Model-guided identification of a therapeutic strategy to reduce hyperammonemia in liver diseases

A liver digital twin for in silico testing of cellular and inter-cellular mechanisms in regeneration after drug-induced damage

This communication presents a mathematical mechanism-based model of the regenerating liver after drug-induced pericentral lobule damage resolving tissue microarchitecture. The consequence of alternative hypotheses about the interplay of different cell types on regeneration was simulated. Regeneration dynamics has been quantified by the size of the damage-induced dead cell area, the hepatocyte density and the spatial-temporal profile of the different cell types. We use deviations of observed trajectories from the simulated system to identify branching points, at which the systems behavior cannot be explained by the underlying set of hypotheses anymore. Our procedure reflects a successful strategy for generating a fully digital liver twin that, among others, permits to test perturbations from the molecular up to the tissue scale. The model simulations are complementing current knowledge on liver regeneration by identifying gaps in mechanistic relationships and guiding the system toward the most informative (lacking) parameters that can be experimentally addressed.

Increased sinusoidal export of drug glucuronides is a compensative mechanism in liver cirrhosis of mice

Rationale: Liver cirrhosis is known to affect drug pharmacokinetics, but the functional assessment of the underlying pathophysiological alterations in drug metabolism is difficult. Methods: Cirrhosis in mice was induced by repeated treatment with carbon tetrachloride for 12 months. A cocktail of six drugs was administered, and parent compounds as well as phase I and II metabolites were quantified in blood, bile, and urine in a time-dependent manner. Pharmacokinetics were modeled in relation to the altered expression of metabolizing enzymes. In discrepancy with computational predictions, a strong increase of glucuronides in blood was observed in cirrhotic mice compared to vehicle controls. Results: The deviation between experimental findings and computational simulations observed by analyzing different hypotheses could be explained by increased sinusoidal export and corresponded to increased expression of export carriers (Abcc3 and Abcc4). Formation of phase I metabolites and clearance of the parent compounds were surprisingly robust in cirrhosis, although the phase I enzymes critical for the metabolism of the administered drugs in healthy mice, Cyp1a2 and Cyp2c29, were downregulated in cirrhotic livers. RNA-sequencing revealed the upregulation of numerous other phase I metabolizing enzymes which may compensate for the lost CYP isoenzymes. Comparison of genome-wide data of cirrhotic mouse and human liver tissue revealed similar features of expression changes, including increased sinusoidal export and reduced uptake carriers. Conclusion: Liver cirrhosis leads to increased blood concentrations of glucuronides because of increased export from hepatocytes into the sinusoidal blood. Although individual metabolic pathways are massively altered in cirrhosis, the overall clearance of the parent compounds was relatively robust due to compensatory mechanisms.

Cognitive Functions, Neurotransmitter Alterations, and Hippocampal Microstructural Changes in Mice Caused by Feeding on Western Diet

Metabolic Dysfunction Associated Steatotic Liver Disease (MASLD) is the most common chronic liver disease in Western countries. It is becoming increasingly evident that peripheral organ-centered inflammatory diseases, including liver diseases, are linked with brain dysfunctions. Therefore, this study aims to unravel the effect of MASLD on brain histology, cognitive functions, and neurotransmitters. For this purpose, mice fed for 48 weeks on standard (SD) or Western diet (WD) were evaluated by behavioral tests, followed by sacrifice and analysis of the liver-brain axis including histopathology, immunohistochemistry, and biochemical analyses. Histological analysis of the liver showed features of Metabolic Dysfunction-Associated Steatohepatitis (MASH) in the WD-fed mice including lipid droplet accumulation, inflammation, and fibrosis. This was accompanied by an elevation of transaminase and alkaline phosphatase activities, increase in inflammatory cytokine and bile acid concentrations, as well as altered amino acid concentrations in the blood. Interestingly, compromised blood capillary morphology coupled with astrogliosis and microgliosis were observed in brain hippocampus of the WD mice, indicating neuroinflammation or a disrupted neurovascular unit. Moreover, attention was impaired in WD-fed mice along with the observations of impaired motor activity and balance, enhanced anxiety, and stereotyped head-twitch response (HTR) behaviors. Analysis of neurotransmitters and modulators including dopamine, serotonin, GABA, glutamate, and acetylcholine showed region-specific dysregulation in the brain of the WD-fed mice. In conclusion, the induction of MASH in mice is accompanied by the alteration of cellular morphology and neurotransmitter expression in the brain, associated with compromised cognitive functions.

