The guide sRNA sequence determines the activity level of box C/D RNPs.

*Fo co espondence: e esa.ca [email protected] hanno e .de Compe ing in e es s: The au ho s decla e ha no compe ing in e es s exis . Funding: See page 24 Recei ed: 08 July 2019 Accep ed: 08 Ma ch 2020 Published: 23 Ma ch 2020 Re iewing edi o : Lewis E Kay, Uni e si y o To on o, Canada Copy igh G aziadei e al. This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s un es ic ed use and edis ibu ion p o ided ha he o iginal au ho and sou ce a e c edi ed. The guide sRNA sequence de e mines he ac i i y le el o box C/D RNPs And ea G aziadei 1,2 , F ank Gabel 3,4 , John Ki kpa ick 2,5 , Te esa Ca lomagno 2,5 * 1 Eu opean Molecula Biology Labo a o y, S uc u al and Compu a ional Biology , Heidelbe g, Ge many; 2 Leibniz Uni e si y Hanno e , Cen e o Biomolecula D ug Resea ch, Hanno e , Ge many; 3 Uni e si y G enoble Alpes, CEA, CNRS IBS, G enoble, F ance; 4 Ins i u Laue-Lange in, G enoble, F ance; 5 Helmhol z Cen e o In ec ion Resea ch, G oup o S uc u al Chemis y, B aunschweig, Ge many Abs ac 2’-O- RNA me hyla ion, which is essen ial in euka yo es and a chaea, is ca alysed by he Box C/D RNP complex in an RNA-guided manne . Despi e he conse a ion o he me hyla ion si es, he abundance o si e-speci ic modi ica ions shows a iabili y ac oss species and issues, sugges ing ha RNA me hyla ion may p o ide a means o con olling gene exp ession. As all Box C/D RNPs a e hough o adop a simila s uc u e, i emains unclea how he me hyla ion e iciency is egula ed. He e, we p o ide he i s s uc u al e idence ha , in he con ex o he Box C/D RNP, he a ini y o he ca aly ic module ib illa in o he subs a e–guide helix is dependen on he RNA sequence ou side he me hyla ion si e, hus p o iding a mechanism by which bo h he subs a e and guide RNA sequences de e mine he deg ee o me hyla ion. To each his esul , we de elop an i e a i e s uc u e-calcula ion p o ocol ha exploi s he powe o in eg a i e s uc u al biology o cha ac e ize con o ma ional ensembles. In oduc ion In a wide a ie y o cellula p ocesses, anging om biosyn hesis o signalling and egula ion o gene exp ession, RNA is chemically modi ied bo h co- and pos - ansc ip ionally. All classes o RNA a e modi ied, and RNA p ocessing and edi ing mechanisms a e highly conse ed, wi h mo e han 140 chemical modi ica ions suppo ing RNA unc ion in all h ee domains o li e (Machnicka e al., 2013). In RNA, he mos abundan modi ica ion is 2’-O-me hyla ion, which impac s p e- RNA p oc- essing, ibosome assembly and unc ion. Func ionally, 2’-O-me hyla ion has been shown o p o ec RNA om ibonucleoly ic clea age (He schlag e al., 1993), s abilize single base-pai s, ac as a chape one (Helm, 2006;Williams e al., 2001) and in luence olding a high empe a u es (Kawai e al., 1992). None heless, he exac ole o posi ion-speci ic 2’-O- ibose me hyla ion is mos ly unknown. Recen e idence shows ha , while me hyla ion si es a e la gely conse ed and clus e in unc ion- ally impo an egions o he ibosome (Deca u and Fou nie , 2002), he abundance o modi ied nucleo ides is no uni o m ac oss species, o e en ac oss issues. In humans, one hi d o me hyla ed si es show a iable le els o modi ica ion acco ding o he cell- ype (K ogh e al., 2016). The he e o- geneous ibosome popula ion esul ing om hese di e en me hyla ion le els is consis en wi h he no ion o specialized ibosomes ha ansla e pa icula genes wi h imp o ed e iciency (Xue and Ba na, 2012). In ag eemen wi h i s pu a i e ole in egula ing ansla ion, he complexi y o RNA 2’-O-me hyla ion has inc eased wi h e olu ion: in bac e ia, a p o ein enzyme ca alyses 2’-O-me hyla- ion a a hand ul o RNA si es, while in yeas and humans a small nucleola ibonucleop o ein com- plex ( he Box C/D snoRNP) uses a se o guide RNAs o deposi me hyl g oups in a sequence- speci ic manne a ~50 and 100 RNA si es, espec i ely. G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 1 o 27 RESEARCH ARTICLE Besides hei ole in guiding 2’-O-me hyla ion, Box C/D RNPs a e in ol ed in a a ie y o o he unc ions, anging om RNA p ocessing ( o example, he U3 snoRNP, Kass e al., 1990) o RNA base ace yla ion (Sha ma e al., 2017). Fu he mo e, nea ly hal o all human snoRNPs ha e no p e- dic able RNA a ge s, sugges ing ha hey may ha e o he oles wi hin he cell (Falalee a e al., 2017). Some o hese so-called o phan snoRNPs ha e been associa ed wi h cance and o he dis- eases (Gong e al., 2017;Williams and Fa zaneh, 2012). The a ying le els o me hyla ion measu ed a di e en si es and he in ol emen o he Box C/D RNPs in p ocesses o he han me hyla ion aise he ques ion as o how he enzyma ic ac i i y is egu- la ed o e en silenced in he a ious Box C/D RNPs. The lack o an in i o econs i u ion p o ocol yielding an ac i e snoRNP cu en ly p ecludes mechanis ic and s uc u al s udies o he euka yo ic Box C/D complex. All s uc u al and in i o unc- ional wo k o da e has ocused on he a chaeal Box C/D sRNP (Figu e 1a). The alidi y o his sys- em as a p oxy o he euka yo ic enzyme is es ablished by hei a chi ec u al simila i y and compa able complexi y o he RNA me hyla ion pa e ns (~115 RNA me hyla ion si es a e p e- dic ed in Py ococcus u iosus). In a chaea, Box C/D sRNPs consis o h ee p o eins assembled a ound he guide sRNA (Fig- u e 1— igu e supplemen 1). Wi hin he guide RNA, he highly conse ed box C/D sequence mo i olds in o he kink- u n (K- u n) (Kiss-La ´szlo ´e al., 1998) s uc u e and ec ui s he p o ein L7Ae (Snu13 and 15.5K in yeas and human, espec i ely) (Moo e e al., 2004). By analogy, he less con- se ed box C’/D’ mo i has been p oposed o old in o he kink-loop (K-loop) s uc u e (Noli os e al., 2005), which also binds L7Ae (Gagnon e al., 2010). The guide RNA–L7Ae complex binds he wo C- e minal domains (CTDs) o he homodime Nop5 (he e odime Nop58–Nop56 in yeas and humans), which hen ec ui s wo copies o he me hyla ion enzyme ib illa in (Nop1 and ib illa in in yeas and human, espec i ely) h ough i s N- e minal domains (NTDs). The guide sRNA ecognizes he RNA subs a e sequences a space egions loca ed be ween boxes C and D0and be ween boxes C0and D; once bound o he subs a e, i di ec s me hyla ion o he i h nucleo ide ups eam o ei he box D (subs a e D) o D’ (subs a e D’) (Reichow e al., 2007). In he absence o subs a e RNA (apo o m), he a chaeal Box C/D sRNP has been ound o assemble mainly as a dime ic RNP, comp ising ou copies o each p o ein and wo copies o he guide sRNA (Bleiche e al., 2009) (di-RNP, Figu e 1). Upon sa u a ion o he subs a e RNA bind- ing si es (holo o m), wo oligome ic s a es ha e been epo ed (Figu e 1— igu e supplemen 2): he monome ic RNP (mono-RNP, Lin e al., 2011), con aining wo copies o each p o ein, one guide sRNA and wo subs a e RNAs (Figu e 1— igu e supplemen 2a), and he dime ic RNP (di-RNP, Lapinai e e al., 2013), con aining ou copies o each p o ein, wo guide sRNAs and ou subs a e RNAs (Figu e 1— igu e supplemen 2b). Whe he he exis ence o bo h mono- and di-RNP o ms is me ely a consequence o he di e en expe imen al se -ups in i o o has a unc ional ele ance in i o emains an open ques ion (Yu e al., 2018). In any case, he monome ic sRNP is belie ed o be a be e ep esen a ion o he euka yo ic sys em, as snoRNPs ha e ne e been shown o assemble in o dime s, and he s uc u e o he U3 snoRNP bound o a p e- ibosomal complex displays a mono-RNP a chi ec u e (Cheng e al., 2017). The le els o me hyla ion ca alysed by sRNP complexes in i o a y acco ding o he subs a e sequence. In ea ly s udies he e iciency o 2’-O-me hyla ion in i o was p oposed o depend on he s abili y o he subs a e–guide duplex and on he o ma ion o an ideal A- o m helical geome y close o he modi ica ion si e (Appel and Maxwell, 2007). Using he Py ococcus u iosus (P ) sR26 guide RNA, whose co esponding sRNP me hyla es subs a e D’ mo e e icien ly han subs a e D, we demons a ed ha me hyla ion le els depend on — among o he ac o s — he na u e o he i s base-pai ed nucleo ide o he subs a e (G aziadei e al., 2016). The obse a ion ha subs a e D’, wi h a 5’-u idine, displays good u no e in all condi ions, while u no e o subs a e D, wi h a 5’- guanosine, equi es binding o subs a e D’ (G aziadei e al., 2016), led us o sugges ha he na u e o he las base-pai be o e he box D (o box D’) egula es p oduc dissocia ion. In ag ee- men wi h he hypo hesis ha me hyla ion le els a e no exclusi ely dependen on he s abili y o he subs a e–guide duplex, a ecen s udy, which quan i ied si e-speci ic RNA me hyla ion in wo di e en human cell lines (K ogh e al., 2016), e ealed ha me hyla ion le els in i o do no co e- la e wi h ei he he numbe o base-pai s o he s abili y o he subs a e–guide helix. He e we demons a e ha he sequence o he subs a e–guide duplex in luences he a ini y o ib illa in o he subs a e and ha he ex en o ib illa in binding co ela es wi h he e iciency o G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 2 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics Fib Nop5-NTD L7Ae Nop5-CTD Nop5-CC sRNA di-RNP apo sRNA Fib illa in Nop5 dime NTD CTD L7Ae 5' 3' BoxC p e- RNA a K- u n K-loop Subs a e D' Subs a e D BoxD BoxD' BoxC' di-RNP holo Subs a e D Subs a e D' mono-RNP holo b 5.2 5.4 s -sR26 apo R g54.3±0.1 Å holo R g44.9±0.1 Å S2 10 100 0.02 0.04 0.06 0.08 0.1 0.12 0.14 S (Å -1) ln(In ensi y) 2×10-4 4×10-4 6×10-4 In ensi y s -sR26 5'-GCGAGCAAUGAUGA ACUGA G A A A 3'-CGCUCG AGUCU AGUAG UA Subs a e D sR26 U Box C Box C' Box D' Box D Subs a e D' G U G A U G G G C G A C U A G U G G A G G C c 5'-GCGAGCA AUGAUGA G U G A U G G G C G A ACUGA G A A A GCUC 3'-CGCUCG AGUC AGUAG UA ACGAG A G C G G G U A G U G ssR26 A C U A G U G G A G G C 5'-GCGAGCA AUGAUGA G U G A U G G G C G A ACUGA G A A A GCUC 3'-CGCUCG AGUC AGUAG UA ACGAG A Figu e 1. Oligome ic assembly s a es o he a chaeal Box C/D RNP. (a) Top-le : molecula componen s o he a chaeal Box C/D sRNP. Top- igh : schema ic model o he apo sRNP. Bo om-le : schema ic model o he holo mono-RNP om Lin e al. (2011) Bo om- igh : schema ic model o he holo di-RNP om Lapinai e e al. (2013). NTD: N- e minal domain; CTD: C- e minal domain; CC: coiled-coil. (b) Two RNA sequences (s -sR26 and ssR26) we e de i ed om he P sR26 RNA and used o assemble he Box C/D sRNPs ei he in his (s -sR26) o p e ious s udies (ssR26, Lapinai e e al., 2013). The sequence o s -sR26 is de i ed om he na i e sR26 RNA by subs i u ion o he apical K-loop elemen wi h he mo e s able K- u n elemen . (c) SAXS cu es wi h Guinie plo s in he inse s o he Box C/D sRNPs econs i u ed wi h s -sR26 be o e (apo) and a e (holo) addi ion o 1.25 equi alen s o each o subs a e D and D’ a a concen a ion o 2 mg/ml. The ansi ion om an apo di-RNP o a holo mono-RNP is e iden om he espec i e R g alues (Figu e 1— igu e supplemen 4). The da a was collec ed a 40˚C. All cu es a e scaled o he same o wa d sca e ing in ensi y. The online e sion o his a icle includes he ollowing igu e supplemen (s) o igu e 1: Figu e supplemen 1. Conse a ion o he Box C/D RNP be ween a chaea and yeas . Figu e 1 con inued on nex page G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 3 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics me hyla ion. Using nuclea magne ic esonance (NMR), small angle X- ay (SAXS) and neu on (SANS) sca e ing da a, we demons a e ha , in he con ex o he sRNP complex, he a ini y o ib illa in o he subs a e depends on he RNA sequence beyond he me hyla ion si e. This di e ence in a ini y is explained by he ene ge ics o a global con o ma ional ansi ion o he sRNP om an inac- i e o an ac i e s a e and p o ides a u he ou e, besides he modula ion o p oduc dissocia ion desc ibed p e iously (G aziadei e al., 2016), o une RNA me hyla ion le els. To de i e hese esul s we de eloped an ensemble s uc u e-calcula ion me hod ha exploi s he abili y o in eg a- i e s uc u al biology in solu ion o e eal and cha ac e ize con o ma ional equilib ia. Resul s S uc u e de e mina ion o he hal -loaded mono-RNPs To unde s and he easons o he highe e iciency o subs a e D’ me hyla ion as compa ed o sub- s a e D in he P sR26 RNP we se ou o de e mine he s uc u e o he co esponding hal -loaded sRNPs, bound o ei he subs a e D o subs a e D’. We used a s abilized e sion o he P sR26 guide RNA, whe e he apical K-loop has been subs i u ed by a K- u n sequence (s abilized sR26, s - sR26, Figu e 1b). This modi ica ion was necessa y o ensu e ha he complex emains s ably assem- bled o e se e al days a 55˚C, as equi ed by he NMR expe imen s, and does no a ec he oligo- me iza ion s a e o he complex (Figu e 1 and Figu e 1— igu e supplemen 3). Fi s , we de e mined he oligome iza ion s a e o he RNP complexes assembled wi h s -sR26 om hei adius-o -gy a ion (R g ), measu ed by SAXS o SANS. To es ima e he compa ibili y o expe imen ally de e mined R g alues wi h he mono- o di-RNP assembly s a es, we e alua ed he heo e ical R g dis ibu ions o 5000 di-RNP models wi h andomized posi ions o he ib illa in copies no bound o he RNA in bo h apo and holo ( ully-loaded) con o ma ions om Lapinai e e al. (2013);Figu e 1— igu e supplemen 4). We ob ained a mean R g alue o 55.9 A ˚wi h a s anda d de ia ion (SD) o 2.0 A ˚ o he apo di-RNP and a mean R g o 58.1 ±3.6 A ˚ o he holo di-RNP. The SAXS cu es o he apo sRNP assembled wi h s -sR26 (Figu e 1c) co espond o a adius-o -gy a ion (R g ) o 54.3 A ˚, which is consis en wi h a di-RNP a chi ec u e (Figu e 1— igu e supplemen 4). Addi- ion o 1.25 mola equi alen s o ei he subs a e D o D’ educes he R g om 54.3 A ˚ o 50.0 o 47.3 A ˚, espec i ely, wi h a u he educ ion o 45.0 A ˚, upon addi ion o bo h subs a es (holo s a e) (G aziadei e al., 2016). These adii a e no longe compa ible wi h a di-RNP, demons a ing ha bo h he hal -loaded and holo s -sR26 complexes a e mono-RNPs (Figu e 1— igu e supplemen 4). The same ansi ion om a di-RNP o a mono-RNP occu ed o he Box C/D RNP assembled wi h sR26 upon subs a e RNA binding (Figu e 1— igu e supplemen 3). This is di e en om he holo complex assembled p e iously in ou labo a o y wi h he ssR26 RNA (symme ic and s abilized sR26), which con ains wo subs a e D’ RNA binding si es o he same sequence (Figu e 1b and Fig- u e 1— igu e supplemen 2b). The RNP assembled wi h ssR26 emained a di-RNP a e sa u a ion o he subs a e RNA binding si es (Lapinai e e al., 2013). Be o e emba king upon he s uc u al s udy o he sRNPs con aining s -sR26, we wan ed o unde s and which elemen s a e esponsible o he di e en oligome iza ion s a es o he holo ssR26- and holo s -sR26-RNPs. The ssR26 and he s -sR26 RNAs di e only in he sequence o he guide RNA a he box D posi ion, which in he case o ssR26 is iden ical o ha o guide D’. Thus, we gene a ed wo addi ional guide RNAs wi h dis inc D and D’ sequences, s -sR26-1 and s -sR26-2: in s -sR26-1 (s -sR26-2), guide sequence D is a chime ic sequence, o med by he 5’ hal o s -sR26 guide D (s -sR26 guide D’) and he 3’ hal o s -sR26 guide D’ (s -sR26 guide D) (Figu e 1— igu e supplemen 5a). In e es ingly, he Box C/D enzyme con aining s -sR26-1 main ained he di-RNP Figu e 1 con inued Figu e supplemen 2. The mono- and di-RNP s a es o he a chaeal holo sRNP. Figu e supplemen 3. The sRNP assembled wi h sR26 has he same oligome iza ion beha iou as he sRNP assembled wi h s -sR26. Figu e supplemen 4. Ranges o adii o gy a ion o he mono- and di-RNP s a es o he a chaeal sRNP. Figu e supplemen 5. Dependence o he oligome ic s a e o he holo sRNPs on he subs a e- ecogni ion sequence o he guide RNA. Figu e supplemen 6. The SANS cu es o 2 H-Fib indica e he p esence o ei he ou o wo copies o ib illa in in he apo and subs a e-loaded s - sR26 RNPs, espec i ely. G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 4 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics a chi ec u e upon binding o ei he subs a e RNAs, while he sRNP con aining s -sR26-2 ansi ioned o he mono-RNP s a e (Figu e 1— igu e supplemen 5b). Mu a ion o he las nucleo ide o s - sR26-1 guide D o ei he C o U (A61C and A61U wi h complemen a y subs a e D) did no pe u b he di-RNP a chi ec u e (Figu e 1— igu e supplemen 5c). We conclude ha he guide sequence s ongly in luences he oligome iza ion s a e o he holo complex. Fu he e idence o he monome ic s a e o hal -loaded and holo s -sR26 complexes eme ges om he P( ) dis ibu ion calcula ed om he SANS cu e o he complexes assembled wi h 2 H- ib il- la in in 42%:58% D 2 O:H 2 O sol en : he numbe and ela i e in ensi ies o he maxima a e compa ible wi h he p esence o wo ib illa in copies bu incompa ible wi h he p esence o ou (Figu e 1— ig- u e supplemen 6). As monome ic complexes, he subs a e-loaded s -sR26 RNPs can se e as p ox- ies o he euka yo ic snoRNP. As we showed p e iously (G aziadei e al., 2016), he sRNP assembled wi h his RNA ca alyses he me hyla ion o he subs a e D’ mo e e icien ly han sub- s a e D, in a simila manne o he na i e P sR26 RNP. To in es iga e whe he he di e ence in me hyla ion e iciency o subs a e D and D’ co ela es wi h s uc u al di e ences, we assembled he Box C/D RNP wi h he s -sR26 guide RNA and sa u- a ed ei he i s D o D’ guide si e (Figu e 1b) o ob ain wo hal -loaded mono-RNPs. We hen de e - mined hei s uc u es in solu ion, whe e he con o ma ional dynamics o he complexes a e p ese ed. The mono-RNPs a e ~190 kDa in size and hus no amenable o s anda d s uc u e de e - mina ion by NMR. In his molecula -weigh ange, solu ion NMR ocuses on me hyl-g oup esonan- ces, which ha e a ou able elaxa ion p ope ies and show s ong signal in ensi y (Sp ange s and Kay, 2007;Tuga ino e al., 2003) Thus, o sol e he s uc u e o he wo hal -loaded sRNPs, we used a combina ion o me hyl-g oup NMR spec oscopy and small-angle sca e ing (see Me hods and Ca lomagno, 2014). As in ou ea lie wo k on he ully-loaded di-RNP complex (Lapinai e e al., 2013), we s a ed om he assump ion ha he in e ac ion in e ace o he Nop5-CTD wi h he L7Ae–K- u n-RNA com- plex and ha o he Nop5-NTD wi h ib illa in do no change wi h espec o hose obse ed in he espec i e c ys al s uc u es (Liu e al., 2007;Xue e al., 2010;Ai aleb e al., 2003). To alida e his assump ion we acqui ed wo-dimensional 1 H- 13 C co ela ion spec a o ib illa in and L7Ae labelled speci ically a he me hyl g oups o Ile, Val and Leu esidues (Tuga ino and Kay, 2003). The chemical shi pe u ba ions measu ed o L7Ae in he Box C/D mono-RNP wi h espec o L7Ae in he L7Ae–K- u n-sRNA complex map o he p e iously desc ibed in e ace be ween L7Ae