Although normally quiescent, the adult mammalian liver possesses a great capacity to regenerate after different types of injuries in order to restore the lost liver mass and ensure maintenance of the multiple liver functions. Major players in the regeneration process are mature residual cells, including hepatocytes, cholangiocytes and stromal cells. However, if the regenerative capacity of mature cells is impaired by liver-damaging agents, hepatic progenitor cells are activated and expand into the liver parenchyma. Upon transit amplification, the progenitor cells may generate new hepatocytes and biliary cells to restore liver homeostasis. In recent years, hepatic progenitor cells have been the subject of increasing interest due to their therapeutic potential in numerous liver diseases as alternative or supportive/complementary tools to liver transplantation. While the first investigations on hepatic progenitor cells have focused on their origin and phenotypic characterization, recent attention has focused on the influence of the hepatic microenvironment on their activation and proliferation. This microenvironment comprises the extracellular matrix, epithelial and non-epithelial resident liver cells, and recruited inflammatory cells as well as the variety of growth-modulating molecules produced and/or harboured by these elements. The cellular and molecular responses to different regenerative stimuli seem to depend on the injury inflicted and consequently on the molecular microenvironment created in the liver by a certain insult. This review will focus on molecular responses controlling activation and expansion of the hepatic progenitor cell niche, emphasizing similarities and differences in the microenvironments orchestrating regeneration by recruitment of progenitor cell populations or by replication of mature cells.

The experimental protocols used in the investigation of stem cell-mediated liver regeneration in rodents are characterized by activation of the hepatic stem cell compartment in the canals of Hering followed by transit amplification of oval cells and their subsequent differentiation along hepatic lineages. Although the protocols are numerous and often used interchangeably across species, a thorough comparative phenotypic analysis of oval cells in rats and mice using wellestablished and generally acknowledged molecular markers has not been provided. In the present study, we evaluated and compared the molecular phenotypes of oval cells in several of the most commonly used protocols of stem cell-mediated liver regeneration-namely, treatment with 2-acetylaminofluorene and partial (70%)

Thy-1, a marker of hematopoietic stem cells, has been reported to be expressed by oval cells proliferating during stem cell-mediated regeneration in rat liver, suggesting a relationship between the two cell populations. Consequently, Thy-1 has become an accepted cell surface marker to sort hepatic oval cells. In the present study we used the well-characterized 2-acetylaminfluorene/partial hepatectomy model to induce transit-amplification of hepatic oval cells in the regenerating liver and characterized Thy-1 expression using Northern hybridization, quantitative reverse transcriptase-polymerase chain reaction analysis, immunofluorescence confocal microscopy, and immunoelectronmicroscopy. We found that Thy-1 expression was induced during transit-amplification of the oval cell population, but Thy-1 mRNA was not present in the ␣-fetoprotein-expressing oval cells.

BackgroundDuring liver development, intrahepatic bile ducts are thought to arise by a unique asymmetric mode of cholangiocyte tubulogenesis characterized by a series of remodeling stages. Moreover, in liver diseases, cells lining the Canals of Hering can proliferate and generate new hepatic tissue. The aim of this study was to develop protocols for three-dimensional visualization of protein expression, hepatic portal structures and human hepatic cholangiocyte tubulogenesis.ResultsProtocols were developed to digitally visualize portal vessel branching and protein expression of hepatic cell lineage and extracellular matrix deposition markers in three dimensions. Samples from human prenatal livers ranging from 7 weeks + 2 days to 15½ weeks post conception as well as adult normal and acetaminophen intoxicated liver were used. The markers included cytokeratins (CK) 7 and 19, the epithelial cell adhesion molecule (EpCAM), hepatocyte paraffin 1 (HepPar1), sex determining region Y (SRY)-box 9 (SOX9), laminin, nestin, and aquaporin 1 (AQP1).Digital three-dimensional reconstructions using CK19 as a single marker protein disclosed a fine network of CK19 positive cells in the biliary tree in normal liver and in the extensive ductular reactions originating from intrahepatic bile ducts and branching into the parenchyma of the acetaminophen intoxicated liver. In the developing human liver, three-dimensional reconstructions using multiple marker proteins confirmed that the human intrahepatic biliary tree forms through several developmental stages involving an initial transition of primitive hepatocytes into cholangiocytes shaping the ductal plate followed by a process of maturation and remodeling where the intrahepatic biliary tree develops through an asymmetrical form of cholangiocyte tubulogenesis.ConclusionsThe developed protocols provide a novel and sophisticated three-dimensional visualization of vessels and protein expression in human liver during development and disease.

Cerebral edema is a feared complication to acute liver failure (ALF), but the pathogenesis is still poorly understood. The water channels Aquaporin-1 (Aqp1) and -4 (Aqp4) has been associated with brain edema formation in several neuropathological conditions, indicating a possible role of Aqp1 and/or Aqp4 in ALF mediated brain edema. We induced acute liver injury and hyperammonemia in mice, to evaluate brain edema formation and the parallel expression of Aqp1 and Aqp4 in ALF. Liver injury and hyperammonemia were induced by +D-galactosamine (GLN) plus lipopolysaccharide (LPS) intraperitoneally and intravenous ammonia-acetate (NH(4)(+)), the GLN+LPS+NH(4)(+) group. The vehicle control group (CONTROL) was treated with NaCl and phosphate-buffered saline. The GLN+LPS+NH(4)(+) group showed significantly elevated p-alanine aminotransferase, p-INR and p-ammonium vs. CONTROL (p < 0.001). Cortical brain water content was significantly elevated in the GLN+LPS+NH(4)(+) group vs. CONTROL, mean (SEM) 80.8(0.3) vs 80.0(0.1) % (p < 0.05). Western blot of membrane enriched cortical brain tissue showed significantly upregulation of Aqp4 in the GLN+LPS+NH(4)(+) group vs. CONTROL, mean AU (SEM) 100775(14820) vs. 58857(6266) (p < 0.05), and stationary levels for Aqp1. Aqp1 and Aqp4 mRNA were stationary. This study indicates that Aqp4, but not Aqp1, may be of importance in the pathogenesis of cortical brain edema in mice with ALF.