Fig. 1: Identification of cell-specific exons and SLED vector construction.

a Diagramatic sketch of SLED vector design strategy. SLED is compatible with any promoter. A frameshifting mutation is introduced into a cell-specific alternative exon to create two potential translational reading frames from an upstream start codon. In most SLED vectors, exon skipping will produce a red fluorescent protein while exon inclusion will shift the reading frame to produce a green fluorescent protein. b Ranking all neuron-enriched exons by the percent spliced-in (PSI) difference between neurons and other cell types. Exons were identified from mouse RNA-Seq datasets analyzed with the ASCOT pipeline²⁹. Approximately 1000 neuron-enriched exons have a ΔPSI greater than 20. c Among these top 1000 exons, approximately 200 candidates reside in introns < 2 kb in length. d, e UCSC genome browser views of the neuron-specific exon in Pls3 that is used in SLED.NPL. Exon incorporation is only observed in neuronal datasets (red arrows) from both mouse (d) and human (e). A similar strategy was used to identify the photoreceptor-specific exon in Atp1b2. These exons were used to generate SLED vectors that were then tested in HEK293, HepG2, and N2a cancer cell lines (f). As predicted, neither vector showed EGFP expression, indicating an absence of cell-specific exon incorporation, except when the neuron-specific SLED was transfected into N2a cells, reflecting the neuronal characteristics of N2a neuroblastoma cells. For ratio calculations in panel f, n = 100 cells except for HepG2 transfected with neuronal SLED (n = 68). Scale bars = 50 μm. Center lines show the medians; box limits indicate the 25th and 75th percentiles as determined by R software; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles, outliers are represented by dots. Source data are provided as a Source Data file.