SEARCH

SEARCH BY CITATION

Keywords:

  • liver progenitor cell;
  • hepatic stellate cell;
  • myofibroblast;
  • sympathetic nervous system;
  • neuroendocrine

Abstract

  1. Top of page
  2. Abstract
  3. LIVER PROGENITOR CELLS
  4. ACTIVATION OF PROGENITOR CELLS
  5. NEUROENDOCRINE FEATURES OF HEPATIC PROGENITOR CELL COMPARTMENT
  6. EXTRACELLULAR MATRIX COMPONENTS INDUCE ENDOCRINE DIFFERENTIATION IN VITRO IN HUMAN LIVER CELL CULTURES THAT HAVE A BILE DUCT PHENOTYPE
  7. HEPATIC STELLATE CELLS AND MYOFIBROBLASTS: HSC ACTIVATION
  8. NEUROENDOCRINE FEATURES OF STELLATE CELLS/MYOFIBROBLASTS
  9. EFFECT OF PARASYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  10. EFFECT OF SYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  11. LITERATURE CITED

Liver progenitor cells as well as hepatic stellate cells have neuroendocrine features. Progenitor cells express chromogranin-A and neural cell adhesion molecule, parathyroid hormone-related peptide, S-100 protein, neurotrophins, and neurotrophin receptors, while hepatic stellate cells express synaptophysin, glial fibrillary acidic protein, neural cell adhesion molecule, nestin, neurotrophins, and their receptors. This phenotype suggests that these cell types form a neuroendocrine compartment of the liver, which could be under the control of the central nervous system. We recently showed that the parasympathetic nervous system promotes progenitor cell expansion after liver injury, since selective vagotomy reduces the number of progenitor cells after chemical injury in the rat. Similarly, after transplantation, which surgically denervates the liver, human livers that develop hepatitis have fewer progenitor cells than native, fully innervated livers with similar degrees of liver injury. There is also accumulating experimental evidence linking the autonomic system, in particular the sympathetic nervous system (SNS), with the pathogenesis of cirrhosis and its complications. Recently, it has been shown that hepatic stellate cells themselves respond to neurotransmitters. Moreover, inhibition of the SNS reduced fibrosis in carbon tetrachloride-induced liver injury. In view of the denervated state of transplanted livers, it is very important to unravel the neural control mechanisms of regeneration and fibrogenesis. Moreover, since there is a shortage of donor organs, a better understanding of the mechanisms of regeneration could have therapeutic possibilities, which could even obviate the need for orthotopic liver transplantation. © 2004 Wiley-Liss, Inc.

The liver consists of two mature epithelial cell compartments, hepatocytes and bile duct epithelial cells, and also harbors an epithelial progenitor cell compartment (Roskams et al., 2003a). Progenitor cells are activated when the mature cell compartments are damaged or inhibited in their replication by, e.g., toxic agents. Progenitor cells can differentiate into hepatocytes and bile duct epithelial cells and hence contribute to the regeneration of these cell compartments in diseased liver (Roskams et al., 2003a). Hepatic stellate cells (HSCs) after activation differentiate into myofibroblasts and play an important role in repair and cicatrization (fibrogenesis, development of cirrosis) (Reeves and Friedman, 2002; Safadi and Friedman, 2002). We have previously shown that the human hepatic progenitor cell compartment displays neural/neuroendocrine features such as chromogranin-A, neural cell adhesion molecule, parathyroid hormone-related peptide, S-100 protein, and neurotrophins and neurotrophin receptors (Sciot et al., 1986; Roskams et al., 1991, 1993a, 1993b; Cassiman et al., 2001a, 2001b; Pirenne et al., 2001). In addition to that, we have extended the list of neural/neuroendocrine features expressed by HSCs (Reynaert et al., 2000). Activation and proliferation of HSCs and hepatic progenitor cells are often observed in close anatomical and temporal relationship (Dabeva and Shafritz, 1993; Miyazaki et al., 1993; Yin et al., 1999, 2002). HSCs are proposed to interact with hepatic progenitor cells, for instance by secreting hepatocyte growth factor (HGF), stem cell factor (SCF), and brain-derived nerve growth factor (BDNF), while the receptors for these growth factors are expressed on hepatic progenitor cells (Hu et al., 1993; Fujio et al., 1994). HSCs have been shown to have close contacts with nerve endings in normal liver (Akiyoshi, 1989; Ueno et al., 1990; Akiyoshi and Terada, 1998). In cirrhotic liver, the innervation of the parenchyma is decreased and the contacts between nerve endings and activated intranodular HSCs are less numerous (Akiyoshi, 1989; Ueno et al., 1990; Akiyoshi and Terada, 1998). A clear effect of the hepatic innervation on regeneration has been shown previously (Ohtake et al., 1993; Kiba et al., 1994, 1995; LeSage et al., 1999; Cassiman et al., 2002b; Oben et al., 2003a). All these facts lead us to hypothesize that progenitor cells and HSCs/myofibroblasts form a neural/neuroendocrine compartment in the liver. In this neural/neuroendocrine compartment, HSCs and hepatic progenitor cells might be considered the effector cells of repair and regeneration, working under control of the central nervous system. Conversely, HSCs and hepatic progenitor cells may also be necessary to create the right environment for reinnervation of diseased or renewed tissue. Understanding the control mechanisms of the central nervous system over the neural/neuroendocrine compartment of the liver is important, since after liver transplantation, the transplanted liver is denervated (the nerves are dissected) and as such has no connection with the nervous system of the recipient. Since progenitor cells and hepatic stellate cells/myofibroblasts play an important role in regeneration and cicatrization, the transplanted liver might have a disturbed regeneration potential compared to the nontransplanted liver.

This review describes the liver progenitor cell compartment and hepatic stellate cells/myofibroblasts and their neuroendocrine features. In addition, the current knowledge on the neural control mechanisms of progenitor cells and stellate cells/myofibroblasts is discussed.

LIVER PROGENITOR CELLS

  1. Top of page
  2. Abstract
  3. LIVER PROGENITOR CELLS
  4. ACTIVATION OF PROGENITOR CELLS
  5. NEUROENDOCRINE FEATURES OF HEPATIC PROGENITOR CELL COMPARTMENT
  6. EXTRACELLULAR MATRIX COMPONENTS INDUCE ENDOCRINE DIFFERENTIATION IN VITRO IN HUMAN LIVER CELL CULTURES THAT HAVE A BILE DUCT PHENOTYPE
  7. HEPATIC STELLATE CELLS AND MYOFIBROBLASTS: HSC ACTIVATION
  8. NEUROENDOCRINE FEATURES OF STELLATE CELLS/MYOFIBROBLASTS
  9. EFFECT OF PARASYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  10. EFFECT OF SYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  11. LITERATURE CITED

