selectively target cells bearing the diseasednetwork (17).

REFERENCES AND NOTES

1. H. Saito, F. Posas, Genetics 192, 289–318 (2012).2. J. L. Brewster, T. de Valoir, N. D. Dwyer, E. Winter, M. C. Gustin,

Science 259, 1760–1763 (1993).3. D. Muzzey, C. A. Gómez-Uribe, J. T. Mettetal,

A. van Oudenaarden, Cell 138, 160–171 (2009).4. U. Alon, An Introduction to Systems Biology: Design Principles

of Biological Circuits (Chapman & Hall/CRC, 2006).5. H. D. Madhani, G. R. Fink, Science 275, 1314–1317

(1997).6. K. Furukawa, S. Hohmann, Mol. Microbiol. 88, 5–19

(2013).7. M. C. Good, J. G. Zalatan, W. A. Lim, Science 332, 680–686

(2011).8. M. Good, G. Tang, J. Singleton, A. Reményi, W. A. Lim, Cell 136,

1085–1097 (2009).9. H. D. Madhani, C. A. Styles, G. R. Fink, Cell 91, 673–684 (1997).

10. S. M. O’Rourke, I. Herskowitz, Genes Dev. 12, 2874–2886(1998).

11. E. A. Elion, J. Cell Sci. 114, 3967–3978 (2001).12. J. G. Zalatan, S. M. Coyle, S. Rajan, S. S. Sidhu, W. A. Lim,

Science 337, 1218–1222 (2012).13. F. Posas, H. Saito, Science 276, 1702–1705

(1997).14. P. Wei et al., Nature 488, 384–388 (2012).15. P. Hersen, M. N. McClean, L. Mahadevan, S. Ramanathan,

Proc. Natl. Acad. Sci. U.S.A. 105, 7165–7170(2008).

16. J. E. Purvis, G. Lahav, Cell 152, 945–956 (2013).17. M. Behar, D. Barken, S. L. Werner, A. Hoffmann, Cell 155,

448–461 (2013).

ACKNOWLEDGMENTS

We thank H. Youk, R. Almeida, S. Coyle, and M. Thomson forinsightful discussions. This work was supported by NIH grantsR01 GM55040, R01 GM62583, PN2 EY016546, and P50GM081879; the NSF Synthetic Biology Engineering Research

Center (SynBERC); and HHMI (to W.A.L.). This work was alsosupported in part by MOST grant 2015CB910300, National NaturalScience Foundation of China grant 31470819, and Peking-TsinghuaCenter for Life Sciences (to P.W.). A.M. is a European MolecularBiology Organization Fellow (ALTF 419-2010) and the recipient of aProgram for Breakthrough Biomedical Research PostdoctoralResearch Award (UCSF).

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/350/6266/1379/suppl/DC1Materials and MethodsFigs. S1 to S5Tables S1 and S2References (18, 19)Movies S1 and S2

10 March 2015; accepted 8 November 2015Published online 19 November 201510.1126/science.aab0892

TRANSCRIPTION

RNA polymerase II–associated factor 1regulates the release and phosphorylationof paused RNA polymerase IIMing Yu,1* Wenjing Yang,2* Ting Ni,3 Zhanyun Tang,1 Tomoyoshi Nakadai,1

Jun Zhu,2 Robert G. Roeder1†

Release of promoter-proximal paused RNA polymerase II (Pol II) during earlyelongation is a critical step in transcriptional regulation in metazoan cells. PausedPol II release is thought to require the kinase activity of cyclin-dependent kinase 9(CDK9) for the phosphorylation of DRB sensitivity–inducing factor, negativeelongation factor, and C-terminal domain (CTD) serine-2 of Pol II. We found thatPol II–associated factor 1 (PAF1) is a critical regulator of paused Pol II release,that positive transcription elongation factor b (P-TEFb) directly regulates theinitial recruitment of PAF1 complex (PAF1C) to genes, and that the subsequentrecruitment of CDK12 is dependent on PAF1C. These findings reveal cooperativityamong P-TEFb, PAF1C, and CDK12 in pausing release and Pol IICTD phosphorylation.

