Biochemical Characterization of the Caenorhabditiselegans FBF ⋅CPB-1 Translational Regulation ComplexIdentifies Conserved Protein Interaction Hotspots

Elena Menichelli1, Joann Wu1, Zachary T. Campbell2, Marvin Wickens2 and James R. Williamson1

1 - Department of Molecular Biology, Department of Chemistry, The Skaggs Institute for Chemical Biology, The Scripps ResearchInstitute, La Jolla, CA 92037, USA2 - Howard Hughes Medical Institute, Department of Biochemistry, University of Wisconsin-Madison, Madison, WI, 53706, USA

Correspondence to James R. Williamson: jrwill@scripps.eduhttp://dx.doi.org/10.1016/j.jmb.2012.11.012Edited by D. E. Draper

Abstract

Caenorhabditis elegans CPB-1 (cytoplasmic polyadenylation element binding protein homolog-1) and FBF(fem-3 mRNA binding factor) are evolutionary conserved regulators of mRNA translation that belong to theCPEB (cytoplasmic polyadenylation element binding) and PUF (Pumilio and FBF) protein families,respectively. In hermaphrodite worms, CPB-1 and FBF control key steps during germline development,including stem cell maintenance and sex determination. While CPB-1 and FBF are known to interact, themolecular basis and function of the CPB-1 ⋅FBF complex are not known. The surface of CPB-1 that interactswith FBF was localized using in vivo and in vitro methods to a 10-residue region at the N-terminus of theprotein and these residues are present in the FBF-binding protein GLD-3 (germline development defective-3).PUF proteins are characterized by the presence of eight α-helical repeats (PUF repeats) arranged side by sidein an elongated structure. Critical residues for CPB-1 binding are found in the extended loop that connectsPUF repeats 7 and 8. The same FBF residues also mediate binding to GLD-3, indicating a conserved bindingmode between different protein partners. CPB-1 binding was competitive with GLD-3, suggestive of mutualexclusivity in vivo. RNA binding measurements demonstrated that CPB-1 alters the affinity of FBF for specificRNA sequences, implying a functional model where the coregulatory protein CPB-1 modulates FBF targetselection.

© 2012 Elsevier Ltd. All rights reserved.

Introduction

Post-transcriptional regulation of gene expression isvital for many diverse biological processes, includingstem cell maintenance and differentiation,1,2 neuro-nal synaptic plasticity,3 and cellular senescence.4

The fate of the mRNA is often determined by thelength of the poly(A) tail, which controls its localiza-tion, stability, and translational efficiency. ManymRNAs are stored with a short poly(A) tail andtheir translation is activated by polyadenylation.5

Shortening of the poly(A) tail triggers mRNAdegradation.6 The regulation of poly(A) tail lengthis often mediated by specific proteins that bind toelements located in the 3′ untranslated region (UTR)

0022-2836/$ - see front matter © 2012 Elsevier Ltd. All rights reserve

of the mRNA and nucleate formation of largemultiprotein complexes.Caenorhabditis elegans CPB-1 (cytoplasmic poly-

adenylation element binding protein homolog-1)7

and FBF (fem-3 mRNA binding factor)7 are evolu-tionary conserved 3′ UTR regulatory proteins thatcontrol key steps in germline development. CPB-1 isrequired for spermatognonia to progress from first tosecond meiosis.7 CPB-1 belongs to the cytoplasmicpolyadenylation element binding (CPEB) family ofproteins, which is found in mammals andinvertebrates.4 CPEB proteins contain two RNArecognition motifs (RRMs) followed by a zinc fingerdomain (Fig. 1a), which are required for theinteraction with the cytoplasmic polyadenylation

d. J. Mol. Biol. (2013) 425, 725–737

726 Characterization of C. elegans FBF•CPB-1 complex

element consensus sequence in the 3′ UTR of targetmRNAs.8 In Xenopus oocytes, CPEB regulates bothmRNA activation and repression.5,9 Both processesrequire the dynamic assembly of a complex ofproteins at the cytoplasmic polyadenylation element.During oocyte maturation, cytoplasmic polyadenyla-tion involves, in addition to CPEB, the scaffoldingprotein symplekin, the poly(A) polymerase GLD-2(germline development factor-2), and the multisubu-nit cleavage and polyadenylation specificity factorCSPF.10 CPEB-mediated translational repressioninvolves the poly(A) ribonuclease PARN.11

In many species, including Xenopus and C.elegans, CPEB proteins interact with relatives ofDrosophila PUM and C. elegans FBF, termed PUF(Pumilio and FBF) proteins.9,12,13 PUF proteinsrecruit the conserved deadenylase complexCCR4–Pop2–Not,14,15 Argonaute (Ago),16 and thetranslational repressor Nanos.5,17 In C. elegans,FBF controls sex determination by interacting withGLD-3 (germline development defective-3).18

FBF belongs to the PUF family of RNA-bindingproteins found in Drosophila, C. elegans, humans,and yeast. FBF plays an important role in germlinestem cell maintenance,2 in spermatogenesis,7 andin the spermatogenesis-to-oogenesis switch.19 PUFproteins are characterized by the presence of eightadjacent repeats (PUF repeats) arranged side byside in an elongated structure.20,21 Each repeatconsists of three α-helices and recognizes a singleRNA base.20,22,23 In addition to binding RNA, thePUF repeats also mediate the interaction with proteinpartners.16,23,24 PUF proteins repress2 or activatemRNA translation1,3 depending on the mRNA targetand the repertoire of associated proteins. PUFprotein–RNA interactions have been extensivelycharacterized in recent years,20,21,23,25–29 but lessis known about mechanisms ofmolecular recognitionof protein partners. An additional open questionabout PUF-mediated regulation of gene expressionis if PUF proteins simply function as a scaffold for therecruitment of protein cofactors to the mRNA or if theinteracting partners modulate their RNA bindingactivity.24,30,31

In this study, the specificity of the interactionbetween CPB-1 and FBF was investigated, as wellas cooperative binding of the multiprotein complex toRNA. Using a combination of in vitro studies withrecombinant proteins and the yeast two-hybrid ap-proach, the FBF-binding region of CPB-1 wasnarrowed down to a short stretch of amino acids andkey residues were identified, whose mutation had adeleterious effect on the interaction. Residues in theFBF loop between repeats 7 and 8 were also found tobe important for the interaction. Residues in the CPB-1/FBF interface are also found in the GLD-3 ⋅FBFcomplex,31 indicating a conservedmode of interactionbetween different protein partners. Finally, the FBF-binding region of CPB-1 enhances binding of FBF to a

putative mRNA target of the CPB-1 ⋅FBF complexboth in vitro and in vivo, supporting a mechanisticmodel where cooperative RNA–protein complexformation can modulate FBF's target selection.

