Within the past decade, tremendous progress has been made in the isolation and culture of human hepatocytes for drug metabolism and toxicology, which could potentially reduce the number of animal experiments performed. However, human hepatocyte cultures are still not widely used for preclinical drug testing, partly due to inconsistent supply and quality of human tissue. Thus, the aim of this study was to evaluate primary cultured human hepatocytes from different patients over a study period of 14 days, by assays that characterise cell quality and function. We found urea production and albumin synthesis in all cell cultures over at least 7 days. Cytochrome P4501A2, CYP2D6, and CYP3A4 protein expression was demonstrated by Western Blot analysis and CYP1A1/2 and CYP3A4 induction by 3-methylcholantrene, phenobarbital or rifampicin over 14 days. In addition, we saw that UDP-glucoronyltransferase activity was preserved in human hepatocytes over 2 weeks.
In conclusion, we could show that primary human hepatocytes isolated from discarded liver tissue can consistently be kept in culture over a long time period and are therefore well suited for preclinical drug testing.
The present work was partially supported by ZEBET/BgVV Z 5.1-1328-154.

The aim of the present work was to establish an in vitro screening assay for drug candidates using human endothelial cells as a model for vascular injury after intravenous application. Different endpoints for viability and functionality of endothelial cells were investigated in human umbilical vein endothelial cells (HUVEC) and in immortalised human endothelial cells (IVEC). Cellular viability was determined by measuring ATP content and by the AlamarBlue assay. For comparison, the toxicity of the selected compounds was also tested in a murine fibroblast cell line (3T3 cells). Selected endpoints for endothelial cell-specific function were vascular permeability, determined by measurement of the transendothelial resistance and the diffusion of tracer molecules (FITC-dextran), and the release of prostaglandin and thromboxane as indicators for prothrombotic or vasoconstrictory action. Five compounds (cyclosporin A, mitomycin C, menadione, amrinone and rolipram) were selected due to their known effects on the vasculature. The cytotoxicity of all compounds was similar in endothelial and 3T3 cells. ATP content and AlamarBlue metabolism did not differ significantly except for amrinone. A dose-dependent decrease of transendothelial resistance and an increase in FITC-dextran permeability could be measured in HUVEC cells for the tested compounds, but the sensitivity was not higher than that of the cytotoxicity assays. Increased prostaglandin or thromboxane release was detected for all compounds at cytotoxic concentrations and for rolipram also at non-toxic concentrations. In conclusion, for a first ranking of drug candidates, cytotoxicity assays on any of the three cell types used are appropriate. For a more detailed characterisation of individual compounds, functional assays on HUVEC cells are proposed.

Concerns have been raised about the potential of natural and man-made chemicals to interfere with the endocrine system of humans and wildlife adversely. This issue has been addressed by intensive research over the past years. Most studies focused on interactions with sex hormone receptors. An estrogenic or (anti)androgenic potential was demonstrated for a number of natural and anthropogenic compounds. This work focused on the detection of compounds with affinity to the androgen receptor (AR). An AR binding assay based on a recombinant rat AR (fusion protein to thioredoxin) properly ranked strong and weak (anti)androgens and other compounds with affinity to the AR. The data correlated excellently with that obtained in binding studies employing a cytosolic AR preparation from rat prostate. Studies with detergents indicated that these are confounders of the assay. They most likely disturb ligand receptor interactions and/or modification of the receptor protein. A transactivation assay based on the androgen-sensitive PALM cell line detected strong and weak androgens, and was able to detect antiandrogenicity when cells had been stimulated with an androgen. These findings indicate that in vitro methods can contribute successfully to the characterisation of compounds with affinity to sex hormone receptors. Whereas binding assays provide the basic information on affinity to the receptors, transactivation assays can discriminate agonists and antagonists and also provide information on cytotoxicity. These in vitro assays could also reduce animal use in endocrine toxicity testing (a) by using recombinant sex hormone receptors instead of cytosolic receptor preparations, (b) by applying intelligent screening strategies that include (potential) metabolites, as these assays probably have only little or no metabolic capacity and by combining several in vitro systems should equivocal results be obtained in one system, and (c) by using in vitro methods to identify mechanisms of action of effects observed in animal studies. In order to support the acceptance of these methods, a thorough validation and the characterisation of confounders is desirable. Whether receptor binding and transactivation assays can predict the outcome of in vivo screening assays reliably remains to be established.