Thousands of developmentally regulated genesin metazoans harbor promoter-proximalpaused RNA polymerase II (Pol II) 30 to50 nucleotides downstream of their tran-scription start sites (TSS) (1–3). Paused Pol

IIs are usually phosphorylated on C-terminaldomain (CTD) Ser5 and are associated with 5,6-dichloro-1-b-D-ribofuranosylbenzimidazole (DRB)sensitivity–inducing factor (DSIF) and negativeelongation factor (NELF). Release of paused Pol II

into productive elongation is believed to requirephosphorylation of CTD Ser2, conversion of DSIFinto a positive elongation factor by phosphoryl-ation of its SPT5 subunit, and disassociation ofNELF (1). Although it was long believed that CTDSer2 phosphorylationwas catalyzed predominant-ly by cyclin-dependent kinase 9 (CDK9), themam-malian ortholog of yeast Bur1, recent studieshave identified CDK12 as the metazoan orthologof Ctk1, the major CTD Ser2 kinase in yeast, andsuggested that CDK9 is a CTD Ser5 kinase (4, 5).The yeast Paf1 complex and the human PAF1complex, of interest here, have been implicatedin transcription elongation on DNA and chro-matin templates, recruitment and activation ofhistonemodifiers, mRNA 3′ formation, etc. (6, 7).However, PAF1C has not been considered a crit-ical elongation factor because depletions of PAF1Csubunits in yeast and fly, while reducing the levelof CTD Ser2-phosphorylated elongating Pol II

(8, 9), did not affect the distribution of total Pol IIon active genes (8, 10).To study the function of human PAF1C, we per-

formed chromatin immunoprecipitation sequenc-ing (ChIP-seq) experiments for PAF1C subunitsPAF1, CDC73, LEO1, and CTR9, as well as totalPol II and CTD Ser2-phosphorylated Pol II [Pol II(ser-2p)], in human acutemyeloid leukemia THP1cells. Similar to Pol II (ser-2p), the four PAF1Csubunits occupied transcribed regions of mostactive genes and exhibited maximum occupancydownstream of transcription end sites (TESs)(Fig. 1, A to C, and fig. S1, A to F). LEO1 (fig. S1B)and CTR9 (fig. S1C) occupancies did not gener-ally overlap with the promoter-proximal Pol IIpeaks, as reported previously (2). However, PAF1and CDC73, the major scaffolding componentswithin human PAF1C (11), did overlap with thepromoter-proximal Pol II peaks (Fig. 1, B and C,and fig. S1, A, D, E, and F). Complementary strand-specific mRNA-seq analyses using RNA fromTHP1 cells identified 19,481 transcripts [readsper kilobase of transcript per million mappedreads (RPKM) > 1], corresponding to 10,664 genes,of which 9823 were bound by PAF1. Notably, thePAF1 binding signals on these genes positivelycorrelated with correspondingmRNA levels (Fig.1C and fig. S1, A to C), suggesting an involvementof PAF1C in Pol II transcription or transcription-coupled events.In further functional analyses, we used two

lentiviral short hairpin RNAs (shRNA #1 andshRNA #2, targeting different regions of thePAF1mRNA) to reduce the level of the keyPAF1 subunit (8, 11) in THP1 cells (Fig. 2A) andassessed global gene expression changes by RNA-seq. With false discovery rate < 0.01 and consid-ering only relative expression change by a factorof >1.5, of the 9823 genes bound by PAF1, only1351 showed changes in expression (table S1).The knockdown of PAF1 also resulted in an in-creased level of promoter-proximal paused Pol IIthat was not limited to genes whose mRNA levelswere affected by PAF1 knockdown (Fig. 2, B andC).Considering only genes with a normalized pro-moter read count change of >2, 5851 exhibitedincreased Pol II pausing and only 344 exhibited

