Reduced Expression of Ribosomal Proteins Relieves MicroRNA ...

Reduced Expression of Ribosomal ProteinsRelieves MicroRNA-Mediated Repression

The MIT Faculty has made this article openly available. Please share how this access benefits you. Your story matters.

Citation Janas, Maja M., Eric Wang, Tara Love, Abigail S. Harris, KristenStevenson, Karlheinz Semmelmann, Jonathan M. Shaffer, et al.“Reduced Expression of Ribosomal Proteins Relieves MicroRNA-Mediated Repression.” Molecular Cell 46, no. 2 (April 2012):171-186. © 2012 Elsevier.

As Published http://dx.doi.org/10.1016/j.molcel.2012.04.008

Publisher Elsevier B.V.

Version Final published version

Citable link http://hdl.handle.net/1721.1/83590

Terms of Use Article is made available in accordance with the publisher'spolicy and may be subject to US copyright law. Please refer to thepublisher's site for terms of use.

https://libraries.mit.edu/forms/dspace-oa-articles.html

http://hdl.handle.net/1721.1/83590

Molecular Cell

Article

Reduced Expression of Ribosomal ProteinsRelieves MicroRNA-Mediated RepressionMaja M. Janas,1,3,4 Eric Wang,5,6 Tara Love,1,3,4,8 Abigail S. Harris,7 Kristen Stevenson,2 Karlheinz Semmelmann,7

Jonathan M. Shaffer,7 Po-Hao Chen,1,3,4 John G. Doench,4 Subrahmanyam V.B.K. Yerramilli,7 Donna S. Neuberg,2

Dimitrios Iliopoulos,1,3 David E. Housman,5,6 Christopher B. Burge,5 and Carl D. Novina1,3,4,*1Department of Cancer Immunology and AIDS, Dana-Farber Cancer Institute2Department of Biostatistics and Computational Biology, Dana-Farber Cancer Institute3Department of Microbiology and Immunobiology

Harvard Medical School, Boston, MA 02115, USA4Broad Institute of Harvard and MIT, Cambridge, MA 02141, USA5Department of Biology6The Koch Institute

Massachusetts Institute of Technology, Cambridge MA 02139, USA7Qiagen, Frederick, MD 21703, USA8Present address: Correlagen Diagnostics, 307 Waverly Oaks Road, Suite 101, Waltham, MA 02452, USA

*Correspondence: [email protected]

DOI 10.1016/j.molcel.2012.04.008

SUMMARY

MicroRNAs (miRNAs) regulate physiological andpathological processes by inducing posttranscrip-tional repression of target messenger RNAs (mRNAs)via incompletely understood mechanisms. To dis-cover factors required for human miRNA activity, weperformed an RNAi screen using a reporter cell lineof miRNA-mediated repression of translation initia-tion. We report that reduced expression of ribosomalprotein genes (RPGs) dissociated miRNA complexesfrom target mRNAs, leading to increased polysomeassociation, translation, and stability of miRNA-tar-geted mRNAs relative to untargeted mRNAs. RNAsequencing of polysomes indicated substantial over-lap in sets of genes exhibiting increasedor decreasedpolysomal association after Argonaute or RPGknockdowns, suggesting similarity in affected path-ways.miRNAprofiling ofmonosomesandpolysomesdemonstrated that miRNAs cosediment with ribo-somes. RPG knockdowns decreased miRNAs inmonosomes and increased their target mRNAs inpolysomes.Our data show thatmostmiRNAs represstranslation and that the levels of RPGs modulatemiRNA-mediated repression of translation initiation.

INTRODUCTION

MicroRNA (miRNA)-mediated translational repression of

messenger RNA (mRNA) targets was first described in

C. elegans, where lin-4 and let-7 were reported to reduce pro-

tein levels without affecting mRNA stability (Olsen and Ambros,

1999; Wightman et al., 1993). Untargeted and miRNA-targeted

mRNAs demonstrated similar polysomal association as as-

sessed by density gradient centrifugation, suggesting that

miRNAs repressed translation after initiation (Olsen and Am-

bros, 1999; Seggerson et al., 2002). In mammals, polysome

profiling studies led to contradictory models of miRNA-

mediated translational repression. Although miRNA-targeted

mRNAs had reduced polysomal association compared to

unrepressed mRNAs in some studies, suggesting repression

of initiation (Bhattacharyya et al., 2006a; Pillai et al., 2005),

miRNAs and their mRNA targets had robust polysomal associ-

ation in other studies, suggesting repression after initiation

(Maroney et al., 2006; Nelson et al., 2004; Nottrott et al.,

2006; Petersen et al., 2006). Additionally, treatment with trans-

lation elongation inhibitors shifted miRNAs and miRNA-targeted

mRNAs from polysomes toward monosomes, indicating that

miRNAs target actively translating mRNAs (Maroney et al.,

2006; Nottrott et al., 2006).

Subsequent studies demonstrated that miRNAs could also

reduce mRNA stability. In C. elegans, lin-4 and let-7 decreased

target mRNA levels (Bagga et al., 2005). In both C. elegans and

mammals, miRNAs were shown to promote deadenylation,

leading to reduced target mRNA stability (Behm-Ansmant

et al., 2006; Wu and Belasco, 2005; Wu et al., 2006). In

mammals, transfected miRNAs globally reduced levels of target

mRNAs (Lim et al., 2005), and ribosome footprinting determined

that miRNA targeting resulted in a decrease in ribosome density

that was only modestly (�15%) greater than the decrease in

mRNA level, suggesting that miRNAs act primarily to destabilize

mRNAs (Guo et al., 2010). More recent ribosome footprinting

analysis at earlier time points demonstrated that repression of

translation initiation precedes deadenylation for the majority of

miR-430-targeted mRNAs in zebrafish and that mRNA decay is

likely a consequence of these earlier activities (Bazzini et al.,

2012).

Here we report that �70% of all miRNA species expressed

in HeLa cells are detected in monosomal and polysomal frac-

tions, implying that the majority of miRNA species can repress

translation. We also demonstrate that ribosomal protein genes

Molecular Cell 46, 171–186, April 27, 2012 ª2012 Elsevier Inc. 171

mailto:[email protected]

http://dx.doi.org/10.1016/j.molcel.2012.04.008

(RPGs) preferentially regulate miRNA-mediated repression of

translation initiation. Specifically, RPG depletion led to

decreased monosomal association of a subset of miRNAs and

increased polysomal association of mRNA targets of those

miRNAs, indicating relief of translation initiation block. This

study establishes a framework for understanding ribosomes as

global regulators of miRNA-mediated repression of translation

initiation.

RESULTS

A Reporter of miRNA-Mediated Repression ofTranslation InitiationWe engineered a clonal HeLa cell line we called D8 that stably

expresses two reporter genes: Firefly Luciferase (FL) with six

binding sites for endogenous miR-21 in its 30 untranslated region

(30 UTR), and Renilla Luciferase (RL) lacking miRNA binding sites

(Figure 1A). Inhibition of miR-21 with a miR-21-specific antago-

mir resulted in up to 7-fold increase in FL relative to RL as

assessed by the dual luciferase assay (Figure 1B). The miR-21-

specific antagomir increased FL protein levels more than FL

mRNA levels, even after 144 hr of antagomir treatment (Fig-

ure S1A available online). These results show that in D8, FL

was strongly repressed by miR-21 and suggest that the mecha-

nism involved was repression of translation.

Additionally, knockdown of miRNA effector genes (Argo-

nautes [Agos] or eIF6) yielded a 2- to 10-fold increase in FL

protein levels accompanied by a proportional yet smaller

increase in FL mRNA levels (Figure 1C). Knockdown of miRNA

biogenesis (Drosha and DGCR8) or mRNA turnover (DCP2 and

PARN) genes also increased FL levels in a small interfering

RNA (siRNA) dose-dependent manner (Figure S1B). These

data show that D8 has a large dynamic range and high sensitivity

to reduced levels of miRNA pathway proteins and thus is

an appropriate tool to discover new factors and regulators of

miRNA-mediated repression by gain-of-signal RNA interference

(RNAi) screening.

To determine which step of FL translation is inhibited by

miR-21 in D8, we performed polysome profiling of control,

Ago1- or eIF6-depleted cells (Figures 1D and 1E). In control

cells, the majority of miR-21-targeted FL mRNA was in mono-

somes, and the majority of untargeted RL mRNA was in poly-

somes. Knockdown of Ago1 or eIF6 led to a striking shift of

FL mRNA from monosomes to polysomes, while RL mRNA

exhibited only a modest shift toward heavier polysome frac-

tions but remained polysomal in all conditions. Consistent

with RT-qPCR data in Figure 1C and with previously reported

miRNA effects on mRNA stability (Bagga et al., 2005; Guo

et al., 2010; Behm-Ansmant et al., 2006; Wu and Belasco,

2005; Wu et al., 2006), knockdown of Ago1 or eIF6 stabilized

miR-21-targeted FL mRNA, leading to a higher combined signal

for FL but not RL mRNA across gradients relative to the Scr

control. Consistent with its role in 60S ribosomal subunit

biogenesis (Basu et al., 2001), eIF6 knockdown strongly

reduced the levels of 60S ribosomal subunits without affecting

the levels of 40S subunits as assessed by A254 absorbance

traces (Figure 1D). These data demonstrate that in D8, miR-

21 repressed initiation of FL translation.

