Architectural and Immunohistochemical Characterization of Biliary Ductules in Normal Human Liver

Abstract

The canals of Hering or biliary ductules have been described to connect the bile canaliculi with the interlobular bile ducts, and thus forming the distal part of the biliary tree. Studies in the last two decades suggested that the cells constructing these ductules could behave as hepatic progenitor cells. The canals of Hering are confined to the periportal space in the rat, while they have been reported to spread beyond the limiting plate in human liver. The distribution of the distal biliary ductules in normal human hepatic tissue has been investigated in our recent experiments. We could demonstrate the presence of interlobular connective tissue septa in a rudimentary form in healthy livers. The canals of Hering run in these septa in line with the terminal branches of the portal vein and hepatic arteries. This arrangement develops in the postnatal period but regresses after early childhood. The canals of Hering can be identified by the unique epithelial membrane antigen (EMA)−/CD56+/CD133+ immunophenotype. The canals of Hering leave the periportal space and spread into the liver parenchyma along rudimentary interlobular septa outlining the hepatic lobules. Our observations refine the original architectural description of the intraparenchymal portion of the canals of Hering in the human liver. The distinct immunophenotype supports their unique biological function.

Introduction

The existence of a progenitor cell population in the liver has become generally accepted [1,2], and the clinical application of these cells have been of tremendous interest [3,4]. Today, liver transplantation is the only available curative treatment for liver failure, either in cirrhosis or in fulminant liver necrosis. However, the number of available donor livers sets the limit for the application of this procedure and alternative treatments are being sought. Transdifferentiation of bone marrow cells into hepatocytes does not seem to be efficient enough for clinical application [5 –7]. Conversely, efficient hepatic regeneration has been recorded from the endogenous liver progenitor cells in human [8,9].

Most data refer to the canals of Hering as the site of the hepatic progenitor cell compartment [1,2,10]. This structure was described by Hering as “hepatic capillaries” [11], which maintained the link between bile ducts and the hepatocyte canalicular system. Later, it had been proposed to be the niche for hepatic progenitor cells [12,13]. The niche is a special microenvironment, which has a major impact on the maintenance and activation of the stem/progenitor cell compartment [14]. Therefore, the exact identification and characterization of this structure is necessary to understand its behavior under normal and pathological conditions.

The canals of Hering are usually shown as short, straight ductules at the border of the periportal connective tissue and liver parenchyma, but probably this conformation is oversimplified. We have recently characterized the hepatic progenitor cell niche in rat liver by laser scanning confocal microscopy [15]. Long, branching ductules have been observed, strictly inside the periportal connective tissue. They had contact with the bile canalicular system at the limiting plate. Their unique CK19+/CK7− immunophenotype has made their identification within the biliary tree easier. However, no CK7− biliary structures could be observed in human liver specimens. Furthermore, the canals of Hering have been described to spread into the hepatic lobule in the normal human liver [16]. These observations suggest that the organization of the canals of Hering in human liver is different from the traditional simple view as well as from the architecture we saw in rat liver.

In our present study, we set out to collect normal human liver tissue from individuals of various ages and analyzed the architecture and immunophenotype of the biliary ductules by confocal microscopy.

Materials and Methods

Normal human liver specimens were collected from cadavers of spontaneous premature birth neonates without developmental abnormalities and individuals who died suddenly in accidents without morphological signs and anamnestic data of any liver disease. (Age and gender of the patients: 3 males, 23rd week of pregnancy; 4 females, 23rd week of pregnancy; 1 male, 39th week of pregnancy; 2 males, 3 years; 1 female, 3 years; 1 female, 13 years; 1 male, 20 years; 1 male, 26 years old.) All autopsies were performed within 24 h following death. The liver samples were thoroughly examined on formalin-fixed, paraffin-embedded liver sections with H&E, diastase PAS, Prussian blue, orcein and Masson’s trichrome stainings; no fibrosis, ductular reaction, or other pathological alterations were observed. Snap frozen liver samples were stored at −80°C.

Frozen sections (10–20 µm) were fixed in methanol and were incubated at room temperature (1 h) with a mixture of primary antibodies (Supplementary Table 1;Supplementary materials are available online at http://www.liebertpub.com/) followed by the appropriate fluorescent secondary antibodies (JacksonImmunoresearch, West Grove, PA). All samples were analyzed by confocal laser scanning microscopy using Bio-Rad MRC-1024 system (Bio-Rad, Richmond CA).

