Background

RNA-binding proteins (RBPs) such as TDP43 regulate post-transcriptional gene regulation by orchestrating RNA stability, transport and processing, including mRNA splicing (Gerstberger et al., 2014). Impairment of RBP function as a result of mutations and/or aggregation has been implicated in the etiology of many neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) (Harrison and Shorter, 2017). Notably, ALS-associated mutations in RBPs like TDP43 can cause transcriptome-wide splicing defects, suggesting that misregulation of RBP splicing function may be a key driver of disease (Arnold et al., 2013; Sun et al., 2015). Thus, an important goal in ALS research is to uncover the molecular mechanisms of RBP function, requiring experimental approaches for the systematic interrogation of RBP splicing function.

Widely-used methods for studying splicing are reverse transcription-PCR (RT-PCR) and RNA sequencing (RNA-seq). Both methods can be used to study the splicing of endogenous transcripts; however, typically only as a bulk measurement of an entire population of cells. RT-PCR is simple and affordable, yet efficient only to study splicing of a select few endogenous transcripts. Moreover, RT-PCR read-outs are often only semi-quantitative. In contrast, RNA-seq provides quantitative insights into global splicing, but is time- and cost-intensive and requires sophisticated data analysis pipelines (Wang et al., 2009). Thus, both RT-PCR and RNA-seq are poorly suited for high-throughput applications, for example to systematically compare the splicing function of multiple RBP variants.

A powerful tool to study splicing are minigenes, which typically are plasmid-based, simplified genes containing introns and exons that recapitulate defined splicing events outside of the native gene context (Baralle and Baralle, 2005). In particular, minigenes have been instrumental in discovering cis- and trans-acting splicing elements (Cooper, 2005). Instead of endogenous targets, minigenes are also routinely used to streamline investigation of RBP splicing function (D'Ambrogio et al., 2009; Kino et al., 2011). Conventionally, the splicing of minigenes is evaluated with analytical or quantitative RT-PCR. However, the low-throughput and bulk read-out of this method precludes leveraging the full potential of minigene-based splicing reporters.

To overcome these hurdles, we developed a minigene that encodes for a fluorescent protein (FP minigene), thus allowing us to convert a defined splicing event into a sensitive, inherently quantitative and readily detectable signal on a single-cell level by flow cytometry (Schmidt et al., 2019). To normalize splicing of the FP minigene to the overall abundance of the parent transcript, we fused it to another, intron-free fluorescent protein (constitutive FP). While this approach is broadly applicable (Gurskaya et al., 2012; Sorenson and Stevens, 2014; Gonatopoulos-Pournatzis et al., 2018), we adapted it specifically to study TDP43-dependent splicing by interrupting the FP minigene with exon 9 of the CFTR gene, which is skipped only in the presence of functional TDP43 (Buratti et al., 2001). To rapidly compare the splicing efficiency of multiple TDP43 variants in TARDBP (i.e., the gene encoding TDP43) knock-out cells, we used a bidirectional promoter to drive the coordinated expression of both our splicing reporter and a given BFP-tagged TDP43 variant from the same plasmid. A graphical overview of our splicing reporter design in shown in Figure 1.