Nucleolar proteins Bfr2 and Enp2 interact withDEAD-box RNA helicase Dbp4 in two differentcomplexesSahar Soltanieh Martin Lapensee and Francois Dragon

Departement des sciences biologiques and Centre de recherche BioMed Universite du Quebec a MontrealMontreal Quebec Canada

Received February 13 2013 Revised November 4 2013 Accepted November 21 2013

ABSTRACT

Different pre-ribosomal complexes are formedduring ribosome biogenesis and the compositionof these complexes is highly dynamic Dbp4 aconserved DEAD-box RNA helicase implicated inribosome biogenesis interacts with nucleolarproteins Bfr2 and Enp2 We show that like Dbp4Bfr2 and Enp2 are required for the early processingsteps leading to the production of 18S ribosomalRNA We also found that Bfr2 and Enp2 associatewith the U3 small nucleolar RNA (snoRNA) theU3-specific protein Mpp10 and various pre-18Sribosomal RNA species Thus we propose thatBfr2 Dbp4 and Enp2 are components of the smallsubunit (SSU) processome a large complex of80S Sucrose gradient sedimentation analysesindicated that Dbp4 Bfr2 and Enp2 sediment in apeak of 50S and in a peak of 80S Bfr2 Dbp4and Enp2 associate together in the 50S complexwhich does not include the U3 snoRNA howeverthey associate with U3 snoRNA in the 80S complex(SSU processome) Immunoprecipitation experi-ments revealed that U14 snoRNA associates withDbp4 in the 50S complex but not with Bfr2 orEnp2 The assembly factor Tsr1 is not part of thelsquo50Srsquo complex indicating this complex is not apre-40S ribosome A combination of experimentsleads us to propose that Bfr2 Enp2 and Dbp4 arerecruited at late steps during assembly of the SSUprocessome

INTRODUCTION

The making of eukaryotic ribosomes is an intricate processthat is highly conserved Our knowledge of ribosomebiogenesis comes mainly from studies in the budding

yeast Saccharomyces cerevisiae (1ndash4) Ribosome biogenesisinitiates within the nucleolus continues in the nucleoplasmand terminates in the cytoplasm This process involvesribosomal RNA (rRNA) transcription processing modi-fication and assembly of rRNAs with ribosomal proteinswhich leads to the synthesis of the small and large riboso-mal subunits (40S and 60S) (1256)

A key process in ribosome biogenesis is the productionof mature rRNAs the functional components of ribo-somes (7) Yeast RNA polymerase I synthesizes a longprecursor of 35S that encodes the 18S 58S and 25SrRNAs whereas the 5S rRNA is independentlytranscribed by RNA polymerase III (28) The 35S pre-rRNA is subjected to an orderly maturation process thatrequires about 200 trans-acting factors (1689)In addition tens of small nucleolar RNAs (snoRNAs)base pair transiently with pre-rRNAs and direct site-specific post-transcriptional modification of rRNAsVery few snoRNAs are required for the endonucleolyticcleavages that remove spacer sequences from pre-rRNAsIn yeast only U3 U14 and snR30 snoRNAs are essentialfor the cleavage reactions that lead to the production of18S rRNA (21011) The functionally active U3ribonucleoprotein (RNP) is a very large complex of80S called the small subunit (SSU) processome whichis formed at the 50 end of nascent pre-rRNA and can beseen under the electron microscope (1213) In yeast theSSU processome is implicated in early pre-rRNA cleav-ages at processing sites A0 A1 and A2 (1213 Figure 1)The SSU processome is an early pre-ribosomal particlethat is necessary for maturation of the 18S rRNA itcontains the U3 snoRNA and about 72 proteins includingribosome biogenesis factors and ribosomal proteins (14)These proteins assemble and interact together to formthe SSU processome A number of studies identified thepresence of sub-complexes of the SSU processome Thesesub-complexes are called UtpAtUTP UtpB UtpCMpp10 Rcl1Bms1 and U3 snoRNP (15ndash25) Howeverproteins identified from these sub-complexes account for

To whom correspondence should be addressed Email dragonfrancoisuqamca

3194ndash3206 Nucleic Acids Research 2014 Vol 42 No 5 Published online 18 December 2013doi101093nargkt1293

The Author(s) 2013 Published by Oxford University PressThis is an Open Access article distributed under the terms of the Creative Commons Attribution License (httpcreativecommonsorglicensesby30) whichpermits unrestricted reuse distribution and reproduction in any medium provided the original work is properly cited

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

43 of the proteins of the SSU processome indicatingthat many proteins of the SSU processome have not yetbeen identified as components of a sub-complex (1426)There are also studies showing that some of the sub-complexes of the SSU processome associate with therRNA precursors in a hierarchical and stepwise manner(222728)

Many of the non-ribosomal factors involved in rRNAmaturation are RNA helicases These enzymes are viewedas molecular motors that rearrange RNA structures in anenergy-dependent fashion (29ndash35) However some can re-arrange RNAndashprotein complexes and many could in factbe RNPases (3036) DEAD-box protein Dbp4 is a RNAhelicase that is phylogenetically conserved and essential foryeast viability Dbp4 was first identified as a multi-copysuppressor of lethal mutations in the Y domain of U14snoRNA (37) More recently it was shown that Dbp4 isrequired for the production of 18S rRNA and more specif-ically for the early cleavages at sites A0 A1 and A2 of thepre-rRNA (38) The C-terminal extension that flanks thecatalytic core of Dbp4 contains a predicted coiled-coilmotif which is conserved in all Dbp4 orthologs (our

unpublished observation) Because this motif is implicatedin proteinndashprotein interactions Dbp4 might function in acomplex with other protein(s) We found that Dbp4 asso-ciates with the essential nucleolar proteins Bfr2 and Enp2(39ndash41)We also show that Bfr2 and Enp2 are implicated inthe early cleavages leading to 18S rRNA production andthat Bfr2 and Enp2 associate with the U3 snoRNA and theU3-specific protein Mpp10 Sucrose gradient analyses andimmunoprecipitation assays revealed that Dbp4 Bfr2 andEnp2 associate together in complexes of 50S and 80SThese proteins do not associate with the U3 snoRNA inthe 50S peak however they interact with the U3 snoRNAin the 80S peak

MATERIALS AND METHODS

Yeast strains and media

All conditional yeast strains and strains expressing taggedproteins were derived fromYPH499 (MATa ura3-52 lys2-80 ade2-101 trp1-D63 his3-D200 and leu2-D1) (42) Wegenerated strain GALHA-BFR2 (alias YSS5) that

Figure 1 The pre-rRNA processing pathway in yeast The structure of the 35S pre-rRNA (primary transcript) is shown on top The rectangles representcellular compartments in which different steps of the processing pathway take place The pre-rRNA cleavage sites are indicated on the transcripts

Nucleic Acids Research 2014 Vol 42 No 5 3195

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

expresses 3HA-tagged Bfr2 under the control of theGAL1 promoter which was substituted for the naturalpromoter by chromosomal integration at the BFR2 locus(43) Strain YSS5 was further engineered to produce9myc-tagged Enp2 expressed from its natural promoter(44) this new strain (YSS7) is hereafter referred to as thedouble-tagged strain Strain GALENP2-myc (alias YSS9)expresses C-terminally 9myc-tagged Enp2 under thecontrol of the GAL1 promoter (43) Strain GALDBP4-HA expresses 3HA-tagged Dbp4 under the control of theGAL1 promoter (43) Strain AH109 was obtained fromClontech (MATa trp1-901 leu2-3 112 ura3-52 his3-200gal4D gal80D LYS2GAL1UASGAL1TATAHIS3GAL2UASGAL2TATAADE2 URA3 MEL1UAS

MEL1TATALacZ MEL1) The strains were grown inrich medium YPD (1 yeast extract 2 peptone 2dextrose) YPGal (1 yeast extract 2 peptone 2 gal-actose) or synthetic minimal media (067 yeast nitrogenbase) complemented with the proper dropout mix and ap-propriate carbon source