An RNAi Screen Identifies RPGs as Regulatorsof the miRNA PathwayRNAi screens for RNAi pathway genes have been used effec-

tively in simpler eukaryotes (Dorner et al., 2006; Eulalio et al.,

2007; Kim et al., 2005; Parry et al., 2007; Saleh et al., 2006; Ulvila

et al., 2006), leading to identification of many effectors of small

RNA function, such as RNA binding proteins, cytoskeletal

proteins, and one RPG in C. elegans (Parry et al., 2007), as

well as several RPGs and proteasomal components in

D. melanogaster (Zhou et al., 2008). However, similar screens

have not yet been reported in mammals. Therefore, we used

D8 to perform an RNAi screen for effectors and regulators of

the human miRNA pathway (Figure 2A). In all, �7,000 genes

were targeted with four individual siRNAs per gene from the

Druggable Genome Library of siRNAs (QIAGEN), and effects of

knockdowns were assessed with the dual luciferase assay.

Using stringent criteria (at least two siRNAs per gene causing

derepression of FL R eIF6 knockdown), 314 genes (4.3%)

scored as hits (Table S1, part A). RPGs were enriched more

than 13-fold in this screen: seven out of 12 RPGs (58%) scored

as hits. Closer inspection of the screening data showed that all

12 RPG knockdowns derepressed FL to some extent. Gene

ontology analysis revealed that the ribosomal pathway was

one of the most significantly enriched pathways in the screen

(Figure 2B and Table S1, part B). Thus, we hypothesized that

as a class RPGs might regulate miRNA repression.

To determine whether depletion of every RPG inhibits

miRNA-mediated repression, we targeted all 79 RPGs with

four individual siRNAs per gene in D8. Knockdown of virtually

all 40S (Figure 2C) and 60S (Figure 2D) RPGs with at least

one siRNA led to derepression of FL that was greater than

derepression after eIF6 knockdown. After three days (the point

of strongest FL derepression), most RPG knockdowns reduced

cell numbers by 20%–50%. The Y chromosome-expressed

S4Y1 and S4Y2 provide natural negative controls in female

cells and did not score in our HeLa cell validation screen (Fig-

ure 2C). The consistent FL derepression after depletion of

almost every 40S and 60S RPG suggests that altered ribo-

somal integrity or altered ribosomal subunit stoichiometry

and not extraribosomal functions of RPGs led to decreased

miRNA activity. Supporting this conclusion, knockdown of

factors involved in either 40S (Bms1 and Tsr1) or 60S (Bop1

and Nip7) subunit biogenesis also led to derepression of

miRNA-targeted mRNAs (Figure S2A), and double knockdown

of 40S and 60S RPGs led to greater derepression of miRNA-

targeted mRNAs compared to single RPG knockdowns

(Figure S2B).

To validate the screening results, we chose five representative

40S (S5, S11, S12, S15, and S18) and 60S (L5, L11, L12, L35A,

and LP2) RPGs for detailed biochemical analysis. Knockdown

of each RPG in D8 (Figure S2C) derepressed FL mostly at the

protein levels, although proportional increases in FL mRNA

levels were also observed (Figure S2D), consistent with knock-

downs of known miRNA pathway genes (Figure 1C). In the lucif-

erase screens, which were not normalized to total protein, deple-

tion of RPGs increased FL levels and decreased RL levels

(Figure S2E). These trends were affected by reduced cell

numbers with, on average, half the number of cells after 72 hr

Molecular Cell

Ribosomal Proteins Modulate miRNA Activity

172 Molecular Cell 46, 171–186, April 27, 2012 ª2012 Elsevier Inc.

Firefly

Renilla

miR-21A

B

miR-21

tRNA

0 5 20 100

0

2

4

6

8

10

12

Fol

d D

erep

ress

ion

ProteinmRNA

C

miR-21 antagomir [nM]

Fol

d D

erep

ress

ion

Ago1Actin

Ago2Actin

eIF6Actin

Scr

Ago

1

Scr

Ago

2

Scr

eIF

6

ProteinmRNA

D

0 5 20 100

Scr Ago1 Ago2 eIF6

Scr

Ago1

eIF6

FL

RL

28S

18S

A254

FL

RL

28S

18S

A254

FL

RL

28S

18S

A254

0

1

2

3

4

5

6

7

8

CMV

CMV

0102030405060708090

1 2 3 4 5 6 7 8 9 10 11 12Fraction number

% F

L m

RN

A

ScrAgo1eIF6

E

0

10

20

30

40

50

60

70

1 2 3 4 5 6 7 8 9 10 11 12Fraction number

% R

L m

RN

A

ScrAgo1eIF6

40S 60S 80S Polysomes



Figure 1. D8 Is a Highly Sensitive Reporter of miR-21-Mediated Repression of Translation Initiation

(A) Schematic representation of CMV-driven luciferase reporters stably expressed in D8. Firefly Luciferase (FL) possesses six imperfect miR-21 binding sites in its

30 UTR, while Renilla Luciferase (RL) lacks miRNA binding sites.

(B) D8 reports on miR-21-mediated repression. A miR-21-specific antagomir was transfected into D8 at indicated concentrations, and FL and RL protein and

mRNA levels were assessed after 24 hr by the dual luciferase assay and RT-qPCR, respectively. Fold derepression was calculated as (FLtest/RLtest)/(FLcontrol/

RLcontrol). The antagomir reduced miR-21 levels in a dose-dependent fashion as assessed by northern blotting.

(C) D8 reports on miRNA effector function. D8 was transfected with Scrambled (Scr), Ago1-, Ago2-, or eIF6-specific siRNAs, and FL and RL protein and mRNA

levels were assessed after 72 hr. Knockdowns of Ago1, Ago2, and eIF6 were confirmed by western blotting.

(D) Polysomeprofiling demonstrates that knockdown ofAgo1 or eIF6 shifts FLmRNA frommonosomes to polysomes. Lysates fromD8 transfected either with Scr

or with Ago1- or eIF6-specific siRNAs were fractionated while absorbance was monitored at 254 nm (A254). FL and RL mRNAs were assessed by northern

blotting, and 28S and 18S rRNAs were assessed by ethidium bromide staining.

(E) Quantification of northern blots with ImageQuant. FL and RLmRNA detected in each fraction is represented as the percent of the total mRNA detected in all

fractions across the gradient.

Bar graphs show the mean ± SD from three independent experiments. See also Figure S1.

Molecular Cell



A

B

C

Scr

eIF

6L

3L

4L

5L

6L

7L

7A

L8

L9

L1

0L

11

L1

2L

13

L1

3A

L1

4L

15

L1

7L

18

L1

8A

L1

9L

21

L2

2L

23

AL

24

L2

6L

27

L2

7A

L2

8L

29

L3

1L

32

L3

4L

35

L3

5A

L3

6L

36

AL

37

L3

7A

L3

8L

39

L4

0L

41

LP

0L

P1

LP

2

Gene ontology p-value Corrected p-value

GPCR protein signaling pathway 6.20E-09 4.80E-06

Protein metabolic processing 8.70E-08 5.20E-07

Inflammatory signaling pathway 4.70E-06 6.50E-05

Ribosomal pathway 2.80E-04 1.70E-03

Proteolysis 1.10E-04 7.50E-03

PDGF signaling pathway 4.50E-03 2.50E-02

TRANSFECT

Fo

ldD

ere

pre

ssio

n

Scr

eIF

6

SA

S2

S3

S3

A

S4

X

S4

Y1

S4

Y2

S5

S6

S7

S8

S9

S1

0

S11

S1

2

S1

3

S1

4

S1

5

S1

5A

S1

6

S1

7

S1

8

S1

9

S2

0

S2

1

S2

3

S2

4

S2

5

S2

6

S2

7

S2

7A

S2

7L

S2

8

S2

9

S3

0

D

Fo

ldD

ere

pre

ssio

nsiRNA library D8 cell line

72 HRS

+FL substrate

+RL substrate

LUCIFERASE

ASSAY

eIF6

Scr

0

1

2

3

4

5

6

7

8

9

10

11

0

1

2

3

4

5

6

7

8

9

10

11

eIF6

Scr

E F

0

1

2

3

4

5

6

Fo

ldD

ere

pre

ssio

n

HMGA2 mRNA

Scr

Ag

o1

S11

S1

2

S1

5

S1

8

L5

L11

L1

2

L1

5

L3

5A

PTEN

Actin

Scr

Ag

o1

eIF

6

S5

S11

S1

2

S1

5

S1

8

L5

Scr

S5

S11

S1

2

S1

5

S1

8

L11

L1

2

L3

5A

LP

2

eIF

6

Ag

o1

Ag

o2

Scr

Scr

S5

S1

2

S1

5

L1

2

L3

5A

Ag

o1

eIF

6

RAS

GAPDH

IMP-1

PDCD4

GAPDH

Figure 2. RNAi Screening for Effectors and Regulators of the miRNA Pathway Identifies RPGs

(A) Schematic representation of the RNAi screen. D8 was transfected with siRNAs against�7,000 genes, and FL and RL protein levels were assessed by the dual

luciferase assay after 72 hr.