and he Nop5-CTD (Xue e al., 2010;Figu e 2— igu e supplemen 1). Simila ly, he chemical-shi pe u - ba ions measu ed o ib illa in in he Nop5-NTD– ib illa in complex wi h espec o ee ib illa in map o he in e ac ion in e ace obse ed in p e ious c ys al s uc u es (Ai aleb e al., 2003). These CSPs a e conse ed in he Nop5– ib illa in complex and in he apo Box C/D mono-RNP (Figu e 2— igu e supplemen 2), demons a ing ha ib illa in in e ac s exclusi ely wi h he Nop5-NTD in all complexes. We hen used he signals om he L7Ae and ib illa in me hyl g oups o measu e pa amagne ic elaxa ion enhancemen s (PREs). In his echnique, a pa amagne ic ag (spin-label) ca ying an unpai ed elec on is coupled o a unique cys eine enginee ed on one p o ein subuni wi hin he com- plex. The PREs elici ed on he me hyl g oups o a second p o ein subuni by he unpai ed elec on a e ansla ed in o dis ance es ain s (Ba is e and Wagne , 2000), which de ine he posi ion and ela i e o ien a ion o he wo subuni s in he complex. Fo he D-loaded (D’-loaded) mono-RNP, we collec ed a o al o 407 (442) PREs using spin-labels on L7Ae-Q45C, L7Ae-E58C/C68S, L7Ae-C68, Nop5-E196C, Nop5-D247C and Nop5-S343C while obse ing he me hyl esonances o ib illa in and on Nop5-E65C while obse ing he me hyl esonances o L7Ae (Figu e 2— igu e supplemen 3a). The PRE da a we e alida ed by means o in a-molecula PREs wi hin he igid ib illa in module (Figu e 2— igu e supplemen 4). The excellen i be ween he expe imen al PRE in ensi y a ios and hose p edic ed om he known dis ances con i ms he eliabili y o he PRE-de i ed in e - molecula dis ances. A second class o s uc u al es ain s was de i ed om SANS cu es acqui ed wi h con as - ma ching. In hese expe imen s one o mo e p o eins in he complex a e 2 H-labelled and con ibu e o he obse ed sca e ing signal, while he sca e ed in ensi y o he unlabelled p o eins is masked by he sol en , which is p epa ed as a 42%:58% D 2 O:H 2 O mix u e. A combina ion o such da ase s p o ides su icien in o ma ion o es ain he ela i e posi ion o se e al molecules wi hin a mul i- subuni complex. In ou case we acqui ed SANS cu es o 2 H-L7Ae, 2 H-Nop5, 2 H-Fib, 2 H-RNA, 2 H- G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 5 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics Fib/ 2 H-RNA and 2 H(70%)-Nop5/ 2 H-RNA in 42%:58% D 2 O:H 2 O (Figu e 2— igu e supplemen 3b). In addi ion, we also collec ed SAXS cu es, which epo on he shape o he en i e complexes. These da a we e hen inco po a ed in o a s uc u e-calcula ion p o ocol adap ed om ha de el- oped in ou p e ious s udy (Lapinai e e al., 2013) ( o a desc ip ion o he adap ed p o ocol, e e o Me hods and Figu e 3— igu e supplemen 1). We used he con o ma ions o he modules L7Ae–K- u n-sRNA–Nop5-CTD and Nop5-NTD– ib illa in obse ed in p e ious c ys al s uc u es, and es ic ed ou con o ma ional sea ch o he ela i e o ien a ions o he h ee domains o Nop5, he con o ma ion o he sRNA in pa s o he han he K- u n mo i s and A- o m helices and he ela i e posi ions o he wo copies o each p o ein in he mono-RNP. Con o ma ion o he hal -loaded mono-RNPs in solu ion The me hyl-g oup NMR spec um o ib illa in in he apo RNP assembled wi h s -sR26 is iden ical o he spec um o he RNP assembled wi h ssR26 (Figu e 2a, le panel). This was expec ed, as in bo h di-RNPs all ou ib illa in copies a e a om he RNA and hus hei chemical shi s a e independen o he RNA sequence used o assemble he complex. Me hyl g oups a e a he spa se in he p o ein su aces in ol ed in ecogni ion o he RNA back- bone; in he RNA-bound o m o ib illa in, only he me hyl g oups o V35, I82, V110, L114, I117, V151 and V185 a e expec ed o be wi hin 8 A ˚o he RNA, while only V110 should be close han 5 A ˚. The e o e, he chemical shi pe u ba ions (CSPs) o ib illa in upon RNA binding should be ew and ela i ely small in magni ude. As expec ed, he me hyl-g oup NMR spec um o he subs a e- bound RNPs showed only mode a e CSPs; none heless, hese we e mainly localized in he spec al egion con aining V110, V151 and V185, hus con i ming ha ib illa in ecognizes he subs a e D’– guide duplex (Figu e 2a, igh panel). Fu he e idence o subs a e–guide ecogni ion by ib illa in was p o ided by he PRE da a. As shown in Figu e 2b o subs a e D’, upon ib illa in binding o he subs a e–guide duplex (on-s a e, uppe le ), he Nop5-E65C spin-label ( ed) comes close o one L7Ae copy and would lead o PRE in ensi y- a ios o less han 0.8 o he L7Ae-ILV esidues shown as yellow sphe es. In con as , when ib illa in is no bound o he subs a e–guide duplex (o -s a e, uppe igh ), he Nop5-E65C spin- label is a om L7Ae and canno cause any PRE a enua ion o L7Ae peaks. Thus, he low PRE in en- si y- a ios obse ed expe imen ally o he me hyl g oups o he esidues ma ked in yellow (Figu e 2b, bo om) indica es he p esence o con o me s in which ib illa in is bound o he sub- s a e–guide duplex. In an hal -loaded mono-RNP, one ib illa in copy is necessa ily in he o -s a e, due o he lack o he co esponding subs a e; he second ib illa in copy could be ei he s ably bound o he sub- s a e–guide duplex (yielding a complex in he [on,o ]-con o ma ion) o exchanging be ween he on- and o -s a es (co esponding o he RNP exchanging be ween he RNP [on,o ]- and [o ,o ]- con o ma ions, Figu e 2c). The NMR da a a e quali a i ely compa ible wi h bo h scena ios, as he b oad line-wid hs and he o e lap o he ib illa in NMR peaks ha show he la ges CSPs upon RNA binding p eclude a quan i a i e analysis o he magni ude o he CSPs in e ms o ela i e p opo - ions o he wo con o ma ions. Thus, we decided o conside bo h scena ios in he in e p e a ion o he s uc u al da a. S uc u e calcula ions To de e mine he [on,o ]- and [o ,o ]-con o ma ions o bo h he subs a e D- and D’-loaded sRNPs, we adap ed ou p e iously de eloped s uc u e-calcula ion p o ocol (Lapinai e e al., 2013). We ini- ially pe o med wo s uc u e calcula ions pe complex: in he i s calcula ion, we imposed he es ain ha one ib illa in copy is in con ac wi h he co esponding subs a e–guide duplex, while he o he copy is no ([on,o ]-s a e); in he second calcula ion, we le bo h ib illa in copies ee o adop any posi ion compa ible wi h he PRE da a ([o ,o ]-s a e). We hen ecu si ely binned he PRE-de i ed dis ance- es ain s in o wo se s, acco ding o hei compa ibili y wi h he he [on,o ]- o [o ,o ]-con o ma ions (Figu e 3— igu e supplemen 1). The majo i y o es ain s we e ound o be consis en wi h bo h s a es and he e o e appea ed in bo h se s. One no able excep ion is he se o PRE es ain s de i ed om he me hyl-g oups o L7Ae in he p esence o spin-labelled Nop5- E65C, which a e compa ible only wi h ib illa in being in con ac wi h he subs a e–guide duplex (Figu e 2b). In o al, we pe o med ou s uc u e-calcula ion uns, wo o each o he hal -loaded G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 6 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics 0.0 5.0-0.1 5.01.5 s -sR26 apo ssR26 apo 13C] (ppm) 1H] (ppm) a s -sR26 apo s -sR26 D'-loaded 0.0 5.0- 5 10 15 20 25 1.001.101.201.30 18.0 19.0 20.0 21.0 V223 V185 V110 0.1 5.01.5 s -sR26 apo s -sR26 D'-loaded V151 b Fib-on Fib-oc k1 k-1 Residue numbe Ipa a Idia / Subs a e D-loaded Subs a e D’-loaded [ [ [on, o ]-s a e [on, o ]-s a e [o , o ]-s a e 0 0.2 0.4 0.6 0.8 1.0 1.2 0 20 40 60 80 100 120 Fib- on b-o  Figu e 2. NMR and SAS o he hal -loaded s -sR26 RNPs. (a) Le , o e lay o ILV-me hyl 1 H- 13 C spec a o ib illa in in he apo ssR26 ( u quoise) and apo s -sR26 (blue) RNPs. In bo h di-RNPs, all ou ib illa in copies a e dis an om he RNA and he wo spec a a e iden ical. Middle, o e lay o ILV-me hyl 1 H- 13 C spec a o ib illa in in he apo s -sR26 (blue) and subs a e D’-loaded s -sR26 (g een) RNPs. Righ , expanded iew o he o e lay o ILV-me hyl 1 H- 13 C spec a o ib illa in in he apo s -sR26 (blue) and subs a e D’-loaded s -sR26 (g een) RNPs. (b) Le , s uc u al snapsho s o he on- (le ) and o - ( igh ) s a es o one ib illa in copy in he subs a e D’-loaded mono-RNP. Upon binding o ib illa in o he subs a e–guide duplex, he Nop5-E65C spin-label ( ed) comes close o one L7Ae copy (g een), leading o PRE in ensi y- a ios below 0.8 o he L7Ae-ILV esidues shown as yellow sphe es. In con as , when ib illa in is in he o -s a e ( igh ), he Nop5-E65C spin-label is a om L7Ae and canno induce any PRE-media ed a enua ion o peak in ensi ies. Colou -code as in Figu e 1. Righ , PRE e ec s (I pa a /I dia , a io o he peak in ensi ies when he spin-label is in he pa amagne ic and diamagne ic s a e, espec i ely) o he Nop5-E65C ag on he L7Ae-ILV peaks in he subs a e D-bound ( ed) and subs a e D’-bound (blue) mono- RNPs. The yellow ba s indica e he esidues ep esen ed as yellow sphe es in he le panel. (c) Le , ca oon ep esen a ion o he [on,o ]-con o me o he subs a e D’-loaded mono-RNP; igh , ca oon ep esen a ion o he con o ma ional equilib ium be ween he [on,o ]- and [o ,o ]-con o me s o he same complex. The online e sion o his a icle includes he ollowing igu e supplemen (s) o igu e 2: Figu e supplemen 1. L7Ae in he Box C/D mono-RNP main ains he p e iously de e mined in e ac ion in e aces wi h Nop5-CTD. Figu e supplemen 2. Fib illa in in he Box C/D mono-RNP main ains he p