Two-hybrid analyses

The ORF encoding Dbp4 was amplified by polymerasechain reaction from genomic DNA isolated from yeaststrain YPH499 following the procedure of Asubel et al(45) Primers DBP4ndashforNco 50-CAT GCC ATG GCCAAA AAA AAT AGA TTG AAC-30 and DBP4ndashrevXma 50-CCC CCC GGG TTA ACC CTG GAT TAATTT AGC TGT C-30 were used and the DNA fragmentwas cloned between the NcoI and XmaI sites of pGBKT7(Clontech) to produce pGBK-DBP4 This plasmid wastransformed into yeast strain AH109 and used as bait ina two-hybrid screen carried out with yeast genomiclibraries (46) Plasmids pGADndashDBP4 pGADndashBFR2 andpGADndashENP2 were prepared as described earlier exceptthat primer pairs DBP4ndashforXma 50-CCC CCC GGG TATGGC CAA AAA AAA TAG ATT GAA-30 and DBP4ndashrevXho 50-CCG CTC GAG TTA ACC ATG GAT TAATTT AGC TGT C-30 BFR2ndashforXma 50-CCC CCC GGGTAT GGA AAA ATC ACT AGC GGA TCA AAT TTCC-30 and BFR2ndashrevXho 50-CGC CTC GAG TCA ACCAAA GAT TTG GAT ATC ATC GTT TTT AAC-30and ENP2ndashforXma 50-CCC CCC GGG TAT GGT TTTGAA ATC TAC TTC CGC AAA TG-30 and ENP2ndashrevXho 50-CGC CTC GAG CTA CAT ACC ACG GAACGC ATT TTT G-30 were used to amplify the ORFs ofDBP4 BFR2 and ENP2 respectively and the DNA frag-ments were individually cloned between de XmaI andXhoI sites of pGADT7 (Clontech) The integrity of two-hybrid constructs was verified by automated sequencing atthe McGill University and Genome Quebec InnovationCentreThe interaction between Dbp4 and various proteins

was assessed by the yeast two-hybrid assay in strainAH109 To this end pGBKndashDBP4 was used as bait andprey plasmids included pGADndashDBP4 pGADndashBFR2pGADndashENP2 and pGAD-NOP6 Bait and prey plasmidswere simultaneously transformed into yeast strain AH109and double transformants were selected onto SDndashTrpndashLeuagar plates (45) For each combination of bait and prey

plasmids transformants were first streaked onto a SDndashTrpndashLeu plate and after 3 days of incubation at 30C thecells were restreaked onto a SDndashTrpndashLeundashHis plate Toincrease the stringency of the two-hybrid assay 3-amino-124-triazole (3-AT) was added to SDndashTrpndashLeundashHisplates at concentrations of 2 mM or 20 mM Empty baitor prey plasmids were used as controls

Antibodies

The antibodies used in this study are as follows anti-HAmouse monoclonal antibody (mAb) (12CA5 hybridomasupernatant) anti-myc mouse mAb (9E10 hybridomasupernatant) anti-Mpp10 rabbit polyclonal (47) anti-Dbp4 rabbit polyclonal antibodies anti-Tsr1 rabbit poly-clonal (48) anti-MBP rabbit polyclonal antibodies (NEB)anti-PentamiddotHis mouse mAb (QIAGEN) and anti-GSTgoat polyclonal antibodies (GE Healthcare)

The anti-Dbp4 antibodies were raised against recombin-ant Dbp4cat which lacks most of the catalytic domainof Dbp4 to avoid cross-reaction with other DEAD-boxRNA helicases His-tagged Dbp4cat was produced inEscherichia coli BL21(DE3) pLysA from the pET23a(+)vector a kind gift of TH King and MJ Fournier(University of Massachusetts Amherst USA) this con-struct encodes a mutant derivative of Dbp4 lacking mostof the catalytic domain due to elimination of the in-frameEcoRI fragment His-tagged Dbp4cat was first isolatedon a HisTrap column using the AKTApurifier as recom-mended by the manufacturer (GE Healthcare) Duringelution fractions of 500 ml were collected and peak frac-tions were pooled recombinant Dbp4cat was furtherpurified by electrophoresis in preparative sodiumdodecyl sulfate (SDS) gels These gels were subjected toreverse staining (49) and the 46-kDa band correspondingto Dbp4cat was excised electro-eluted and concentrated(Microcon filters Millipore) The purified protein wasquantified with the Bio-Rad Protein Assay and stored at80C Immunization of rabbits was carried out in-houseat the Animal Care Facility

Immunoprecipitations

Immunoprecipitation experiments (IPs) were conductedwith whole cell extracts (WCEs) prepared from exponen-tially growing cells Cells were harvested by centrifuga-tion washed with sterile water and broken with glassbeads in TMN100 buffer (25mM Tris-HCl pH 75100mM NaCl 10mM MgCl2 and 01 NP-40) ForRNase treatment the WCEs were pre-incubated with30 mg RNase A (Sigma) for 10min at 37C and themock experiments were incubated similarly except thatno RNase A was added Thirty A600 units of cells werecollected and after preparation of cellular extract theequivalent of five A600 units were used for each IP experi-ment when IPs were done to verify association of largeRNA precursors 30 A600 units were used IPs were alsocarried out on fractions from sucrose density gradientsfractions 3 4 and 5 were pooled together and formedthe lsquo50Srsquo peak whereas pooled fractions 7 and 8 formedthe lsquo80Srsquo peak Cell lysates were incubated with protein-Aagarose beads (Roche) saturated with anti-Dbp4 anti-

3196 Nucleic Acids Research 2014 Vol 42 No 5

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

Mpp10 anti-HA or anti-myc antibodies IPs were done at4C for 1 h on a Nutator and immunoprecipitates werewashed five times with 1ml of TMN100 buffer Forprotein analyses the immunoprecipitates were eithermixed with 2 SDS loading buffer or eluted withelution buffer (25mM Tris-HCl pH 75 10mM EDTAand 05 SDS) for 10min at 65C and 2 SDS loadingbuffer was added afterwards For RNA analysis theimmunoprecipitates were eluted with the elution bufferextracted with phenolchloroform and precipitated withethanol The precipitated RNA was either resuspendedin 95 formamide or in 51 formamide and 17 for-maldehyde to analyze the U3 snoRNA or large RNAsrespectively

Western blotting

Protein samples were separated by SDS-polyacrylamidegel electrophoresis transferred onto a polyvinylidenedifluoride membrane and subjected to immunodetectionwith anti-HA (1100) anti-myc (1100) anti-Mpp10(110 000) or anti-Dbp4 (13000) anti-MBP (110 000)anti-His (11000) and anti-GST (11500) and theappropriate horse radish peroxidase-conjugated second-ary antibodies were used (GE Healthcare) Immunoblotswere revealed by chemiluminescence with theImmmobilon Western kit (Millipore)

Northern blotting

To analyze precursor and mature rRNAs total RNA wasextracted with hot acidic phenol (45) To detect the U3snoRNA in either sucrose gradient fractions or IP assaysRNA was isolated using phenolchloroform extraction asdescribed by Ausubel et al (45) Large RNAs wereseparated on 12 formaldehyde-agarose gels and smallRNAs were separated on 8 denaturing polyacrylamidegels Northern hybridization was carried out withradiolabeled oligonucleotide probes complementary tothe U3 snoRNA or to different rRNA precursors Theoligonucleotides used are as follows anti-U3 50-CCAAGT TGG ATT CAG TGG CTC-3 50-A0 50-CGCTGC TCA CCA ATG G-30 D-A2 50-GCT CTC ATGCTC TTG CC-30 A2-A3 50-TTG TTA CCT CTGGGC CC-30 anti-18S 50-CAT GGC TTA ATC TTTGAG AC-30 anti-25S 50-CTC CGC TTA TTG ATATGC-30 anti-58S 50-GCG TTG TTC ATC GAT GC-30 anti-U14 50-CGA TGG GTT CGT AAG CGT ACTCCT ACC GTG G-30 The mature 18S and 25S rRNAswere visualized by staining with GelRedTM (Biotium)Membranes were exposed to a phosphor screen andrevealed with a Molecular Imager FX (Bio-Rad)

Sucrose density gradients

WCEs were fractionated on 7ndash47 linear sucrose gradi-ents as described by Lemay et al (50) except that the lysisbuffer was TMK100 (25mM Tris-HCl pH 75 100mMKCl 10mM MgCl2 and 01 NP-40) Sixteen fractionswere collected with an ISCO density gradient fraction-ation system coupled to a UA-6 detector to producecontinuous absorbance profiles at 254 nm Eighty

microliters of each fraction was used for proteinanalyses and 200 ml used for RNA analyses