Results

Mapping the FBF-binding site within CPB-1indicates that the interaction is mediated by ashort stretch of amino acids

Yeast two-hybrid experiments previously showedthat CPB-1 binds to FBF-1 through the first 80N-terminal amino acids.7 This interaction was con-firmed in vitro with a pull-down assay with the RNA-binding domain of FBF-1 and additionally with thesame domain of the protein FBF-219 (not shown).Since these two proteins are almost identical insequence, have overlapping function in vivo,19 andyielded the same results in the binding assays withCPB-1, they will collectively be referred to as FBF.Analysis of the CPB-1(1–80) peptide by circular

dichroism (CD) showed that the FBF-binding regionof CPB-1 is largely unordered (Supplementary Fig.2a), in agreement with results from secondary-structure prediction algorithms†. Moreover, theobservation that CPB-1 was sensitive to lowconcentrations of protease further supported theabsence of a compact structure (see below).Many protein interactions are mediated by short

peptide segments that bind to a globulardomain.32,33 Such peptide motifs often lie withindisordered regions.33 The observation that CPB-1(1–80) is mostly unfolded led us to investigatewhether a shorter motif within this segment wassufficient to bind FBF. Two approaches were used inparallel to accurately define the FBF-binding region:limited proteolysis experiments with recombinantproteins and yeast two-hybrid assays with CPB-1deletion mutants. Limited proteolysis with trypsinwas performed on free CPB-1 and CPB-1 that hadbeen chemically cross-linked to FBF. Comparison ofthe digestion pattern of CPB-1 alone and CPB-1cross-linked to FBF identified regions of the proteinthat were no longer accessible to the protease uponbinding to FBF. FBF was refractory to mild proteol-ysis with trypsin, which facilitated the analysis of theCPB-1 fragments. This crude approach identifiedCPB-1 residues 26–63 as the shortest fragmentprotected from proteolysis when bound to FBF(Supplementary Fig. 1). At the same time, a complexof FBF with the longest recombinant CPB-1 con-struct that could be expressed and purified (residues1–363) was digested with trypsin. The CPB-1fragments bound to FBF were isolated and identifiedby mass spectrometry (not shown). The CPB-1fragments were then produced recombinantly as

Fig. 1. Mapping of the FBF-bind-ing site within CPB-1. (a) Domainorganization of CPB-1 and FBF.CPB-1 contains two RRM domainsand a zinc finger domain. Thepreviously identified FBF-bindingregion (amino acids 1–80) is indi-cated. FBF contains eight PUF re-peats (R1 to R8) flanked byconserved regions Csp1a andCsp2 indicated as R1′ and R8′.The bar represents 100 aminoacids. (b) In vitro binding assay ofCPB-1 deletion fragments to FBF.His-tagged CPB-1 fragments weremixed with FBF and applied onto Ni-NTA resin. The resin was washedwith buffer containing increasingconcentrations of imidazole, andthe eluate was analyzed by SDS-PAGE. Load indicates the samplesthat were used in the binding assay.FBF was eluted from the Ni-NTAresin with low imidazole concentra-tion (upper panel) and was retainedon the resin only when bound to theHis-tagged CPB-1. (c) Yeast two-hybrid analysis of the interaction ofCPB-1 fragments with FBF. N- andC-terminal CPB-1 deletion mutantsfused to the LexA DNA-bindingdomain were tested with FBF

fused to the Gal4 transcriptional activation domain. CPB-1(1–560) indicates full-length protein; vector indicates noCPB-1. The shortest CPB-1 fragment that strongly activated the lacZ reporter consisted of residues 40–70. Takentogether, the in vitro and in vivo experiments indicate that the FBF-binding region of CPB-1 is contained within residues40–60. Statistical analysis was performed using a two-tailed Student's t test; P values were computed relative to the noCPB-1 control. P values less than 0.05 were considered statistically significant and are indicated with an asterisk(*P valueb0.05). Results are representative of three independent experiments; all the P values are reported inSupplementary Table 1.

727Characterization of C. elegans FBF•CPB-1 complex

hexahistidine (His)-tagged proteins and tested forbinding to FBF with a pull-down binding assay (Fig.1b). In the pull-down assay, the His-tagged protein isimmobilized on Ni-NTA resin and eluted with buffercontaining high concentration of imidazole. Thesecond protein, the putative interacting partner, isproduced without the affinity tag and is therefore notretained by the Ni-NTA resin. When protein–proteininteraction occurs, both proteins are retained on theresin and co-elute with high imidazole concentration.With this approach, the FBF-binding region withinCPB-1 was narrowed to amino acid residues 32–60(lower panel in Fig. 1b), although CPB-1 deletionfragments that did not contain residues 32–47 didnot bind to FBF. In parallel, a number of CPB-1deletion mutants were tested for binding with theyeast two-hybrid method. The smallest CPB-1fragment that still showed high β-galactosidaseactivity consisted of residues 40–70, in goodagreement with the in vitro data (Fig. 1c). Removalof the 40–50 stretch of amino acids abolished

binding. A summary of all the constructs tested forbinding in vitro and with the yeast two-hybrid methodis shown in Supplementary Fig. 2b. Taken together,the data indicate that the minimal FBF-binding regionis contained within the 40–60 stretch of amino acids.Upon narrowing down the FBF-binding region of

CPB-1, we proceeded to investigate the bindingmode of the CPB-1/FBF system. Upon binding totheir specific target, some disordered protein seg-ments may transition to regular secondary or tertiarystructures, inducing long-range structural rearrange-ments, while others remain ordered coils in theirbound form. To gain insights into the binding modeof CPB-1, we acquired heteronuclear single quan-tum coherence [15N,1H]HSQC spectra using 15N-labeled CPB-1(32–80). The spectra were recordedin the presence and absence of the FBF ⋅RNAcomplex. In the absence of RNA, FBF tended toprecipitate out of solution under NMR conditions athigh concentration for an extended period of time atroom temperature. CPB-1 and RNA form a stable

728 Characterization of C. elegans FBF•CPB-1 complex

ternary complex on FBF as discussed below. FreeCPB-1(32–80) is flexibly disordered in solution, asindicated by the low chemical shift dispersion in the[15N,1H]HSQC spectrum shown in SupplementaryFig. 2c. The spectrum is characterized by a crowdedset of backbone amide moieties located between the7.5 and 8.5 ppm proton-frequency range, which istypically populated by residues found in flexiblydisordered polypeptide segments. In the presence ofFBF, the [15N,1H]HSQC spectrum of CPB-1 exhibitsa similar pattern with a reduced number of peaks.The conserved pattern indicates that no majorstructural rearrangement occurs for the residuesobserved in these conditions (Supplementary Fig.2c). Such residues can be confidently mapped in theprotein sequence to the regions 32 to 37 and 64 to 80using the characteristic backbone HN Gly and theside-chain H2N Gln and Asn peaks as guides. Themissing peaks in the [15N,1H]HSQC of the CPB-1 ⋅FBF complex most likely correspond to residues

Fig. 2. Sequence alignment with GLD-3 shows sequence coCPB-1. CPB-1 and GLD-3 do not co-exist on FBF. (a) CPB-1 acomplexes were covalently linked before analysis by SDS-PAGindicated on the left. The first panel from the left shows GLD-3concentration is indicated with a triangle above the gel). Thewith CPB-1 (constant amounts of protein are indicated with aCPB-1 and GLD-3 to FBF was reversed. Increasing amounincreasing amounts of CPB-1 were added to FBF pre-incubated3 on FBF. (b) Sequence alignment of CPB-1 with the FBF-residues are highlighted. The alignment was performed withconserved residue, colons indicate conservation between grconservation between groups of weakly similar properties. (mutants. Residues that most affected binding to FBF were loc

bound to FBF or proximal to the binding site, mappedto residues 38 to 63. Although the current data do notallow us to rule out a structural rearrangement in thesegment 38 to 63, they indicate that FBF does notinduce long-range conformational changes in otherregions of the CPB-1 sequence employed.