Support for the liver progenitor cell hypothesis has come mainly from cell culture data (Fausto, 1994) and from rodent models of chemical hepatocarcinogenesis (Opie, 1944; Farber, 1956; Grisham and Hartroft, 1961; Fausto, 1990; Sell, 1990) and liver cell regeneration after chemical injury (Wilson and Leduc, 1958; Lemire et al., 1991). In these models, a periportal population of small primitive epithelial cells proliferates in association with or before hepatocyte multiplication. These cells were called oval cells because of their shape (Opie, 1944; Farber, 1956), and this name is widely used for hepatic progenitor cells in animal liver. An extensive review of the currently available data on oval cells is beyond the scope of this review. Oval cells are related to terminal biliary ductules and the so-called canals of Hering, which represent the terminal branches of the biliary tree that connect the interhepatocytic bile canaliculi with the bile ducts in the portal tract (Grisham, 1980; Germain et al., 1988; Lemire et al., 1991; Lenzi et al., 1992; Paku et al., 2001). Oval cells express phenotypic markers of both (immature) hepatocytes (such as α-fetoprotein) and bile duct cells (such as bile duct-type cytokeratins) (Germain et al., 1985; Hixson and Allison, 1985; Dunsford et al., 1989; Dabeva et al., 1993). They constitute a heterogeneous cell population (Germain et al., 1985; Hixson and Allison, 1985; Dunsford and Sell, 1989; Dabeva and Shafritz, 1993; Dabeva et al., 1993; Mandache et al., 2002), but at least a subset of oval cells is pluripotent and has the capacity to differentiate toward hepatocytes, bile ductular cells, and intestinal epithelium and can give rise to hepatocellular carcinoma and cholangiocellular carcinoma (Grisham, 1980; Evarts, 1987, 1996; Tsao, 1987; Sell, 1990; Yasui, 1997; Alison, 2001). A new light was shed on the liver progenitor cell field with studies suggesting that bone marrow-derived stem cells can give rise to hepatocytes, oval cells, and bile duct cells (Petersen et al., 1999; Alison et al., 2000; Theise et al., 2000a; Korbling et al., 2002). Similar mechanisms were described in human liver (Theise et al., 2000b). At least in certain animal models, liver repair by bone marrow-derived cells takes place by fusion (Vassilopoulos et al., 2003; Wang et al., 2003). This new evolution in the field needs further study, since fusion of cells could induce genome instability and loss of chromosomes potentiating malignant transformation (Holden, 2003).

Although a heresy for a long time, it is now generally accepted that in human liver also, progenitor cells (the equivalents of rat oval cells) exist and are activated in different liver diseases (Haque et al., 1996; Roskams et al., 1998, 2003a; Sell, 1998; Theise et al., 1999; Libbrecht and Roskams, 2002). Human liver progenitor cells have mainly been studied in regeneration after severe hepatocellular necrosis (Gerber, 1983; Roskams et al., 1991, 1998; Haque et al., 1996; Lowes et al., 1999; Fujita, 2000), but recent studies show that this cell compartment is also activated in chronic viral hepatitis (Sakamoto et al., 1975; Libbrecht et al., 2000; Xiao et al., 2003), alcoholic liver disease (Ray, 1993; Roskams, 2003), and nonalcoholic fatty liver disease (Roskams, 2003), the most frequent chronic liver diseases in the Western world.

The only possible therapy for end-stage chronic liver diseases is liver transplantation. The shortage of donor organs necessitates the search for alternative treatment strategies. Cell therapy could be a possible alternative in the future. Because of their bipotential differentiation abilities, progenitor cells are ideal candidates for cell therapy. Therefore, studies on their activation and differentiation mechanisms are of high priority.

ACTIVATION OF PROGENITOR CELLS

  1. Top of page
  2. Abstract
  3. LIVER PROGENITOR CELLS
  4. ACTIVATION OF PROGENITOR CELLS
  5. NEUROENDOCRINE FEATURES OF HEPATIC PROGENITOR CELL COMPARTMENT
  6. EXTRACELLULAR MATRIX COMPONENTS INDUCE ENDOCRINE DIFFERENTIATION IN VITRO IN HUMAN LIVER CELL CULTURES THAT HAVE A BILE DUCT PHENOTYPE
  7. HEPATIC STELLATE CELLS AND MYOFIBROBLASTS: HSC ACTIVATION
  8. NEUROENDOCRINE FEATURES OF STELLATE CELLS/MYOFIBROBLASTS
  9. EFFECT OF PARASYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  10. EFFECT OF SYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  11. LITERATURE CITED

Activation of hepatic progenitor cells (oval cells in rodents) is a term that is used for an increase in the number of progenitor cells and their differentiation toward the hepatocytic and/or biliary lineage (Dabeva et al., 1993; Alison, 1998; Alison and Sarraf, 1998). In human liver, differentiation toward the biliary lineage leads to formation of reactive ductules, while differentiation toward the hepatocytic lineage occurs via intermediate hepatobiliary cells. Reactive ductules are anastomosing strands of immature biliary cells with an oval nucleus and a small rim of cytoplasm, located at the mesenchymal (portal/septal)-parenchymal interface (Roskams et al., 1996a, 1998; Roskams and Desmet, 1998; Theise et al., 1999; Libbrecht et al., 2001). Intermediate hepatobiliary cells are polygonal cells with a size and phenotype intermediate between progenitor cells and hepatocytes (Roskams et al., 1991, 1998; Demetris et al., 1996; Theise et al., 1999). Different subtypes of progenitor cells have also been recognized by electron microscopy: the most immature progenitor cell type besides cells already featuring some biliary or hepatocytic features (De Vos and Desmet, 1992; Xiao et al., 1999, 2003; Mandache et al., 2002). This spectrum of cells ranging from the most immature phenotype to ductules and intermediate hepatocytes forms a cell compartment with a specific phenotype and is often referred to as the progenitor cell compartment.

From animal models, we have learned that progenitor cells (oval cells) are activated when damage or loss of hepatocytes and/or cholangiocytes is combined with impaired regeneration of the mature cell types involved (Alison, 1998). Virtually every acute and chronic human liver disease is associated with damage to and loss of hepatocytes and/or cholangiocytes, the two major epithelial cell compartments of the liver. Injury can be caused by numerous triggers, including viruses, alcohol, toxic substances, metabolic errors, and unknown factors. Hence, it is not surprising that hepatic progenitor cell activation has been described in a variety of human liver diseases (Libbrecht and Roskams, 2002; Roskams et al., 2003a).

