Willem G.E.J. Schoonen

Gene expression analysis of precision-cut human liver slices indicates stable expression of ADME-Tox related genes☆

In the process of drug development it is of high importance to test the safety of new drugs with predictive value for human toxicity. A promising approach of toxicity testing is based on shifts in gene expression profiling of the liver. Toxicity screening based on animal liver cells cannot be directly extrapolated to humans due to species differences. The aim of this study was to evaluate precision-cut human liver slices as in vitro method for the prediction of human specific toxicity by toxicogenomics. The liver slices contain all cell types of the liver in their natural architecture. This is important since drug-induced toxicity often is a multi-cellular process. Previously we showed that toxicogenomic analysis of rat liver slices is highly predictive for rat in vivo toxicity. In this study we investigated the levels of gene expression during incubation up to 24 h with Affymetrix microarray technology. The analysis was focused on a broad spectrum of genes related to stress and toxicity, and on genes encoding for phase-I, -II and -III metabolizing enzymes and transporters. Observed changes in gene expression were associated with cytoskeleton remodeling, extracellular matrix and cell adhesion, but for the ADME-Tox related genes only minor changes were observed. PCA analysis showed that changes in gene expression were not associated with age, sex or source of the human livers. Slices treated with acetaminophen showed patterns of gene expression related to its toxicity. These results indicate that precision-cut human liver slices are relatively stable during 24h of incubation and represent a valuable model for human in vitro hepatotoxicity testing despite the human inter-individual variability.

Validation of in vitro screening models for progestagenic activities: Inter-assay comparison and correlation with in vivo activity in rabbits

The PR CALUX® cell line is a stably transfected human U2-OS cell line expressing the human PR and a luciferase reporter construct containing three progesterone-responsive elements coupled to a minimal promoter. The validity of this assay has been studied as an alternative to the McPhail assay in rabbits, an in vivo assay to detect progestins. The PR CALUX assay was characterized by its stable expression of PR protein which leads to induction of endogenous PR target genes by progestins. It was found to have a highly selective response to low levels of different progestins, as well as an insignificant response to other nuclear hormone receptor ligands. As an important step in their validation, the PR CALUX bioassay was compared with another earlier described in vitro bioassay, a Chinese Hamster Ovary (CHO) cell-based PR-CHO reporter gene assay as well as with an in vitro PR-binding (PR-BIN) assay, and the in vivo McPhail assay. This was done using 35 (with the most accurate potency determinations in all tests) and 50 (with less reliable potency determinations in some tests) compounds tested in all assays. The correlation scores between PR CALUX and PR-CHO were r(2)=0.77, and 0.93, respectively; between PR CALUX and PR-BIN r(2)=0.69 and 0.80. Comparison between either the PR CALUX or the PR-CHO transactivation assay and the in vivo McPhail assay revealed very good correlations of r(2)=0.68 (n=35), and 0.85 (n=50). The transactivation assays can discriminate very potent, from potent, weak and inactive compounds rather easily. Besides testing the biological activity of pure chemicals and pharmaceuticals in vitro, the PR CALUX and PR-CHO transactivation assays proved to be relatively good predictors of in vivo progestagenic activity, allowing the use of these assays as prescreening methods or in vitro alternatives.

High content screening analysis of phospholipidosis: Validation of a 96-well assay with CHO-K1 and HepG2 cells for the prediction of in vivo based phospholipidosis

Drug-induced phospholipidosis is marked by an excessive accumulation of phospholipids in lysosomes which can occur after exposure to cationic amphiphilic drugs. Phospholipidosis is considered as an adverse side effect and may delay or negatively affect registration of drug candidates. Currently, the gold standard method of phospholipidosis detection is electron microscopy on tissue samples. This technique is time consuming and only performed relatively late in drug development. Therefore, in vitro screening methods for phospholipidosis are essential in early drug development. In this study, an in vitro phospholipidosis detection assay is developed with CHO-K1 and HepG2 cells by using the fluorescent marker NBD-PE and high content screening analysis. Lysosomal localization of NBD-PE was demonstrated by colocalization with Lysotracker and lamellar body formation by electron microscopy. Upon drug exposure, lysosomal NBD-PE accumulation can be visualized and quantified. Validation with 56 reference compounds, divided in 25 phospholipidosis inducers and 31 negative compounds, showed that this new in vitro assay has a high sensitivity (CHO-K1=92.0% and HepG2=88.0%) and specificity (CHO-K1=87.1% and HepG2=80.6%) for predicting phospholipidosis in vivo. Thus a selective screening tool has been developed for early selection of drug candidates with low probability for phospholipidosis.

