Molecular requirements for FMRP and RNA helicase MOV10 in the translational regulation of co-bound mRNAs

The fragile X mental retardation protein (FMRP), regulates translation of its bound mRNAs through an incompletely defined mechanism. Historically, FMRP has been known to directly associate with Argonaute (AGO), a key effector protein of the RNA induced silencing complex (RISC). MicroRNAs regulate key cellular processes in mammalian genomes by post-transcriptionally regulating gene expression. However, FMRP's role in microRNA-mediated translational regulation has remained unclear. In this work, we studied the interaction of FMRP with RNA helicase MOV10. The purpose is to gain insight into how the formation of the FMRP-MOV10 RiboNucleoProtein (RNP) complex can facilitate or inhibit microRNA-mediated translational regulation. In chapter 2 of this work, we identified the mRNA targets of MOV10 by individual nucleotide Cross-linking Immunoprecipitation (iCLIP) as well as identified FMRP-MOV10 co-bound targets by comparison of this data with published FMRP CLIP data. We then showed that FMRP recruits MOV10 to a subset of mRNAs, presumably to resolve RNA secondary structure and facilitate AGO accessibility to MRE sites. In support of this hypothesis, transcriptome analysis by RNA-seq revealed an increase in mRNA abundance of MOV10 targets RNAs as well as in FMRP-MOV10 co-bound mRNAs in MOV10 knockdown (KD) HEK293 cells. We then strengthened our hypothesis that FMRP functions in microRNA-mediated translational regulation by comparing CLIP binding sites of FMRP, MOV10, and AGO2 in the 3' UTR of co-bound mRNAs. All three proteins exhibit an enrichment of binding proximal to commonly known microRNA Recognition Elements (MRE) sites in HEK293 cells. However, we also observed that in a subset of mRNAs, FMRP and MOV10 binding at G-Quadruplex structures containing an embedded MRE site suppressed AGO2mediated translational regulation. We concluded that FMRP recruits MOV10 to a subset of mRNAs, which usually leads to regulation by AGO; however, in a subset of mRNAs where the FMRP-MOV10 complex binds on a G-Quadruplex, AGO association is blocked. In chapter 3, I studied the dynamic interaction of the FMRP-MOV10 RNP complex and its effect on the translational fate of mRNA. To determine how the FMRP-MOV10 complex protected G-Quadruplex-embedded MREs from AGO2 association, I mapped the interactive domains of the three proteins. I found that the N-terminus of MOV10 directly interacted with the KH1 domain of FMRP, strengthening recent data proposing that FMRP's KH1 domain is capable of protein-protein interaction. I also found that the N-terminus of FMRP directly interacted with AGO. As predicted in our earlier work, I showed that MOV10 resolved G-Quadruplex structures using an RNA unwinding assay that I developed. I then showed that FMRP globally facilitates AGO2 binding to all regions of target mRNAs in P0 mouse brain via eCLIP. The number of AGO2 binding sites on mRNAs was reduced by approximately 75% in the absence of FMRP. Lastly, I was able to determine the mechanism by which the FMRP-MOV10 complex at G-Quadruplexes acted to inhibit AGO2 association with a subset of mRNAs. This characteristic is modulated through FMRP's RGG box, which increases affinity for a G-Quadruplex through FMRP's interaction with the N-terminus of MOV10. Thus, the N-terminus of MOV10 has a function independent of its helicase activity and is required for neurite outgrowth in Neuro2A cells (N2a)