Uncovering Actions of Type 3 Deiodinase in the Metabolic Dysfunction-Associated Fatty Liver Disease (MAFLD)

Metabolic dysfunction-associated fatty liver disease (MAFLD) has gained worldwide attention as a public health problem. Nonetheless, lack of enough mechanistic knowledge restrains effective treatments. It is known that thyroid hormone triiodothyronine (T3) regulates hepatic lipid metabolism, and mitochondrial function. Liver dysfunction of type 3 deiodinase (D3) contributes to MAFLD, but its role is not fully understood. Objective: To evaluate the role of D3 in the progression of MAFLD in an animal model. Methodology: Male/adult Sprague Dawley rats (n = 20) were allocated to a control group (2.93 kcal/g) and high-fat diet group (4.3 kcal/g). Euthanasia took place on the 28th week. D3 activity and expression, Uncoupling Protein 2 (UCP2) and type 1 deiodinase (D1) expression, oxidative stress status, mitochondrial, Krebs cycle and endoplasmic reticulum homeostasis in liver tissue were measured. Results: We observed an increase in D3 activity/expression (p < 0.001) related to increased thiobarbituric acid reactive substances (TBARS) and carbonyls and diminished reduced glutathione (GSH) in the MAFLD group (p < 0.05). There was a D3-dependent decrease in UCP2 expression (p = 0.01), mitochondrial capacity, respiratory activity with increased endoplasmic reticulum stress in the MAFLD group (p < 0.001). Surprisingly, in an environment with lower T3 levels due to high D3 activity, we observed an augmented alpha-ketoglutarate dehydrogenase (KGDH) and glutamate dehydrogenase (GDH) enzymes activity (p < 0.05). Conclusion: Induced D3, triggered by changes in the REDOX state, decreases T3 availability and hepatic mitochondrial capacity. The Krebs cycle enzymes were altered as well as endoplasmic reticulum stress. Taken together, these results shed new light on the role of D3 metabolism in MAFLD.

Digital twin demonstrates significance of biomechanical growth control in liver regeneration after partial hepatectomy

Partial liver removal is an important therapy option for liver cancer. In most patients within a few weeks, the liver is able to fully regenerate. In some patients, however, regeneration fails with often severe consequences. To better understand the control mechanisms of liver regeneration, experiments in mice were performed, guiding the creation of a spatiotemporal 3D model of the regenerating liver. The model represents cells and blood vessels within an entire liver lobe, a macroscopic liver subunit. The model could reproduce the experimental data only if a biomechanical growth control (BGC)-mechanism, inhibiting cell cycle entrance at high compression, was taken into account and predicted that BGC may act as a short-range growth inhibitor minimizing the number of proliferating neighbor cells of a proliferating cell, generating a checkerboard-like proliferation pattern. Model-predicted cell proliferation patterns in pigs and mice were found experimentally. The results underpin the importance of biomechanical aspects in liver growth control.

Spatial modeling reveals nuclear phosphorylation and subcellular shuttling of YAP upon drug-induced liver injury

The Hippo signaling pathway controls cell proliferation and tissue regeneration via its transcriptional effectors yes-associated protein (YAP) and transcriptional coactivator with PDZ-binding motif (TAZ). The canonical pathway topology is characterized by sequential phosphorylation of kinases in the cytoplasm that defines the subcellular localization of YAP and TAZ. However, the molecular mechanisms controlling the nuclear/cytoplasmic shuttling dynamics of both factors under physiological and tissue-damaging conditions are poorly understood. By implementing experimental in vitro data, partial differential equation modeling, as well as automated image analysis, we demonstrate that nuclear phosphorylation contributes to differences between YAP and TAZ localization in the nucleus and cytoplasm. Treatment of hepatocyte-derived cells with hepatotoxic acetaminophen (APAP) induces a biphasic protein phosphorylation eventually leading to nuclear protein enrichment of YAP but not TAZ. APAP-dependent regulation of nuclear/cytoplasmic YAP shuttling is not an unspecific cellular response but relies on the sequential induction of reactive oxygen species (ROS), RAC-alpha serine/threonine-protein kinase (AKT, synonym: protein kinase B), as well as elevated nuclear interaction between YAP and AKT. Mouse experiments confirm this sequence of events illustrated by the expression of ROS-, AKT-, and YAP-specific gene signatures upon APAP administration. In summary, our data illustrate the importance of nuclear processes in the regulation of Hippo pathway activity. YAP and TAZ exhibit different shuttling dynamics, which explains distinct cellular responses of both factors under physiological and tissue-damaging conditions.