The procedure has been approved by the ethical committee of Semmelweis University.

Quantitative analysis of immunohistochemical staining

The livers of two 3-year-old children and the 20-, 26-year-old adults were used for quantitation. Consecutive frozen sections were co-stained for CK7/EMA, CK7/CD56, and CK7/CD133. The number of CK7-stained structures was determined. The double-stained structures were counted and the results were given as percent of the CK7+ structures.

Results

Distribution of hepatic ductules in the liver parenchyma

When sections from the livers of 3-year-old healthy children were stained for panCK, CK7, and laminin, hepatic ductules surrounded by basement membrane could be observed in the parenchyma far from the portal spaces (Fig. 1A and 1B). Low power examination revealed that these ductules were not randomly arranged. They outlined dimly polygonal structures with terminal veins in the centers and portal triads at the corners, that is, the classical hepatic lobules. When micrographs of 40 serial sections stained for cytokeratin-7 were digitally aligned and merged (Fig. 1C), this kind of perilobular arrangement of the ductules was even more obvious. Two other characteristics could be observed on this composite image: (i) no CK7+ biliary ductules were present inside the hepatic lobules; (ii) the ductules at the interlobular border spread until the half of the porto-portal distances, which resulted in watershed-like empty gaps in the middle of these stretches.

FIG. 1.

Confocal images of normal human liver from a 3-year-old child. (A and B) Triple labeling for CK7 (red), laminin (blue), and pan CK (green). (A) CK7 (red) and laminin (blue) staining. (B) Merged image. Comparing the two images, the perilobular arrangement of the laminin-framed ductules is clearly discernible. (C) CK7 (green)-stained biliary ductules sharply outline the hepatic lobule when 40 thick serial sections are merged. Note the “gaps” (arrows) halfway of the porto-portal distances. (D) Horizontal view of 42 optical sections stained for panCK. Note the numerous connections of the bile ductules with the liver parenchyma (arrowheads). The collecting bile ductule is running in an “empty” space toward the portal area. Scale bars for A and B = 500 µm; C = 100 µm; and D = 50 µm.

High power examination of individual biliary ductules showed that these narrow tubules did really extend beyond the limiting plate. The ductules spread in virtual “empty” spaces among hepatocytes on cytokeratin antibody stained sections (Fig. 1D), where only the epithelial elements of the hepatic tissue were decorated. When the ductules terminated on hepatocytes, they were surrounded by typical U-shaped basement membrane (Fig. 2B inset), and the hepatocytes did not participate in the composition of the ductules beyond these connections. The “empty” space around the ductules was filled by collagenous matrix (Fig. 2A) and contained CD31+ small blood vessels (Fig. 2B), some of which were also labeled by the arterial marker NG-2 [17] (Fig. 2C). Taken together, confocal analysis of normal human liver revealed the deposition of small amounts of extracellular matrix between hepatic lobules, with expanding biliary ductules and blood vessels. This kind of arrangement of the blood vessels was described earlier as the vascular septum [18,19]. The small amount of matrix could not be visualized on traditional histological sections by special stainings on any of our liver samples.

FIG. 2.

Confocal images of normal human livers from a (A and B) 3- and a (C) 13-year-old child. (D–F) hepatic lobules in (D) 23- and (E) 39-week-old fetuses and (F) in a 26-year-old adult liver. (A) Double staining for CK7 and collagen I. CK7 (green)-stained perilobular ductules are embedded in rudimentary collagenous matrix (red) in the portal spaces and in the interlobular septa. The inset shows the vascular septum at a higher magnification on a different section. (B and C) High power view of the “vascular septum.” (B) Laminin (blue) surrounded CK7+ (green) ductules are accompanied by CD31 (red)-decorated blood vessels (arterioles). The larger vessel (arrow) probably represents a terminal branch of the portal vein. The inset shows the connection of a ductule on the hepatic plate (cannot be seen with this staining); note the sharp ending of the U-shaped basement membrane at the ducto-parenchymal border (arrowheads). (C) Note the proximity of NG-2+ (red) arterioles to the CK7 (green)-stained ductules. The empty laminin (blue) circles (arrows) probably represent portal vein branches. (D) The portal tracts are “closed”; laminin (red) surrounded CK7+ (green) ductules are confined to the border of the portal space. Note the high number of ductules marking the ductogenesis from the ductal plate. (E) Thy-1 (red)-stained myofibroblasts initiate the formation of the interlobular septa with a few CK7+ (green) ductules (arrows). The CYP450 (blue) staining shows the zonality of the hepatic lobule with a Thy-1-positive terminal vein in the center. (F) Laminin (red) and CK7 (green) marked ductules outline the hepatic lobule in the adult liver, but they are rarer in the septa than in children (compare with Fig. 1C, 1D). Scale bar for A, D, E, and F = 100 µm; and B and C = 50 µm.