Molecular underpinnings of the CPB-1 ⋅FBFcomplex

FBF interacts with GLD-3.24,31 Previous experi-ments demonstrated that loss of the FBF loop(residues 479–485) joining PUF repeats 7 and8 disrupts binding to both GLD-3 and CPB-1,suggesting that the two proteins share a commonFBF-binding site.31,34 If CPB-1 and GLD-3 bind tothe same site on FBF, this would prevent formationof a ternary complex and binding of the two proteinsmay be competitive. In order to test this hypothesis,we bound CPB-1 and GLD-3 to FBF, varying the

nservation between the FBF-binding region of GLD-3 andnd GLD-3 do not form a ternary complex with FBF. ProteinE. The in-gel mobility of proteins alone and in complex istitration into a constant amount of FBF (increasing GLD-3second panel shows GLD-3 titration to FBF pre-incubatedrectangle). In the last two panels, the order of addition ofts of CPB-1 were added to FBF (third panel) and finallywith GLD-3. In the last experiment, CPB-1 replaced GLD-

binding region of GLD-3 (residues 864–874). ConservedClustalW;35 asterisks indicate positions that have a fullyoups of strongly similar properties, and periods indicatec) Pull-down assay with FBF and CPB-1 double-alanineated in the conserved 40–47 region.

Fig. 3. Alanine scan of the minimal FBF-binding region of CPB-1 identifies key residues. (a) Determination of CPB-1/FBF binding specificity by competition-binding assay. The relative affinity of CPB-1 alanine mutants for FBF wasmeasuredby competition-binding assay using fl-CPB-1. Fl-CPB-1 was bound to FBF in order to achieve 90% complex formation andwas competed off with increasing concentrations of competitor CPB-1. The plot of the normalized polarization as a functionof competitor CPB-1 concentration is shown for wild-type CPB-1 (wt) self-competition and representative CPB-1 alaninemutants. (b) Fluorescence polarization data were fit to a Hill equation to determine the apparent IC50. The relative IC50 ofeach mutant relative to wild-type CPB-1 is shown. Experiments were performed at least in triplicate; asterisks indicatestatistically significant P values computed versus wt CPB-1 (*P valueb0.05). (c) The same alanine mutants were alsotested for binding to FBF in the yeast two-hybrid system. (d) Additional mutations of the key CPB-1 residues were screenedin the yeast-two hybrid assay for interaction with FBF. Asterisks indicate statistically significant P values (*P valueb0.05)computed versus the vector control (no CPB-1).

729Characterization of C. elegans FBF•CPB-1 complex

order of addition. Upon complex formation, theproteins were covalently cross-linked in order toallow their identification by denaturing protein gelelectrophoresis (SDS-PAGE). The chemically cross-linked GLD-3 ⋅FBF and CPB-1 ⋅FBF complexesshowed different migration patterns on the gel (Fig.2a). GLD-3 bound to free FBF but not to FBF pre-incubated with CPB-1. When CPB-1 was added toFBF, a complex with a 1:1 stoichiometry was formed.When added to the GLD-3 ⋅FBF complex, which alsohas a 1:1 stoichiometry, CPB-1 replaced GLD-3 onFBF. No ternary complex formation was observed,independent of the order of protein addition. CPB-1and GLD-3 may therefore bind to the same site, withCPB-1 binding with a greater affinity.

The sequence alignment of the FBF-bindingregion of GLD-3 with CPB-1 indicated that themost important residues in the GLD-3 FBF-bindingregion are also present in CPB-131 and are foundwithin the newly identified FBF-binding region (Fig.2b). Mapping of the residues in CPB-1 involved inbinding to FBF was done by substituting pairs ofamino acids with alanines. The CPB-1 double-alanine mutants were produced recombinantly asHis-tagged proteins and subjected to a pull-downassay with FBF. The residues with the greatest effecton binding to FBF were located in the 40–47 region(Fig. 2c). The effect of each residue in this regionwas then quantified by systematic single-alaninescanning mutagenesis. The relative affinity of the

730 Characterization of C. elegans FBF•CPB-1 complex

CPB-1 alanine mutants for FBF was measured bycompetition titration experiments using fluorescencepolarization assay (Fig. 3a). In the competition-binding assay, FBF was bound to fluorescentlylabeled wild-type CPB-1(19–80) (fl-CPB-1) andincreasing concentrations of unlabeled competitor

Fig. 4. Identification of FBF-2 amino acids that are necessmutants were tested for binding to CPB-1 with the Ni-NTA puCPB-1 or FBF only; mutation of FBF Tyr479 to Ala completelyaffinity of fl-CPB-1 for FBF wild-type and FBF alanine mutanFluorescence polarization is plotted as a function of FBF conapparent equilibrium dissociation constant Kd (shown in the(*P valueb0.05) was determined by Student's t test versus walanine mutations in the Y479–T485 loop is shown. (c) FBF recrystal structure of FBF-2 in complex with PMERNA.29 The eighconnecting repeats 7 and 8 encompassing residues 479–485 isshown as spheres.

CPB-1 (wild type or mutant) were titrated into thecomplex. The plot of the normalized polarization as afunction of competitor CPB-1 concentration wasused to derive the half-maximal inhibitory concen-tration (IC50) (Fig. 3b). The IC50 of self-competitor(wild type) CPB-1 was 1.15±0.05 μM. The reduction

ary for the interaction with CPB-1. (a) Single-alanine FBFll-down assay. The first two panels show the controls withdisrupted the interaction with CPB-1 (fourth panel). (b) Thets was determined by fluorescence polarization assays.centration and fitted to the Hill equation to determine thetable; the Hill coefficient was set to 1). The significanceild-type FBF. A graphical summary of the effect of single-sidues that affect binding to CPB-1 are highlighted in thet PUF repeats (R1–R8) are colored blue and gray, the loopin red, and key residues for the interaction with CPB-1 are

Fig. 5. CPB-1 alters the affinity of FBF for specific RNAs. (a) Representative EMSA gel determining FBF associationwith cyb-1 RNA in the absence (left) or presence (right) of CPB-1. In-gel migration of free RNA, binary, and ternarycomplexes is indicated on the left with a black line, blue diamond, and purple square, respectively. (b) Quantification of gelshifts shown on the left. The fraction of bound RNA was plotted as a function of FBF concentration and fitted to the Hillequation to determine Kd. (c) Schematic representation of the modified yeast three-hybrid assay. (d) Cell-based assaysconfirm the cooperative effects observed for the cyb-1RNA. Four additional RNAs were examined for stimulated binding inthe presence of CPB-1. The “none” construct lacks an insert; an additional negative control, PME U9C, is included. TheRNA sequences are shown in the table as is the fold activation determined by dividing binding activity in the presence ofCPB-1 by the corresponding negative control experiment conducted in the absence of CPB-1. All of the values arenormalized to cell density. Asterisks indicate P valuesb0.05.

731Characterization of C. elegans FBF•CPB-1 complex

in competition efficiency relatively to wild-type CPB-1 ranged from ~3- to 40-fold for the most significantsingle-alanine mutants. Some double-alanine mu-tants were also tested for loss of binding and theeffects were additive (not shown). Similar effectswere observed when the alanine-scanning strategywas used in the yeast two-hybrid assay (Fig. 3c).Residues whose mutations caused the greatest lossof binding were Leu40, Lys44, Thr45, Leu47, andIle49.To test if additional amino acid substitutions at key

sites in CPB-1 compromised binding to FBF, weconducted additional mutagenesis. Random mu-tants were introduced into four key sites in CPB-1using mutagenesis primers containing three randomoligonucleotides (Fig. 3d). The resulting mutantswere transformed into yeast and their relative

interactions were quantified using a yeast-two hybridassay. The identity of each mutant was then inferredbased on the DNA composition obtained by Sangersequencing.36 The majority of mutations at positions40, 45, and 47 reduce binding to FBF. However,multiple mutations of Ile49 were permissible. A smallgain in apparent binding activity was observed for aconservative Ile-to-Leu mutation. We conclude thatthe majority of mutations at key sites in CPB-1 aredeleterious for binding of FBF.