Pull-down assays

The ORFs encoding Bfr2 Dbp4 and Enp2 were clonedinto the following plasmids pMAL-c5 (NEB) pET-23a(+) (Novagen) and pGEX-4T-1 (GE Healthcare)Proteins were expressed in RosettaTM(DE3) pLysS cells(Novagen) Chloramphenicol and ampicillin were supple-mented to the LB medium Overnight cultures were grownat 37C then diluted and grown again to an A600 of 06before induction of 1mM IPTG After 2ndash4 h of inductionat 30C the cells were harvested and the pellet was resus-pended in lysis buffer (BugBuster Novagen) The MBPndashBfr2 extract was precipitated with ammonium sulfate(40) and the pellet was resuspended in TMN100 Thebinding and elution of MBP or MBPndashBfr2 fusion proteinwas carried out according to pMAL protein fusion andpurification system manual (NEB) using amylosemagnetic beads (NEB) MBPndashBfr2-coated beads wereincubated with Dbp4ndashHis or GSTndashEnp2 washed withTMN100 and eluted with maltose Pull-down experimentswere also done in the presence of yeast total RNA isolatedby the hot acidic phenol procedure (45) Eluted proteinswere analyzed by SDS-polyacrylamide gel electrophoresis(8 polyacrylamide)

RESULTS

Bfr2 interacts with Dbp4 and Enp2

The function of many RNA helicases is likely modulatedby interacting protein(s) (51) Our bioinformatics searchesrevealed that the C-terminal extension of Dbp4 harbors acoiled-coil motif that is conserved in all orthologs of Dbp4(data not shown) This suggested that Dbp4 could interactwith other protein(s) through its coiled-coil motif Toidentify potential partners of Dbp4 we carried out exten-sive two-hybrid screens with yeast genomic libraries (46)Among the two-hybrid hits that were identified (unpub-lished data) Bfr2 was a very attractive candidate becauseit is a nucleolar protein that has a role in ribosome bio-genesis (39) Database mining further suggested that Bfr2was a likely partner of Dbp4 together with Enp2 (52ndash54)All three proteins are essential for yeast growth they arephylogenetically conserved and contain at least onecoiled-coil motif (data not shown) Like Bfr2 Enp2 is anucleolar protein that has been classified as a non-SSUprocessome component (39) We carried out directedtwo-hybrid assays using full-length Dbp4 as bait andfull-length Bfr2 and Enp2 as prey We also includedDbp4 as prey because DEAD-box RNA helicases canfunction as dimers (30) and Dbp4 might do so as well(20) Controls with empty prey plasmid or empty baitplasmid (Figure 2) did not grow on selective mediaruling out a possible auto-activation of the reporter geneby the bait or preys We used the ribosome biogenesisfactor Nop6 as an additional negative prey control Wecould not use Bfr2 as bait because it is an auto-activator(our unpublished observation) Cell growth was observedwhen Dbp4 (bait) was tested together with Bfr2 Enp2 or

Nucleic Acids Research 2014 Vol 42 No 5 3197

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

Dbp4 as preys on selective medium lacking 3-AT (data notshown) However when adding 2 mM 3-AT to eliminatebackground activation of the HIS3 reporter gene or up to20 mM 3-AT to increase the stringency of the selectivemedium (55) only the Dbp4ndashBfr2 combination couldgrow (Figure 2) indicating that Dbp4 interacts morestrongly with Bfr2 than with the other proteinsSimilar experiments were carried out using Enp2 as bait

(Figure 2) Growth on selective medium was seen whenEnp2 was co-transformed with either Bfr2 or Dbp4 butonly the Enp2 and Bfr2 combination could grow in thepresence of 20mM 3-AT This result indicates that Enp2and Bfr2 strongly interact together (55) Our data are sup-ported by previous studies that identified Bfr2 and Enp2 aspotential partners of Dbp4 and Bfr2 as a potential partnerof Enp2 (56ndash58) Taken together our two-hybrid analysessuggest that the association between Dbp4 and Enp2 mightbe dependent on the presence of Bfr2

Dbp4 is associated with Bfr2 and Enp2 in vivo

To validate the two-hybrid results we verified the inter-action between Dbp4 Bfr2 and Enp2 in vivo We were notable to tag Bfr2 at its C-terminus (see also reference 39)therefore we generated a strain that expresses HA-taggedBfr2 (HA-Bfr2) under the control of the GAL1 promoterand myc-tagged Enp2 (Enp2-myc) from its natural

promoter this strain was named double-tagged strainWe carried out IPs with extracts prepared from thedouble-tagged strain grown in galactose-containingmedium (Figure 3A) IPs were done using the anti-HAmAb for Bfr2 IPs an anti-myc mAb for Enp2 IPs and

Figure 3 Analyzing interaction between Bfr2 Dbp4 and Enp2 by IPs(A) Dbp4 associates with Bfr2 and Enp2 in vivo IPs were carried outwith anti-HA anti-myc and anti-Dbp4 antibodies using extracts preparedfrom the double-tagged strain that expresses HA-tagged Bfr2 under thecontrol of the GAL1 promoter and myc-tagged Enp2 from its naturalpromoter Control IPs were done in absence of antibodies (beads aloneBA) Lane 1 is whole cell extract (T is 65 input) and lanes 2ndash5 are IPswith beads alone (BA) anti-HA mAb (Bfr2) anti-Dbp4 antibodies andanti-myc mAb (Enp2) The same blot was subjected to immunodetectionwith various antibodies recognizing proteins identified on the right(B) Dbp4 associates with Bfr2 and Enp2 in an RNA-dependentmanner Control IPs (lanes 1ndash3) were done as in Figure 3A In themock (lanes 4ndash5) the cellular extract was incubated at 37C for 10min before IP and in lanes 6 and 7 the cellular extract was treatedwith RNase A for 37C for 10 min IPs were done in absence ofantibodies (BA lane 2) or with anti-Dbp4 antibodies (lanes 3ndash7) andimmunoblotting was performed using anti-myc (Enp2) and anti-HAmAbs (Bfr2) T is 65 of input (C) Association of Bfr2 with Enp2 isnot RNA-dependent IPs were carried out as in Figure 3B except thatanti-myc mAb (Enp2) was used for IP and immunodetection was per-formed with anti-HA mAb (Bfr2) T is 65 of input (D) Bfr2 isrequired for the association of Dbp4 with Enp2 Cellular extracts wereprepared from undepleted cells (0 h lanes 1ndash2) or Bfr2-depleted cells (8h lanes 3ndash4) IPs were carried out with anti-Dbp4 antibodies and westernblotting analyses for Enp2 and Dbp4 were done with anti-myc mAb andanti-Dbp4 antibodies respectively The asterisk indicates the overexposedblot T is 65 of input

Figure 2 Directed yeast two-hybrid assays Yeast strain AH109 wastransformed with bait plasmid pGBKT7 (Vec) or its derivative pGBK-DBP4 or PGBK-ENP2 and prey plasmid pGADT7 (Vec) or its deriva-tives pGAD-DBP4 pGAD-ENP2 pGAD-BFR2 and pGAD-NOP6The bait and prey plasmids respectively carry TRP1 and LEU2 auxo-trophic markers that allow growth on medium lacking tryptophan andleucine (upper panel) Interactions between bait and prey hybridproteins activate transcription of the HIS3 reporter gene which ismonitored by growth on medium lacking histidine addition of 2 or20 mM 3-AT to this medium enhances the stringency of the HIS3reporter allowing detection of the strongest two-hybrid interactions(middle and lower panels respectively)

3198 Nucleic Acids Research 2014 Vol 42 No 5

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

rabbit polyclonal antibodies raised against Dbp4 (here-after named anti-Dbp4) Control IPs were done withuncoated agarose beads (BA) These experiments showthat Dbp4 is associated with Bfr2 and Enp2 in vivo (lane4 in Figure 3A) and Bfr2 and Enp2 also interact togetherin vivo (lanes 3 and 5 in Figure 3A) Thus IPs confirm thetwo-hybrid assay results showing a strong interactionbetween Bfr2 and Enp2

DEAD-box RNA helicases use the energy of ATP tobind and remodel RNA or RNAndashprotein complexes (34)We tested whether the association between Dbp4 and itstwo partners was dependent on the presence of RNA IPswere carried out with cellular extracts pre-treated withRNase A As shown in Figure 3B the association ofDbp4 with either Bfr2 or Enp2 was lost when usingRNase-treated extracts (compare lane 7 with lanes 3 and5) showing that their association is RNA-dependent inagreement with the recent demonstration that Dbp4 needsadditional contacts with the extension flanking the RNAduplex for optimal helicase activity (60) In contrast theinteraction between Bfr2 and Enp2 was not affectedby RNase treatment showing that their association isnot dependent on the presence of RNA (Figure 3C)