The rapid development of molecular toxicology is providing innovative approaches to an improved investigation and recognition of toxic substances. Proteome analysis offers, with 2DE/MS (two-dimensional gel electrophoresis and mass spectrometry) and SELDI (surface enhanced laser desorption/ionisation), a promising discipline to classify molecular changes caused by toxic exposure. The Rat Liver Foci Bioassay (RLFB) is a detailed, well-described model for the investigation of liver carcinogenesis induced by chemical substances. Based on this model, we examined whether proteomic methods of molecular toxicology can be used for the early recognition of toxic and/or carcinogenic characteristics of toxic substances. In addition, identification and subsequent prevalidation of new hepatocellular biomarkers was performed, enabling better prediction of toxic and/or carcinogenic effects. This could lead to a more meaningful RLFB and thus to an improved risk assessment of chemicals.
2DE analysis in this study showed that deregulated proteins are assigned to mainly anabolic and catabolic metabolism pathways in the cell. Beyond this, individual proteins were identified which play a key role in the carcinogenic process. A comparison of the differentially expressed proteins in tissue from tumour-bearing animals and tissue derived from the start of the study revealed that protein expression changes (biomarkers) were already detectable shortly after exposure.
In addition, analysis by SELDI clearly showed several differentially expressed proteins and/or derived masses. The spectra represented specific differences in tissues, which could be assigned to the same histopathological endpoints. With bioinformatics analysis it was possible to identify individual discriminating mass peaks, which were indicative of tumour formation. Group specific changes can be illustrated and/or represented in more detail with further cluster analysis methods.
These results give hope for an improved prediction of hepatotoxicity and carcinogenicity by means of protein markers, which could in the future lead to a shortening of carcinogenicity studies and to a reduction in the use of experimental animals.

Pharmaceutical and chemical industries are facing new challenges for hazard and risk assessment from regulatory agencies. Especially for potential embryotoxicity of active compounds, conclusions from animal testing remain problematic due to numerous species-specific effects. Developmental toxicity screening preferentially should be performed with human material.
Appropriate models are scarce or missing, and the development of a human in vitro model for the molecular characterisation of embryotoxic effects appears to be highly desirable. The outstanding advantages of a human embryonic stem cell (hESC) based in vitro screening model for embryonic neurotoxicity become clear from corresponding results from a murine ESC-screening system. This in vitro test system is based on neuronal differentiated murine embryonic stem cells and quantitative differential proteomic display techniques to identify biomarkers for neurotoxicity. Results are superior to those of conventional array technologies (nucleic acids), because the proteomic analysis covers posttranslational modifications.
Under the new strict guidelines for stem cell importation of the German Ministry of Health and a Central Ethics Commission for Stem Cell Research, it is now possible for the first time to exploit the outstanding features of human embryonic stem cells to establish an innovative screening method for embryo- and neurotoxicity and to identify toxicity biomarkers without using animal-based in vitro or in vivo systems.

The inflammation model of the isolated hemoperfused bovine uterus was used to introduce a new in vitro model for the investigation of anti-inflammatory substances.
As previous studies demonstrated both an increase in PGE2 synthesis and an up-regulation of COX-2 and iNOS mRNA by ischemia-reperfusion injury in the model (Braun and Kietzmann, 2004), inhibitory effects of the glucocorticoid dexamethasone, the NSAID flunixin and the selective COX-2 inhibitor DFU (5,5-dimethyl-3-(3-fluorophenyl)-4-(4-methylsulphonyl)-phenyl-2-(5H)-furanone) were studied. All substances caused a significant decrease in tissue PGE2 production, while none induced down-regulation of COX-2 mRNA. A slight decrease in the mRNA level of iNOS was observed after 300 minutes of perfusion with dexamethasone-supplemented perfusion medium.
In conclusion, the suitability of the isolated hemoperfused bovine uterus for the investigation of anti-inflammatory substances, especially regarding their COX-2 selectivity, was demonstrated. Use of the isolated hemoperfused bovine uterus in pharmacological research and drug screening may contribute to a reduction of animal testing.

Cell culture models are useful tools to study the uptake of drugs across the barriers of the human body, like the intestine, the skin or the blood-brain barrier. Cell-based in vitro models not only help to reduce the number of animals used but are also much faster to perform, more cost effective and give more reproducible data than animal studies. Given the increasing number of new drugs and chemicals under development, there is an urgent need for the establishment of such in vitro models. However, the validity of such in vitro models is reflected by its ability to accurately predict the behaviour of a substance at the corresponding in vivo barrier. Here, we compare a well-established cell culture model for the intestine, based on Caco-2 colon carcinoma cells, with a primary cell culture model of the blood-brain barrier. We find that Caco-2 cells and cells of the blood-brain barrier have different barrier properties. Therefore, cells used for cell-based assays should be derived from the corresponding tissue to reflect the in vivo barrier characteristics.