Identification of key FBF residues for CPB-1binding

Removal of the long loop connecting PUF repeats7 and 8 of FBF disrupts binding to CPB-1.37 In orderto investigate the contribution of each residue in the

732 Characterization of C. elegans FBF•CPB-1 complex

loop to CPB-1 binding, we performed a mutationalanalysis in vitro with recombinant proteins (Fig. 4a).To ensure that the reduced affinity between CPB-1and mutant FBF was due to genuine loss of bindingcontacts at the protein–protein interface rather thanbeing caused by protein misfolding, we tested theFBF mutants for folding defects using RNA bindingas a readout. Electrophoretic mobility shift assay(EMSA) experiments showed that the mutants wereproperly folded and active in binding RNA (notshown). Pull-down assays indicated that mutation ofTyr479 to Ala completely disrupted binding to CPB-1(Fig. 4a). This qualitative study also indicated thatIle480 and Thr485 play an important role in binding(Fig. 4a). The effects of the alanine mutations werequantified with the fluorescence polarization assayby direct titration into fl-CPB-1 (Fig. 4b). Fl-CPB-1bound to FBF with a Kd=49±3 nM; residues thatmost significantly affected binding to CPB-1 whenmutated to alanines were Tyr479 (which showedN60 fold increase in Kd), Ile480, with a Kd=778±49 nM, and Thr485, with a Kd=1261±146 nM, inagreement with the pull-down assay. Importantresidues for the interaction with CPB-1 are mappedonto the three-dimensional structure of FBF-2 incomplex with the protein binding element of fem-3(UGUGUCAUU)29 in Fig. 4c (Protein Data Bankaccession code: 3K64).

CPB-1 alters the affinity of FBF for cyb-1 mRNA

EMSA experiments were performed to investigatebinding of FBF to RNA in the presence of CPB-1.The mRNA target for the CPB-1 ⋅FBF complex isunknown. However, in Xenopus, CPEB and PUFfamilies act together to regulate expression of cyclinB.9 Since cyclin B mRNA is a putative conservedtarget of FBF,38 we used cyclin B of C. elegans (cyb-1) in our in vitro studies. FBF bound to cyb-1 RNAwith a Kd=24 nM (Hill =1.5), but in the presence ofCPB-1, binding was about 4 times tighter (Kd=5.8 nM, Hill =1.0) (Fig. 5a and b). The effect of CPB-1 on FBF binding to cyb-1 RNA in vitro is modest butsuggestive that CPB-1 binding and RNA binding toFBF may be cooperative.To corroborate these results and examine the

effects of CPB-1 binding on other RNA targets, weutilized a modified yeast-three hybrid assay (Fig. 5cand d). In these experiments, binding of FBF tocandidate RNAs is measured in the presence orabsence of CPB-1 fused to an SV40 NLS. Thus, theeffects of complex formation on the affinity of FBF for agiven RNA can be determined in a cell-based assay.CPB-1 appears to stimulate binding to the NRE(Nanos response element), cyb-1, and fog-1a (femi-nization of germline) RNAs. Based on comparison ofthe various RNAs, we conclude that there is asequence dependence to the observed cooperativity.Intriguingly, expression of CPB-1 appeared to consis-

tently enhance binding to the target RNA, consistentwith previous results on synthetic targets.30,34

Discussion

CPEB and PUF proteins are key regulators ofpost-transcriptional gene expression and are con-served from mammals to invertebrates.4,39 Gener-ally, post-transcriptional regulators controltranslation of multiple mRNA targets by nucleatingthe assembly of dynamic protein complexes. In thisstudy, the specificity determinants of the interactionbetween the C. elegans member of the CPEB familyCPB-1 and PUF protein FBF were determined.CPEB proteins contain an N-terminal sequence ofno conserved structural motif and two RRMs andzinc finger motifs that mediate their binding tomRNA.40 PUF proteins bind target mRNAs as wellas protein partners through their PUF repeats. Inrecent years, a wealth of biochemical and structuralinformation on the mRNA binding specificity of PUFproteins has been published,27 but less is knownabout how they interact with their protein partners.Biochemical analysis of the CPB-1 and FBF surfaceallowed the identification of key residues for theinteraction. CPB-1 binds FBF through a short stretchof amino acids located at the N-terminus of theprotein that has no regular secondary structure freein solution and undergoes no major structuralrearrangement upon binding to FBF. In vitro analysisof recombinant proteins and yeast two-hybrid assaysshowed that CPB-1 residues important for theinteraction with FBF are found within the 10-merpeptide LSTDKTNLDI encompassing residues 40–49. Amino acids whose mutation to alanine mostlydisrupted binding to FBF were Leu40, Lys44, Thr45,Leu47, and Ile49. In particular, the Leu47-to-Alamutation had a deleterious effect on the interaction.This newly identified FBF-binding region of CPB-1

is similar in sequence to the FBF-binding region ofGLD-3.31 In GLD-3, residues important for bindingare also contained in a short stretch of amino acidsand include Lys864, Thr865, and, most importantly,Leu867, whose mutation to Ala completely disruptsbinding to FBF.31 The conserved KTXL (X=anyresidue) motif plays a key role in both the GLD-3 andCPB-1 interaction with FBF. Although a KTXLsequence is important for both proteins to bind toFBF, CPB-1 and GLD-3 have different affinities forFBF, with CPB-1 binding with a Kd in the lownanomolar range and GLD-3 showing a Kd about 20times higher.31 Binding of both proteins to FBF ismediated by the loop between PUF repeats 7 and 8,and mutation of Tyr479, Ile480, and Thr485 to Alasignificantly impacts the affinity of FBF for bothproteins,31 showing a conserved mode of binding.In vitro, CPB-1 and GLD-3 compete with one

another for binding to the same region of FBF. The in

733Characterization of C. elegans FBF•CPB-1 complex

vivo consequences of their binding to the same sitein FBF are unclear and a key area for future analysis.CPB-1 and GLD-3 might antagonize one another incontrolling FBF activity in vivo; alternatively, theycould act additively on FBF, or in different cells andtimes. CPB-1 and GLD-3 are both expressed inearly-stage spermatocytes,7,41 but their co-localiza-tion with one another, and with FBF, has not beenanalyzed. Similarly, loss of either CPB-1 or GLD-3leads to the same spermatogenesis defect, arrest asprimary spermatocytes,7,18 but further genetic anal-ysis is required to understand their regulatoryrelationships and the mechanisms leading to thatarrest. In particular, similar mutant phenotypes couldarise from triggering a common checkpoint, or fromadditional, as yet unexplored roles of the twoproteins. Use of the mutations described here asC. elegans transgenes should provide preciseprobes for the biological roles of the interactionsbetween FBF and its partners. The CPB-1/FBFinteraction is also disrupted or considerably reducedby point mutations in FBF (Tyr479 to Ala and Thr485to Ala), although these mutations will certainly affectFBF's interaction with multiple proteins31,34 andtherefore will not be the ideal candidate for in vivostudies.FBF physically associates with ~7% of C. elegans