To determine if Bfr2 is required for the association ofDbp4 with Enp2 we carried out IPs with Bfr2-depletedcellular extracts The double-tagged strain was grown toexponential phase in medium containing galactose(YPGal) and then shifted to dextrose-containing medium(YPD) for 8 h We chose the 8-h time point for our experi-ments because western blot analysis showed no detectableBfr2 in the cellular extract (Figure 5D lower panel) Cellswere collected from both culture media and IPs were donewith anti-Dbp4 antibodies (Figure 3D) These experimentsshowed that the interaction between Dbp4 and Enp2 wasdecreased in Bfr2-depleted cells and this was not due toloss of Dbp4 in the immunoprecipitate (Figure 3D lowerpanel) These data corroborate our two-hybrid resultssuggesting that Bfr2 bridges Dbp4 and Enp2

Bfr2 and Enp2 are necessary for early cleavages leadingto 18S rRNA maturation

It has been shown that Dbp4 is necessary for early pre-rRNA cleavages at sites A0 A1 and A2 [(38) Figure 1]Because Bfr2 and Enp2 associate with Dbp4 we decidedto investigate their involvement in rRNA maturation

Cells were grown to exponential phase in YPGal usingthe following two strains GALHA-BFR2 expressingHA-tagged Bfr2 and GALENP2-myc encoding myc-tagged Enp2 both under the control of the GAL1promoter The cells were then shifted to YPD and har-vested at different time points after depletion total RNAwas extracted and used for northern analyses Results ofBfr2 depletion are shown in Figure 4A on depletion ofBfr2 there is a decrease in the production of the 27SA2precursor consistent with the loss of cleavage at site A2We also observed an increase in the amount of 35S and23S pre-rRNAs compared with the non-depleted sampleThe 35S and 23S pre-rRNA usually accumulate in absenceof early cleavages at sites A0ndashA2 (2) The levels of 20S pre-rRNA and the mature 18S rRNA were decreased

consistent with impaired cleavages at sites A0ndashA2 Therewere no changes observed in the abundance of the mature25S and 58S rRNA The same type of results wereobtained with Enp2-depleted cells (Figure 4B) (i) highlevels of 35S and 23S pre-rRNAs (ii) low levels of27SA2 20S pre-rRNAs and mature 18S rRNA and (iii)no change in the levels of 25S and 58S rRNAs Takentogether these results indicate that Bfr2 and Enp2 areimplicated in early processing events that lead to 18SrRNA productionPolysome profiles of Bfr2- and Enp2-depleted cells

were analyzed by sucrose density gradient sedimentationwe observed decreased amounts of 40S and 80S ribosomesand an increase of free 60S subunits (data not shown)These defects are consistent with impaired 40S subunit bio-genesis and the altered pre-rRNA processing events seen inBfr2- and Enp2-depleted cells (Figure 4A and B)

Bfr2 and Enp2 associate with the U3 snoRNA and Mpp10

We know that Dbp4 associates specifically with the U3snoRNA and the U3-specific protein Mpp10 (our unpub-lished data) so we decided to verify if Bfr2 and Enp2also associate with these SSU processome componentsIPs were carried out with Mpp10 antibodies followed bywestern analysis (Figure 5A) The results show that Mpp10associates with Bfr2 and Enp2 We also immunopre-cipitated Bfr2 and Enp2 and observed that Bfr2 associateswith Mpp10 (Figure 5B) The fact that Enp2 co-immunoprecipitates with Mpp10 but Mpp10 was notdetected in Enp2 IPs suggests that the bulk of Enp2 isnot in complex with Mpp10 or that the amount ofco-immunoprecipitated Mpp10 is below detection limitNevertheless these results show that Bfr2 and Enp2 canassociate with Mpp10 To verify the association of the U3and U14 snoRNAs with Bfr2 Dbp4 and Enp2 IPs weredone using WCEs as described in Figure 5B followed bynorthern analysis (Figure 5C) The results indicate that

Figure 4 Bfr2 and Enp2 are required for pre-rRNA processing TotalRNA was extracted from depletion strains GALHA-BFR2 (A) andGALENP2-myc (B) grown in YPGal (0h lane 1) and at differentdepletion times after the shift in YPD (lanes 2ndash4) RNAs wereanalyzed by northern hybridization using probes directed against dif-ferent rRNA precursors indicated on the right Mature 18S and 25SrRNAs were visualized by staining with GelRedTM The short and longforms of 58S rRNA were detected by northern hybridization

Nucleic Acids Research 2014 Vol 42 No 5 3199

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

Bfr2 Dbp4 and to a lesser extent Enp2 associate with theU3 snoRNA There was no association between U14snoRNA and Bfr2 Dbp4 or Enp2 (Figure 5C upperpanel) We then asked if the absence of Bfr2 affected theinteractions between U3 snoRNA and Dbp4 or Enp2Cellular extracts were prepared from the double-taggedstrain after growth in YPD to deplete Bfr2 and IPs weredone as described earlier In the absence of Bfr2 the inter-action between Enp2 and the U3 snoRNA was lostwhereas the association of Dbp4 with U3 was decreasedabout 2-fold (Figure 5C lower panel) These data indicatethat Bfr2 is necessary for the association of Enp2 withthe U3 snoRNA The absence of Bfr2 also affected theDbp4ndashU3 snoRNA interaction (but to a lesser extent)Note that the efficiency of Dbp4 and Enp2 IPs withextracts from undepleted and Bfr2-depleted cells was thesame (Figure 5D upper and middle panel)

Dbp4 Bfr2 and Enp2 associate with pre-rRNAs

Our results suggest that Bfr2 Dbp4 and Enp2 could beSSU processome components To further investigate this

possibility we tested whether these proteins associate withrRNA precursors Extracts were prepared from undepletedand Bfr2-depleted cells and we carried out IPs followed bynorthern analyses (Figure 6) The results show that in thepresence of Bfr2 the 23S pre-rRNA associates with Bfr2Dbp4 and Enp2 (lanes 3ndash5) Interestingly we observed thatBfr2 Dbp4 and Enp2 also interact with the 20S pre-rRNA(lanes 3ndash5) This result suggests that Bfr2 Dbp4 and Enp2stay associated with the pre-rRNA after its cleavage at siteA2We were also able to detect the association of Bfr2 withthe 35S and 32S pre-rRNA (lane 3) In the absence of Bfr2there was a loss of association of Enp2 with the pre-rRNAs(lane 10) In contrast Dbp4 remained associated with the23S pre-rRNA and to a lesser extent with the 35Spre-rRNA (see upper panel in Figure 6)

Depletion of Bfr2 alters the sedimentation profile ofDbp4 and Enp2

We carried out sucrose gradient sedimentation analyses todetermine the sedimentation behavior of Bfr2 Dbp4 andEnp2 The double-tagged strain was grown in YPGal and

Figure 5 Bfr2 and Enp2 associate with Mpp10 and the U3 snoRNA (A) Mpp10 associates with Bfr2 and Enp2 IPs were carried out with anti-Mpp10 antibodies and immunoblotting was done with anti-myc (Enp2) anti-Mpp10 and anti-HA (Bfr2) antibodies (B) Bfr2 interacts with Mpp10IPs were carried out with anti-HA (Bfr2) and anti-myc (Enp2) antibodies and western blotting was done with anti-Mpp10 antibodies (C) Associationof U3 snoRNA with Bfr2 Dbp4 and Enp2 in presence or absence of Bfr2 IPs were carried out with beads alone (BA) anti-HA (Bfr2) anti-Dbp4and anti-myc (Enp2) antibodies Northern analysis was done with a radiolabeled oligonucleotide complementary to the U3 and U14 snoRNAs Inthe top panel cellular extracts were prepared from undepleted cells (0 h) In the bottom panel cellular extracts were obtained form Bfr2-depletedcells (8 h) T is the input (10) S is the supernatant (10) and IP is the immunoprecipitated RNA (D) IPs of Dbp4 and Enp2 in undepleted andBfr2-depleted cells IPs were done with undepleted (0 h) and Bfr2-depleted cells (8 h) using anti-Dbp4 and anti-myc (Enp2) antibodies andimmunoblotting was done with anti-myc (Enp2) anti-Dbp4 and anti-HA (Bfr2) antibodies The asterisks indicate the overexposed blots

