Microrna-Repressed Mrnas Contain 40S but Not 60S Components

MicroRNA-repressed mRNAs contain 40S but not 60S components

Bingbing Wang*, Adrienne Yanez*, and Carl D. Novina*†‡

*Department of Cancer Immunology and AIDS, Dana–Farber Cancer Institute and Department of Pathology, Harvard Medical School, Boston, MA 02115; and †Broad Institute of Harvard and Massachusetts Institute of Technology, Cambridge, MA 02141

Communicated by Phillip A. Sharp, Massachusetts Institute of Technology, Cambridge, MA, February 5, 2008 (received for review June 4, 2007) MicroRNAs (miRNAs) are small noncoding RNAs that may target Northern blot analysis, Western blot analysis, and toeprinting more than one-third of human genes, yet the mechanisms used analyses of miRNA-targeted mRNAs. by miRNAs to repress translation of target mRNAs are obscure. Using a recently described cell-free assay of miRNA function, we Results observe that miRNA-targeted mRNAs are enriched for 40S but miRNA-Targeted mRNAs Contain Reduced Amounts of 60S Ribosome not 60S ribosome components. Additionally, toeprinting analy- Components. To investigate the mechanism of repressed translation sis of miRNA-targeted mRNAs demonstrates that Ϸ18 nt 3؅ initiation, we tested the ability of the 40S and 60S ribosome subunits relative to the initiating AUG are protected, consistent with 40S to physically associate with miRNA-targeted mRNAs in the cell- ribosome subunits positioned at the AUG codon. Our results free reactions described first in Wang et al. (6). To enable recovery suggest that miRNAs repress translation initiation by preventing of miRNA-targeted mRNAs, firefly luciferase reporter mRNAs 60S subunit joining to miRNA-targeted mRNAs. were subjected to polyadenylation such that, on average, two biotinylated adenosines were incorporated into a polyA tail con- Ϸ argonaute ͉ translation ͉ ribosome sisting of 200 adenosines. These reporter mRNAs contained six imperfectly complementary binding sites to the CXCR4 siRNA (FL6X). Consistent with previous observations (6), the FL6X icroRNAs (miRNAs) are endogenous small RNAs that may reporter mRNA demonstrated translation repression upon addi- Mregulate large networks of genes in several species. miRNAs tion of CXCR4 siRNAs when normalized to an untargeted renilla bind to target mRNAs with imperfect sequence complementarity luciferase reporter mRNA lacking CXCR4 siRNA-binding sites and repress translation through mechanisms that are incompletely (RL0X, Fig. 1A). understood (reviewed in ref. 1). Intensive efforts have focused on Upon completion of the translation repression reactions, re- determining the precise stage of translation repressed by miRNAs. porter mRNAs were precipitated with streptavidin beads, and Early observations suggested that lin-4, the first miRNA described precipitates were subjected to Northern blot analysis for ribosomal in animals, represses translation of its target mRNAs lin-14 or lin-28 RNAs (Fig. 1B). In all analyses in Fig. 1B, a fixed amount of RNA after initiation, because the distribution of target mRNAs in was loaded in each lane, and Northern blot analysis signals were polysomes is similar to that of untargeted mRNAs (2, 3). Similar to normalized by calculating the ratio of 60S rRNA:40S rRNA and these early studies in worms, a recent report in mammals also tRNAi-Met:40S rRNA. Northern blot analysis probes directed indicated that miRNA-targeted mRNAs are found in the same against 18S rRNA, a component of the 40S ribosome subunit, polysomal fractions as their untargeted counterparts (4). In con- detected negligible differences in 40S content between reactions trast, another report in mammals indicated that polysome profiles containing or lacking siRNAs. However, Northern blot analysis of miRNA-targeted mRNAs shift toward monosomes, indicating a probes directed against 28S, 5.8S, and 5S rRNAs of the 60S translation initiation block (5). The causes and consequences of ribosome subunit detected significant reductions in 60S ribosome components relative to the 18S rRNA in reactions containing these discrepancies in the mechanism of miRNA-dependent trans- BIOCHEMISTRY lation repression have not been resolved (1). mRNAs targeted by miRNAs (Fig. 1B). Additionally, the reduction Recently, several cell-free miRNA-dependent translation repres- in 28S, 5.8S, and 5S rRNAs associated with FL6X was approxi- sion reactions have been described. We reported the first cell-free mately the same as the degree of translational repression observed translation repression reactions (6), which faithfully recapitulate in Fig. 1A (Ϸ60%). These results indicate that miRNAs promote important properties of miRNA function in cells including require- reduced 60S ribosome subunit loading on target mRNAs. Con- ments for 5Ј phosphates on miRNAs (7, 8) and perfect seed region versely, Northern blot analysis probes directed against tRNAi-Met complementarity between miRNAs and target mRNAs (9–11). detected no change relative to the 18S rRNA in reactions contain- Importantly, translation is repressed without reduction in target ing mRNAs targeted by miRNA compared with untargeted mR- mRNA levels. However, significant reduction in target mRNA NAs. This result indicates that miRNAs permit 43S ribosome levels is observed when perfectly complementary siRNAs are subunit loading on target mRNAs. added to these reactions. Additionally, these translation repression Several control reactions confirm the specificity of reduced 60S reactions directly demonstrated a dependence on a 7-methyl ribosome recruitment to miRNA-targeted mRNAs: (i) There was no significant change in the 40S or 60S ribosomes in Northern blots guanosine cap (5, 12) and a polyA tail (12) on target mRNAs for of total lysates without precipitation of miRNA-targeted mRNAs translational repression as observed in cells. Other cell-free trans- from translation repression reactions, suggesting that the ribosome lation repression reactions have been described recently. These subunits are not being lost because of degradation. Relative to reactions also demonstrate a requirement for 7-methyl guanosine capped target mRNAs for translational repression in mouse (13), human (14), and fly (15) extracts, further supporting a model of Author contributions: B.W., A.Y., and C.D.N. designed research; B.W. and A.Y. performed miRNA repression of translation initiation. Still, the precise mech- research; B.W., A.Y., and C.D.N. analyzed data; and C.D.N. wrote the paper. anisms of miRNA function are unknown. We define a mechanism The authors declare no conﬂict of interest. of miRNA-directed repression of translation initiation by decreased ‡To whom correspondence should be addressed. E-mail: carl࿝[email protected]. 60S ribosome recruitment to target mRNAs using cell-free mono- This article contains supporting information online at www.pnas.org/cgi/content/full/ cistronic and bicistronic miRNA reporter assays, ribosome-binding 0801102105/DCSupplemental. assays, precipitation of miRNA-targeted mRNAs followed by © 2008 by The National Academy of Sciences of the USA

