Effect of periportal- and centrilobular-equivalent oxygen tension on liver specific functions in long-term rat hepatocyte cultures

Defining Physiological Normoxia for Improved Translation of Cell Physiology to Animal Models and Humans

The extensive oxygen gradient between the air we breathe (Po₂ ~21 kPa) and its ultimate distribution within mitochondria (as low as ~0.5-1 kPa) is testament to the efforts expended in limiting its inherent toxicity. It has long been recognized that cell culture undertaken under room air conditions falls short of replicating this protection in vitro. Despite this, difficulty in accurately determining the appropriate O₂ levels in which to culture cells, coupled with a lack of the technology to replicate and maintain a physiological O₂ environment in vitro, has hindered addressing this issue thus far. In this review, we aim to address the current understanding of tissue Po₂ distribution in vivo and summarize the attempts made to replicate these conditions in vitro. The state-of-the-art techniques employed to accurately determine O₂ levels, as well as the issues associated with reproducing physiological O₂ levels in vitro, are also critically reviewed. We aim to provide the framework for researchers to undertake cell culture under O₂ levels relevant to specific tissues and organs. We envisage that this review will facilitate a paradigm shift, enabling translation of findings under physiological conditions in vitro to disease pathology and the design of novel therapeutics.

Liver lobe and strain differences in the activity of murine cytochrome P450 enzymes

The cytochrome P450 (CYP) enzyme superfamily is the most important enzyme system for phase I biotransformation. For toxico- and pharmacokinetic studies, use of liver-based microsomes, including those of mice, is state-of-the-art to study CYP-dependent metabolism. However, reproducibility and interpretation of these data is still very variable, partly because current testing guidelines do not cover details on organ sampling and potential liver lobe differences. Hence, we analyzed CYP activity, CYP protein content, mRNA expression of CYP1A, CYP2C, CYP2D and CYP3A isozymes, and cytochrome P450 reductase (CPR) activity of the four different liver lobes and processus papillaris of male C57BL/6J mice in comparison to whole liver. Additionally, we used whole liver of Balb/cJ and 129S1/SvImJ for strain comparison. Our data show significant differences in CYP activity, being most prominent in lobus sinister lateralis and lobus medialis, and lowest in processus papillaris. These differences were not caused by varying Cyp gene expression or CYP protein level, but partly correspond with lobe specific CPR activities. We also observed significant strain differences in CYP mRNA expression and activities with overall high activities in 129S1/SvImJ mice and low activities in Balb/cJ mice compared to C57BL/6J mice. In addition, strain specific differences in CYP2C and CYP2D activity seem to be reflected in strain dependent differences in CPR activity. In summary, our results indicate that in mice CYP activity and gene expression are strain dependent and may vary highly between liver lobes. To ensure reproducibility and comparability of different probes and studies, this should be taken into account when liver samples are collected for the analysis of CYP-dependent metabolism.

Loss of Nrf2 in mice evokes a congenital intrahepatic shunt that alters hepatic oxygen and protein expression gradients and toxicity

The transcription factor Nrf2 (Nfe2l2 nuclear factor, erythroid 2-like 2) regulates gene expression directly, controlling pharmacological and toxicological responses. These processes may also be influenced by the structure of the hepatic vasculature, which distributes blood flow through compartmentalized microenvironments to maintain organismal stability. Castings of the hepatic portal vasculature of albino C57BL/6J but not ICR Nrf2(-/-) mice revealed a congenital intrahepatic shunt that was present in two thirds of Nrf2-disrupted mice. This shunt directly connected the portal vein to the inferior vena cava and displayed characteristics of a patent ductus venosus. Immunohistochemistry revealed that Nrf2(-/-) mice with an intrahepatic shunt manifest changes to hepatic oxygen and protein expression gradients when compared with wild-type (WT) and non-shunted Nrf2(-/-) mice. Centrilobular hypoxia found in WT and Nrf2(-/-) mice without shunts was reduced in Nrf2(-/-) livers with a shunt. Hepatic protein expression of phosphoenolpyruvate carboxykinase (Pepck), normally confined to the periportal zone, exhibited both periportal and centrilobular zonal expression in livers from Nrf2(-/-) mice with an intrahepatic shunt. Centrilobular expression of Cytochrome P450 2E1 (Cyp2e1) was diminished in shunted Nrf2(-/-) livers compared with WT and Nrf2(-/-) livers without shunts. The intrahepatic shunt in Nrf2(-/-) mice was further found to diminish acetaminophen hepatoxicity compared with WT and Nrf2(-/-) non-shunted mice following a 6 h challenge with 250 mg/kg acetaminophen. The presence of an intrahepatic shunt influences several physiological and pathophysiological properties of Nrf2(-/-) mice through changes in blood flow, hepatic oxygenation, and protein expression that extent beyond loss of canonical transactivation of Nrf2 target genes.

