokeratin 19 is expressed at low levels in hepatoblasts and at increasing levels in differentiating cholangiocytes, nonetheless, it is not a good indicator of cell fate commitment. On the contrary, cytokeratin 7 becomes detectable when cholangiocytes are already committed to a cholangiocyte lineage.⁵⁴ Much is yet to uncover regarding other pathways for which there is some evidence of biliary participation. The need is clear for better models that will enable the study and precise characterization of the factors that mediate the embryogenesis of the biliary system (Fig. 1).

Functional maturation stimulus from the bile acids

Physiologically, cholangiocytes carry out the alkalinization of the bile produced by hepatocytes, and contribute to the modification of this fluid with other secretory and absorptive functions, all of which accounts an approximate 30% of total bile production.⁵⁵ The processing of bile acids is certainly essential to digestion, nonetheless, it also represents an important signaling system which can guide proliferation and growth in cholangiocytes. This has been shown in the case of taurocholic, taurolithocholic, and glycodeoxycholic acid, while ursodeoxycholic acid has the opposite effect of inhibiting proliferation.⁵⁶ Interestingly, some of these bile acids, namely tauroursodeoxycholic acid and ursodeoxycholic acid, additionally indirectly stimulate cholangiocytic secretion of fluid and electrolytes through changes in cytosolic calcium, which are caused by ATP released into bile by induced hepatocytes.⁵⁷ Furthermore, taurocholate when infused in rats was found to induce structural changes of the bile ducts, including an increase in canalicular branching and redundancy of canalicular microvilli.⁵⁸

Since patients with primary biliary cholangitis, a disease that targets the bile ducts, display histologically hyperplasia of cholangiocytes at early stages, it can be inferred that bile acids exert favoring activity on the biliary epithelium.⁵⁹ This has been replicated with the so-called bile duct ligation models which have demonstrated an increase in the number of cholangiocytes due to ductular proliferation.⁶⁰ Experiments on hepatocyte transplantation have even shown that when transplanted in the spleen and by ligating the bile duct, hepatocytes develop bile canaliculi resembling Hering's canals onto which they cluster.⁶¹ In a contradictory manner, bile stasis is an important factor leading to cytoxicity, oxidative stress, and apoptosis, as is seen for example in cirrhotic livers, nonetheless, in cholangiocytes it appears to have a carcinogenic effect by promoting proliferation and inflammation.⁶² In addition, bile acids bring about the benefits mentioned above concerning secretion, which in the case of taurocholic and taurolithocholic acid include an increase in gene expression of the secretin receptor, cAMP levels, and Cl−/HCO3− exchanger activity.⁶³

Figure 1. During generation of iPSCs derived cholangiocytes, the key transcription factors that guide the specification of definitive endoderm cells (DE) and hepatoblasts stages are pointed out in the figure, which should be well evaluated. Those factors that favor commitment of hepatoblasts to the hepatocytic lineage should be down regulated during the cholangiocyte differentiation. Several markers are showed in the figure that can be used to recognize mature cholangiocytes, and thus evaluate the differentiation process; these also include cell ultrastructural components and transporters that are essential for normal cholangitic function.

As to the molecular factors activated through bile acid signaling, it appears that the key initial triggers are their binding to the nuclear receptor farnesoid X-receptor (Fxr) and/or the plasma membrane G protein-coupled bile acid receptor TGR5.⁵⁶ In hepatocytes, for instance, Fxr activation by bile salts causes the short heterodimer protein (Shp) to block retinoid-X nuclear receptor/retinoic acid receptor (RXR/RAR) from binding to the promoter of the sodium taurocholate-cotransporting polypeptide (NTCP), a bile acid transporter protein expressed in the basolateral membrane of hepatocytes.⁵⁵ In cholangiocytes, the knockout of Tgr5 prevented their proliferation in mice bile duct ligation models, and on the other hand, bile salt stimulation of this receptor was shown to result in an increase in cAMP, biliary bicarbonate secretion, and secretin receptors in the human gallbladder epithelium.⁶⁴ Regarding Fxr activation in cholangiocytes, it is linked to cytoprotective and anti-cholestatic properties, which is attributed indirectly to the inhibition of CYP27 expression, a rate determining enzyme of bile acid synthesis.⁶⁵