SCIENCE sciencemag.org 11 DECEMBER 2015 • VOL 350 ISSUE 6266 1383

1Laboratory of Biochemistry and Molecular Biology, TheRockefeller University, New York, NY 10065, USA. 2SystemsBiology Center, National Heart, Lung, and Blood Institute,Bethesda, MD 20892, USA. 3State Key Laboratory of GeneticEngineering and Ministry of Education Key Laboratory ofContemporary Anthropology, Collaborative Innovation Centerof Genetics and Development, School of Life Sciences, FudanUniversity, Shanghai 200438, P.R. China.*These authors contributed equally to this work. †Correspondingauthor. E-mail: roeder@rockefeller.edu

RESEARCH | REPORTSon July 1, 2020

http://science.sciencemag.org/

Dow

nloaded from

decreased Pol II pausing (table S2). The increasedPol II pausing, which led to an average factor of 2increase in Pol II occupancy on promoters (Fig.2C), was confirmedby comparison of the travelingratio of total Pol II in control and knockdown cells(Fig. 2D).The apparent and seemingly paradoxical in-

crease of PAF1 occupancy near promoters inPAF1 knockdown cells relative to control cells(Fig. 2B) raised the possibility that the increasedPol II pausing might not be a direct effect of thedecreased PAF1 chromatin association. However,normalization of the PAF1 levels to Pol II levelsin control and knockdown cells revealed a rela-tive decrease in PAF1 chromatin association inknockdown cells, suggesting that the increasedPol II pausing is indeed a direct effect of reducedPAF1 association at the promoter-proximal re-gion (Fig. 2E). In a further validation of theChIP-seq results for total Pol II, ChIP-seq forCTD Ser5-phosphorylated Pol II revealed, as ex-pected, a corresponding increase in Pol II (ser-5p)in PAF1 knockdown cells relative to control cells(Fig. 2F and fig. S2, A and B). Thus, PAF1C is acritical elongation factor that regulates elonga-tion as early as the Pol II pausing release step.To determine whether the facilitation of Pol II

pausing release by PAF1C reflects a generalmecha-nism that is not cell type–specific, we knockeddown PAF1 (by shRNA #2) in human acute lym-phoblastic leukemia CCRF-CEM cells and com-pared PAF1 and Pol II occupancies in control andknockdown cells. Surprisingly, and in contrastto the results in THP1 cells, PAF1 knockdown inthese cells resulted in an increase in Pol II paus-ing on only 142 genes and a decrease in Pol IIpausing on 1244 genes (fig. S3, A to E, and tableS3). To rule out any off-target effects, the resultswere validated by PAF1 knockdown using shRNA#1 (fig. S3, F and G). With respect to the apparentcell-dependent or context-dependent variabilityin effects of PAF1C functions, we note thatwhereas PAF1C is generally considered a positiveeffector (6, 7), there are earlier (12) and more re-cent (13) reports of PAF1C function as a negativeregulator of Pol II pausing release. Therefore, thedifferential effects of PAF1 knockdown on Pol IIpausing inTHP1 cells (carrying theMLL-AF9 fusiongene) and CCRF-CEM cells (bearing TP53 muta-tions) likely reflect thedistinct genetic backgroundsand physiological states of the two cell types. Thus,thediverse results in current andpublished studies(12, 13) indicate variable context-dependent effectsof PAF1C components as either positive or negativeregulators, as further exemplified by a switch inthe CDC73 subunit from a positive regulator (on-coprotein) to a negative regulator (tumor suppres-sor) by tyrosine phosphorylation (14).The effect of PAF1 knockdown onPol II pausing

resembles that of pan-CDK inhibition by flavo-piridol (2, 15). To determine whether flavopiridoltreatment affects Pol II pausing in part throughPAF1C, we compared the genomic occupancy ofPol II and the LEO1 subunit of PAF1C in di-methyl sulfoxide (DMSO)– and flavopiridol-treatedTHP1 cells. Flavopiridol significantly increasedglobal Pol II pausing (Fig. 2G) and markedly re-

duced the chromatin occupancy of LEO1 (Fig. 2Hand fig. S4, A to B). These results implicatePAF1C as a key factor for the release of promoter-proxmal paused Pol II. The reduced occupancyof PAF1C as a result of flavopiridol treatmentsuggests that pan-CDK inhibition increases PolII pausing in part through compromising therecruitment of PAF1C.