The viability and function of primary rat hepatocytes cultured on polymeric membranes developed for hybrid artificial liver devices

Bioartificial liver devices require membranes to support the function and viability of hepatocytes because they are anchorage-dependent cells. This study investigated the ability of several polymeric membranes to support the functions of primary hepatocyte cultures. Tailor-made membranes were sought by synthesizing acrylonitrile copolymers with different comonomers resulting in ionic, hydrophilic, or reactive functional groups on the polymer surface. Hepatocyte morphology and viability were assessed by confocal microscopy, and function by the content and activities of cytochrome P450, and the expression of glutathione S-transferases. Hydrophilic membranes (polyacrylonitrile and acrylonitrile copolymerized with 2-acrylamino-2-methyl-propane sulfonic acid) were more biocompatible than hydrophobic membranes such as polysulfone. The chemistry of the hydrophilic group was important; amine groups had a deleterious effect on maintenance of the primary hepatocytes. The biocompatibility of hydrophobic membranes was improved by collagen coating. Improving the chemistry of membranes for artificial liver devices will enhance the phenotypic stability of the cells, enabling us to prolong treatment times for patients.

Time course of primary liver cell reorganization in three-dimensional high-density bioreactors for extracorporeal liver support: an immunohistochemical and ultrastructural study

To enable extracorporeal liver support based on the use of primary liver cells, culture models supporting the maintenance of cell integrity and function in vitro are required. In this study the cell organization and ultrastructure of primary porcine hepatocytes cocultured with nonparenchymal cells in three-dimensional high-density bioreactors were analyzed after 10, 20, and 30 days of culture by immunohistochemistry and transmission electron microscopy. Biochemical data showed that metabolic activity of the cells in the system was relatively stable over at least 20 days. Immunohistochemical studies were performed in comparison with donor organ biopsies. They showed that hepatocytes and nonparenchymal cells reaggregated in bioreactors, forming structures partly resembling natural liver parenchyma. Bile duct-like structures characterized by cytokeratin 7 (CK-7) immunoreactivity (IR) were regularly detected. Nonparenchymal cells (vimentin IR) formed sinusoidal-like structures within parenchymal cell aggregates. Proliferative activity (Ki-67 IR) increased over time. The detection of collagen I and laminin indicated the production of extracellular matrix components within bioreactors. The results showed that primary liver cell reorganization and long-term maintenance of their differentiated state were achieved within the bioreactors The findings on cell proliferation indicated that the culture model is also of interest for further in vitro studies on cell regeneration and tissue formation.

Short-term assay for the identification of neurotoxic compounds and their liver derived stable metabolites

The release of stable neurotoxic metabolites from liver after metabolic activation of xenobiotics can be investigated in cultures of primary rat hepatocytes as metabolic activation system and embryonic chicken brain cell cultures as target. It was shown that adult rat hepatocyte cell cultures are a reliable bioactivating system for xenobiotics such as cyclophosphamide and isophenphos resulting in the release of stable metabolites into the supernatant. The cultured embryonic chicken brain cells were able to discriminate between the toxicity of parent drugs and its metabolite(s) and between metabolites with an unspecific cytotoxic activity (cyclophosphamide) and metabolites with a high potential to damage specific nerve cell population(s) (isophenphos). The observed neurotoxicity of the isophenphos metabolites is not an acute effect but induced only after a prolonged exposure period. The present study provides evidence that the subsequent use of hepatocytes and brain cell cultures has the potential to be used as an in vitro screening system for the identification of specific and non-specific neurotoxic compounds. Solely stable metabolites are taken into account, since in the in vivo situation only stable metabolites have a change to reach the nervous system. Our data suggest that the present approach is able to discriminate (a) between cell-type and organ specific toxicity, (b) between neurotoxicity derived after long-term or acute exposure, and (c) between the neurotoxicity induced by the native test compound or stable metabolites.