Inhibition of cytochrome P450 enhances the nephro- and hepatotoxicity of ochratoxin A

The mycotoxin ochratoxin A (OTA) is a contaminant in food that causes nephrotoxicity and to a minor degree hepatotoxicity. Recently, we observed that OTA induces liver damage preferentially to the cytochrome P450 (CYP)-expressing pericentral lobular zone, similar to hepatotoxic substances known to be metabolically toxified by CYP, such as acetaminophen or carbon tetrachloride. To investigate whether CYP influences OTA toxicity, we used a single dose of OTA (7.5 mg/kg; intravenous) with and without pre-treatment with the pan CYP-inhibitor 1-aminobenzotriazole (ABT) 2 h before OTA administration. Blood, urine, as well as liver and kidney tissue samples were collected 24 h after OTA administration for biochemical and histopathological analyses. Inhibition of CYPs by ABT strongly increased the nephro- and hepatotoxicity of OTA. The urinary kidney damage biomarkers kidney injury molecule-1 (KIM-1) and neutrophil gelatinase-associated lipocalin (NGAL) were increased > 126-fold and > 20-fold, respectively, in mice treated with ABT and OTA compared to those receiving OTA alone. The blood biomarkers of liver damage, alanine transaminase (ALT) and aspartate transaminase (AST) both increased > 21- and 30-fold, respectively, when OTA was administered to ABT pre-treated mice compared to the effect of OTA alone. Histological analysis of the liver revealed a pericentral lobular damage induced by OTA despite CYP-inhibition by ABT. Administration of ABT alone caused no hepato- or nephrotoxicity. Overall, the results presented are compatible with a scenario where CYPs mediate the detoxification of OTA, yet the mechanisms responsible for the pericental liver damage pattern still remain to be elucidated.

From Seeing to Simulating: A Survey of Imaging Techniques and Spatially-Resolved Data for Developing Multiscale Computational Models of Liver Regeneration

Liver regeneration, which leads to the re-establishment of organ mass, follows a specifically organized set of biological processes acting on various time and length scales. Computational models of liver regeneration largely focused on incorporating molecular and signaling detail have been developed by multiple research groups in the recent years. These modeling efforts have supported a synthesis of disparate experimental results at the molecular scale. Incorporation of tissue and organ scale data using noninvasive imaging methods can extend these computational models towards a comprehensive accounting of multiscale dynamics of liver regeneration. For instance, microscopy-based imaging methods provide detailed histological information at the tissue and cellular scales. Noninvasive imaging methods such as ultrasound, computed tomography and magnetic resonance imaging provide morphological and physiological features including volumetric measures over time. In this review, we discuss multiple imaging modalities capable of informing computational models of liver regeneration at the organ-, tissue- and cellular level. Additionally, we discuss available software and algorithms, which aid in the analysis and integration of imaging data into computational models. Such models can be generated or tuned for an individual patient with liver disease. Progress towards integrated multiscale models of liver regeneration can aid in prognostic tool development for treating liver disease.

Treatment of Hyperammonemia by Transplanting a Symbiotic Pair of Intestinal Microbes

Hyperammonemia is a deleterious and inevitable consequence of liver failure. However, no adequate therapeutic agent is available for hyperammonemia. Although recent studies showed that the pharmabiotic approach could be a therapeutic option for hyperammonemia, its development is clogged with poor identification of etiological microbes and low transplantation efficiency of candidate microbes. In this study, we developed a pharmabiotic treatment for hyperammonemia that employs a symbiotic pair of intestinal microbes that are both able to remove ammonia from the surrounding environment. By a radioactive tracing experiment in mice, we elucidated how the removal of ammonia by probiotics in the intestinal lumen leads to lower blood ammonia levels. After determination of the therapeutic mechanism, ammonia-removing probiotic strains were identified by high-throughput screening of gut microbes. The symbiotic partners of ammonia-removing probiotic strains were identified by screening intestinal microbes of a human gut, and the pairs were administrated to hyperammonemic mice to evaluate therapeutic efficacy. Blood ammonia was in a chemical equilibrium relationship with intestinal ammonia. Lactobacillus reuteri JBD400 removed intestinal ammonia to shift the chemical equilibrium to lower the blood ammonia level. L. reuteri JBD400 was successfully transplanted with a symbiotic partner, Streptococcus rubneri JBD420, improving transplantation efficiency 2.3×10³ times more compared to the sole transplantation while lowering blood ammonia levels significantly. This work provides new pharmabiotics for the treatment of hyperammonemia as well as explains its therapeutic mechanism. Also, this approach provides a concept of symbiotic pairs approach in the emerging field of pharmabiotics.