(B) Ribosomal pathway was significantly enriched among hits from the screen as assessed by gene ontology analysis. The p values were calculated with Fisher

exact test analysis and corrected p values were calculated with the Bonferroni method.

Molecular Cell



of RPG knockdown, presumably because RPG depletion

reduced cellular proliferation rates as previously reported (Fly-

gare et al., 2005; Kirn-Safran et al., 2007;Miller et al., 2003; Oliver

et al., 2004; Oristian et al., 2009; Pani�c et al., 2007; Volarevic

et al., 2000). When normalized to total protein by western blot-

ting, we observed increased FL and unchanged RL levels (Fig-

ure S2F). To further validate these findings and to control for

possible siRNA off-target effects, we expressed five siRNA-

resistant isoforms of RPGs (S5, S12, S15, L12, and LP2) in D8

cells knocking down these RPGs. In all cases, transfection of

constructs expressing RPGs with silent mutations in regions

complementary to siRNAs partially rescued the FL derepression

(Figure S2G), demonstrating that RPG depletion directly led to

miRNA-targeted mRNA derepression.

We next assessed the effect of RPG depletion on various

reporter and endogenous miRNA-targeted mRNAs. FL and RL

reporters containing six imperfectly complementary binding

sites for the artificial CXCR4 miRNA were derepressed in RPG

knockdowns only in the presence of the CXCR4 miRNA (Figures

S2H and S2I). However, an FL reporter containing one perfectly

complementary site for the CXCR4 miRNA was only modestly

affected by RPG knockdowns (Figure S2J), indicating that

RPGs preferentially regulate miRNA-mediated translational

repression and not siRNA-mediated mRNA cleavage. The RL

reporter containing HMGA2 30 UTR with seven seed matches

to the let-7 miRNA (Mayr and Bartel, 2009) was derepressed

upon RPG knockdowns (Figure S2K). Consistent with previous

studies showing that repression of HMGA2 correlates with its

mRNA level (Lee and Dutta, 2007), endogenous HMGA2

mRNA levels increased afterRPG knockdowns (Figure 2E). Simi-

larly, an RL reporter containing the KRAS 30 UTRwith seven let-7

sites (Johnson et al., 2005) was derepressed in HeLa cells upon

RPG knockdowns (Figure S2L), as was endogenous RAS protein

(Figure 2F). Other validated endogenous miRNA-targeted

mRNAs, such as miR-21-targeted PTEN (Lewis et al., 2003;

Talotta et al., 2009), let-7-targeted IMP-1 (Boyerinas et al.,

2008), and miR-21-targeted PDCD4 (Asangani et al., 2008)

were all derepressed upon RPG knockdowns (Figure 2F).

Although the degree and pattern of derepression varied between

different miRNA-targeted mRNAs, as a class RPGs regulated

miRNA repression.

RPG Knockdowns Relieve Repression of TranslationInitiationTo explore the mechanism of derepression of miRNA-targeted

mRNAs in RPG knockdown cells, we first examined miRNA

biogenesis. We did not detect any defects in the biogenesis of

any miRNAs tested (Figures S3A and S3B). Additionally, RPG

knockdowns did not decrease the levels of Ago1 or Ago2

proteins (Figure S3C). Polysome profiling of D8 cells knocking

down S15 (Figures 3A and 3C) or L12 (Figures 3B and 3C)

demonstrated a strong shift of FL mRNA but not RL mRNA

from monosomes to polysomes, indicating that RPG knock-

downs relieve miRNA-mediated repression of translation initia-

tion. Consistent with knockdowns of miRNA pathway genes

(Figure 1D), RPG knockdowns increased the combined signal

for FL but not RL mRNA across gradients relative to the Scr

control, indicating preferential stabilization of FL mRNA. Heavy

complexes formed on miRNA-targeted mRNAs after RPG

depletion were actively translating polysomes because FL

mRNA shifted to lighter fractions after puromycin treatment

which dissociates only ribosomes actively translocating on

mRNAs (Figures S3D and S3E). Thus, reduced levels of RPGs

increased loading of active ribosomes onto miRNA-targeted FL

mRNA but not untargeted RL mRNA.

Importantly, the effects of RPG knockdowns on ribosomal

subunit levels were similar among all tested 40S and 60S

RPGs. Knockdown of 40S RPGs consistently decreased the

levels of free 40S subunits and 18S ribosomal RNA (rRNA) and

increased the levels of free 60S subunits (Figures 3A, S4A, and

S4B). On the other hand, knockdown of 60S RPGs or eIF6

consistently decreased the levels of free 60S subunits and 28S

rRNA (Figures 3B, S4A, and S4B). These data are consistent

with established contributions of 40S and 60S RPGs to biogen-

esis and/or stability of ribosomal subunits (Lempiainen and

Shore, 2009) and further indicate that defects in ribosome

biogenesis rather than extraribosomal functions of RPGs led to

reduced miRNA-mediated repression.

To resolve the 40S, 60S, and 80S peaks, polysome profiling

was performed with lower-density gradients, higher-speed

spins, and longer times, which resulted in reduced resolution

of the highest-density complexes (Figures 3A and 3B). These

conditions enabled detection of reduced 40S and 60S peaks

and thus verified efficient and functional RPG knockdowns. To

resolve the highest density polysomes and prevent pelletting

of the heaviest complexes, we performed polysome profiling

using higher-density gradients, lower-speed spins, and shorter

times. Under these conditions, RPG knockdowns led to a con-

sistent shift of FL but not RLmRNAs to polysomes (Figure S4C).

Together, these data show that RPG knockdowns inhibited

miRNA-mediated repression of translation initiation and that

miRNA-targeted mRNA derepression could be a result of

perturbed stoichiometry between 40S and 60S ribosomal

subunits, either directly through physical interactions with

miRNP complexes or indirectly through signaling pathways

that sense ribosome biogenesis.

(C and D) Subgenomic RNAi screen in D8 identifies almost all RPGs as hits in themiRNApathway. Small and large ribosomal subunitRPGswere knocked down in

D8 and the dual luciferase assay was performed after 72 hr. Each point represents an individual siRNA. RPGs included in the original screen are highlighted.

Dotted horizontal gray lines represent derepression of FL in the negative control (Scr) and the positive control (eIF6) transfections.

(E) Knockdown of RPGs derepresses endogenous HMGA2 at the mRNA level. HeLa was transfected with siRNAs against indicated genes, and after 72 hr,

endogenousHMGA2mRNA levels were assessed by RT-qPCR and normalized toGAPDHmRNA levels. Bar graphs show themean ± SD from three independent

experiments.

(F) Knockdown of RPGs derepresses endogenous miRNA targets at the protein level. HeLa was transfected with siRNAs against the indicated genes, and RAS,

PTEN, IMP-1, PDCD4, and GAPDH protein levels were assessed by western blotting after 72 hr.

See also Figure S2.

Molecular Cell



A

BScr L12

Scr S15

FL

RL

A254


FL

RL

28S

18S

A254

C

0

10

20

30

40

50

60

70

80

1 2 3 4 5 6 7 8 9 10Fraction number

% F

L m

RN

A

S15ScrL12

Scr

0

10

20

30

40

50

60

70

% R

L m

RN

A

Scr

S1

5

L1

2

Scr

S1

5

L1

2

SCR CXCR4

Input

35S-FLAG IP

E

S

Scr

Scr+

CH

X

S5

S11

S1

2

S1

5

S1

8

L5

L11

L1

2

L3

5A

LP

2

eIF

6

Ag

o1

Co

om

assie

35

FLAG-FL6X

35S-FLAG IP/Input 1.0 1.3 1.5 0.3 0.8 0.9

D

28S

18S


40S 60S 80S Polysomes 40S 60S 80S Polysomes

35S/Coomassie 1.0 0.2 1.1 1.1 1.2 1.2 1.0 0.9 0.7 0.9 0.9 0.9 1.1 1.1

Figure 3. RPG Knockdowns Selectively Increase Translation of miRNA-Targeted mRNAs

(A and B) RPG knockdowns relieve miR-21-mediated repression of translation initiation. Polysome profiles demonstrate that knockdown of S15 (A) and L12 (B)

shift miR-21-targeted FL mRNA but not untargeted RL mRNA from monosomes to polysomes.

