Transcription and splicing dynamics during early Drosophila development

T ansc ip ion and splicing dynamics du ing ea ly D osophila de elopmen PEDRO PRUDÊNCIO, 1,2,5 ROSINA SAVISAAR, 1,5 KENNY REBELO, 1,4 , 5 RUI GONÇALO MARTINHO, 1,2,3 and MARIA CARMO-FONSECA 1 1 Ins i u o de Medicina Molecula João Lobo An unes, Faculdade de Medicina, Uni e sidade de Lisboa, 1649-028 Lisboa, Po ugal 2 Alga e Biomedical Cen e Resea ch Ins i u e (ABC-RI), Uni e sidade do Alga e, 8005-139 Fa o, Po ugal 3 Depa men o Medical Sciences and Ins i u e o Biomedicine (iBiMED), Uni e sidade de A ei o, 3810-193 A ei o, Po ugal ABSTRACT Widesp ead co ansc ip ional splicing has been demons a ed om yeas o human. Howe e , mos s udies o da e ad- d essing he kine ics o splicing ela i e o ansc ip ion used ei he Saccha omyces ce e isiae o me azoan cul u ed cell lines. He e, we adap ed na i e elonga ing ansc ip sequencing echnology (NET-seq) o measu e co ansc ip ional splic- ing dynamics du ing he ea ly de elopmen al s ages o D osophila melanogas e emb yos. Ou esul s e eal he posi ion o RNA polyme ase II (Pol II) when bo h canonical and ecu si e splicing occu . We ound he e ogenei y in splicing dynam- ics, wi h some RNAs spliced immedia ely a e in on ansc ip ion, whe eas o o he ansc ip s no splicing was obse ed o e he i s 100 n o he downs eam exon. In ons ha show splicing comple ion be o e PolII has eached he end o he downs eam exon a e necessa ily in on-de ined. We s udied he splicing dynamics o bo h nascen p e-mRNAs an- sc ibed in he ea ly emb yo, which ha e ew and sho in ons, as well as p e-mRNAs ansc ibed la e in emb yonic de el- opmen , which con ain mul iple long in ons. As expec ed, we ound a ela ionship be ween he p opo ion o spliced eads and in on size. Howe e , in on de ini ion was obse ed a all in on sizes. We u he obse ed ha genes an- sc ibed in he ea ly emb yo end o be isola ed in he genome whe eas genes ansc ibed la e a e o en o e lapped by a neighbo ing con e gen gene. In isola ed genes, ansc ip ion e mina ion occu ed soon a e he polyadenyla ion si e, while in o e lapped genes, Pol II pe sis ed associa ed wi h he DNA empla e a e clea age and polyadenyla ion o he nascen ansc ip . Taken oge he , ou da a un a el no el dynamic ea u es o Pol II ansc ip ion and splicing in he de eloping D osophila emb yo. Keywo ds: D osophila melanogas e emb yo; NET-seq; ansc ip ion e mina ion; splicing kine ics INTRODUCTION I is by now la gely accep ed ha splicing can happen, and o en does happen, while ansc ip ion is s ill in p og ess (Beye and Osheim 1988; Ca illo Oes e eich e al. 2010; Ameu e al. 2011; Khodo e al. 2011; Windhage e al. 2012; B ugiolo e al. 2013; Nojima e al. 2015; Alpe e al. 2017). I is also known ha he p ocesses o splicing and ansc ip ion a e igh ly in e linked. The RNA polyme ase II (Pol II) elonga ion a e can a ec exon inclu- sion (de la Ma a e al. 2003; Fong e al. 2014; Aslanzadeh e al. 2018; Maslon e al. 2019), and many o he p o eins in ol ed in splicing associa e wi h he Pol II la ge subuni ca boxy- e minal domain (CTD) (Mo is and G eenlea 2000; Emili e al. 2002; Da id e al. 2011; Gö nemann e al. 2011; Hsin and Manley 2012; Nojima e al. 2018). Con e sely, splicing may also a ec ansc ip ion, wi h e idence sugges ing ha Pol II slows down a exons (Alexande e al. 2010; Ca illo Oes e eich e al. 2010; Jonke s e al. 2014; Veloso e al. 2014; Maye e al. 2015), po en ially o allow o splicing o comple e. Se e al s udies ha e add essed he iming o in on exci- sion ela i e o Pol II elonga ion. A nascen RNA sequencing s udy (Ca illo Oes e eich e al. 2016) showed ha in Saccha omyces ce e isiae, splicing could occu as soon as he 3′splice si e had eme ged om he ansc ip ion machin- e y, sugges ing ha splicing may be comple ed immedia ely a e in on ansc ip ion. P e ious a emp s o de e mine he du a ion o splicing in i o had e u ned es ima es anging om a ew seconds o se e al minu es (Alpe e al. 2017). The lowes o hese es ima es, such as he ew-second es i- ma e epo ed in Ma in e al. (2013), a e consis en wi h 4 P esen add ess: Donnelly Cen e, Uni e si y o To on o, To on o, On a io M5S 3E1, Canada 5 These au ho s con ibu ed equally o his wo k. Co esponding au ho s: ca mo. [email protected] , [email p o ec ed] A icle is online a h p://www. najou nal.o g/cgi/doi/10.1261/ na. 078933.121. F eely a ailable online h ough he RNA Open Access op ion. © 2022 P udêncio e al. This a icle, published in RNA, is a ailable unde a C ea i e Commons License (A ibu ion 4.0 In e na ional), as desc ibed a h p://c ea i ecommons.o g/licenses/by/4.0/. RNA (2022) 28:139–161; Published by Cold Sp ing Ha bo Labo a o y P ess o he RNA Socie y 139 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om splicing being comple ed igh a e he 3′splice si e is an- sc ibed. The mos ecen b eak h oughs in he ield ollowed he de elopmen o long- ead nascen RNA sequencing echnologies (D exle e al. 2020; Reime e al. 2021; Sousa-Luís e al. 2021). Collec i ely, hese s udies pe o med in human and D osophila cul u ed cell lines show ha al- hough many in ons a e immedia ely excised as soon as he downs eam exon eme ges om Pol II, a subse emains unspliced and p og essi ely unde goes delayed splicing while Pol II ansc ibes u he (D exle e al. 2020; Reime e al. 2021; Sousa-Luís e al. 2021). He e, we s udied he dynamic p ope ies o ansc ip ion and co ansc ip ional splicing du ing he ea ly s ages o de elopmen in D osophila melanogas e emb yos using na i e elonga ing ansc ip sequencing (NET-seq) (Chu chman and Weissman 2011; Nojima e al. 2015). An ad an age o using D osophila emb yos is ha hey a e mo e physiological han cul u ed cell lines. Mo eo e , compa ed o human, he D osophila genome is mo e com- pac (G a eley e al. 2011) and hus he co e age o NET- seq eads on in agenic egions is highe . Ano he ad an- age o he D osophila model is ha genes ansc ibed ini- ially in he emb yo ha e ew and sho in ons (like yeas genes), whe eas genes ansc ibed la e con ain mul iple long in ons (mo e simila o human genes). An addi ional ea u e o D osophila is he p esence o many genes wi h excep ionally long in ons ha a e subdi ided by a nonca- nonical mechanism e med ecu si e splicing (Du e al. 2015; Joseph e al. 2018; Pai e al. 2018). Fo all hese ea- sons, D osophila is an a ac i e model o s udy co an- sc ip ional splicing dynamics. D osophila ea ly de elopmen is cha ac e ized by apid mi o ic cycles ha lack cy okinesis, esul ing in nuclea p o- li e a ion in a syncy ial cy oplasm (Campos-O ega 1985; La e e al. 2015). Du ing hese ini ial mi o ic di isions, which impose signi ican cons ain s on ansc ip ion and splicing (She moen and O’Fa ell 1991; Ro he e al. 1992; Guilgu e al. 2014; Ma inho e al. 2015; Sandle e al. 2018; Kwasnieski e al. 2019), he emb yo la gely e- lies on ma e nally deposi ed mRNAs. Subsequen ly, he du a ion o he cell cycle is p og essi ely expanded, mem- b anes o m be ween he nuclei and seg ega e hem in o cells, and he zygo ic ansc ip ome s a s o be ully ex- p essed (Campos-O ega 1985; Bly he and Wieschaus 2015; Yuan e al. 2016). By pe o ming dNET-seq in he de eloping D osophila emb yo, we ha e un a eled no el dynamic ea u es o Pol II ansc ip ion and splicing. RESULTS Na i e elonga ing ansc ip sequencing in D osophila emb yos (dNET-seq) Ou i s ask was o adap he NET-seq echnology o use in D osophila emb yos (hence o h e e ed o as dNET- seq). The NET-seq echnique in ol es isola ion o an- sc ip ion complexes o med by Pol II, he DNA empla e and he nascen RNA by immunop ecipi a ion, wi hou c osslinking (Chu chman and Weissman 2011; Nojima e al. 2015). A e solubiliza ion o Pol II complexes unde na i e condi ions by ex ensi e mic ococcal nuclease (MNase) diges ion o isola ed na i e ch oma in, elonga ion complexes we e immunop ecipi a ed using an ibodies ha speci ically ecognize di e en phospho yla ion s a es o he Pol II CTD (Nojima e al. 2015). We collec ed emb yos a 2–3 h a e e iliza ion ( e- e ed o as ea ly emb yos) and 4–6 h a e e iliza ion ( e- e ed o as la e emb yos) (Fig. 1A). Analysis o emb yos s ained wi h a luo escen dye o isualize DNA (Fig. 1B) e- ealed ha ea ly emb yos we e p edominan ly in mi o ic cycle 14 (s age 5), which is when massi e ac i a ion o zy- go ic ansc ip ion occu s (Campos-O ega 1985; La e e al. 2015). The majo i y o la e emb yos we e in he la e s age o ge m-band ex ension (s age 10; Fig. 1C), when he emb yo unk (also known as he ge m-band) elonga es in he an e o-pos e io axis and na ows in he do so- en- al axis (Campos-O ega 1985; La e e al. 2015). In adap ing mNET-seq o D osophila emb yos, we op i- mized bu e s and washing condi ions o pu i y he ch o- ma in ac ion om manually so ed emb yos, solubilize he ansc ip ion complexes wi h MNase diges ion and im- munop ecipi a e Pol II wi h an ibodies. We used abbi polyclonal an ibodies aised agains syn he ic pep ides o he YSPTSPS epea o he CTD o he la ges Pol II sub- uni in Saccha omyces ce e isiae, phospho yla ed a ei he S5 (ab5131 Abcam) o S2 (ab5095 Abcam). Bo h an ibod- ies ha e been ex ensi ely used o ch oma in immunop e- cipi a ion expe imen s in D osophila melanogas e (Yan e al. 2014; Dahlbe g e al. 2015; Lee e al. 2015; Boija e al. 2017; Vizcaya-Molina e al. 2018; Akh a e al. 2019; A za e-Mejía e al. 2020). To enable di ec ional sequencing, he 5′hyd oxyl (OH) gene a ed by MNase diges ion o RNA was i s con e ed o a 5′phospha e by T4 polynucleo ide kinase (Fig. 1D). RNA was hen pu i ied om he immunop ecipi a ed Pol II complexes and size-selec ed using an RNA pu i ica ion ki p ocedu e ha combines a unique bu e sys em wi h a column echnology (see he Ma e ials and Me hods sec- ion o mo e de ail). RNAs wi h a size abo e 60 n we e used o subsequen liga ion o speci ic adap e s o he 5′P and 3′OH ends o each RNA agmen ollowed by PCR-based p epa a ion o a cDNA lib a y o high- h oughpu Illumina sequencing (Fig. 1D). A e sequenc- ing, adap e sequences we e immed and pai ed-end eads wi h sequence o e laps we e me ged in o a single ead ha spans he ull leng h o he o iginal RNA agmen (da k o ange; Fig. 1E). The esul ing single eads we e aligned o he D osophila e e ence genome. The nucleo- ide a he 3′end o each RNA agmen was iden i ied and i s genomic posi ion eco ded (as e isk; Fig. 1E). P udêncio e al. 140 RNA (2022) Vol. 28, No. 2 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om Two o h ee dNET-seq lib a ies we e independen ly p epa ed om ea ly and la e emb yos using S5P an ibody; h ee addi ional lib a ies we e p epa ed om la e emb yos using S2P an ibody. Each lib a y was sequenced o a high co e age wi h a ead leng h o 150 bp (Supplemen al Fig. S1; see he Ma e ials and Me hods sec ion o mo e de ail). Expe imen al ep oducibili y was demons a ed by s ong ag eemen o