Automated Detection of Portal Fields and Central Veins in Whole-Slide Images of Liver Tissue

Many physiological processes and pathological phenomena in the liver tissue are spatially heterogeneous. At a local scale, biomarkers can be quantified along the axis of the blood flow, from portal fields (PFs) to central veins (CVs), i.e., in zonated form. This requires detecting PFs and CVs. However, manually annotating these structures in multiple whole-slide images is a tedious task. We describe and evaluate a fully automated method, based on a convolutional neural network, for simultaneously detecting PFs and CVs in a single stained section. Trained on scans of hematoxylin and eosin-stained liver tissue, the detector performed well with an F1 score of 0.81 compared to annotation by a human expert. It does, however, not generalize well to previously unseen scans of steatotic liver tissue with an F1 score of 0.59. Automated PF and CV detection eliminates the bottleneck of manual annotation for subsequent automated analyses, as illustrated by two proof-of-concept applications: We computed lobulus sizes based on the detected PF and CV positions, where results agreed with published lobulus sizes. Moreover, we demonstrate the feasibility of zonated quantification of biomarkers detected in different stainings based on lobuli and zones obtained from the detected PF and CV positions. A negative control (hematoxylin and eosin) showed the expected homogeneity, a positive control (glutamine synthetase) was quantified to be strictly pericentral, and a plausible zonation for a heterogeneous F4/80 staining was obtained. Automated detection of PFs and CVs is one building block for automatically quantifying physiologically relevant heterogeneity of liver tissue biomarkers. Perspectively, a more robust and automated assessment of zonation from whole-slide images will be valuable for parameterizing spatially resolved models of liver metabolism and to provide diagnostic information.

Spatio-Temporal Multiscale Analysis of Western Diet-Fed Mice Reveals a Translationally Relevant Sequence of Events during NAFLD Progression

Mouse models of non-alcoholic fatty liver disease (NAFLD) are required to define therapeutic targets, but detailed time-resolved studies to establish a sequence of events are lacking. Here, we fed male C57Bl/6N mice a Western or standard diet over 48 weeks. Multiscale time-resolved characterization was performed using RNA-seq, histopathology, immunohistochemistry, intravital imaging, and blood chemistry; the results were compared to human disease. Acetaminophen toxicity and ammonia metabolism were additionally analyzed as functional readouts. We identified a sequence of eight key events: formation of lipid droplets; inflammatory foci; lipogranulomas; zonal reorganization; cell death and replacement proliferation; ductular reaction; fibrogenesis; and hepatocellular cancer. Functional changes included resistance to acetaminophen and altered nitrogen metabolism. The transcriptomic landscape was characterized by two large clusters of monotonously increasing or decreasing genes, and a smaller number of 'rest-and-jump genes' that initially remained unaltered but became differentially expressed only at week 12 or later. Approximately 30% of the genes altered in human NAFLD are also altered in the present mouse model and an increasing overlap with genes altered in human HCC occurred at weeks 30-48. In conclusion, the observed sequence of events recapitulates many features of human disease and offers a basis for the identification of therapeutic targets.

Epigenomic and transcriptional profiling identifies impaired glyoxylate detoxification in NAFLD as a risk factor for hyperoxaluria

Epigenetic modifications (e.g. DNA methylation) in NAFLD and their contribution to disease progression and extrahepatic complications are poorly explored. Here, we use an integrated epigenome and transcriptome analysis of mouse NAFLD hepatocytes and identify alterations in glyoxylate metabolism, a pathway relevant in kidney damage via oxalate release-a harmful waste product and kidney stone-promoting factor. Downregulation and hypermethylation of alanine-glyoxylate aminotransferase (Agxt), which detoxifies glyoxylate, preventing excessive oxalate accumulation, is accompanied by increased oxalate formation after metabolism of the precursor hydroxyproline. Viral-mediated Agxt transfer or inhibiting hydroxyproline catabolism rescues excessive oxalate release. In human steatotic hepatocytes, AGXT is also downregulated and hypermethylated, and in NAFLD adolescents, steatosis severity correlates with urinary oxalate excretion. Thus, this work identifies a reduced capacity of the steatotic liver to detoxify glyoxylate, triggering elevated oxalate, and provides a mechanistic explanation for the increased risk of kidney stones and chronic kidney disease in NAFLD patients.

