Koskela M. The multisubunit translation elongation factor eEF1H in human tumours.

The dissertation elucidates whether the subunits of the eEF1B complex occur independently from complex in different human tumours. For the first time we performed a systematic analysis of the eEF1H subunits expression at mRNA and protein levels in the same samples of human cardioesophageal, lung and renal carcinomas. We have observed the lack of correlated changes in the expression level of all eEF1B subunits that suggests the presence of the individual, non-complexed subunits of eEF1B in human cardioesophageal and lung carcinomas. Cardioesophageal and lung carcinomas showed increased level of expression of all eEF1H subunits observed in 8-36% of all samples, depending on the subunit and tumour type. However, as the changes in the subunits expression are not coordinated, the elevated protein levels of one or another subunit expression were found in 72% of cardioesophageal and 52% of lung carcinomas. Interestingly, we did not observe marked difference in the eEF1H proteins expression in human renal carcinomas where the overexpression of the eEF1A and eEF1Bgamma proteins was found in two diverse tumour samples out from eighteen specimens researched. This finding favours the specificity of the phenomenon in human cardioesophageal and lung carcinomas, suggesting that the combined increase of different eEF1B subunits of the eEF1H complex may be considered as useful addition to the panel of markers for lung and cardioesophageal carcinomas. Subcellular localization of the eEF1H subunits was investigated in the cardioesophageal and lung cancer tissue samples and lung adenocarcinoma cell line A549. All eEF1H subunits were found in nuclei of A549 cells, whereas the only eEF1Bbeta and eEF1Bgamma nuclear localization was found in the lung and cardioesophageal cancer specimens by immunohistochemical investigation. This difference may be patient specific or represent a genuine divergence in the eEF1B localization between cultivating cells and cancer tissues. More importantly, the immunoprecipitation analysis did not reveal the classic eEF1H complex in nucleus, as the eEF1A-eEF1B interaction was not detected there despite the presence of all factors in the nuclear fraction. Molecular reasons for this are under investigation and may be related to a different level of post-translational modifications of the eEF1A or eEF1B components. Notably, the phosphorylation of eEF1Bbeta can influence its interaction with eEF1A. A reason for the presence of the eEF1H subunits in nucleus is unknown. The data on the existence of nuclear protein synthesis remain controversial. As eEF1A and eEF1B did not interact in the nuclear fraction of the cancer cells, this argues against nuclear translation and suggests independent roles of eEF1A and eEF1B.