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0801102105 PNAS ͉ April 8, 2008 ͉ vol. 105 ͉ no. 14 ͉ 5343–5348 Downloaded by guest on September 24, 2021 A - + siCXCR4 C T S P - + - + - + siCXCR4 FL RPL18 (60S) RL RPS7 (40S) 73 ± 2 % Reduction FL

B siCXCR4 5’ unphospho. siCXCR4 5’ seed mut. FL6X T S P T S P T S P - + - + - + - + - + - + - + - + - + siRNA 28S rRNA

60S 5.8S rRNA 5S rRNA

18S rRNA 40S

tRNAi-Met T S P T S P T S P 2 1 0 28S rRNA

e 2 g n g 1 5.8S rRNA

a h C h a 0

d 2 l

o 1 F 0 5S rRNA 2 1 0 tRNAi-Met

Fig. 1. miRNA-targeted mRNAs have reduced 60S ribosome components. (A) Translation repression reactions using FL6X and RL0X mRNAs without (Ϫ) and with (ϩ) CXCR4 siRNAs (siCXCR4). Both mRNAs possessed 7-methyl-guanosine caps and polyA tails. The FL6X mRNA contained biotinyl-adenosines (99:1 adenosines: biotinyl-adenosines). Translation repression of ﬁreﬂy luciferase (FL, % reduction FL) is normalized by renilla luciferase (RL) and calculated by the equation [1ϪFL/RL(ϩsiCXCR4):FL/RL(ϪsiCXCR4)] x 100%. (B) Northern blot analysis for 40S and 60S rRNAs in total (T) translation repression reactions, supernatants (S), and precipitates (P) after streptavidin precipitation of biotinylated mRNA reporters from total translation repression reactions. Streptavidin-precipitated FL6X mRNAs from reactions with (ϩ) CXCR4 siRNA demonstrate reduced amounts of 60S rRNAs relative to reactions lacking (Ϫ) CXCR4 siRNA (Left). Streptavidin-precipitated FL6X mRNAs from reactions with (ϩ) and without (Ϫ) unphosphorylated CXCR4 siRNA (5Ј unphospho. siCXCR4, Middle). Streptavidin-precipitated FL6X mRNAs with point mutations in the 5Ј seed region (5Ј seed mut. FL6X) from reactions with (ϩ) and without (Ϫ) CXCR4 siRNA (Right). Signal intensities were normalized by calculating 60S (5S, 5.8S, and 28S):40S (18S) rRNAs and initiator methionine tRNA (tRNAi-Met):40S rRNA. The fold changes in signal intensities between reactions containing and lacking siRNAs were calculated by the equation [(60S/40SϩsiCXCR4):(60S/40SϪsiCXCR4)] or [(tRNAiϪMet/40SϩsiCXCR4):(tRNAiϪMet/ 40SϪsiCXCR4)]. Results are presented as bar graphs below each image as an average of n ϭ 3 trials (Left), n ϭ 2 trials (Middle), and n ϭ 2 trials (Right). (C) Western blot analysis for 40S- and 60S-associated proteins in T, S, and P after streptavidin precipitation of biotinylated mRNA reporters from total translation repression reactions. The precipitate from reactions with (ϩ) CXCR4 siRNA demonstrates a slightly increased amount of 40S-associated protein (RPS7) and a strongly decreased amount of 60S-associated protein (RPL18). In corresponding lanes between reactions without (Ϫ) or with (ϩ) CXCR4 siRNAs in T, S, and P, identical total protein amounts were loaded.

reactions with siRNAs, there was no change in the amounts of 60S amounts of 40S-associated proteins were detected on miRNA- ribosome subunits between reactions using (ii) FL6X mRNAs with repressed mRNAs, strongly decreased amounts of 60S-associated unphosphorylated CXCR4 siRNAs (Fig. 1B Middle), (iii) FL6X proteins were detected on these same mRNAs. Together, these mRNAs containing point mutations in the 5Ј seed region of the observations demonstrate that miRNA-targeted mRNAs have miRNA-binding sites in the 3Ј UTR (Fig. 1B Right), (iv) FL6X steady-state levels of 40S ribosome components but reduced levels mRNAs with nonspecific control siRNAs [supporting information of 60S ribosome components relative to untargeted mRNAs. (SI) Fig. S1C], and (v) FL0X mRNAs with CXCR4 siRNAs (Fig. S1C). The absence of significant changes in 40S and 60S ribosomal Chemical Inhibitors Identify High Molecular Mass Complex Formation RNA content in these control reactions is consistent with the on miRNA-Targeted mRNAs That Depends on 40S but Not 60S Ribo- absence of translational repression in these control reactions (Fig. somes. To define the stage of translation initiation affected by 1 A and Fig. S1B) and indicates that the reduction in 60S ribosome miRNAs more precisely, we used ribosome-binding assays to RNAs is specific to miRNA-repressed mRNAs. analyze the sedimentation profiles of radiolabeled miRNA- To independently confirm these observations, precipitated re- targeted mRNAs from translation repression reactions (Fig. 2). porter mRNAs were subjected to Western blot analysis for 40S- and Similar to all reports of polysome profiling of miRNA targeted 60S-associated proteins (Fig. 1B). Whereas slightly increased mRNAs (2–5, 15, 16), ribosome-binding assays require chemicals to