Subcellular spatio-temporal intravital kinetics of aflatoxin B1 and ochratoxin A in liver and kidney

Local accumulation of xenobiotics in human and animal tissues may cause adverse effects. Large differences in their concentrations may exist between individual cell types, often due to the expression of specific uptake and export carriers. Here we established a two-photon microscopy-based technique for spatio-temporal detection of the distribution of mycotoxins in intact kidneys and livers of anesthetized mice with subcellular resolution. The mycotoxins ochratoxin A (OTA, 10 mg/kg b.w.) and aflatoxin B1 (AFB1, 1.5 mg/kg b.w.), which both show blue auto-fluorescence, were analyzed after intravenous bolus injections. Within seconds after administration, OTA was filtered by glomeruli, and enriched in distal tubular epithelial cells (dTEC). A striking feature of AFB1 toxicokinetics was its very rapid uptake from sinusoidal blood into hepatocytes (t1/2 ~ 4 min) and excretion into bile canaliculi. Interestingly, AFB1 was enriched in the nuclei of hepatocytes with zonal differences in clearance. In the cytoplasm of pericentral hepatocytes, the half-life (t1/2~ 63 min) was much longer compared to periportal hepatocytes of the same lobules (t1/2 ~ 9 min). In addition, nuclear AFB1 from periportal hepatocytes cleared faster compared to the pericentral region. These local differences in AFB1 clearance may be due to the pericentral expression of cytochrome P450 enzymes that activate AFB1 to protein- and DNA-binding metabolites. In conclusion, the present study shows that large spatio-temporal concentration differences exist within the same tissues and its analysis may provide valuable additional information to conventional toxicokinetic studies.

Ammonia inhibits energy metabolism in astrocytes in a rapid and glutamate dehydrogenase 2-dependent manner

Astrocyte dysfunction is a primary factor in hepatic encephalopathy (HE) impairing neuronal activity under hyperammonemia. In particular, the early events causing ammonia-induced toxicity to astrocytes are not well understood. Using established cellular HE models, we show that mitochondria rapidly undergo fragmentation in a reversible manner upon hyperammonemia. Further, in our analyses, within a timescale of minutes, mitochondrial respiration and glycolysis were hampered, which occurred in a pH-independent manner. Using metabolomics, an accumulation of glucose and numerous amino acids, including branched chain amino acids, was observed. Metabolomic tracking of 15N-labeled ammonia showed rapid incorporation of 15N into glutamate and glutamate-derived amino acids. Downregulating human GLUD2 [encoding mitochondrial glutamate dehydrogenase 2 (GDH2)], inhibiting GDH2 activity by SIRT4 overexpression, and supplementing cells with glutamate or glutamine alleviated ammonia-induced inhibition of mitochondrial respiration. Metabolomic tracking of 13C-glutamine showed that hyperammonemia can inhibit anaplerosis of tricarboxylic acid (TCA) cycle intermediates. Contrary to its classical anaplerotic role, we show that, under hyperammonemia, GDH2 catalyzes the removal of ammonia by reductive amination of α-ketoglutarate, which efficiently and rapidly inhibits the TCA cycle. Overall, we propose a critical GDH2-dependent mechanism in HE models that helps to remove ammonia, but also impairs energy metabolism in mitochondria rapidly.

Highlight report: New insights in liver physiology: Canalicular bile flux is diffusion dominated

One of the central functions of the liver is excretion of bile into the intestine. Currently, bile excretion is explained by the osmotic model, according to which bile acids are excreted by hepatocytes into the bile canaliculi and since bile acids are osmotically active they draw water into the canalicular lumen. Bile canaliculi are closed at the central side. Therefore, bile was postulated to flow to the open side into the ducts. However, bile flow in canaliculi has never been measured because of the small canalicular diameter which does not allow analysis of flux by conventional methods. Recently, methods have been developed that allow flow analysis in bile canaliculi and ducts. Interestingly, no measurable directed flow was observed in the canaliculi. Instead, small molecules in bile canaliculi reached the larger bile ducts by diffusion. Only there measurable flow sets in. The pathophysiological implications of this novel observation are discussed.