In yeast, Ser2 and Ser5 phosphorylation of the PolII CTD, as well as phosphorylation of the Spt5subunit of DSIF by Bur1 (ortholog of metazoanCDK9), are critical for the recruitment of PAF1Cto target genes (16, 17). To determine whether,as in yeast, CDK9 or SPT5 is required for therecruitment of PAF1C in THP1 cells, we per-formed independent SPT5 and CDK9 knockdown

1384 11 DECEMBER 2015 • VOL 350 ISSUE 6266 sciencemag.org SCIENCE

Fig. 2. PAF1C is a criti-cal regulator of promoter-proximal pausing release

of Pol II. (A) Comparison of PAF1levels in control and PAF1 knockdownTHP1 cells by Western blot. b-Actinwas used as a loading control. (B andC) Comparison of the occupancy ofPAF1 (B) and Pol II (C) on an average

gene. (D) Comparison of the traveling ratios of genes bound by Pol II. (E) Comparison of the normalizedoccupancy of PAF1 on an average gene. (F) Comparison of the occupancy of Pol II (ser-5p) on an averagegene. (G and H) Comparison of the occupancy of Pol II (G) and LEO1 (H) on an average gene in DMSO-and flavopiridol-treated THP1 cells.

Fig. 1. PAF1C occupancy positively correlates withgene expression level. (A) An integrative genomicsviewer (IGV) browser snapshot comparing occupancyof PAF1, CDC73, LEO1, CTR9, Pol II (ser-2p), and totalPol II within the CA2 locus in THP1 cells. (B) Occu-pancy of PAF1, Pol II (ser-2p), and total Pol II on anaverage gene. (C) Metagene analyses showing posi-tive correlation between PAF1 occupancy and themRNA level of genes.

RESEARCH | REPORTSon July 1, 2020

http://science.sciencemag.org/

Dow

nloaded from

analyses and evaluated the effects on Pol II andPAF1 occupancy. As anticipated (2), SPT5 knock-down reduced the occupancy of Pol II and PAF1(fig. S5, A toD). Like flavopiridol treatment, CDK9knockdown (~80%; Fig. 3, A and B) reduced PAF1occupancy (Fig. 3C) but, unlike flavopiridol,had little effect on Pol II occupancy (Fig. 3D). Inaddition, and unexpectedly, CDK9 knockdownminimally increased global Pol II pausing (fig.S6A). However, a comparison of the levels ofBRD4-associated CDK9 (CDK9-active complex)and LARP7-associated CDK9 [CDK9-inactive 7SKsmall nuclear ribonucleic protein (snRNP) com-plex] (18) in control and CDK9 knockdown THP1cells revealed a preferential reduction of theCDK9 fraction in the 7SK snRNP complex rel-ative to the CDK9 fraction in the BRD4 com-plex (fig. S6B). Thus, the minimal effect of theCDK9 knockdown on Pol II pausing may be dueto the minimal effect on the kinase-active CDK9fraction, and if that is the case, the regulation ofPAF1C recruitment by CDK9 is likely indepen-dent of the CDK9 kinase activity.In a further analysis of the role of CDK9 in