NEUROENDOCRINE FEATURES OF HEPATIC PROGENITOR CELL COMPARTMENT

  1. Top of page
  2. Abstract
  3. LIVER PROGENITOR CELLS
  4. ACTIVATION OF PROGENITOR CELLS
  5. NEUROENDOCRINE FEATURES OF HEPATIC PROGENITOR CELL COMPARTMENT
  6. EXTRACELLULAR MATRIX COMPONENTS INDUCE ENDOCRINE DIFFERENTIATION IN VITRO IN HUMAN LIVER CELL CULTURES THAT HAVE A BILE DUCT PHENOTYPE
  7. HEPATIC STELLATE CELLS AND MYOFIBROBLASTS: HSC ACTIVATION
  8. NEUROENDOCRINE FEATURES OF STELLATE CELLS/MYOFIBROBLASTS
  9. EFFECT OF PARASYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  10. EFFECT OF SYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  11. LITERATURE CITED

We have previously shown that the hepatic progenitor cell compartment has neuroendocrine features such as expression of chromogranin-A, neural cell adhesion molecule, parathyroid hormone-related peptide, and neuron-specific enolase (Fig. 1) (Fukuda et al., 1989; Roskams et al., 1990, 1991, 1993a; Roskams and Desmet, 1997).

thumbnail image

Figure 1. a: Chromogranin-A staining in a liver biopsy of a patient with primary biliary cirrhosis. A large network of reactive ductules is present. b-c: Electron micrograph showing neuroendocrine granules (arrows) in the cytoplasm of hepatocytes (Fig. 1b) and epithelial cells with features of both hepatocytes and progenitor cells (Fig. 1c).

Download figure to PowerPoint

Chromogranin-A is a molecule present in the matrix of neuroendocrine granules where it is thought to play a role in the packaging and processing of hormones and neuropeptides and in stabilizing the granule content. Chromogranin-A may have extracellular roles as well, since the intact protein or proteolytic fragments derived thereof exert biological effects on target cells, e.g., as a prohormone of pancreastatin (Simon and Aunis, 1989). In contrast to other molecules, chromogranin-A has been found to be a highly reliable marker for neuroendocrine cells (Simon and Aunis, 1989; Wiedenmann and Huttner, 1989).

The neuroendocrine character of the progenitor cell compartment was furthermore supported by their reactivity for N-CAM (Roskams et al., 1991, 1996a, 1998). Differential expression of NCAM on either a temporal or a topographic basis is of fundamental importance to neuro-ontogenesis. Although the cell distribution of N-CAM is widespread in early embryonic life, it shows a more restricted distribution in adult tissues, as it is limited mainly to neural cells and endocrine cells (Crossin et al., 1985; Murray et al., 1986; Langley et al., 1989). N-CAM is a cell surface adhesion molecule that is known to play an important regulatory role in development through cell-cell and cell-matrix interactions. It works through homotypic binding of two N-CAM molecules. Therefore, N-CAM could play an important role in modulating the extent of progenitor cells/reactive ductules in the surrounding matrix. This could be through cell-cell interaction, via binding with N-CAM on hepatic stellate cells, which have been shown to express N-CAM. On the other hand, N-CAM has been shown to have binding sites for extracellular matrix components such as heparan sulfate (Werz and Schachner, 1988), which is present around the progenitor cell compartment; we showed strong reactivity for heparan sulfate proteoglycan perlecan around activated progenitor cells and reactive ductules in different diseases (Roskams et al., 1996b).

Neuron-specific enolase (NSE) is a glycolytic enzyme found in neural and neuroendocrine cells. It is used as a broad-range marker that reacts with most neuroendocrine cells and neoplasms. Reactive bile ductules have been shown to express NSE (Fukuda et al., 1989)

The immunohistochemical evidence for a neuroendocrine phenotype was confirmed by the electron microscopic demonstration of dense-cored secretory granules in the progenitor cell compartment (Fig. 1b and c). Previous electron microscopic studies focused mainly on the effect of biliary constituents on the ultrastructural appearance of reactive bile ductules but have failed to identify dense-cored granules in these cells. This may be due to the relative scarcity of these granules. The existence of a well-developed peribiliary capillary plexus led already to the hypothesis that biliary epithelial cells subserve an endocrine function and that these cells might release biologically active peptides in the surrounding liver tissue or in the general circulation (Ohtani, 1979).

In addition to progenitor cells/reactive bile ductules, clusters of periportal hepatocytes express diffuse cytoplasmic positivity for chromogranin-A, OV-6, and cytokeratin 7 (biliary cytokeratin, also marking progenitor cells). These hepatocytes are smaller than normal hepatocytes and seemed in continuity with bile ductular structures. They lack N-CAM reactivity. These intermediate hepatobiliary cells are seen in livers with long-standing cholestasis, in regenerating livers after submassive necrosis, in chronic hepatitis, and in alcoholic and nonalcoholic steatohepatitis (Roskams et al., 1990, 1991, 1998, 2003a, 2003b; Crosby et al., 1998; Roskams and Desmet, 1998; Lowes et al., 1999). Electron microscopy confirmed the presence of neuroendocrine granules in the cytoplasm of hepatocytes (Fig. 1b). Moreover, epithelial cells presenting features of both hepatocytes and progenitor cells/bile ductular cells in variable proportions were found in the parenchyma. Hepatocyte characteristics included glycogen rosettes, abundant mitochondria with short cristae, and many endoplasmatic reticulum cisternae; progenitor cell/bile duct cell characteristics included interdigitations of their adjoining lateral membranes, the presence of basement membranes, and many thick bundles of tonofilaments. These cells contained one or more small dense-cored granules, localized preferentially near the peripheral membrane (Fig. 1b and c). The presence of cells intermediate between hepatocytes and progenitor cells/bile duct cells supports the concept of differentiation of progenitor cells into hepatocytes. We found chromogranin-A-positive hepatocytes never without the presence of immunoreactive bile ductules/progenitor cells. These results hint at a (bile ductule-related) progenitor cell origin of the intermediate cells and small hepatocytes (Roskams et al., 1990, 1991, 1996a, 1998, 2003a).

The presence of cells with neuroendocrine features in regenerating and cholestatic liver tissue suggested that the substance(s) in the dense-cored granules might play a role in the growth and/or differentiation of liver epithelial cells through an autocrine and/or paracrine pathway. In search of the substance(s) presumed to be present in the granules, initially we performed immunohistochemistry for the following hormones and growth factors: somatostatin, gastrin, vasoactive intestinal peptide, protein gene product, glucagon, secretin, growth hormone, pancreatic polypeptide, motilin, metenkephalin, neurotensin, and parathyroid hormone. All these immunohistochemical stainings were negative in the progenitor cell compartment, at least with the techniques available at that time (the 1990s).

Parathyroid hormone-related peptide (PTHrP), a peptide that has been purified and sequenced in 1987 (Suva et al., 1987), is the major factor responsible for humoral hypercalcemia of malignancy (HHM). PTHrP has been demonstrated in neoplasms associated with HHM, but also in tumors not associated with HHM (Rabbani et al., 1986; Ikeda et al., 1988; Deftos et al., 1989; Drucker et al., 1989; Asa et al., 1990). The detection of PTHrP mRNA transcripts and peptide in many tissues (Roskams and Desmet, 1997) raised the possibility that PTHrP may play a role in the biology of nonneoplastic cells, different from its role in the regulation of calcium balance. Current evidence supports an important role for PTHrP in growth and differentiation of neoplastic as well as nonneoplastic cells (Roskams and Desmet, 1997). Moreover, the gene encoding PTHrP shares several features with members of the immediate early gene family. In addition, PTHrP was detected in fetal but not in normal adult rat liver. This prompted us to examine the expression of PTHrP in conditions associated with progenitor cell activation and to compare its distribution with the results obtained with neuroendocrine markers.