Liver Bioreactor Design Issues of Fluid Flow and Zonation, Fibrosis, and Mechanics: A Computational Perspective

Tissue engineering, with the goal of repairing or replacing damaged tissue and organs, has continued to make dramatic science-based advances since its origins in the late 1980’s and early 1990’s. Such advances are always multi-disciplinary in nature, from basic biology and chemistry through physics and mathematics to various engineering and computer fields. This review will focus its attention on two topics critical for tissue engineering liver development: (a) fluid flow, zonation, and drug screening, and (b) biomechanics, tissue stiffness, and fibrosis, all within the context of 3D structures. First, a general overview of various bioreactor designs developed to investigate fluid transport and tissue biomechanics is given. This includes a mention of computational fluid dynamic methods used to optimize and validate these designs. Thereafter, the perspective provided by computer simulations of flow, reactive transport, and biomechanics responses at the scale of the liver lobule and liver tissue is outlined, in addition to how bioreactor-measured properties can be utilized in these models. Here, the fundamental issues of tortuosity and upscaling are highlighted, as well as the role of disease and fibrosis in these issues. Some idealized simulations of the effects of fibrosis on lobule drug transport and mechanics responses are provided to further illustrate these concepts. This review concludes with an outline of some practical applications of tissue engineering advances and how efficient computational upscaling techniques, such as dual continuum modeling, might be used to quantify the transition of bioreactor results to the full liver scale.

Holland C, Zaza A, Murad W, Hassan R, A. Ahmed Y, Abbas T, A. Abdelrahim E, Schneider KM, Matz-Soja M, Reinders J, Gebhardt R, Berres ML, Hatting M, Drasdo D, Saez-Rodriguez J, Trautwein C, G. Hengstler J. Influence of Liver Fibrosis on Lobular Zonation

Little is known about how liver fibrosis influences lobular zonation. To address this question, we used three mouse models of liver fibrosis, repeated CCl4 administration for 2, 6 and 12 months to induce pericentral damage, as well as bile duct ligation (21 days) and mdr2-/- mice to study periportal fibrosis. Analyses were performed by RNA-sequencing, immunostaining of zonated proteins and image analysis. RNA-sequencing demonstrated a significant enrichment of pericentral genes among genes downregulated by CCl4; vice versa, periportal genes were enriched among the upregulated genes. Immunostaining showed an almost complete loss of pericentral proteins, such as cytochrome P450 enzymes and glutamine synthetase, while periportal proteins, such as arginase 1 and CPS1 became expressed also in pericentral hepatocytes. This pattern of fibrosis-associated 'periportalization' was consistently observed in all three mouse models and led to complete resistance to hepatotoxic doses of acetaminophen (200 mg/kg). Characterization of the expression response identified the inflammatory pathways TGFβ, NFκB, TNFα, and transcription factors NFKb1, Stat1, Hif1a, Trp53, and Atf1 among those activated, while estrogen-associated pathways, Hnf4a and Hnf1a, were decreased. In conclusion, liver fibrosis leads to strong alterations of lobular zonation, where the pericentral region adopts periportal features. Beside adverse consequences, periportalization supports adaptation to repeated doses of hepatotoxic compounds.

Prediction of human drug-induced liver injury (DILI) in relation to oral doses and blood concentrations

