Nature Genetics: doi: /ng Supplementary Figure 1. Alternative splicing events in the 5K panel.

Transcription

1 Supplementary Figure 1 Alternative splicing events in the 5K panel. The majority of cryptic splicing occurred by creation of an AG or GT (Type I). While some other mutations increased the usage of a nearby weaker splice-site (type II). Very few mutations were found to abolish alternative splice-site usage (type III).

2 Supplementary Figure 2 MaPSy performance. (a d) Agreement between allelic splicing ratios (log 2 ) of three cell culture replicates of MaPSy in vivo (a c) and two experimental replicates of MaPSy in vitro (d). (e) Stacked histogram of mutant (red) and wild-type (blue) relative splicing efficiency in MaPSy in vivo (top) and in vitro (bottom). (f) Full gel of output (spliced species) from MaPSy in vivo.

3 Supplementary Figure 3 MaPSy validation in patient samples and ENCODE data. (a f) MaPSy s identified ESMs in mutations causing inosine triphosphatase deficiency (a), galactosemia (b), haemorrhagic telangiectasia (c), Menkes syndrome (d) and Barth syndrome (e,f) were shown to exhibit splicing aberrations (exon skipping and/or intron retention) in RNAs derived from patient tissue samples. (g) Splicing efficiency in MaPSy corresponds to splicing in ENCODE data.

4 Supplementary Figure 4 Mode of inheritance in the 5K panel. (a) Percent ESM in the 5K panel stratified by modes of inheritance in haploinsufficient genes (prediction score = 1), haplosufficient genes (prediction score < 0.7) and moderately haploinsufficient genes (1 > prediction score 0.7) 8. Error bars, 95% confidence intervals. (b) Number of mutations in the different modes of inheritance in the 5K panel.

5 Supplementary Figure 5 Genes intolerant to protein-truncating variants (PTVs) in the ExAC population are predisposed to disease-associated splicing mutations. (a) Mean fraction of ESMs in PTV-intolerant (pli 0.9), semitolerant (0.1 < pli < 0.9) and tolerant (pli 0.1) genes in dominant and recessive traits. Error bars, s.e.m. (b) PTV-intolerant genes also have more introns than other genes, similar to disease genes that lose function via splicing mutations.

6 Supplementary Figure 6 Features of splicing. (a) The mean of relative splicing efficiency of wild-type species in vivo (n = 2,086) is plotted against increasing mean of feature measures in sliding window (size = 200, step = 1). Shaded regions represent 95% confidence intervals. Intron length is plotted on a log 10 scale. The mean of PhastCons score for all bases of the exon was used to measure conservation. Genomic features that have previously been associated with splicing are shown to display similar trends in MaPSy. P values were obtained from linear regression analyses. (b) The 5K panel is divided into five bins of increasing feature measures, and percent ESM in each bin is plotted. Error bars, 95% confidence intervals. Low differential GC content between exon and intron, less ESE, more ESS and less agreement with splicesite consensus sequence, which are all associated with weaker splicing are shown to sensitize exons to ESM. The Kruskal Wallis test was used to obtain P values.

7 Supplementary Figure 7 The role of PTBP1 and SRSF1 in ESM phenotypes. (a) The splicing phenotype of a mutation in exon 20 of COL1A2 that creates a PTBP1-binding motif was partially rescued when PTBP1 was knocked down. (b) A mutation that weaken a SRSF1-binding motif in exon 8 of MLH1 caused a modest but not significant increase of skipping events in the absence of SRSF1, whereas the wild-type exon that contains a SRSF1-binding site had a significant increase in skipping events when SRSF1 was knocked down.

8 Supplementary Figure 8 Overlap of intronic and exonic splicing regulatory motifs. (a) The density for each RBP motif was calculated in all wild-type species (n = 2,048). (b) Clustering of intronic data reveals similar trends in vivo and in vitro. (c) Intronic splicing activators and exonic splicing repressors show a high degree of overlap. (d) Intronic splicing repressor motifs and exonic splicing activator motifs display a high degree of overlap. (e) Table of exonic splicing repressors and exonic splicing activators that exhibit the same function in vivo and in vitro.

9 Supplementary Figure 9 In vitro functional SELEX. (a) Series of functional SELEX with MaPSy. (b) Mutant/wild type ratio in the B/C fraction in comparison to spliced species (left) and in the A fraction in comparison to spliced species (right). Enrichment in B/C complex is positively correlated with splicing, while enrichment in A complex is negatively correlated with splicing. (c) Clustering the effects of exonic mutation disruptions on different stages of spliceosomal assembly revealed mechanistic signatures of ESM. Only clusters with 8 members are shown.