(C) Quantification of northern blots with ImageQuant. FL and RLmRNA detected in each fraction is represented as the percent of the total mRNA detected in all

fractions across the gradient.

(D) Translation rates of CXCR4 miRNA-targeted FL6X are increased after RPG knockdowns only in the presence of the targeting CXCR4 miRNA. HeLa was

transfected with Scr or with S15- or L12-specific siRNAs, and after 48 hr FLAG-tagged FL6X was transfected either with nontargeting Scr (control) or targeting

CXCR4 (test) miRNAs. After 24 hr, cells were 35S-pulse-labeled and FL was IP-ed with anti-FLAG agarose. Inputs and precipitates were resolved by PAGE, and

the amount of 35S incorporation into FL6X was quantified with ImageQuant and normalized to the Scr transfection.

(E) Overall translation rates are not reduced by depletion of RPGs as assessed by 35S-pulse-metabolic-labeling. HeLa transfectedwith indicated siRNAswas 35S-

pulse-labeled, resolved by PAGE, and either stained with Coomassie as a loading control or visualized with a phosphorimager. As a positive control, pulse

labeling was performed in the presence of a known translation inhibitor cycloheximide (CHX).

See also Figures S3 and S4.

Molecular Cell



To further validate that RPG knockdowns increased transla-

tion rates of miRNA-targeted but not untargeted mRNAs, we

performed 35S pulse labeling followed by immunoprecipitation

(IP) of FLAG-tagged FL protein targeted by the CXCR4 miRNA.

AS

cr

S1

5

Scr

S1

5

Scr

S1

5

Scr

S1

5

FLAG-Ago1 FLAG-Ago2

Input IP

RN

A

Pro

tein

FLAG

Actin

FL

miR-21

let-7a

tRNA

B

Ago1

Ago2

Scr

S1

5

L1

2

eIF

6

Input

Scr

S1

5

L1

2

eIF

6

Ribosomal Pellet

S6

L7a

Actin

FLAG-Ago1 FLAG-Ago2

0

0.5

1

1.5

2

2.5

3

3.5

Fo

ldC

han

ge

FLmiR-21let-7a

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0

1

2

3

Fo

ldC

han

ge

Ago1Ago2

0

1

2

3

4

*

*

**

*

** ** ** **

**

Figure 4. RPG Knockdown Reduces Ago Associa-

tion with miRNA-Targeted mRNAs and Ribosomes

(A) Knockdown of S15 decreases association of Ago1

and Ago2 with target mRNAs but not miRNAs. D8 was

transfectedwith Scr or withS15-specific siRNAs, and after

24 hr, with vectors expressing FLAG-tagged Ago proteins.

After 48 hr, inputs and anti-FLAG IPs were blotted for

proteins (FLAG-Ago and Actin) and RNAs (FL, miR-21, let-

7a, and transfer RNA) and quantified with ImageQuant.

(B) Knockdown of S15 or L12 dissociates endogenous

Agos from ribosomes. HeLa was transfected with indi-

cated siRNAs and ribosomes were pelleted and western

blotted for endogenous Ago1 and Ago2 proteins. The

efficiency and specificity of ribosomal recovery was

assessed by western blotting for Actin, S6, and L7a. Ago1

and Ago2 in inputs and ribosomal pellets was quantified

with ImageQuant.

Bar graphs show the mean ± SD from three independent

experiments. *p < 0.05, **p < 0.01. See also Figure S4.

RPG knockdowns resulted in up to 3-fold

increased incorporation of the 35S label into

FL protein in the presence of the targeting

CXCR4 miRNA, while RPG knockdowns did

not significantly affect 35S incorporation into FL

protein in the absence of the CXCR4 miRNA

(Figure 3D), demonstrating increased translation

rates of amiRNA-targeted but not an untargeted

mRNA. Indeed, gross effects on overall transla-

tion rates were not observed after RPG knock-

downs (Lai et al., 2009; Volarevic et al., 2000),

as indicated by unchanged polysome peak

heights in A254 absorbance traces (Figures 3A,

3B, and S4A) and pulse 35S labeling of bulk

proteins (Figure 3E). Therefore, depletion of indi-

vidual RPGs specifically increased translation

rates of miRNA-targeted mRNAs but not untar-

geted mRNAs.

RPG Knockdowns Reduce AgoAssociation with miRNA-TargetedmRNAsTo address themechanism ofRPG knockdowns

leading to derepression of miRNA-targeted

mRNAs, we assessed Ago1 and Ago2 associa-

tion with miRNAs and target mRNAs before

and after RPG depletion. RPG knockdown

followed by IP of FLAG-tagged Ago1 or Ago2

did not affect Ago association with miR-21 or

let-7a miRNAs but significantly decreased Ago

association with miR-21-targeted FL mRNA

relative to control cells (Figure 4A). These data

indicate that Ago association with miRNAs

remains stable and that intact miRNPs disso-

ciate from miRNA-targeted mRNAs. As observed for FL mRNA

in D8, RPG knockdowns reduced association of Ago1 and

Ago2 with endogenous miRNA-targeted mRNAs (KRAS and

PTEN) while increasing their stability (Figure S4D). Supporting

Molecular Cell



Polysome / Input RatioNo. of genes up-regulated > 2-fold

Polysome / Input RatioNo. of genes down-regulated > 2-fold

1.51.20.90.60.30.0-0.3-0.6-0.9-1.2-1.5

A

B

C

D

No. of poly

som

e-

up

-re

gu

late

d g

en

es

eIF6

L11

L12

S15

Ago1

Ago2

Ago3

Ago4

Ag

o4

Ag

o3

Ag

o2

Ag

o1

S1

5

L1

2

L11

eIF

6

Ag

o4

Ag

o3

Ag

o2

Ag

o1

S1

5

L1

2

L11

eIF

6

vs. S

cr1

vs. Scr2

log

2 (

Ob

se

rve

d/E

xp

ecte

d)

10

8

6

4

2

0

log

2 (

Ob

se

rve

d/E

xp

ecte

d)

2 3 4 5 6 7 8No. of knockdowns

1800

1600

1400

1200

1000

800

600

400

200

0

Observed

Expected

Observed/Expected

10

8

6

4

2

0

log

2 (

Ob

se

rve

d/E

xp

ecte

d)

No. of knockdowns

1800

1600

1400

1200

1000

800

600

400

200

0

Observed

Expected

Observed/Expected

Fraction of polysome up-regulated genes with

HeLa miRNA sites (conserved 8mers)

Observed

Expected

No

. o

f sh

uffle

d t

ria

ls 60

50

40

30

20

10

0

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

Fraction of polysome down-regulated genes with

HeLa miRNA sites (conserved 8mers)

Observed

Expected

No

. o

f sh

uffle

d t

ria

ls

90

80

70

60

50

40

30

20

10

0

Observed

Expected

No. of knockdowns

0.35

0.30

0.25

0.20

0.15

0.10

No. of knockdowns

0.35

0.30

0.25

0.20

0.15

0.10

Observed

Expected

Fra

ction o

f poly

som

e

do

wn

-re

gu

late

d g

en

es w

ith

He

La

miR

NA

site

s

2 3 4 5 6 7 8

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

2 3 4 5 6 7 8 2 3 4 5 6 7 8

No. of poly

som

e-

do

wn

-re

gu

late

d g

en

es

Fra

ction o

f poly

som

e

up

-re

gu

late

d g

en

es w

ith

He

La

miR

NA

site

s

Figure 5. Reduced RPG Expression Upregulates Polysomal Association of a Consistent Set of Genes that Is Enriched in miRNA-Targeted

mRNAs

(A) Genes whose polysome/input ratios change after knockdown of Ago, RPGs, or eIF6 are shared. The numbers of genes whose polysome/input ratios are

increased or decreased more than 2-fold in pairs of knockdowns were counted and are displayed in matrix format. The numbers of genes expected to be

regulated in the same manner by chance are shown in parentheses. The observed/expected ratio for each comparison is displayed in color.

Molecular Cell



the model of miRNP dissociation from miRNA-targeted mRNAs,

RPG depletions reduced Ago1 and Ago2 association with ribo-

somes (Figure 4B). Together, these data demonstrate that RPG

knockdown dissociates miRNPs from miRNA-targeted mRNAs

and does not affect miRNP assembly or stability.