transcriptome;38 interacts with a number of proteinsincluding CPB-1, GLD-3, NOS-317 (Nanos related),and CCF-1 (the C. elegans CAF-1);1 and forms aternary complex with EF1A and CSR-1 (a translationini t iat ion factor and Ago family member,respectively).16 The key FBF residues for the interac-tion with CPB-1 are contained within the loopconnecting PUF repeats 7 and 8. These resultsindicate that relatively small surfaces are involved inbinding, consistent with the fact that these aretransient interactions. Mutation of single residues oneither one of the protein interaction surfaces has adeleterious effect on their interaction. Protruding loopson the concave surface of PUF proteins mediate aplethora of interactions.31,34 For instance, DrosophilaBrain Tumor (Brat) protein is recruited by Pumilio torepress translation of hunchback mRNA.42 Thestructural model of the Pumilio ⋅Brat complex showedthat an extended loop in PUF repeat 8 fits in theentrance to the central channel of the β-propellerformed by Brat's NHL domain.23 Flexible loops on theouter concave surface of PUF proteins bind specif-ically to both structured proteins such as Brat anddisordered proteins such as CPB-1 or GLD-3.23

The direct, physical interaction of CPEB andPUF proteins is broadly conserved from humans toC. elegans.12,37 On many 3′ UTRs in vivo, PUF andCPEB proteins are likely to be bound to their ownelements and physically interact with one another aswell. Our data show that CPEB also can influenceFBF specificity directly, independent of binding RNA.Two key observations support this hypothesis. First,

the affinity of a PUF protein, FBF-2, for specific RNAsequences is influenced by a fragment of CPEB thatlacks RNA recognition domains and possesses nospecificity for RNA itself.37 Second, different RNAsvary in their sensitivity to CPEB's effects on PUFspecificity: for example, CPB-1 increases the affinityof FBF for the binding element in fog-1 but not gld-1RNAs. In a parallel study, we examined the globalspecificity for RNA of FBF-2 alone, CPB-1 alone, andthe protein complex.30 In those studies, the shortsegment of CPB-1 broadened the specificity of FBF-2. We suggest that CPEB facilitates binding of PUFproteins to RNAs that lack the highest-affinity PUFbinding elements. Indeed, suboptimal sites appearto be common among the targets of FBF-2 andhuman Pumilio.38,43 We propose that on somemRNAs, CPEB acts and influences PUF proteinactivity without that CPEB molecule binding to theRNA. Such mRNA targets would lack CPEs, butnonetheless respond to the protein.PUF proteins bind RNA though a conserved

mechanism. The PUF domain forms a curved,elongated structure where each of the eight PUFrepeats recognizes a single nucleotide20,21,29 (Fig.4c). Despite the conserved binding mode, the lengthof the consensus sequence is protein specific. Forinstance, human Pumilio binds eight nucleotide RNAsequences, while the target recognized by FBF is ninenucleotides in length. The flatter curvature of FBFrelative to human Pumilio requires an additionalspacer nucleotide, whose base flips out and pointsaway from the protein. TheRNAspecificity is thereforeobtained through perturbations of the curvature of thePUFdomain.44 By binding to the outer surface of FBF,opposite the concave RNA binding surface, CPB-1may alter the curvature of the protein, causing a tighterbinding to specific RNA sequences. This resultsuggests a functional model where the FBF targetselection is modulated by the coregulatory factorCPB-1, consistent with the observation that the FBF-binding region of CPB-1 is sufficient for promotingtranslational repression in the presence of FBF.34

Structural analysis of the FBF ⋅CPB-1 ⋅RNA complexis needed to understand howCPB-1modulates FBF'sRNA binding activity and is now an important goal.

Materials and Methods

Protein expression and purification

CPB-1(1–80) and CPB-1(19–80) constructs werecloned into pET-22b(+) (Novagen) between restrictionsites NdeI and XhoI in order to introduce a C-terminalhexahistidine (His) tag. Strains used for cloning andexpression of recombinant proteins were Escherichia coliDH5α (Invitrogen) and E. coli BL21-Gold(DE3) (Strata-gene), respectively. Cells were grown at 37 °C in LuriaBertani medium supplemented with 50 μg/ml ampicillin

734 Characterization of C. elegans FBF•CPB-1 complex

(LB/Amp), and protein expression was induced with 1 mMIPTG for 3 h. Cell pellets were resuspended in lysis buffer[20 mM Tris–HCl (pH 7.0), 300 mM NaCl, 20 mM imidaz-ole, and 5% glycerol] supplemented with Complete, EDTA-free Protease Inhibitor Cocktail Tablets (Roche) andsubjected to mild sonication followed by 30 min incubationat 4 °C with 0.5× FastBreak Cell Lysis Reagent (Pro-mega). The cleared lysate was loaded onto TALON MetalAffinity Resin (Clontech). The resin was washed with lysisbuffer supplemented with 1 M NaCl and the protein waseluted with lysis buffer containing 150 mM imidazole(pH 7.0). Only fractions where protein purity was N95%were pooled and concentrated in Amicon Ultra-15 Centrif-ugal Filter Units (Millipore).The CPB-1(19–80) construct used in the EMSA experi-

ments and the CPB-1(1–363) construct contained an N-terminal maltose binding protein (MBP) tag. The proteinswere cloned into a modified pMAL-c2x vector (New EnglandBiolabs) in which a tobacco etch virus (TEV) cleavage sitehad been introduced between the MBP tag and the multiplecloning site. A C-terminal His tag was introduced by PCR.Protein expression was carried out in E. coli BL21-Gold(DE3) cells as described for the previous proteins.Cell pellets were resuspended in lysis buffer [0.5 M NaCl,20 mM Tris–HCl (pH 8.5), 20 mM imidazole, and 5 mMbeta-mercaptoethanol (βME)] supplementedwithComplete,EDTA-free Protease Inhibitor Cocktail Tablets and lysed bysonication. The cleared lysate was mixed with Ni-NTASuperflow resin (Qiagen) and the protein was purified withthe batch-binding procedure. The resin was washed withlysis buffer containing1 MNaCl, equilibratedagainwith lysisbuffer, and finally eluted with 20 mM Tris–HCl (pH 8.5),300 mM imidazole, 250 mM NaCl, and 10 mM βME. Theeluate was applied to amylose resin (New England Biolabs)equilibrated with lysis buffer. The resin was washed with20 mM Tris–HCl (pH 8.5), 200 mM NaCl, and 10 mM βME,and the protein was eluted with the same buffer supple-mented with 10 mM maltose. The buffer was exchanged to20 mMTris–HCl (pH 8.5), 100 mMNaCl, and 5 mMDTT bydialysis, and the protein was stored at −80 °C.Double- and single-alanine mutations were introduced in

the CPB-1(19–80) construct by PCR mutagenesis andproteins were purified as wild-type CPB-1(19–80).CPB-1 deletion mutants shorter than the CPB-1(19–80)

construct were also expressed as N-terminal, TEV-cleavable MBP fusions and with a C-terminal His tag.The proteins were purified over Ni-NTA and amyloseresins; all the buffers were supplemented with 10%glycerol. The MBP tag was removed by overnight digestionat 4 °C with recombinant His-tagged TEV protease. TheMBP tag was separated from CPB-1 by centrifugation inAmicon Ultra-15 Centrifugal Filter Units with a molecularmass cutoff of 30 kDa (Millipore).FBF-1(164–566) and FBF-2(164–566) were cloned into

a modified pETDuet-1 expression plasmid (Novagen) thatencoded for an N-terminal, TEV-cleavable His tag. Proteinexpression was carried out in E. coli BL21(DE3) cells(Stratagene) for 12 h at 18 °C upon induction with 0.5 mMIPTG. Cell pellets were resuspended in lysis buffer [20 mMTris–HCl (pH 8.6), 200 mM NaCl, 10% glycerol, 10 mMβME, 5 mM imidazole, 0.5 M urea, and Complete, EDTA-free Protease Inhibitor Cocktail Tablets] and disrupted bysonication. The proteins were purified by Ni-NTA affinitychromatography and the His tag was removed overnight at