10 Supplementary Figure 10 Mutant feature analyses in different clusters revealed distinct ESM mechanistic signatures. Horizontal dotted lines indicate the mean value of the features in the 5K panel. Box plots of feature values that are significantly different than background (permuted cluster assignment) are colored red. The medians are indicated as horizontal bold lines, and the means as black hollow dots.

11 Supplementary Figure 11 ESM visualization browser. A web browser was developed to visualize raw counts and information on individual mutations from original publications. Mutations can be queried by HGMD ID, gene or author.

12 A b Supplementary Figure 12 Common sequences of the 5K panel reporters. (a) In vivo reporter sequence: CMV enhancer and promoter sequence (blue), adenovirus sequence (green; exon in uppercase and intron in lowercase), 200-mer library (red), ACTN1 intron 15 (lowercase, purple) and exon 16 (uppercase, purple), bgh poly(a) (cyan). (b) In vitro reporter sequence: adenovirus sequence (green; including T7 promoter sequence in bold), 200-mer oligo library (red) and additional intronic sequence (purple).

13 Supplementary Table 2. Summary of MaPSy validation in patient samples HGMD or SNP id gene hg19 position nucleotide change AA change MaPSy result validation result Patient tissue/ RNA source class CM ATM chr11: :+ c.6095g>a p.r2032k loss of splicing in vivo (0 fold) and in vitro (0 fold) exon 44 skipping lymphoblastoid cell line TP loss of splicing in vivo (0 fold) and in vitro (0 fold), but 46 CM ATM chr11: :+ c.5932g>t p.e1978* exon 42 skipping lymphoblastoid cell line FN neither were significant due to low counts CM ATP7A chrx: :+ c.1933c>t p.r645* loss of splicing in vivo (0.4 fold) and in vitro (0.4 fold) exon 8 skipping fibroblasts TP Coriell CM BRCA1 chr17: :- c.211a>g p.r71g loss of splicing in vivo (0 fold) and in vitro (0.5 fold) exon 5 skipping T-lymphocytes TP CM BRCA1 chr17: :- c.5123c>a p.a1708e loss of splicing in vivo (0.3 fold) and in vitro (0.3 fold) exon 18 skipping T-lymphocytes TP CM BRCA1 chr17: :- c.5080g>t p.e1694* loss of splicing in vivo (0.03 fold) and in vitro (0.1 fold) exon 18 skipping T-lymphocytes TP Reference / source CM BRCA1 chr17: :- c.131g>a p.c44y no change no change lymphoblastoid cell line TN Amanda Toland CM BRCA2 chr13: :+ c.673a>g p.t225a no change no change T-lymphocytes TN CM BRCA2 chr13: :+ c.7007g>a p.r2336h loss of splicing in vivo (0.03 fold) and in vitro (0.03 fold) exon 13 skipping T-lymphocytes TP CM COL1A1 chr9: :- c.608g>t p.g203v no change no change fibroblast TN Joan Marini CM COL1A1 chr9: :- c.634g>a p.g212r Loss of splicing in vivo (0.1 fold) and in vitro (0.2 fold) no change fibroblast FP Joan Marini CM ENG chr9: :- c.1687g>t p.e563* loss of splicing in vivo (0.38 fold) and in vitro (0.2 fold) loss of exon 13 T-lymphocytes TP Authors CM FBN1 chr15: :- c.6339t>g p.y2113* loss of splicing in vivo (0.04 fold) and in vitro (0.2 fold) exon 51 skipping fibroblasts TP CM FBN1 chr15: :- c.4766g>t p.c1589f loss of splicing in vivo (0.3 fold) and in vitro (0.6 fold) no change fibroblasts FP Coriell CM GALT chr9: :+ c.563a>g p.q188r loss of splicing in vivo (0.01 fold) and in vitro (0.4 fold) missplicing lymphoblastoid cell line TP Coriell CM HPRT1 chrx: :+ c.590a>t p.e197v loss of splicing in vivo (0 fold) and in vitro (0.3 fold) exon 8 skipping lymphoblastoid cell line TP CM HPRT1 chrx: :+ c.602a>t p.d201v loss of splicing in vivo (0.2 fold) and in vitro (0.3 fold) exon 8 skipping T-lymphocytes TP CM HPRT1 chrx: :+ c.569g>a p.g190e no change no change lymphoblastoid cell line TN CM HPRT1 chrx: :+ c.601g>a p.d201n no change no change lymphoblastoid cell line TN CM HPRT1 chrx: :+ c.601g>t p.d201y loss of splicing in vivo (0.18 fold) but not in vitro no change CM ITPA chr20: :+ c.94c>a p.p32t loss of splicing in vivo (0.1 fold) and in vitro (0.3 fold) exon 3 skipping lymphoblastoid cell line and T lymphocytes lymphoblastoid cell line and T-lymphocytes CM MLH1 chr3: :+ c.677g>a p.r226q loss of splicing in vivo (0 fold) and in vitro (0.01 fold) loss of exon 8 whole blood TP CM MLH1 chr3: :+ c.350c>t p.t117m no change no change lymphoblastoid cell line TN CM MLH1 chr3: :+ c.338t>a p.v113d loss of splicing in vivo (0.6 fold) but no change in vitro no change lymphoblastoid cell line TN CM MLH1 chr3: :+ c.637g>a p.v213m loss of splicing in vivo (0.6 fold) and in vitro (0.65 fold) no change lymphoblastoid cell line FP CM PSEN1 chr14: :+ c.548g>t p.g1883v loss of splicing in vivo (0.3 fold) and in vitro (0.1 fold) brain-specific loss of exons 6 and 7 frontal cortex (post mortem) TP CM PSEN1 chr14: :+ c.871a>c p.t291p no change exon 9 skipping Whole blood FN CM PSEN1 chr14: :+ c.509c>t ps170f no change in vivo but loss of splicing in vitro (0.5 fold) no change frontal cortex and cerebellum (post mortem) TN rs RBM23 chr14: :- c.408g>a p.r136 loss of splicing in vivo (0.3 fold) and in vitro (0.5 fold) exon 6 skipping lymphoblastoid cell line TP CM SLC12A3 chr16: :+ c.433c>t p.r145c loss of splicing in vivo (0.4 fold) and in vitro (0.6 fold) but neither were significant CM TAZ chrx: :+ c.497t>a p.l166* loss of splicing in vivo (0.3 fold) and in vitro (0.5 fold) TN exon 7 skipping whole blood FN exon 6 and 7 skipping TP Coriell Alison Goate lymphoblastoid cell line TP Coriell CM TAZ chrx: :+ c.367c>t p.r123* loss of splicing in vivo (0.6 fold) and in vitro (0.4 fold) intron retention lymphoblastoid cell line TP Coriell 57 58