3200 Nucleic Acids Research 2014 Vol 42 No 5

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

then shifted to YPD and cellular extracts were preparedfor ultracentrifugation through sucrose gradients Thegradients were fractionated into 16 fractions and eachfraction was subjected to western and northern analysesAs shown in Figure 7A Dbp4 Bfr2 and Enp2 co-sedimentin a peak of about 50S in sucrose gradients Bfr2 and Enp2are also enriched in the 80S region of the gradient whichcontains very little Dbp4 The distribution of Dbp4 couldreflect the transient nature of its interactions with compo-nent(s) of the 80S complex (see further text) We alsoanalyzed the sedimentation profile of Mpp10 which wasenriched at the top of the gradient and in the 80S region ofthe gradient When cells were depleted of Bfr2 for 8 hDpb4 was distributed in a wide peak of 40ndash80Sthe fact that Dbp4 appears in complexes of various sizeson depletion of Bfr2 implies that dynamic rearrangementsof Dbp4 complexes require the presence of Bfr2Depletion of Bfr2 also changed the sedimentation profileof Enp2 which sedimented in low-density fractions sug-gesting that Bfr2 is required for association of Enp2 withcomplexes of about 50S and 80S In contrast the sedimen-tation profile of Mpp10 remained almost unchangedThese data indicate that depletion of Bfr2 alters the sedi-mentation profiles of Dbp4 and Enp2 but not that ofMpp10

We also analyzed the sedimentation pattern of theU3 and U14 snoRNAs in the presence or the absence ofBfr2 (Figure 7B) The U3 snoRNA is normally detectedin low-density fractions and in the 80S region of thegradient (top panel in Figure 7B) In the absence ofBfr2 there was no change in the overall sedimenta-tion pattern of the U3 snoRNA (bottom panel inFigure 7B) This is similar to what was observed withMpp10 in Bfr2-depleted cells (Figure 7A) Howeverthere was an important change in the distributionpattern of U14 snoRNA with Bfr2-depleted extractsU14 accumulated to a much higher extent in the 80Sregion and this was accompanied by a decrease in itsabundance in fractions 3ndash5 (Figure 7C) These resultssuggest that Bfr2 affects the release of U14 snoRNAfrom pre-rRNAs by Dbp4

Molecular interactions of Bfr2 Dbp4 and Enp2 in the 50Sand 80S complexes

We conducted a more refined analysis to investigate theassociation between Bfr2 Dbp4 and Enp2 in the 50S and80S peaks Sucrose gradient fractions were obtained fromundepleted and Bfr2-depleted cells fractions 3ndash5 (lsquo50Srsquocomplex) or 7ndash8 (lsquo80Srsquo complex) were pooled togetherand IPs were carried out on the 50S pool and the 80Spool followed by western blot analyses (Figure 8A)The intensity of the signals in Bfr2 Dbp4 and Enp2

inputs from 50S and 80S peaks in undepleted and Bfr2-depleted cells correlated with their sedimentation profilesin sucrose gradients for example on Bfr2 depletionthe amount of Enp2 was reduced in the 80S peakcompared with undepleted cells (compare lanes 2 and 4in Figure 8A)IPs with the lsquo50Srsquo and lsquo80Srsquo peak of undepleted cells

revealed that Bfr2 Dbp4 and Enp2 co-precipitated (seelanes 5 9 13 and 6 in Figure 8A) These results suggestthat Bfr2 Dbp4 and Enp2 associate together in the 50S and80S peak When Bfr2 was depleted Dbp4 could no longerassociate with Enp2 in the 50S and 80S (lanes 7and 8)We investigated the association of the U3 snoRNA with

Bfr2 Dbp4 and Enp2 in the lsquo50Srsquo and lsquo80Srsquo peaks IPs

Figure 7 Sedimentation patterns on Bfr2-depletion (A) Sedimentationprofiles of Dbp4 Bfr2 Enp2 and Mpp10 Cellular extracts wereprepared from undepleted (0h) and Bfr2-depleted cells (8h) andfractionated on 7ndash47 sucrose density gradients Fractions 1ndash16 weresubjected to western blot analysis using anti-myc (Enp2) anti-Dbp4anti-HA (Bfr2) and anti-Mpp10 antibodies The positions of 40S and60S ribosomal subunits 80S ribosome and polysomes are indicated(B) Sedimentation profile of the U3 and U14 snoRNAs Sucrosegradient fractions were prepared as in Figure 7A except that RNAswere extracted from fractions 1ndash16 and subjected to northern blotanalysis with radiolabeled oligonucleotides complementary to the U3or U14 snoRNA

Figure 6 Bfr2 Dbp4 and Enp2 associate with pre-rRNAs Cellularextracts were prepared from undepleted and Bfr2-depleted cells IPswere done without antibodies (BA) and with anti-HA (Bfr2) anti-Dbp4 and anti-myc (Enp2) antibodies Northern analysis was donewith a radiolabeled oligonucleotide to detect pre-rRNAs The asteriskindicates the overexposed blot

Nucleic Acids Research 2014 Vol 42 No 5 3201

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

were done as described in Figure 8A using undepleted cellsand the U3 snoRNA was detected by northern hybridiza-tion (Figure 8B) U3 could be detected in the 50S peak butit did not co-immunoprecipitate with Bfr2 Enp2 or Dbp4However the U3 snoRNA present in the 80S peak (SSUprocessome) did co-immunoprecipitate with Bfr2 andEnp2 and to lesser extent with Dbp4 (detectable onoverexposure see the bottom panel with the asterisk)Thus the lsquo50Srsquo complex containing Bfr2 Dbp4 andEnp2 does not include the U3 snoRNA but Bfr2 Dbp4and Enp2 associate with U3 in the SSU processomeWe also verified if Bfr2 Dbp4 and Enp2 are associated

with U14 snoRNA in the lsquo50Srsquo and lsquo80Srsquo peaks There isno association between U14 snoRNA and Bfr2 or Enp2 inthese peaks (data not shown) In contrast U14 snoRNAwas associated with Dbp4 in the lsquo50Srsquo peak of undepletedcells and in the lsquo80Srsquo peak of Bfr2-depleted cells(Figure 8C) These results correlate well with the su-crose gradient sedimentation profiles (Figure 7C) In

Bfr2-depleted cells Dbp4 and U14 snoRNA remainedassociated in the 80S peak suggesting the release of U14snoRNA from the 80S complex was impaired in theabsence of Bfr2

To determine whether the lsquo50Srsquo complex could be a pre-40S ribosome we verified if Bfr2 and Enp2 wereassociated with Tsr1 a GTPase-like protein involved inassembly of pre-40S ribosomes (6162) IPs conductedwith the 50S and 80S peaks isolated from undepletedcells revealed that Tsr1 did not co-immunoprecipitatewith Enp2 nor with Bfr2 (Figure 9) Therefore the lsquo50Srsquocomplex containing Enp2 and Bfr2 is not a pre-40Sribosome

The binding partners of Bfr2

To better define the nature of the interaction betweenBfr2 Dbp4 and Enp2 we carried out pull-down experi-ments using bacterially expressed recombinant proteinsThe results show that Bfr2 binds directly to Enp2 butnot to Dbp4 (Figure 10 left panel) Adding Enp2 to themixture did not improve Dbp4 binding to Bfr2 (data notshown) Interestingly when yeast total RNA extractedwith hot acidic phenol (and devoid of proteins) wasadded to the mixture Dbp4 could bind Bfr2 (Figure 10right panel) These results are in perfect agreement withour IP experiments showing that association of Dbp4with Bfr2 is RNA-dependent and that the interaction ofEnp2 with Bfr2 is not dependent on the presence of RNA(Figure 3)

Association of U3 snoRNA with Mpp10 in depleted cells

To test the order of recruitment of Bfr2 and Dbp4 intothe SSU processome complex we determined whether theMpp10ndashU3 snoRNA association was perturbed in theabsence of Bfr2 or Dbp4 (Figure 11) These experimentsshowed that U3 snoRNA and Mpp10 remained associatedin Bfr2- or Dbp4-depleted cells Thus our results suggestthat Bfr2 and Dbp4 are recruited into the SSU processomeafter the incorporation of the U3 snoRNP and Mpp10sub-complex