uniquely aligned ead densi y be ween bio- logical eplica es p epa ed wi h an ibodies aised agains he CTD phospho yla ed on ei he se ine 5 (S5P) o se ine 2 (S2P) posi ions (Supplemen al Fig. S1B–E). This sugges s ha bo h S5P and S2P an ibodies ecognize he CTD o elonga ing Pol II in D osophila emb yos. dNET-seq cap u es splicing in e media es and spliceosomal snRNAs NET-seqcap u esno only he inal(3′OHend)nucleo ideo nascen RNA bu also he 3′OH end o RNAs ha associa e wi h he PolII elonga ion complex (Nojima e al. 2015, 2018; Schlackowe al.2017).No ably,inhumans,mNET-seqo Pol II phospho yla ed on CTD se ine 5 de ec ed splicing in e - media es o med by clea age a he 5′splice si e a e he i s splicing eac ion. The p esence o such in e media es mani es s as an en ichmen o eads whose 3′ends map p e- cisely o he las nucleo ide o anexon. In bo h ea ly and la e D osophila emb yos, we indeed obse ed la ge peaks o dNET-seq/S5P eads mapping o he las nucleo ide o spliced exons, as shown o he kuk gene (g een as e isk; Fig. 2A). We also de ec ed dNET-seq/S5P peaks a he las nucleo ide o in ons, as shown o he eEF1alpha1 gene (pink as e isk; Fig. 2B); en ichmen o hese eads esul s om coimmunop ecipi a ion o eleased in on la ia s a e comple ion o he splicing eac ion (Fig. 2B). In addi ion, we obse ed eads co esponding o ma u e snRNAs en- gaged in co ansc ip ional spliceosome assembly, sugges - ing ha dNET-seq was cap u ing he ee 3′OH ends o he snRNAs (blue as e isk; Fig. 2C). We ound p ominen peaks a he end o spliceosomal U1, U2, U4, and U5 snRNAs. As expec ed, no peak was de ec ed mapping o he end o he U3 snRNA, which is in ol ed in he p ocessing o p e- RNA syn hesized by Pol I. No ewo hy, we obse ed an ac- cumula ion o dNET-seq signal a he end o U6 snRNA (Supplemen al Fig. S2A), con as ing wi h a lack o peak ob- se edinmammaliancells(Nojimae al.2018).Thisisconsis- en wi h he inding ha mos mammalian U6 snRNAs con ain a 2′,3′-cyclic phospha e e minal g oup a he 3′ end, whe eas U6 3′ends in D osophila cells consis o ei he acyclic2 ′,3′-phospha e, a 3′-phospha e o a 2′,3′-hyd oxyl g oup (Lund and Dahlbe g 1992). To quan i y how many cons i u i ely spliced exons ha e dNET-seq peaks a he end, we applied an algo i hm ha inds nucleo ides whe e he NET-seq ead densi y is a leas h ee s anda d de ia ions abo e he ansc ip mean in a local egion de ined by 100 bp ups eam and downs eam (Chu chman and Weissman 2011; P udêncio e al. 2020). Upon analyzing eplica es o dNET-seq/S5P and dNET-seq/S2P lib a ies, we iden i ied o e 10,000 exons showing splicing in e media e peaks (Fig. 2D). As peaks we e mo e equen ly de ec ed on ex- ons o genes wi h highe ead densi y (Supplemen al Fig. S2B), we classi ied he exons in o ou g oups (qua iles) based on he dNET-seq ead densi y o he co espond- ing gene and es ic ed he analysis o exons in he ou h qua ile, ha is, om genes wi h he highes ead densi y. E B A C D FIGURE 1. Na i e elonga ing ansc ip sequencing in D osophila emb yos. (A) Timeline o D osophila ea ly emb yonic de elopmen , which s a s wi h 13 apid syncy ial mi o ic cycles. Du ing in e phase o cycle 14, memb anes o m be ween he nuclei loca ed a he pe- iphe y o he emb yo (cellula iza ion). The new cells s a mo phoge- ne ic mo emen s leading o elonga ion o he emb yo unk (ge m- band ex ension). (B) Rep esen a i e images (s ained o DNA) o em- b yos in mi o ic cycle 14 (s age 5) and la e ge m-band expansion (s age 10). (C) The g aph depic s he de elopmen al s age o emb yos so ed in o he “ea ly”and “la e”g oups. App oxima ely 30,000 ea ly emb yos and 15,000 la e emb yos we e analyzed. (D) Ou line o he dNET-seq expe imen al p o ocol. (E) Ou line o dNET-seq da a analysis. T ansc ip ion and splicing dynamics in D osophila www. najou nal.o g 141 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om The esul s show ha splicing in e media e peaks a e de- ec ed in ∼80% o all cons i u i ely spliced exons. We u - he analyzed he so-called p e-MBT (mid-blas ula ansi ion) genes, which a e he i s zygo ic genes o be- come ansc ip ionally ac i e in ea ly emb yos (Chen e al. 2013), and genes exp essed in la e emb yos. We ound simila p opo ions o splicing in e media e peaks associ- a ed wi h p e-MBT and la e genes in he S5P and he S2P da a se s (Fig. 2E). We hen used he same me hodology and he same se o genes o de ec peaks a he las in onic nucleo ide, co esponding o eleased in on la ia s. Such peaks we e de ec ed in <10% o in ons (Fig. 2F). Taken oge he , hese esul s demons a e ha dNET- seq wi h an ibodies ha ecognize he Pol II CTD phos- pho yla ed a ei he S5 o S2 posi ions is capable o de ec ing splicing in e media es and spliceosomal snRNAs in D osophila emb yos, as p e iously epo ed in mammalian cells (Nojima e al. 2015, 2018; Schlackow e al. 2017). dNET-seq speci ically cap u es nascen RNA To alida e ha we we e de ec ing nascen ansc ip ion in D osophila emb yos, we analyzed ma e nal mRNAs ha a e ansc ibed du ing oogenesis and loaded in o he egg (Fig. 3A). As expec ed, ma e nal ansc ip s such as bicoid (Fig. 3B), nanos (Supplemen al Fig. S3A), gu ken (Supplemen al Fig, S3B), and Rab32 (Supplemen- al Fig. S3C) we e de ec ed by RNA-seq in emb yos col- lec ed 2–3 h a e e iliza ion, bu he dNET-seq signal o e hese genes was negligible. Howe e , a obus dNET-seq/S5P signal was ound a he pumilio gene (Fig. 3C). Exp ession o his gene was conside ed o be exclu- si ely ma e nal based on RNA-seq (Lo e al. 2011) and Pol II ChIP-seq da a (Chen e al. 2013), bu a GRO-seq s udy de ec ed pumilio nascen ansc ip s in emb yos col- lec ed 2–2.5 h pos e iliza ion (Saunde s e al. 2013). The de ec ion o pumilio RNA in ea ly emb yos by GRO-seq and dNET-seq/S5P highligh s he sensi i i y o hese wo echniques in cap u ing low-le el nascen ansc ip s. EF BAC D FIGURE 2. dNET-seq cap u es splicing in e media es and spliceosomal snRNAs. (A–C) The diag ams ou line he 3′OH ends gene a ed by co ansc ip ional clea age a he 5′splice si e (A), he 3′splice si e (B), and he ee 3′OH end o spliceosomal snRNAs (C). Below each diag am, dNET-seq/S5P p o iles o e he indica ed genes a e depic ed (da a om la e emb yos). The g een as e isk deno es he peak a he end o he exon (A). The pink as e isk deno es he peak a he end o he in on (B). The blue as e isk deno es he peak a he end o he U2 snRNA gene (C). A ows indica e he di ec ion o ansc ip ion. Exons a e ep esen ed by boxes. Thinne boxes ep esen UTRs. In ons a e ep esen ed by lines connec ing he exons. (D) Compa ison (Venn diag ams) o exons wi h a splicing in e media e peak de ec ed in biological eplica es o dNET-seq/S5P and dNET-seq/S2P lib a ies. (E,F) F equency o peaks co esponding o splicing in e media es (E) and eleased in on la ia s (F) in p e-MBT genes and genes exp essed in la e emb yos. Only genes wi h he highes ead densi y ( ou h qua ile) we e conside ed. P udêncio e al. 142 RNA (2022) Vol. 28, No. 2 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om E F B A CD G HI FIGURE 3. dNET-seq p o iles in ea ly and la e emb yos. (A) The diag am illus a es he empo al exp ession o ma e nal, p e-MBT, MBT, and pos - MBT genes du ing D osophila emb yonic de elopmen . dNET-seq/S5P and RNA-seq p o iles o e he ma e nal genes bicoid (B) and pumilio (pum) (C), he p e-MBT gene snail (D), and he pos -MBT gene Akap200 (E). Reads ha aligned o he posi i e s and a e in blue, and eads ha aligned o he nega i e s and a e in ed. (F) Me a-analysis o mean dNET-seq/S5P ead densi y a ound he ansc ip ion s a si e (TSS) in ma- e nalandp e-MBTgenes( eplica e1).(G–I) No malized me agene analysis in a bi a y uni s (A.U.). ThedNET-seq/S5P signalisdepic edalong he no malized gene leng h (g ay backg ound), as well as 500 bp ups eam o he ansc ip ion s a si e (TSS) and 500 bp downs eam om he poly- adenyla ion (pA) si e. (G) P e-MBT genes in ea ly emb yos. (H) T ansc ip ionally ac i e genes in la e emb yos; he signal o e genes ha ha e he 3′- UTR o e lapped by an an isense gene is depic ed in da k ed, while he signal o e genes wi h no o he genes wi hin 500 bp is depic ed in ligh ed. (I) T ansc ip ionallyac i egenesin la eemb yos; he signal o e genes ha ha e he 3′-UTR o e lapped by a ansc ip ionally ac i e an isense gene is depic ed in da k ed, while he signal o e genes wi h he 3′-UTRo e lapped by a ansc ip ionally inac i e an isense geneis depic ed in ligh ed. www. najou nal.o g 143 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om The s age a which he emb yo swi ches om elying on ma e nally deposi ed mRNAs and p o eins o unde going i s own ansc ip ion is e med he mid-blas ula ansi ion (MBT). Howe e , a ew genes (known as p e-MBT genes) become ansc ip ionally ac i e be o e MBT (Kwasnieski e al. 2019). A s ong dNET-seq signal was de ec ed in ea - ly emb yos o e he bodies o p e-MBT genes such as snail (Fig. 3D), ushi a azu (Supplemen al Fig. S3D) and odd skipped (Supplemen al Fig. S3E). Al oge he , we exam- ined 117 p e iously iden i ied p e-MBT genes and 35 ma- e nal mRNAs (Fig. 3F; Supplemen al Fig. S3F). The inding ha a obus dNET-seq/S5P signal was eco e ed om zygo ically bu no om ma e nally exp essed an- sc ip s indica es ha dNET-seq is speci ically a ge ing he nascen ansc ip ome. dNET-seq e eals di e ences in ansc ip ion e mina ion p o iles be ween genes exp essed in ea ly and la e emb yos Ha ing con i med ha dNET-seq was cap u ing nascen RNA, we nex in es iga ed he dis ibu ion o Pol II densi y o e ansc ip egions in ea ly and la e emb yos. dNET- seq/S5P p o iles (Fig. 3C–E; Supplemen al Fig. S3A–E) do no show he cha ac e is ic highe ead densi y nea he p omo e , as p e iously desc ibed in mammalian cells (Maye e al. 2015; Nojima e al. 2015). This is mos likely because Pol II ypically pauses ∼30–60 bp downs eam om he ansc ip ion s a si e (TSS) (Kwak e al. 2013) and in ou dNET-seq app oach we en ich o RNAs longe han 60 n (Supplemen al Fig. S1F); hus, we only eco d he posi ion o polyme ases ha ha e ansc ibed a leas 60 bp pas he TSS. We hen u ned ou a en ion o he dNET-seq signal a ound he polyadenyla ion (pA) si e. We ound ha in p e-MBT genes such as snail (Fig. 3D), ailless ( ll) (Supplemen al Fig. S4A), and nullo (Supplemen al Fig. S4B), he dNET-seq/S5P signal ends soon a e he pA si e. In con as , genes exp essed in la e emb yos such as Akap200 (Fig. 3E), His3.3A (Supplemen al Fig. S3H) and s (Supplemen al Fig. S3I), ha e a widesp ead dNET-seq/ S5P signal on in ons and egions downs eam om he pA si e. Analysis o RNA-seq da a se s e ealed ha mRNAs encoded by hese genes a e e icien ly spliced and 3′-end p ocessed (Supplemen al Fig. S3H,I). Thus, dNET-seq/S5P is cap u ing newly syn hesized ansc ip s ha ha e no ye been spliced, as well as RNAs syn hesized by Pol II complexes ha con inued o ansc ibe he DNA empla e a e he