5344 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0801102105 Wang et al. Downloaded by guest on September 24, 2021 0.18 0.18 A Hippuristanol 5’ unphospho. siRNA 0.16 D 0.16 -siRNA -siRNA 100 0.14 +siRNA 0.14 +siRNA G )%( LF fo n fo LF )%( 0.12 0.12 75 0.10 0.10

0.08 0.08 50 o ita 0.06 0.06 l snarT 0.04 0.04 25

0.02 0.02 0 0.00 0.00 - + siCXCR4

0.18 0.18 .

GMP-PNP . 5’ seed mut. FL6X B m.p. 0.16 0.16 E m.p.c FL6X, Cycloheximide -siRNA -siRNA H c la c 0.14 +siRNA 0.14 +siRNA l T S P tot fo %( X %( fo tot atot f atot 0.12 0.12 - + - + - + siCXCR4

0.10 % o 0.10 28S rRNA

0.08 X ( 0.08

0.06 0.06 60S

delebal delebal 0.04 deleb 0.04 5.8S rRNA 0.02 0.02 5S rRNA a l

F6) PFL6 2 0.00 ) PFL6 0.00 3 2

3 18S rRNA 40S 0.18 0.18 Cycloheximide Uncapped FL6X C 0.16 F 0.16 tRNAi-Met -siRNA -siRNA 0.14 +siRNA 0.14 +siRNA T S P 0.12 0.12 2

1 0.10 0.10 0 28S rRNA egna 0.08 0.08 2 1 0 5.8S rRNA

0.06 0.06 h C 2

0.04 0.04 d 1 loF 0 5S rRNA 0.02 0.02 2 1 0.00 0.00 0 tRNAi-Met 0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10121416182022 Fraction Fraction

Fig. 2. A high molecular mass complex containing 40S but lacking 60S ribosome subunits forms on miRNA-repressed mRNAs. (A–F) Ribosome-binding assays using 7-methyl-guanosine capped or uncapped FL6X mRNAs containing 32P-labeled polyA tails. (A) Ribosome-binding analysis of translation repression reactions containing hippuristanol identifies a peak (fraction 4) corresponding to unbound mRNAs in reactions without (Ϫ, dotted line) and with (ϩ, solid line) CXCR4 siRNAs. (B) Ribosome-binding analysis of translation repression reactions containing GMP-PNP identifies a 48S complex (fraction 8) in reactions without (Ϫ, dotted line) and a high molecular mass complex in reactions with (ϩ, solid line) CXCR4 siRNAs. (C–F) Ribosome-binding assays of translation repression reactions containing cycloheximide identify an 80S complex (fraction 12) in all reactions without CXCR4 siRNAs (Ϫ, dotted line). (C) A high molecular mass complex is formed on FL6X mRNAs in reactions with CXCR4 siRNAs (ϩ, solid line). (D) An 80S complex is formed on FL6X mRNAs in reactions with unphosphorylated CXCR4 siRNAs (ϩ, solid line, 5Ј unphospho. siRNA). (E) An 80S complex is formed on FL6X mRNAs with three point mutations in the 5Ј seed region (5Ј seed mut. FL6X) in reactions with CXCR4 siRNAs (ϩ, solid line). (F) An 80S complex is formed on uncapped FL6X mRNAs in reactions with CXCR4 siRNAs (ϩ, solid line). Horizontal axis indicates fraction number. Vertical axis indicates the cpm (c.p.m) of each fraction as a percent of total counts recovered (% of total c.p.m.). In B–F, the upward pointing arrows indicate peak fractions of mRNA sedimentation in glycerol gradients. (G) Dual luciferase assay using FL6X mRNAs containing 32P-labeled polyA tails in the presence of CXCR4 siRNAs. Firefly luciferase (FL) measurements were normalized to renilla luciferase expressed from RL0X mRNAs with an unlabeled cap and BIOCHEMISTRY a polyA. Bars indicate percentage translation of FL (%,y axis). (H) Northern blot analysis of translation repression reactions in the presence of cycloheximide using FL6X mRNAs with 7-methyl-guanosine caps and biotinylated polyA tails from reactions without (Ϫ) or with (ϩ) CXCR4 siRNA. Total translation repression reactions (T), supernatants (S), and precipitates (P) from total translation repression reactions after streptavidin precipitation of biotinylated FL6X mRNA were Northern blotted for 40S and 60S rRNAs, and initiator methionine tRNA (tRNAi-Met) as in Fig. 1B. Results are presented as a bar graphs with error bars indicating the standard deviation of n ϭ 2 trials.