Probing enhanced cytochrome P450 2B1/2 activity in rat hepatocyte spheroids through confocal laser scanning microscopy

Cytochrome P450 (CYP450) enzymes are essential for xenobiotic metabolism. Although CYP450s are found in many tissues, CYP2B1/2 are primarily expressed in the rat liver. The constitutive expression in vivo of CYP2B1/2 is low but it is induced in the presence of various drugs such as phenobarbital (PB). In this study, CYP2B1/2 activity in cultured hepatocytes was assessed in situ with the introduction of a fluorogenic substrate, pentoxyresorufin. The product of 7-pentoxyresorufin-O-dealkylation (PROD), which is catalyzed specifically by CYP2B1/2, was detected using confocal laser scanning microscopy (CLSM). Primary hepatocytes cultured as monolayers on collagen-coated surfaces exhibited background PROD activity and minimal PB inducibility after 4 days in culture. In contrast, rat hepatocytes organized in compacted aggregates, or spheroids, exhibited higher levels of PROD activity and retained their ability for PB induction. The results from the CLSM analysis were verified by RT-PCR and Western immunoblotting analysis. Furthermore, CLSM in conjunction with image processing techniques and three-dimensional reconstruction revealed the localization of enhanced PROD activity in the center of spheroids. The results support the use of CLSM as a powerful tool for investigating CYP2B1/2 activity in cultured rat hepatocytes.

Effect of oxygen concentration on the expression of glutathione S-transferase activity in periportal and perivenous rat hepatocyte cultures

Cultures of perivenous (PV) and periportal (PP) hepatocytes could provide suitable in vitro models for studying the zone-specific hepatotoxic potential of xenobiotics. However, it is not known whether cultured PP and PV hepatocytes keep their phenotypes when the microcirculation of the liver changes. This question has been studied by culturing rat hepatocytes at 13 and 4% (v/v) O(2), respectively, mimicking the acinar oxygen gradient. PP and PV adult rat hepatocytes were isolated by digitonin-collagenase in situ perfusion and cultured on plastic Falcon and gas-permeable Petriperm dishes in Williams' E medium and kept at 13 and 4% (v/v) O(2), respectively. Cultures at 20% (v/v) O(2) on plastic dishes served as a control. Two types of cultures were studied, namely conventional cultures either unsupplemented or supplemented with 30 mM pyruvate. The activities of glutamine synthetase (GS) and glutathione S-transferase (GST) were measured in freshly isolated PP and PV hepatocytes and all cultures. The heterogeneous expression of GS (PV>PP), observed in freshly isolated hepatocytes, was kept for at least 4 days in culture. Total, Mu and Alpha class GST activities were predominantly expressed in PV freshly isolated cells. However, no beneficial effect could be observed in culture by exposing the cells to their specific in vivo oxygen concentration. The best maintenance of GST PV predominance in culture was observed in Petriperm dishes at 20% (v/v) O(2), as well in pyruvate-supplemented as unsupplemented cultures. PV GST predominance was thus kept in particular when the highest oxygen concentration was used and made available to the cells through the gas-permeable membranes. The results on GS PV predominance support these findings.

