Human introns contain conserved tissue-specific cryptic poison exons

Abstract

Eukaryotic cells express a large number of transcripts from a single gene due to alternative splicing. Despite hundreds of thousands of splice isoforms being annotated in databases, it has been reported that the current exon catalogs remain incomplete. At the same time, introns of human protein-coding (PC) genes contain a large number of evolutionarily conserved elements with unknown function. Here, we explore the possibility that some of them represent cryptic exons that are expressed in rare conditions. We identified a group of cryptic exons that are similar to the annotated exons in terms of evolutionary conservation and RNA-seq read coverage in the Genotype-Tissue Expression dataset. Most of them were poison, i.e. generated an nonsense-mediated decay (NMD) isoform upon inclusion, and many showed signs of tissue-specific and cancer-specific expression and regulation. We performed RNA-seq in A549 cell line treated with cycloheximide to inactivate NMD and confirmed using quantitative polymerase chain reaction that seven of eight exons tested are, indeed, expressed. This study shows that introns of human PC genes contain cryptic poison exons, which reside in conserved intronic regions and remain not fully annotated due to insufficient representation in RNA-seq libraries.

Figure 1.

Identification of cryptic exons. Transcript models built by StringTie for each RNA-seq experiment are parsed to identify cryptic cassette exons supported by splice junctions (red), read coverage (gray) and evolutionary conservation score (green). For each of these metrics, the 10th percentile of the distribution for annotated exons is used as a threshold. CDF denotes cumulative distribution function.

Figure 2.

Properties of cryptic exons. (A) Venn diagram of the set of cryptic exons passing the 10th percentile threshold by evolutionary conservation, split read support and read coverage support. (B) The size of the intersection in (A) as a function of the percentile. (C) The support of cryptic exons by the number of samples. (D) The support of cryptic exons by the number of tissues. (E) The number of exons in each tissue supported by a different number of samples. (F) The inclusion levels (Ψ) of exons from VastDB, CHESS and exons reported here (NEXON). (G) The average phastCons scores of exons from VastDB, CHESS and NEXON.

Figure 3.

Cryptic exons are mostly poison. (A) The proportion of poison, PC and non-coding exons with respect to the MANE Select transcript isoform of the gene. (B) The response (ΔΨ = Ψ_CHX − Ψ_CTL) of cryptic PEs and annotated PC cassette exons to NMD inhibition by CHX. (C) Same as in (B) but in Hela cell line (43) (top) and for ΔΨ = Ψ_EJC − Ψ_{RNA − seq} in the RIPiT experiment (44) (bottom). (D) The phyloP scores of annotated and cryptic PEs and PC exons in the first, second and the third codon position (phase 1, 2 and 3, respectively). (E) The average phastCons scores of annotated and cryptic poison and PC exons (top) and those of their respective flanking introns (bottom). (F) The splicing index I/(I + R), where I and R are the number of reads supporting intron splicing and retention, respectively, of introns flanking cryptic exons (NEXON) versus exons from other catalogs. Symbols **, ***, **** and ns. denote statistically discernible differences at the 1%, 0.1%, 0.01% significance level and not significant differences, respectively.

Figure 4.

Expression and regulation patterns of cryptic exons. (A) The pattern of splicing of cryptic exons across GTEx tissues. Colored squares represent Ψ value. Gray squares represent missing Ψ values. PEs and PC exons, as well as exons present in VastDB and CHESS databases are indicated in the leftmost columns. (B) The pattern of splicing changes (ΔΨ = Ψ_KD − Ψ_CTL) in RBP inactivation experiments. (C) The pattern of splicing changes (ΔΨ = Ψ_tumor − Ψ_normal) in paired samples from TCGA cohorts.

Figure 5.

RT-qPCR validation of cryptic PEs in NSD1 (A), PATL1 (B), PRRC2B (C), SENP7 (D), SPAG9 (E) and UBR5 (F) genes. Boxplots represent Ψ values across five biological replicates. The SJ track shows the number of split reads supporting splice junctions in NMD inactivation by CHX. Symbols * and ** denote statistically discernible differences at the 5% and 1% significance level. The diagram below shows the MANE Select PC isoform with exon numbers, the unproductive isoform (NEXON) and conservation scores (from 0 to 1) across 100 vertebrate species (Cons100) and 30 placental mammals (Cons30).