Biochemical and cellular characterization of the coding region determinant-binding protein (CRD-BP)-mRNA interaction.

RNA-binding proteins play critical role in the post-transcriptional processing of mRNAs. One such RNA binding protein is termed as the-Coding Region Determinant Binding Protein (CRD-BP). CRD-BP is an onco-fetal protein whose overexpression has been reported in various types of human cancers including, breast, colon, liver, skin, ovary, lung, brain, chorion, and testicular cancers. CRD-BP is a member of VICKZ family of RNA-binding proteins. In many instances, RNA binding leads to stabilization of the transcripts and an increase in their corresponding protein levels; the result is manifest in downstream effects and the cancer phenotype. The primary goal of this study was to obtain a better understanding of the interaction between CRD-BP and its three target mRNAs: GLI1, MDR1 and CD44. Radiolabelled electrophoretic mobility shift assay (EMSA) was performed with [³²P]-labeled truncated GLI1 and MDR1 RNAs to find smaller region of the transcripts which can still bind CRD-BP. It was found that GLI1 320-380 RNA is the minimum region required for binding CRD-BP, while MDR1 779-881 RNA is the minimum region which still has high affinity for CRD-BP. Previous deletion studies of CRD-BP orthologs revealed that the KH domains, and not the RRM domains, are critical for binding RNA substrates. However, it was unclear as to what extent each KH domain plays nor is it known if different KH domains are important in binding different RNAs. In this study, I used site-directed mutagenesis to mutate the GXXG to DXXG in each of the KH domains as an approach to investigate the role of each KH domains, in the context of the entire protein, in binding GLI1 and MDR1 RNAs. The K[subscript]d values of all the single and double KH variants that are capable of binding to GLI1 and CD44 RNAs was determined. In general, it was found that single mutation in KH domains may or may not affect the binding affinity of transcript, while mutations at any two KH domains totally abrogated the binding of RNA to CRD-BP, with the exception of KH3-4 which binds CD44 RNA but not GLI1 RNA. This finding also supports the hypothesis that KH domains generally work in tandem. The result also clearly showed that different RNAs bind CRD-BP differently in vitro. The secondary goal of my thesis was to design RNA oligonucleotides capable of breaking the specific CRD-BP-GLI1 RNA interaction in vitro and in cells. For in vitro studies, competition studies using [³²P] labelled GLI1 230-420 RNA and MFOLD designed RNA/DNA oligonucleotides were utilized. Amongst eight RNA oligonucleotides and one DNA oligonucleotide, S1 RNA was the best competitor against [³²P] labelled GLI 230-420 RNA in vitro. The T47D human breast cancer cell and HCT116 human colorectal cancer cell which expressed detectable level of GLI1 mRNA were chosen for further studies with the S1 RNA oligonucleotide. In both T47D and HCT116 cells where CRD-BP-GLI1 mRNA interaction was demonstrated to exists, S1 RNA oligonucleotide significantly and specifically down-regulate GLI1 mRNA expression. The results obtained support the model that CRD-BP protects GLI1 mRNA from degradation in T47D and HCT116 cells, and suggest that breaking CRD-BP-GLI1 mRNA interaction is a feasible approach to inhibit GLI1 expression. In summary, this work shows that different mRNAs indeed bind to CRD-BP differently and it is feasible to design/discover molecules capable of breaking specific CRD-BP-RNA interaction in vitro. Most importantly, molecule that breaks CRD-BP-RNA interaction in vitro was also capable of down-regulating specific the mRNA in cells. This work has provided further evidence to support the development of a new class of anti-cancer drugs that act by breaking specific protein-RNA interaction.