Lead stimulates intercellular signalling between hepatocytes and Kupffer cells

The role of intercellular signalling between liver cells in lead (Pb)(1)-induced liver toxicity was investigated in cocultures of freshly isolated and cultured rat hepatocytes and Kupffer cells. The Kupffer cells (seeded onto culture dish inserts), the hepatocytes or the two in cocultures were exposed to Pb acetate (2-50 microM) in combination with lipopolysaccharide (0.1-1000 ng/ml). In hepatocyte cultures, the combined Pb/lipopolysaccharide treatment induced no significant increase in the release of the proinflammatory cytokine tumor necrosis factor-alpha (TNF-alpha) whereas in Kupffer cell cultures and in cocultures, at low lipopolysaccharide levels (0.1 and 1 ng/ml), TNF-alpha release was synergistically increased (up to 30-fold) when compared to lipopolysaccharide exposure alone. This stimulation of Kupffer cell-derived TNF-alpha release was specific for Pb or not detectable with mercury and cadmium. As a response to the Pb/lipopolysaccharide induced release of TNF-alpha, the cocultured hepatocytes increased their nitric oxide (NO) content sixfold when compared with lipopolysaccharide alone and downregulated the negatively regulated acute phase protein albumin. This downregulation was also detectable without lipopolysaccharide and without TNF-alpha release, indicating that Pb induces additional thus far unidentified Kupffer cell-derived factors, which interact with the cocultured hepatocytes. At the time of TNF-alpha release, the viability of the hepatocytes and the Kupffer cells was not affected. However, after a 48-h treatment period, Pb induced a Kupffer cell specific toxicity without affecting the hepatocytes. Loss of hepatocyte viability after lipopolysaccharide/Pb stimulation was only detectable in the presence of cocultured Kupffer cells together with human-derived granulocytes. It is concluded that Pb stimulates intercellular signalling between Kupffer cells and hepatocytes which is synergistically enhanced in the presence of low lipopolysaccharide levels. The released Kupffer cell-derived signals (e.g. cytokines) promotes most likely proteolytic hepatocyte killing in combination with a direct cellular interaction between the granulocytes and the hepatocytes.

Manipulation of the phenotype of immortalised rat hepatocytes by different culture configurations and by dimethyl sulphoxide

The liver-specific phenotype of immortalised rat hepatocytes is not irretrievably lost as they age in culture but can be manipulated by modifying the culture environment. Testosterone metabolism was used to investigate the profile of cytochrome P450 isoenzymes present in two immortalised cell lines, P9 and LQC, and in primary cultures of rat hepatocytes, cultured on collagen films, gels and double gel cultures (sandwich configuration). The extent of testosterone metabolism, and the range of metabolites produced, was increased in immortalised cells by the presence of collagen as a substratum film or gel but survival was poorer and the range of metabolites was reduced in sandwich culture. In contrast, testosterone metabolism was retained in primary hepatocytes in sandwich cultures at a higher level than in collagen film or gel cultures. Expression of alpha class glutathione-S-transferases (GSTs) increased and that of GSTP1 decreased (changes which indicate a recovery of normal liver GST phenotype) when the medium of immortalised cell cultures was supplemented with dimethyl sulphoxide (DMSO). DMSO also improved ethoxyresorufin O-deethylation (EROD) and testosterone metabolism in immortalised cells. It also markedly inhibited proliferation, DNA, RNA and protein synthesis. Maximal testosterone metabolism was observed in immortalised cells cultured on collagen gels in the presence of 1% (v/v) DMSO. Development of a protocol for treating immortalised liver cells cultured on collagen gels with DMSO to switch between proliferation and differentiation may provide a convenient system expressing the xenobiotic metabolising enzymes required for in vitro toxicity testing.

Kupffer cell-mediated differential down-regulation of cytochrome P450 metabolism in rat hepatocytes

Nonparenchymal cells, particularly Kupffer cells, might play an important role in the modulation of xenobiotic metabolism in liver and its pharmacological and toxicological consequences. This intercellular communication via the exchange of soluble factors was investigated in primary rat Kupffer cells and hepatocytes. Freshly isolated rat Kupffer cells were seeded onto cell culture inserts and cocultured with 5 day old serum-free rat hepatocyte monolayer cultures at a ratio of 1:1 for 2 days. Hepatocyte cultures, Kupffer cell cultures or cocultures were treated with 0.1 ng/ml-10 microg/ml lipopolysaccharide (LPS). Within this concentration range, no significant toxicity was observed in either cell type. In LPS-exposed cocultures, tumor necrosis factor alpha (TNFalpha) levels rose up to 5 ng/ml within 5 h; nitric oxide (NO) levels increased up to 70 microM within 48 h of treatment, both in a dose-dependent fashion. The release of negative (albumin) and positive (alpha1-acid-glycoprotein) acute phase proteins from the hepatocytes was strongly down- and up-regulated, respectively. The simultaneous treatment of the cocultures with phenobarbital and LPS (10 ng/ml) or 3-methylcholanthrene and LPS (10 ng/ml) resulted in a strong down-regulation (85%) of the phenobarbital-induced cytochrome P450 (CYP) isoform CYP2B1 in the hepatocytes whereas the 3-methylcholanthrene-induced isoform CYP1A1 was only weakly affected (15%). This specific down-regulation of CYP2B1 was mediated exclusively by TNFalpha, released from the Kupffer cells. It was not linked with NO release from or inducible NO synthase activity in the hepatocytes. The TNFalpha release was not affected by the two xenobiotics. Acetaminophen tested in these cocultures showed no direct interaction with the Kupffer cells. The use of liver cell cocultures is therefore a useful approach to investigate the influence of intercellular communication on xenobiotic metabolism in liver.