We showed that reactive bile ductules, but not portal bile ducts of all sizes, stained positive for chromogranin-A and neural cell adhesion molecule, providing further evidence for their neuroendocrine phenotype. In liver biopsies characterized by intermittent or partial obstruction, regeneration after submassive necrosis, as well as in primary biliary cirrhosis and primary sclerosing cholangitis, the majority of reactive bile ductular cells expressed PTHrP. In focal nodular hyperplasia, smaller numbers of ductular cells expressed PTHrP and unstained cells were found among immunoreactive cells. Bile ductular cells and hepatocytes in normal human liver tissue showed no immunoreactivity for PTHrP or for neuroendocrine markers (Roskams et al., 1993a).

Our finding that PTHrP is present in reactive human bile ductules/progenitor cell compartment that have a neuroendocrine phenotype suggested that PTHrP may play a role in the early cellular response of progenitor cells to growth factor stimulation in an autocrine and/or paracrine manner. Since PTHrP stimulates DNA synthesis in a variety of cells only in the presence of epidermal growth factor (EGF) or transforming growth factor-β (TGF-β), PTHrP has been suggested to contribute to a multifactorial growth factor loop. Also in the liver, PTHrP may be part of such a multifactorial growth factor loop. We suggested that PTHrP may act as a possible autocrine growth factor for progenitor cells. PTHrP induces EGF-dependent transformation of fibroblast cell lines with concomitant significant increase in the production of fibronectin, a property it shares with TGF-β. Therefore, it is possible that PTHrP plays also a role in the fibrogenesis that accompanies ductular reaction by acting in an paracrine way on matrix-producing cells (Roskams and Desmet, 1997).

EXTRACELLULAR MATRIX COMPONENTS INDUCE ENDOCRINE DIFFERENTIATION IN VITRO IN HUMAN LIVER CELL CULTURES THAT HAVE A BILE DUCT PHENOTYPE

  1. Top of page
  2. Abstract
  3. LIVER PROGENITOR CELLS
  4. ACTIVATION OF PROGENITOR CELLS
  5. NEUROENDOCRINE FEATURES OF HEPATIC PROGENITOR CELL COMPARTMENT
  6. EXTRACELLULAR MATRIX COMPONENTS INDUCE ENDOCRINE DIFFERENTIATION IN VITRO IN HUMAN LIVER CELL CULTURES THAT HAVE A BILE DUCT PHENOTYPE
  7. HEPATIC STELLATE CELLS AND MYOFIBROBLASTS: HSC ACTIVATION
  8. NEUROENDOCRINE FEATURES OF STELLATE CELLS/MYOFIBROBLASTS
  9. EFFECT OF PARASYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  10. EFFECT OF SYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  11. LITERATURE CITED

Which factors induce the neuroendocrine phenotype of the progenitor cell compartment? An in vitro study of de Bruine et al. (1993) showed that the endocrine phenotype of cells in culture (a colon tumor cell line with endocrine differentiation) could be induced by adding basic fibroblast growth factor (FGF) to the culture medium. A similar effect was obtained by seeding them on plates coated with both collagen type IV and matrix heparan sulfate proteoglycan (HSPG) perlecan, while each of these matrix components separately did not have an inducing effect. Interestingly, collagen type IV and perlecan are basement membrane components and reactive bile ductules are always surrounded by a basement membrane (Abdel-Aziz et al., 1990).

Therefore, we tested the hypothesis that collagen type IV and basement membrane-type HSPG perlecan could induce the neuroendocrine phenotype of bile ductular epithelial cells. HepG2 cells (cell line with a hepatocyte phenotype) and A16 and H1 cells (cell lines with a bile duct phenotype) were cultured on basement membrane matrigel, a medium that contains these components, and on minimal essential medium.

As expected, A16 and H1 cells were clearly immunoreactive for chromogranin-A when cultured on basement membrane matrigel, while HepG2 cells remained negative for this neuroendocrine marker. When cultured on minimal essential medium, all cell lines remained unreactive for chromogranin-A (data not shown).

This in vitro study indicates that interaction of bile duct epithelial cells with the surrounding matrix and in particular with collagen type IV and perlecan has important implications on their phenotype. In a recent in vitro study, Shiojiri and Mizuno (1993) used primary organ cultures of fetal mouse liver fragments without evidence of bile ducts to demonstrate that the addition of dexamethasone to the medium stimulated the development of both mature hepatocytes and bile duct epithelium. When the immature liver fragments were cultured on substrata such as collagen gel, Millipore filter, or Spongel, the presence of dexamethasone dramatically stimulated glycogen storage, which is a feature of hepatocytes, but did not induce bile duct differentiation. On the other hand, when the fragments were cultured on basement membrane matrigel, dexamethasone stimulated the expression of bile duct markers. This study confirms our results that basement membrane components are important for the phenotype of bile duct epithelial cells.

Since it was unknown whether matrix HSPG perlecan is present in human liver and whether its expression is enhanced in progenitor cell activation, we studied the expression of perlecan and integral membrane HSPGs in both normal and diseased human liver. We showed expression of perlecan in normal liver around the bile ducts, around progenitor cells, and also in the space of Disse. We confirmed this expression by immunoelectron microscopy (Roskams et al., 1995, 1996b). The in vivo expression fits very nicely with our in vitro experiments.

In multicellular organisms, the maintenance of tissue architecture, cellular polarization, and processes such as embryogenesis, proliferation, differentiation, inflammation, and wound healing are dependent on interactions of the cells with the components of the extracellular matrix. Extracellular matrix molecules interact with cell surface receptors that signal cells to change behavior. In many instances, signaling by these highly specific protein receptors requires the assistance of a complex carbohydrate, principally heparan sulfate (HS). This HS exists as proteoglycan (PG), as it is covalently bound to a variety of core proteins. These HSPGs are ubiquitous and are found inside cells, principally within storage vesicles of various secretory cells and possibly in the nucleus, at the cell surface as part of integral membrane PGs, and extracellularly in the pericellular matrix and basement membrane. According to the structure of the core proteins and their location, different families of HSPGs have been recognized: HSPGs of the extracellular matrix/basement membrane, related to the high-molecular-weight proteoglycan perlecan, and two families of integral membrane HSPGs. The first family of membrane HSPGs include syndecan-like proteoglycans, with core proteins that span the membrane and that share sequence motifs in highly conserved cytoplasmic domains. The second is made up by glypican-related integral membrane proteoglycans that are linked to the cell surface via glycosyl phosphatidyl inositol (David and Bernfield, 1998).

The biological roles of HSPGs are highly diversified, ranging from relatively simple mechanical support functions to more intricate effects on various cellular processes such as cell adhesion, motility, proliferation, differentiation, and tissue morphogenesis. Most of these effects mainly depend on binding of ligands to the HS chains (David and Bernfield, 1998).