mRNA was clea ed and polyadenyla ed a he pA si e. We u he no ed ha he 3′-UTR o genes wi h a dNET-seq signal ex ending pas he pA si e is o e - lapped by ano he gene (gu ken,Nepl3, and In S1, espec- i ely), which is ansc ibed in he opposi e di ec ion. Nex , we gene a ed dNET-seq me ap o iles o p e-MBT and la e genes. Me agene analysis o he dNET-seq/S5P signal on p e-MBT genes exp essed in ea ly emb yos con- i med ha in his g oup o genes, he dNET-seq signal is sha ply educed pas he pA si e (Fig. 3G). The majo i y (65) o p e-MBT genes a e isola ed in he genome, wi h no o he gene on ei he s and wi hin a egion o 500 bp downs eam om he pA si e, as shown o ailless ( ll) (Supplemen al Fig. S4A). Ano he 26 p e-MBT genes a e embedded in la ge genes, as shown o nullo, which is lo- ca ed wi hin a long in on o he CG12541 gene (Supplemen al Fig. S4B). A smalle g oup o p e-MBT genes (21) ha e neighbo ing genes loca ed on ei he s and wi hin a egion o 500 bp downs eam om he pA si e, as shown o Elba2 (Supplemen al Fig. S4C). We u - he iden i ied 5 p e-MBT genes ha ha e he 3′-UTR o e - lapped by an an isense con e gen gene, as shown o spook (spo) (Supplemen al Fig. S4D). No ably, in hese genes, he dNET-seq/S5P signal ex ended pas he pA si e (Supplemen al Fig. S4C,D). To iden i y all he genes ha a e ansc ip ionally ac i e in la e emb yos, we used a s a egy adap ed om GRO- seq analysis (Co e e al. 2008) ha elies on ead densi y in gene dese egions as backg ound e e ence o ab- sence o ansc ip ion. Ve y la ge in e genic egions (gene dese s) we e di ided in o 50 kb windows, and ead densi ies we e calcula ed by di iding ead coun s in each window by he window leng h in bp (Supplemen al Fig. S4E). Genes wi h dNET-seq signal o e he gene body (in RPKM) abo e he 90 h pe cen ile o ead densi y o all in e genic egions analyzed we e conside ed o be ansc ip ionally ac i e (Supplemen al Fig. S4F). We iden i ied ∼7000 ac i e genes, wi h simila esul s ob ained om dNET-seq/S5P and dNET-seq/S2P da a se s (Supplemen al Fig. S4G). This se o ac i e genes includes o e 85% o he 3500 genes p e iously iden i ied as ac i ely ansc ibed a e he mid-blas ula ansi ion (MBT) based on ChIP-seq expe imen s (Chen e al. 2013). Nex , we di ided he genes ansc ibed in la e emb yos in o wo g oups, depending on whe he hei 3′-UTR was o was no o e lapped by ano he con- e gen gene. The me agene analysis shows ha when a e aged ac oss all ansc ibed genes, a e y low dNET- seq signal is de ec ed a he pA si e (Fig. 3H; Supplemen al Fig. S3G), as expec ed assuming ha when Pol II eaches his si e, he nascen ansc ip is clea ed and polyadenyla ed and he e o e he e is no RNA le a ached o he polyme ase o be sequenced. Howe e , in he case o la e genes ha ha e he 3′-UTR o e lapped by an an isense (con e gen ) gene, he dNET-seq/S5P and dNET-seq/S2P signals inc ease again a e he pA si e (Fig. 3H; Supplemen al Fig. S3G). Using he me hodology desc ibed abo e o iden i y ansc ibed genes, we ound ha he as majo i y (>80%) o o e lap- ping con e gen genes we e ansc ip ionally ac i e and only 219 genes we e silen . The me agene analysis shown in Figu e 3I clea ly indica es ha he de ec ion o dNET- P udêncio e al. 144 RNA (2022) Vol. 28, No. 2 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om seq/S5P signal pas he pA si e is independen om an- sc ip ion o he con e gen o e lapping gene. In conclusion, he dis ibu ion p o iles o dNET-seq/S5P and dNET-seq/S2P eads a ound he pA si e sugges dis- inc pa e ns o ansc ip ion e mina ion o genes ha a e ei he isola ed in he genome o o e lapped by ano h- e con e gen gene. Howe e , we canno exclude he pos- sibili y ha speci ically in isola ed genes, he lack o dNET- seq signal esul s om loss o CTD phospho yla ion as Pol II ansc ibes pas he pA si e. Analysis o dNET-seq ead densi y p o iles We nex ocused on he dis ibu ion o dNET-seq eads on exons and in ons. The numbe o nascen RNA eads whose 3′ends map a a pa icula genomic posi ion is p o- po ional o he numbe o Pol II molecules a ha posi ion. Thus, Pol II pause si es can be de ec ed as local peaks in NET-seq ead densi y (Chu chman and Weissman 2011; La son e al. 2014). Howe e , because we pe o med dNET-seq wi h an ibodies ha ecognize he Pol II CTD phospho yla ed a ei he S5 o S2 posi ions, changes in ead densi y may e lec dynamic phospho yla ion o he CTD a he han Pol II pausing. We excluded signal esul - ing om splicing in e media es (i.e., eads ha map o he e y las nucleo ide o in ons and exons we e disca ded) in o de o analyze only eads whose 3′ends associa e wi h he Pol II ac i e si e. We analyzed bo h S5P and S2P da a. Howe e , wewe e pa icula ly in e es ed in explo ing he mul iple peaks o dNET-seq/S5P signal obse ed on bo h exons and in ons (Fig. 4A), because co ansc ip ional splicing has been linked o S5P in humans (Nojima e al. 2015). A di icul y when pe o ming he sys ema ic iden i ica- ion o dNET-seq peaks is ha ansc ip s wi h highe ini i- a ion a es will con ain mo e eads and hus peaks a e mo e likely o be de ec ed han in mo e lowly ansc ibed genes. To con ol o his con ound, we de eloped a peak calling algo i hm ha de ec s egions whe e he local ead densi y is signi ican ly highe han expec ed by chance, gi en he o e -all ead densi y o he ansc ip (Fig. 4A; see also he Ma e ials and Me hods sec ion). We emphasize ha o de- ec ing eads co esponding o splicing in e media es (Fig. 2D–F), we used a peak calling me hod ha looks o signi - ican single-nucleo ide posi ions (Chu chman and Weissman 2011), whe eas his new me hod de ec s highe ead egions o a iable leng h. We hen aligned exons and in ons on he splice si es and calcula ed he a e age peak densi y a each posi ion. Exons ha e a highe o e -all peak densi y han in ons (mean p opo ion o nucleo ides in peaks ∼0.022/∼0.015 o eplica e 1/ eplica e 2 in ons and ∼0.043/∼0.034 o eplica e 1/ eplica e 2 exons; P< 2.2 × 10 −16 o bo h ep- lica es; wo- ailed Mann–Whi ney U- es wi h he peak densi ies o indi idual in ons/exons as da a poin s) (Fig. 4B). This is consis en wi h p e ious epo s indica ing ha he Pol II elonga ion a e is dec eased o e exons in mammalian (Jonke s e al. 2014; Maye e al. 2015) and D osophila cells (Kwak e al. 2013). In addi ion, peak den- si y is sha ply inc eased a ound he 5′splice si e (Fig. 4C). This could indica e ei he Pol II pausing o inc eased S5 phospho yla ion associa ed wi h splice si e ecogni ion. We also canno exclude ha misaligned splicing in e me- dia e eads a e con ibu ing o he obse ed inc ease in peak densi y a ound he 5′splice si e. The dNETseq/S5P p o ile a ound he 3′splice si e shows a highe peak den- si y egion jus a e he in on-exon bounda y, and a p o- g essi e inc ease in a e age exonic peak densi y is u he obse ed s a ing oughly 60 n a e he 3′splice si e (Fig. 4C). A e y simila peak densi y me ap o ile is obse ed wi h he dNETseq/S2P da a se (Fig. 4D). A po en ial ca ea o ead densi y analysis is ha nucle- o ide composi ion a ies sys ema ically ac oss exons and in ons. Fo example, exons end o ha e a highe GC con- en han in ons (Zhu e al. 2009). This could be p oblem- a ic as NET-seq elies on MNase diges ion o DNA and RNA o solubilize ch oma in. MNase diges ion o DNA is known o be sequence-biased, wi h mos no ably a p e e - ence o clea ing jus 5′o an adenine (Dingwall e al. 1981; Hö z and Al enbu ge 1981; Ga ney e al. 2012). An anal- ysis o he 5′ends o ou eads e ealed simila biases o MNase diges ion o RNA (Fig. 4D). This sequence p e e - ence could lead o a e ac ual a ia ion in ead densi y, wi h mo e eads being sampled om ansc ip s and an- sc ip egions whose nucleo ide composi ion is mo e sim- ila o MNase diges ion biases. To e i y o wha ex en ou esul s we e a ec ed by his con ound, we pe o med a simula ion o de e mine he expec ed dis ibu ion o eads based on he diges ion bias alone (Supplemen al Me hods). We concluded ha MNase biases a e unlikely o explain ei he he en ichmen o peaks in exons o he gene al p o ile o peak densi ies pas he 3′splice si e. dNET-seq cap u es ecu si e splicing in e media es Ha ing shown ha he spliceosome o ms a complex wi h he elonga ing Pol II in D osophila emb yos, we asked when splicing akes place ela i e o ansc ip ion. We i s looked a ecu si e splicing o long in ons because his p ocess in ol es he o ma ion o inhe en ly uns able in e - media es ha a e mo e likely o be o med soon a e he ansc ip ion o each in onic splice si e (Pai e al. 2018). In ecu si e splicing, long in ons a e emo ed by sequen- ial excision o adjacen sec ions in ol ing sepa a e splic- ing eac ions, each p oducing a dis inc la ia (Ha on e al. 1998). Recu si ely spliced in on segmen s a e bounded a one o bo h ends by ecu si e si es o a che poin s (Bu ne e e al. 2005), which co espond o ze o nu- cleo ide exons consis ing o jux aposed 3′and 5′splice T ansc ip ion and splicing dynamics in D osophila www. najou nal.o g 145 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om si es a ound a cen al AG|GT mo i , whe e he e ical line ep esen s he splice junc ion (Fig. 5A). To cap u e ecu si e splicing in e media es using dNET-seq, i is essen ial o ha e a good co e age o eads co esponding o nascen ansc ip s and spanning he splice junc ions. The o al numbe o eads esul ing om nascen RNA in each dNET-seq da a se is depic ed in Supplemen al Figu e S1G. By me ging he sequencing in o ma ion o o e lapped pai ed-end eads (Fig. 1E), we we e able o sequence on a e age ∼103 n pe nascen RNA (Supplemen al Fig. S1F,H). Focusing on p e iously iden i ied D osophila a che poin s (Du e al. 2015; Joseph e al. 2018), we ound dNET-seq/S5P eads ha span he junc ion be ween he canonical 5′splice si e a he end o he exon and he i s a che poin (RP1) in- e nal o he downs eam in on, as shown o he second in on o he Megalin gene (Fig. 5B). Reads spanning he subsequen in onic RPs we e also obse ed (Fig. 5B). O e all, we de ec ed dNET-seq/S5P and dNET-seq/S2P spliced eads suppo ing mos o he p e iously iden i ied ecu si e splicing e en s (Fig. 5C; Supplemen al Fig. S5A). Analysis o dNET-seq p o iles a ound a RP e eals an en ichmen o eads in a egion loca ed a ew nucleo ides EB AC D FIGURE 4. Analysis o dNET-seq ead densi y p o iles. (A)dNET-seq/S5P and RNA-seq p o iles o e he pos -MBT gene smoke ala m (smal). The ead numbe is depic ed a wo magni ica ion le els in wo biological eplica es. Fo eplica e 1, he line “Peak Calle 1”shows peaks called using he “la ge peaks”se ing, which is app op ia e o de ec ing la ge egions o pu a i e Pol II pausing. The line “Peak Calle 2”shows peaks called using he “small peaks”se ing, which p o ides highe spa ial esolu ion and has been used o subsequen analyses. RNA-seq da a o he same egions is also shown. (B) Peak densi y in he exons and in ons o ansc ip ionally ac i e