Alterations of the hepatic ductules with age

The maturation of the biliary system continues in postnatal life in humans [20,21], and major changes are also observed in rats [15,22]. Therefore, we decided to examine the distribution of the canals of Hering in healthy livers of individuals of various ages.

The portal areas in the liver of immature neonates born on the 23rd week of pregnancy were “closed.” There were numerous bile ducts in the periportal connective tissue, especially at the periphery, as the remnants of the ductal plate. However, no signs of vascular septa were seen; no matrix deposition, blood vessels, or biliary ductules could be observed outside the limiting plate in any of the examined specimens (Fig. 2D).

Early signs of vascular septum formation could already be recognized in a liver sample derived from a fetus of the 39th week of pregnancy. It was mostly outlined by Thy-1-positive myofibroblasts, but a few CK7+ ductules were also present in these septum fundaments outside the portal fields (Fig. 2E).

The vascular septa and all of its components were mostly developed in the livers of young children of the age of 3 years (Fig. 1A and 1B). The only available liver specimen from a 13-year-old girl contained relatively regressed vascular septa, and although all the elements described earlier were present in the livers of young adults (20 and 26 years), they were more scarce (Fig. 2F).

Immunophenotypic characterization of bile ductules

There are several proposed markers for hepatic progenitor cells in human liver (1,2), but most of them did not distinguish hepatic ductules of the vascular septa from larger interlobular bile ducts in our hands. Some of the markers (AFP, chromogranin, synaptophysin, DMBT, DLK, CEA, CK20, CK14) did not label any biliary structures, while others (EpCAM, E-cadherin, CK7, CK19) stained the complete biliary tree (data not shown). Only three markers reacted differentially with bile ducts and ductules. Epithelial membrane antigen (EMA) resulted in a very sharp characteristic linear apical staining in the interlobular bile ducts. Conversely, it was absent in the ductules even on cross sections (Fig. 3A and 3B). The staining pattern of CD133 (Fig. 3C) and CD56 (Fig. 3D, 3E, and 3F) was opposite. In specimens up to the age of 3 years, there was a consistent apical CD133 staining in all segments of the biliary tree. However, in samples from older individuals the staining was strictly confined to the small ductules of the vascular septa. The distribution of CD56 in all specimens was similar to this latter case: it labeled exclusively the small ductules.

FIG. 3.

Immunophenotypical characterization of the hepatic ductules in (A–C) adult and (D–F) 3-year-old livers. (A and B) There is a sharp apical epithelial membrane antigen (EMA) (red) staining in CK7 (green) and laminin (blue) marked bile ducts (arrows) inside the portal space, while EMA staining is absent in the small ductules (arrowheads) at the periphery. For better visibility of the EMA staining, B shows only the red and blue channels. (C) There is an apical CD133 (red) staining in a CK7 (green) marked ductule (arrowhead), while this marker is not present in an interlobular bile duct (arrow). The inset shows the apical CD133 positivity of ductules. (D) CD56 (red) stains only the small nerves (arrowheads) in the portal space, the CK7 (green) and laminin (blue) labeled bile duct is negative (arrow). (E) High magnification of a CK7 (green) and laminin (blue) stained ductule within the septum reveals membranous CD56 (red) positivity (arrowheads). (F) CD56+ (green) bile ductules (arrowheads) are attached to hepatocytes (not highlighted by this staining), are surrounded by laminin-positive (red) U-shaped basement membrane (small arrows). The laminin (red)-positive “empty” structure surrounded by basement membrane (large arrow) represents a blood vessel. Scale bar = 50 µm.