stabilize intermediates of monosome (80S) assembly. In ribosome- performed in the presence of the nonhydrolyzable GTP analog binding assays, addition of CXCR4 siRNAs to translation repres- GMP-PNP, which blocks 60S ribosome recruitment, resulting in the sion reactions without chemical inhibitors is not sufficient to capture of 48S initiation complexes (fraction 8) in ribosome-binding capture complexes at specific stages of ribosome assembly during assays. When CXCR4 siRNAs were added to translation repression translational repression (data not shown). Complexes stabilized at reactions containing GMP-PNP, the mRNA sedimentation peak specific stages of ribosome assembly on radiolabeled mRNAs are was shifted from fraction 8 to fraction 14, indicating the formation sedimented through a glycerol gradient and detected by Cerenkov of a high molecular mass complex (Fig. 2B). These results suggest scintillation counting of individual glycerol gradient fractions (re- that this high molecular mass complex is formed on miRNA- viewed in ref. 17). targeted mRNA after 43S recruitment but before 60S recruitment. First, translation repression reactions were performed in the Then, translation repression reactions were performed in the presence of the eIF4A inhibitor hippuristanol (18). Hippuristanol presence of the translation elongation inhibitor cycloheximide, blocks 43S recruitment to mRNAs, resulting in the majority of the which traps fully assembled 80S monosomes at the initiation codon mRNA migrating at the top of the glycerol gradient as unbound of mRNAs (fraction 12). Reactions without CXCR4 siRNA led to mRNA (fraction 4) in ribosome-binding assays. Addition of the expected mRNA sedimentation, consistent with captured 80S CXCR4 siRNA to translation repression reactions did not alter the complexes (Fig. 2 C–F). Addition of CXCR4 siRNAs to translation sedimentation of mRNAs in the presence of hippuristanol (Fig. repression reactions containing cycloheximide, however, generated 2A), indicating that any complexes that form on repressed mRNAs an mRNA sedimentation profile identical to the profile observed require 43S recruitment. Next, translation repression reactions were with GMP-PNP (Fig. 2C). This observation indicates that 80S

Wang et al. PNAS ͉ April 8, 2008 ͉ vol. 105 ͉ no. 14 ͉ 5345 Downloaded by guest on September 24, 2021 monosomes do not form in these reactions, and that the repression AB occurs at an earlier step of translation initiation. -+-++ -- -Hipp. 4 8 12 14 Formation of the high molecular mass complex was specific to -+-+--- + GMP-PNP siCXCR4 ACGU -+-- -+ - + siCXCR4 - the CXCR4 siRNA and its interaction with its target site. Inclusion eIF2α of unphosphorylated CXCR4 siRNAs (Fig. 2D), a triple point + Ј mutant in the 5 seed region of the miRNA-binding site in the FL6X - mRNA 3Ј UTR (Fig. 2E), nonspecific control siRNA (Fig. S2C)or eIF3g fully phosphorylated CXCR4 siRNAs and the FL0X mRNA (Fig. + S2D) all resulted in the formation of an 80S complex in the presence - eIF4E of cycloheximide and had no effect in translation repression reac- + tions (Fig. 2G and Fig. S1B). Consistent with the lack of transla- A U 1