genes (dNET-seq/S5P, eplica e 1). Peak densi y has been de ined as he pe cen age o nucleo ides wi hin a gi en exon o in on ha o e lap wi h a signi ican peak. (C,D) Me agene analysis o peak den- si y es ima ed om dNET-seq/S5P (C) and dNET-seq/S2P (D) da a om la e emb yos. To calcula e peak densi y o each posi ion, we di ided he numbe o in ons ha o e lap wi h a peak a ha posi ion by he o al numbe o in ons. The las 50 n o exons, he i s 50 n o in ons, he las 25 n o in ons, and he i s 100 n o exons a e shown. Only in e nal and ully coding exons om ansc ip ionally ac i e genes ha a e a leas 100 n long a e shown. Exons sho e han 150 n con ibu e o bo h he exon end and s a . Only in ons ha we e a leas 50 n long we e conside ed. (E) Sequence logo o nucleo ide equencies wi hin a 10-n window a ound he 5′ends o NET-seq eads. The combined heigh o he bases a each posi ion is p opo ional o he in o ma ion con en . Posi ion 6 co esponds o he 5′-mos nucleo ide o he ead. Pu a i e in e nal p iming eads, as well as eads mapping o he las nucleo ide o exons o in ons (possible splice in e media e and in on la ia eads) we e igno ed. P udêncio e al. 146 RNA (2022) Vol. 28, No. 2 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om downs eam om he RP, as shown o RP2 in he i s in- on o he Tenascin majo gene (Fig. 5D). No ewo hy, mos o hese eads a e al eady spliced o he p e ious RP (Fig. 5D). A me a-analysis o dNET-seq/S5P eads a ound 137 RPs con i ms ha many spliced eads can be obse ed jus downs eam om RPs (Fig. 5E,F), dem- ons a ing ha splicing occu s soon a e he ansc ip ion o in onic ecu si e si es. The obse ed en ichmen o dNET-seq/S5P eads a ound RPs u he sugges s ha e- cu si e splicing and Pol II elonga ion a e may be kine i- cally coupled. In ag eemen wi h his iew, a slow Pol II mu an enhanced ecu si e splicing o Ubx ansc ip s in D osophila emb yos (de la Ma a e al. 2003). Splicing akes place as Pol II ansc ibes pas he 3′′′′′ splice si e Ha ing es ablished ha dNET-seq cap u es ecu si e splic- ing, we hen asked whe he dNET-seq eads spanning ca- nonical exon-exon junc ions we e also de ec ed (Fig. 6A). To iden i y splicing e en s, we conside ed all in e nal and ully coding exons ha a e a leas 100 n long in ac i ely ansc ibed genes in ea ly and la e emb yos. Fo each splice junc ion, we coun ed how many eads had he 3′end mapped o he i s 100 n o he exon. Only exons wi h a leas 10 eads mapping o his egion we e conside ed (see he Ma e ials and Me hods sec ion o jus i ica ion o EF B A C D FIGURE 5. dNET-seq cap u es ecu si e splicing in e media es. (A) Schema ic illus a ing ecu si e splicing. A a che poin (RP) wi h jux aposed accep o and dono splice si e mo i s is indica ed. (B) Visualiza ion o dNET-seq/S5P eads ha align o he second in on o he Megalin (mgl) gene. Recu si ely spliced eads align o exon 2 (da k blue) and he in on a e RP1 (ligh blue). Unspliced eads a e depic ed in g ay. The numbe o spliced and unspliced eads a each RP in he in on is indica ed. (C) Venn diag am compa ing RPs iden i ied in wo dNET-seq/S5P biological eplica es and in p e iously epo ed s udies (Du e al. 2015; Joseph e al. 2018). (D) Numbe o dNET-seq/S5P eads ha ha e he 3′end mapped a ound RP2 in he i s in on o Tenascin majo (Ten-m) gene. The op panel depic s all eads, and he bo om panel depic s only eads ha ha e been spliced o RP1. (E,F) Me a-analysis wi h single nucleo ide esolu ion o no malized dNET-seq/S5P eads a ound RPs (n= 137) using all eads (E) o only eads spliced o he p e ious RP o exon (F). T ansc ip ion and splicing dynamics in D osophila www. najou nal.o g 147 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om MATERIALS AND METHODS Emb yo collec ion D osophila melanogas e lies (O egon R [O R] s ain) we e aised a 25°C, in polyp opylene ials con aining s anda d en iched cul- u e medium (co nmeal, molasses, yeas , soya lou , and bee oo sy up). Th ee-day-old lies (coun ing om pupae eclosion) we e a ened in cul u e medium supplemen ed wi h esh yeas o 2 d. Emb yos we e collec ed in o apple juice-aga pla es supple- men ed wi h esh yeas using app op ia e cages con aining ∼200 lies each. To a oid emale e en ion o olde emb yos, h ee p ecollec ions o 30 min each we e made be o e he i s collec ion o emb yos. To maximize egg laying, and a oid la ae con amina ion, adul lies we e ans e ed o clean emb yo col- lec ion cages e e y day o e 5 d. Fo ea ly s age emb yos (2–3 h a e egg-laying), adul emales we e allowed o lay eggs o 1 h in apple juice-aga pla es. Pla es we e subsequen ly collec ed and emb yos we e aged a 25°C o 90 min. Du ing he ollowing 30 min, emb yos we e ha es ed om he pla es, decho iona ed in 50% bleach solu ion o 2 min, and washed once in Phospha e Bu e ed Saline supplemen ed wi h 0.1% Tween-20 (PBT) and wice in deionized wa e . In o de o disca d olde emb yos (s age 6 and olde ), manual s aging o collec ed emb yos was pe o med wi h he help o o ceps and unde a magni ie scope. Emb yos we e hen esuspended in a solu ion con aining 120 mM NaCl and 0.04% T i on, and washed wice wi h 120 mM NaCl solu ion. A he end o he 3 h collec ion, he solu ion was emo ed and emb yos we e ozen in liquid ni ogen and s o ed a −80°C. Fo he la e s age (4–6 h), eggs we e laid o 2 h and aged a 25°C o 3.5 h. Emb yo collec ion and p ocessing was simila o he ea ly s age emb yos, bu in his case, no manual s aging was pe o med. Emb yo DNA s aining Fo each emb yo collec ion, and a e decho iona ion, a ep esen- a i e emb yo sample was collec ed om he o al pool and ixed in a scin illa ion lask, using a solu ion con aining one olume o 4% o maldehyde in PBT and ou olumes o hep ane, o 20 min a 100 pm. The lowe aqueous phase solu ion was subse- quen ly emo ed, 4 mL o me hanol was added and emb yos we e shaken igo ously o 1 min. Emb yos we e hen collec ed om he bo om o he scin illa ion lask, washed wice wi h me h- anol, and ozen a −20°C in me hanol. To ehyd a e he emb yos, hey we e washed o 5 min each, wi h 3:1, 1:1, and 1:3 mix solu- ions o me hanol:PBT. Emb yos we e subsequen ly washed wice in PBT and incuba ed wi h 1:5000 Sy ox g een (In i ogen), sup- plemen ed wi h 5 µg/mL RNase A (Sigma-Ald ich) in PBT o 15 min. A e washing wi h PBT, emb yos we e moun ed in luo es- cence moun ing medium (Dako) and examined in a Zeiss AxioZoom V16 Fluo escence S e eo Mic oscope o image acqui- si ion and emb yo s aging. Images we e p ocessed using ImageJ so wa e (NIH). dNET-seq and lib a y p epa a ion The dNET-seq p o ocol was adap ed om mNET-seq (Nojima e al. 2016). B ie ly, 300 µL o ozen emb yos was esuspended in 3.5 mL o Bu e B1 (15 mM HEPES-KOH, pH 7.6; 10 mM KCl; 5 mM MgCl 2 ; 1 mM DTT; 0.1 mM EDTA; 0.35 M suc ose; 4 µg/mL peps a in; 10 mM sodium ne abisul i e; 0.5 mM EGTA sup- plemen ed wi h comple e EDTA ee p o ease inhibi o [Roche] and PhoSTOP [Roche]). Emb yos we e homogenized in a Dounce homogenize wi h 11× s okes using a igh pes le on ice. The suspension was cen i uged a 7700g o 15 min a 4°C, he supe na an was disca ded, and he whi e pelle con aining he nuclei was esuspended in 500 µL o Bu e B1. The suspen- sion was again homogenized in he Dounce wi h 4× s okes and loaded wi hou mixing on he op o bu e B2 (15 mM HEPES- KOH, pH 7.6; 10 mM KCl; 5 mM MgCl 2 ; 1 mM DTT; 0.1 mM EDTA; 0.8 M suc ose; 4 µg/mL peps a in; 10 mM sodium me abi- sul i e; 0.5 mM EGTA supplemen ed wi h comple e EDTA ee p o ease inhibi o [Roche] and PhosSTOP [Roche]). The suspen- sion was cen i uged a 1310g o 30 min a 4°C, and he pelle was esuspended wi h 125 µL o NUN1 bu e (20 mM T is-HCl [pH 7.9]; 75 mM NaCl; 0.5 mM EDTA and 50% Glyce ol). An amoun o 1.2 mL o Bu e NUN2 (300 mM NaCl, 7.5 mM MgCl 2 , 1% NP-40, 1 M U ea supplemen ed wi h comple e EDTA ee p o ease inhibi o [Roche] and PhoSTOP [Roche]) was mixed wi h he nuclei and incuba ed on ice o 15 min p e- o ming a sho o ex e e y 3 min. Ch oma in was hen cen i- uged a 10,000g o 10 min a 4°C, washed wi h 100 µL o 1× MNase Bu e , and incuba ed in 100 µL o MNase eac ion mixing (1× MNase bu e and 30 gel uni /µL MNase [New England Biolabs]) o 3 min a 37°C wi h mixing a 1400 pm. The eac ion was s opped wi h 10 µL o 250 mM EGTA, cen i uged a 10,000g o 5 min a 4°C, and he supe na an con aining he solubilized ch oma in was eco e ed. Fo he ea ly emb yos sample, 2 × 300 µL o emb yos we e p epa ed in pa allel and pooled oge he a e he ch oma in solubiliza ion. Immunop ecipi a ion o Pol II– RNA complexes was pe o med using 50 µL o P o ein G Dynabeads (The mo Fishe Scien i ic), p eincuba ed o e nigh wi h 5 µg o he co esponden an ibody: an i-Pol II CTD S5P (ab5131 Abcam) o an i-Pol II CTD S2P (ab5095 Abcam) in 100 µL NET2 (50 mM T is-HCl pH 7.4; 150 mM NaCl and 0.05% NP-40) and washed h ee imes wi h NET2. Beads we e incuba ed wi h he solubilized ch oma in in 1 mL o al olume o NET2 o 1 h a 4°C, washed se en imes wi h 500 µL o NET2 and once wi h 100 µL o PNKT (1× PNK bu e and 0.1% ween) be o e incuba ion o 6 min in 50 µL o PNK eac- ion mix (1× PNKT, 1 mM ATP and 0.05 U/mL T4 PNK 3′phospha ase minus [NEB] in a he momixe a 37°C and 1400 pm. A e washing he beads wi h NET2, long RNA ag- men s we e isola ed using Quick-RNA Mic oP ep (Zymo e- sea ch): 300 µL o RNA Lysis Bu e in 33% E OH was mixed o he beads by pipe ing. Beads we e disca ded and he sus- pension was loaded in o a Zymo spin column ha was cen i- uged a 10,000g o 30 sec. The column was washed once wi h 400 µL RNA p ep bu e and wice wi h 700 µL and 400 µL RNA wash bu e , espec i ely. RNA was hen elu ed in 15 µL o DNase/RNase- ee wa e (Zymo) and s o ed a −80°C. An amoun o 100 ng o RNA was used o p epa e each lib a y, ollowing he s anda d p o ocol o he T uSeq Small RNA Lib a y P ep ki (Illumina). A e adap e liga ion and e e se ansc ip- ion, he lib a ies we e PCR ampli ied using 16 PCR cycles, and cDNA lib a ies we e ac iona ed in he gel be ween 130 o 300 bp. The lib a ies we e sequenced using PE-150 on he Illumina HiSeq X pla o m by No ogene Co., L d. P udêncio e al. 154 RNA (2022) Vol. 28, No. 2 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om Quan i ica ion and s a is ical analysis dNET-seq da a p ocessing Adap e sequences we e emo ed om all dNET-seq pai ed-end samples using Cu adap ( e sion 1.18) (Ma in 2011) wi h he ol- lowing pa ame e s: -a TGGAATTCTCGGGTGCCAAGG -A GATCGTCGGACTGTAGAACTCTGAAC -m 10 -e 0.05 –ma ch- ead-wildca ds -n 1. Pai ed-end ead me ging was pe o med us- ing bbme ge.sh om BBMap (Bushnell e al. 2017) wi h he “xloose”pa ame e . Me ged eads we e hen aligned o he D osophila e e ence genome (dm6; Ensembl elease 95) (Cunningham e al. 2019) using STAR ( e sion 2.6.0b) (Dobin e al. 2013) wi h –chimSegmen Min se o 20. Only uniquely mapped eads we e conside ed, ex ac ed using SAM ools ( e - sion 1.7) (Li e al. 2009) wi h -q se o 