PAF1C recruitment, a coimmunoprecipitationassay revealed strong association of endoge-

nous PAF1 and P-TEFb (a complex of CDK9 andcyclin T1) in THP1 cells (Fig. 3E). Themuch strongerassociation between PAF1C and P-TEFb relativeto the reported (19) PAF1-AF9 interaction (Fig.3E) raised the possibility of an AF9-independentinteraction between these two complexes. ThePAF1 and CDC73 subunits of PAF1C, as well asPol II, were reciprocally coimmunoprecipitatedwith CDK9 by a CDK9 antibody (Fig. 3F). Moreimportant, binding assays using purified P-TEFband PAF1C complexes established a direct AF9-independent interaction between these two com-plexes (Fig. 3G), indicating that P-TEFb contributesto the recruitment of PAF1C through directinteraction.These results raised the possibility that PAF1C

may regulate promoter-proximal pausing releaseof Pol II in part by facilitation of P-TEFb extrac-tion from 7SK snRNP. However, PAF1C was un-able to releaseP-TEFb from7SKsnRNP ina releaseassay (fig. S7). To further analyze the functionalconsequences of the PAF1C-CDK9 interaction,we conducted ChIP-seq experiments for CDK9in control and PAF1 knockdown cells. These re-vealed that (i) CDK9 is mainly associated withboth enhancers and promoters (Fig. 3, H and I)

(20, 21), and (ii) PAF1 depletion reduces normal-ized CDK9 occupancy on promoters (Fig. 3, J andK), consistent with a previous report (19). Theenhancer and promoter association of CDK9,along with Pol II imaging data (21), makes it lesslikely, as proposed in another study (13), thatP-TEFb generally travels with Pol II during elon-gation.Therefore,wepropose that (i) the interactionbetween PAF1C and P-TEFb is required mainly forthe initial recruitment of PAF1C but may alsostabilize the P-TEFb promoter association, and(ii) CDK9 and other kinases subsequently phos-phorylate the CTD of Pol II (16) and the CTR ofSPT5 (17), thus creating PAF1C binding sites onPol II and the associated DSIF that enable PAF1Cto facilitate release of paused Pol II into produc-tive elongation.In agreement with previous studies (8, 9), we

observed decreased Pol II CTD Ser2 phosphoryl-ation but increased CTD Ser5 phosphorylationafter PAF1 knockdown (fig. S8, A to C). Thus,PAF1C may be responsible either for the recruit-ment of CTD Ser2 kinases or for the regulation oftheir activity. Recent studies have suggested thatCDK12 is the predominant Ser2 kinase (4) andthat bromodomain-containing protein 4 (BRD4)

SCIENCE sciencemag.org 11 DECEMBER 2015 • VOL 350 ISSUE 6266 1385

Fig. 3. P-TEFb contributes to the recruitment of PAF1C. (A) Comparison of CDK9 levels in control and CDK9 knockdown THP1 cells. (B toD) ChIP-qPCRdatacomparing the occupancy of CDK9 (B), PAF1 (C), and Pol II (D). Error bars indicate SD (N = 3). (E) Coimmunoprecipitation of P-TEFb and AF9 with CDC73.(F) Coimmunoprecipitation of PAF1, CDC73, total Pol II, and Pol II (ser-2p) with CDK9. (G) Pulldown assay using immobilized P-TEFb as bait and PAF1C asprey. (H) CDK9 peak distribution in control and PAF1 knockdown cells.Total peak numbers are shown at the top of each column. (I) An IGV browser snapshotcomparing CDK9 occupancy within the c-MYC locus. (J and K) Comparison of CDK9 occupancy (J) and normalized CDK9 occupancy (K) on an average gene,respectively.

RESEARCH | REPORTSon July 1, 2020

http://science.sciencemag.org/

Dow

nloaded from

is an atypical CTD Ser2 kinase (22). A comparisonof CDK12 and BRD4 occupancies on the c-MYCgenebyChIP-qPCR (quantitative polymerase chainreaction) revealed reduced occupancy of CDK12,but not BRD4, in PAF1 knockdown cells (fig. S9,A and B). Genomic analysis of the effect of PAF1Con CDK12 recruitment revealed a global decreasein CDK12 occupancy, especially when normalizedto Pol II occupancy, in PAF1 knockdown cells(Fig. 4, A to C). These results are consistent withthe reduced level of Pol II (ser-2p) and suggest arole for PAF1C-dependent recruitment of CDK12in Pol II CTD Ser2 phosphorylation.Experiments to determine whether the PAF1C-