e iously de e mined in e ac ion in e aces wi h Nop5-NTD. Figu e supplemen 3. Schema ic summa y o he expe imen al da a. Figu e supplemen 4. Valida ion o he PRE-de i ed dis ances. G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 7 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics complexes. The [on,o ]-con o ma ions we e compa ible wi h nea ly all PRE-de i ed es ain s (400 ou o 407 o he subs a e D-loaded and 436 ou o 442 o he subs a e D’-loaded RNP, espec- i ely), while he [o ,o ]-con o ma ions we e compa ible wi h 364 and 414 es ain s o he sub- s a e D- and subs a e D’-loaded RNP, espec i ely. Each indi idual s uc u e calcula ion p oceeded h ough a global and a local sea ch s age. A each s age, he o al and dis ance- es ain ene gies as well as he back-calcula ed i s o he SANS cu es we e used o s uc u e selec ion. The wo inal s uc u e ensembles co esponding o he [on,o ]-s a es (Figu e 3) a e de ined o a p ecision o be e han 2.5 A ˚( oo -mean-squa e-de ia ion, RMSD, o he p o ein Caand RNA P a oms, excluding lexible egions). When compa ed o he exis ing s uc u e o he holo mono-RNP om Sul olobus sol a a icus (PDB en y 3pla, Lin e al., 2011), he subs a e D- and subs a e D’-loaded complexes show a easonable simila i y (Figu e 3— igu e supplemen 2). All majo ea u es o he subs a e-bound si e a e conse ed: he RNA-guide sequences lie on he coiled-coil Nop5 domain a an angle o abou 70˚and he C- e minal ip o L7Ae is in p oximi y o he sho Nop5 b-shee 77–79 and a-helix 64–73. Howe e , he solu ion s uc- u es di e om he c ys allog aphic s uc u e in many de ails, demons a ing ha he sRNP a chi- ec u e is lexible enough o adap o di e en guide- and subs a e-RNAs. As expec ed, a signi ican di e gence om he s uc u e o PDB en y 3pla is obse ed in he subs a e-unbound hal o he complexes. Impo an ly, nei he he [on,o ]- no he [on,on]-ensemble a e able o ep oduce he combina ion o PRE and SAS da a sa is ac o ily o each o he subs a e D- o he subs a e D’-loaded RNPs. The PRE in ensi y- a ios measu ed o he Nop5-NTD-E65C mu an on he me hyl-g oups o L7Ae indi- ca e he p esence o con o me s in he [on,o ]-s a e. In ag eemen wi h his, he [on,o ]-s uc u es o Figu e 3 ep oduce he PRE da a easonably well bo h o he subs a e D- and subs a e D’- loaded complexes (Figu e 3— igu e supplemen s 3 and 4). Howe e , hese s uc u es a e unable o i he 2 H-Fib SANS, 2 H-Fib/ 2 H-RNA SANS and SAXS cu es in a sa is ac o y manne (Figu e 3— igu e supplemen 5). Thus, he combina ion o PRE and SAS da a is incompa ible wi h a single s a e o each o he subs a e D- o subs a e D’-loaded RNPs, bu a he e eals he p esence o con o ma ional ensembles. Con o ma ional ensembles Because he SAS da a ha a e in disag eemen wi h he [on,o ]-con o ma ions o Figu e 3 all epo on he posi ion o he ib illa in copies in he complexes, we deduced ha he con o ma ional equi- lib ia p esen in solu ion mus be ela ed o he posi ion o ib illa in. Di e en ypes o con o ma- ional equilib ia a e concei able. In he simples scena io, only he ib illa in in he o -s a e samples mul iple con o ma ions, wi h he second ib illa in emaining s ably in he on-s a e; in a mo e com- plex scena io, he second ib illa in copy may sample bo h he on- and o -s a es (in addi ion o he con o ma ional lexibili y o he ib illa in copy in he o -s a e). To ep esen bo h scena ios and ob ain s uc u al ensembles compa ible wi h bo h PRE and SAS expe imen al da a, we de eloped an ensemble sco ing p o ocol (Figu e 3— igu e supplemen 1b, Me hods). Fo bo h he subs a e D- and subs a e D’-loaded RNPs, we used ep esen a i e s uc- u es o he [on,o ]- and [o ,o ]-s a e ensembles (Figu e 3) — de ined as he s uc u e closes o he mean s uc u e — as s a ing poin s o gene a e ou se s o ~4000 con o ma ions, in which he posi ions o he Nop5-NTD– ib illa in uni s no bound o he subs a e–guide duplex we e andom- ized, in o de o accoun o hei lexibili y. We hen used a pseudo-gene ic algo i hm o selec ensembles o ei he exclusi ely [on,o ]-con o me s o o bo h [on,o ]- and [o ,o ]-con o me s ha bes i he PRE da a, as well as he 2 H-Fib and 2 H-Nop5 SANS, 2 H-Fib/ 2 H-RNA SANS, 2 H(70%)- Nop5/ 2 H-RNA SANS and SAXS cu es (Figu e 3— igu e supplemen 1). Con o ma ional ensemble o he subs a e D’-loaded sRNP Despi e he easonable i o he PRE in ensi y a ios o he subs a e D’-loaded sRNP wi h he ep e- sen a i e s uc u e o he [on,o ]-con o me s o (Figu e 3;Figu e 3— igu e supplemen 3), he la ge R g o he expe imen al 2 H-Fib SANS cu e wi h espec o he heo e ical one indica ed he p esence o con o me s whe e he wo copies o ib illa in a e mo e dis an om each o he han in his se o [on,o ]-con o me s (Figu e 3— igu e supplemen 5). G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 8 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics [on,o ]-con o me s 12 s uc u es (2.4 Å) [o ,o ]-con o me s 3 s uc u es (4.6 Å) [o ,o ]-con o me 1 s uc u e (N/A) [on,o ]-con o me s 20 s uc u es (2.4 Å) s -sR26 + subs a e D' s -sR26 + subs a e D me s 2.4 Å ) [ on,o ] -con o me s 20 s uc u e s (2.4 Å ) ubs a e D s -sR26 + subs a e D [ on o ] con o m me s Figu e 3. Ensembles o s uc u es in ag eemen wi h he expe imen al da a o he [on,o ]- and [o ,o ]-s a es o subs a e D’- and subs a e D-loaded sRNPs. The RMSD alues o each ensemble (in pa en heses) a e calcula ed as he a e age o he RMSD alues o he ensemble s uc u es wi h espec o he s uc u e closes o he mean o e he Caand P a oms o he p o ein and RNA s uc u ed domains, including he ib illa in uni s no bound o he RNA. Colou -code as in Figu e 1. The online e sion o his a icle includes he ollowing igu e supplemen (s) o igu e 3: Figu e supplemen 1. S uc u e-calcula ion algo i hms. Figu e supplemen 2. S uc u es o he hal -loaded sRNPs in he [on,o ]-s a e. Figu e supplemen 3. Fi o indi idual [on,o ]- o [o ,o ]-con o me s o he PRE da a o he subs a e D’-loaded sRNP. Figu e supplemen 4. Fi o indi idual [on,o ]- o [o ,o ]-con o me s o he PRE da a o he subs a e D-loaded sRNP. Figu e supplemen 5. Fi o indi idual [on,o ]- o [o ,o ]-s a es o he SAS da a. G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 9 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics Ma e ials and me hods Key esou ces able Reagen ype (species) o esou ce Designa ion Sou ce o e e ence Iden i ie s Addi ional in o ma ion S ain, s ain backg ound (Esche ichia coli) BL21 (DE3) EMBL p o ein exp ession acili y NA S ain, s ain backg ound (Esche ichia coli) BL21 Rose a 2 Me ck Millipo e Ca #71400–3 Recombinan DNA eagen pETM-11 Fib illa in (plasmid) Lapinai e e al. (2013) N e minal His6 + TEV si e Recombinan DNA eagen pETM-11 Nop5 (plasmid) Lapinai e e al. (2013) N e minal His6 + TEV si e; L113K V223E mu an . Codon-op imised syn he ic gene (GeneA ) Recombinan DNA eagen pETM-11 Nop5 E65C (plasmid) Lapinai e e al. (2013) Mu a ion o pETM-11 Nop5 Recombinan DNA eagen pETM-11 Nop5 E196C (plasmid) Lapinai e e al. (2013) Mu a ion o pETM-11 Nop5 Recombinan DNA eagen pETM-11 Nop5 D247C (plasmid) Lapinai e e al. (2013) Mu a ion o pETM-11 Nop5 Recombinan DNA eagen pETM-11 Nop5 S343C (plasmid) Lapinai e e al. (2013) Mu a ion o pETM-11 Nop5 Recombinan DNA eagen pETM-11 L7Ae (plasmid) Lapinai e e al. (2013) N e minal His6 + TEV si e Recombinan DNA eagen pETM-11 L7Ae Q45C (plasmid) Lapinai e e al. (2013) Mu a ion o pETM-11 L7Ae also ca ying C68S mu a ion Recombinan DNA eagen pETM-11 L7Ae E58C (plasmid) Lapinai e e al. (2013) Mu a ion o pETM-11 L7Ae also ca ying C68S mu a ion Sequence- based eagen s -sR26 G aziadei e al. (2016) In i o ansc ibed RNA Sequence- based eagen s -sR26-1 This pape In i o ansc ibed RNA Me hod sec ion: RNA syn hesis Sequence- based eagen s -sR26-1 subs a e This pape In i o ansc ibed RNA Me hod sec ion: RNA syn hesis Sequence- based eagen s -sR26-1 A61C This pape In i o ansc ibed RNA Me hod sec ion: RNA syn hesis Sequence- based eagen s -sR26-1 A61U This pape In i o ansc ibed RNA Me hod sec ion: RNA syn hesis Sequence- based eagen s -sR26-2 This pape In i o ansc ibed RNA Me hod sec ion: RNA syn hesis Sequence- based eagen s -sR26-2 subs a e This pape In i o ansc ibed RNA Me hod sec ion: RNA syn hesis Sequence- based eagen sR26 G aziadei e al. (2016) In i o ansc ibed RNA Sequence- based eagen ssR26 Lapinai e e al. (2013) In i o ansc ibed RNA Comme cial assay o ki TLAM-ILVp oS labelling NMR-Bio NA Con inued on nex page G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 16 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics Con inued Reagen ype (species) o esou ce Designa ion Sou ce o e e ence Iden i ie s Addi ional in o ma ion Chemical compound, d ug Iodoace oamido- PROXYL Sigma-Ald ich Ca # 253421–25 MG Chemical compound, d ug (me hyl-13C, 99%; 3,3-D2, 98%) a- ke obu y ic acid Camb idge Iso ope Labs CDLM-7318-PK Chemical compound, d ug (3-me hyl-13C, 99%; 3,4,4,4-D4, 98%) a-ke oiso ale ic acid Camb idge Iso ope Labs CDLM-7317-PK Chemical compound, d ug [3–2 H2,4–2H, 5–13C, 5’2 H3]-a-ke oiso- cap oa e Lich enecke e al. (2013) So wa e, algo i hm CNS This pape Me hod sec ion: S uc u e calcula ion and selec ion. Adap a ion o p o ocol om Lapinai e e al. (2013) So wa e, algo i hm Py hon-based SAS-PRE sco ing algo i hm This pape So