255. Excep ionally, in Supplemen al Figu e S2A, HiSa 2 was used, using he same dm6 genome anno a ion ile o genome indexing and using de- aul pa ame e s, wi h he –no-disco dan –no-mixed lags se . PCR in e nal p iming e en s gene a ed du ing lib a y p epa a ion we e emo ed using a cus om Py hon sc ip (P udêncio e al. 2020) wi h he ollowing pa ame e s: -a TGG.. -s pai ed. To ob ain single-nucleo ide esolu ion, a cus om Py hon sc ip (P udêncio e al. 2020) was used o ex ac he 5′end nucleo ide o ead 2 (a - e imming) in each sequencing pai , wi h he di ec ionali y indi- ca ed by ead 1 (Fig. 1E). Publicly a ailable RNA-seq da a se s used Publicly a ailable D osophila emb yonic ansc ip ome sequenc- ing da a [Poly(A) RNA-seq], pe o med in de elopmen al s ages simila o he dNET-seq ea ly and la e samples, we e used in his s udy. RNA-seq da a se s co esponding o cycle 14B (Lo e al. 2011) we e ob ained om he Gene Exp ession Omnibus (GEO) (samples GSM618409, GSM618410, GSM618421, and GSM618422 om da a se GSE25180). RNA-seq da a se s om 4–6 h old emb yos we e ob ained om modENCODE p ojec PRJNA75285 (G a eley e al. 2011) (accessions SRR023696, SRR023746, SRR023836, SRR035220, SRR023669, SRR035405, SRR035406, SRR024014, and SRR023539). RNA-seq da a p ocessing Adap e s we e emo ed om all da a se s using T im Galo e ( e - sion 0.4.4) (h p://www.bioin o ma ics.bab aham.ac.uk/p ojec s/ im_galo e/; las accessed 26 Ap il 2020). Da a se s om modENCODE and GEO we e aligned o he dm6 D osophila e - e ence genome (Ensembl elease 95) (Cunningham e al. 2019) using STAR ( e sion 2.6.0b) (Dobin e al. 2013) wi h –chimSegmen Min se o 20. S ing ie ( e sion 1.3.3b) (Pe ea e al. 2015) was used o quan i y no malized gene exp ession as T ansc ip s Pe Kilobase Million (TPM) alues wi h he ollowing pa ame e s: -a 5 -e. In addi ion, he iso o m lis was p o ided o- ge he wi h he -G pa ame e co esponding o he dm6 (Ensembl elease 95) GTF ile. Genes wi h TPM alues abo e 2 we e consid- e ed o be exp essed. Selec ion o genes and iso o ms o analysis Simila o a p e ious GRO-seq analysis (Co e e al. 2008), we used he ead densi y o e y la ge in e genic egions (gene dese e- gions) o de ine he e e ence o absence o ansc ip ion. Gene dese s we e di ided in o 50kb windows, and dNET-seq ead densi ies we e calcula ed by di iding he ead coun s in each win- dow by he window leng h (in bp). Read coun s pe window we e ob ained wi h bed ools genome co e age ( e sion 2.27.1-1- gb87c465) (Quinlan and Hall 2010), and an a bi a y densi y h eshold was de ined as he 90 h pe cen ile o he ead densi y dis ibu ion (Supplemen al Fig. S4A). T ansc ip s whose gene body dNETseq ead densi y exceeded his h eshold we e con- side ed o be ansc ip ionally ac i e (Supplemen al Fig. S4B). Fo all o he analyses pe o med on la e genes, only ansc ip ion- ally ac i e genes we e conside ed. In addi ion, only one iso o m pe gene was conside ed o all analysis, selec ed as he iso o m wi h he highes RPKM alue in he RNA-seq da a se o he co - esponding de elopmen al s age. The coo dina es o p e iously iden i ied p e-MBT genes we e ob ained om Chen e al. (2013) and con e ed o dm6 coo di- na es using he li O e ool (h ps://genome.ucsc.edu/cgi-bin/ hgLi O e ). The mos ep esen a i e iso o m o each gene was manually selec ed h ough isualiza ion o indi idual p o iles. Splicing in e media e and la ia de ec ion Exons con aining splicing in e media es o in ons con aining la - ia s we e iden i ied using a peak inde algo i hm (NET_sn Peak- Finde ) (Chu chman and Weissman 2011; P udêncio e al. 2020) ha de ec s he p esence o a peak in he las nucleo ide o an exon (splicing in e media e) o an in on (la ia ), by compa ing he accumula ion o 3′end eads mapping a ha posi ion wi h he mean ead densi y o he lanking 200 n . A peak is called when he ead densi y a he peak is supe io o he mean o his su ounding egion plus 3 s anda d de ia ions (Chu chman and Weissman 2011). Since gene ead densi y in luences peak de ec- ion, he exons we e di ided in o qua iles based on he dNET-seq ead densi y o he co esponding gene. Only exons om he highes qua ile (i.e., om genes wi h he highes ead densi y) we e conside ed in Figu e 2 o dNET-seq La e analysis. Analysis o ead densi y and peak calling RPKM alues o he me ged dNET-seq da a se s we e calcula ed in he ollowing manne : RPKM ( ansc ip ) = eads ×104×106 ( o al uniquely mapped eads) ×(gene leng h in bp) , whe e 10 4 no malizes o gene leng h and 10 6 no malizes o se- quencing dep h. Que ies o gene 3′UTR o e laps (Fig. 3) be ween genes we e pe o med wi h bed ools in e sec ( e sion 2.27.1-1- gb87c465) (Quinlan and Hall 2010). To de ec splicing in e media e and in on la ia peaks, he al- go i hm om Chu chman and Weissman (2011) was adap ed as desc ibed in P udêncio e al. (2020). To be able o call la ge e- gions as peaks, a cus om algo i hm was de eloped and imple- men ed in Py hon 3.7. No e ha like o all o he cus om Py hon code epo ed he e, he e is hea y eliance on bed ools 2.29.0 (Quinlan and Hall 2010) o ope a ions on coo dina e in- e als. In addi ion, NumPy 1.17.2 (Ha is e al. 2020) and SciPy 1.4.0 (Vi anen e al. 2020) we e used. Ou peak calle de ec s T ansc ip ion and splicing dynamics in D osophila www. najou nal.o g 155 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om egions whe e he local ead densi y is signi ican ly highe han expec ed by chance gi en he o e all ead densi y o he an- sc ip . Almos all o he nume ical pa ame e s used by he peak calle can be adjus ed. In Figu e 4A, esul s ob ained wi h wo pa - icula pa ame e iza ions a e shown. Peak Calle 1 (“la ge peaks”) is adap ed o de ec ing la ge peaks and p o ides esul s ha a e mo e in ui i e o a human obse e . Peak Calle 2 (“small peaks”) p o ides a ine spa ial esolu ion, and co esponds o he se ings used in all o he analyses in his s udy. The peak calle akes as inpu a BED ile wi h he 3′ends o eads (single-nucleo ide eso- lu ion), as well as a GTF ile wi h ansc ip and exon anno a ions and a lis o ansc ip s o analyze. I unc ions by calcula ing a slid- ing a e age o ead densi y wi hin each ansc ip (window size 5/ 21 o small/la ge; only eads mapping o he same s and as he anno a ed ansc ip a e conside ed). I hen andomly shu les he posi ions o he eads wi hin he ansc ip and ecalcula es he sliding a e ages o de e mine he andom expec a ion. This can be epea ed se e al imes ( i e in his s udy) o mo e obus ness. Windows ob ained wi h he ue ead dis ibu ion a e called as signi ican i hei ead densi y is highe han he 99 h pe cen ile o he simula ed windows. No e ha in his s udy, we used a se - ing whe eby his h eshold is calcula ed sepa a ely o each exon (and i s ups eam in on, o exons o he han he i s ), by exclud- ing he in on-exon pai o in e es and eads o e lapping i du ing he simula ion s ep. This is necessa y so ha when calling peaks wi hin a gi en exon (and i s ups eam in on), he h eshold se would no be a ec ed by he eads wi hin ha pa icula exon and i s ups eam in on. This way, o ins ance, he calling o a peak in he beginning o he exon is no a ec ed by he calling o a peak in he middle o he same exon (excep h ough po en- ial me ging, see below). A e he ini ial peaks a e called, hey a e il e ed o emo e peaks whe e mo e han 90% o he eads come om a single nucleo ide (p obable PCR duplica es), ha a e sho - e han 5 n , o ha o e lap wi h ewe eads han a speci ied h eshold (10/5 o la ge/small). Finally, peaks ha a e wi hin a speci ied dis ance o each-o he (21/5 n o la ge/small) a e me ged oge he . Indi idual gene p o iles Indi idual dNET-seq gene p o iles we e gene a ed by sepa a ing eads by s and using SAM ools ( e sion 1.7) (Li e al. 2009). S and-sepa a ed ead da a was con e ed o bedG aph o ma using bed ools genomeco wi h he -bg lag ( e sion 2.27.1-1- gb87c465) (Quinlan and Hall 2010). Co e age alues we e no - malized pe nucleo ide accoun ing o he o al numbe o uniquely aligned eads and wi h he scale se o eads pe 10 8 se- quences. The ou come was con e ed o bigwig iles h ough he bedG aphToBigWig ool (Ken e al. 2010) and uploaded o he UCSC Genome B owse (James Ken e al. 2002). Me agene analysis The ead densi y me agene plo s in Figu e 3F–I and Figu e 5E,F we e c ea ed wi h deepTools ( e sion 3.0.2) (Ramí ez e al. 2016). Me agenes wi h no malized gene size (Fig. 3) ha e bins o 10 bp while all o he me agene plo s in his s udy ha e single nucleo ide esolu ion. No malized gene and in on leng hs (Fig. 3F–I) we e ob ained h ough he scale- egions op ion. Exon-in- on junc ions wi hou no malized leng hs we e ob ained using he e e ence-op ion se o he 3′SS o he 5′SS. Fo no maliza- ion, we di ided he numbe o eads a each nucleo ide (o bin) by he o al numbe o eads in he en i e genomic egion unde analysis. These alues we e hen used o calcula e he mean o each nucleo ide, and he esul s we e plo ed in a bi a y uni s (A.U.) anging om 0 o 1. The peak o ead densi y me agene plo s in Figu es 3D, 4B and 7D, and Supplemen al Figu es S3G, S7A–G, and S7E–Jwe e p e- pa ed using cus om Py hon and R sc ip s (h ps://gi hub.com/ osinaSa /dNETseq_code). The peak densi y alue ep esen s he p opo ion o in ons/exons ha o e lap wi h a peak a ha posi ion. Only in e nal ully coding exons ha we e a leas 100 n long we e included. In addi ion, he in on jus ups eam o he exon had o be a leas 50 n long. Fo Figu e 7D, Supplemen al Figu es S7A–G and S7E–H, u he il e ing based on ead co e age was pe o med (see below). No e ha exons sho e han 150 nucleo ides con ibu e bo h o he ups eam and downs eam exonic p opo ion o he plo . Immedia e splicing analysis The immedia e splicing analysis was pe o med solely on he S5P da a se s. Only he 117 p e iously anno a ed p e-MBT genes (Chen e al. 2013) we e analyzed o ea ly da a se s. Reads we e conside ed as spliced i hey con ained “N”s in he CIGAR s ing, in a posi ion co esponding o an anno a ed in on. Reads ha o e lapped bo h he (unspliced) in on and he downs eam exon we e conside ed as unspliced. In bo h cases, only eads whe e he 3′end was loca ed a leas 5 n downs eam om he 3′ss we e included, o a oid analyzing misaligned eads whose 3′end should ha e mapped o he end o he ups eam exon in- s ead. Spliced eads ha had he 5′end mapped o he ups eam exon and he 3′end mapped o he in on we e conside ed indic- a i e o ecu si e splicing i he i s nucleo ide o he downs eam end (indica i e o he a che poin posi ion) ma ched he second G in he AGGT canonical splicing mo i . I he las nucleo ide o a ead ma ched he las nucleo ide o an exon, i was conside ed a splicing in e media e ead and no ep esen a i e o nascen RNA. A splicing a io was calcula ed by di iding he numbe o na- scen RNA spliced eads by he sum o he numbe o spliced and unspliced nascen RNA eads, only including eads whose 3′ends mapped o he i