Quantitative evaluation of the immunohistochemical reactions (Table 1) showed that only a small portion of the biliary structures were stained for EMA, and this staining was restricted to the periportal area. No such preferential distribution was noticed with the two other markers. Almost all ducts/ductules were decorated by CD133 in the livers of the children, while in adulthood the ratio of CD56+ and CD133+ ductules were similar. The CK7 antibody reacted sometimes with very small bile ductules occasionally appearing as single cells especially along the vascular septa. Since the CD56 and CD133 reaction was not as strong and diffuse as the CK7, the number of the CD56 and CD133 ductules was probably underestimated.

Table 1. Immunophenotype of Biliary Structures in Normal Liver

Sample	CK7+	EMA+	CD56+	CD133+
3 years	100% (1,112)	4.8%	62.9%	96.3%
Adult	100% (1,125)	6.7%	54.4%	59%

Abbreviation: EMA, epithelial membrane antigen.

( ) Total number of counted bile ducts.

Discussion

We have analyzed the architecture of biliary ductules in normal human livers and observed them to circumscribe the classical hepatic lobules by participating in the formation of the so-called vascular septa (Fig. 4). This arrangement develops in postnatal life and can be most obviously seen in early childhood. The hepatic ductules are characterized by a unique EMA−/CD56+/CD133+ immunophenotype.

FIG. 4.

Schematic representation of hepatic lobule (brown), with bile ducts/ductules (green) and accompanying arterioles (red). Note that these structures extend only halfway into the porto-portal distances; however, they cover the whole “lateral” surface of the lobule. For simplicity, the portal vein branches are not shown.

There are several candidates for the liver stem/progenitor cell niche. Kuwahara et al. [10] proposed four structures to harbor such cells: the canals of Hering, intralobular bile ducts, periductal “null” mononuclear cells, and peribiliary hepatocytes. After all, stemness has been proposed to be not an entity but function [23] and—depending on the situation—different cell populations can behave as hepatic progenitor cells. Insofar, most evidence shows that the canals of Hering have the highest potential to behave as hepatic progenitor cells [1,2,6,13]. Therefore, the accurate architecture of these structures is a key issue to understand their behavior under normal and pathological conditions.

The canals of Hering were originally described [11] as short straight ducts at the limiting plate, which connect the bile canaliculi to the interlobular bile ducts. However, Theise et al. [16] demonstrated the extension of hepatic ductules through the limiting plate into the hepatic lobule. Our results confirm the presence of these structures deep in the hepatic parenchyma; moreover, a clear orientation of the ductules could be observed on our confocal images. The ductules spread into the parenchyma along the porto-portal axis. The hexagonal structures outlined by hepatic ductules correspond to the classical hepatic lobules. High power examination of individual ductules revealed their close correlation with bile canaliculi enabling their drainage.

In addition to former observations [19,24] that venular branches are present in vascular septa, we noticed NG-2-stained arterioles running in line with the ductules. There have been speculations about bile ductule escorting hepatic arterioles [18,19,25,26,27] but no convincing evidence has been published so far. Gouw et al. [19] and van der Heuvel et al. [28] emphasize the importance of the microvascular compartment for the efficient regeneration of ductules. Since—contrary to several other species—no arterio-portal anastomoses exist in the human liver, the presence of arterial blood would be advantageous for the regenerative function of the ductules. The proximity of the blood vessels to the bile ductules corresponds very well to the close correlation between the development of biliary and vascular structures [29]. This architecture of the intraparenchymal ductules and the escorting vessels is in full agreement with the proposed model of Matsumoto [18] based on 3D reconstruction of the human liver from thousands of serial sections.

According to the original description [18], the “vascular septum” is not a fibrous septum but a vascular surface from which sinusoids originate. However, we observed a collagenous matrix in this location. Since all the studied liver samples were normal and no fibrosis could be seen by traditional connective tissue staining, we suggest that a minimal amount of matrix material in the vascular septum, which can be visualized only by careful immunohistochemical analysis, is a component of the normal human hepatic tissue. Hepatic lobules are separated by well-defined connective tissue septa in several species [30], and the vascular septum of the human liver with its matrix components can be regarded as a rudimentary interlobular septum.