wa e, algo i hm ATSAS 2.7.5 Pe oukho e al., 2012 So wa e, algo i hm Py hon-based SAS-PRE sco ing algo i hm This pape Me hod sec ion: Ensemble Sco ing P o ein exp ession, labelling and pu i ica ion L7Ae (UniP o KB accession code Q8U160), Nop5 (Q8U4M1) and a chaeal ib illa in (Q8U4M2) we e exp essed, pu i ied and econs i u ed wi h sRNAs as desc ibed p e iously (G aziadei e al., 2016). Nop5 was exp essed wi h he L113K and V223E mu a ions in o de o p e en he o ma ion o agg ega es. Deu e a ed p o eins we e exp essed in 100% D 2 O M9 minimal medium using 2 H-glyc- e ol as he sole ca bon sou ce. Deu e a ed p o eins wi h 1 H, 13 C-labelled ILV me hyl g oups we e p oduced ollowing p o ocols de eloped in he Kay labo a o y (Tuga ino and Kay, 2003). S e eo- speci ic p o-S 1 H, 13 C-labelling o aline and leucine me hyl g oups was ob ained by exp ession wi h he app op ia e me abolic p ecu so acco ding o he speci ica ions o he manu ac u e (TLAM- I d1 LV p oS ; Nm Bio). Leucine-speci ic labelling was achie ed using he p o ocol desc ibed by Lich enecke e al. (2013). All NMR samples we e assembled wi h 2 H-Nop5, and, in he case o 1 H, 13 C -ILV me hyl-labelled L7Ae, wi h bo h 2 H-Nop5 and 2 H- ib illa in. The 2 H(70%)-Nop5 sample o SANS expe imen s was ob ained by exp ession in 100% D 2 O M9 minimal medium wi h 1 H-glu- cose as he sole ca bon sou ce; deu e a ion le els o his sample we e e i ied by MALDI mass spec ome y. RNA syn hesis Guide-RNAs we e p oduced by in i o ansc ip ion om double-s anded plasmid DNA empla es using T7 RNA polyme ase p oduced in-house and NTPs (Ro h). RNAs we e pu i ied by dena u ing 12–20% polyac ylamide gel elec opho esis, and ex ac ed by elec o-elu ion. Fo 2 H-RNA samples, RNA syn hesis was pe o med using 2 H-labelled NTPs (Silan es). s -sR26: 5’-GCGAGCAAUGAUGAGUGAUGGGCGAACUGAGCUCGAAAGAGCAAUGAUGACG- GAGGUGAUCACUGAGCUCGC-3’ s -sR26-1: 5’-CGAGCAAUGAUGAGUGAUGGGCGAACUGAGC UCGAAAGAGCAAUGAUGACGGAGGGGCGAACUGAGCUGCG-3’ s -sR26-2: 5’-CGAGCAAUGAUGAGUGAUGGGCGAACUGAGCUCGAAAGAGCAAUGAUGAG UGAUGUGAUCACUGAGCUGCG-3’ sR26: 5’-GCGAGCAAUGAUGAGUGAUGGGCGAACUGAAA UAGUGAUGACGGAGGUGA UCUCUGAGCUCGC-3’ Subs a e RNAs o s -sR26 we e p oduced in-house using syn he ic DNA oligonucleo ides: Subs a e D0: 50-GCUUCGCCCAUCAC-3’ G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 17 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics Subs a e D: 50-GUAGAUCACCUCCG-3’ s -sR26-1 subs a e D: 5’-GUAUCGCCCCUCCG-3’ s -sR26-2 subs a e D: 5’-GUAGAUCACAUCAC-3’ T ans e o NMR me hyl-g oup assignmen s In he ee s a e, ib illa in me hyl esonances we e s e eospeci ically assigned by means o 3D NOE- SY– 13 C-HMQC spec a, acqui ed on ILV and ILV p oS -labelled samples, in combina ion wi h 3D TOC- SY– 13 C-HMQC spec a and by compa ison o he NOEs expec ed om he ib illa in s uc u e. The assignmen was ans e ed s epwise om he ee ib illa in o he Nop5-NTD– ib illa in complex, he Nop5– ib illa in complex and inally o he ull Box C/D complex. Fo he ILV-labelled Nop5- NTD– ib illa in complex, we also acqui ed a 3D NOESY– 13 C-HMQC spec um; o all complexes we acqui ed 13 C-HMQC spec a on ILV-labelled, ILV p oS -labelled and L-labelled samples. Fo he ILV- labelled Nop5– ib illa in complex, pai ings o HMQC peaks om he dias e eo opic me hyl-g oups o leucine and aline esidues we e e i ied wi h he assis ance o a 3D expe imen in which he 1 H and 13 C esonances o he me hyl g oups we e co ela ed wi h he 13 C esonances o he di ec ly bonded me hine ca bon (Cgand Cb o leucine and aline esidues, espec i ely), he eby allowing me hyl-pai s o be iden i ied om hei common me hine esonance. The pulse-sequence o his expe imen comp ises an ou -and-back magne iza ion- ans e -pa hway s a ing and ending on he me hyl p o ons, using COSY- ype ans e s be ween he me hyl and me hine ca bons and cons an - ime chemical-shi e olu ion pe iods o bo h indi ec 13 C dimensions. PRE measu emen s Mu an s we e gene a ed ollowing he QUIKCHANGE-XL p o ocol (Agilen Technologies) and pu i- ied in he p esence o 5 mM b-me cap oe hanol in o de o p e en disul ide bond o ma ion. Fo L7Ae, he na i e C68 was mu a ed o se ine p io o he in oduc ion o cys eine esidues a o he si es. The pu i ied p o ein was hen bu e exchanged in o 50 mM NaPi, 500 mM NaCl, pH 6.6 using a HiP ep 26/10 desal ing column (GE Heal hca e) and elu ed di ec ly in o ubes con aining a 10- old mola excess o he 3-(2-iodoace oamido)-PROXYL adical (Sigma-Ald ich) in he da k. The spin- labelling eac ion was allowed o p oceed o e nigh a oom empe a u e. Spin-labelled p o eins we e used o complex econs i u ion; he ee spin-label was emo ed du ing he gel- il a ion s ep. The inal econs i u ion s ep was ca ied ou in 100% D 2 O bu e (50 mM NaPi, 500 mM NaCl, pH 6.6), p io o concen a ion wi h a 10 kDa-cu o Amicon cen i ugal concen a o (Me ck Millipo e). All subs a e-loaded sRNPs we e ob ained by addi ion o 1.25 mola equi alen s o subs a e RNA. This a io yields ull sa u a ion o he subs a e RNA-binding si es o he guide RNA. We e i- ied his by moni o ing he appea ance o peaks indica i e o ee RNA (sha p peaks) in one-dimen- sional 1 H spec a o he sRNP upon addi ion o inc easing concen a ions o subs a e RNA. Sha p peaks began o appea a e a 1:1 mola a io o subs a e:guide RNA was eached. 13 C-HMQC spec a we e acqui ed on B uke A ance 800 and 850 MHz spec ome e s, equipped wi h TCI c yop obes, a 55˚C wi h sample concen a ions be ween 10 and 40 mM (2–8 mg/ml). Dia- magne ic spec a we e eco ded a e educ ion o he spin-label by addi ion o asco bic acid o a inal concen a ion o 5 mM. All spec a we e p ocessed using apodiza ion wi h an exponen ial unc ion in o de o p ese e Lo en zian line-shapes. Peaks we e i ed wi h he p og am FUDA (h p://www.ucl.ac.uk/hansen-lab/ uda/) assuming Lo en zian line-shapes. When necessa y, o e lapped peaks we e i ed as g oups. The i ed olumes and line-wid hs we e hen con e ed in o peak-heigh s. The heigh s in he pa a- magne ic and diamagne ic s a es we e used o calcula e he dis ance be ween he ni oxide g oup o he pa amagne ic ag and he espec i e me hyl-g oup (see below). The diamagne ic R 2 a es co esponding o he ans e se elaxa ion a es o 1 H single-quan um cohe ence (R 2diaH ) and 1 H- 13 C mul iple-quan um cohe ence (R 2diaHC ) o each indi idual peak we e quan i ied using he pulse-schemes om he Kay labo a o y (Tuga ino and Kay, 2006; Tuga ino and Kay, 2013), modi ied o emo e he as - elaxing-componen pu ging-elemen . Relaxa ion delays we e 0, 2, 3, 4, 6, 7, 10 and 16 ms o ib illa in, and 0, 2, 3, 4, 6, 7 and 10 ms o L7Ae. The peak-heigh s we e i ed o a mono-exponen ial decay unc ion o ex ac R 2diaH and R 2diaHC . G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 18 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics In o de o de i e he co ela ion- ime o he elec on-nucleus in e ac ion ec o , C , we quan i- ied pa amagne ic (I pa a : oxidized, pa amagne ic s a e o he spin-label) and diamagne ic (I dia : educed, diamagne ic s a e o he spin-label) peak-heigh s co esponding o known dis ances wi hin ib illa in in complexes econs i u ed wi h he Fib-R109C mu an . Fo L7Ae, we used known dis ances be ween he Nop5-CTD and L7Ae in complexes econs i u ed wi h he Nop5-S343C mu an . The a ios o peak-heigh s we e con e ed in o PREs (G 2 ), using Equa ion 2 and he R 2diaHC and R 2diaH a es measu ed o he espec i e peaks. Ipa a Idia ¼exp G2 HMQC ð ÞRdiaH 2RdiaHC 2 RdiaH 2þG2   RdiaHC 2þG2   (2) whe e HMQC ep esen s he magne iza ion ans e ime in he HMQC sequence (7.6 ms). As his equa ion is non-in e ible, G 2 was de i ed by plo ing he simula ed bleaching a io, I pa a /I dia , as a unc ion o G 2 o a gi en se o diamagne ic a es, wi h he expe imen al e o s on I pa a /I dia , R 2diaH and R 2diaHC used o de e mine he uppe and lowe bounds o he de i ed PRE. These PREs we e hen used as es ain s in he p o ocol de eloped in he Clo e Lab (Iwaha a e al., 2004), which op i- mizes an ensemble o mul iple spin-label con o ma ions in combina ion wi h C . Fo L7Ae, we used isoleucine esonances only. The minimiza ion was un using he ecommended ‘obsig’ se ing o he weigh ing o he di e en PREs. A e minimiza ion o 20 s uc u es, C was 51.8 ±5.7 ns o ib illa in and 50.4 ±9.4 ns o L7Ae. Fo a gi en alue o C , dis ances be ween he unpai ed elec on and he me hyl p o ons we e ex ac ed om he equa ion: ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi K G2 4 Cþ3 C 1þ!2 2 c   6 s(3) whe e K is a cons an (1:23 1023 cm 6 s 2 ) and wis he p o on La mo equency in ad/s. The e o s on he dis ances we e again es ima ed by using he e o s in C , expe imen al I pa a /I dia a ios and R 2 a es o yield uppe and lowe bounds on a calib a ion cu e. A lowe -bound o 10% was used o he e o s o I pa a /I dia , as ecommended by Ba is e & Wagne (Ba is e & Wagne , 2000). A lowe - bound o 2 A ˚was imposed o he e o s on he dis ances in o de o accoun o ag lexibili y. Finally, a minimum e o o 4 A ˚was used as lowe bound o he dis ances ex ac ed om he PRE a ios in he calcula ion o he [on,o ]-s uc u es, o accoun o he possibili y ha he me hyl g oup o only one ib illa in copy is close he pa amagne ic ag: in his case, he e ec i e dis ance o he me hyl g oup o he one