tional repression of uncapped mRNAs (ref. 6; Fig. S3), FL6X 8 G -

n eIF4A lacking a 7-methyl guanosine cap did not result in formation of the t + high molecular mass complex in reactions containing CXCR4 C - siRNAs but instead resulted in formation of an 80S complex (Fig. Ago2 2F). These results demonstrate that the high molecular mass + complex forms only on translationally repressed mRNAs and - RPS7 possesses the 40S ribosome subunit but lacks the 60S ribosome + subunit, further supporting a model that miRNAs repress transla- - tion by preventing 60S ribosome subunit recruitment to target RPL18 + mRNAs. Additionally, the ability of the eIF4A inhibitor, hippurista- 1 2 3 4 5 6 7 8 9 10 11 12 nol, to prevent formation of the high molecular mass complex suggests that miRNA-directed repression of translation occurs after Fig. 3. miRNAs repress translation after 48S scanning but before 60S subunit cap-facilitated, 40S ribosome subunit recruitment. joining. (A) Toeprinting assay of FL6X mRNAs with a 7-methyl-guanosine cap To rule out any effect of cycloheximide in the reactions in Fig. and polyA tail. Compared with reactions without (Ϫ) CXCR4 siRNAs (siCXCR4, ϩ 2, we performed Northern blot analysis on precipitated complexes lane 5), reactions with ( ) CXCR4 siRNA (lane 6) demonstrate a strong toeprint that protects a region encompassing the initiating AUG codon. Addition of from translation repression reactions containing cycloheximide hippuristanol (hipp.) to translation repression reactions does not lead to a (Fig. 2H). Consistent with data presented in Fig. 1B, Northern blot toeprint (lanes 7 and 8) and blocks the CXCR4 siRNA-induced toeprint (lane 9). analysis of precipitated FL6X demonstrated similar levels of 18S Addition of GMP-PNP to translation repression reactions indicates the 40S ribosomal RNA but reduced 28S, 5.8S, and 5S ribosomal RNAs toeprint (lanes 10 and 11) and augments the CXCR4 siRNA-induced toeprint (but no change in tRNAi-met) associated with target mRNAs in (lane 12). A labeled ladder (lanes 1–4) indicates nucleotide positions relative reactions with CXCR4 siRNA compared with reactions without to AUG. (B) Western blot analysis using antibodies to eIF2␣, eIF3g, eIF4E, CXCR4 siRNAs. Also consistent with data presented in Fig. 1B, eIF4A, Ago2, RPS7 (40S-associated protein), and RPL18 (60S-associated pro- 60S ribosome association was reduced to a similar degree (Ϸ70%) tein) detects proteins that coprecipitate with FL6X mRNAs from glycerol gradient fractions 4 (free mRNA), 8 (48S peak), 12 (80S peak), and 14 (high as translational repression observed in Figs. 1A and 2G. molecular mass complex peak) in ribosome-binding assays. In corresponding lanes between reactions without (Ϫ) or with (ϩ) CXCR4 siRNAs, identical total 48S Complexes Are Positioned at AUG on miRNA-Repressed miRNAs. protein amounts were loaded. Although images are presented one above the To identify the position of 40S ribosomes assembled on miRNA- other, for each antibody, data were obtained from the same blot. repressed mRNAs, primer extension analysis was performed (Fig. 3A). In this assay, a radiolabeled primer hybridizing to sequences 3Ј relative to the AUG codon of miRNA-targeted mRNAs was used eIF2 and eIF3 Are Associated with miRNA-Targeted mRNAs. To inves- to initiate reverse transcription without extraction from associated tigate the complement of translation initiation factors associated proteins. A ‘‘toeprint’’ of bound protein complexes is generated with mRNAs in fraction 4 (free mRNA), fraction 8 (48S com- when steric hindrance prevents reverse transcriptase from tran- plexes), fraction 12 (80S complexes), and fraction 14 (high molec- scribing cDNA from regions of the mRNA. Translation repression ular mass complex), mRNA precipitates from glycerol gradient reactions with CXCR4 siRNAs generated bands at 18 nt 3Ј relative fractions in ribosome-binding assays were subjected to Western blot to the AUG codon (compare lanes 5 and 6). This toeprint was analysis. Because many translation initiation and miRNA- identical to translation repression reactions containing GMP-PNP interacting factors are highly conserved through evolution, anti- (lanes 10–12), which marks 40S ribosomes positioned at the start bodies against these human and mouse proteins crossreact with codon after completion of scanning. Consistent with the ability of their rabbit homologs. hippuristanol to block formation of the high molecular mass The translation initiation factors eIF2 and eIF3 are recruited to complex in ribosome-binding assays (Fig. 2A), the CXCR4 siRNA- 43S ribosome complexes before joining mRNAs and dissociate induced toeprint was blocked in the presence of hippuristanol from 48S ribosome complexes just before (or concomitant with) (compare lanes 6 and 9). 60S ribosome subunit joining mRNAs (reviewed in ref. 19). There- Toeprinting was quantified by using the ratio of the 3Ј (specific) fore, we probed ribosome-binding assay fractions for these factors band protected in toeprinting relative to the 5Ј (nonspecific) band known to assemble with the 40S subunit during translation initia- relative to AUG. RNA secondary structure causes MMLV reverse tion (Fig. 3B). The eIF2 subunit, eIF2␣, and the eIF3 subunit, transcriptase to dropoff of its template, thus generating a nonspe- eIF3g, were significantly enriched in fractions 12 and 14 from cific 5Ј band that can be used to quantify specific miRNA toeprint reactions with CXCR4 siRNAs compared with reactions without formation. By this measure, the CXCR4 siRNA-induced toeprint CXCR4 siRNAs. Together with toeprinting analysis, Western blot ratio was 2.3 (lane 6), the GMP-PNP-induced toeprint ratio was 6.9 analyses support a model in which miRNAs block translation after (lanes 10 and 11), and the combined ratio was 7.1 (lane 12), 3-fold 43S subunit joining and scanning but before eIF2 and eIF3 release more than the toeprint ratio induced by CXCR4 siRNA alone. and 60S ribosome subunit joining. These data indicate that GMP-PNP more strongly stabilizes 40S To interrogate the cap dependency of miRNA-directed transla- complexes positioned at AUG compared with miRNAs alone and tional repression and high molecular mass complex formation, we may help explain why the addition of miRNAs alone is not sufficient performed Western blot analysis of glycerol gradient fractions from to capture complexes in ribosome-binding assays. ribosome-binding assays for the cap-binding protein (eIF4E) and