Although the ubiquitous presence of HSPGs has long been apparent, recent interest in these molecules stems from increasing awareness of the functional implications of their interactions. This applies especially to their function as essential cofactors in receptor-growth factor interactions (e.g., as coreceptor for acidic and basic fibroblast growth factor), in cell-cell recognition systems, and in cell-matrix adhesion processes (e.g., by binding different collagen types, fibronectin, tenascin) (David and Bernfield, 1998; De Cat and David, 2001).

HEPATIC STELLATE CELLS AND MYOFIBROBLASTS: HSC ACTIVATION

  1. Top of page
  2. Abstract
  3. LIVER PROGENITOR CELLS
  4. ACTIVATION OF PROGENITOR CELLS
  5. NEUROENDOCRINE FEATURES OF HEPATIC PROGENITOR CELL COMPARTMENT
  6. EXTRACELLULAR MATRIX COMPONENTS INDUCE ENDOCRINE DIFFERENTIATION IN VITRO IN HUMAN LIVER CELL CULTURES THAT HAVE A BILE DUCT PHENOTYPE
  7. HEPATIC STELLATE CELLS AND MYOFIBROBLASTS: HSC ACTIVATION
  8. NEUROENDOCRINE FEATURES OF STELLATE CELLS/MYOFIBROBLASTS
  9. EFFECT OF PARASYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  10. EFFECT OF SYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  11. LITERATURE CITED

HSCs are described to activate in all pathological conditions affecting the liver (hepatitis, fibrosis, cirrhosis). HSC activation is characterized by the loss of lipid droplets and the acquisition of α-smooth muscle actin (αSMA) expression (Hautekeete and Geerts, 1997). Based on their ultrastructural characteristics and their expression of vimentin, desmin (in rats), and αSMA, activated HSCs are referred to as myofibroblast-like cells or myofibroblasts (MFs). HSC activation is accompanied by a major switch in expression profile, which is amply documented at the level of intermediate filament expression, extracellular matrix production, and degradation. Activated human and rat HSCs express αSMA (Hautekeete and Geerts, 1997), activated human HSCs express GFAP (Levy et al., 1999), activated rat HSCs express N-CAM (Fig. 2a) (Knittel et al., 1996) and intermediate filament nestin (Niki et al., 1999).

thumbnail image

Figure 2. a: Reactive ductules at the edge of a portal tract (PT) and hepatic stellate cell in the parenchyma (P) of a cirrhotic liver are positive for NCAM. b-c: Expression of synaptophysin (Fig. 2b) and neurotrophin 3 (Fig. 2c) by hepatic stellate cells in a cirrhotic nodule. d: GFAP-positive myofibroblasts at the interface between a septum and the parenchyma in the liver of a rat with carbon-tetrachloride-induced cirrhosis. e: Electron micrograph of normal human liver biopsy showing a stellate cell in the Disse space, flanked by part of two hepatocytes (H). The stellate cell contains fat droplets. Inset: Detail of a cytoplasmic processus containing many small clear vesicles, comparable to the synaptic vesicles found in neurons.

Download figure to PowerPoint

The expression of αSMA in activated HSCs is held responsible for the contractile properties of these cells. This contractility, together with the unique localization of HSCs around the sinusoids, is currently held responsible for the active component of portal hypertension (Rockey and Weisiger, 1996).

HSCs are known to proliferate and migrate toward areas of necrosis and areas of regeneration in varying pathological conditions (Johnson et al., 1992; Dabeva and Shafritz, 1993). In acute hepatitis with activation of the stem cell compartment and in bile ductular reaction, coproliferation of HSCs with hepatic progenitor cells has been described (Dabeva et al., 1993; Miyazaki et al., 1993). Interactions between HSC and stem cells have been suggested. For instance, HSCs produce SCFs, while oval cells carry the c-kit SCF receptor; HSCs also produce HGFs, while oval cells carry the c-met HGF receptor (Hu et al., 1993; Fujio et al., 1994).

NEUROENDOCRINE FEATURES OF STELLATE CELLS/MYOFIBROBLASTS

  1. Top of page
  2. Abstract
  3. LIVER PROGENITOR CELLS
  4. ACTIVATION OF PROGENITOR CELLS
  5. NEUROENDOCRINE FEATURES OF HEPATIC PROGENITOR CELL COMPARTMENT
  6. EXTRACELLULAR MATRIX COMPONENTS INDUCE ENDOCRINE DIFFERENTIATION IN VITRO IN HUMAN LIVER CELL CULTURES THAT HAVE A BILE DUCT PHENOTYPE
  7. HEPATIC STELLATE CELLS AND MYOFIBROBLASTS: HSC ACTIVATION
  8. NEUROENDOCRINE FEATURES OF STELLATE CELLS/MYOFIBROBLASTS
  9. EFFECT OF PARASYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  10. EFFECT OF SYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  11. LITERATURE CITED

Human and Rat Hepatic Stellate Cells Express Synaptophysin

Synaptophysin (SYN) is a major transmembrane glycoprotein of small (30–80 nm) electron-translucent (SET) vesicles. This class of vesicles includes the presynaptic vesicles in neuronal cells and the somewhat larger synaptic-like microvesicles (SLMVs) in neuroendocrine cells (Edelmann et al., 1995). The SYN protein is implicated in the control of exocytosis (Edelmann et al., 1995) and neurotransmitter release, e.g., in the neuromuscular synapse (Alder et al., 1992, 1995). Immunohistochemical detection of SYN is commonly used in conjunction with chromogranin-A, neuron-specific enolase, and Leu-7 immunohistochemistry to determine neuroendocrine origin of or differentiation in tissues and tumors throughout the body (Schurmann et al., 1990; Poola and Graziano, 1998). To the best of our knowledge, only neural and neuroendocrine cell types have been demonstrated to express SYN (Bargou and Leube, 1991), with the exception of rabbit thrombocytes (Bahler et al., 1990).

Since SYN staining is a well-established diagnostic tool in clinical practice and the presence of SYN protein is clearly linked to neural/neuroendocrine differentiation, we decided to check the expression profile of SYN in the liver (Cassiman et al., 1999).

We showed that quiescent as well as activated human and rat HSCs express synaptophysin (Fig. 2b) (Cassiman et al., 1999). The SYN protein is known to be present in membranes of neuronal synaptic vesicles containing neurotransmitters. Functionally, SYN is most probably involved in membrane fusion leading to exocytosis (Edelmann et al., 1995). Ultrastructurally, synaptic vesicles were shown to be present in both quiescent and activated stellate cells (Fig. 2e).

