A dominant negative over expression model of mammalian MED12 function

Hans Levi Packer

doi:10.17077/etd.uvcq82no

Although schizophrenia has been shown to have a substantial component, there is a paucity of known risk alleles. Furthermore, all of the known risk genes are of small effect sizes. Previously it has been shown that a 12 base pair insertional polymorphism in the in the C-terminal, opposite paired (Opa) domain of the MED12 gene known as MED1212bp represents a small but significant risk for a positive syndrome psychosis. In addition, the MED1212bp polymorphism is found in approximately 1.6 percent of X-chromosomes of northern European decent. Studies in zebrafish show that alterations in MED12 reduces staining for monoaminergic neuronal populations, including dopaminergic and serotonergic populations. However, precise mechanisms through which these changes occur are not known. My goal for this study was to use PC6-3 cells as a mammalian cell culture for studying cellular and transcriptional effects of MED12 in a dopaminergic model system. The approach I took was based on studies that have shown that overexpression of C-terminal proline, glutamine and leucine rich (PQL) and Opa domain constructs interact in a dominant negative manner with several transcriptional regulatory proteins that interact with MED12. GFP tagged PQL and Opa domain constructs were placed into a tetracycline inducible T-REx™ regulated expression vector and introduced into a previously generated PC6-3, TR156 cell line that expresses the Tet-Repressor molecule. In this study, I report a selection bias against stably transfected cell lines strongly expressing constructs containing the two C-terminal PQL-Opa protein domains of MED12. I also show that the described low levels of induction of that construct are associated with small, but significant alterations in nuclear morphology, possibly due to nuclear reorganization. Induction of PQL-Opa domains also increases in cell metabolism as measured by a tetrazolium salt assay, typically associated with increases in proliferation compared to the GFP controls or Opa domain alone. Interestingly, the MTS results in the stable cell lines were not reflected changes in cell numbers from direct cell counts performed by light microscopy, or changes in cell cycle distribution as measured by propidium iodide staining and fluorescence activated cell sorting (FACS). In addition I also show microarray gene expression data for both the stable tetracycline inducible lines, as well as transiently electroporated PC6-3 cells. For both the stable and transient expression experiments, the arrays were characterized by small fold changes, which were not validated by RT-PCR. The stable arrays did not produce any robust findings. However, gene ontology (GO) data, as determined by GoMiner analysis, from the transiently electroporated cells shows that 9 of the top 31 GO categories are related to changes in proliferation and cytoskeletal reorganization. However, despite this trend, the data from the GoMiner analysis was above the level of statistical significance (á = 0.05), as is indicated by the false discovery rates (FDR > 0.3). Analysis of the directionality of expression proved intriguing and demonstrated significant evidence of skewing in the pattern of differential expression of annotated genes where there was a significant tendency for the most significantly differentially expressed probes belonging to 13568 annotated genes to be more highly expressed genes in the electroporated GFP control construct cells than those with the PQL/Opa construct. This is also consistent with a broad overlap of the expression data with ChIP-seq data suggesting that the dominant negative effect may be spread over many MED12 regulated genes, in which case the low expression levels are particularly problematic. While the data from these experiments do not present a clear mechanism for MED12 function, they are informative in developing models of MED12 alteration, and potential improvements are discussed.

A dominant negative over expression model of mammalian MED12 function

Files and links (1)

Abstract

Details

Metrics