5346 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0801102105 Wang et al. Downloaded by guest on September 24, 2021 ribosome recruitment to translationally repressed mRNAs in worms, humans (16), and flies (15). In worm and human cells, the S06 60S antiassociation factor eIF6 (23–26) associates with RNA- induced silencing complexes but not necessarily with miRNA- targeted mRNAs. Like the data presented here, in fly extracts, pseudopolysomes, nonpolysomal complexes of a molecular mass 4E 4A m7G Ͼ80S, form on miRNA-targeted mRNAs in the presence of both 4G 40S cycloheximide and GMP-PNP, indicating the absence of 60S sub- 2 units (15). Contrary to the cap dependency of the high molecular PABP miRNP Ago2 mass complex presented here, pseudopolysomes form on mRNAs lacking a 7-methyl guanosine cap. These observations suggest important similarities between miRNA-mediated translation repression across species but also imply distinguishing details in the mechanisms of miRNA-mediated repression in these organisms. Fig. 4. A model of miRNA-directed repression of translation initiation. Several translation initiation factors may interact with a recruited Ago protein The formation of a high molecular mass complex on miRNA- to repress translation including the cap-binding factor, eIF4E; the protein targeted mRNAs containing 40S but lacking 60S ribosome com- associated with the polyA tail, PABP; the bridging protein between cap ponents in ribosome-binding assays described here provides one structures and the polyA tails, eIF4G; the RNA helicase that unwinds local possible explanation for the rapid sedimentation of miRNA- mRNA secondary structure, eIF4A; and the multicomponent proteins associ- targeted mRNAs in polysome profiling assays observed in worms ated with the 40S ribosome, eIF3 and eIF2. (2, 3) and humans (4). Further analyses in cell-based and -free systems will more precisely define the mechanism(s) of miRNA function in mammals and their similarities and differences across the RNA helicase that facilitates 40S recruitment (eIF4A). In- species. creases in both of these factors were observed in reactions con- Ago2 (co-eIF2A) was originally defined as a ribosome-associated taining CXCR4 siRNAs relative to those lacking CXCR4 siRNAs protein that eluted in high salt (27) and that stabilized 40S- (compare fractions 12 and 14 without CXCR4 siRNAs vs. with containing complexes in the presence of mRNAs (28). The high CXCR4 siRNAs). Together, these observations indicate that eIF4E molecular mass complex formed on translationally repressed mR- and eIF4A are still bound to translationally repressed mRNAs after NAs possesses 40S ribosome subunits but lacks 60S ribosome 40S subunit joining and suggest that interaction of these proteins subunits. Consistent with a role in stabilizing 40S ribosomes asso- with the cap is important for translational repression by miRNAs. ciated with mRNAs, Ago2 is recruited to unrepressed mRNAs In all species capable of small RNA-directed gene silencing, [fraction 8, Ago2 (Ϫ), Fig. 3B Upper]. Ago2 is also recruited to microribonucleoprotein (miRNP) complexes possess a member of translationally repressed mRNA (fraction 14, Fig. 3B), possibly the Ago family of proteins (reviewed in ref. 20). To determine because these mRNAs possess increased amounts of 40S subunits whether Ago proteins are recruited to miRNA-targeted mRNAs in without joined 60S subunits. It has been shown that Ago2 interacts the reactions reported here, Western blot analysis was performed with the antiassociation factor eIF6 and thus 60S through TRBP with antibodies against Ago2 (Fig. 3B). Consistent with the notion and prevents 60S subunit joining to translationally repressed mR- that the high molecular mass complex formed on miRNA-repressed NAs (16). Our data present a complimentary mechanism that Ago2 mRNAs is a bona fide miRNP, fractions 12 and 14 were significantly interacts with translationally repressed mRNAs and prevents 60S enriched for Ago2 in reactions containing CXCR4 siRNAs relative subunit joining. Ago2 has also been shown to directly interact with to reactions lacking CXCR4 siRNAs. To demonstrate that Ago caps of translationally repressed mRNAs (22). Together, these data proteins recruited to miRNAs preannealed to mRNAs are func- suggest that Ago2 may function in more than one way to repress tional, Ago2-dependent RNA cleavage assays were performed. Our translation. data indicate that Ago2-mediated cleavage of target RNAs in vitro maps to the exact position as reported for Ago2-dependent cleavage Materials and Methods BIOCHEMISTRY in cells (Fig. S4 and SI Materials and Methods). These data indicate Translation Repression Reactions. All mRNA reporters used in these studies were that these reaction conditions permit formation of functional prepared and used as described in ref. 6. Plasmids expressing all of these mRNA miRNP/RISC on miRNA-repressed mRNAs. reporters are available from (www.addgene.org). The sequences of the CXCR4 siRNA were 5ЈP-GUUUUCACUCCAGCUAACACA-3 (sense strand) and 5ЈP- Discussion UGUUAGCUGGAGUGAAAACUU-3Ј (antisense strand). The sequences of the GFP The process of translation initiation is typically regulated at one of siRNA were 5ЈP-GGCUACGUCCAGGAGCGCACC-3Ј (sense strand) and 5ЈP- two steps: either at the 43S preinitiation complex formation or at UGCGCUCCUGGACGUAGCCUU-3Ј(antisense strand). The control mRNA reporter the ribosome recruitment phase (19). However, more specialized used in Supporting Online Fig. 3 was human CD3 (kind gift of Chenqi Xu, Dana-Farber Cancer Institute). mechanisms of translational control have been reported. The Translation repression reactions were performed as described in ref. 6. Brieﬂy, mechanism for miRNA-directed translation repression proposed preannealed mRNA reporter (0.025 pmol) and CXCR4 siRNA (0.15 pmol) were here is analogous to a previously identified 3Ј UTR regulatory incubated with a master mix containing 7 ␮l of nuclease-treated rabbit reticulo- ribonucleoprotein complex that represses translation by inhibiting cyte lysate (RRL, Promega), 4–8 units RNase Out (Invitrogen), 20 ␮M amino acid 60S subunit joining with the 40S subunit positioned at the AUG mixture (complete or minus methionine and cysteine, Promega), and 0.4 ␮l (5.7 codon of lipooxygenase mRNA (21). Because miRNAs may reg- ␮Ci) Promix L-[35S] in vitro cell labeling mix (Amersham Biosciences) at 30°C for 10 ulate large networks of genes, the mechanism of blocked 60S min. Reaction products were separated on 12% SDS/PAGE and transferred onto recruitment may be far more prevalent than originally anticipated. PVDF (BioRad) or subjected to dual luciferase assay. A model integrating the observations reported here is presented in Fig. 4. It is important to note that this model makes no Dual Luciferase Assay. Dual luciferase assays were performed according to the manufacturer’s protocols (Promega). Fireﬂy luciferase activity was measured by conclusions about whether the 7-methyl guanosine cap-associated adding 2 ␮l of each reaction with LAR I (20 ␮l) into one well of a 96-well plate and eIF4F components or Ago2 are part of the miRNA-dependent high read in Victor3 V (PerkinElmer) for 5 sec. Renilla luciferase activity was measured molecular mass complex. Indeed, it was recently shown that eIF4E by adding Stop & Glo (20 ␮l, Promega) to each well and reread for 5 sec. (13) and Ago2 (22) bind to 7-methylguanosine caps to mediate miRNA-directed repression of translation. Recently, two other Ribosome-Binding Assay. Ribosome-binding assays were performed as described groups reported miRNA repression consistent with reduced 60S in ref. 18. In vitro translation repression reactions supplemented with cyclohex-