We have observed a peculiar age dependence of the vascular septa. Obviously, a more detailed age-related analysis of these structures is required. The interlobular bile ducts develop from the ductal plate [21], but ductal plates disappear shortly after birth and new bile ducts/ductules arise from pre-existent ducts by branching and elongation [31]. This “cholangiogenesis” could follow the primitive septa we observed in the liver of the 39-week-old fetus. The postnatal maturation of the biliary tree is well documented in humans [21] and rats [22] as well. Furthermore, the interlobular septum also develops postnatally in pigs [30].

The progressing scarcity of the ductular system with age should also be analyzed in more detail. We do not know if this process is absolute due to the apoptosis of biliary cells or just relative. The size of the hepatic lobules increases during ontogeny [32], and if the growth of the ductules is arrested earlier, it may be responsible for their relative regression.

We were curious if the biliary ductules could be characterized by a special immunophenotype. Three different antibodies were able to distinguish reliably the canals of Hering from larger bile ducts: the canals of Hering were EMA−/CD56+/CD133+; whereas, interlobular bile ducts were EMA+/CD56−/CD133−. All of these markers have already been mentioned in connection with the hepatic progenitor cell compartment. Atypical ductular reactions have been reported EMA−/CD56+, while the typical ones, which are similar to the interlobular bile ducts, are EMA+/CD56− [33]. CD56 has been demonstrated in proliferating ductules, while it could not be observed in normal canals of Hering [34,35], but recent studies found CD56 mRNA and protein in ductules of normal human livers [36,37]. CD133 has originally been described as a hematopoietic stem cell marker, and its mRNA has been detected in the liver by Northern hybridization, but no immunostaining was identified in paraffin sections by Miraglia et al. [38]. However, the protein could be detected by immunohistochemistry in the canals of Hering of normal human liver [39] and in regenerating ductules related to fulminant liver failure [40]. In our present experiments, the distribution of this marker showed an age-dependent change. This is similar to our results on rat liver, where the immunophenotype of the canals of Hering developed postnatally [15]. Interestingly, CD133+ cells isolated from HCC proved to be highly tumorigenic and have been reported as tumor stem cells [41]. Increased expression of CD133 has also been reported in a subset of cholangiocellular carcinomas, which were claimed to have a progenitor cell origin [42].

The combined application of these three antibodies provides an efficient tool for the identification of the canals of Hering in the normal human liver. Furthermore, the distinct immunophenotype of the hepatic ductules supports their different biological potential.

In conclusion, we present a refinement for the widely cited architectural description [16,27,43] of the intraparenchymal biliary ductules in normal human liver. The canals of Hering with escorting vessels are situated in the vascular septum and are components of a rudimentary interlobular septum. They can be distinguished from larger bile ducts by a unique immunophenotype. Better comprehension of canals of Hering’s architecture in normal human liver may promote our understanding of their behavior in various pathological/biological reactions.

Footnotes

Acknowledgments

The authors appreciate the help of Dr. Nóra Szlávik, Dr. Éva Görbe, and Dr. Júlia Hajdú in collecting the samples. The article is supported by OTKA K 67697.

Author Disclosure Statement

We declare that we have no duality of interest.

References

Alison

, Vig

, Russo

, Bigger

, Amofah

, Themis

and Forbes

. (2004). Hepatic stem cells: from inside and outside the liver?. Cell Prolif, 37:1–21.

Santoni-Rugiu

, Jelnes

, Thorgeirsson

and Bisgaard

. (2005). Progenitor cells in liver regeneration: molecular responses controlling their activation and expansion. APMIS , 113:876–902 .

Dahlke

, Popp

, Larsen

, Schlitt

and Rasko

JEJ

. (2004). Stem cell therapy of the liver-fusion or fiction? Liver Transpl 10: 471–479.

Tosh

and Strain

. (2005). Liver stem cells-prospects for clinical use. J Hepatol , 42 :, S75–S84 .

Thorgeirsson

and Grisham

. (2006). Hematopoietic cells as hepatocyte stem cells: a critical review of the evidence. Hepatology , 43 :, 2–8.