Drug metabolism in hepatocyte sandwich cultures of rats and humans

Adult hepatocytes from rat and man were maintained for 2 weeks between two gel layers in a sandwich configuration to study the influence of this culture technique on the preservation of basal activities of xenobiotic-metabolizing phase I and phase II enzymes. The response of these enzyme activities to an enzyme inducer was investigated using rifampicin (RIF). Basal levels of cytochrome P-450 (CYP) isozymes were characterized by measuring ethoxyresorufin O-deethylation (EROD), ethoxycoumarin O-deethylation (ECOD), and the specific oxidation of testosterone (T). In hepatocytes from untreated rats, CYP isozyme levels, including the major form CYP 2C11, increased during the first 3 days in culture. After this period of recovery, the levels of CYP 2C11, CYP 2A1, and CYP 2B1 decreased, whereas CYP 3A1 increased. In contrast to these dynamic changes, CYP activities such as CYP 1A2 and the major isozyme CYP 3A4 were largely preserved until day 9 in cultures of human hepatocytes. In measuring phase II activities, a distinct increase in glucuronosyltransferase (UDP-GT) activity toward p-nitrophenol (PNP) was found for rat and human hepatocytes over 2 weeks in culture. Sulfotransferase (ST) activity toward PNP showed an initial increase, with a maximum at day 7 and day 9 in culture, respectively, and then decreased until day 14. Glutathione S-transferase (GST) activity decreased constantly during the time of culture. Effects of the enzyme-inducing drug rifampicin on phase I and phase II enzymes were investigated using cultured human hepatocytes. Rifampicin treatment (50 micromol/L) for 7 days resulted in a 3.7-fold induction of CYP 3A4 at day 9 in culture. ECOD activity was increased sixfold and phase II ST activity increased twofold compared to the initial value at day 3. No effect of rifampicin on CYP 3A was found in cultures of rat hepatocytes. These results demonstrate that rat and human hepatocytes preserve the major forms of CYP isozymes and phase II activities and respond to inducing drugs such as rifampicin. The novel hepatocyte sandwich culture is suitable for investigating drug metabolism, drug-drug interactions and enzyme induction.

A multicompartment geometric model of the liver in relation to regional induction of cytochrome P450s

A geometric, multicompartment model of the liver was developed to examine regional protein induction and to provide model output suitable for predicting the degree of induction in both the whole liver and in specific regions. The model was based on functional hexagonal arrays within the liver. A geometric representation was used to divide these functional units into five zones: a concentric periportal zone, a fenestrated periportal region that interconnects among multiple functional units, and three concentric centrilobular areas, referred to, respectively, as zones 1 through 5. The surface areas (and volumes for hexagonal cylinders) of these live zones were, respectively, 13.5, 25.2, 33.9, 20.3, and 6.8% of the total liver. The pharmacokinetic model for induction had dissociation constants (Kd) and Hill constants (n) for interactions of transcriptional activator-ligand complexes with response elements on DNA. Estimates of regional induction were converted to color intensities to "paint" the two-dimensional liver for a visual comparison with immunohistochemical observations. To obtain sharp moving boundaries of induced areas with increasing dose (as noted in various experiments), n values in each subcompartment must be large. To create realistic total induction curves that are relatively smooth, the differences in Kd values between adjacent subcompartments must be less than fivefold. Because of the high n values, the low-dose induction characteristics predicted with the multicompartment liver model differ significantly from those predicted with a model that considers the liver as a single homogeneous compartment.

