1. Genome Analysis Research Group Leibniz Institute for Age Research – Fritz Lipmann Institute Violating the splicing rules: TG dinucleotides function as 3' splice sites in U2-dependent introns Karol Szafranski presented by Rileen Sinha to appear soon in Genome Biology 8 (2007)

2. subtle alternative splicing overview Cartegni et al., Nat Rev Genet 3, 285 (2002)

3. subtle alternative splicing Exceptional splice sites? Quan & Forte MCB 10,910 (1990) Pollard et al. JBC 277,15241 (2002)...UGGAGA gu.....uuacug cag...UGGAGA gu.....uuacug CAG UGAGAA...UGAGAA... adenylyl cyclase stimulatory G-protein G  S GNAS

4. subtle alternative splicing Exceptional splice sites? Cartegni et al., Nat Rev Genet 3, 285 (2002) GU-AG rule for U2-dependent introns

5. subtle alternative splicing Exceptional splice sites? ` G/G A/G A/A 132 number of protein isoforms Hiller et al., Am J Hum Genet 78, 291 (2006) genotype

6. subtle alternative splicing Exceptional splice sites? adenylyl cyclase stimulatory G-protein G  S unusual UG acceptor site! Quan & Forte MCB 10,910 (1990) Pollard et al. JBC 277,15241 (2002)...UGGAGA gu.....uuacug cag...UGGAGA gu.....uuacug CAG UGAGAA...UGAGAA... NAGNAG no SNP! no GNAS

7. subtle alternative splicing Exceptional splice sites? Identification and validation of 36 TG acceptors in the human genome

8. mining spliced alignments Approach...NNNNNN nn.....nnnnnn nnn...NNNNNN nn.....nnnnnn NNN NNNNNN...NNNNNN... neutral approach to search for 3' splice variants: splice variant pairs (SVP) spliced alignment of ESTs: 3’ splice site (“E”) 3’ splice site (“I”)

9. mining spliced alignments Approach II...NNNNNN nnn.....nnnnnnnug NNNNNN... spliced alignment of RefSeq transcripts: UG 3’ splice site neutral approach to search for non-canonical splice sites based on RefSeq transcripts (independent of alternative splicing)

10. mining spliced alignments Pipeline systematic in silico screening for unusual alternative 3‘ splice sites of human introns human EST-to-genome alignmenthuman RefSeq-to-genome alignment quality filtering BLAST-validation of EST-supported splice variants manual inspection

11. mining spliced alignments Pipeline

12. mining spliced alignments Pipeline filter for paralogous gene loci validate putative splice variants by explicit BLAST searches; choose only high-quality BLAST alignments around the exon-exon junction require at least two ESTs as evidence for a splice variant quality aspects of EST-based screen for splice variants

13. mining spliced alignments ESTs vs. RefSeqs RefSeq-basedEST-based 44 raw cases of unusual intron termini 39 8artifacts30 34 (77%) considered valid 9 (23%) 6 3 28 non-canonical TG 3'ss in U2 introns 92% reproducible in RT-PCR

14. mining spliced alignments Experimental validation CNBP (cellular nucleic acid binding protein) UG-derived splice variants were experimentally validated! AG-derived splice variants UG-derived splice variants exon3exon4

15. functional relevance More than just noise? noise produced by error-prone spliceosomes? variants awaiting functionalization during evolution? functionally relevant alternative splicing? flanking sequence conservation splice site conservation tissue-specific splicing pattern Indicators :

16. functional relevance intron sequence conservation conservation of flanking intron sequence H.sapiens M.musculus ? exon intron

17. functional relevance intron sequence conservation efficient TG alternative splicing under purifying selection

18. functional relevance intron sequence conservation RYK intron 7

19. functional relevance tissue-dependent splice patterns CNBPBRUNOL4

20. structural and mechanistical implications structural properties and mechanistic implications

21. structural and mechanistical implications TGs were exclusively found as alternative acceptors, part of TG-AG splice site tandems structural pattern Significant structural constraint observed: 6 AG-TG, up to  4 nt 31 TG-AG, up to  28 nt

22. functional relevance Mechanism of SS choice

23. structural and mechanistical implications Mechanism of SS choice AG-AG tandems may have separate branchpoints BPAG...... allowing large-distance splice sites. BP TG-AG tandems share the same branchpoint (BP) BPTGAG...... limiting the distance of splice sites.

24. functional relevance Mechanism of SS choice rare but high-confidence TG acceptors TG acceptors always alternative TG-AG acceptor distance constraint high conservation of alternative splicing at TG acceptors... specific mechanism?

25. functional relevance Mechanism of SS choice...nnCUaAcnnnnnnnnnnn UG nnnnnnnnn AG NNNNNNNNN... U2AF 35 exon U2 snRNP typically 15-40 nt U2AF 65 splicing step II splice site selection TG splice site selection takes place in splicing step II AG is required for intron recognition during splicing step I intron recognition splicing step I SPF45...or another factor? Lallena et al. 2002, Cell 109,285

26. Acknowledgments Stefanie Schindler Stefan Taudien Klaus Huse Niels Jahn Matthias Platzer Genome Analysis, FLI Jena Michael Hiller Rolf Backofen Bioinformatics, Univ. Freiburg Stefan Schreiber Philip Rosenstiel Inst. Clin. Mol. Biol., Univ. Kiel