Cytotoxic effects of 109 reference compounds on rat H4IIE and human HepG2 hepatocytes. III: Mechanistic assays on oxygen consumption with MitoXpress and NAD(P)H production with Alamar Blue™.

In vitro toxicity screening can reduce the attrition rate of drug candidates in the pharmaceutical industry in the early development process. The focus in this study is to compare the sensitivity for cytotoxicity of a time-resolved fluoro metric oxygen probe with that of a fluoro metric Alamar Blue™ (AB) assay. Both assays measure mitochondrial activity by either oxygen consumption (LUX-A65N-1 (MitoXpress, Luxcel) probe) or NADH/FADH conversion (AB). Both assays were carried out with increasing concentrations of 109 reference compounds using rat H4IIE and human HepG2 hepatocytes at incubation periods of 24, 48 and 72 h. Prior to this study, the influence on medium with either glucose or galactose was studied to analyze the rate of glycolysis and oxygen consumption, which latter process may be impaired in hepatoma cells. Inhibitors of oxygen consumption in combination with a glucose up-take inhibitor showed the largest consumption rate differences in the presence of 5mM of glucose. The choice for the 109 reference compounds was based on the so-called Multicentre Evaluation for In vitro Cytotoxicity (MEIC) and on diverse drug categories. For 59 toxic reference compounds, an evaluation for both assays was carried up to 10(-3)M. Toxicity was demonstrated with MitoXpress for 23 (39%) and 36 (61%) compounds in H4IIE and HepG2 cells, respectively, and with AB for 44 (75%) and 40 (68%) compounds. For 50 more pharmaceutical drugs more physiological concentrations were used up to 3.16×10(-5)M, and only 19 (38%) of these compounds appeared to be toxic in both assays. In conclusion, overall 63 (58%) and 60 (55%) compounds showed toxic effects with the MitoXpress and AB assays on rat H4IIE and human HepG2 hepatocytes, respectively. AB assays were more sensitive with respect to H4IIE cells and MitoXpress assays with respect to HepG2 cells. At all tested time intervals, MitoXpress showed its sensitivity, while AB is more sensitive at 48 and 72 h. With AB more toxic compounds were identified, whereas MitoXpress was more sensitive for a few compounds. A species specific difference was clearly found with digoxin, a human specific potassium channel inhibitor. Thus both assays are valuable identifiers of early toxicity with discrimination in time, compounds and species.

Acute toxicity of CCl4 but not of paracetamol induces a transcriptomic signature of fibrosis in precision-cut liver slices.

In rat in vivo, both paracetamol (APAP) and carbon tetrachloride (CCl4) induce liver necrosis, but long-term treatment with CCl4, in contrast to paracetamol, causes liver fibrosis. The aim of this study was to perform transcriptomic analysis to compare the early changes in mRNA expression profiles induced by APAP and CCl4 in the rat precision-cut liver slice model (PCLS) and to identify early markers that could predict fibrosis-inducing potential. Microarray data of rat PCLS exposed to APAP andCCl4was generated using a toxic dose based on decrease in ATP levels. Toxicity pathway analysis using a custom made fibrosis-related gene list showed fibrosis as one of the predominant toxic endpoints in CCl4-treated, but not in APAP-treated PCLS. Moreover, genes which have a role in fibrosis such as alpha-B crystallin, jun proto-oncogene, mitogen-activated protein kinase 6, serpin peptidase inhibitor and also the transcription factor Kruppel-like-factor-6 were up-regulated by CCl4, but not by APAP. Predicted activation or inhibition of several upstream regulators due to CCl4 is in accordance with their role in fibrosis. In conclusion, transcriptomic analysis of PCLS successfully identified the fibrotic potential of CCl4 as opposed to APAP. The application of PCLS as an ex vivo model to identify early biomarkers to predict the fibrogenic potential of toxic compounds should be further explored.