RPG, eIF6, and Ago Knockdowns Alter PolysomeAssociation of Common mRNAsTo globally assess effects on translation of endogenous mRNAs,

we conducted RNA sequencing (RNA-Seq) analysis of polyso-

mal fractions and total RNA (‘‘input’’ to the gradient) before

and after knockdowns of RPGs, Agos, or eIF6. A remarkably

large number of common genes demonstrated a 2-fold or

greater increase in polysome association (polysome/input ratio)

after knockdown of RPGs, Agos, or eIF6 (Figure 5A). Many

common genes also demonstrated a 2-fold or greater decrease

in polysome association for many pairs of knockdowns, particu-

larly between knockdowns of differentAgos. These observations

suggested the surprising conclusion that large and small subunit

RPGs, Agos, and eIF6 have related effects on translational

regulation globally, perhaps through common regulatory path-

ways. This idea was further supported by analysis of mRNAs

impacted in three or more of these knockdowns. Instead of

decreasing rapidly to zero, as expected if the knockdowns

affected independent sets of genes, the numbers of mRNAs

with at least 2-fold increased polysome association remained

in the several hundred range, even as the number of compared

knockdowns increased from three to eight, with 479 mRNAs

exhibiting an increase in all eight knockdowns (Figure 5B and

Table S2, part A). A smaller but still highly significant set of 57

mRNAs with consistently decreased polysome association in

all knockdowns was also observed (Figure 5B and Table S2,

part B). Thus RPGs, Agos, and eIF6 inhibited polysome associ-

ation of a large common cohort of mRNAs, and promoted poly-

some association of a smaller common cohort of mRNAs.

miRNA Target Sites Are Enriched in Polysome-ShiftedmRNAsTo explore potential connections to the miRNA pathway, we

analyzed the fraction of mRNAs containing conserved 8-mer

seed matches to HeLa-expressed miRNAs, a relatively stringent

set of putative targets that likely excludesmany authentic targets

with weaker or less stringently conserved seed matches (Fried-

man et al., 2009). Significant enrichment of conserved target

sites relative to controls was observed in the set of mRNAs

with consistently increased polysomal association in all knock-

downs (Figure 5C). The extent of target site enrichment

increased as the number of intersected knockdowns increased

(Figure 5D). The observed enrichment for miRNA target sites in

this common set of polysome-increased mRNAs suggested

that miRNAs are involved in repression of this gene set via

a mechanism involving translation, and that the perturbations

introduced by RPG, Ago, and eIF6 knockdowns alter ribosome

loading onto a cohort of mRNAs that is strongly enriched

for miRNA-targeted mRNAs. Mechanisms involving miRNA-

directed changes in stability of polysome-associated versus

non-polysome-associatedmRNAs are also possible. In contrast,

the set of genes with consistently decreased polysomal associ-

ation in the knockdowns was not enriched for conserved miRNA

target sites (Figures 5C and 5D), suggesting that this set of

mRNAs is less often directly impacted by miRNAs. Although

RPG, Ago, and eIF6 knockdowns also altered mRNA levels of

a common cohort of genes, these sets of genes were not en-

riched for conserved 8-mer seed matches to HeLa-expressed

miRNAs (Figure S5). Together, these data show that RPGs,

Agos, and eIF6 contributed to reducing polysome association

of a common subset of mRNAs enriched for miRNA-targeted

mRNAs. Because miRNA-targeted mRNAs were not enriched

in a common subset of mRNAs exhibiting increased mRNA

levels after knockdowns, we observed translational derepres-

sion and not stabilization of miRNA-targeted mRNAs.

miRNAs Predominantly Cosediment with PolysomesThe preceding studies demonstrated that miRNA-targeted

mRNAs associate with monosomes and that RPG knockdowns

increase association of miRNA-targetedmRNAswith polysomes

without changing miRNA levels or miRNA association with Ago

proteins. To gain additional insights into RPG regulation of

miRNA-mediated repression, we performed global miRNA

profiling of total, monosomal, and polysomal RNA after sucrose

gradient fractionation of HeLa cells before and after RPG

knockdown. Out of 669 unique miRNA species expressed in

HeLa cells, 462 miRNAs (70%) were reliably detected in

monosomal and polysomal fractions at steady state (Table S3,

part A). The majority of miRNAs were equally abundant in

monosomal and polysomal fractions. Specifically, out of 462

miRNAs detected in monosomes and polysomes, 265 miRNAs

(57%) had no statistically significant enrichment in either frac-

tion. Thus, the majority of miRNAs are equally distributed across

monosomal and polysomal fractions and may contribute to

(B) A common set of mRNAs shifts to and away from polysomes in Ago, RPG, and eIF6 knockdowns. The numbers of genes whose polysome/input ratios are

increased or decreased more than 2-fold in two-way, three-way, . eight-way comparisons of knockdowns are displayed in red or blue, respectively. The

numbers of genes expected to be regulated in the same manner by chance are shown in black. Ninety percent confidence intervals are shown in pale red, blue,

and black.

(C) The upregulated genes have a significantly increased proportion of HeLa-expressed miRNA binding sites (p < 0.003). The fraction of genes with conserved 8-

mermiRNA sites was obtained for genes whose polysome expression is upregulated (red) or downregulated (blue) in all eight knockdowns (Ago1–Ago4, L11, L12,

S15, and eIF6). For assessment of significance, the same fraction was determined for 1,000 sets of control genes that matched the original up- or downregulated

sets in 7-mer conservation and UTR length (black histogram).

(D) The proportion of upregulated genes with HeLa-expressed miRNA-targeted mRNAs increases as the number of knockdowns used to define the set of

coregulated genes increases. The fraction of genes with HeLa-expressed conserved 8-mer miRNA sites was obtained for those genes whose polysome

expression is upregulated (red) or downregulated (blue) in two, three, . eight knockdowns. Control genes (black) were selected to match the up- or down-

regulated genes.

See also Figure S5.

Molecular Cell



multiple modes of translational repression, suggesting mecha-

nistic variation for molecules of the same miRNA species.

Interestingly, 197 of these miRNAs (43%) were differentially

present in monosomal versus polysomal fractions (Figure 6A

and Table S3, part B). Strikingly, 195 of those 197 miRNAs

(99%) were enriched in polysomal fractions, while only two

miRNAs (miR-21 and miR-126*) were enriched in monosomal

fractions (Figure 6B and Table S3, part B). These data demon-

strate that (1) the majority of HeLa miRNA species cofractionate

with ribosomes and thus can repress translation; (2) miR-21

predominantly represses translation initiation, consistent with

biochemical characterization of the D8 cell line where miR-21-

repressed FL mRNA accumulated in monosomes (Figures 1D,

3A, and 3B); (3) individual miRNA species can be found preferen-

tially in polysomal relative to monosomal fractions, indicating

that miRNA or target mRNA identity can affect the mode of

translational repression; and (4) most miRNAs preferentially co-

sediment with polysomal fractions. Because puromycin treat-

ment did not significantly affect sedimentation of the 195miRNAs

enriched in polysomal fractions (Figure S6A), thesemiRNAswere

either associated with heavy processing body (P body) aggre-

gates that cosedimented with polysomes or with ribosomes

stalled on target mRNAs, suggesting a postinitiation block.

RPG Knockdowns Decrease miRNA Associationwith MonosomesTo assess RPG regulation of miRNA-mediated repression, we