dependent recruitment of CDK12 is through adirect interaction showed that CDK12 and cyclinK, as well as Pol II and BRD4, were coimmuno-precipitated with an antibody to the CDC73subunit of PAF1C (Fig. 4D) and, reciprocally, thatthe PAF1 andCDC73 subunits of PAF1C, as well asPol II, were coimmunoprecipitated with CDK12–cyclin K by an antibody to CDK12 (Fig. 4E). ThePAF1C association with BRD4 was significantlyweaker than its association with CDK12 (Fig. 4D),partially explaining why the recruitment of BRD4is less dependent on PAF1C (fig. S9B). Analyseswith purified recombinant CDK12–cyclin K andPAF1C complexes revealed a robust direct bind-ing of PAF1C to CDK12–cyclin K under stringentconditions (Fig. 4F). A parallel binding assay withpurified proteins under similar conditions revealeda very weak interaction of CDK12–cyclin K withPol II (Fig. 4G) relative to the strong interactionwith PAF1C (Fig. 4F). These results strongly suggestthat (i) PAF1C, in addition to regulating the releaseof paused Pol II, is directly involved in the re-cruitment of CDK12, and (ii) along with theknown interaction of human PAF1C with Pol II(11), the association of CDK12–cyclin K with PolII is likely mediated by Pol II–bound PAF1C.

We next investigated whether CDK12–cyclin Kaffects the recruitment of PAF1C by comparingPAF1 and Pol II occupancy in control and cyclinK knockdown cells. Despite effecting a globallydecreased Pol II (ser-2p) (fig. S8A) and reducedCDK12 occupancy, cyclin K knockdown had littleeffect on PAF1 or Pol II occupancy (fig. S10, A toD). These results, which agreewith published datashowing that a knockout of yeast Ctk1 (orthologof CDK12) does not affect Paf1 occupancy (23),support a model in which Pol II–bound PAF1Crecruits CDK12.Our results indicate a critical role for PAF1C in

Pol II pausing release, a direct role for P-TEFb inPAF1C recruitment, and a PAF1C-CDK12–cyclinK interaction that is important for CTD Ser2

phosphorylation. These findings complement andextend previous results demonstrating functionsfor P-TEFb, DSIF/NELF, and Pol II CTD Ser2 phos-phorylation in Pol II pausing release and aresummarized in an updated model (fig. S11) thatwill guide further mechanistic studies of bothpositive and negative functions of PAF1C in tran-scriptional control.

REFERENCES AND NOTES

1. K. Adelman, J. T. Lis, Nat. Rev. Genet. 13, 720–731(2012).

2. P. B. Rahl et al., Cell 141, 432–445 (2010).3. L. J. Core, J. J. Waterfall, J. T. Lis, Science 322, 1845–1848

(2008).4. B. Bartkowiak et al., Genes Dev. 24, 2303–2316

(2010).5. N. Czudnochowski, C. A. Bösken, M. Geyer, Nat. Commun. 3,

842 (2012).6. B. N. Tomson, K. M. Arndt, Biochim. Biophys. Acta 1829,

116–126 (2013).7. J. A. Jaehning, Biochim. Biophys. Acta 1799, 379–388

(2010).8. C. L. Mueller, S. E. Porter, M. G. Hoffman, J. A. Jaehning,

Mol. Cell 14, 447–456 (2004).9. K. Nordick, M. G. Hoffman, J. L. Betz, J. A. Jaehning,

Eukaryot. Cell 7, 1158–1167 (2008).