Figure 8 Association of Bfr2 Dbp4 and Enp2 in complexes of lsquo50Srsquoand lsquo80Srsquo isolated from sucrose gradients (A) Cellular extracts obtainedfrom undepleted and Bfr2-depleted cells were fractionated on sucrosegradients as in Figure 7A and two series of inputs (In) were preparedfor IPs pooled fractions 3ndash5 correspond to the lsquo50Srsquo complex andfractions 7ndash8 are the lsquo80Srsquo complex IPs were done with anti-Dbp4(lanes 5ndash8) anti-myc (lanes 9ndash12) and anti-HA antibodies (lanes 13and 14) Western blot analyses were carried out using the sameantibodies to detect the presence of Enp2 (myc) Bfr2 (HA) andDbp4 Input lanes correspond to 12 of pooled fractions (B)Gradient fractions were prepared from undepleted cells and IPs weredone as in Figure 8A except that RNAs were extracted and subjectedto northern hybridization with a radiolabeled oligonucleotide comple-mentary to the U3 snoRNA Inputs (In) correspond to 10 Theasterisk indicates the overexposed blot of Dbp4 IP (C) IPs weredone with anti-Dbp4 antibodies as in Figure 8B except that northernhybridization was carried out with a radiolabeled oligonucleotide com-plementary to the U14 snoRNA

Figure 9 Bfr2 and Enp2 do not associate with Tsr1 IPs were carriedout with anti-myc (Enp2 upper panel) and anti-HA (Bfr2 lower panel)mAbs as in Figure 8A except that western blot analyses were doneusing anti-Tsr1 polyclonal antibodies

3202 Nucleic Acids Research 2014 Vol 42 No 5

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

DISCUSSION

There are more than 200 non-ribosomal factors requiredfor processing modification and assembly reactionsduring ribosome biogenesis (16926) A large number ofthese proteins are part of the SSU processome complex(14) which is necessary for the maturation of 18S rRNA(1213) Some proteins of the SSU processome formspecific sub-complexes (15ndash25) but more than a half ofits components are not categorized into known sub-complexes (26) Moreover most of the proteinndashproteininteractions between SSU processome components havenot been identified yet (26) Thus studying the proteininteractions of the SSU processome is important torefine our understanding of the assembly architectureand activity of this complex during ribosome biogenesis(1426) Dbp4 is one of the DEAD-box RNA helicasesnecessary for the early cleavages of the pre-rRNA atsites A0ndashA2 cleavages that lead to the production of18S rRNA [(38) Figure 1] To get a better understandingof the assembly and function of the SSU processome wedecided to analyze the role of Dbp4 in molecular inter-actions leading to the production of 18S rRNA

We identified Bfr2 and Enp2 as partners of Dbp4 usingyeast two-hybrid assays (Figure 2) and we showed byimmunoprecipitation with antibodies to Dbp4 that Bfr2and Enp2 associate with Dbp4 in vivo (Figure 3) With theyeast two-hybrid system there is always a risk that thebait protein binds a secondary factor that mediates (or

bridges) the interaction with the prey protein Pull-downassays with bacterially expressed recombinant proteinsrevealed that Bfr2 binds directly to Enp2 but not toDbp4 however when adding yeast total RNA to themixture Dbp4 could bind Bfr2 (Figure 10) The RNAused in these experiments is devoid of proteins rulingout the possible involvement of a third protein mediatingthe interaction As Bfr2 does not contain an RNA-bindingmotif it is unlikely that RNA acts as a mediator of theinteraction with Dbp4 Thus the simplest explanation isthat RNA binding to Dbp4 could induce a conformationalchange that facilitates its interaction with Bfr2When IPs were done via the Brf2 or Enp2 component

the results showed that Bfr2 and Enp2 interacted witheach other but not with Dbp4 (Figure 3A) It is possiblethat the amount of co-precipitated Dbp4 in IPs for eitherBfr2 or Enp2 was under the detection limit This may alsoreflect differences in the stoichiometry or differential ac-cessibility of the tags within the complex Depletion ofBfr2 impaired the association of Dbp4 with Enp2(Figure 3D) Note that the association between Dbp4and Enp2 was not completely lost possibly becausesmall amounts of Bfr2 could still be present after 8 h ofdepletion Based on the results from two-hybrid assaysIPs and pull-down assays we propose a model for theinteraction between these three proteins Bfr2 and Enp2interact directly together in an RNA-independent manner(Figures 3 and 10) RNA binding to Dbp4 induces a con-formational change which allows interaction with Bfr2 Inthis scenario Bfr2 would acts as a bridge between Dbp4and Enp2Previous studies showed that Dbp4 is involved in the

maturation of 18S rRNA Our findings indicate that Bfr2and Enp2 are also implicated in this process (Figure 4) Infact the processing defects observed in either Bfr2- orEnp2-depleted cells are consistent with the involvementof Bfr2 and Enp2 in the early processing events atcleavage sites A0 A1 and A2 The hallmark of such pro-cessing defects is the strong accumulation of 23S pre-rRNA which was observed in Bfr2- and Enp2-depletedcells (Figure 4) Li et al (41) reported that Bfr2 and Enp2are involved in pre-rRNA processing because their deple-tion led to accumulation of the 35S pre-rRNA howeverthey did not see strong accumulation of 23S pre-rRNA ondepletion The phenotypes observed by Li et al (41) couldbe due to degradation of 23S pre-rRNA on long depletiontimes [see also (39)]Formation of the SSU processome is necessary for the

maturation of the 18S rRNA (13) The SSU processomecomplex consists of the U3 snoRNA Mpp10 (U3-specificprotein) and many other nucleolar factors (121439)Previous investigations indicated that Dbp4 associateswith U3 snoRNA and Mpp10 (unpublished data)and here we showed that Bfr2 and Enp2 also associatewith U3 and Mpp10 (Figure 5) We were able toco-immunoprecipitate Mpp10 with Bfr2 but not withEnp2 (Figure 5B) (although the amount of Mpp10co-precipitated with Enp2may be too small to be detectableby our western analyses) These analyses suggest thatDbp4 Bfr2 and Enp2 could be SSU processomecomponents

Figure 10 Pull-down assays with recombinant proteins Pull-down ex-periments were carried out using MBP (lanes 2 and 5) or MBPndashBfr2(lanes 3 and 6) bound to amylose beads After incubation and elutionthe presence of proteins Dbp4ndashHis GSTndashEnp2 and MBPndashBfr2 wasdetected by immunoblotting Experiments were done in the absenceof RNA (wo RNA left panels) or in the presence of yeast totalRNA (with RNA right panels)

Figure 11 Association of U3 snoRNA and Mpp10 is not affected bydepletion of Bfr2 or Dbp4 IPs with anti-Mpp10 antibodies were donewith extracts from undepleted cells or cells depleted of Bfr2 or Dbp4for 8 h (as in Figure 3C)

Nucleic Acids Research 2014 Vol 42 No 5 3203

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

We showed that Bfr2 Dbp4 and Enp2 associate withvarious pre-rRNAs in non-depleted cells (Figure 6)Interestingly these three proteins associate with the 20Spre-rRNA suggesting that they remain associated withthe rRNA precursor after the A2 cleavage until nuclearexport in line with the findings of Li et al (41) whoreported that Bfr2 and Enp2 were required for smallsubunit export In Bfr2-depleted cells the interaction ofEnp2 with 23S and 20S pre-rRNA was lost Thus thepresence of Bfr2 is required for the association of Enp2with these pre-RNAs In contrast Dbp4 interacts with the35S pre-rRNA and stays associated with the 23Spre-rRNA in Bfr2-depleted cells Therefore it appearsthat Bfr2 affects the molecular interactions of Dbp4 withpre-18S rRNAs during processing events leading to thematuration of 18S rRNAAnalyzing the sedimentation pattern of ribosome bio-

genesis factors by sucrose gradient sedimentation isuseful because co-sedimentation of non-ribosomal factorswith the pre-ribosomal particles may suggest physical inter-action with these particles The SSU processome has a sedi-mentation coefficient of about 80S (12) The data obtainedfrom sucrose gradient sedimentation and IPs on pooledfractions of the gradient (Figure 7 and 8) indicate thatthe 50S complex contains Bfr2 Dbp4 Enp2 and U14snoRNA (and possibly additional nucleolar factors)whereas the 80S complex (the SSU processome) containsBfr2 Dbp4 Enp2 and U3 snoRNA (Figure 8A and B)These results highlight the dynamic reorganization oflarge complexes during maturation of 18S rRNA Notethat the lsquo50Srsquo complex is not a pre-40S ribosomalparticle as there was no association of Tsr1 with Bfr2 orEnp2 (Figure 9) It has been shown that actinomycin Dtreatment induces accumulation of a 50S U3 snoRNPparticle that contains DDX10 (human homolog of Dbp4)in HeLa cells (63) Given that the lsquo50Srsquo complex describedin this study does not contain the U3 snoRNA it appearsthat the 50S complex seen in HeLa cells is not the same asthe one we characterized here