4 °C with recombinant His-tagged TEV protease. Asecond Ni-NTA purification step was performed to removethe released tag and the TEV protease, followed by size-exclusion chromatography (Superdex 200, GE Health-care) in storage buffer [20 mM Tris–HCl (pH 8.6), 200 mMNaCl, 5 mM DTT, and 10% glycerol]. FBF-2 constructencompassing residues 164–575 was overexpressed andpurified as described in Wu et al.31

Qualitative protein binding assay (Ni-NTA pull down)

Protein complexes were reconstituted as follows: 10 to20 μMFBFwas incubated for 3 hwith 1.5–2.0molar excessCPB-1 in reconstitution buffer [20 mM Tris–HCl (pH 7.5 orpH 8.6), 100 mMNaCl, 10 mM βME, and 5%glycerol]. Pull-down assays were performed by applying the reconstitutedprotein–protein complexes or the individual proteins ascontrols, to His SpinTrap columns (GE Healthcare) pre-equilibrated in reconstitution buffer. Columns were washedthree timeswith 0.6 ml of buffer and the proteins were elutedwith 0.4 ml of buffer containing increasing concentrations ofimidazole (up to 300 mM). The content of each elutionfractionwas visualized by SDS-PAGE.Wild-type FBF-1 andFBF-2 (residues 164–566) were used for pull-down exper-iments with the His-tagged CPB-1 deletion mutants. Wild-type FBF-2 (residues 164–575) was used for pull-downassays with the single- and double-alanine mutant CPB-1constructs. Single-alanine mutations were introduced in theFBF-2(164–575) construct. The reason for using slightlydifferent FBF constructs is that for initial experiments, wedesigned the FBF-1(164–566) and FBF-2(164–566) con-structs based on sequence alignments with other PUFproteins. Since in parallel we were also pursuing crystalli-zation attempts, when the crystal structure of FBF-2(164–575) was published,29 we switched to this slightly longerconstruct, and for consistency, we also used it for thebiochemical experiments. It is worth noticing that the lasteight amino acids (residues 568–575) are not included in thecrystal structure because of poor electron density29 and donot affect binding to CPB-1.

Yeast two- and three-hybrid assays

Point mutants were generated using site-directedmutagenesis (Invitrogen). Mixed base oligonucleotideswere obtained by machine mixed random codons (IDT). Inthe yeast two-hybrid assays, CPB-1(1–80) was cloned intopBTM116 and FBF-2(121-C-term) was cloned into pACT2and expressed in strain L40U−. Modified yeast three-hybrid experiments were conducted as previouslydescribed.30,34 Yeast strain YBZ-1 was transformed withan additional plasmid encoding p414TEF CPB-1(40–80)fused to an SV40 nuclear localization signal. Measure-ments were determined in triplicate from three individualyeast transformants. Luminescence data were collectedusing the β-Glo reagent (Promega).

Tryptic digestion of CPB-1

His-tagged FBF-1(164–566) (11 μM) was incubated onice for 2 hwith 15 μMCPB-1(1–80) in binding buffer [20 mM

735Characterization of C. elegans FBF•CPB-1 complex

Na-phosphate (pH 8.6), 100 mMNaCl, and 10 mM βME]. Afreshly prepared solution of ethylene glycol bis(succinimi-dylsuccinate) (EGS) in dimethyl sulfoxide was added to afinal concentration of 1.25 mM to half of the reactionmix andthe samplewas incubated on ice for 5 min. The cross-linkingreactionwas quenched by adding 50 mMTris–HCl (pH 8.6).Trypsin was added to both samples to a final concentrationof 0.5 ng/μl and the reactions were incubated at roomtemperature for 50 min. The tryptic digestion was stoppedby adding 3% formic acid. The trypsinized fragments wereidentified by matrix-assisted laser desorption/ionizationtime-of-flight mass spectrometry. The CPB-1 fragmentsthat were protected upon binding to FBF were identified bycomparison of the digestion pattern of the cross-linkedFBF ⋅CPB-1 sample with the non-cross-linked sample.

Protein cross-linking

Cross-linking experiments were performed in parallel withboth FBF-1(164–566) and FBF-2(164–566) and yielded thesame results regardless of the protein used. A constantconcentration of FBF (7.5 μM) was equilibrated with a 2-foldmolar excess of GLD-3(860–949) in 20 mM sodiumphosphate (pH 8.5), 200 mM NaCl, and 10 mM βME in20 μl total volume. Increasing concentrations of CPB-1(1–80) were added to a final concentration of 3.8 μM, 15.0 μM,and 30.0 μM. Half a microliter of freshly prepared 40 mMsolution of EGS in dimethyl sulfoxide was added and thecross-linking reaction was allowed to proceed for 1 min atroom temperature. The reaction was quenched by adding1 μl of 100 mM Tris–HCl (pH 8.6), and the samples wereanalyzed by SDS-PAGE. The same type of experiment wasrepeated with FBF incubated with constant amount of CPB-1 andwhereGLD-3 was added in a 0.5-, 2-, and 4-foldmolarexcess prior to EGS cross-linking.

Labeling of CPB-1 with Alexa Fluor 488

A cysteine residue was introduced by PCRmutagenesisbetween residues Ser35 and Phe36 in the CPB-1(19–80)construct. Protein expression was performed as for wild-type protein. The cell pellet was resuspended in lysis buffersupplemented with 2 mM DTT and lysed as described forwild-type CPB-1. The cleared lysate was loaded onto Ni-NTA resin. The resin was washed with labeling buffer[20 mM Hepes (pH 7.0), 150 mM NaCl, and 2% glycerol]to remove the reducing agent. Precautions were taken toavoid cysteine oxidation (all the buffers were thoroughlydegassed and the resin was flushed with nitrogen). Theprotein was incubated with a 5-fold molar excess of AlexaFluor 488 C5 maleimide (Invitrogen) while on the resin.Labeled CPB-1 was eluted and dialyzed in 20 mM Tris–HCl (pH 7.0), 50 mM NaCl, 10% glycerol, and 2 mM DTT.The protein was loaded onto Q column (GE Healthcare)equilibrated with the same buffer. CPB-1(19–80) did notbind to the column in these conditions but was present inthe flow through N95% free of contaminants.