ib illa in copy o he pa amagne ic ag would be smalle han he dis ance calcula ed om he sum o he wo o e lapping ib illa in peaks (one wi h PRE in ensi y- a ios < 0.8 and one wi h PRE in ensi y- a ios close o 1). In he s uc u e calcula ions (CNS), dis ances we e imposed om he ni ogen a om o he ni o- xide g oup o he pa amagne ic ag o he ca bon a oms o ib illa in me hyl g oups. Fo L7Ae, whe e s e eospeci ic assignmen o LV me hyl g oups was no a ailable, he dis ance es ain was imposed o bo h me hyl g oup ca bons wi h an ‘OR’ s a emen . Fo complexes wi h bo h ib illa in copies posi ioned away om he RNA, he same se o dis ance es ain s was imposed on each ib illa in copy; o complexes wi h one ib illa in copy close o he RNA, dis ance es ain s we e imposed wi h an ‘OR’ s a emen . Small-angle X- ay sca e ing (SAXS) Box C/D sRNPs econs i u ed in 50 mM NaPi pH 6.6, 500 mM NaCl we e eco ded a 40˚C and con- cen a ions a ying om 0.4 o 5 mg/ml, unless o he wise speci ied. In mos expe imen s a empe a- u e o 40˚C ins ead o 55˚C was used o SAXS measu emen s due o he di icul y in collec ing da a wi h high sal concen a ions a he highe empe a u e. Fo all measu emen s, 2 mM di hio h ei ol (DTT) was added o mi iga e adia ion damage. Da a collec ion was pe o med a he ESRF bioSAXS beamline BM29 wi h exposu e o 10 ames each o 1 s du a ion. The cu es we e compa ed, me ged, and he bu e con ibu ion sub ac ed by he beamline so wa e BsxCube (Pe no e al., 2013). Fo wa d sca e ing in ensi y I(0) alues we e no malized ela i e o an ideal p o ein in an ideal solu ion, and we e epo ed as 288, 194, 215 and 197 o he apo s -sR26 RNP, he subs a e D’- bound s -sR26 RNP, he subs a e D-bound s -sR26 RNP and he holo s -sR26 RNP, espec i ely, all G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 19 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics a 5 mg/ml. The R g and I(0) alues we e ex ac ed acco ding o he Guinie app oxima ion using PRI- MUS in ATSAS 2.7.5 (Kona e e al., 2003). All R g alues we e compu ed using an s.R g uppe limi o 1.3 (whe e s is he modulus o he sca e ing ec o ), as ecommended o globula pa icles. To es ima e he compa ibili y o he expe imen ally de e mined R g alues wi h he mono- o di- RNP assembly s a es, we e alua ed he heo e ical R g dis ibu ions o 5000 di-RNP models in bo h apo and holo con o ma ions om Lapinai e e al. (2013) and 500 hal -loaded mono-RNP models gene a ed in bo h [on,o ] and [o ,o ]-s a es using he o sion-angle simula ed-annealing p o ocol desc ibed below. The apo di-RNP showed a mean R g alue o 55.9 A ˚wi h a s anda d de ia ion (SD) o 2.0 A ˚; he holo di-RNP showed a mean R g o 58.1 ±3.6 A ˚; he [on,o ]-s a e o he mono-RNP showed a mean R g o 44.7 ±1.4 A ˚; and he [o ,o ]-s a e o he mono-RNP showed a mean R g o 48.5 ±1.7 A ˚(Figu e 1— igu e supplemen 3). Small-angle neu on sca e ing (SANS) 2 H-L7Ae, 2 H-Nop5, 2 H- ib illa in, 2 H-RNA, 2 H- ib illa in/ 2 H-RNA and 2 H(70%)-Nop5/ 2 H-RNA samples we e measu ed in 50 mM NaPi pH 6.6, 500 mM NaCl, 42%:58% D 2 O:H 2 O solu ions, in o de o mask he con ibu ion o he 1 H-p o eins. The cu es co esponding o 2 H-L7Ae, 2 H-Nop5, 2 H-RNA and 2 H(70%)-Nop5/ 2 H-RNA we e acqui ed a D22 a he Ins i u e Laue Lange in (ILL, G enoble, F ance), wi h a neu on wa eleng h o 6 A ˚. The 2 H- ib illa in and 2 H- ib illa in/ 2 H-RNA cu es we e acqui ed a KWS-1 a JCNS (Munich, Ge many) (Feok ys o e al., 2015) wi h a neu on wa eleng h o 5 A ˚. Bo h ins umen s we e con igu ed wi h sample-de ec o dis ances o 4 m and collima ion leng hs o 4 m. Da a educ ion and adial in eg a ion we e done wi h s anda d p ocedu es using beamline-speci ic so wa e. Bu e sub ac ion was done in PRIMUS. Pai -wise dis ance-dis ibu ion unc ions P( ) we e calcula ed om expe imen al da a using GNOM in ATSAS 2.7.5 (S e gun, 1992). All SANS cu es we e acqui ed a 55˚C. S uc u e calcula ion and selec ion S uc u es we e calcula ed using an adap ed e sion o he p o ocol desc ibed in Lapinai e e al. (2013);Nilges, 1995 acco ding o he wo k low desc ibed in Figu e 3— igu e supplemen 1. The s a ing s -sR26 RNA s uc u es, bound o ei he subs a e D o subs a e D’, we e gene a ed in sep- a a e calcula ion uns using es ain s o impose an A- o m helical geome y on he subs a e–guide duplex, and o yield he app op ia e K- u n s uc u es. S a ing p o ein con o ma ions we e gene - a ed om he PDB en y 3nmu and assembled in o wo L7Ae–Nop5–Fib p o ome s, in which he L7Ae–Nop5-CTD and Nop5-NTD–Fib in e ac ion in e aces o 3nmu we e p ese ed, bu no he el- a i e o ien a ion o he Nop5-NTD and CTD, which we e andomised. The wo copies o he p o o- me s wi hin he sRNP we e sepa a ed and andomly o a ed wi h espec o each o he . The building-blocks L7Ae–Nop5-CTD, Nop5-NTD–Fib and he Nop5 coiled-coil domain we e kep igid h oughou he calcula ions. S uc u es we e calcula ed o bo h he subs a e D- and subs a e D’- loaded sRNPs. Fo each sRNP he p o eins and RNA we e subjec ed o wo se s o pa allel o sion- angle simula ed-annealing p ocedu es; one included a se o es ain s posi ioning one ib illa in copy on he me hyla ion si e o he subs a e–guide duplex ([on,o ]-s a e); in ano he no es ain s we e imposed be ween ib illa in and he RNA ([o ,o ]-s a e). The con o ma ional sampling was d i en by PRE-de i ed dis ance es ain s, dis ance es ain s posi ioning he wo L7Ae–Nop5-CTD modules on o he RNA K- u ns and a loose dis ance es ain be ween he cen es o mass o he wo L7Ae modules (90 ±15 A ˚), which was de i ed om he P( ) cu e o 2 H-L7Ae in 42%:58% D 2 O: H 2 O. Res ain s posi ioning he Nop5-a9’ helix be ween he wo guide egions ( om Nop5-K301 and K304 o he phospha e backbone o he nucleo ide linking he K- u n and subs a e–guide helix) we e also used. Wi h his se up, we s a ed an i e a i e p ocedu e, o gene a e wo lis s o PRE- de i ed dis ance- es ain s compa ible wi h ei he he [on,o ]- o [o ,o ]-s a e. 500 s uc u es we e calcula ed pe i e a ion. A he end o each i e a ion, es ain iola ions we e e alua ed: es ain s iola ed by mo e han 10 A ˚in ei he se o calcula ions we e classi ied, elimina ed om ha pa icu- la se , bu kep in he o he . A e 5 i e a ions, his led o wo es ain -lis s pe sRNP, co espond- ing o he [on,o ]- and [o ,o ]-s a es o he sRNP. Wi h hese ou se s o es ain s ( wo o he subs a e D-loaded and wo o he subs a e D’- loaded sRNP), ou sepa a e uns o o sion-angle simula ed-annealing calcula ions we e pe o med; we gene a ed 2500 s uc u es pe un, using he se ings desc ibed in Lapinai e e al. (2013). G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 20 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics The i ness o he expe imen al SAS and PRE da a wi h espec o he calcula ed s uc u es was assessed by calcula ion o he c 2 s a is ic (Equa ion 4) and by isual inspec ion o i s be ween back- calcula ed and expe imen al da a: 2¼1 NX N i¼1 Iexp si ð Þ  cIcalc si ð Þ ssi ð Þ  2 (4) whe e I calc ep esen s he back-calcula ed da a-poin (I pa a /I dia in esi y- a ios o SAS in ensi ies), I exp is he co esponding expe imen al alue, N is he numbe o expe imen al poin s, s ep esen s he expe imen al e o and c is he scaling ac o : c¼PN i¼1 Iexp si ð ÞIcalc si ð Þ ssi ð Þ2 h i PN i¼1 Icalc si ð Þ s2 si ð Þ   (5) The s uc u es anking in he op 2% in bo h o al ene gy and es ain ene gy we e selec ed. To u he na ow down he selec ion on he basis o he SAS da a, we e alua ed he c 2 dis ibu ion o he 2 H-Nop5, 2 H-L7Ae and 2 H-RNA SANS cu es. The SAS cu es including he con ibu ion om ib illa in we e le ou , because we expec ed he posi ion o ib illa in o be a iable when i is no in con ac wi h he RNA. SAS i ness was calcula ed wi h he p og ams CRYSOL and CRYSON, om he ATSAS sui e, e sion 2.7.5 (S e gun e al., 1998). Based on he dis ibu ion o i ness o all s uc u es in each o he uns, we se loose cu -o s, which excluded only s uc u es beyond he smoo h, linea ly inc easing po ion o he dis ibu ion cu e. Fo he subs a e D’-loaded complex, we selec ed s uc u es wi hin he op 90% anking by 2 H-RNA i ness c 2 <1.4 c 2min in he [o ,o ]- s a e, c 2 <2.2 c 2min in he [on,o ]-s a e), he op 50% by 2 H-L7Ae i ness (c 2 < 1.3 c 2min in he [o , o ]-s a e, c 2 <1.3 c 2min in he [on,o ]-s a e), and he op 80% by 2 H-Nop5 i ness (c 2 < 6.1 c 2min in he [o ,o ]-s a e, c 2 <6.8 c 2min in he [on,o ]-s a e); o he subs a e D-loaded complex, we selec ed s uc u es wi hin he op 90% anking by 2 H-RNA i ness (c 2 < 2.5 c 2min in he [o ,o ]-s a e, c 2 <3.4 c 2min in he [on,o ]-s a e), he op 80% by 2 H-L7Ae i ness (c 2 < 2.0 c 2min in he [o ,o ]- s a e, c 2 <1.8 c 2min in he [on,o ]-s a e) and he op 90% by 2 H-Nop5 i ness (c 2 < 6.5 c 2min in he [o ,o ]-s a e, c 2 <5.3 c 2min in he [on,o ]-s a e). The a e age pai -wise RMSD o he s uc u es o each ensemble, calcula ed o e he Caand P a oms o he p o ein and RNA s uc u ed domains, including he ib illa in uni s no bound o he RNA, was below 5 and 7 A ˚ o he [on,o ] and [o ,o ] con o me s, espec i ely, wi h a maximum RMSD alue o less han 10 A ˚in all cases. Among he selec ed s uc u es o each o he ou uns ([on,o ]- and [o ,o ]-s a es o bo h sub- s a e D- and subs a e D’-loaded sRNPs), he one wi h he lowes es ain - iola ion ene gy ha main ained