Maturation in extracellular matrix

A key component that is also relevant and should be taken into account when culturing and differentiating pluripotent cell lines, is the extracellular matrix. The use of specific basal lamina components such as collagen or fibronectin, and not only Matrigel whose conformation isn't completely known and derives from mouse Engelbreth–Holm–Swarm tumors, has proved to favor cell-specific morphologic changes and gene expression as well.⁶⁶ In regard to this, the team of Tanimizu et al. demonstrated that by culturing a liver progenitor cell line with laminin- 111 (α1β1γ1), one of the major components of Matrigel, cyst formation was promoted and apicobasal polarity that is typical for cholangiocytes could be observed.⁶⁷ Laminin is a heterotrimer glycoprotein composed of α, β, and γ chains, of which five, four and three variants exist respectively, and a total of 18 isoforms can be found in mammals; laminin receptors include syndecans and dystroglycans, but the main adhesion receptor of human embryonic stem cells are integrins which are made of α and β chains.⁶⁸ In a further study, it was shown that β1 integrins are the main ones that activate the signaling system from laminins that leads to cyst formation in hepatic progenitor cells, being α6β1 the most characterized. Moreover, α−1 containing laminin is an important but dispensable extracellular matrix component needed to promote differentiation into cholangiocytes, while α−5 containing laminin is not only necessary for cyst formation, but for bile duct formation as well since knockout livers of this laminin presented a decrease in mature ductal structures and a reduced size of the luminal space of the remaining ducts.⁶⁸ This can be expected, considering that α−1 containing laminins are relevant to the regenerating liver, as was shown with the transient expression of laminin 111 in the liver lobules of mice after two-thirds partial hepatectomy, and α−5 containing laminins (specifically laminins 511 and 521) are expressed in the basal lamina of mature epithelial tissues including bile ducts, central veins, and hepatic arteries.⁶⁹ Taking into account that the extracellular matrix is equally relevant to growth factors and the culture media for the differentiation of iPSCs into cholangiocytes, Tanimizu et al.⁶⁸ set out to prove the utility of various laminin isoforms. Due that the mass production of intact, full-length laminins is technically difficult and requires weeks for purification, shorter recombinant E8 fragments were used as reported by Mizayaki et al. which can be purified in days, have a higher yield, and retain full α6β1 integrin-binding capacity; in addition, E8 laminin fragments promote a greater adhesion of human iPSCs than substrates like Matrigel and intact laminin 511.⁶⁶ Thus, it was proved that culture with laminins 411 E8 and 511 E8 supported the cholangiocyte differentiation from hepatoblast-like cells, and on the contrary, laminin 111 E8 and also 121 E8, 211 E8, and 221 E8 inhibit this transition. Further studies will be needed to characterize the pathways activated by these laminins and to optimize culture conditions.

Future applications of human iPSCs cholangiocytes in basic and clinical research

Modeling of cholangiocyte disease

Cholangiopathies are a set of disorders of the epithelial cells that line the bile ducts, commonly known as cholangiocytes, apparently caused by genetic and environmental factors and that can progress to end-stage liver disease, or even suffer malignant transformation.³ Their significance resides in the fact that they are the leading cause in up to a third of all liver transplants done on adults, with primary biliary cholangitis and primary sclerosing cholangitis being the principal leading cholangiopathies. In children, biliary atresia represents the main reason for liver transplant.⁷⁰ The progressive developments of generating mature cholangiocyte from human iPSCs are opening new possibilities in the field of cholangiopathies, encouraging the conception of experiments that more closely resemble pathology under human conditions and holds great promise for new and precise human models that will enable a better understanding of disease and pathophysiology, as virtually any type of anomaly can be recreated, either artificially by gene manipulation or by obtaining and reprogramming patient cells. The two recent studies of genarating iPSCs derived cholangiocytes also explored this area. Ogawa et al. developed cholangiocytes from patients with cystic fibrosis (CF) carrying the characteristic F508del mutation of this disorder, which interestingly had delayed cyst formation even after the addition of the folding defect corrector molecule VX-809 (now FDA approved as Lumacaftor). Nonetheless, treated cysts that were potentiated with VX-770 (also FDA approved, Ivacaftor) showed significant swelling compared to untreated cysts after induction with forskolin/IBMX.⁴⁴ In a similar manner, the team of Sampaziotis et al. decreased CFTR function with improvement and increased organoid size after the addition of corrector molecules on CF patient derived cholangiocytes.⁴⁵ Such an effect is achieved clinically in the polycystic liver disease (PLD) cholangiopathy, characterized by cystic lesions due to benign cholangiocyte hyperproliferation, but with octreotide, a synthetic somatostatin analog.³ Of worth noting, both disease models have shown valid applications of this iPSCs not only in the proving of pathophysiological interactions, but in biliary-specific pharmacological screening as well. In the close future, functional grafts all from iPSCs are also expected, potentially serving as treatment for diseases with no other cure available.