Quantitative analysis of protein–protein interactionsby fluorescence polarization

Fluorescencepolarization experimentswere carriedout in96-well opaque fluotrak 200-μl plates (Greiner); the polari-

zation was determined using an Envision plate reader(Perkin Elmer). In the direct titration experiments, 6 nM fl-CPB-1 was incubated for at least 1 h at room temperaturewith increasing concentrations of wild-type or mutant FBF-2(164–575) in a total volume of 100 μl. The equilibrationbuffer contained 20 mM Tris–HCl (pH 8.6), 100 mM NaCl,2 mM DTT, 0.10 mM ethylenediaminetetraacetic acid,0.10 mg/ml tRNA, 0.05 mg/ml bovine serum albumin, and5% glycerol. The experiments were performed at least intriplicate and data were fitted with the Hill equation usingIGOR (Wavemetric) to determine the Kd. For the binding-competition experiments, 6 nM fl-CPB-1 was mixed with200 nM wild-type FBF-2(164–575) in the same buffer as forthe direct titration experiments and increasing concentra-tions of unlabeled competitor CPB-1 were added. IC50valueswere calculated by fitting the data to theHill equation.

RNA binding assay

A constant concentration of 75 pM 32P-labeled cyb-1RNA (5′-CGAAAUAAACAUUUUGUACCAUUCAGUC-3′)was incubated for 1 h at room temperature with increasingconcentrations of FBF-2(164–575) alone or pre-incubatedfor 2 h with 6 μM MBP-CPB-1(1–80) in 10 mM Tris–HCl(pH 7.5), 0.5 mM ethylenediaminetetraacetic acid, 50 mMNaCl, 2 mM DTT, 0.1 mg/ml bovine serum albumin, 0.02%Tween, and 0.2 mg/ml tRNA. Samples were loaded ontonative polyacrylamide gel, visualized, and analyzed essen-tially as described in Wu et al.31 The FBF Y479A pointmutant and theFBFΔ(479–485) deletionmutantwere testedfor binding to 32P-labeled FBE RNA in the same conditions.

NMR spectroscopy

15N-labeled CPB-1(32–80) was expressed in M9 min-imal medium containing 1 g/l (15NH4)2 SO4 (CambridgeIsotope Laboratories) as the source of nitrogen. Theprotein was purified as described above and bufferexchanged into 10 mM Tris–HCl (pH 6.0), 50 mM NaCl,and 5 mM DTT. For complex reconstitution, 15N-labeledCPB-1(32–80) was mixed with 1.5 molar excess of FBF-2(164–575)/FBE RNA (5′-UGUGCCAUA-3′) and the sam-ples were concentrated to approximately 0.7 mM. NMRsamples were prepared by adding 15% (v/v) D2O. Thetwo-dimensional [15N,1H]HSQC spectra were recorded at298 K and 700 MHz on a Bruker DRX spectrometer.

CD measurements

The CD spectrum of CPB-1(1–80) was recorded on aJASCO J-815 CD spectropolarimeter at a protein concen-tration of 47 μM in 20 mM sodium phosphate (pH 8.0) and20 mM NaCl at 20 °C in a 0.1-cm cuvette.

Acknowledgements

We wish to thank Pedro Serrano Navarro for helpwith [15N,1H]HSQC data acquisition and usefuldiscussions and Christine Beuck and John Hammond

736 Characterization of C. elegans FBF•CPB-1 complex

for comments on the manuscript. This work wassupported by the National Institutes of Health grantsF32 GM095169 (Z.T.C.), GM53320 (J.R.W),GM031892, and GM050942 (M.W.).

Supplementary Data

Supplementary data to this article can be foundonline at http://dx.doi.org/10.1016/j.jmb.2012.11.012

Received 26 July 2012;Received in revised form 1 November 2012;

Accepted 7 November 2012Available online 15 November 2012

Keywords:Pumilio;

PUF protein;FBF;

protein–protein interactions;germline development

† http://bioinf.cs.ucl.ac.uk/psipred/

Abbreviations used:PUF, Pumilio and FBF; FBF, fem-3 mRNA binding factor;

CPEB, cytoplasmic polyadenylation element binding;CPB-1, cytoplasmic polyadenylation element bindingprotein homolog-1; UTR, untranslated region; GLD-3,

germline development defective-3; RRM, RNA recognitionmotif; fl-CPB-1, fluorescently labeled wild-typeCPB-1(19–80); EMSA, electrophoretic mobility

shift assay; cyb-1, cyclin B of Caenorhabditis elegans;MBP, maltose binding protein; TEV, tobacco etch virus;

βME, beta-mercaptoethanol; EGS, ethylene glycolbis(succinimidylsuccinate).

References

1. Suh, N., Crittenden, S. L., Goldstrohm, A., Hook, B.,Thompson, B., Wickens, M. & Kimble, J. (2009). FBFand its dual control of gld-1 expression in the Caenor-habditis elegans germline. Genetics, 181, 1249–1260.

2. Crittenden, S. L., Bernstein, D. S., Bachorik, J. L.,Thompson, B. E., Gallegos, M., Petcherski, A. G.et al. (2002). A conserved RNA-binding proteincontrols germline stem cells in Caenorhabditiselegans. Nature, 417, 660–663.

3. Kaye, J. A., Rose, N. C., Goldsworthy, B., Goga, A. &L'Etoile, N. D. (2009). A 3′ UTR pumilio-bindingelement directs translational activation in olfactorysensory neurons. Neuron, 61, 57–70.

4. Richter, J. D. (2007). CPEB: a life in translation.Trends Biochem. Sci. 32, 279–285.

5. Radford, H. E., Meijer, H. A. & de Moor, C. H. (2008).Translational control by cytoplasmic polyadenylation

in Xenopus oocytes. Biochim. Biophys. Acta, 1779,217–229.

6. Goldstrohm, A. C. & Wickens, M. (2008). Multifunc-tional deadenylase complexes diversifymRNA control.Nat. Rev. Mol. Cell Biol. 9, 337–344.

7. Luitjens, C., Gallegos, M., Kraemer, B., Kimble, J. &Wickens, M. (2000). CPEB proteins control two keysteps in spermatogenesis in C. elegans. Genes Dev.14, 2596–2609.

8. Hake, L. E., Mendez, R. & Richter, J. D. (1998).Specificity of RNA binding by CPEB: requirement forRNA recognition motifs and a novel zinc finger. Mol.Cell. Biol. 18, 685–693.

9. Pique, M., Lopez, J. M., Foissac, S., Guigo, R. &Mendez, R. (2008). A combinatorial code for CPE-mediated translational control. Cell, 132, 434–448.

10. Barnard, D. C., Ryan, K., Manley, J. L. & Richter, J. D.(2004). Symplekin and xGLD-2 are required forCPEB-mediated cytoplasmic polyadenylation. Cell,119, 641–651.

11. Kim, J. H. & Richter, J. D. (2006). Opposingpolymerase–deadenylase activities regulate cytoplas-mic polyadenylation. Mol. Cell, 24, 173–183.

12. Nakahata, S., Katsu, Y., Mita, K., Inoue, K.,Nagahama, Y. & Yamashita, M. (2001). Biochemicalidentification of Xenopus Pumilio as a sequence-specific cyclin B1 mRNA-binding protein that phys-ically interacts with a Nanos homolog, Xcat-2, and acytoplasmic polyadenylation element-binding protein.J. Biol. Chem. 276, 20945–20953.

13. Nakahata, S., Kotani, T., Mita, K., Kawasaki, T.,Katsu, Y., Nagahama, Y. & Yamashita, M. (2003).Involvement of Xenopus Pumilio in the translationalregulation that is specific to cyclin B1 mRNA duringoocyte maturation. Mech. Dev. 120, 865–880.