Wang et al. PNAS ͉ April 8, 2008 ͉ vol. 105 ͉ no. 14 ͉ 5347 Downloaded by guest on September 24, 2021 imide (600 ␮M), GMP-PNP (1 mM), or hippuristanol (50 ␮M) were loaded onto anti-Ago2 antibody (Upstate) recognizes residues 7–48 of human Ago2, which 10–30% glycerol gradient containing 1ϫ HSB (500 mM NaCl; 20 mM Hepes-KOH, are conserved amino acids between human, mouse, cattle, dog, and frog. Anti- pH 7.5; 30 mM MgOAc; and 2 mM DTT). Glycerol gradients were ultracentrifuged eIF2␣, -eIF4A, and -eIF4E were kind gifts from Jerry Pelletier. The anti-eIF3g by using an SW41 rotor (Beckman) at 39,000 rpm for 3.5 h, sequentially fraction- antibody was a kind gift from Hiroaki Imataka, Riken Genomic Sciences Center, ated (500 ␮l) from the top, and subjected to Cerenkov scintillation counting. Wako, Japan. Anti-RPS7 and -RPL18 antibodies were used according to the manufacturer’s protocol (Abnova). Membranes were washed three times with Precipitation of Biotinylated mRNAs. Streptavidin agarose (SAA) beads (Invitro- PBST, incubated with horseradish peroxidase-conjugated secondary antibodies gen) were used to precipitate biotinylated mRNA reporters. SAA beads (100 ␮l) (Jackson ImmunoResearch) at 1:5,000 in 1% nonfat milk powder–PBST, and were washed in 1ϫ HSB buffer three times and incubated with glycerol gradient developed by ECL (Pierce). fractions (200 ␮l) or whole lysate reactions at 4°C for 60 min. Reactions were To strip Western blots of antibody complexes, membranes were incubated in centrifuged, and supernatants were removed. Precipitates were washed twice in stripping buffer (100 mM 2-mercaptoethanol; 2% SDS; and 62.5 mM Tris⅐HCl, pH 1ϫ HSB and subjected to RNA extraction and precipitation or to Western blot 6.7) at 50°C for 30 min. These membranes were washed with PBST for 2 ϫ 10 min, analysis. blocked in 5% milk–PBST, and reprobed with appropriate antibodies.

Northern Blot Analysis. The Northern blot analysis was performed by PAGE as Toeprinting Assay. Translation repression reactions containing mRNA (0.1 pmol) described in ref. 6. RNAs extracted from SAA precipitates were separated on and CXCR4 siRNA (0.6 pmol), RRL (7 ␮l), and MgOAc (2 mM) with or without 8% PAGE containing urea (7 M) and transferred to Hybond n ϩ membranes GMP-PNP (1 mM) or Hisppuristanol (50 ␮M) proceeded for 5 min at 30°C. Then, (Amersham Biosciences) for 2.5 h at 350 mA. After UV cross-linking, mem- reverse transcription (RT) mix containing dNTPs (5 mM), 1ϫ reconstitution buffer Ј Ј branes were hybridized with 5 end-labeled primers for 5S rRNA, 5 - (20 mM Tris⅐HCl, pH 7.5; 100 mM KCl; and 1 mM DTT), 5Ј end-labeled primer (0.2 Ј Ј TTAGCTTCCGAGATCA-3 ; 5.8S rRNA, 5 -GCTAGCGCTGCGTTCTTCATCGACGC- pmol, 5Ј-TTATGCAGTTGCTCTCCAGCG-3Ј), and M-MLV RT (1 ␮l, Invitrogen) was Ј Ј Ј Ј 3 ; 28S rRNA5 -AACGATCAGAGTAGTGGTATTTCACC-3 ; 18S rRNA, 5 - added to translation repression reactions. These mixtures were incubated for 15 Ј Ј CGGAACTACGACGGTATCTG-3 ; and tRNAi-Met, 5 -GGTAGCAGAGG- min at 30°C and subjected to deproteinization and ethanol precipitation. RNAs Ј ATGGTTTCGATCC-3 . Membranes were washed, visualized, and analyzed by were resolved on a 10% sequencing gel (National Diagnotics) and visualized by PhosphorImager (Molecular Dynamics). PhosphorImager analysis (Molecular Dynamics). See SI Materials and Methods for additional details. Western Blot Analysis. The SAA precipitates were resuspended in 1ϫ SDS loading buffer, boiled at 95°C for 5 min, and centrifuged. Supernatants were resolved on ACKNOWLEDGMENTS. We thank Jerry Pelletier, John Doench, Steffen Schubert, SDS–10% PAGE and transferred onto PVDF membranes (BioRad). Membranes and Tara Love for critically reading this manuscript. We thank Helen Cargill and were blocked in 5% nonfat milk powder in PBST (10 mM phosphate buffer, pH Etienne Gagnon for help with Fig. 4. A.Y. was supported by National Institutes of 7.2; 150 mM NaCl; and 0.1% Tween 20) for 60 min, washed twice with PBST, and Health Training Grant GM07266. This work was supported by a grant from the incubated with antibodies in 1% nonfat milk powder–PBST at 4°C overnight. The Claudia Adams Barr Program in Cancer Research (C.D.N.).