In present literature, no other cholangiopathy has been modeled with iPSCs technology, though sufficient examples exist that could serve as a basis. The advances that have been made with CF and PLD prove that innovative experiments can be designed, as compared to earlier models, which relied only on animals with induced deficiency or mutation of characteristic genes, Cftr or ΔF508 in CF, and sec 63, prkcsh, pkd1, or pkhd1 in PLD.^71-73 Modeling with iPSCs permits the direct introduction of disease-specific genes in the target cells, with consequent observation and follow up until pathology settles. This also will allow the testing of new drugs before introducing them on animal models, which may be easier to achieve given that pathophysiology can be inferred with more detail. In biliary atresia, perinatal obstructive fibrosis of the bile ducts due to unexplained phenomena occurs, though attributed to both the environment and genetics. Currently, models for this disease have found that sensitization to viral particles, including rotavirus, reovirus or cytomegalovirus can lead to biliary obstruction. Nonetheless, certain susceptibility yet to decipher triggers an exaggerated immune response that induces fibrosis, perhaps related to autoimmunity although patients don't benefit from corticosteroid use.^74,75 Thus, it will be interesting to confirm such an etiological hypothesis with iPSCs cholangiocytes derived from biliary atresia cases, and genetic alteration may define potential therapeutic targets. Primary biliary cholangitis and primary sclerosing cholangitis are similar cholangiopathies in the sense that both are characterized by chronic immune-mediated inflammation and destruction of the bile ducts, although their differences are evident including a female preponderance and distinctive antimitochondrial antibodies in the former, and male preponderance and probable HLA antigen-B locus related susceptibility in the latter.^3,76 As with the other cholangiopathies discussed, their etiology is still not clearly elucidated, in part because of the complex interaction between environment and genes. Current animal models have recurred mainly to genetic alterations, xenobiotic immunization, infectious agents, and chemical induction, though neither one can fully show all disease-specific characteristics.^76,77 The use of iPSCs technology will allow the possibility of directly studying pathological mechanisms with patient-specific cholangiocytes, and perhaps solve what determines each clinical condition. The idea of personalized medicine is also an attractive one, as for example, not all patients with primary sclerosing cholangitis respond to ursodeoxycholic acid therapy, and long-term consequences of the use of this drug in patients with primary biliary cholangitis are unknown.³ In the case of cholangiocarcinoma, no specific etiology can be attributed to the disease, although chronic inflammation with concomitant malignant transformation is an important factor considering the fact that primary sclerosing cholangitis and polycystic liver disease are the main risk factors involved.^3,78 It appears that cholangiocarcinoma is the result of genetic alterations, induced by any of the risks factors, as was proven in a hamster model where infection with the helminth parasite Clonorchis sinensis in hamsters caused upregulation of oncogenes and downregulation of tumor suppressor genes.⁷⁹ Accordingly, it is tantalizing to recreate an inflammatory microenvironment in iPS cholangiocyte cultures that favors dysplasia. Multiple ways to achieve this setting are to be found, cholestasis already being a common tested variable in bile duct obstruction models, including direct exposure to inflammatory cytokines or IL-33, a biliary mitogen.^80,81

iPSCs cholangiocytes in cell therapy, scaffolding, and bioprinting

As the generation of functional differentiated cholangiocytes is becoming an easier goal to reach, it is an attractive idea to use them as part of cell therapy, as has been intensively researched with hepatocytes. Although non-parenchymal cells represent only a small fraction of the total liver mass (6.5%), they are an essential component with functions that benefit and enhance overall hepatocyte metabolism, thus possibly favoring their survival which in regard to cell therapy could represent reduced hepatocyte amounts needed to achieve a therapeutic effect.^82,83 In the case of cholangiocytes, a conceivable advantage is the complementing of bile canaliculi with concomitant elimination of hepatocyte end products of metabolism.⁸⁴ The extent to which cholangiocyte transplant has been explored is short, reflecting the broader focus on hepatocytes. After finding that intrasplenic transplantation of autologous adult hepatocytes in Lewis rat recipients formed hepatocyte cords but not bile canaliculi when pretreated with a mixture of irradiation, intravenous anti-asialo GM1 antiserum, and syngeneic bone marrow cell reconstitution, Olszewski et al.⁸⁴ added to their protocol the ligation of the common bile duct, which resulted in the development of these biliary structures. Interestingly, it was found that transplanted hepatocytes and not autologous cholangiocytes responded to an unknown signal in the bile stasis model, creating canaliculi with CK9 and CK19 and gamma-glutamyl transpeptidase positive cells. Thus, it appeared that hepatocyte intrasplenic transplantation was sufficient for laying out the basis of the biliary architecture as long as there is ligation of the common bile duct, probably due to a fraction of hepatocytes with cholangiocyte progenitor cell properties, while cholangiocyte transplantation did not lead to either hepatocyte or canaliculi development, probably indicating a more mature and differentiated phenotype.⁸⁴ Considering the seeding of scaffolds for bioengineered organs using both iPSCs derived cholangiocytes and hepatocytes, the question arises if cholangiocytes would react to the bile canaliculi formed by the latter and establish a functional biliary system that would promote even more hepatocyte growth.