he co ec RNA opology was chosen as he s a ing poin o e inemen in Ca esian space. The ou e inemen uns comp ised 1500 s uc u es each spanning up o 10 A ˚RMSD o Ca and P a oms ela i e o he s a ing s uc u e (numbe calcula ed o he subs a e D’-loaded [on, o ]-s a e). A he end o he e inemen , we applied s ingen selec ion c i e ia wi h espec o he SAS cu es and loose c i e ia wi h espec o he ene gy. The cu -o s o he SAS da a we e se upon isual inspec ion o he c 2 dis ibu ions o each un and cu e, whe eby we allowed mo e s uc u es o be selec ed when he c 2 dis ibu ion was la . Fo he subs a e D’-loaded sRNP he cu -o s a e as ollows: op 33% o es ain - iola ion, an de Waals and o al ene gy; op 83% o 2 H-RNA (c 2 < 1.3 c 2min o he [o ,o ]-s a e, c 2 <2.0 c 2min o he [on,o ]-s a e); op 67% o 2 H-L7Ae (c 2 < 1.8 c 2min o he [o ,o ]-s a e, c 2 <1.1 c 2min o he [on,o ]-s a e); op 33% o 2 H-Nop5 (c 2 < 2.7 c 2min o he [o ,o ]-s a e, c 2 <3.4 c 2min o he [on, o ]-s a e); op 10% o 2 H(70%)-Nop5-RNA (c 2 < 6.1 c 2min o he [o ,o ]-s a e, c 2 <3.3 c 2min o he [on,o ]-s a e). Applying hese c i e ia we selec ed 1 s uc u e o he subs a e D’-loaded [o ,o ]- s a e and 12 s uc u es o he [on,o ]-s a e. The [on,o ]-s a e s uc u es displayed an a e age RMSD o 2.4 A ˚, calcula ed on all Caand P a oms (Figu e 3) excluding he ully lexible egions, namely he ee guide egion o he RNA (nucleo ides 51–62), he loops connec ing he Nop5-NTD o he coiled-coil domain ( esidues 116–122), and he loops connec ing he coiled-coil domain o he Nop5-CTD ( esidues 249–251). Fo he subs a e D-loaded sRNP he cu -o s a e as ollows: op 33% o es ain - iola ion, an de Waals and o al ene gy; op 83% o 2 H-RNA (c 2 < 2.1 c 2min o he [o ,o ]-s a e, c 2 <2.2 c 2min G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 21 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics o he [on,o ]-s a e); 66% o 2 H-L7Ae (c 2 < 2.7 c 2min o he [o ,o ]-s a e, c 2 <1.2 c 2min o he [on, o ]-s a e); 33% o 2 H-Nop5 (c 2 < 3.0 c 2min o he [o ,o ]-s a e, c 2 <1.9 c 2min o he [on,o ]-s a e); 10% o 2 H(70%)-Nop5-RNA (c 2 < 3.2 c 2min o he [o ,o ]-s a e, c 2 <2.8 c 2min o he [on,o ]-s a e). The inal ensembles o he subs a e D-loaded [o ,o ]- and [on,o ]-s a es consis o 3 and 20 s uc- u es, espec i ely, wi h a Caand P RMSDs o 4.6 and 2.4 A ˚, espec i ely. Rep esen a i e s uc u es in he inal ensembles we e minimized in explici wa e using Ambe 14 and he co esponding Ambe 99SB o ce ield (Ho nak e al., 2006). Ensemble sco ing The PRE da a and he SAS da a indica ed he p esence o a con o ma ional equilib ium be ween he [on,o ]- and [o ,o ]-s a es, as discussed in he main ex . The 2 H- ib illa in, 2 H- ib illa in/ 2 H-RNA SANS and SAXS cu es we e he e o e i ed o a mix u e o s uc u es in he [on,o ]- and [o ,o ]- s a es. In o de o add ess he lexibili y o he Nop5-NTD– ib illa in modules no in con ac wi h he RNA, we sough o gene a e ensembles con aining di e en o ien a ions o hese modules ha would imp o e he i o he SAS cu es. This con o ma ional di e si y is in addi ion o he equilib- ium be ween he [on,o ]- and [o ,o ]-s a es, esul ing in a pool o s uc u es con aining bo h [on, o ]- and [o ,o ]- s a es and mul iple con o ma ions o Nop5-NTD– ib illa in modules in each s a e. To gene a e hese ensembles we p oceeded as ollows. S a ing om he ep esen a i e s uc u e o each ensemble o Figu e 3, co esponding o he s uc u e closes o he mean o he ensemble, we pe o med a u he simula ed-annealing s ep, whe e he loops connec ing he Nop5- NTD– ib illa in modules o he es o he Box C/D pa icle we e allowed o adop andom o ien a- ions, while he es o he pa icle was kep igid. A his s age, we gene a ed 4000 s uc u es wi h andomised Nop5-NTD– ib illa in posi ions, om which we emo ed s uc u es con aining s e ic clashes. The s uc u es also con ained all spin-labels, which we e le lexible, in o de o allow back- calcula ion o PREs (see below). In a sepa a e un comp ised o 300 s uc u es, he empla e s uc u es we e kep en i ely igid while he spin-label side-chains we e allowed o o a e in o de o gene a e di e en o ien a ions, as mul iple con o ma ions o he spin-label ha e been demons a ed o i he PRE da a mo e accu- a ely han a single con o ma ion (Iwaha a e al., 2004). Ensemble sco ing was ca ied ou o subs a e D’- and subs a e D-loaded sRNPs ia he pseudo-gene ic algo i hm shown in Figu e 3— igu e supplemen 1b. Fi s , we g ouped he s uc- u es in o ou pools, con aining 3500, 3500, 300 and 300 s uc u es: [on,o ]-s a e wi h andomised Nop5-NTD ib illa in posi ions, [o ,o ]-s a e wi h andomised Nop5-NTD ib illa in posi ions, [on, o ]-s a e wi h andomised spin-label o ien a ions and [o ,o ]-s a e wi h andomised spin-label o ien- a ions. The algo i hm gene a ed ou ‘pa en ’ ensembles, each comp ising o 2–10 con o me s an- domly chosen om he pools. These ensembles we e me ged and sub-sampled, yielding 20 ‘child en’ sub-ensembles anging om 3 o 10 con o me s in size. Each sub-sampling e en had a 30% p obabili y o duplica ing a con o me o eplacing one wi h ano he om he main pool. The p ocess o pa en selec ion, sub-sampling and sco ing was epea ed 250 imes. The heo e ical sca e ing cu e o he ensemble was compu ed as he linea combina ion o he sca e ing cu es o each indi idual con o me (scaling he popula ions o ep esen mola ac ions a he han olume ac ions, which is he s anda d ATSAS ou pu ). The c 2 alue wi h espec o he expe imen al da a was calcula ed by OLIGOMER (Kona e e al., 2003). The no maliza ion o c 2 o all sub-sampled ensembles and ac oss i e a ions was done acco ding o Equa ion 6 (Ka aca e al., 2017): 2 no m ¼2 ensemble 2 min 2 max 2 min (6) whe e c 2ensemble is he i ness o an indi idual ensemble, and c 2min and c 2max a e he espec i e mini- mum and maximum alues ac oss he i e a ions o sub-ensembles being conside ed. Fi e SAS cu es we e used o sco ing: 2 H-Nop5, 2 H-Fib, 2 H-Fib/ 2 H-RNA, 2 H(70%)-Nop5/ 2 H-RNA and SAXS. The no - malized c 2 alues o each cu e we e hen summed and eno malized in o a single alue, ob ained wi h he same Equa ion 6, which hen ep esen ed he o e all SAS- i ness. G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 22 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics The calcula ion o he heo e ical I pa a /I dia a ios om mixed [on,o ]- and [o ,o ]-s a e ensembles equi es an es ima ion o he imescale o he exchange a e k ex (k ex =k 1 +k -1 ) be ween he [on,o ]- and [o ,o ]-con o me s. This can be easily done by inspec ing he ILV-me hyl 1 H- 13 C spec a o ib il- la in: in he case o slow con o ma ional exchange, he me hyl g oups in he ib illa in copy sampling he on- and o -s a es should each yield wo sepa a e NMR peaks, while o as con o ma ional exchange hese me hyl g oups should each show only a single peak, a a posi ion co esponding o he popula ion-weigh ed a e age o he posi ions co esponding o he on- and o -s a es. To in es- iga e his, we used he spec um o he RNP assembled wi h ssR26 and loaded wi h subs a e RNA as a e e ence o he slow-exchange si ua ion: in his complex, wo o he ou ib illa in copies adop a s able on-s a e, while he o he wo a e in he o -s a e, and a subse o he ib illa in me hyl g oups show sepa a e and esol able peaks co esponding o he wo s a es. In he spec a o he hal -loaded s -sR26 RNPs we did no de ec any peak a he posi ions co esponding o RNA-bound ib illa in in he holo ssR26 RNP spec a, indica ing ha in he hal -loaded mono-RNP, ei he he k ex is as e han he di e ences in he esonance equencies o he ib illa in me hyl g oups in he on- and o -s a es (~40–100 Hz), o he popula ion o he on-s a e is oo small o be de ec ed. In he sec- ond case, one would expec no CSPs upon subs a e RNA binding, which does no co espond wi h he obse ed spec a (Figu e 2a, igh panel), Thus, we back-calcula ed he PREs o he ib illa in copy ha can be in con ac wi h he subs a e–guide duplex using < 6 >ensemble a e aged dis an- ces o e he [on,o ]- and [o ,o ]-s a es, as app op ia e o he as exchange egime. Each me hyl g oup o each ib illa in o L7Ae copy is in luenced by wo PRE ags (SL1 and SL2). The esul ing G 2 alues o he me hyl g oups o he wo copies a e gi en by: GMe hyl1 2¼GMe hyl1 2;SL1þGMe hyl1 2;SL2 GMe hyl2 2¼GMe hyl2 2;SL1þGMe hyl2 2;SL2(7) whe e Me hyl1 and Me hyl2 e e o he wo copies o L7Ae o ib illa in. Because Me hyl1 and Me hyl2 ha e almos indis inguishable chemical shi s, he esul ing I pa a /I dia a ios o Me hyl1 and Me hyl2, calcula ed om Equa ions (2) and (3), we e a e aged be o e compa ison o he expe i- men al da a. The PRE i ness was quan i ied using c 2 o all expe imen al PRE alues using Equa- ion (4). Dis ances we e compu ed om he PDB iles using he Biopy hon Bio.PDB module (Cock e al., 2009). The i ness o PRE da a was no malized using Equa ion (6) and summed wi h he SAS- i ness sco e, o yield a consensus PRE-SAS sco e o each ensemble wi hin he 20 sub-sam- pling e en s, and ac oss he 250 i e a ions. Th ee independen uns o he sco ing algo i hm we e pe o med o subs a e D’- and subs a e D-loaded sRNPs, wi h he op sco ing ensemble, judged by he consensus PRE-SAS sco e, displayed in Figu e 4. A e his selec ion, he con o ma ions o each indi idual ag we e e ined by gene a ing