Based on the expression of vimentin, desmin, and αSMA (Yokoi et al., 1984; Burt et al., 1986; Ballardini et al., 1988; Takase et al., 1988; Schmitt-Gräff et al., 1991), HSCs have been considered to be of mesenchymal origin. From Kupffer's first description of Sternzellen (HSC) in liver, using gold chloride staining in search of nerve fibers onward (Kupffer, 1876), indications of neural/neuroendocrine differentiation of HSCs are accumulating. HSCs show GFAP reactivity (Neubauer et al., 1996; Niki et al., 1996, 1999), N-CAM expression (Knittel et al., 1996), nestin reactivity (Niki et al., 1999), and SYN expression (Cassiman et al., 1999). All these can be considered as arguments in favor of a neural/neuroendocrine differentiation or origin. Neural tissue originates from the ectodermal layer. Neuroendocrine tissue is thought to originate from either the neural crest, migrating out of the ectodermal layer, or from the endodermal layer. Expression of so-called mesenchymal features (vimentin, desmin, αSMA) is not contradictory to the suggested neural/neuroendocrine origin, since for example in the aorticopulmonary septum, smooth muscle cells supposedly originating from the neural crest also express αSMA (Beall and Rosenquist, 1990) and desmin (Sumida et al., 1987). In view of the evidence mentioned above, unavoidable questions about the mesenchymal origin and functional differentiation of HSCs rise. As a consequence, new studies focusing on the expression of neural and neuroendocrine markers in embryonic as well as adult liver become necessary.

Human and Rat Hepatic Stellate Cells Express Neurotrophins and Their Receptors

The family of neurotrophins (NTs) consists of nerve growth factor (NGF), BDNF, neurotrophin 3 (NT-3), and neurotrophin 4/5 (NT-4/5). Their known receptors (NTr) are the high-affinity tyrosine kinases Trk-A, Trk-B, and Trk-C and the low-affinity nerve growth factor receptor p75. The NTs are known to bind to the Trks in a specific manner, NGF binding to the Trk-A receptor, BDNF to Trk-B, NT-3 to Trk-C and, to a lesser degree, Trk-B, and NT-4/5 to Trk-B. All NTs bind to p75, a member of the tumor necrosis factor receptor family, which lacks an intracellular signal transduction domain. P75 modulates survival and death decisions (Chao et al., 1998; Casaccia-Bonnefil et al., 1999).

Originally, research on NTs focused on their role in the development and pathology of the central and peripheral nervous system (Barde, 1989). The expression of NTs by organs or systems other than the central or peripheral nervous system was considered to be aimed at stimulation of outgrowth and maintenance of the peripheral nervous system (Shalizi et al., 2003).

On the other hand, a rapidly growing body of literature is suggesting a more fundamental and more widespread role for the different NTs in varying nonneural organ systems in normal and pathological conditions. For instance, stromal smooth muscle cells in human prostate express NGF, BDNF, and Trk-C, while the prostate epithelial tumor cell line LNCaP expresses Trk-A, Trk-B, and Trk-C. These findings were considered indicative of a local paracrine interaction between stromal smooth muscle cells and tumor cells (Dalal and Djakiew, 1997). Secondly, rat vascular smooth muscle cells express NGF, BDNF, NT-3, and NT-4/5 as well as Trk-A, Trk-B, and Trk-C. A dramatic increase in NGF, BDNF, Trk-B, and Trk-C expression in areas of vascular injury was noted, as well as a migratory influence, exerted by NGF, on vascular smooth muscle cells in vitro (Donovan et al., 1995). In murine skin, finally, epidermal keratinocytes were shown to express Trk-B and to proliferate in response to stimulation with BDNF, NT-3, and NT-4/5 (Botchkarev et al., 1999). In accordance with the cited evidence, a fundamental role for NTs outside the nervous system (a role in differentiation, tissue remodeling, proliferation, and migration) is becoming widely accepted.

We showed that human and rat HSCs express the neurotrophins NGF, BDNF, NT-3, and NT-4/5 as well as the neurotrophin receptors p75 and Trk-C in normal and various pathological conditions (Fig. 2) (Cassiman et al., 2001a, 2002a). In addition, HSCs express Trk-A mRNA and Trk-B protein. The functional significance of the present data is supported by the finding that expression of NT and NTr in HSCs is maintained during activation, migration, and proliferation of HSCs in various pathological and experimental conditions. For instance, the numbers of HSCs immunoreactive to NGF, NT-3, Trk-C, and p75 following galactosamine (Gal) intoxication increased significantly over the first 48 hr by a factor of 5, 9, 2, and 7, respectively. The proliferation of HSCs in the Gal intoxication model, measured by bromodeoxyuridine incorporation, has been described to reach a maximum around 48 hr postintoxication, the number of desmin-reactive HSCs reaching a maximum at the same point. The number of HSCs was described to increase by a factor of 2–3 over the first 48 hr, with the increase of desmin-reactive cells ascribed, at least in part, to proliferation. Combination of our quantitative data with these published data strongly suggests upregulation of NT-3 and p75 expression in HSCs following Gal intoxication, since their respective increases (factors of 9 and 7) clearly exceed the expected increase in HSC number (factors of 2 to 3).

The immunohistochemical expression of both NTs and their receptors by HSCs suggests an autocrine and/or paracrine growth factor action, in analogy with findings in other organs and systems (Chalazonitis et al., 1994; Lamballe et al., 1994). RT-PCR results and results from immunohistochemistry (also at lower antibody dilutions), however, suggest additional expression sites. Low titers of BDNF, NT-4/5, and Trk-A mRNA transcripts and high titers of NT-3 transcripts were found in rat hepatocytes as well as immunoreactivity to BDNF, NT-3, and NT-4/5. Thus, a growth factor action by NTs, emanating from hepatocytes, could be exerted on HSCs and vice versa. We also observed staining of bile duct epithelial cells for NT-4/5 and Trk-B. Since coproliferation of reactive bile ductules/progenitor cells with HSCs is a well-known feature in regenerating liver, after submassive necrosis and in animal models of bile ductular proliferation (Miyazaki et al., 1993), a paracrine interaction between HSCs and the progenitor cell compartment is probable. Since NT-4/5 binds to Trk-B, autocrine action of NT-4/5 in progenitor cells and bile duct epithelial cells is also a possibility.

There were significant differences in the number of HSCs reactive for the different NT/NTr, as demonstrated for NGF, NT-3, p75, and Trk-C. The question as to whether every single HSC produces every single NT or carries every single NTr cannot be answered, however. HSCs are a heterogeneous population of cells, as was already shown for desmin (Niki et al., 1996) and αSMA (Schmitt-Gräff et al., 1991; Hautekeete and Geerts, 1997). What can be stated on the basis of our observations, however, is that each of the NT/NTr is expressed in at least a subset of HSCs. This is convincingly shown by colocalization of the different NT/NTr in lobular desmin or αSMA-immunoreactive cells, while KP-1 and neurofilament staining showed no lobular colocalization with any of the NT/NTr in parallel sections, excluding the expression of NT/NTr in Kupffer cells and lobular nerve endings (Cassiman et al., 2001a).