s 100 n o he downs eam exon. Only ully coding in e nal exons a leas 100 n long we e conside ed (ex- cep ionally, in Fig. 6E and K and Supplemen al Fig. S6O–Q, he 3′mos coding exon was also analyzed). Finally, we pe o med il- e ing o emo e exons whe e he ead co e age was oo low o allow o obus es ima ion o he splicing a io. The ele an h eshold was calcula ed o each da a se sepa a ely. We calcu- la ed he o al p opo ion o spliced eads ou o all spliced/ unspliced eads o he da a se o ob ain he expec ed splicing a io. We hen pe o med a binomial es o know he p obabili y o sampling only spliced/unspliced eads by chance unde he null ha he ue splicing a io equaled his expec a ion. We se he h eshold as he lowes numbe o eads ha had o be sam- pled o he p obabili y o be below 0.01. Th ough his p ocedu e, he h eshold was se a ≥10 eads o eplica e 1 and 2 o he la e da a se (no e minal coding exons), a ≥11/10 eads o eplica e 1/2 o he la e da a se including e minal coding exons, and a ≥14/9 eads o eplica e 1/2 o he ea ly da a se ( e minal coding exons always included). P udêncio e al. 156 RNA (2022) Vol. 28, No. 2 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om Gene a chi ec u e and nucleo ide composi ion analysis Gene a chi ec u e and nucleo ide composi ion pa ame e s we e calcula ed using cus om Py hon and R sc ip s based on Ensembl anno a ions o dm6.18 (Cunningham e al. 2019). Splice si e s eng h sco es we e calcula ed using MaxEn Scan (Yeo and Bu ge 2004) wi h de aul pa ame e s. Signi icance es - ing was pe o med using K uskal–Wallis es s wi h a Bon e oni co ec ion o mul iple compa isons, wi h he co ec ion applied sepa a ely o ei he eplica e. Fo p edic o s whe e he co ec ed P- alue was <0.05, Dunn’s es was pe o med on he pai wise compa isons using Rpackage dunn. es (Dinno 2017). The se- quence logo in Figu e 4E was gene a ed using he seqLogo pack- age e sion 1.52.0 in R (Bembom 2019). DATA DEPOSITION All aw and p ocessed sequencing da a gene a ed in his s udy ha e been submi ed o he NCBI Gene Exp ession Omnibus (GEO; h ps://www.ncbi.nlm.nih.go /geo/) unde accession num- be GSE152585. The Py hon code used is a ailable a h ps:// gi hub.com/ osinaSa /dNETseq_code and h ps://gi hub.com/ kenny ebelo. SUPPLEMENTAL MATERIAL Supplemen al ma e ial is a ailable o his a icle. ACKNOWLEDGMENTS We hank Takayuki Nojima and Nicholas P oud oo (Uni e si y o Ox o d, UK) o c i ical discussions. This wo k was suppo ed by unding o M.C.-F. (Fundação pa a a Ciência e Tecnologia, FCT/Minis é io da Ciência, Tecnologia e Ensino Supe io - Fundos do O çamen o de Es ado [UIDB/50005/2020], and FCT/ FEDER/POR Lisboa 2020, P og ama Ope acional Regional de Lisboa PORTUGAL 2020, g an LISBOA-01-0145-FEDER- 016394) and o R.G.M (FCT g an PTDC/BIA-BID/28441/2017). P.P. was a ecipien o an FCT ellowship (SFRH/BD/109689/ 2015). R.S. was a ecipien o an EMBO Long-Te m Fellowship (EMBO ALTF 101-2019). This p ojec has ecei ed unding om he Eu opean Union’s Ho izon 2020 esea ch and inno a ion p o- g amme unde g an ag eemen s no. 842695 (Ma ie Skłodowska- Cu ie Ac ions) and no. 857119 (RiboMed). Au ho con ibu ions: P.P. pe o med all molecula biology ex- pe imen s. R.S. pe o med he majo i y o splicing kine ics analy- ses and de eloped he peak calle o de ec ing pu a i e pause si es. K.R. pe o med addi ional bioin o ma ics analyses. P.P., R. S., R.G.M., and M.C.-F concei ed he s udy and w o e he pape . Recei ed July 26, 2021; accep ed Sep embe 23, 2021. REFERENCES Akh a J, K eim N, Ma ini F, Mohana G, B üne D, Binde H, Roignan J- Y. 2019. P omo e -p oximal pausing media ed by he exon junc- ion complex egula es splicing. Na Commun 10: 521. doi:10 .1038/s41467-019-08381-0 Alexande RD, Innocen e SA, Ba ass JD, Beggs JD. 2010. Splicing- dependen RNA polyme ase pausing in yeas . Mol Cell 40: 582– 593. doi:10.1016/j.molcel.2010.11.005 Alpe T, He zel L, Neugebaue KM. 2017. Pe ec iming: splicing and ansc ip ion a es in li ing cells. WIREs RNA 8: e1401. doi:10 .1002/w na.1401 Ameu A, Zaghlool A, Hal a dson J, We e bom A, Gyllens en U, Ca elie L, Feuk L. 2011. To al RNA sequencing e eals nascen an- sc ip ion and widesp ead co- ansc ip ional splicing in he human b ain. Na S uc Mol Biol 18: 1435–1440. doi:10.1038/nsmb.2143 A ie i CG, F ase HB. 2014. T ansc ip leng h media es de elopmen- al iming o gene exp ession ac oss D osophila.Mol Biol E ol 31: 2879–2889. doi:10.1093/molbe /msu226 A za e-Mejía RG, Josué Ce ecedo-Cas illo A, Gue e o G, Fu lan- Maga il M, Recillas-Ta ga F. 2020. In si u dissec ion o domain bound- a ies a ec genome opology and gene ansc ip ion in D osophila. Na Commun 11: 894. doi:10.1038/s41467-020-14651-z Aslanzadeh V, Huang Y, Sanguine i G, Beggs JD. 2018. T ansc ip ion a e s ongly a ec s splicing ideli y and co ansc ip ionali y in budding yeas . Genome Res 28: 203–213. doi:10.1101/g .225615.117 Ba sché E, Yani M, Mucha d C. 2006. The human SWI/SNF subuni B m is a egula o o al e na i e splicing. Na S uc Mol Biol 13: 22–29. doi:10.1038/nsmb1030 Bembom O. 2019. seqLogo: sequence logos o DNA sequence alignmen s. R package e sion 1.52.0. Be ge SM. 1995. Exon ecogni ion in e eb a e splicing. J Biol Chem 270: 2411–2414. doi:10.1074/jbc.270.6.2411 Beye AL, Osheim YN. 1988. Splice si e selec ion, a e o splicing, and al e na i e splicing on nascen ansc ip s. Genes De 2: 754–765. doi:10.1101/gad.2.6.754 Bly he SA, Wieschaus EF. 2015. Coo dina ing cell cycle emodeling wi h ansc ip ional ac i a ion a he D osophila MBT. Cu Top De Biol 113: 113–148. doi: 10.1016/bs.c db.2015.06.002 Bly he SA, Wieschaus EF. 2016. Es ablishmen and main enance o he i able ch oma in s uc u e du ing ea ly D osophila emb yogen- esis. Eli e 5: e20148. doi:10.7554/eLi e.20148 Boija A, Maha DB, Za e A, Holmq is P-H, Philip P, Meye s DJ, Cole PA, Lis JT, S enbe g P, Manne ik M. 2017. CBP egula es e- c ui men and elease o p omo e -p oximal RNA polyme ase II. Mol Cell 68: 491–503.e5. doi:10.1016/j.molcel.2017.09.031 B ugiolo M, He zel L, Neugebaue KM. 2013. Coun ing on co- an- sc ip ional splicing. F1000P ime Rep 5: 9. doi:10.12703/P5-9 Bu ne e JM, Miyamo o-Sa o E, Schaub MA, Conklin J, Lopez AJ. 2005. Subdi ision o la ge in ons in D osophila by ecu si e splic- ing a nonexonic elemen s. Gene ics 170: 661–674. doi:10.1534/ gene ics.104.039701 Bushnell B, Rood J, Singe E. 2017. BBMe ge: accu a e pai ed sho - gun ead me ging ia o e lap. PLoS One 12: e0185056. doi:10 .1371/jou nal.pone.0185056 Campos-O ega JA. 1985. Gene ics o ea ly neu ogenesis in D osophila melanogas e .T ends Neu osci 8: 245–250. doi:10 .1016/0166-2236(85)90097-9 Ca illo Oes e eich F, P eibisch S, Neugebaue KM. 2010. Global analysis o nascen RNA e eals ansc ip ional pausing in e minal exons. Mol Cell 40: 571–581. doi:10.1016/j.molcel.2010.11.004 Ca illo Oes e eich F, He zel L, S aube K, Huje K, Howa d J, Neugebaue KM. 2016. Splicing o nascen RNA coincides wi h in- on exi om RNA polyme ase II. Cell 165: 372–381. doi:10.1016/ j.cell.2016.02.045 Cha ho h KT, Ba ass JD, Webb S, Beggs JD. 2014. A splicing-depen- den ansc ip ional checkpoin associa ed wi h p espliceosome o - ma ion. Mol Cell 53: 779–790. doi:10.1016/j.molcel.2014.01.017 Chen K, Johns on J, Shao W, Meie S, S abe C, Zei linge J. 2013. A global change in RNA polyme ase II pausing du ing he T ansc ip ion and splicing dynamics in D osophila www. najou nal.o g 157 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om D osophila midblas ula ansi ion. Eli e 2: e00861. doi:10.7554/ eLi e.00861 Chu chman LS, Weissman JS. 2011. Nascen ansc ip sequencing isualizes ansc ip ion a nucleo ide esolu ion. Na u e 469: 368–373. doi:10.1038/na u e09652 Co e LJ, Wa e all JJ, Lis JT. 2008. Nascen RNA sequencing e eals widesp ead pausing and di e gen ini ia ion. Science 322: 1845–1849. doi:10.1126/science.1162228 Cunningham F, Achu han P, Akanni W, Allen J, Amode MR, A mean IM, Benne R, Bhai J, Billis K, Boddu S, e al. 2019. Ensembl 2019. Nucleic Acids Res 47: D745–D751. doi:10.1093/na /gky1113 Dahlbe g O, Shilko a O, Tang M, Holmq is P-H, Manne ik M. 2015. P-TEFb, he supe elonga ion complex and media o egula e a subse o non-paused genes du ing ea ly D osophila emb yo de- elopmen . PLoS Gene 11: e1004971. doi:10.1371/jou nal .pgen.1004971 Da id CJ, Boyne AR, Millhouse SR, Manley JL. 2011. The RNA poly- me ase II C- e minal domain p omo es splicing ac i a ion h ough ec ui men o a U2AF65–P p19 complex. Genes De 25: 972– 982. doi:10.1101/gad.2038011 de Almeida SF, Ca mo-Fonseca M. 2008. The CTD ole in co ansc ip- ional RNA p ocessing and su eillance. FEBS Le 582: 1971– 1976. doi:10.1016/j. ebsle .2008.04.019 De Con i L, Ba alle M, Bu a i E. 2013. Exon and in on de ini ion in p e-mRNA splicing. Wiley In e discip Re RNA 4: 49–60. doi:10 .1002/w na.1140 de la Ma a M, Alonso CR, Kadene S, Fededa JP, Blaus ein M, Pelisch F, C ame P, Ben ley D, Ko nblih AR. 2003. A slow RNA polyme ase II a ec s al e na i e splicing in i o. Mol Cell 12: 525–532. doi:10.1016/j.molcel.2003.08.001 Dingwall C, Lomonosso GP, Laskey RA. 1981. High sequence speci- ici y o mic ococcal nuclease. Nucleic Acids Res 9: 2659–2674. doi:10.1093/na /9.12.2659 Dinno A. 2017. dunn. es : Dunn’s es o mul iple compa isons using ank sums. R package e sion 1.3.5. h ps://c an. -p ojec .o g/ web/packages/dunn. es /index.h ml DobinA,Da isCA,Schlesinge F,D enkowJ,ZaleskiC,JhaS,Ba u P, Chaisson M, Ginge as TR. 2013. STAR: ul a as uni e sal RNA-seq aligne . Bioin o ma ics 29: 15–21. doi:10.1093/bioin o ma ics/b s635 D exle HL, Choque K, Chu chman LS. 2020. Splicing kine ics and co- o dina ion e ealed by di ec nascen RNA sequencing h ough nanopo es. Mol Cell 77: 985–998.e8. doi:10.1016/j.molcel.2019 .11.017 Du MO, Olson S, Wei X, Ga e SC, Osman A, Bolise y M, Plocik A, Celnike SE, G a eley BR. 2015. Genome-wide iden i ica ion o ze o nucleo ide ecu si e splicing in D osophila.Na u e 521: 376–379. doi:10.1038/na u e14475 Emili A, Shales M, McC acken S, Xie W, Tucke PW, Kobayashi R, Blencowe BJ, Ingles CJ. 2002. Splicing and ansc ip ion-associa - ed p o eins PSF and p54 n b /NonO bind o he RNA polyme ase II CTD. RNA 8: 1102–1111. doi:10.1017/S1355838202025037 E ickson JW, Cline TW. 1993. A bZIP p o ein, sis e less-a, collabo a es wi h bHLH ansc ip ion ac o s ea ly in D osophila de elopmen o de e mine sex. Genes De 7: 1688–1702. doi:10.1101/gad.7.9.1688 Fong N, Kim H, Zhou Y, Ji X, Qiu J, Saldi T, Diene K, Jones K, Fu XD, Ben ley DL. 2014. P e-mRNA splicing is acili a ed by an op imal RNA polyme ase II elonga ion a e. Genes De 28: 2663–2676. doi:10.1101/gad.252106.114 Fox-Walsh KL, Dou Y, Lam BJ, Hung SP, Baldi PF, He el KJ. 2005. The a chi ec u e o p e-mRNAs a ec s mechanisms o splice-si e pai - ing. P oc Na l Acad Sci 102: 16176–16181. doi:10.1073/pnas .0508489102 Fukaya