addi- ional 3000 con o me s pe ag and by using he same pseudo-gene ic algo i hm o selec he ensembles o ag con o ma ions ha bes i ed each indi idual PRE da ase . Du ing his e inemen s ep he posi ions o all p o eins and RNA, as well as he popula ions o ib illa in con o me s in he ensemble, we e le in a ian , in o de no o al e he i o he SAS da a. Molecula dynamics Molecula dynamics simula ions o he subs a e D’- and subs a e D-bound s uc u es ep esen ing he [on,o ]-s a es we e ca ied ou in AMBER 2018 (Case e al., 2018). The simula ions we e ca ied ou in explici TIP3P wa e using a cubic box wi h a 14 A ˚wa e laye and he 14SB pa ame e se . The sys em was subjec ed o 20,000 cycles o sol en minimiza ion wi h posi ional es ain s on he complex (NPT), ollowed by hea ing o 328 K (NVT). The comple e sys em was subjec ed o an addi- ional 20,000 cycles o ene gy minimisa ion, and hen allowed o elax, keeping es ain s on he p o- eins and hea y a oms (NPT a 328 K, 0.5 ns). Subsequen ly, he wo s uc u es we e subjec ed o a 150-ns molecula dynamics. Con ac s we e ex ac ed using CPPTRAJ (Roe and Chea ham, 2013). G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 23 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics Acknowledgemen s The au ho s hank D . A em Feok ys o (MLZ Munich) o assis ance wi h eco ding and p ocessing SANS da a a KWS-1; D . Roman Lich enecke (Uni e si y o Vienna), o kindly p o iding he leucine me hyl labelling p ecu so sodium [3- 2 H 2 ,4- 2 H, 5- 13 C, 5’- 2H3 ]-a-ke oiso-cap oa e; D . Pawel Masie- wicz (EMBL Heidelbe g) and Susanne Zu Lage (HZI B aunschweig) o RNA p oduc ion and D . Be nd Simon (EMBL Heidelbe g) o assis ance wi h s uc u e calcula ions. Addi ional in o ma ion Funding Funde G an e e ence numbe Au ho Eu opean Commission FP7 ITN p ojec RNPne (con ac numbe 289007 And ea G aziadei Deu sche Fo schungsge- meinscha CA294/3-2 Te esa Ca lomagno The unde s had no ole in s udy design, da a collec ion and in e p e a ion, o he decision o submi he wo k o publica ion. Au ho con ibu ions And ea G aziadei, Resou ces, Da a cu a ion, So wa e, Fo mal analysis, In es iga ion, Visualiza ion, Me hodology; F ank Gabel, John Ki kpa ick, Da a cu a ion, Fo mal analysis; Te esa Ca lomagno, Concep ualiza ion, Da a cu a ion, Supe ision, Funding acquisi ion Au ho ORCIDs And ea G aziadei h ps://o cid.o g/0000-0001-7709-6002 Te esa Ca lomagno h ps://o cid.o g/0000-0002-2437-2760 Decision le e and Au ho esponse Decision le e h ps://doi.o g/10.7554/eLi e.50027.sa1 Au ho esponse h ps://doi.o g/10.7554/eLi e.50027.sa2 Addi ional iles Supplemen a y iles .T anspa en epo ing o m Da a a ailabili y All da a unde lying he igu es, including ele an s uc u es, is a ailable in D yad wi h DOI h ps:// doi.o g/10.5061/d yad.q573n5 d . All SAS cu es ha e been deposi ed o he SASBDB unde acces- sion codes SASDGV2, SASDGW2, SASDGX2, SASDGY2, SASDGZ2, SASDG23, SASDG33, SASDG43, SASDG53, SASDG63, SASDG73, SASDG83, SASDG93, SASDGA3 (p ojec accession code 860). The ollowing da ase s we e gene a ed: Au ho (s) Yea Da ase i le Da ase URL Da abase and Iden i ie G aziadei A, Gabel F, Ki kpa ick J, Ca lomagno T 2020 Da a om: The guide sRNA sequence de e mines he ac i i y le el o Box C/D RNPs h ps://doi.o g/10.5061/ d yad.q573n5 d D yad Digi al Reposi o y, 10.5061/ d yad.q573n5 d G aziadei A, Gabel F, Ki kpa ick J, Ca lomagno T 2019 SAS da a om: The guide sRNA sequence de e mines he ac i i y le el o Box C/D RNPs h ps://www.sasbdb.o g/ p ojec /860/ Small Angle Sca e ing Biological Da a Bank, p ojec 860 G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 24 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics Re e ences Ai aleb M, Rashid R, Chen Q, Palme JR, Daniels CJ, Li H. 2003. S uc u e and unc ion o a chaeal box C/D sRNP co e p o eins. Na u e S uc u al & Molecula Biology 10:256–263. DOI: h ps://doi.o g/10.1038/nsb905 Appel CD, Maxwell ES. 2007. S uc u al ea u es o he guide: a ge RNA duplex equi ed o a chaeal box C/D sRNA-guided nucleo ide 2’-O-me hyla ion. RNA 13:899–911. DOI: h ps://doi.o g/10.1261/ na.517307 Ba is e JL, Wagne G. 2000. U iliza ion o Si e-Di ec ed Spin Labeling and High-Resolu ion He e onuclea Nuclea Magne ic Resonance o Global Fold De e mina ion o La ge P o eins wi h Limi ed Nuclea O e hause E ec Da a † .Biochemis y 39:5355–5365. DOI: h ps://doi.o g/10.1021/bi000060h Bleiche F, Gagnon KT, B own BA, Maxwell ES, Leschzine AE, Unge VM, Base ga SJ. 2009. A Dime ic S uc u e o A chaeal Box C/D Small Ribonucleop o eins. Science 325:1384–1387. DOI: h ps://doi.o g/10. 1126/science.1176099 Ca lomagno T. 2014. P esen and u u e o NMR o RNA–p o ein complexes: A pe spec i e o in eg a ed s uc u al biology. Jou nal o Magne ic Resonance 241:126–136. DOI: h ps://doi.o g/10.1016/j.jm .2013.10. 007 Case DA, Ben-Shalom IY, B ozell SR, Ce u i DS, Chea ham I, C uzei o TE. 2018. AMBER 2018. San F ancisco: Uni e si y o Cali o nia. Cheng J, Kellne N, Be ninghausen O, Hu E, Beckmann R. 2017. 3.2-A ˚- esolu ion s uc u e o he 90S p e ibosome be o e A1 p e- RNA clea age. Na u e S uc u al & Molecula Biology 24:954–964. DOI: h ps:// doi.o g/10.1038/nsmb.3476 Clo e GM, Iwaha a J. 2009. Theo y, P ac ice, and Applica ions o Pa amagne ic Relaxa ion Enhancemen o he Cha ac e iza ion o T ansien Low-Popula ion S a es o Biological Mac omolecules and Thei Complexes. Chemical Re iews 109:4108–4139. DOI: h ps://doi.o g/10.1021/c 900033p Cock PJA, An ao T, Chang JT, Chapman BA, Cox CJ, Dalke A, F iedbe g I, Hamel yck T, Kau F, Wilczynski B, de Hoon MJL. 2009. Biopy hon: eely a ailable Py hon ools o compu a ional molecula biology and bioin o ma ics. Bioin o ma ics 25:1422–1423. DOI: h ps://doi.o g/10.1093/bioin o ma ics/b p163 Deca u WA, Fou nie MJ. 2002. RNA modi ica ions and ibosome unc ion. T ends in Biochemical Sciences 27: 344–351. DOI: h ps://doi.o g/10.1016/S0968-0004(02)02109-6 E ales J, Ma chand V, Pan hu B, Gillo S, Belin S, Ghayad SE, Ga cia M, La o eˆ s F, Ma cel V, Baudin-Baillieu A, Be in P, Cou e´ Y, Ad ai A, Meye M, The izols G, Yusupo M, Namy O, Ohlmann T, Mo o in Y, Ca ez F, e al. 2017. E idence o RNA 20-O-me hyla ion plas ici y: Con ol o in insic ansla ional capabili ies o human ibosomes. PNAS 114:12934–12939. DOI: h ps://doi.o g/10.1073/pnas.1707674114 Falalee a M, Welden JR, Duncan MJ, S amm S. 2017. C/D-box snoRNAs o m me hyla ing and non-me hyla ing ibonucleop o ein complexes: Old dogs show new icks. BioEssays 39:1600264. DOI: h ps://doi.o g/10.1002/ bies.201600264 Feok ys o AV, F ielinghaus H, Di Z, Jaksch S, Pipich V, Appa ou M-S, Babcock E, Hanslik R, Engels R, Kemme ling G, Kleines H, Io e A, Rich e D, B u ¨ckel T. 2015. KWS-1 high- esolu ion small-angle neu on sca e ing ins umen a JCNS: cu en s a e. Jou nal o Applied C ys allog aphy 48:61–70. DOI: h ps://doi. o g/10.1107/S1600576714025977 Gagnon KT, Zhang X, Qu G, Biswas S, Su yadi J, B own BA, Maxwell ES. 2010. Signa u e amino acids enable he a chaeal L7Ae box C/D RNP co e p o ein o ecognize and bind he K-loop RNA mo i . RNA 16:79–90. DOI: h ps://doi.o g/10.1261/ na.1692310 Gong J, Li Y, Liu CJ, Xiang Y, Li C, Ye Y, Zhang Z, Hawke DH, Pa k PK, Diao L, Pu key JA, Yang L, Guo AY, Lin C, Han L. 2017. A Pan-cance analysis o he exp ession and clinical ele ance o small nucleola RNAs in human Cance . Cell Repo s 21:1968–1981. DOI: h ps://doi.o g/10.1016/j.cel ep.2017.10.070,PMID: 29141226 G aziadei A, Masiewicz P, Lapinai e A, Ca lomagno T. 2016. A chaea box C/D enzymes me hyla e wo dis inc subs a e RNA sequences wi h di e en e iciency. RNA 22:764–772. DOI: h ps://doi.o g/10.1261/ na.054320. 115 Helm M. 2006. Pos - ansc ip ional nucleo ide modi ica ion and al e na i e olding o RNA. Nucleic Acids Resea ch 34:721–733. DOI: h ps://doi.o g/10.1093/na /gkj471,PMID: 16452298 He schlag D, Ecks ein F, Cech TR. 1993. The impo ance o being ibose a he clea age si e in he Te ahymena ibozyme eac ion. Biochemis y 32:8312–8321. DOI: h ps://doi.o g/10.1021/bi00083a035,PMID: 7688573 Ho nak V, Abel R, Oku A, S ockbine B, Roi be g A, Simme ling C. 2006. Compa ison o mul iple ambe o ce ields and de elopmen o imp o ed p o ein backbone pa ame e s. P o eins: S uc u e, Func ion, and Bioin o ma ics 65:712–725. DOI: h ps://doi.o g/10.1002/p o .21123,PMID: 16981200 Iwaha a J, Schwie e s CD, Clo e GM. 2004. Ensemble app oach o NMR s uc u e e inemen agains (1)H pa amagne ic elaxa ion enhancemen da a a ising om a lexible pa amagne ic g oup a ached o a mac omolecule. Jou nal o he Ame ican Chemical Socie y 126:5879–5896. DOI: h ps://doi.o g/10.1021/ ja031580d,PMID: 15125681 Ka aca E, Rod igues J, G aziadei A, Bon in A, Ca lomagno T. 2017. M3: an in eg a i e amewo k o s uc u e de e mina ion o molecula machines. Na u e Me hods 14:897–902. DOI: h ps://doi.o g/10.1038/nme h.4392, PMID: 28805795 Kass S, Tyc K, S ei z JA, Sollne -Webb B. 1990. The U3 small nucleola ibonucleop o ein unc ions in he i s s ep o p e ibosomal RNA p ocessing. Cell 60:897–908. DOI: h ps://doi.o g/10.1016/0092-8674(90)90338-F, PMID: 2156625 G aziadei e al. eLi e 2020;9:e50027. DOI: h ps://doi.o g/10.7554/eLi e.50027 25 o 27 Resea ch a icle Biochemis y and Chemical Biology S uc u al Biology and Molecula Biophysics