Another interesting observation was that when usingWCEs for IPs there was no association between U14snoRNA and Bfr2 Dbp4 or Enp2 (Figure 5C)However Dbp4 was associated with U14 snoRNA inthe lsquo50Srsquo peak in undepleted cells (Figure 8C) Whencells were depleted of Bfr2 U14 snoRNA was associatedwith Dbp4 in the 80S peak suggesting that Bfr2 could beimplicated in the release of U14 snoRNA from the lsquo80Srsquocomplex Discrepancies between IPs with WCEs andsucrose gradient fractions could be explained by theamount of material used or simply by the fact that frac-tions isolated from sucrose gradients are partially purifiedcomplexes which may enhance the efficiency of IPs byeliminating interfering components

Based on the following data we propose that Bfr2Dbp4 and Enp2 are SSU processome components(i) Bfr2 Dbp4 and Enp2 are nucleolar proteins (ii) theyare involved in pre-rRNA processing at cleavage sites A0A1 and A2 (iii) these proteins associate with the U3snoRNA and Mpp10 and also interact with differentpre-18S rRNA species and (iv) Bfr2 Dbp4 and Enp2co-sediment in a peak of about 80S and they associatewith U3 snoRNA in that peak

Nan1 is a component of the t-UTP sub-complex whichassembles at a very early step during SSU processomeformation (even before U3 snoRNP incorporation)Depletion of Nan1 changes the sedimentation profile ofU3 snoRNA (27) The absence of Bfr2 did not alter thesedimentation profile of the U3 snoRNA and Mpp10(Figure 7) Moreover the absence of Bfr2 or Dbp4 didnot affect the association between U3 snoRNA andMpp10 (Figure 11) These results suggest that Bfr2 andDbp4 might assemble into the SSU processome afterassembly of the U3 snoRNP and Mpp10 sub-complexesAccording to our results and the studies of (2864) wepropose a simplified model for the assembly steps of theSSU processome (Figure 12) First the UtpAt-Utp bindsto the pre-rRNA followed by two mutually independentsteps one step includes the assembly of the UtpB and theother involves the association of UtpC with the 35S during

Figure 12 Simplified model for the assembly steps of the SSU processome (i) The UtpAt-Utp sub-complex assembles on the 50ETS of the nascentpre-rRNA (ii) UtpB then associates with the pre-rRNA followed by binding of the U3 snoRNP complex (iii) (iv) UtpC interacts with the pre-rRNAindependent of UtpB and U3 snoRNP (v) Mpp10 sub-complex assembly takes place after U3 snoRNP binding (vi) The Bfr2ndashEnp2 dimer and Dbp4incorporate the SSU processome particle

3204 Nucleic Acids Research 2014 Vol 42 No 5

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

later assembly steps The U3 snoRNP base pairs with thepre-rRNA after UtpB binding which in turn allowsthe assembly of the Mpp10 sub-complex on the nascentpre-rRNA Bfr2 Enp2 and Dbp4 then incorporate intothe SSU processome particle The results from IPs andsucrose gradient sedimentation combined with the obser-vation that the absence of Bfr2 did not affect the associ-ation of Dbp4 and U3 snoRNA in the 80S peak (data notshown) suggest that Bfr2 and Enp2 are recruited togetherbut Dbp4 could be incorporated independently from Bfr2and Enp2

Our results provide new insight into the order ofassembly of three nucleolar proteins into the nascentSSU processome These additional data refine our under-standing of SSU processome structure and function

ACKNOWLEDGEMENTS

The authors thankKorka Ba for carrying the original yeasttwo-hybrid screen Sabrina Makhlouf for two-hybridanalyses Karen Wehner for the anti-HA mAb SusanBaserga for anti-Mpp10 antibodies Katrin Karbstein forthe anti-Tsr1 antibodies Tom King and Skip Fournier forthe pET23-Dbp4cat construct Vincent Lemay for helpwith artwork and Ann Beyer for comments on themanuscript

FUNDING

The National Sciences and Engineering Research Councilof Canada (NSERC) [RGPIN 249792] Funding for openaccess charge NSERC

Conflict of interest statement None declared

REFERENCES

1 HenrasAK SoudetJ GerusM LebaronSCaizergues-FerrerM MouginA and HenryY (2008) The post-transcriptional steps of eukaryotic ribosome biogenesis Cell MolLife Sci 65 2334ndash2359

2 VenemaJ and TollerveyD (1999) Ribosome synthesis inSaccharomyces cerevisiae Annu Rev Genet 33 261ndash311

3 WoolfordJL Jr (1991) The structure and biogenesis of yeastribosomes Adv Genet 29 63ndash118

4 WarnerJR (1989) Synthesis of ribosomes in Saccharomycescerevisiae Microbiol Rev 53 256ndash271

5 LafontaineDL and TollerveyD (2001) The function andsynthesis of ribosomes Nat Rev Mol Cell Biol 2 514ndash520

6 Fromont-RacineM SengerB SaveanuC and FasioloF (2003)Ribosome assembly in eukaryotes Gene 313 17ndash42

7 MoorePB and SteitzTA (2002) The involvement of RNA inribosome function Nature 418 229ndash235

8 KresslerD LinderP and de La CruzJ (1999) Proteintrans-acting factors involved in ribosome biogenesis inSaccharomyces cerevisiae Mol Cell Biol 19 7897ndash7912

9 KresslerD HurtE and BasslerJ (2010) Driving ribosomeassembly Biochim Biophys Acta 1803 673ndash683

10 MaxwellES and FournierMJ (1995) The small nucleolarRNAs Annu Rev Biochem 64 897ndash934

11 HenrasAK DezC and HenryY (2004) RNA structure andfunction in CD and HACA s(no)RNPs Curr Opin StructBiol 14 335ndash343

12 DragonF GallagherJE Compagnone-PostPA MitchellBMPorwancherKA WehnerKA WormsleyS SettlageREShabanowitzJ OsheimY et al (2002) A large nucleolar U3

ribonucleoprotein required for 18S ribosomal RNA biogenesisNature 417 967ndash970

13 OsheimYN FrenchSL KeckKM ChampionEASpasovK DragonF BasergaSJ and BeyerAL (2004) Pre-18Sribosomal RNA is structurally compacted into the SSUprocessome prior to being cleaved from nascent transcripts inSaccharomyces cerevisiae Mol Cell 16 943ndash954

14 LimYH CharetteJM and BasergaSJ (2011) Assembling aprotein-protein interaction map of the SSU processome fromexisting datasets PLoS One 6 e17701

15 ChampionEA LaneBH JackrelME ReganL andBasergaSJ (2008) A direct interaction between the Utp6 half-a-tetratricopeptide repeat domain and a specific peptide in Utp21 isessential for efficient pre-rRNA processing Mol Cell Biol 286547ndash6556

16 FreedEF and BasergaSJ (2010) The C-terminus of Utp4mutated in childhood cirrhosis is essential for ribosomebiogenesis Nucleic Acids Res 38 4798ndash4806

17 LeeSJ and BasergaSJ (1999) Imp3p and Imp4p two specificcomponents of the U3 small nucleolar ribonucleoprotein that areessential for pre-18S rRNA processing Mol Cell Biol 195441ndash5452

18 WegierskiT BillyE NasrF and FilipowiczW (2001) Bms1pa G-domain-containing protein associates with Rcl1p and isrequired for 18S rRNA biogenesis in yeast RNA 7 1254ndash1267

19 DosilM and BusteloXR (2004) Functional characterization ofPwp2 a WD family protein essential for the assembly of the 90 Spre-ribosomal particle J Biol Chem 279 37385ndash37397

20 KroganNJ PengWT CagneyG RobinsonMD HawRZhongG GuoX ZhangX CanadienV RichardsDP et al(2004) High-definition macromolecular composition of yeastRNA-processing complexes Mol Cell 13 225ndash239

21 RudraD MallickJ ZhaoY and WarnerJR (2007) Potentialinterface between ribosomal protein production and pre-rRNAprocessing Mol Cell Biol 27 4815ndash4824

22 GallagherJE DunbarDA GrannemanS MitchellBMOsheimY BeyerAL and BasergaSJ (2004) RNA polymerase Itranscription and pre-rRNA processing are linked by specific SSUprocessome components Genes Dev 18 2506ndash2517

23 GrannemanS KudlaG PetfalskiE and TollerveyD (2009)Identification of protein binding sites on U3 snoRNA andpre-rRNA by UV cross-linking and high-throughput analysis ofcDNAs Proc Natl Acad Sci USA 106 9613ndash9618

