U2AF2 binds IL7R exon 6 ectopically and represses its inclusion

Abstract

Interleukin 7 receptor α-chain is crucial for the development and maintenance of T cells and is genetically associated with autoimmune disorders including multiple sclerosis (MS), a demyelinating disease of the CNS. Exon 6 of IL7R encodes for the transmembrane domain of the receptor and is regulated by alternative splicing: inclusion or skipping of IL7R exon 6 results in membrane-bound or soluble IL7R isoforms, respectively. We previously identified a SNP (rs6897932) in IL7R exon 6, strongly associated with MS risk and showed that the risk allele (C) increases skipping of the exon, resulting in elevated levels of sIL7R. This has important pathological consequences as elevated levels of sIL7R has been shown to exacerbate the disease in the experimental autoimmune encephalomyelitis mouse model of MS. Understanding the regulation of exon 6 splicing provides important mechanistic insights into the pathogenesis of MS. Here we report two mechanisms by which IL7R exon 6 is controlled. First, a competition between PTBP1 and U2AF2 at the polypyrimidine tract (PPT) of intron 5, and second, an unexpected U2AF2-mediated assembly of spicing factors in the exon. We noted the presence of a branchpoint sequence (BPS) (TACTAAT or TACTAAC) within exon 6, which is stronger with the C allele. We also noted that the BPS is followed by a PPT and conjectured that silencing could be mediated by the binding of U2AF2 to that tract. In support of this model, we show that evolutionary conservation of the exonic PPT correlates well with the degree of alternative splicing of exon 6 in two non-human primate species and that U2AF2 binding to this PPT recruits U2 snRNP components to the exon. These observations provide the first explanation for the stronger silencing of IL7R exon 6 with the disease associated C allele at rs6897932.

Keywords: Alternative splicing; IL7R; U2AF2; multiple sclerosis.

FIGURE 1.

Exon 6 of IL7R contains elements that resemble a 3′ splice site. (A) IL7R RNA sequence spanning the last 30 nt of intron 5, expanded sequences within exon 6, and the first 2 nt of intron 6. Intron sequences are shown in lower case letters and exon sequences in upper case letters. The first sequence (a) shows the consensus sequence with the location of the nonsynonymous SNP rs6897932C > T in exon 6 indicated in red (Y = pyrimidine). SVM-BPfinder was used to predict the location and scores of potential branch point sequences (BPS) and polypyrimidine tracts (PPT) within intron 5 and exon 6. The predicted BPS in intron 5 and exon 6 (exBPS) are indicated in black bolded letters, whereas the predicted intronic (PPT) and exonic (exPPT) PPTs with corresponding scores are indicated in blue. Sequences b and c show sequences containing either the risk C allele (b) or the protective T allele (c) of the SNP rs6897932. Sequences d–g show mutations or deletions introduced around the SNP, which disrupt the exBPS. ND (not detected) indicates no BPS was detected. (B) Schematics of IL7R exon 6 minigene. The locations of T7 and SP6 vector-specific primers are indicated by black arrows. Sequences in blue of the wild-type (wt) PPT within intron 5 and two mutant versions (mt1 and mt2) are shown below. The mutated residues are underlined. The PPT score for each sequence is displayed on the right. (C) RT-PCR analysis of exon 6 splicing in transcripts from the IL7R minigenes transfected into HeLa cells (+E6, exon 6 included; −E6, exon 6 skipped). (D) Binding of U2AF2 and PTBP1 to reporter RNAs in RNA affinity chromatography (RAC) assays. The abundance of U2AF2 and PTBP1 pulled down with each RNA was determined by western blot (WB). Shown here is the ratio of U2AF2 to PTBP1 enrichment for mt1 and mt2 IL7R RNAs relative to the wt RNA (see Materials and Methods). (****) P < 0.0001.

FIGURE 2.