Fausto

. (2004). Liver regeneration and repair: hepatocytes, progenitor cells, and stem cells. Hepatology, 39:1477–1487.

Grompe

. (2003). The role of bone marrow stem cells in liver regeneration. Semin Liver Dis, 23:363–371.

Fujita

, Furukawa

, Hattori

, Todo

, Ishida

and Nagashima

. (2000). Sequential observation of liver cell regeneration after massive hepatic necrosis in auxiliary partial orthotopic liver transplantation. Mod Pathol, 13:152–157.

Eleazar

, Memeo

, Jhang

, Mansukhani

, Chin

, Park

, Lefkowitch

and Bhagat

. (2004). Progenitor cell expansion: an important source of hepatocyte regeneration in chronic hepatitis. J Hepatol, 41. 983–991.

10.

Kuwahara

, Kofman

, Landis

, Swenson

, Barendswaard

and Theise

. (2008). The hepatic stem cell niche: identification by label-retaining cell assay. Hepatology, 47:1994–2002.

11.

Hering

. (1867). Über den Bau der Wirbelthierleber. Arch Mikrosk Anat, 3:88–114.

12.

Grisham

and Porta

. (1964). Origin and fate of proliferated hepatic ductal cells in rat: electron microscopic and autoradiographic studies. Exp Mol Pathol, 3:242–261.

13.

Paku

, Schnur

, Nagy

and Thorgeirsson

. (2001). Origin and structural evolution of the early proliferating oval cells in rat liver. Am J Pathol, 158:1313–1323.

14.

Scheres

. (2007). Stem-cell niches: nursery rhymes across kingdoms. Nature Rev Mol Cell Biol, 8:345–354.

15.

Paku

, K

Dezsõ

, Kopper

and Nagy

. (2005). Immunohistochemical analysis of cytokeratin 7 expression in resting and proliferating biliary structures of rat liver. Hepatology, 42:863–870.

16.

Theise

, Saxena

, Portmann

, Thung

, Yee

, Chiriboga

, Kumar

and Crawford

. (1999). The canals of Hering and hepatic stem cells in humans. Hepatology, 30:1425–1433.

17.

. Fukushi

, Makagiansar

and Stallcup

. (2004). NG2 proteoglycan promotes endothelial cell motility and angiogenesis via engagement of Galectin-3 and α3β1 integrin. Mol Biol Cell, 15:3580–3590.

18.

Matsumoto

, Komori

, Magara

, Ui

, Kawakami

, Tokuda

, Takasaki

, Hayashi

, Jo

, Hano

and Tanaka

. (1979). A study on the normal structure of the human liver, with special reference to its angioarchitecture. Jikeikai Med J, 26:1–40.

19.

Gouw

, van der Heuvel

, Boot

, Slooff

, Poppema

and de Jong

. (2006). Dynamics of the vascular profile of the finer ranches of the biliary tree in normal ad diseased human livers. J Hepatol, 45:393–400.

20.

Crawford

. (2002). Development of the intrahepatic biliary tree. Semin Liver Dis, 22:213–226.

21.

Lemaigre

. (2003). Development of the biliary tract. Mech Dev, 120:81–87.

22.

Van

Eyken P

, Sciot

and Desmet

. (1988). Intrahepatic bile duct development in the rat. Lab Invest, 59:52–59.

23.

Blau

, Brazelton

and Weimann

. (2001). The evolving concept of a stem cell: entity or function?. Cell, 105:829–841.

24.

Teutsch

. (2005). The modular microarchitecture of human liver. Hepatology, 42:317–325.

25.

Yamamoto

, Sherman

, Phillips

and Fisher

. (1985). Three-dimensional observations of the hepatic arterial terminations in rat, hamster and human liver by scanning electron microscopy of microvascular casts. Hepatology, 5:452–456.

26.

. MacSween

RNM

, Desmet

, Roskams

and Scothorne

. (2002). Developmental anatomy and normal structure. In: Pathology of the Liver. MacSween

RNM

, Burt

, Portmann

, Ishak

, Scheuer

, Anthony

, eds. Churchill Livingstone, London, pp 1–66.

27.