Hepatocyte-derived interleukin-6 and tumor-necrosis factor alpha mediate the lipopolysaccharide-induced acute-phase response and nitric oxide release by cultured rat hepatocytes

The regulation of acute-phase protein production and nitric oxide (NO) release in lipopolysaccharide-induced liver injury is thought to occur in response to monocytes/macrophages and Kupffer-cell-derived cytokines. In this study, we used primary cultured rat hepatocytes maintained as a differentiated phenotype to investigate the direct effects of endotoxin (lipopolysaccharide) on the production of the acute-phase proteins and on NO release. Lipopolysaccharide (10 micrograms/ml) increased the production of alpha 2-macroglobulin 2.5-fold compared to untreated cultures and decreased the production of albumin by 50%. The effect of lipopolysaccharide was mimicked by adding interleukin-6 (IL-6) and tumor-necrosis factor alpha (TNF-alpha), cytokines being induced by treatment of hepatocytes with lipopolysaccharide. Maximal TNF-alpha (600 pg/ml) and IL-6 (1800 pg/ml) concentrations were observed 4 h and 6 h after lipopolysaccharide stimulation, respectively. The lipopolysaccharide-induced acute-phase protein response was blocked by anti-(IL-6) but not by anti-(TNF-alpha) IgG. The latter reduced the lipopolysaccharide-induced IL-6 production by 60%. Besides its effects on the acute-phase proteins, endotoxin caused a significant increase in NO production in cultured rat hepatocytes. Unlike anti-(IL-6) IgG, anti-(TNF-alpha) IgG reduced the lipopolysaccharide-induced NO production by 50% indicating that endotoxin-induced NO production is partially mediated by TNF-alpha but not by IL-6. Preculture with gadolinium chloride (GdCl3), an inhibitor of Kupffer cells, did not change the response of hepatocytes to lipopolysaccharide indicating that the observed findings are direct endotoxin effects on hepatocytes. The data demonstrate that by their production of TNF-alpha and IL-6 rat hepatocytes respond to lipopolysaccharide treatment with an IL-6 mediated acute-phase protein and a TNF-alpha-mediated NO production. These features have previously been attributed to monocytes/macrophages and Kupffer cells.

Tumor necrosis factor alpha differentially modulates the cellular response of rat hepatocytes in periportal- and pericentral-equivalent cultures

Alterations of cellular functions induced by recombinant human tumor necrosis factor alpha (TNF alpha) were compared in rat hepatocytes cultured under either periportal-equivalent (10 nM insulin; 10 nM glucagon; 13% O2) or perivenous-equivalent conditions (10 nM insulin; 1 nM glucagon; 4% O2). TNF alpha induced a time- and dose-dependent increase in nitric oxide (NO) production and an acute phase response (inhibition of albumin secretion and elevation of alpha 2-macroglobulin production) under both culture conditions. NO production was more pronounced in periportal cultures, while the acute phase response was stronger in pericentral cultures. This suggests that NO production and the acute phase response are controlled by different pathways. After exposure to TNF alpha, DNA content was measured fluorimetrically and biochemically. A marked decrease in nuclear DNA content was found exclusively in pericentral cultures after an 8-h exposure, followed by an elevation of lactic dehydrogenase (LDH) release after a 12-h exposure. Aurintricarboxylic acid (100 microM), an inhibitor of endonuclease, significantly inhibited the TNF alpha-induced decrease in nuclear DNA content but only partially inhibited the LDH release. This indicates that the loss of nuclear DNA content in pericentral cultures is due to an activation of endonuclease and the resulting DNA fragmentation and does not correlate with NO production. Furthermore, the release of LDH seems to be only partially associated with DNA damage. Dexamethasone (100 nM) completely inhibited both TNF alpha-induced DNA fragmentation and the elevation of LDH release. The results clearly indicate that the toxicity of TNF alpha is influenced by the metabolic state of hepatocytes. Accordingly, the preferential perivenous cell injury observed after exposure to endotoxins in vivo seems to be due to a higher sensitivity of the pericentrally localized hepatocytes towards TNF alpha rather than a TNF alpha concentration gradient.