24 VenemaJ VosHR FaberAW van VenrooijWJ andRaueHA (2000) Yeast Rrp9p is an evolutionarily conserved U3snoRNP protein essential for early pre-rRNA processingcleavages and requires box C for its association RNA 61660ndash1671

25 WehnerKA GallagherJE and BasergaSJ (2002) Componentsof an interdependent unit within the SSU processome regulateand mediate its activity Mol Cell Biol 22 7258ndash7267

26 PhippsKR CharetteJM and BasergaSJ (2011) The smallsubunit processome in ribosome biogenesis-progress andprospects Wiley Interdiscip Rev RNA 2 1ndash21

27 Perez-FernandezJ RomanA De Las RivasJ BusteloXR andDosilM (2007) The 90S preribosome is a multimodular structurethat is assembled through a hierarchical mechanism Mol CellBiol 27 5414ndash5429

28 Perez-FernandezJ Martin-MarcosP and DosilM (2011)Elucidation of the assembly events required for the recruitment ofUtp20 Imp4 and Bms1 onto nascent pre-ribosomes Nucleic AcidsRes 39 8105ndash8121

29 StaleyJP and GuthrieC (1998) Mechanical devices of thespliceosome motors clocks springs and things Cell 92315ndash326

30 TannerNK and LinderP (2001) DExDH box RNA helicasesfrom generic motors to specific dissociation functions Mol Cell8 251ndash262

31 CordinO BanroquesJ TannerNK and LinderP (2006)The DEAD-box protein family of RNA helicases Gene 36717ndash37

32 BleichertF and BasergaSJ (2007) The long unwinding road ofRNA helicases Mol Cell 27 339ndash352

Nucleic Acids Research 2014 Vol 42 No 5 3205

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018

33 RajkowitschL ChenD StampflS SemradK WaldsichCMayerO JantschMF KonratR BlasiU and SchroederR(2007) RNA chaperones RNA annealers and RNA helicasesRNA Biol 4 118ndash130

34 JankowskyE (2011) RNA helicases at work binding andrearranging Trends Biochem Sci 36 19ndash29

35 JankowskyA GuentherUP and JankowskyE (2011) TheRNA helicase database Nucleic Acids Res 39 D338ndashD341

36 JankowskyE and BowersH (2006) Remodeling ofribonucleoprotein complexes with DExHD RNA helicasesNucleic Acids Res 34 4181ndash4188

37 LiangWQ ClarkJA and FournierMJ (1997) TherRNA-processing function of the yeast U14 small nucleolarRNA can be rescued by a conserved RNA helicase-like proteinMol Cell Biol 17 4124ndash4132

38 KosM and TollerveyD (2005) The putative RNA helicase Dbp4pis required for release of the U14 snoRNA from preribosomes inSaccharomyces cerevisiae Mol Cell 20 53ndash64

39 BernsteinKA GallagherJE MitchellBM GrannemanS andBasergaSJ (2004) The small-subunit processome is a ribosomeassembly intermediate Eukaryot Cell 3 1619ndash1626

40 ChabaneS and KepesF (1998) Expression of the yeast BFR2gene is regulated at the transcriptional level and throughdegradation of its product Mol Gen Genet 258 215ndash221

41 LiZ LeeI MoradiE HungNJ JohnsonAW andMarcotteEM (2009) Rational extension of the ribosomebiogenesis pathway using network-guided genetics PLoS Biol 7e1000213

42 SikorskiRS and HieterP (1989) A system of shuttle vectorsand yeast host strains designed for efficient manipulation ofDNA in Saccharomyces cerevisiae Genetics 122 19ndash27

43 LongtineMS McKenzieA III DemariniDJ ShahNGWachA BrachatA PhilippsenP and PringleJR (1998)Additional modules for versatile and economical PCR-based genedeletion and modification in Saccharomyces cerevisiae Yeast 14953ndash961

44 KnopM SiegersK PereiraG ZachariaeW WinsorBNasmythK and SchiebelE (1999) Epitope tagging of yeast genesusing a PCR-based strategy more tags and improved practicalroutines Yeast 15 963ndash972

45 AusubelBR KingstonRE MooreDD SeidmanJGSmithJA and StruhlK (eds) (1999) Short Protocols InmolecularBiology John Wiley amp Sons Inc New York

46 JamesP HalladayJ and CraigEA (1996) Genomic librariesand a host strain designed for highly efficient two-hybrid selectionin yeast Genetics 144 1425ndash1436

47 DunbarDA WormsleyS AgentisTM and BasergaSJ (1997)Mpp10p a U3 small nucleolar ribonucleoprotein componentrequired for pre-18S rRNA processing in yeast Mol Cell Biol17 5803ndash5812

48 StrunkBS LoucksCR SuM VashisthH ChengSSchillingJ BrooksCL III KarbsteinK and SkiniotisG(2011) Ribosome assembly factors prevent premature translationinitiation by 40S assembly intermediates Science 3331449ndash1453

49 OrtizML CaleroM Fernandez PatronC PatronCFCastellanosL and MendezE (1992) Imidazole-SDS-Zn reverse

staining of proteins in gels containing or not SDS andmicrosequence of individual unmodified electroblotted proteinsFEBS Lett 296 300ndash304

50 LemayV HossainA OsheimYN BeyerAL and DragonF(2011) Identification of novel proteins associated withyeast snR30 small nucleolar RNA Nucleic Acids Res 399659ndash9670

51 SilvermanE Edwalds-GilbertG and LinRJ (2003) DExDH-box proteins and their partners helping RNA helicases unwindGene 312 1ndash16

52 RiffleM MalmstromL and DavisTN (2005) The yeastresource center public data repository Nucleic Acids Res 33D378ndashD382

53 CollinsSR KemmerenP ZhaoXC GreenblattJFSpencerF HolstegeFC WeissmanJS and KroganNJ (2007)Toward a comprehensive atlas of the physical interactome ofSaccharomyces cerevisiae Mol Cell Proteom 6 439ndash450

54 NashR WengS HitzB BalakrishnanR ChristieKRCostanzoMC DwightSS EngelSR FiskDGHirschmanJE et al (2007) Expanded protein information atSGD new pages and proteome browser Nucleic Acids Res 35D468ndashD471

55 TobyGG and GolemisEA (2001) Using the yeast interactiontrap and other two-hybrid-based approaches to study protein-protein interactions Methods 24 201ndash217

56 GavinAC BoscheM KrauseR GrandiP MarziochMBauerA SchultzJ RickJM MichonAM CruciatCM et al(2002) Functional organization of the yeast proteome bysystematic analysis of protein complexes Nature 415 141ndash147

57 KroganNJ CagneyG YuH ZhongG GuoXIgnatchenkoA LiJ PuS DattaN TikuisisAP et al (2006)Global landscape of protein complexes in the yeastSaccharomyces cerevisiae Nature 440 637ndash643

58 HazbunTR MalmstromL AndersonS GraczykBJ FoxBRiffleM SundinBA ArandaJD McDonaldWH ChiuCHet al (2003) Assigning function to yeast proteins by integrationof technologies Mol Cell 12 1353ndash1365

59 GarciaI and UhlenbeckOC (2008) Differential RNA-dependentATPase activities of four rRNA processing yeast DEAD-boxproteins Biochemistry 47 12562ndash12573

60 GarciaI AlbringMJ and UhlenbeckOC (2012) Duplexdestabilization by four ribosomal DEAD-box proteinsBiochemistry 51 10109ndash10118

61 GelperinD HortonL BeckmanJ HensoldJ andLemmonSK (2001) Bms1p a novel GTP-binding protein andthe related Tsr1p are required for distinct steps of 40S ribosomebiogenesis in yeast RNA 7 1268ndash1283

62 CampbellMG and KarbsteinK (2011) Protein-proteininteractions within late pre-40S ribosomes PLoS One 6e16194

63 TurnerAJ KnoxAA PrietoJL McStayB and WatkinsNJ(2009) A novel small-subunit processome assembly intermediatethat contains the U3 snoRNP nucleolin RRP5 and DBP4Mol Cell Biol 29 3007ndash3017

64 DutcaLM GallagherJE and BasergaSJ (2011) The initial U3snoRNApre-rRNA base pairing interaction required for pre-18SrRNA folding revealed by in vivo chemical probing Nucleic AcidsRes 39 5164ndash5180

3206 Nucleic Acids Research 2014 Vol 42 No 5

Downloaded from httpsacademicoupcomnararticle-abstract42531941056324by gueston 15 February 2018