14. Goldstrohm, A. C., Seay, D. J., Hook, B. A. &Wickens,M. (2006). PUF protein-mediated deadenylation iscatalyzed by Ccr4p. J. Biol. Chem. 282, 109–114.

15. Goldstrohm, A. C., Hook, B. A., Seay, D. J. &Wickens, M. (2006). PUF proteins bind Pop2p toregulate messenger RNAs. Nat. Struct. Mol. Biol. 13,533–539.

16. Friend, K., Campbell, Z. T., Cooke, A., Kroll-Conner,P., Wickens, M. P. & Kimble, J. (2012). A conservedPUF–Ago–eEF1A complex attenuates translationelongation. Nat. Struct. Mol. Biol. 19, 176–183.

17. Kraemer, B., Crittenden, S., Gallegos, M., Moulder,G., Barstead, R., Kimble, J. & Wickens, M. (1999).NANOS-3 and FBF proteins physically interact tocontrol the sperm-oocyte switch in Caenorhabditiselegans. Curr. Biol. 9, 1009–1018.

18. Eckmann, C. R., Crittenden, S. L., Suh, N. & Kimble, J.(2004). GLD-3 and control of the mitosis/meiosisdecision in the germline of Caenorhabditis elegans.Genetics, 168, 147–160.

19. Zhang, B., Gallegos, M., Puoti, A., Durkin, E., Fields,S., Kimble, J. & Wickens, M. P. (1997). A conservedRNA-binding protein that regulates sexual fates in theC. elegans hermaphrodite germ line. Nature, 390,477–484.

20. Edwards, T. A., Pyle, S. E., Wharton, R. P. &Aggarwal, A. K. (2001). Structure of Pumilio revealssimilarity between RNA and peptide binding motifs.Cell, 105, 281–289.

737Characterization of C. elegans FBF•CPB-1 complex

21. Wang, X., Zamore, P. D. & Hall, T. M. (2001). Crystalstructure of a Pumilio homology domain. Mol. Cell, 7,855–865.

22. Wang, X., McLachlan, J., Zamore, P. D. & Hall, T. M.(2002). Modular recognition of RNA by a humanpumilio-homology domain. Cell, 110, 501–512.

23. Edwards, T. A., Wilkinson, B. D., Wharton, R. P. &Aggarwal, A. K. (2003). Model of the brain tumor–Pumilio translation repressor complex. Genes Dev.17, 2508–2513.

24. Eckmann, C. R., Kraemer, B., Wickens, M. & Kimble,J. (2002). GLD-3, a bicaudal-C homolog that inhibitsFBF to control germline sex determination inC. elegans. Dev. Cell, 3, 697–710.

25. Dong, S., Wang, Y., Cassidy-Amstutz, C., Lu, G.,Bigler, R., Jezyk, M. R. et al. (2011). Specific andmodular binding code for cytosine recognition inPumilio/FBF (PUF) RNA-binding domains. J. Biol.Chem. 286, 26732–26742.

26. Filipovska, A., Razif, M. F., Nygard, K. K. & Rackham,O. (2011). A universal code for RNA recognition byPUF proteins. Nat. Chem. Biol. 7, 425–427.

27. Kaymak, E., Wee, L. M. & Ryder, S. P. (2010).Structure and function of nematode RNA-bindingproteins. Curr. Opin. Struct. Biol. 20, 305–312.

28. Bernstein, D., Hook, B., Hajarnavis, A., Opperman, L.& Wickens, M. (2005). Binding specificity and mRNAtargets of a C. elegans PUF protein, FBF-1. RNA, 11,447–458.

29. Wang, Y., Opperman, L., Wickens, M. & Hall, T. M.(2009). Structural basis for specific recognition ofmultiple mRNA targets by a PUF regulatory protein.Proc. Natl Acad. Sci. USA, 106, 20186–20191.

30. Campbell, Z. T., Bhimsaria, D., Valley, C. T.,Rodriguez-Martinez, J. A., Menichelli, E., Williamson,J. R. et al. (2012). Cooperativity in RNA–proteininteractions: global analysis of RNA binding specific-ity. Cell Rep. 1, 570–581.

31. Wu, J., Campbell, Z. T., Menichelli, E., Wickens, M. &Williamson, J. R. (2012). Protein ⋅protein interactionsequences specifying the recruitment of GLD-3 to theFBF ⋅ fem-3 mRNA complex. J. Mol. Biol. http://dx.doi.org/10.1016/j.jmb.2012.11.013.

32. Dyson, H. J. & Wright, P. E. (2005). Intrinsicallyunstructured proteins and their functions. Nat. Rev.Mol. Cell Biol. 6, 197–208.

33. Petsalaki, E. & Russell, R. B. (2008). Peptide-mediated interactions in biological systems: new

discoveries and applications. Curr. Opin. Biotechnol.19, 344–350.

34. Campbell, Z. T., Menichelli, E., Friend, K., Wu, J.,Kimble, J., Williamson, J. R. & Wickens, M. (2012). Aconserved interface between PUF and CPEB pro-teins. J. Biol. Chem. 287, 18854–18862.

35. Thompson, J. D., Higgins, D. G. & Gibson, T. J.(1994). CLUSTAL W: improving the sensitivity ofprogressive multiple sequence alignment throughsequence weighting, position-specific gap penaltiesand weight matrix choice. Nucleic Acids Res. 22,4673–4680.

36. Sanger, F., Nicklen, S. & Coulson, A. R. (1977). DNAsequencing with chain-terminating inhibitors. Proc.Natl Acad. Sci. USA, 74, 5463–5467.

37. Campbell, Z. T., Menichelli, E., Friend, K., Wu, J.,Kimble, J., Williamson, J. R. & Wickens, M. (2012).Identification of a conserved interface betweenPUF and CPEB proteins. J. Biol. Chem. 287,18854–18862.

38. Kershner, A. M. & Kimble, J. (2010). Genome-wideanalysis of mRNA targets for Caenorhabditis elegansFBF, a conserved stem cell regulator. Proc. Natl Acad.Sci. USA, 107, 3936–3941.

39. Zamore, P. D., Williamson, J. R. & Lehmann, R.(1997). The Pumilio protein binds RNA through aconserved domain that defines a new class of RNA-binding proteins. RNA, 3, 1421–1433.

40. Walker, J., Minshall, N., Hake, L., Richter, J. &Standart, N. (1999). The clam 3′ UTR maskingelement-binding protein p82 is a member of theCPEB family. RNA, 5, 14–26.

41. Wang, L., Eckmann, C. R., Kadyk, L. C., Wickens, M.& Kimble, J. (2002). A regulatory cytoplasmic poly(A)polymerase in Caenorhabditis elegans. Nature, 419,312–316.

42. Sonoda, J. & Wharton, R. P. (2001). Drosophila braintumor is a translational repressor. Genes Dev. 15,762–773.

43. Galgano, A., Forrer, M., Jaskiewicz, L., Kanitz, A.,Zavolan, M. & Gerber, A. P. (2008). Comparativeanalysis of mRNA targets for human PUF-familyproteins suggests extensive interaction with themiRNA regulatory system. PLoS One, 3, e3164.

44. Cooke, A., Prigge, A., Opperman, L. & Wickens, M.(2011). Targeted translational regulation using thePUF protein family scaffold. Proc. Natl Acad. Sci.USA, 108, 15870–15875.