Saxena

, Theise

and Crawford

. (1999). Microanatomy of the human liver-exploring the hidden interfaces. Hepatology, 30:1339–1346.

28.

Van

der Heuvel MC

, de Jong

, Boot

, Sloff

, Poppema

and Gouw

. (2006). Preservation of bile ductules mitigates bile duct loss. Am J Transplant, 6:2660–2671.

29.

Libbrecht

, Cassiman

and Desmet

. (2002). The correlation between portal myofibroblasts and development of intrahepatic bile ducts and arterial branches in human liver. Liver, 22:252–258.

30.

Ekataksin

and Wake

. (1997). New concepts in biliary and vascular anatomy of the liver. Prog Liver Dis, 15:1–30.

31.

Roskams

and Desmet

. (2008). Embryology of extra- and intrahepatic bile ducts, the ductal plate. Anat Rec, 291:628–635.

32.

. Papp

, Dezso˝

, László

, Nagy

and Paku

. (2009). Architectural changes during regenerative and ontogenic liver growth in the rat. Liver Transplant, 15:177–183.

33.

Sonzogni

, Colloredo

, Fabris

, Cadamuro

, Paris

, Roffi

, Pozzi

, Bovo

, Del Poggio

, Portmann

and Strazzabosco

. (2004). Isolated idiopathic bile ductular hyperplasia in patients with persistently abnormal liver function tests. J Hepatol, 40:592–598.

34.

Van

der Heuvel MC

, Slooff

MJH

, Visser

, Muller

, DeJong

, Poppema

and Gouw

. (2001). Expression of anti-OV6 antibody and anti-N-CAM antibody along the biliary line of normal and diseased human livers. Hepatology, 33:1387–1393.

35.

Zhou

, Rogler

, Teperman

, Morgan

and Rogler

. (2007). Identification of hepatocytic and bile ductular cell lineages and candidate stem cells in bipolar ductular reactions in cirrhotic human liver. Hepatology, 45:716–724.

36.

Schmelzer

, Wauthier

and Reid

. (2006). The phenotypes of pluripotent human hepatic progenitors. Stem Cells, 24:1852–1858.

37.

Zhang

, Theise

, Chua

and Reid

. (2008). The stem cell niche of human livers. Hepatology, 48:1598–1607.

38.

Miraglia

, Godfrey

and Yin

. (1997). A novel five-transmembrane hematopoietic stem cell antigen: isolation, characterization, and molecular cloning. Blood, 90:5013–5021.

39.

Karbanová

, Missol-Kolka

, Fonseca

, Lorra

, Janich

, Hollerová

, Jászai

, Ehrmann

, Kolar

, Liebers

, Arl

, Subrtová

, Freund

, Mokry

, Huttner

and Corbelli

. (2008). The stem cell marker CD 133 (Prominin-1) is expressed in various human glandular epithelia. J Histochem Cytochem, 56:977–993.

40.

Craig

CEH

, Quaglia

, Selden

, Lowdell

, Hodgson

and Dhillon

. (2004). The histopathology of regeneration in massive hepatic necrosis. Semin Liver Dis, 24:49–64.

41.

, Chan

, Hu

, Lee

, Wo

, Ng

, Zheng

and Guan

. (2007). Identification and characterization of tumorgenic liver cancer stem/progenitor cells. Gastroenterology, 132:2542–2556.

42.

Komuta

, Spee

, Borght

, De Vos

, Verslype

, Aerts

, Yano

, Suzuki

, Matsuda

, Fujii

, Desmet

, Kojiro

and Roskams

. (2008). Clinicopathological study on cholangiolocellular carcinoma suggesting hepatic progenitor cell origin. Hepatology, 47:1544–1556.

43.

Roskams

, Theise

, Balabaud

, Bhagat

, Bhathal

, Bioulac-Sage

, Brunt

, Crawford

, Crosby

, Desmet

, Finegold

, Geller

, Gouw

, Hytiroglou

, Knisely

, Kojiro

, Lefkowitch

, Nakanuma

, Olynyk

, Park

, Portmann

, Saxena

, Scheuer

, Strain

, Thung

, Wanless

and West

. (2004). Nomenclature of the finer branches of the biliary tree: canals, ductules, and ductular reactions in human livers. Hepatology, 39:1739–1745.