10. K. Adelman et al., Mol. Cell. Biol. 26, 250–260(2006).

11. J. Kim, M. Guermah, R. G. Roeder, Cell 140, 491–503(2010).

12. X. Bai et al., Cell 142, 133–143 (2010).13. F. X. Chen et al., Cell 162, 1003–1015

(2015).14. A. Takahashi et al., Mol. Cell 43, 45–56

(2011).15. I. Jonkers, H. Kwak, J. T. Lis, eLife 3, e02407

(2014).16. H. Qiu, C. Hu, N. A. Gaur, A. G. Hinnebusch, EMBO J. 31,

3494–3505 (2012).17. Y. Liu et al., Mol. Cell. Biol. 29, 4852–4863

(2009).18. B. M. Peterlin, D. H. Price, Mol. Cell 23, 297–305

(2006).19. N. He et al., Proc. Natl. Acad. Sci. U.S.A. 108, E636–E645

(2011).20. J. Lovén et al., Cell 153, 320–334 (2013).21. A. Ghamari et al., Genes Dev. 27, 767–777

(2013).22. B. N. Devaiah et al., Proc. Natl. Acad. Sci. U.S.A. 109,

6927–6932 (2012).23. S. H. Ahn, M. Kim, S. Buratowski, Mol. Cell 13, 67–76

(2004).

ACKNOWLEDGMENTS

We thank G. Morin for providing the cyclin K cDNA, andS. Malik, S. Murphy, and M. Guermah for critical reading ofthe manuscript. Supported by a Leukemia and LymphomaSociety SCOR grant (R.G.R.), the intramural researchprogram of the National Heart, Lung, and Blood Institute(W.Y. and J.Z.), and a National Science Foundation of Chinagrant (T. Ni). ChIP-seq and RNA-seq data have been submittedto GEO under accession number GSE62171.

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/350/6266/1383/suppl/DC1Materials and MethodsSupplementary TextFigs. S1 to S11Tables S1 to S5References (24–29)

12 August 2015; accepted 21 October 201510.1126/science.aad2338

1386 11 DECEMBER 2015 • VOL 350 ISSUE 6266 sciencemag.org SCIENCE

Fig. 4. PAF1C is respon-sible for the recruitmentof CDK12. (A and B)Comparison of CDK12occupancy (A) and nor-malized CDK12 occupancy(B) on an average genein control and PAF1 knock-down THP1 cells, respec-tively. (C) An IGV browsersnapshot comparingCDK12 occupancy withinthe TMEM30A locus.(D) Coimmunoprecipitationof CDK12, cyclin K, totalPol II, Pol II (ser-2p), andBRD4 with CDC73. (E) Co-immunoprecipitation ofPAF1, CDC73, total Pol II,and Pol II (ser-2p) withCDK12. (FandG) Pulldownassays using immobilizedCDK12–cyclin K com-plex as bait, and PAF1C(F) and purified Pol II(G) as prey, respectively.

RESEARCH | REPORTSon July 1, 2020

http://science.sciencemag.org/

Dow

nloaded from

RNA polymerase IIassociated factor 1 regulates the release and phosphorylation of paused−RNA polymerase II

Ming Yu, Wenjing Yang, Ting Ni, Zhanyun Tang, Tomoyoshi Nakadai, Jun Zhu and Robert G. Roeder

DOI: 10.1126/science.aad2338 (6266), 1383-1386.350Science 

, this issue p. 1383Sciencephosphorylation of Pol II on its C-terminal domain, freeing it to start transcription in earnest.complexes. The positive transcription elongation factor b helps recruit PAF1 to the paused Pol II. This facilitates the

associated factor 1 (PAF1) plays a central role in regulating the activation of these paused Pol II− show that Pol IIet al.Yu genes have a ''paused'' Pol II near their promoters, waiting to be released so they can start messenger RNA synthesis.

RNA polymerase II (Pol II) is the principal protein complex required for gene transcription in metazoan cells. Many''Please release me, let me go.''

ARTICLE TOOLS http://science.sciencemag.org/content/350/6266/1383

MATERIALSSUPPLEMENTARY http://science.sciencemag.org/content/suppl/2015/12/09/350.6266.1383.DC1

REFERENCES

http://science.sciencemag.org/content/350/6266/1383#BIBLThis article cites 29 articles, 8 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience

Copyright © 2015, American Association for the Advancement of Science

on July 1, 2020

http://science.sciencemag.org/

Dow

nloaded from