1. Jackson RJ, Standart N (2007) How do microRNAs regulate gene expression? Sci STKE 16. Chendrimada TP, et al. (2007) MicroRNA silencing through RISC recruitment of eIF6. 2007:re1. Nature 447:823–828. 2. Olsen PH, Ambros V (1999) The lin-4 regulatory RNA controls developmental timing in 17. Wang B, Doench JG, Novina CD (2007) Analysis of microRNA effector functions in vitro. Caenorhabditis elegans by blocking LIN-14 protein synthesis after the initiation of Methods 43:91–104. translation. Dev Biol 216:671–680. 18. Bordeleau ME, et al. (2006) Functional characterization of IRESes by an inhibitor of the 3. Seggerson K, Tang L, Moss EG (2002) Two genetic circuits repress the Caenorhab- RNA helicase eIF4A. Nat Chem Biol 2:213–220. ditis elegans heterochronic gene lin-28 after translation initiation. Dev Biol 19. Pestova TV, Lorsch JR, Hellen CUT (2007) in Translational Control in Biology and 243:215–225. Medicine, eds Mathews MB, Sonenberg N, Hershey JWB (Cold Spring Harbor Lab Press, 4. Petersen CP, Bordeleau ME, Pelletier J, Sharp PA (2006) Short RNAs repress translation Cold Spring Harbor, NY), pp 87–128. after initiation in mammalian cells. Mol Cell 21:533–542. 20. Parker JS, Barford D (2006) Argonaute: A scaffold for the function of short regulatory 5. Pillai RS, et al. (2005) Inhibition of translational initiation by Let-7 MicroRNA in human RNAs. Trends Biochem Sci 31:622–630. cells. Science 309:1573–1576. 21. Ostareck DH, Ostareck-Lederer A, Shatsky IN, Hentze MW (2001) Lipoxygenase mRNA 6. Wang B, Love TM, Call ME, Doench JG, Novina CD (2006) Recapitulation of short silencing in erythroid differentiation: The 3ЈUTR regulatory complex controls 60S RNA-directed translational gene silencing in vitro. Mol Cell 22:553–560. ribosomal subunit joining. Cell 104:281–290. 7. Nykanen A, Haley B, Zamore PD (2001) ATP requirements and small interfering RNA 22. Kiriakidou M, et al. (2007) An mRNA m7G cap binding-like motif within human Ago2 structure in the RNA interference pathway. Cell 107:309–321. represses translation. Cell 129:1141–1151. 8. Schwarz DS, Hutvagner G, Haley B, Zamore PD (2002) Evidence that siRNAs function as 23. Ceci M, et al. (2003) Release of eIF6 (p27BBP) from the 60S subunit allows 80S ribosome guides, not primers, in the Drosophila and human RNAi pathways. Mol Cell 10:537–548. assembly. Nature 426:579–584. 9. Lewis BP, Shih IH, Jones-Rhoades MW, Bartel DP, Burge CB (2003) Prediction of 24. Russell DW, Spremulli LL (1979) Purification and characterization of a ribosome dissocia- mammalian microRNA targets. Cell 115:787–798. tion factor (eukaryotic initiation factor 6) from wheat germ. J Biol Chem 254:8796–8800. 10. Stark A, Brennecke J, Russell RB, Cohen SM (2003) Identification of Drosophila mi- 25. Valenzuela DM, Chaudhuri A, Maitra U (1982) Eukaryotic ribosomal subunit anti- croRNA targets. PLoS Biol 1:E60. association activity of calf liver is contained in a single polypeptide chain protein of 11. Doench JG, Sharp PA (2004) Specificity of microRNA target selection in translational Mr ϭ 25,500 (eukaryotic initiation factor 6). J Biol Chem 257:7712–7719. repression. Genes Dev 18:504–511. 26. Raychaudhuri P, Stringer EA, Valenzuela DM, Maitra U (1984) Ribosomal subunit 12. Humphreys DT, Westman BJ, Martin DI, Preiss T (2005) MicroRNAs control translation antiassociation activity in rabbit reticulocyte lysates. Evidence for a low molecular initiation by inhibiting eukaryotic initiation factor 4E/cap and poly(A) tail function. weight ribosomal subunit antiassociation protein factor (Mr ϭ 25,000). J Biol Chem Proc Natl Acad Sci USA 102:16961–16966. 259:11930–11935. 13. Mathonnet G, et al. (2007) (2007) MicroRNA inhibition of translation initiation in vitro 27. Chakravarty I, Bagchi MK, Roy R, Banerjee AC, Gupta NK (1985) Protein synthesis in by targeting the cap-binding complex eIF4F. Science 317:1764–1767. rabbit reticulocytes. Purification and properties of an Mr 80,000 polypeptide (Co-eIF- 14. Wakiyama M, Takimoto K, Ohara O, Yokoyama S (2007) Let-7 microRNA-mediated 2A80) with Co-eIF-2A activity. J Biol Chem 260:6945–6949. mRNA deadenylation and translational repression in a mammalian cell-free system. 28. Roy AL, Chakrabarti D, Datta B, Hileman RE, Gupta NK (1988) Natural mRNA is required Genes Dev 21:1857–1862. for directing Met-tRNA(f) binding to 40S ribosomal subunits in animal cells: involve- 15. Thermann R, Hentze MW (2007) Drosophila miR2 induces pseudo-polysomes and ment of Co-eIF-2A in natural mRNA-directed initiation complex formation. Biochem- inhibits translation initiation. Nature 447:875–878. istry 27:8203–8209.

5348 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0801102105 Wang et al. Downloaded by guest on September 24, 2021