14 Supplementary Table 4. ENCODE Accession numbers used for the purpose of validating MaPSy results ENCODE Accession Number SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR SRR545700

15 References 46. Teraoka, S.N. et al. Splicing defects in the ataxia- telangiectasia gene, ATM: underlying mutations and consequences. American Journal of Human Genetics 64, (1999). 47. Sanz, D.J. et al. A high proportion of DNA variants of BRCA1 and BRCA2 is associated with aberrant splicing in breast/ovarian cancer patients. Clin Cancer Res 16, (2010). 48. Mazoyer, S. et al. A BRCA1 nonsense mutation causes exon skipping. Am J Hum Genet 62, (1998). 49. Sharp, A., Pichert, G., Lucassen, A. & Eccles, D. RNA analysis reveals splicing mutations and loss of expression defects in MLH1 and BRCA1. Hum Mutat 24, 272 (2004). 50. Dietz, H.C. et al. The skipping of constitutive exons in vivo induced by nonsense mutations. Science 259, (1993). 51. Tu, M., Tong, W., Perkins, R. & Valentine, C.R. Predicted changes in pre- mrna secondary structure vary in their association with exon skipping for mutations in exons 2, 4, and 8 of the Hprt gene and exon 51 of the fibrillin gene. Mutat Res 432, (2000). 52. Arenas, M., Duley, J., Sumi, S., Sanderson, J. & Marinaki, A. The ITPA c.94c>a and g.ivs2+21a>c sequence variants contribute to missplicing of the ITPA gene. Biochim Biophys Acta 1772, (2007). 53. Pagenstecher, C. et al. Aberrant splicing in MLH1 and MSH2 due to exonic and intronic variants. Human Genetics 119, 9-22 (2006). 54. Auclair, J. et al. Systematic mrna analysis for the effect of MLH1 and MSH2 missense and silent mutations on aberrant splicing. Hum Mutat 27, (2006). 55. Dermaut, B. et al. A novel presenilin 1 mutation associated with Pick's disease but not beta- amyloid plaques. Ann Neurol 55, (2004). 56. Dumanchin, C. et al. Biological effects of four PSEN1 gene mutations causing Alzheimer disease with spastic paraparesis and cotton wool plaques. Hum Mutat 27, 1063 (2006). 57. Hull, J. et al. Identification of common genetic variation that modulates alternative splicing. PLoS Genet 3, e99 (2007). 58. Riveira- Munoz, E. et al. Transcriptional and functional analyses of SLC12A3 mutations: new clues for the pathogenesis of Gitelman syndrome. J Am Soc Nephrol 18, (2007).