performed global miRNA profiling of total, monosomal, and

polysomal RNA after S15 knockdown in HeLa cells (Figure 6C

and Table S3, part C). S15 knockdown did not globally alter

A

hsa−miR−21hsa−miR−7hsa−miR−92ahsa−miR−1280hsa−miR−24hsa−miR−1260hsa−miR−30chsa−miR−125a−5phsa−miR−224hsa−miR−25hsa−miR−26ahsa−let−7ghsa−miR−191hsa−miR−222hsa−miR−20ahsa−miR−16hsa−miR−31hsa−miR−17 // hsahsa−let−7dhsa−miR−221hsa−miR−365hsa−miR−92bhsa−let−7ihsa−miR−27ahsa−miR−454hsa−miR−197hsa−let−7bhsa−miR−148bhsa−miR−26bhsa−miR−27bhsa−miR−151−3phsa−let−7chsa−miR−182hsa−miR−423−5phsa−miR−19bhsa−miR−29chsa−miR−29ahsa−miR−103ahsa−miR−423−3phsa−miR−18ahsa−miR−20bhsa−miR−106bhsa−miR−19ahsa−miR−125bhsa−miR−30ahsa−miR−30ehsa−miR−195hsa−miR−93hsa−miR−99bhsa−miR−100hsa−miR−186hsa−miR−193bhsa−miR−425hsa−miR−34ahsa−miR−484hsa−miR−4291hsa−miR−4286hsa−miR−505hsa−miR−574−3phsa−miR−128hsa−miR−378hsa−miR−1307hsa−miR−342−3phsa−miR−424hsa−miR−30bhsa−miR−30dhsa−miR−99ahsa−miR−200bhsa−miR−876−3phsa−miR−16−1*hsa−miR−335hsa−miR−1255bhsa−miR−19b−1*hsa−miR−132*hsa−miR−1909*hsa−miR−589hsa−miR−1227hsa−miR−3916hsa−miR−296−3phsa−miR−196b*hsa−miR−1249hsa−miR−3909hsa−miR−183*hsa−miR−7−1*hsa−miR−425*hsa−miR−29a*hsa−miR−3200−5phsa−miR−4323hsa−miR−550ahsa−miR−27b*hsa−miR−33a*hsa−miR−450ahsa−miR−3618hsa−miR−1260bhsa−miR−22hsa−miR−18a*hsa−miR−339−5phsa−miR−330−3phsa−miR−4289hsa−miR−107hsa−miR−324−3phsa−miR−378*hsa−miR−301ahsa−miR−324−5phsa−miR−500ahsa−miR−29bhsa−miR−185hsa−miR−101hsa−miR−126*hsa−miR−10bhsa−miR−132hsa−miR−335*hsa−let−7a*hsa−miR−183hsa−miR−942hsa−miR−126hsa−miR−1281hsa−miR−1271hsa−miR−452hsa−miR−28−5phsa−miR−1233hsa−miR−532−3phsa−miR−1180hsa−miR−374ahsa−miR−193a−5phsa−miR−10ahsa−miR−3651hsa−miR−148ahsa−miR−130ahsa−miR−200chsa−miR−877*hsa−miR−422ahsa−miR−92b*hsa−miR−486−5phsa−miR−1287hsa−miR−3131hsa−miR−941hsa−miR−1226hsa−miR−221*hsa−miR−1301hsa−miR−140−3phsa−miR−194hsa−miR−629*hsa−miR−1247hsa−miR−455−3phsa−miR−130bhsa−miR−3607−5phsa−miR−28−3phsa−miR−20a*hsa−miR−675*hsa−miR−195*hsa−miR−378bhsa−miR−31*hsa−miR−24−2*hsa−miR−497hsa−miR−1296hsa−miR−96hsa−miR−188−5phsa−miR−920hsa−miR−502−5phsa−miR−140−5phsa−miR−660hsa−miR−662hsa−miR−1285hsa−miR−146b−5phsa−miR−340hsa−miR−361−3phsa−miR−193a−3phsa−miR−1231hsa−miR−491−5phsa−miR−874hsa−miR−99b*hsa−miR−181ahsa−miR−181chsa−miR−3676hsa−miR−22*hsa−miR−615−3phsa−miR−331−3phsa−miR−15ahsa−miR−330−5phsa−miR−301bhsa−miR−582−5phsa−miR−224*hsa−miR−149hsa−miR−105hsa−miR−532−5phsa−miR−362−5phsa−miR−135bhsa−miR−1238hsa−miR−17*hsa−miR−93*hsa−miR−326hsa−miR−596hsa−miR−135ahsa−miR−190hsa−miR−370hsa−miR−501−5p

−1 0 1 2

Polysomes Monosomes

R1 R2 R3 R4 R1 R2 R3 R4

Ct

B

18

19

20

21

22

Ct m

iR-2

1

25

27

29

31

Ct m

iR-1

26

*

C

hsa−miR−23ahsa−miR−720hsa−miR−24hsa−miR−23bhsa−let−7ehsa−let−7ahsa−miR−196bhsa−miR−1260hsa−miR−25hsa−miR−26ahsa−miR−34ahsa−miR−452hsa−miR−126hsa−miR−340*hsa−miR−455−3phsa−miR−3647−3phsa−miR−590−3phsa−miR−29bhsa−miR−576−5phsa−miR−502−3phsa−miR−342−5phsa−miR−422ahsa−miR−1203hsa−miR−193b*hsa−miR−192hsa−miR−200chsa−miR−30b*hsa−miR−378bhsa−let−7f−1*hsa−miR−324−3phsa−miR−221*hsa−miR−185hsa−miR−500a*hsa−miR−23a*hsa−let−7fhsa−miR−30a*hsa−miR−188−3phsa−let−7ihsa−let−7bhsa−miR−320ahsa−let−7dhsa−miR−221hsa−miR−93hsa−miR−31hsa−miR−151−5phsa−miR−26bhsa−let−7d*hsa−miR−30e*hsa−miR−151−3phsa−miR−423−5phsa−miR−106bhsa−miR−629hsa−miR−103ahsa−miR−99ahsa−miR−532−3phsa−miR−342−3phsa−miR−126*hsa−miR−10bhsa−miR−942hsa−miR−335*hsa−miR−320bhsa−miR−128hsa−miR−421hsa−miR−10ahsa−miR−1180hsa−miR−30dhsa−miR−505*

−1.5 −1.0 −0.5 0.0 0.5 1.0 1.5Ct

S15 Scr

R1 R2 R3 R1 R2 R3 R4

Monosomes

Monosomes Polysomes Monosomes Polysomes

Figure 6. miRNAs Cosediment with Ribosomes and Dissociate from Monosomes after RPG Knockdowns

(A) Supervised hierarchical clustering of the 197 HeLa miRNAs differentially expressed in monosomal versus polysomal fractions demonstrates preferential

polysomal cosedimentation of miRNAs. One hundred ninety-five miRNAs were preferentially present in polysomal fractions.

(B) Dot plots of miR-21 and miR-126*, the only two miRNAs that are enriched in monosomes.

(C) Supervised hierarchical clustering of the 67 miRNAs differentially expressed in monosomes in HeLa transfected with control (Scr) versus S15 siRNA

demonstrates thatS15 knockdown preferentially dissociatesmonosomal miRNAs. Out of 462miRNAs detected in HeLamonosomes and polysomes, 67miRNAs

(15%) were reduced in monosomes after S15 knockdown, suggesting that RPG depletion relieves miRNA-mediated repression of translation initiation.

Comparisons are based on Ct values (1.5-fold or 0.6 Ct differences across quadruplicate experiments based on RT-qPCR Ct values). p values were calculated

significant at 0.05 using a two-sided Welch t test adjusted using the Benjamini and Hochberg method.

See also Figure S6.

Molecular Cell



miRNA levels (Figure S6B), consistent with results from northern

blotting experiments (Figure S3B). S15 knockdown also did not

affect miRNA abundance in polysomes, where the average

mean fold change (0.99 with a range 0.95–1.06) was not statisti-

cally significant according to the p values adjusted for multiple

comparisons. These data indicate that RPG depletion does not

affect miRNA-mediated repression of translation after initiation

or the abundance of miRNAs in P bodies.

In contrast, S15 knockdown reduced monosome association

of 67 out of 462 (15%) miRNAs detected in monosomes

and polysomes, demonstrating that RPGs regulate miRNA-

mediated repression of translation initiation (Figure 6C). Impor-

tantly, all 67 of these miRNAs were reduced in monosomes after

S15 knockdown and none of these miRNAs were increasingly

associated with polysomes. These data suggest that miRNPs

containing these miRNAs dissociated from target mRNAs,

consistent with reduced association of miRNA-targeted

mRNAs with Agos after RPG knockdowns (Figure 4A, 4B,

and S4D). Notably, multiple let-7 family members were among

the 67 miRNAs with reduced expression in monosomes

after S15 knockdown, consistent with derepression of validated

let-7 targets (Figures 2E and 2F). miR-21 was consistently

reduced in monosomes after S15 knockdown (average 3.5-

fold decrease) and was not affected in polysomes (average

1.1-fold increase). However, we note that the decreased

association of miR-21 with monosomes was not considered

AAAAAAA

60S

40S

miRNP

AUG

miRNP

AAAAAAA

60S

40S

60S60S

40S 40S

p53/63/73 p53/63/73

Normal Nucleolar Stress

pre-40S

pre-60S

A

0

0.5

1

1.5

2

2.5

Scr S5 S15 L11 L12

Fo

ld D

ere

pre

ssio

n

Scrp53

p53

GAPDH

Scr p53

Scr

S5

S1

5

L11

L1

2

Scr

S5

S1

5

L11

L1

2

0

0.5

1

1.5

2

2.5

3

3.5

Fold

Change

FLRL

B

p53

GAPDH

DM

SO

0.0

01

0.0

02

5.0

10

ActD DRB

DMSO 0.0005 0.001 0.002 2.5 5.0 10

ActD DRB

C

Figure 7. Reduced RPGs Regulate miRNA Function through p53 Pathways

(A) Upregulation of p53mediates derepression ofmiRNA-targetedmRNAs afterRPG knockdowns. A549was transfected with either control (Scr) orRPG-specific

siRNAs and with either Scr (black) or p53 (gray) siRNAs. After 48 hr, vectors expressing FL6X reporter with six imperfect CXCR4 miRNA binding sites and RL

reporter with nomiRNA sites were transfected along with the CXCR4miRNA. Dual luciferase assays (top) and western blotting for p53 and GAPDH (bottom) were

performed after 24 hr.

(B) Chemical induction of nucleolar stress phenocopies RPG knockdowns. A549 expressing FL6X and RL reporters was treated with DMSO (control) or indicated

concentrations (mg/ml) of Actinomycin D (ActD) or 5,6-dichloro-1-b-D-ribofuranosylbenzimidazole (DRB). After 24 hr, dual luciferase assays (top) and western

blotting for p53 and GAPDH (bottom) were performed. Bar graphs show the mean ±SD from three independent experiments.