Based on the new markers for HSCs, three different subpopulations of HSCs/MFs were described: lobular HSCs, interface myofibroblasts, and portal/septal myofibroblasts (Knittel et al., 1999a, 1999b; Cassiman and Roskams, 2002; Cassiman et al., 2002a). These subpopulations can predominate in certain diseases/animal models and are possibly important for drug targeting.

α-B-Crystallin Expression in HSCs

The list of neural/neuroendocrine features described in HSCs is extending (Reynaert et al., 2000). Reactive gliosis, a condition characterized by proliferation of glial cells in response to injury to the central nervous system, shows striking morphological similarities with the behavior of HSCs in varying pathological conditions. Besides apparent morphological parallels, HSCs and glial cells share the expression of nestin, neural cell adhesion molecule, and GFAP (Niki et al., 1996, 1999). Reactive glial cells show marked upregulation of ABCRYS in neurodegenerative diseases such as Alzheimer's, multiple sclerosis, and Creutzfeld-Jacob disease (Oertel et al., 2000; van Rijk and Bloemendal, 2000). The most abundant alpha-beta-cristallin (ABCRYS) expression in the central nervous system is found in glial cells, but ABCRYS is also found in neurons (Oertel et al., 2000; van Rijk and Bloemendal, 2000). In the liver, ABCRYS expression is not restricted to HSCs, but is also found in hepatocytes, albeit at lower expression levels (Cassiman et al., 2001b). ABCRYS expression and induction in the activated/reactive cells constitutes yet another parallel between HSCs and glial cells.

Although the exact function of ABCRYS protein is not entirely clear, it is known that ABCRYS is implicated in the cellular defense against stress, through, for instance, protection of intermediate filaments (Van den Ijssel et al., 1999). It is clear, however, that ABCRYS expression is not merely a common endpoint, induced by all kinds of cellular stress. For instance, in glial cells, ABCRYS could only be induced by tumor necrosis factor-α, while other heat-shock proteins were in turn induced by other, evenly specific cytokines (Bajramovic et al., 2000).

EFFECT OF PARASYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER

  1. Top of page
  2. Abstract
  3. LIVER PROGENITOR CELLS
  4. ACTIVATION OF PROGENITOR CELLS
  5. NEUROENDOCRINE FEATURES OF HEPATIC PROGENITOR CELL COMPARTMENT
  6. EXTRACELLULAR MATRIX COMPONENTS INDUCE ENDOCRINE DIFFERENTIATION IN VITRO IN HUMAN LIVER CELL CULTURES THAT HAVE A BILE DUCT PHENOTYPE
  7. HEPATIC STELLATE CELLS AND MYOFIBROBLASTS: HSC ACTIVATION
  8. NEUROENDOCRINE FEATURES OF STELLATE CELLS/MYOFIBROBLASTS
  9. EFFECT OF PARASYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  10. EFFECT OF SYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  11. LITERATURE CITED

Hepatic branch vagotomy suppresses liver regeneration after partial hepatectomy (Ohtake et al., 1993) and causes an impairment of the ductular reaction after bile duct ligation (LeSage et al., 1999). Conversely, stimulation of the nervus vagus by ventromedial hypothalamic lesions promotes liver regeneration after partial hepatectomy (Kiba et al., 1994). HSCs are known to coproliferate and interact with progenitor cells (Dabeva and Shafritz, 1993; Hu et al., 1993; Miyazaki et al., 1993; Fujio et al., 1994; Alison et al., 1996; Yin et al., 1999). HSCs show close contacts with hepatic nerve endings (Ueno et al., 1997; Akiyoshi and Terada, 1998) and express GFAP (Gard et al., 1985; Neubauer et al., 1996; Niki et al., 1996), N-CAM (Knittel et al., 1996; Nakatani et al., 1996), nestin (Niki et al., 1999), SYN (Cassiman et al., 1999), and neurotrophins and their receptors (Cassiman et al., 2001). Functional complexes of myofibroblasts, mast cells, and cholinergic nerve terminals have been suggested to play a role in the development of cirrhosis (Akiyoshi and Terada, 1998). Reactive bile ductules/progenitor cells display neural/neuroendocrine features such as chromogranin-A, neural cell adhesion molecule, parathyroid hormone-related peptide, S-100 protein, and neurotrophins and neurotrophin receptors (Sciot et al., 1986; Roskams et al., 1990, 1991, 1993a, 1993b; Cassiman et al., 2001a, 2001b).

In view of the possible function of neural/neuroendocrine factors in the regulation of progenitor cell differentiation and proliferation and the role of the nervus vagus in liver regeneration, we studied the expression of the five known muscarinic acetylcholine receptors in normal and diseased human liver.

We showed that human hepatic progenitor cells express the M3 receptor (Fig. 3). A minority of intermediate hepatobiliary cells faintly expresses the M3 receptor. These findings suggest that the expression of the M3 receptor rapidly decreases upon differentiation of hepatic progenitor cells toward the hepatocytic lineage, while it is retained upon differentiation toward the biliary lineage. Immunohistochemistry could not demonstrate any other type of muscarinic receptor (M1, M2, M4, M5) on hepatic progenitor cells or their progeny. Low abundance of M2 or M5 receptor expression cannot be excluded, however, since M2 and M5 messenger RNA transcripts were shown to be present in human liver homogenates, but protein could not be localized by immunohistochemistry.

thumbnail image

Figure 3. a: Double immunostaining for CK7 and M3R in hepatitis C cirrhosis. Colocalization in HPCs and reactive ductules is shown in yellow. An intermediate hepatocyte-like cell (arrow) is positive for CK7, but negative for the M3 receptor. b: Immunohistochemistry for CK7 on a liver biopsy from a kidney transplant patient with chronic hepatitis C (innervated) and on a matched biopsy from a transplant liver with recurrent chronic hepatitis C (denervated). The activation of the progenitor cell compartment is much more pronounced in the innervated compared to the denervated liver.

Download figure to PowerPoint

We observed expression of the M3 receptor in septal and interlobular bile ducts in human liver. Alvaro et al. (1997) recently showed that cholangiocytes of both large and small bile ducts in rat liver express the M3 receptor. Further experiments revealed that acetylcholine secreted by the hepatic branch of the nervus vagus acts as a trophic factor for proliferating cholangiocytes via binding to the M3 receptor (Alvaro et al., 1997). Binding of acetylcholine to the M3 receptor also stimulates the proliferation of other epithelial and neural cells such as human colon cancer cells (Frucht et al., 1999), rat astrocytes, human astrocytoma cells (Guizzetti and Costa, 1996; Guizzetti et al., 1996), and human prostate cancer cells (Rayford et al., 1997).

