Repeat-Associated Non-Aug Translation Initiation at Expanded Ggggcc Repeats in c9orf72-Associated Amyotrophic Lateral Sclerosis and Frontotemporal Dementia

Abstract

A GGGGCC repeat expansion in C9orf72 causes amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). Through a process termed repeat-associated non-AUG (RAN) translation, the expanded repeat is translated in all three readings frame, in the absence of an AUG start codon, to produce dipeptide repeat-containing proteins (DPRs). DPRs accumulate in patient neurons and several are toxic in diverse disease models. However, the non-canonical translation initiation mechanism through which DPRs are synthesized is poorly understood. I developed luciferase-based GGGGCC RAN translation-specific reporters to quantitatively assess RAN translation levels in all three reading frames. Both in vitro in a rabbit reticulocyte lysate (RRL) and in cultured cells and neurons, I found that RAN translation of 70 GGGGCC repeats occurs most readily on mRNAs containing a functional 5′ mRNA cap, and through eIF4A-mediated ribosomal scanning. Additionally, GGGGCC RAN translation occurs most efficiently in the glycine-alanine (GA) reading frame, followed by the glycine-proline (GP) and glycine-arginine (GR) frames. This is due in part to use of a CUG initiation codon located upstream of the repeat in the GA frame, as mutating this codon greatly decreases polyGA production. Interestingly, unlike global translation, C9RAN translation is not inhibited by cellular stress. Instead, eIF2α phosphorylation during stress enhances C9RAN translation in a manner that depends upon its non-AUG initiation. These findings identify aspects of RAN translation that are mechanistically distinct from canonical translation, suggesting the possibility of developing RAN translation-selective inhibitors that do not impair global translation. To assess this possibility, I adapted our RRL in vitro RAN translation assay to perform a high-throughput screen of 3253 small molecules. From this screen, I identified five small molecules that more significantly inhibit RAN translation than AUG-initiated translation across multiple reading frames of the GGGGCC repeat, as well as the CGG repeat causative of fragile X-associated tremor/ataxia syndrome. I further showed that while three of these inhibitors directly interact with repeat RNAs, two do not, suggesting multiple mechanisms by which RAN translation can be selectively targeted. Additionally, as a potential mechanism by which C9RAN translation evades downregulation following eIF2α phosphorylation, I investigated the role of eIF2 alternatives in supporting C9RAN translation. eIF2A, eIF2D, and DENR/MCTS1 can function in place of eIF2 and deliver the initiator methionine tRNA to the pre-initiation ribosome. In HEK293 cells, DENR/MCTS1 knockdown more significantly reduces RAN translation than AUG-initiated translation, as does deletion of eIF2A in in vitro translation lysates. Furthermore, knockdown of each factor modestly reduces some repeat-associated toxicity in model organisms. Lastly, while C9RAN translation reporters are a powerful tool, it is critical to understand how RAN translation occurs on endogenous repeat containing RNAs in patient cells. To determine the C9orf72 sequence found upstream of the normally intronic GGGGCC repeat, I performed repeat-primed 5′ RNA ligation mediated rapid amplification of cDNA ends on actively translated RNA isolated from control and C9ALS/FTD patient-derived iNeurons following polysome profiling. This identified novel C9orf72 5′ ends that position the expanded GGGGCC repeat in a non-intronic context, providing evidence for one pathway through which the repeat becomes available to ribosomes. Together, these studies support a cap- and scanning-dependent model of C9RAN translation, that uses a CUG start codon, occurs on 5′ truncated C9orf72 transcripts, is enhanced by cellular stress, and is selectively inhibited with small molecules or genetic manipulations, and provide groundwork for future RAN translation-targeting therapeutic development.