(C) A model of translational derepression of miRNA-targeted mRNAs resulting from altered ribosome subunit biogenesis. When normal RPG expression

generates stoichiometric 40S and 60S subunits, the p53 stress response pathway is not activated and miRNAs repress translation (left). Reduced expression of

RPGs leads to perturbed biogenesis of 40S and 60S ribosomal subunits, leading to nucleolar stress, activation of the p53 pathway, and dissociation of miRNPs

from miRNA-targeted mRNAs (right). While the translation of untargeted mRNAs remains unchanged, miRNA-mediated repression of translation initiation is

relieved, resulting in increased ribosome loading specifically onto miRNA-targeted mRNAs.

See also Figure S7.

Molecular Cell



statistically significant using strict limits (the raw p value for Scr

versus S15 siRNA was 0.03, but the FDR p value [q value] was

0.07).

To determine whether dissociation of the 67 miRNAs from

monosomes after S15 knockdown affected repression, we per-

formed miRNA target prediction analysis using TargetScan on

the set of 479 mRNAs that commonly shift toward polysomes

after RPG, Ago, or eIF6 knockdowns as assessed by RNA-Seq

(Figure 5B and Table S2, part A). Only favorable targets, defined

asmRNAswith TargetScan context score in the top two quartiles

for each miRNA, were considered. We observed a highly sig-

nificant enrichment of mRNA targets of those 67 miRNAs (Chi-

square value of 103.107; p value of 1.973 10�24) amongmRNAs

that shifted to polysomes following RPG knockdowns (Table S3,

part D). These data indicate that RPG knockdowns relieve

repression of translation initiation by decreasing miRNP associ-

ation and increasing ribosome association of these miRNA-

targeted mRNAs.

p53 Pathway Activation Relieves miRNA-MediatedRepressionGlobal analysis of transcriptome changes (Figure S5) identified

robust effects on p53 pathway genes after RPG (S15, L11, and

L12) but not Ago knockdowns (Figures S7A and S7B and Table

S4). Specifically, gene network analysis of mRNAs commonly

dysregulated in RPG knockdowns revealed enrichment for

a DNA repair network (p value of 6.2 3 10�41; Figure S7B) and

a cell-cycle network (p value of 8.1 10�28; Figure S7B). Because

perturbation of ribosome subunit biogenesis induces nucleolar

stress and leads to p53 pathway activation (Deisenroth and

Zhang, 2010; Holzel et al., 2010; Pani�c et al., 2007; Rudra and

Warner, 2004; Volarevic et al., 2000), we hypothesized that

reduced RPG expression led to p53 pathway activation, which,

in turn, reduced repression of translation initiation of miRNA-

targeted mRNAs. Notably, the human papillomavirus 16/18 E6

protein degrades endogenous p53 in HeLa cells (Scheffner

et al., 1990; Werness et al., 1990). Thus, RPG depletion activates

p53 pathway genes even in the absence of the p53 protein,

possibly by activating other p53 family members (such as p63

and p73).

To test the hypothesis that reduced RPG expression induces

p53 pathways that inhibit miRNA-mediated repression, we

used the A549 human lung cancer cell line which expresses

wild-type p53. RPG knockdowns upregulated p53 at the protein

level (Figures 7A and S7C), suggesting induction of the nucleolar

stress response. Importantly, simultaneous knockdown of RPG

and p53 reversed derepression of miRNA-targeted mRNAs

(Figure 7A), directly implicating the p53 pathway in RPG regula-

tion of miRNA activity.

To test the effects of nucleolar stress on miRNA activity

independently of RPG knockdowns, we chemically induced

nucleolar stress with low concentrations of Actinomycin

D (ActD) or 5,6-dichloro-1-b-D-ribofuranosylbenzimidazole

(DRB) (David-Pfeuty et al., 2001; Holzel et al., 2010). Similar to

RPG knockdowns, ActD andDRB treatments led to dose-depen-

dent p53 activation, reduced cell numbers, and specifically

increased the expression of miRNA-targeted mRNAs without

affecting miRNA levels (Figure 7B and S7D). Derepression of

CXCR4 miRNA-targeted FL mRNA was observed at 8 hr after

treatment (Figure S7E), suggesting that decreased miRNA

activity is a direct result of nucleolar stress induction and not an

indirect result of altered cell metabolism. Taken together, our

data demonstrate that induction of nucleolar stress (triggered

by RPG knockdowns or by small molecules) activates the p53

pathway, which leads to decreased miRNA activity (Figure 7C).

DISCUSSION

RPGs Regulate miRNA-Mediated Repressionof Translation InitiationAn RNAi screen for effectors and regulators of miRNA function

identified an unexpected role for RPGs in regulating miRNA-

mediated repression of translation initiation. Previous reports

have implicated RPGs in small RNA pathways in other organ-

isms, suggesting RPG regulation of small RNA pathways may

be conserved across phyla. In Drosophila, 11 RPGs scored in

a screen of small RNA pathways, constituting 6% of all 177

hits (Zhou et al., 2008). However, RPG knockdown did not affect

miRNA-mediated repression but instead increased endo-siRNA-

mediated repression. Depending upon the RPG tested, siRNA-

mediated repression increased in some cases but decreased

in others. In C. elegans, a genome-wide RNAi screen for miRNA

pathway genes identified one RPG, though this hit was not

confirmed (Parry et al., 2007). In contrast to these model organ-

isms, our data functionally implicates RPGs (as a class) as regu-

lators of miRNA-mediated repression of translation initiation in

human cells.

We showed that RPG depletion reduced the association of 67

miRNAs with monosomes without increasingmiRNA association

with polysomes or reducing total miRNA levels. Importantly,

RPG depletion specifically increased polysomal association

of mRNAs targeted by these miRNAs. Together, these high-

throughput data independently confirm the biochemical data

and support a model in which RPG depletion inhibits repression

of translation initiation mediated by miRNAs. Moreover, we

propose that parallel miRNA and mRNA expression profiling

from monosomes and polysomes could be an accurate method

of target mRNA identification.

miRNAs have been shown to activate translation under certain

stress conditions. Steitz and colleagues reported that cells

forced into quiescence can switch miRNA activity from transla-

tional repression to translational activation (Vasudevan et al.,

2007). It is important to note that RPG knockdowns did not acti-

vate translation but rather derepressed miRNA-targeted

mRNAs. In D8, it was impossible to distinguish derepression

from activation because FL was constitutively repressed by

endogenous miR-21, and thus the expression level of unre-

pressed FL was unknown. Therefore, we knocked down RPGs

in HeLa cells transiently transfected with luciferase reporters

with imperfect binding sites to the artificial CXCR4 miRNA

(Figures S2H and S2I). The expression of unrepressed luciferase

(in the absence of CXCR4 miRNA) was consistently higher than

the expression of derepressed luciferase (in the presence of

CXCR4 miRNA) after RPG knockdowns. These data demon-

strate that RPG depletion derepressed but did not activate trans-

lation of miRNA-targeted mRNAs.

Molecular Cell



p53 Rathways Regulate miRNA FunctionRPG and eIF6 knockdowns led to substantial changes in abun-

dance of ribosomal subunits. The p53 pathway and nucleolar

stress have been implicated in sensing perturbed ribosomal

subunit stoichiometry, leading to cell-cycle arrest (Bachand

et al., 2006; Deisenroth and Zhang, 2010; Holzel et al., 2010;

Pani�c et al., 2007; Rudra and Warner, 2004; Volarevic et al.,

2000). Interestingly, gene ontology analysis of transcriptional

changes revealed that cell cycle, DNA replication, and p53

signaling were affected in all RPG knockdowns but not Ago

knockdowns (Figures S7A and S7B and Table S4). Although

the connection between ribosomal subunit imbalance, nucleolar

stress, and the p53 pathway activation was already known, the

connection between ribosomal subunit imbalance and global

miRNA activity has not been established prior to this study.

We propose that global reduction in miRNA-mediated repres-

sion may be an adaptive response allowing cells to increase

translation of subsets of mRNAs in response to nucleolar stress.

miRNA-mediated repression has been linked to oxidative-,

endoplasmic reticulum-, and nutrient deprivation-stress re-

sponses (Bhattacharyya et al., 2006a, 2006b). In contrast to

nucleolar stress, these stress responses do not induce the

p53 pathway and are distinguished by their effects on general

translation. For instance, amino acid starvation increases eIF2a

phosphorylation and leads to global translation inhibition (Bhat-

tacharyya et al., 2006a, 2006b), which we did not observe in

RPG knockdowns (Figure S6C). Amino acid starvation also

does not affect miRNA activity unless an AU-rich element is

present. In contrast, RPG knockdowns derepress both reporter

and endogenous miRNA-targeted mRNAs even in the absence

of AU-rich elements.