Nonconserved nucleotides within the exonic polypyrimidine-tract of exon 6 dictate its splicing. (A) Alignment of IL7R exon 6 (nt 1–76). Nucleotides in red indicate the differences observed between the different sequences; bolded red nt are located within the exPPT, regular red nt represent differences in other locations of the exon. Bolded nucleotides in blue indicate where the predicted exPPTs are located (nt 41–61 in human and marmoset; nt 32–37 in macaque). The svm-score was predicted by the SVM-BPfinder software. (B) RT-PCR analysis of exon 6 splicing in transcripts from PBMCs from humans stratified by the genotypes of rs6897932 (human CC [11 subjects], human CT [13 subjects], human TT [10 subjects]), rhesus macaques (10 animals), and marmosets (10 animals). Percent exon 6 skipping was calculated, and the blue line indicates the median value. Statistical significance was determined by Mann–Whitney test. (C) Sequences displayed indicate the region between nucleotides 33 and 61 in IL7R exon 6 where mutations were performed with mutated nucleotides indicated in red in the GFP marmoset-IL7R minigene. (D) RT-PCR analysis of exon 6 splicing in transcripts from the GFP marmoset-IL7R minigenes. Nucleotides 33, 42, 47, 53, and 55 of exon 6 were mutated to those of macaque or human. Percent exon 6 skipping for WT or mutated reporters (mean ± S.D., n = 3) was quantified as before. (*) P < 0.05, (**) P < 0.01, (***) P < 0.001, and (****) P < 0.0001.

FIGURE 3.

U2AF2 binds the polypyrimidine tract in exon 6 and represses the exon. (A) Schematics of mutations to polypyrimidine stretch (exPPT) within exon 6 introduced in the IL7R exon 6 minigene. The mutated residues are underlined. For each sequence, the exPPT score is indicated. (B) RT-PCR analysis of exon 6 splicing in transcripts from the wt or mutant IL7R minigenes in A. Percent exon 6 skipping was quantified and is shown under the gels as mean ± S.D. (n = 3). (C) Schematics of RNA baits used for RNA affinity chromatography with HeLa nuclear extracts. The black bar in exon 6 indicates the location of exPPT mutations described in A. (D) Representative WB measuring abundance of U2AF2 in HeLa NE input, or associated with matrix only, aptamer only, or wt, ex6 mt1, ex6 mt2 or ex6 mt3 IL7R RNAs in RNA affinity chromatography. (E) Fold-enrichment of U2AF2 pulled down with each RNA bait is shown relative to the wild-type bait (see Materials and Methods). Enrichment is shown as mean from three independent experiments. (*) P < 0.05, (**) P < 0.01, and (****) P < 0.0001.

FIGURE 4.

U2AF2 recruits U2 snRNP to IL7R exon 6. (A) Schematics of the different RNA baits used for RAC. The black bar in exon 6 indicates the location of the exPPT mutations, and their sequence is shown on the right for each construct. (B) Representative WB illustrating abundance of U2AF2 in HeLa NE, or associated with the matrix, the aptamer RNA, or the different IL7R RNA baits: exon 6 and flanking introns (ex6 + introns), exon 6 only wild-type (ex6 wt), and exon 6 only mutant (ex6 mt4); quantification of U2AF2 pulled down is shown as mean from three experiments in D. (C) Representative WB illustrating abundance of U2AF2 and SF3B1 in HeLa NE, or associated with the aptamer RNA, or IL7R exon 6 RNA after 30, 45, or 60 min incubation at 30°C; quantification of U2AF2 or SF3B1 pulled down is shown as mean from three independent experiments in E (U2AF2) and F (SF3B1). (*) P < 0.05 and (**) P < 0.01.

FIGURE 5.

Exonic assembly of U2AF2-driven splicing complexes silences IL7R exon 6. (A) U2AF2 binds to the polypyrimidine tract in exon 6 creating a negative complex on the exon leading to exon 6 skipping. (B) The exonic BPS recruits U2 snRNP. This interaction is more stable when the risk allele C of rs6897932 is present which leads to more exon 6 skipping. (C) DDX39B remodels U2AF2-recruited U2 snRNP complexes.