At the moment, the hepatic nerves are the only known source of acetylcholine in the liver, since only they contain the necessary choline acetyltransferase enzyme for the production of acetylcholine and the vesicular acetylcholine transporter for the packaging of acetylcholine (Houwing et al., 1996; Schafer et al., 1998; Xue et al., 2000). It is not impossible that additional sites of acetylcholine production are present in the liver, in nonneural cells, as has been demonstrated in other tissues (Carmeliet and Denef, 1989; Kummer and Haberberger, 1999; James and Nijkamp, 2000). Because in animal experiments, a clear trophic influence of the hepatic vagus branch on regeneration and typical ductular reaction was shown (Ohtake et al., 1993; Kiba et al., 1994, 1995; LeSage et al., 1999), we can conclude that the acetylcholine secreted by the hepatic nerves must be of importance and cannot be compensated by any additional—at the moment hypothetical—local source. How acetylcholine, secreted by the hepatic nerves, reaches the M3 receptor expressed on hepatic progenitor cells is not elucidated at the moment. Electron microscopic studies to date have not shown any contacts between nerve endings and hepatic progenitor cells (De Vos and Desmet, 1992; Scoazec et al., 1993), but acetylcholine is known to be capable of mediating its effects in a paracrine manner (Wessler et al., 1999).

Hepatocytes produce and secrete cholinesterase (Brown et al., 1981; Schuman et al., 1984; Berninsone et al., 1989) and acetylcholinesterase is present at the cell surface membrane (Perelman et al., 1989). The concentration of serum cholinesterase accurately reflects the hepatocyte mass (Brown et al., 1981) and can be used to assess the prognosis of patients with liver cirrhosis (Adler et al., 1997). The pattern of expression of the M3 receptor, in combination with the absence of direct innervation of the M3 expressing hepatic progenitor cells and the secretion of cholinesterase by hepatocytes, has led us to propose the following model (Fig. 4). In normal liver and after partial hepatectomy, each hepatic progenitor cell is surrounded by a normal number of hepatocytes that abrogates binding of acetylcholine to the M3 receptor on the hepatic progenitor cell by producing and secreting cholinesterase. As a consequence, the hepatic progenitor compartment is not activated in these conditions. When there is loss and impaired proliferation of hepatocytes, however, as is the case in almost all human liver diseases, the presence of cholinesterase activity will decrease proportionally to the localization and severity of hepatocyte loss. This would allow acetylcholine to exert its trophic effects on the hepatic progenitor cells, until hepatocyte mass is restored again. A similar mechanism seems to be present in the hematopoietic system. Acetylcholinesterase, which is synthesized by mature red blood cells, enhances apoptosis and reduces proliferation of mouse hematopoietic progenitor cells committed to erythroid and other lineages in vitro (Soreq et al., 1994; Deutsch et al., 2002). Hepatic progenitor cells have a phenotypic overlap with hematopoietic progenitor cells (Fujio et al., 1994).

thumbnail image

Figure 4. Hypothetical model in which the interplay between acetylcholine secreted by the nervus vagus, the M3 receptor expressed by hepatic progenitor cells (HPCs), and acetylcholinesterase secreted by hepatocytes explains HPC behavior in different conditions. a: Normal liver or PHx. b: Liver diseases. c: Liver diseases in transplant liver.

Download figure to PowerPoint

During liver transplantation, the hepatic branch of the nervus vagus is sectioned and reinnervation of the transplant liver occurs only at the hilum (Boon et al., 1992; Dhillon et al., 1992). Thus, acetylcholine can no longer exert its beneficiary effects on hepatic progenitor cells, neither in normal nor in diseased transplant liver. Since, on the one hand, animal experiments strongly indicate a role for the hepatic vagus branch and cholinomimetic substances in liver regeneration, and, on the other hand, our present results raise the possibility of a comparable interaction in human liver (Fig. 3b), further research to elucidate the function of the M3 receptor expressed by human hepatic progenitor cells and a possible role for cholinomimetic substances in patients with denervated livers is indicated.

EFFECT OF SYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER

  1. Top of page
  2. Abstract
  3. LIVER PROGENITOR CELLS
  4. ACTIVATION OF PROGENITOR CELLS
  5. NEUROENDOCRINE FEATURES OF HEPATIC PROGENITOR CELL COMPARTMENT
  6. EXTRACELLULAR MATRIX COMPONENTS INDUCE ENDOCRINE DIFFERENTIATION IN VITRO IN HUMAN LIVER CELL CULTURES THAT HAVE A BILE DUCT PHENOTYPE
  7. HEPATIC STELLATE CELLS AND MYOFIBROBLASTS: HSC ACTIVATION
  8. NEUROENDOCRINE FEATURES OF STELLATE CELLS/MYOFIBROBLASTS
  9. EFFECT OF PARASYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  10. EFFECT OF SYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  11. LITERATURE CITED

Recent reports show that inhibition of the sympathetic nervous system, either via α1-adrenergic antagonism with prasozin or via chemical sympathectomy with 6-hydroxydopamin, promotes progenitor cell activation and reduces liver injury in a mouse model of progenitor cell activation (methionine/choline-deficient diet) (Oben et al., 2003a). Similar experiments on CCL4-intoxicated rats showed inhibition of fibrosis in chronic CCL4-intoxicated rats by the sympathetic inhibitors prasozin and 6-hydroxydopamine (Dubuisson et al., 2002). Oben et al. (2003b) have also shown that stellate cells themselves produce norepinephrin and react to norepinephrine.

We recently studied both acute and chronic CCL4 intoxication and acute galactosamine intoxication (data not shown) and showed that in these acute and chronic rat models, prasozin induced significantly higher numbers of progenitor cells (marked by OV-6) and significantly lower number of hepatic stellate cells (marked by GFAP, desmin, and αSMA). The degree of fibrosis was lower in the prasozin-treated animals than in the control animals, confirming the results of Dubuisson et al. (2002). Both progenitor cells and isolated stellate cells express α-adrenergic receptors. Since prasozin is a well-tolerated drug, this opens interesting perspectives for future treatment modalities.

In conclusion, progenitor cells and hepatic stellate cells form a cell compartment with neuroendocrine features in the liver. There is increasing evidence that both the sympathetic and parasympathetic nervous system influence this cell compartment. This is important in the development of future therapeutic strategies to stimulate progenitor cell activation and to inhibit stellate cell activation. Such sustaining therapies could eventually delay or even replace liver transplantation.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. LIVER PROGENITOR CELLS
  4. ACTIVATION OF PROGENITOR CELLS
  5. NEUROENDOCRINE FEATURES OF HEPATIC PROGENITOR CELL COMPARTMENT
  6. EXTRACELLULAR MATRIX COMPONENTS INDUCE ENDOCRINE DIFFERENTIATION IN VITRO IN HUMAN LIVER CELL CULTURES THAT HAVE A BILE DUCT PHENOTYPE
  7. HEPATIC STELLATE CELLS AND MYOFIBROBLASTS: HSC ACTIVATION
  8. NEUROENDOCRINE FEATURES OF STELLATE CELLS/MYOFIBROBLASTS
  9. EFFECT OF PARASYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  10. EFFECT OF SYMPATHETIC NERVOUS SYSTEM ON NEUROENDOCRINE COMPARTMENT OF LIVER
  11. LITERATURE CITED