Evaluation of microarray technology and cell line models in modern toxicology

Modern toxicology puts together existing knowledge of classical biology with new technologies to study effects of perturbations and to predict adverse outcomes resulting from those perturbations. Modern toxicology employs high throughput technology as microarray and cell lines as in vitro models to investigate the effect of potential toxic compounds. In my thesis I evaluate the usefulness and reproducibility of microarray technology and cell line models. I have been mainly interested in the study of the effects on the regulation of gene transcription in cell lines. I used HepG2 cell line as a model for liver to study cadmium-induced cytotoxicity, Caco-2 as model for intestinal epithelial to study nanoparticles and HeLa cells from cervical cancer to evaluate genomic instability. Cadmium is currently classified as carcinogen for human. In my first study, gene expression is used to explain the possible toxic mechanism of cadmium carcinogenity, using HepG2 cell line. A suggested mechanism of cadmium-induced carcinogenicity involves defects in the cell response to DNA damage and in the resistance to apoptosis. In this regard, I focused on the tumour suppressor protein P53, since its inactivation is a common feature found in human cancers and it is a crucial component of the cellular response to DNA damage. The results presented in this thesis demonstrate that in HepG2 cells exposed to cadmium, P53 was correctly moved and accumulated into the nucleus to accomplish its function as a transcription factor. However, in spite of this correct nuclear localization, the signals for the cell cycle arrest were not activated. In this context, the important mediator of cell cycle arrest P21, a P53 downstream protein, was upregulated at the gene level but not at the protein level. These results could be explained by the involvement of a post-transcriptional activity mediated by miRNA, as demonstrated by the upregulation of mir-372 in cadmium-treated HepG2 cells, which was able to affect p21 expression and to promote cell proliferation. Recent advances in materials science have resulted in the creation of particles in the nano-scale range and their use is spreading. This creates the need to evaluate their possible toxic effect on human health. In the second study, gene expression is applied to evaluate nanoparticle effect on Caco-2 cells as in vitro model of intestinal epithelial cells. The main focus was to compare the effect of gold nanoparticles (AuNP) of two sizes (30 nm and 5 nm), using microarray technology. Smaller AuNPs (5 nm) inhibit Caco-2 cell growth. Gene expression analysis shows a broad range of responses induced by small size AuNPs. A bioinformatics reconstruction of the possible pathways regulated by smaller AuNPs indicates that Caco-2 cells activate defence responses even if it was not enough to prevent the observed toxic effect. The response at the transcriptome level upon exposure to larger AuNPs (30 nm) is very weak. This effect can be due to the low uptake of these larger AuNPs. The third study focuses on the issue concerning cell line variability using different batches of HeLa cells obtained from different laboratories. The microarray expression analysis was performed on different batches of HeLa cells exposed to hypoxic conditions. In response to a hypoxic stimulus, each cell line batch activated different pathways, although the regulation of genes related to hypoxia is conserved. A genetic analysis show the high level of extensive chromosome instability in HeLa clones obtained from different laboratories. Each clone accumulates genomic variability in a time-dependent manner. The large differences in gene expression profiles suggest that the use of uncharacterized clones may lead to faulty conclusions and to irreproducible results in studies of gene function and pathway analysis. In the three papers presented in this thesis I have not only shown the reproducibility of microarray by comparison to real time PCR, the gold standard technique for mRNA quantification, but also the potential to create possible model for the mechanistic effects on biological pathways. In my thesis I have used different cell lines as in vitro models. They have shown that they have many advantages and that can give strong support for modern toxicology and be used to study biological mechanism. However, a possible drawback, which should be taken into account, is highlighted in the third study showing a high variability in the responses of different batches of the same cell line.