Grey arrows denote micrococcal nuclease cleavage and red asterisks indicate free 3 OH at the end of the nascent transcript and cleaved 5 ss splicing intermediate. (B) Volcano plots comparing S2P, S5P, and T4P CTD Pol II IPs to mock IPs. elongating Pol II-spliceosome complexes form strong interactions with nascent transcripts, resulting in footprints of approximately 60 nucleotides. Also, splicing intermediates formed by cleavage at the 5?splice site are associated with nearly all spliced exons. These spliceosome-bound intermediates are frequently ligated to upstream exons, implying a sequential, constitutive, and U12-dependent splicing process. Finally, lack of detectable spliced products connected to the Pol II active site in human HeLa or murine lymphoid cells suggests that splicing does not occur immediately following 3 splice site synthesis. Our results imply that most mammalian splicing requires exon definition for completion. has been shown to occur within a short kinetic window following Pol II elongation through an intron and into the adjoining exon (Oesterreich et?al., 2016). In mammals, the prevalence and larger size of introns instead predicts an exon definition mechanism. Binding of regulatory proteins to specific RNA sequences (splicing enhancers and inhibitors) act to select exons for splicing by promoting communication between 3 ss and 5 ss flanking the exon (Ast, 2004, Chen and Manley, 2009). The interplay between these regulatory factors dictates both constitutive and alternative splicing. Notably, transcription kinetics strongly impact splicing decisions. Slow Pol II elongation rates favor Grem1 splicing by allowing more time for spliceosome assembly (Dujardin et?al., 2014, Mu?oz et?al., 2010). Exons in general display a higher nucleosome density, suggesting that chromatin structure acts as a kinetic barrier to Pol II elongation (Nieto Moreno et?al., 2015, Schwartz and Ast, 2010). We have developed a strategy for native elongation transcript sequencing using mammalian cells (mNET-seq). We found an accumulation of transcripts mapping precisely to the 3 end of exons, as expected for intermediates formed after the first transesterification splicing reaction (Nojima et?al., 2015). This indicates that splicing must occur within a stable complex formed between the spliceosome and Pol II. Unexpectedly, splicing intermediates were preferentially detected by antibodies specific for S5P CTD, suggesting that splicing occurs in association with this Pol II phospho-isoform. We now show biochemically and by mass spectroscopy that Calcium D-Panthotenate both protein and snRNA components of the spliceosome associate with Pol II complexes containing S5P CTD. Notably, initial cleavage of the intron creates a dominant 5 ss intermediate that remains embedded in the Pol II-associated spliceosome. In contrast, 3 ss intermediates Calcium D-Panthotenate corresponding to released intron lariats were detected at very low levels, suggesting fast dissociation of the spliceosome from Pol II upon completion of splicing. Results Catalytically Active Spliceosome Associates with S5P CTD Pol II mNET-seq involves immunoprecipitation (IP) of human Pol II elongation complexes from chromatin solubilized by micrococcal nuclease (MNase) digestion with Pol II antibodies specific for different CTD isoforms (Nojima et?al., 2015, Schlackow et?al., 2017). RNA is isolated from these complexes and sequenced by linker ligation onto the RNA 3 ends derived either from nascent RNA in the Pol II active site or co-transcriptional RNA processing intermediates. Notably, we showed that co-transcriptional splicing is associated with Pol II phosphorylated on the CTD serine 5 position (S5P) (Nojima et?al., 2015, Schlackow et?al., 2017). We reasoned that, as well as sequencing IPed RNA, we should also be able to establish the protein composition of these specifically IPed Pol II complexes by mass spectroscopy (MS), termed the mNET-MS method (Figure?1A). A label-free quantitative proteomics approach (Hubner et?al., 2010, Hubner and Mann, 2011) was used to determine the abundance of proteins enriched in Pol II complexes IPed with antibodies specific Calcium D-Panthotenate for S5P, S2P, and T4P CTD relative to a mock (nonspecific IgG) IP control. The fold increase of spectral intensities was compared to p values determined by t tests, as previously described (Harlen et?al., 2016), generating volcano plots (Figure?1B). Significant interactors were taken to be proteins identified as enriched using a false discovery rate of 0.05. We show that complexes precipitated by all three antibodies contained Pol II subunits, transcription factors, histones, chromatin-associated proteins, RNA-binding proteins, and RNA-processing.