T, Lim B, Le ine M. 2017. Rapid a es o Pol II elonga ion in he D osophila emb yo. Cu Biol 27: 1387–1391. doi:10.1016/j.cub .2017.03.069 Ga ney DJ, McVicke G, Pai AA, Fondu e-Mi endo YN, Lewellen N, Michelini K, Widom J, Gilad Y, P i cha d JK. 2012. Con ols o nu- cleosome posi ioning in he human genome. PLoS Gene 8: e1003036. doi:10.1371/jou nal.pgen.1003036 Gel man S, Cohen N, Yea im A, As G. 2013. DNA-me hyla ion e ec on co ansc ip ional splicing is dependen on GC a chi ec u e o he exon-in on s uc u e. Genome Res 23: 789–799. doi:10 .1101/g .143503.112 Gö nemann J, Ba andon C, Huje K, Ru z B, Rigau G, Ko o ic KM, Faux C, Neugebaue KM, Sé aphin B. 2011. Co ansc ip ional spli- ceosome assembly and splicing a e independen o he P p40p WW domain. RNA 17: 2119–2129. doi:10.1261/ na.02646811 G a eley BR, B ooks AN, Ca lson JW, Du MO, Landolin JM, Yang L, A ie i CG, Van Ba en MJ, Boley N, Boo h BW, e al. 2011. The de elopmen al ansc ip ome o D osophila melanogas e . Na u e 471: 473–479. doi:10.1038/na u e09715 Guilgu LG, P udêncio P, Sob al D, Liszeko a D, Rosa A, Ma inho RG. 2014. Requi emen o highly e icien p e-mRNA splicing du ing D osophila ea ly emb yonic de elopmen . Eli e 3: e02181. doi:10.7554/eLi e.02181 Ha len KM, Chu chman LS. 2017. The code and beyond: ansc ip ion egula ion by he RNA polyme ase II ca boxy- e minal domain. Na Re Mol Cell Biol 18: 263–273. doi:10.1038/n m.2017.10 Ha is CR, Millman KJ, an de Wal SJ, Gomme s R, Vi anen P, Cou napeau D, Wiese E, Taylo J, Be g S, Smi h NJ, e al. 2020. A ay p og amming wi h NumPy. Na u e 585: 357–362. doi:10 .1038/s41586-020-2649-2 Ha on AR, Sub amaniam V, Lopez AJ. 1998. Gene a ion o al e na i e Ul abi ho ax iso o ms and s epwise emo al o a la ge in on by esplicing a exon-exon junc ions. Mol Cell 2: 787–796. doi:10 .1016/S1097-2765(00)80293-2 He zel L, S aube K, Neugebaue KM. 2018. Long- ead sequencing o nascen RNA e eals coupling among RNA p ocessing e en s. Genome Res 28: 1008–1019. doi:10.1101/g .232025.117 Hö z W, Al enbu ge W. 1981. Sequence speci ic clea age o DNA by mic ococcal nuclease. Nucleic Acids Res 9: 2643–2658. doi:10 .1093/na /9.12.2643 HoskinsRA,LandolinJM,B ownJB,Sandle JE,TakahashiH, Lassmann T, Boo h BW, Zhang D, Wan KH, Yang L, e al. 2011. Genome-wide analysis o p omo e a chi ec u e in D osophila mela- nogas e .Genome Res 21: 182–192. doi:10.1101/g .112466.110 Hsin JP, Manley JL. 2012. The RNA polyme ase II CTD coo dina es ansc ip ion and RNA p ocessing. Genes De 26: 2119–2137. doi:10.1101/gad.200303.112 James Ken W, Sugne CW, Fu ey TS, Roskin KM, P ingle TH, Zahle AM, Haussle D. 2002. The human genome b owse a UCSC. Genome Res 12: 996–1006. doi:10.1101/g .229102 Ji J-Y, Squi ell JM, Schubige G. 2004. Bo h Cyclin B le els and DNA- eplica ion checkpoin con ol he ea ly emb yonic mi oses in D osophila Jun-Yuan. De elopmen 131: 401–411. doi:10.1242/ de .00944 Jonke s I, Kwak H, Lis JT. 2014. Genome-wide dynamics o Pol II elon- ga ion and i s in e play wi h p omo e p oximal pausing, ch oma- in, and exons. Eli e 3: e02407. doi:10.7554/eLi e.02407 Joseph B, Kondo S, Lai EC. 2018. Sho c yp ic exons media e ecu - si e splicing in D osophila.Na S uc Mol Biol 25: 365–371. doi:10 .1038/s41594-018-0052-6 Ken WJ, Zweig AS, Ba be G, Hin ichs AS, Ka olchik D. 2010. BigWig and BigBed: enabling b owsing o la ge dis ibu ed da ase s. Bioin o ma ics 26: 2204–2207. doi:10.1093/bioin o ma ics/b q351 Ke en H, Le -Mao G, As G. 2010. Al e na i e splicing and e olu ion: di e si ica ion, exon de ini ion and unc ion. Na Re Gene 11: 345–355. doi:10.1038/n g2776 Khodo YL, Rod iguez J, Ab uzzi KC, Tang CHA, Ma MT, Rosbash M. 2011. Nascen -seq indica es widesp ead co ansc ip ional p e- P udêncio e al. 158 RNA (2022) Vol. 28, No. 2 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om mRNA splicing in D osophila.Genes De 25: 2502–2512. doi:10 .1101/gad.178962.111 Khodo YL, Mene JS, Tolan M, Rosbash M. 2012. Co ansc ip ional splicing e iciency di e s d ama ically be ween D osophila and mouse. RNA 18: 2174–2186. doi:10.1261/ na.034090.112 Kwak H, Fuda NJ, Co e LJ, Lis JT. 2013. P ecise maps o RNA polyme - ase e eal how p omo e s di ec ini ia ion and pausing. Science 339: 950–953. doi:10.1126/science.1229386 Kwasnieski JC, O -Wea e TL, Ba el DP. 2019. Ea ly genome ac i a- ion in D osophila is ex ensi e wi h an ini ial endency o abo ed ansc ip s and e ained in ons. Genome Res 29: 1188–1197. doi:10.1101/g .242164.118 La son MH, Mooney RA, Pe e s JM, Windgassen T, Nayak D, G oss CA, Block SM, G eenlea WJ, Landick R, Weissman JS. 2014. A pause sequence en iched a ansla ion s a si es d i es ansc ip ion dynamics in i o. Science 344: 1042–1047. doi:10 .1126/science.1251871 La e JD, Ma solais AJ, Smibe CA, Lipshi z HD. 2015. Regula ion and unc ion o ma e nal gene p oduc s du ing he ma e nal- o-zy- go ic ansi ion in D osophila.Cu Top De Biol.113: 4384. 10 .1016/bs.c db.2015.06.007 Lee JEA, Mi chell NC, Zay se a O, Chahal A, Mendis P, Ca ie - Michaud A, Pa sons LM, Poo inga G, Le ens DL, Hannan RD, e al. 2015. De ec i e H p-dependen ansc ip ional ep ession o dMYC is undamen al o issue o e g ow h in D osophila XPB models. Na Commun 6: 7404. doi:10.1038/ncomms8404 Li H, Handsake B, Wysoke A, Fennell T, Ruan J, Home N, Ma h G, Abecasis G, Du bin R. 2009. The Sequence Alignmen /Map o - ma and SAM ools. Bioin o ma ics 25: 2078–2079. doi:10.1093/ bioin o ma ics/b p352 Lo SE, Villal a JE, Sch o h GP, Luo S, Tonkin LA, Eisen MB. 2011. Noncanonical compensa ion o zygo ic x ansc ip ion in ea ly D oso- phila melanogas e de elopmen e ealed h ough single-emb yo RNA-seq. PLoS Biol 9: e1000590. doi:10.1371/jou nal.pbio.1000590 Lund E, Dahlbe g JE. 1992. Cyclic 2′,3 ′-phospha es and non em- pla ed nucleo ides a he 3′end o spliceosomal U6 small nuclea RNAs. Science 255: 327–330. doi:10.1126/science.1549778 Ma in M. 2011. Cu adap emo es adap e sequences om high- h oughpu sequencing eads. EMBne .J 17: 10–12. doi:10.14806/ ej.17.1.200 Ma in RM, Rino J, Ca alho C, Ki chhausen T, Ca mo-Fonseca M. 2013. Li e-cell isualiza ion o p e-mRNA splicing wi h single-mol- ecule sensi i i y. Cell Rep 4: 1144–1155. doi:10.1016/j.cel ep .2013.08.013 Ma inho RG, Guilgu LG, P udêncio P. 2015. How gene exp ession in as -p oli e a ing cells keeps pace. Bioessays 37: 514–524. doi:10 .1002/bies.201400195 Maslon MM, B aunschweig U, Ai ken S, Mann AR, Kilanowski F, Hun e CJ, Blencowe BJ, Ko nblih AR, Adams IR, Cáce es JF. 2019. A slow ansc ip ion a e causes emb yonic le hali y and pe - u bs kine ic coupling o neu onal genes. EMBO J 38: e101244. doi:10.15252/embj.2018101244 Maye A, Di Iulio J, Male i S, Ese U, Vie s a J, Reynolds A, Sands om R, S ama oyannopoulos JA, Chu chman LS. 2015. Na i e elonga ing ansc ip sequencing e eals human ansc ip- ional ac i i y a nucleo ide esolu ion. Cell 161: 541–554. doi:10 .1016/j.cell.2015.03.010 Mo is DP, G eenlea AL. 2000. The splicing ac o , P p40, binds he phospho yla ed ca boxyl- e minal domain o RNA polyme ase II. J Biol Chem 275: 39935–39943. doi:10.1074/jbc.M004118200 Nojima T, Gomes T, G osso ARF, Kimu a H, Dye MJ, Dhi S, Ca mo- Fonseca M, P oud oo NJ. 2015. Mammalian NET-seq e eals ge- nome-wide nascen ansc ip ion coupled o RNA p ocessing. Cell 161: 526–540. doi:10.1016/j.cell.2015.03.027 Nojima T, Gomes T, Ca mo-Fonseca M, P oud oo NJ. 2016. Mammalian NET-seq analysis de ines nascen RNA p o iles and as- socia ed RNA p ocessing genome-wide. Na P o oc 11: 413–428. doi:10.1038/np o .2016.012 Nojima T, Rebelo K, Gomes T, G osso AR, P oud oo NJ, Ca mo- Fonseca M. 2018. RNA polyme ase II phospho yla ed on CTD se - ine 5 in e ac s wi h he spliceosome du ing co- ansc ip ional splic- ing. Mol Cell 72: 369–379.e4. doi:10.1016/j.molcel.2018.09.004 Pai AA, Hen iques T, McCue K, Bu kholde A, Adelman K, Bu ge CB. 2017. The kine ics o p e-mRNA splicing in he D osophila ge- nome and he in luence o gene a chi ec u e. Eli e 6: e32537. doi:10.7554/eLi e.32537 Pai AA, Paggi JM, Yan P, Adelman K, Bu ge CB. 2018. Nume ous e- cu si e si es con ibu e o accu acy o splicing in long in ons in lies. PLoS Gene 14: e1007588. doi:10.1371/jou nal.pgen .1007588 Pe ea M, Pe ea GM, An onescu CM, Chang TC, Mendell JT, Salzbe g SL. 2015. S ingTie enables imp o ed econs uc ion o a ansc ip ome om RNA-seq eads. Na Bio echnol 33: 290– 295. doi:10.1038/nb .3122 P i cha d DK, Schubige G. 1996. Ac i a ion o ansc ip ion in D osophila emb yos is a g adual p ocess media ed by he nucleo- cy oplasmic a io. Genes De 10: 1131–1142. doi:10.1101/gad.10 .9.1131 P oud oo NJ. 2016. T ansc ip ional e mina ion in mammals: s op- ping he RNA polyme ase II jugge nau . Science 352: aad9926. doi:10.1126/science.aad9926 P udêncio P, Rebelo K, G osso AR, Ma inho RG, Ca mo- onseca M. 2020. Analysis o mammalian na i e elonga ing ansc ip se- quencing (mNET-seq) high- h oughpu da a. Me hods 178: 89– 95. doi:10.1016/j.yme h.2019.09.003 Quinlan AR, Hall IM. 2010. BEDTools: a lexible sui e o u ili ies o compa ing genomic ea u es. Bioin o ma ics 26: 841–842. doi:10.1093/bioin o ma ics/b q033 Ramí ez F, Ryan DP, G üning B, Bha dwaj V, Kilpe F, Rich e AS, Heyne S, Dünda F, Manke T. 2016. deepTools2: a nex gene a ion web se e o deep-sequencing da a analysis. Nucleic Acids Res 44: W160–W165. doi:10.1093/na /gkw257 Reime KA, Mimoso CA, Adelman K, Neugebaue KM. 2021. Co- an- sc ip ional splicing egula es 3′end clea age du ing mammalian e y h opoiesis. Mol Cell 81: 998–1012.e7. doi:10.1016/j.molcel .2020.12.018 Robbe son BL, Co e GJ, Be ge SM. 1990. Exon de ini ion may acil- i a e splice si e selec ion in RNAs wi h mul iple exons. Mol Cell Biol 10: 84–94. doi:10.1128/MCB.10.1.84 Ro he M, Pehl M, Taube H, Jäckle H. 1992. Loss o gene unc ion h ough apid mi o ic cycles in he D osophila emb yo. Na u e 359: 156–159. doi:10.1038/359156a0 Sandle JE, I iza y J, S epanik V, Dunipace L, Am hein H, S a hopoulos A. 2018. A de elopmen al p og am unca es long ansc ip s o empo ally egula e cell signaling. De Cell 47: 773–784.e6. doi:10.1016/j.de cel.2018.11.019 Saunde s A, Co e LJ, Su cli e C, Lis JT, Ashe HL. 2013. Ex ensi e po- lyme ase pausing du ing D osophila axis pa e ning enables high- le el and pliable ansc ip ion. Genes De 27: 1146–1158. doi:10 .1101/gad.215459.113 Schlackow M, Nojima T, Gomes T, Dhi A, Ca mo-Fonseca M, P oud oo NJ. 2017. Dis inc i e pa e ns o ansc ip ion and RNA p ocessing o human lincRNAs. Mol Cell 65: 25–38. doi:10 .1016/j.molcel.2016.11.029 Schulz KN, Bond a ER, Moshe A, Villal a JE, Lieb JD, Kaplan T, McKay DJ, Ha ison MM. 2015. Zelda is di e en ially equi ed o ch oma in accessibili y, ansc ip ion ac o binding, and gene exp ession in he ea ly D osophila emb yo. Genome Res 25: 1715–1726. doi:10.1101/g .192682.115 T ansc ip ion and splicing dynamics in D osophila