Entwicklung eines in vitro Systems zur Untersuchung von substanzinduzierten Genexpressionsänderungen bei primären Hepatozyten der Ratte

Abstract

Development of an in vitro system for the detection of substance induced gene expression alterations in primary rat hepatocytes Following the new European legislation for chemicals (REACH) a large amount of chemicals need to be evaluated from the toxicological perspective. With commonly used techniques, the current testing capacities will be overloaded. Therefore, there is a requirement for faster and more reliable test systems for toxicology. In the last couple of years there has been a lot of progress in the field of toxicogenomics. After exposing animals to toxic substances, it is possible to detect the mode of action based on the substance induced gene expression alterations. This kind of analysis requires animal experiments and is not feasible for high throughput testing. An alternative test method would be the investigation of cultured hepatocytes. Approx. 600 million hepatocytes can be isolated from one rat liver, which would be sufficient for roughly 100 six well plates. At the beginning of this thesis the majority of groups working in this field saw no feasible option to perform gene expression analysis of cultured hepatocytes. This is based on the fact, that cultured hepatocytes undergo a (“spontaneous”) change of more then 1000 genes caused only by the culture conditions. Nonetheless, in this thesis an in vitro system based on primary rat hepatocytes has been established which is able to detect similar substance induced gene expression changes as observed in the in vivo rat model. To achieve this, three commonly used culture systems for the cultivation of primary cells were investigated: (1) the classical collagen monolayer culture, where hepatocytes are cultured on a thin film of collagen, (2) the collagen sandwich culture, where hepatocytes are cultivated between two layers of a collagen gel and (3) Matrigel®-culture, a commercially available smooth extracellular matrix in which the hepatocytes are embedded. The hepatocytes were incubated with methapyrilene as a test substance in a concentration of 100 and 200 μM and the gene expression of four markergenes (Abat, Gsk3b, Myd116 and Sult1a1) was measured. These genes were already known as characteristically up- (Gsk3b, Myd116) and downregulated (Abat, Sult1a1) under the influence of methapyrilene. It could be shown that these characteristic deregulations were the most pronounced and the most reproducible in the collagen sandwich culture. Therefore, the collagen sandwich was used in subsequent experiments. Besides the choice of the collagen sandwich culture as the optimal culture system, it was discovered that there is an advantage to cultivate both control and treated cells on the same six well dish. The variability on the same plate was less then the plate-to-plate variability. When developing a new in vitro system, the choice of the correct concentration of test substances is one critical aspect. This concentration should be in the range of a concentration which can also be obtained in vivo. The results, using the optimized system, show a deregulation at a level of 0.39 μM and a clear deregulation at a level of 6.25 μM methapyrilene. It is known from the literature, that an i.p.-injection of 0.7 mg/kg BW methapyrilene results in a plasma level of approx. 0.5 μM 10 minutes after injection. This led to the conclusion that this test system is able to detect a substance induced effect on gene expression, even at a level close to in vivo relevant concentrations. This was confirmed by measuring additional genes (Bax, Cdkn1, Gsta2, Hsf1, Mdm2 und Nqo1). A concentration dependent effect could also be shown for aflatoxin B1, piperonylbutoxide and 2-nitrofluorene as additional liver-carcinogens. Based on these encouraging results, a larger study was performed to investigate if the substance induced gene expression changes in the in vivo rat liver and in vitro primary rat hepatocytes are comparable. Two genotoxic (aflatoxin B1, 2-nitrofluorene) and two non-genotoxic (methapyrilene, piperonylbutoxide) compounds were used as test substances. Two gene groups were analyzed: one gene group associated with stress response, DNA-repair and metabolism (gene group 1: Abcb1, Ugt1a6, Gsta5, Gsta2, Hsf1, Nqo1, Apex1, Bax, Cdkn1a, Gadd45a, Mt1a, Sds, Mdm2, Myc) and one gene group which is active during cell proliferation (gene group 2: Mcm6, Cdc2, Hdc, 3-Pgdh, Cdc20, Igfbp1, Top2a, Map3k1, Atf3). The in vitro experiments were performed related to this thesis. The results from the in vivo experiments were provided by a cooperation-partner. Interestingly, there was a very good correlation between the in vivoand the in vitro-results when analysing gene group 1 (p< 0.001). For gene group 2 no significant correlation could be found (p=0.286). This could be explained by the fact that stress-response, DNArepair and metabolism are similar processes both in vitro and in vivo. The missing correlation of proliferation-associated genes (gene group 2) could be due to the fact, that there is no replacement proliferation in vitro. Replacement proliferation in organs means that damaged cells are replaced by the proliferation of surrounding healthy cells. With the current state of the art in culturing primary hepatocytes, the effect of replacement-proliferation can not be seen in vitro. The investigation of gene groups, whose substance induced gene expression alterations correlate between in vitro and in vivo systems, will facilitate the development of toxicological test-systems by focusing only on the relevant genes. At present, the use of primary rat hepatocytes is very often criticized. One reason is that hepatocytes produce artificial or false positive results in vitro. For example, the genes Gsk3b and Myd116 were altered by the influence of methapyrilene in vitro only, but not in rat liver in vivo. If this argumentation is correct, a further use of cultivated rat hepatocytes in toxicogenomics would not make any sense. This was the basis for investigating the apparently occurring in vitro/in vivo discrepancy. The result of these investigations showed that this discrepancy is simply based on the different pharmacokinetics in vitro and in vivo. Methapyrilene has a half-life of 2.8 h. The previous studies focused on 24 h as the earliest time point for sampling, therefore the substance was already completely excreted and the gene expression change had already returned to basal level. By investigation of earlier timepoints, it could be shown, that an induction of Gsk3b and Myd116 also occurred in vivo . Due to this fact, the argument of a discrepancy between in vitro and in vivo is no longer valid. In summary, it has been shown that the culture system described in this thesis, is able to detect substanceinduced gene expression alterations at in vivo relevant concentrations. These changes in gene expression correlate well with the in vivo situation concerning the group of stress-associated genes.