A massively parallel reporter assay reveals focused and broadly encoded RNA localization signals in neurons

Abstract

Asymmetric subcellular localization of mRNA is a common cellular phenomenon that is thought to contribute to spatial gene regulation. In highly polar neurons, subcellular transcript localization and translation are thought to enhance cellular efficiency and timely responses to external cues. Although mRNA localization has been observed in many tissues and numerous examples of the functional importance of this process exist, we still lack a systematic understanding of how the transcript sorting machinery works in a sequence-specific manner. Here, we addressed these gaps by combining subcellular transcriptomics and rationally designed sequence libraries. We developed a massively parallel reporter assay (MPRA) for mRNA localization and tested ~50,000 sequences for their ability to drive RNA localization to neurites of neuronal cell lines. By scanning the 3'UTR of >300 genes we identified many previously unknown localization regions and mapped the localization potential of endogenous sequences. Our data suggest two ways the localization potential can be encoded in the 3'UTR: focused localization motifs and broadly encoded localization potential based on small contributions. We identified sequence motifs enriched in dendritically localized transcripts and tested the potential of these motifs to affect the localization behavior of an mRNA. This assay revealed sequence elements with the ability to bias localization towards neurite as well as soma. Depletion of RNA binding proteins predicted or experimentally shown to bind these motifs abolished the effect on localization, suggesting that these motifs act by recruiting specific RNA-binding proteins. Based on our dataset we developed machine learning models that accurately predict the localization behavior of novel sequences. Testing this predictor on native mRNA sequencing data showed good agreement between predicted and observed localization potential, suggesting that the rules uncovered by our MPRA also apply to the localization of native transcripts. Applying similar systematic high-throughput approaches to other cell types will open the door for a comparative perspective on RNA localization across tissues and reveal the commonalities and differences of this crucial regulatory mechanism.

Mammalian Neuronal mRNA Transport Complexes: The Few Knowns and the Many Unknowns

“…As recent evidence suggests that 3'UTRs often contain multiple regions that trigger localization to neurites (Mikl et al, 2021), it is possible that a single mRNA is transported by more than one motor. If this was the case, an mRNA could potentially recruit different numbers of adaptormotor complexes based on the amount of localization motifs encoded in its 3'UTR.…”

RNA localization mechanisms transcend cell morphology

“…For a given RNA, much of what is known about the regulation of its localization, especially in terms of sequence motifs and the RBPs that bind them, has been elucidated in a single cell type. Many examples of the molecular rules that govern RNA transport to neuronal projections (Arora et al, 2021; Goering et al, 2020; von Kügelgen et al, 2021; Mikl et al, 2021; Zappulo et al, 2017) or across the posterior/anterior axis of the Drosophila embryos (Ephrussi and Lehmann, 1992; Hachet and Ephrussi, 2004; Jambor et al, 2014; Lécuyer et al, 2007; Lipshitz and Smibert, 2000) have been defined. However, how these RBPs and RNA elements would regulate RNA localization in cell types or morphologies outside of those that they were discovered in is generally unknown.…”

The Role of Alternative Polyadenylation in the Regulation of Subcellular RNA Localization

“…For example, one such RNAi screen surveyed 489 RBPs to identify regulators of PAS usage (Jenal et al, 2012). Similarly, massively parallel reporter assays have recently been used to identify sequence elements that regulate RNA localization (Arora et al, 2021b;Mikl et al, 2021;von Kügelgen et al, 2021). The synthesis of these large scale datasets has the potential to identify more cases in which the regulation of APA and RNA localization are explicitly linked.…”