Drug-induced liver injury (DILI) cannot be accurately predicted by animal models. In addition, currently available in vitro methods do not allow for the estimation of hepatotoxic doses or the determination of an acceptable daily intake (ADI). To overcome this limitation, an in vitro/in silico method was established that predicts the risk of human DILI in relation to oral doses and blood concentrations. This method can be used to estimate DILI risk if the maximal blood concentration (C_max) of the test compound is known. Moreover, an ADI can be estimated even for compounds without information on blood concentrations. To systematically optimize the in vitro system, two novel test performance metrics were introduced, the toxicity separation index (TSI) which quantifies how well a test differentiates between hepatotoxic and non-hepatotoxic compounds, and the toxicity estimation index (TEI) which measures how well hepatotoxic blood concentrations in vivo can be estimated. In vitro test performance was optimized for a training set of 28 compounds, based on TSI and TEI, demonstrating that (1) concentrations where cytotoxicity first becomes evident in vitro (EC₁₀) yielded better metrics than higher toxicity thresholds (EC₅₀); (2) compound incubation for 48 h was better than 24 h, with no further improvement of TSI after 7 days incubation; (3) metrics were moderately improved by adding gene expression to the test battery; (4) evaluation of pharmacokinetic parameters demonstrated that total blood compound concentrations and the 95%-population-based percentile of C_max were best suited to estimate human toxicity. With a support vector machine-based classifier, using EC₁₀ and C_max as variables, the cross-validated sensitivity, specificity and accuracy for hepatotoxicity prediction were 100, 88 and 93%, respectively. Concentrations in the culture medium allowed extrapolation to blood concentrations in vivo that are associated with a specific probability of hepatotoxicity and the corresponding oral doses were obtained by reverse modeling. Application of this in vitro/in silico method to the rat hepatotoxicant pulegone resulted in an ADI that was similar to values previously established based on animal experiments. In conclusion, the proposed method links oral doses and blood concentrations of test compounds to the probability of hepatotoxicity.

Assessment of stem cell differentiation based on genome-wide expression profiles

In recent years, protocols have been established to differentiate stem and precursor cells into more mature cell types. However, progress in this field has been hampered by difficulties to assess the differentiation status of stem cell-derived cells in an unbiased manner. Here, we present an analysis pipeline based on published data and methods to quantify the degree of differentiation and to identify transcriptional control factors explaining differences from the intended target cells or tissues. The pipeline requires RNA-Seq or gene array data of the stem cell starting population, derived 'mature' cells and primary target cells or tissue. It consists of a principal component analysis to represent global expression changes and to identify possible problems of the dataset that require special attention, such as: batch effects; clustering techniques to identify gene groups with similar features; over-representation analysis to characterize biological motifs and transcriptional control factors of the identified gene clusters; and metagenes as well as gene regulatory networks for quantitative cell-type assessment and identification of influential transcription factors. Possibilities and limitations of the analysis pipeline are illustrated using the example of human embryonic stem cell and human induced pluripotent cells to generate 'hepatocyte-like cells'. The pipeline quantifies the degree of incomplete differentiation as well as remaining stemness and identifies unwanted features, such as colon- and fibroblast-associated gene clusters that are absent in real hepatocytes but typically induced by currently available differentiation protocols. Finally, transcription factors responsible for incomplete and unwanted differentiation are identified. The proposed method is widely applicable and allows an unbiased and quantitative assessment of stem cell-derived cells.This article is part of the theme issue 'Designer human tissue: coming to a lab near you'.

Bile Microinfarcts in Cholestasis Are Initiated by Rupture of the Apical Hepatocyte Membrane and Cause Shunting of Bile to Sinusoidal Blood

Bile duct ligation (BDL) is an experimental procedure that mimics obstructive cholestatic disease. One of the early consequences of BDL in rodents is the appearance of so-called bile infarcts that correspond to Charcot-Gombault necrosis in human cholestasis. The mechanisms causing bile infarcts and their pathophysiological relevance are unclear. Therefore, intravital two photon-based imaging of BDL mice was performed with fluorescent bile salts (BS) and non-BS organic anion analogues. Key findings were followed up by matrix-assisted laser desorption ionization imaging, clinical chemistry, immunostaining, and gene expression analyses. In the acute phase, 1-3 days after BDL, BS concentrations in bile increased and single-cell bile microinfarcts occurred in dispersed hepatocytes throughout the liver caused by the rupture of the apical hepatocyte membrane. This rupture occurred after loss of mitochondrial membrane potential, followed by entry of bile, cell death, and a "domino effect" of further death events of neighboring hepatocytes. Bile infarcts provided a trans-epithelial shunt between bile canaliculi and sinusoids by which bile constituents leaked into blood. In the chronic phase, ≥21 days after BDL, uptake of BS tracers at the sinusoidal hepatocyte membrane was reduced. This contributes to elevated concentrations of BS in blood and decreased concentrations in the biliary tract. Conclusion: Bile microinfarcts occur in the acute phase after BDL in a limited number of dispersed hepatocytes followed by larger infarcts involving neighboring hepatocytes, and they allow leakage of bile from the BS-overloaded biliary tract into blood, thereby protecting the liver from BS toxicity; in the chronic phase after BDL, reduced sinusoidal BS uptake is a dominant protective factor, and the kidney contributes to the elimination of BS until cholemic nephropathy sets in.