www. najou nal.o g 159 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om She idan RM, Fong N, D’Alessand o A, Ben ley DL. 2019. Widesp ead back acking by RNA Pol II is a majo e ec o o gene ac i a ion, 5′ pause elease, e mina ion, and ansc ip ion elonga ion a e. Mol Cell 73: 107–118.e4. doi:10.1016/j.molcel.2018.10.031 She moen AW, O’Fa ell PH. 1991. P og ession o he cell cycle h ough mi osis leads o abo ion o nascen ansc ip s. Cell 67: 303–310. doi:10.1016/0092-8674(91)90182-X Sousa-Luís R, Duja din G, Zukhe I, Kimu a H, Weldon C, Ca mo- Fonseca M, P oud oo NJ, Nojima T. 2021. POINT echnology illumi- na es he p ocessing o polyme ase-associa ed in ac nascen an- sc ip s. Mol Cell 81: 1935–1950.e6. doi:10.1016/j.molcel.2021.02.034 S e ne DA, Ca lo T, Be ge SM. 1996. A chi ec u al limi s on spli genes. P oc Na l Acad Sci 93: 15081–15085. doi:10.1073/pnas.93.26.15081 Veloso A, Ki kconnell KS, Magnuson B, Biewen B, Paulsen MT, Wilson TE, Ljungman M. 2014. Ra e o elonga ion by RNA polyme - ase II is associa ed wi h speci ic gene ea u es and epigene ic mod- i ica ions. Genome Res 24: 896–905. doi:10.1101/g .171405.113 Vi anen P, Gomme s R, Oliphan TE, Habe land M, Reddy T, Cou napeau D, Bu o ski E, Pe e son P, Weckesse W, B igh J, e al. 2020. SciPy 1.0: undamen al algo i hms o scien i ic compu ing in Py hon. Na Me hods 17: 261–272. doi:10.1038/s41592-019-0686-2 Vizcaya-Molina E, Klein CC, Se as F, Mish a RK, Guigó R, Co ominas M. 2018. Damage- esponsi e elemen s in D osophila egene a ion. Genome Res 28: 1852–1866. doi:10.1101/g .233098.117 Wan Y, Anas asakis DG, Rod iguez J, Palanga M, Gudla P, Zaki G, Tandon M, Pego a o G, Chow CC, Ha ne M, e al. 2021. Dynamic imaging o nascen RNA e eals gene al p inciples o ansc ip ion dynamics and s ochas ic splice si e selec ion. Cell 184: 2878–2895.e20. doi:10.1016/j.cell.2021.04.012 Windhage L, Bon e T, Bu ge K, Ruzsics Z, K ebs S, Kau mann S, Mal e e G, L’He naul A, Schilhabel M, Sch eibe S, e al. 2012. Ul asho and p og essi e 4sU- agging e eals key cha ac e is ics o RNA p ocessing a nucleo ide esolu ion. Genome Res 22: 2031–2042. doi:10.1101/g .131847.111 Yan D, Neumülle RA, Buckne M, Aye s K, Li H, Hu Y, Yang-Zhou D, Pan L, Wang X, Kelley C, e al. 2014. A egula o y ne wo k o D osophila ge mline s em cell sel - enewal. De Cell 28: 459– 473. doi:10.1016/j.de cel.2014.01.020 Yeo G, Bu ge CB. 2004. Maximum en opy modeling o sho se- quence mo i s wi h applica ions o RNA splicing signals. J Compu Biol 11: 377–394. doi:10.1089/1066527041410418 Yuan K, Selle CA, She moen AW, Fa ell PHO. 2016. Timing he D osophila mid-blas ula ansi ion: a cell cycle-cen e ed iew. T ends Gene 32: 496–507. doi:10.1016/j. ig.2016.05.006 Zhu L, Zhang Y, Zhang W, Yang S, Chen J-Q, Tian D. 2009. Pa e ns o exon-in on a chi ec u e a ia ion o genes in euka y- o ic genomes. BMC Genomics 10: 47. doi:10.1186/1471-2164- 10-47 MEET THE FIRST AUTHORS Ped o P udêncio Rosina Sa isaa Kenny Rebelo Mee he Fi s Au ho (s) is a newedi o ial ea u ewi hinRNA,in which he i s au ho (s) o esea ch-based pape s in each issue ha e he oppo uni y o in oduce hemsel es and hei wo k o eade s o RNA and he RNA esea ch communi y. Ped o P udêncio, Rosina Sa isaa , and Kenny Rebelo a e co- i s au- ho s o his pape , “T ansc ip ion and splicing dynamics du ing ea ly D osophila de elopmen .”Ped o is a pos doc o al ellow a he Ins i u e o Molecula Medicine, Lisbon in he lab o D . Ca mo-Fonseca, wi h a ocus on he in e play be ween an- sc ip ion and splicing. Rosina is also a pos doc o al ellow in he Ca mo-Fonsecag oup, wi hanin e es in hekine icso splicing and how i is linked o he e olu ion o splicing egula o y in o - ma ion. Kenny is a molecula gene ics g adua e s uden a he Uni e si y o To on o in he lab o Benjamin Blencowe, wi h an in e es in unde s anding how nuclea p o ein complexes p o- cess RNA o enable cellula unc ion. Wha a e he majo esul s desc ibed in you pape and how do hey impac his b anch o he ield? In his pape , we b ough he NET-seq me hod ou o he cell dish and in o whole D osophila emb yos (dNETseq), allowing us o ex- plo e se e al aspec s o RNA biology in his model. We s udied ansc ip ion e mina ion and ound e idence ha his p ocess may exhibi mechanis ic di e ences depending on whe he he ansc ip ion uni is isola ed o o e laps a con e gen gene. As genes exp essed ea lie in emb yonic de elopmen a e mo e likely o be isola ed, his means ha p e e ed mechanisms o ansc ip- ion e mina ion may e ol e as de elopmen p og esses. We also add essed p oblems o splicing kine ics. We obse ed ha many in ons a e spliced immedia ely a e ansc ip ion, including e y la ge in ons, sugges ing ha splicing ia in on de ini ion is no limi ed o small in ons. Con inued P udêncio e al. 160 RNA (2022) Vol. 28, No. 2 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om Wha led you o s udy RNA o his aspec o RNA science? PP: My desi e is o unde s and he complex egula ion o gene ex- p ession. I s a ed my scien i ic ca ee wo king in D osophila em- b yogenesis in Rui Ma inho’s lab, whe e we came ac oss an obse a ion ha led o ques ions ha we didn’ ha e he ools o answe . Luckily, and in collabo a ion wi h he Ca mo-Fonseca lab, I had he oppo uni y o lea n his inc edible echnique and o adap i o D osophila emb yos, and wi h his, ha e he chance o look a gene exp ession om a pe spec i e ne e done be o e in his sys em. RS: Splicing is inc edibly complex and inc edibly con using o s udy, and his is wha always a ac ed me o he ield. Fo my PhD, I wo ked on he molecula e olu ion o splicing signals. Se e al g oups, including he Ca mo-Fonseca lab, we e making ascina ing no el indings ega ding he kine ics o splicing. I became cu ious as o how he dynamic na u e o splicing was a - ec ing hese e olu iona y p ocesses, and hence decided o ocus on splicing kine ics o my pos doc o al wo k. KR: Fascina ion o e how simila o ganized and egula ed biolog- ical p ocesses can be o assembly lines in a ac o y. I joined he Ca mo-Fonseca lab du ing my M.Sc. whe e I was able o explo e his in e es when analyzing da a ha opened he cell’s“ ac o y doo ” o s udy ansc ip ion and splicing. Du ing he cou se o hese expe imen s, we e he e any su p ising esul s o pa icula di icul ies ha al e ed you hinking and subsequen ocus? RS: AdNETseq expe imen gi es ise o a bo omless easu e o e o in o ma ion. Howe e , he analysis equi ed o con e ha in o ma ion in o insigh is complex and w ough wi h pe ils. Fo ins ance, an inc ease in he ansc ip ion ini ia ion a e, a dec ease in he elonga ion a e, and a phospho yla ion ho spo can all lead o simila signa u es in he sequencing da a, and one mus be e y ca e ul o keep all o hese al e na i es in mind when pe o ming he analysis. Being con on ed wi h he complex- i y o his da a has made me ealize how much mo e insigh I could gain i I had a la ge oolki o my analyses. I ha e hus sough ou aining in mo e ad anced s a is ical modeling echniques, some- hing ha I may ha e ne e done i I had no encoun e ed dNETseq da a. PP: Explo ing he ea u es o nascen RNA wi h dNETseq equi es a signi ican amoun o biological sample. This made he analysis o ea lie D osophila de elopmen al s ages as planned ex emely di icul . We he e o e ocused on la e s ages o de elopmen . Howe e , NET-seq da a is so ull o in o ma ion ha we s ill ended up e ealing se e al exci ing new ea u es o RNA ansc ip ion and p ocessing. Wha a e some o he landma k momen s ha p o oked you in e es in science o you de elopmen as a scien is ? KR: I was imp essed, in mo e han one unique momen , by my col- leagues’ eadiness o adap and p oduce solu ions when con on - ed wi h unexpec ed o con ounding esul s. By cons an ly challenging mysel and keeping an open mind, I aspi e o ma ch his p oblem-sol ing abili y, which is co e o spea heading you own esea ch. I you we e able o gi e one piece o ad ice o you younge sel , wha would ha be? KR: De elop a p o essional and e icien communica ion me hod. I belie e his is one o he mos pe inen skill se s o es ablish ea ly on because i can wo k o p oac i ely a oid misunde s andings. Fo example, when p esen ing an idea o a colleague o you supe iso be su e o no imply ha you idea is be e and ins ead explain he esul s/e idence o guide he o he pe son o connec he same do s ha you did. Wha a e you subsequen nea - o long- e m ca ee plans? RS: I am cu en ly looking o ways o keep on doing he kinds o hings ha scien is s do, bu as a eelance , away om he classical academic ca ee pa h. This is so as o ha e mo e lexibili y wi h ega d o whe e in he wo ld I can li e bu also o be able o ake on mo e di e se p ojec s. My p ima y in e es a he momen is in me hods o eaching s a is ics, and I am in he p ocess o o - ganizing a ga gan uan s a is ics cou se a my ins i u e. Exci ing imes! PP: The ield has exci ing open ques ions, and al hough echnol- ogy is de eloping exponen ially, di e en echniques a e s ill poin ing in di e en di ec ions. This means ha he e is s ill wo k o be done in he ield o unde s and how hese inc edible and complex mac omolecules ha syn hesize and modi y RNA a e eg- ula ed. Taking ad an age o he la es echnology o analyze hese p ocesses, I expec o use he nex yea s o my pos doc o explo e hese ea u es in mo e de ail and speci ically in he con ex o a mul icellula o ganism. Wha we e he s onges aspec s o you collabo a ion as co- i s au ho s? This was an amazing collabo a ion, p ima ily because ou back- g ounds a e so complemen a y. Each o us was lea ning new hings in e e y con e sa ion ha we had. T ansc ip ion and splicing dynamics in D osophila www. najou nal.o g 161 Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om 10.1261/ na.078933.121Access he mos ecen e sion a doi: 2022 28: 139-161 o iginally published online Oc obe 19, 2021RNA Ped o P udêncio, Rosina Sa isaa , Kenny Rebelo, e al. de elopmen D osophilaT ansc ip ion and splicing dynamics du ing ea ly Ma e ial Supplemen al h p:// najou nal.cshlp.o g/con en /suppl/2021/10/19/ na.078933.121.DC1 Re e ences h p:// najou nal.cshlp.o g/con en /28/2/139. ull.h ml# e -lis -1 This a icle ci es 103 a icles, 30 o which can be accessed ee a : Open Access Open Access op ion.RNAF eely a ailable online h ough he License Commons C ea i e .h p://c ea i ecommons.o g/licenses/by/4.0/ (A ibu ion 4.0 In e na ional), as desc ibed a , is a ailable unde a C ea i e Commons LicenseRNAThis a icle, published in Se ice Email Ale ing click he e. op igh co ne o he a icle o Recei e ee email ale s when new a icles ci e his a icle - sign up in he box a he h p:// najou nal.cshlp.o g/subsc ip ions go o: RNATo subsc ibe o © 2022 P udêncio e al.; Published by Cold Sp ing Ha bo Labo a o y P ess o he RNA Socie y Cold Sp ing Ha bo Labo a o y P ess on No embe 11, 2022 - Published by najou nal.cshlp.o gDownloaded om