In regard to liver engineering, the team of Baptista et al. set out a method for making liver 3D scaffolds by decellularization and attempted to recellularize them.⁸⁵ Although they used human fetal hepatocytes and human umbilical vein endothelial cells for the recellularization of the hepatic parenchyma, they prove that reepithelization of the bile ducts occurs as immunostaining for CK-19 was positive in biliary tubular structures, and surrounding them were cells that either stained for CK18, a marker of hepatocytes, or both CK-19 and CK-18, indicating also immature cells with hepatoblast properties.⁸⁵ Their results are the only ones in current literature that point out to the possibility of biliary reseeding, and in this case show the importance of the extracellular matrix provided by the scaffold in the engraftment of specific cells which react to the cues given by the various matrix molecules. Coinciding in a way with the results of Olszewski,⁸⁴ it is clear that hepatocytes remain an essential factor in cell therapy and bioengineering, due to their plasticity when either in their adult or fetal forms. Thus, for future experiments that involve iPSCs cholangiocytes, it might be useful to include both mature hepatocytes and hepatoblasts, in an effort to set up a basis of biliary architecture that will promote cholangiocyte engraftment and growth.

An alternative to decellularization methods in order to obtain hepatic scaffolds is to recur to bioprinting, an emerging area of research that promises the creation of 3D tissue structures that resemble those in the normal organ of origin. Despite the technical difficulties that this method implies, including cell damage due to biomechanical forces, the maintaining of the cell culture in the printed structure, and preserving normal cell metabolism,⁸⁶ bioprinting might perhaps become a faster and more accurate procedure to generate disease models than organic scaffolds. Whole organ manufacture is one possibility, nonetheless, small scale samples (associated with the concept of “liver on a chip”) are more accessible and may mimic the tissue's normal physiology.⁸⁷ Few publications are at present focused specifically on liver tissue bioprinting, and as is seen with existing reports on liver scaffolds, the biliary system is for now left out. An early indication on the incorporation of cholangiocytes is mentioned briefly in a recent article that reports the use of digital light processing-based 3D printing to recreate liver lobules which supported iPSCs hepatic progenitor cells, endothelial cells and adipose-derived stem cells; although biliary epithelium wasn't characterized in this model, the authors suggest the hepatoblasts printed on the lobules may have been favored to differentiate into this tissue as well, which would indicate maturation and functionality of the 3D structures.⁸⁸ Certainly, the addition of iPSCs derived cholangiocytes in bioprinting may help maintaining the threatened hepatocyte mass on the manufactured scaffolds, though their interaction with printed biliary ducts and the rest of the cells will be an interesting aspect to demonstrate.

As emphasized by Mubbach et al, there are three important goals to accomplish, if a true liver is to be produced by bioengineering methods: the reseeding of the parenchyma, vascular endothelization, and biliary endothelization.⁸⁹ Unfortunately, too few attention has been given to the latter goal, perhaps owed to the feat of obtaining a sufficient amount of cholangiocytes with a stable phenotype. Until recently, the creation of iPSCs derived cholangiocytes through high efficiency protocols should represent a means to curve this problem. Nonetheless, once this is settled, the problem of cell mass production will soon be needed to solve. The advances accomplished in these fields of research will be in much need not only for cholangiopathies, but for many other liver diseases with no other hope than liver transplant.

CONCLUSION

As we have mentioned throughout, cholangiocyte-specific research is scarce at present, though it's clear that a greater interest is now being directed toward their generation and application in regenerative medicine and toward the creation of better and improved models of disease. While hepatocytes may be the fundamental piece when aiming for any of these objectives, one cannot leave behind the rest of the cells that also conform the liver mass, and whose function is clearly essential in many house-keeping tasks. Cholangiocytes may be the second most important cell in this organ, with bile ducts forming part of the portal triads that demarcate the hepatocyte-rich lobules. Bile is a toxic substance with corrosive properties when not processed adequately, as is evident in patients with cholangiopathies, thus, the inclusion of biliary epithelial cells should be pursued if a true liver graft is to be bioengineered; this also points out to the importance of giving continuity to the models already made for these pathologies, so as to find new therapeutic targets that can be altered either pharmacologically or with an innovative regenerative approach. The review we have presented should inspire the developing of better cholangiocyte differentiation and culture methods, which will be of much need in this continually expanding stem cell study field.

DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST

No potential conflicts of interest were disclosed.

FUNDING

This work was supported by the Natural Science Foundation of China grant # 81470874, 81470874 for the stipend of Y.W.