Clostridium (C.) perfringens is a Gram-positive anaerobic spore-forming bacterium. Disease caused by C. perfringens infection is called enterotoxaemia. C. perfringens strains are classified on the basis of the lethal exotoxins formed by the bacteria. Epsilon toxin is one of the major lethal toxins and is formed by C. perfringens types B and D.
C. perfringens is an ubiquitous bacterium. Infection occurs via food, water, animal litter or soil. Affected animals include mainly sheep, pigs and cattle. C. perfringens infection manifests as pulpy kidney disease and diarrhoea in suckling lambs.
Enterotoxaemia development is peracute in most cases. Animals die suddenly while grazing on the pasture, without any prior signs of disease. Therefore, treatment is possible only in very rare cases.
Suitable immunoprophylactic measures are the treatment of choice to combat the disease: Vaccines and immunosera have therefore been used extensively for a long time.
The requirements for quality, efficacy and safety testing of the inactivated vaccines are laid down in the Ph. Eur. in the monograph: Clostridium perfringens vaccines for veterinary use. After a marketing authorisation is attained, the product batches must be tested in laboratory animal models for their potency against all vaccine components (Pharmeuropa, 1997).
For potency testing (batch control) of C. perfringens types B and D, the induction of specific antibodies against epsilon toxin in rabbits must be verified. For this purpose, 10 rabbits are immunised twice with the product to be tested. Their blood is taken 14 days after the last immunisation and the serum is pooled. The pooled serum is then tested for its protective effect. This is done by means of the toxin neutralisation test in mice (optionally also in guinea pigs) in comparison with an international reference serum. The evaluation criterion is the death rate of the mice in the test and reference groups after administration of lethal doses of epsilon toxin. The exact efficacy of the test serum is given in International Units (IU). The tested serum must show a minimum content of 5 IU.
This in vivo method requires a very high number of experimental animals. Approximately 400 mice (or 50 guinea pigs) are used per vaccine batch.
The monograph for C. perfringens vaccines, which has recently been revised, expressly indicates that a validated serological method may be used for batch testing. In addition, a reference serum known as clostridium multicomponent serum has been available since 2000. The objective is to test vaccine batches against this reference and by means of a competitive ELISA developed in the precursor project, using a monoclonal antibody for direct determination of specific antitoxins in rabbit sera.
This ELISA method was subjected to an international validation to verify whether the protocol and the precision can be transferred within and between the participating laboratories.

An alternative assay for replacing animal experiments should serve the specific microphysiological needs of the cells and be endowed with multiparametric signal monitoring. These requirements are provided by a test system in which the key elements are biocompatible electronic sensor-chips. It is also connected to a medium perfusion set-up, which allows to control the supply of nutrients and test compounds, and the removal of culture medium. The chips are equipped with sensors that continuously monitor basic metabolic parameters and membrane-associated changes of living cells: pH-ISFETs for extracellular acidification, amperometric O2-sensors for oxygen consumption, and IDES for electrical impedance of the cell layer. Experiments with LS174T colon carcinoma cells in culture show the metabolic and electrical changes upon incubation with Zytochalasin B and chloroacetaldehyde. The signal patterns vary and indicate different mechanisms of action for these test compounds. With this test system it is possible to detect effects of unknown substances and mixtures, and to analyse the cellular probe for prolonged times.

The local delivery of drugs to the cochlea is a promising alternative to systemic treatment of inner ear disorders. Whilst new drugs are being developed for this purpose, it is important to determine the time course and total dose required for the various target regions within the inner ear. Due to the small fluid spaces of the inner ear and the resulting experimental and analytical difficulties, many animal studies have only obtained one sample per animal. This results in limited information about drug time courses at specific locations in the inner ear.
We show here how computer models considering general pharmacokinetic principles and inner ear geometry are used for application of the 3R-principle in animal research while avoiding experimental sampling artefacts. This can be achieved by: (1) careful planning and interpretation of experiments to study pharmacokinetics in the inner ear, (2) optimising volume sampling techniques, (3) facilitating the use of advantageous, continuous sampling methods like microdialysis and (4) developing a 3D-model that will permit consideration of the complex geometry of the inner ear when transferring results from one species to another.