TheMajority of miRNA Species Can Repress TranslationPrevious studies have identified miRNA-targeted mRNAs cose-

dimenting with polysomes suggesting repression of translation

postinitiation (Kim et al., 2004; Maroney et al., 2006; Nelson

et al., 2004; Nottrott et al., 2006; Olsen and Ambros, 1999;

Petersen et al., 2006; Seggerson et al., 2002). In agreement

with more recent studies (Bhattacharyya et al., 2006a; Pillai

et al., 2005), we detected miR-21-targeted FL mRNAs predom-

inantly in monosomes, suggesting repression of translation

initiation. There has been much speculation about the causes

and implications of these different results (reviewed in Filipowicz

et al., 2008). One possible explanation may be the efficiency of

repression. Less efficient repression of translation initiation

may result in a larger population of miRNA-targeted mRNAs in

polysomes, whereas more efficient repression of translation

initiation might result in a larger population of miRNA-targeted

mRNAs in monosomes. In D8, where FL is repressed by miR-

21 via six target sites, strong repression and monosome

association was observed.

We report global miRNA expression profiles from HeLa

monosomal and polysomal fractions. Detection of all miRNAs

in both monosomal and polysomal fractions suggests that the

mechanisms of miRNA-mediated translational repression are

not uniform. Differential sedimentation of particular miRNAs

implies that miRNA and/or mRNA identity may affect the mech-

anism of translational repression. miRNAs cosedimenting with

polysomes did not appear to associate with actively translating

ribosomes. Puromycin treatment reduced polysomes (Fig-

ure S3D) but did not affect sedimentation of these miRNAs (Fig-

ure S6A), suggesting that miRNAs may associate with stalled

ribosomes or non-translating multimegadalton complexes

(e.g., P bodies).

eIF6 Indirectly Affects the Human miRNA PathwayOur screen identified eIF6, a eukaryotic translation initiation

factor that has been implicated in growth and transformation

(Gandin et al., 2008). eIF6 binds to 60S subunits and functions

as an antiassociation factor by preventing 60S subunits from

joining with 40S subunits (Ceci et al., 2003). eIF6 has also been

implicated in miRNA-mediated translational repression in worms

and human cells (Chendrimada et al., 2007). Human eIF6 has

been shown to associate with Agos, miRNAs, and 60S subunits

(Chendrimada et al., 2007). In contrast to these observations, we

did not detect any interaction between eIF6 and Agos, miRNAs,

or miRNA-targeted mRNAs (data not shown). Indeed, we did not

detect any association of eIF6 with Agos even by the highly

sensitive Multidimensional Protein Identification Technology

analysis (data not shown). Our data suggest that eIF6 may affect

miRNA-targeted mRNA repression indirectly by altering ribo-

some subunit stoichiometry. eIF6 is required for 18S and 5.8S

rRNA maturation which, in turn, is required for generating the

60S ribosomal subunit (Basu et al., 2001). Indeed, depletion of

eIF6 reduced 60S ribosomal subunits (Figure 1D), the same

phenotype as in 60S RPG knockdowns.

Reduced RPGs in Ribosomopathies and CancersCancer pathway genes were significantly enriched in polysomes

after RPG, Ago, and eIF6 knockdowns (Table S4). Downregula-

tion ofRPGs has been identified in precancerous states, cancers

in situ, and metastatic cancers (van Riggelen et al., 2010). Addi-

tionally, several genetic diseases that predispose patients to

cancers (Avondo et al., 2009; Campagnoli et al., 2008; Gazda

et al., 2008) are characterized by mutations in or reduced

expression of RPGs, including dyskeratosis congenita (DKC1),

cartilage-hair hypoplasia (RMRP), Shwachman-Diamond

syndrome (SBDS), Turner syndrome (S4X), Noonan syndrome

(L6), Camurati-Englemann disease (S18), and, most notably,

Diamond-Blackfan anemia (S7, S15, S17, S19, S24, S27A, L5,

L11, L35A, and L36). Interestingly, decreased expression of

individual RPGs in zebrafish (Amsterdam et al., 2004; Lai et al.,

2009; MacInnes et al., 2008) and flies (Stewart and Denell,

1993; Watson et al., 1992) promotes tumorigenesis, suggesting

that RPGs may act as haploinsufficient tumor suppressors.

This relationship is counterintuitive because rapid growth and

proliferation of tumors must require robust translational activity.

Our data offer a possible resolution in that reduced levels of

RPGs may preferentially derepress protein production from a

cohort of miRNA-targeted messages, many of which contribute

to cellular proliferation and oncogenesis.

EXPERIMENTAL PROCEDURES

Detailed experimental procedures can be found in the Supplemental Experi-

mental Procedures.

Molecular Cell



Polysome Profiling

HeLa cells were incubated with 100 mg/ml cycloheximide (CHX) for 5 min at

37�C and washed on ice twice with 5 ml cold PBS containing 100 ug/ml

CHX. Cells were scraped in 500 ml lysis buffer (15 mM Tris [pH 7.4], 15 mM

MgCl2, 150 mM NaCl, 1% Triton X-100, 100 mg/ml CHX, and 1 mg/ml heparin)

and centrifuged at 12,000 g for 5 min at 4�C. Supernatant was loaded onto

12 ml 4.5%–45% sucrose gradients in 15 mM Tris (pH 7.4), 15 mM MgCl2,

150mMNaCl, and 100 mg/ml CHX. Gradients were centrifuged in SW41Ti rotor

at 39,000 rpm for 2.5 hr at 4�C and 1 ml fractions were collected. RNA was

extracted with Trizol LS (Invitrogen) or by addition of 750 ml guanidinium

hydrochloride and 800 ml isopropanol to 500 ml of a fraction and incubation

at �20�C overnight. Samples were centrifuged at 10,000 rpm for 25 min at

4�C, pellets washed with 70% ethanol, and resuspended in 180 ml TE buffer.

After addition of 20 ml 3 M sodium acetate (pH 5.2) and 600 ml 100% ethanol,

RNA was precipitated at �80�C overnight. Samples were centrifuged at

10,000 rpm for 25 min at 4�C, pellets washed with 70% ethanol, and RNA

was resuspended in water.

RNA-Seq Read Mapping and RPKM Calculation

Short reads were mapped to the human genome and a precomputed set of

splice junctions using Bowtie (Langmead et al., 2009). Reads per kilobase of

exon model per million mapped reads (RPKM) for each Entrez gene was

computed by using all reads mapping to constitutive Refseq exons.

Global miRNA Expression Profiling

RNA from unfractionated HeLa cells and from fractionated monosomes and

polysomes was extracted with Trizol LS (Invitrogen). Complementary DNA

(cDNA) synthesis, cDNA preamplification, and real-time PCR were performed

with the miScript PCR System (QIAGEN). miRNA Ct measures were compared

between monosomes and polysomes over all samples and separately for Scr

and S15 knockdown groups using a two-sided Welch t test. p values were

adjusted with the method of Benjamini and Hochberg and considered sig-

nificant at the 0.05 level. Supervised hierarchical clustering was performed

based on the Euclidean distance function with the complete linkage agglom-

eration method for the miRNAs found to be differentially expressed between

groups. Only comparisons based on Ct were used, rather than DCt, because

expression of the 5S control varies between monosomes and polysomes,

and is altered in S15 knockdown.

SUPPLEMENTAL INFORMATION

Supplemental Information includes four tables, Supplemental Experimental

Procedures, and seven figures and can be found with this article online at

doi:10.1016/j.molcel.2012.04.008.

ACKNOWLEDGMENTS

Wewould like to thank Dr. Steffen Schubert and The Institute of Chemistry and

Cell Biology (ICCB) Screening Facility at Harvard Medical School for helping

with designing and performing the RNAi screen. We also thank Dr. Stephen

Buratowski, Dr. Danesh Moazed, Dr. Kai Wucherpfennig, and Dustin Grie-

semer for insightful discussions. This work was supported by a Distinguished

Young Scholars Award from the W.M. Keck Foundation to C.D.N.

Received: March 17, 2011

Revised: February 16, 2012

Accepted: April 5, 2012

Published online: April 26, 2012

REFERENCES

Amsterdam, A., Sadler, K.C., Lai, K., Farrington, S., Bronson, R.T., Lees, J.A.,

and Hopkins, N. (2004). Many ribosomal protein genes are cancer genes in

zebrafish. PLoS Biol. 2, E139.

Asangani, I.A., Rasheed, S.A., Nikolova, D.A., Leupold, J.H., Colburn, N.H.,

Post, S., and Allgayer, H. (2008). MicroRNA-21 (miR-21) post-transcriptionally

downregulates tumor suppressor Pdcd4 and stimulates invasion, intravasa-

tion and metastasis in colorectal cancer. Oncogene 27, 2128–2136.

Avondo, F., Roncaglia, P., Crescenzio, N., Krmac, H., Garelli, E., Armiraglio,

M., Castagnoli, C., Campagnoli, M.F., Ramenghi, U., Gustincich, S., et al.

(2009). Fibroblasts from patients with Diamond-Blackfan anaemia show

abnormal expression of genes involved in protein synthesis, amino acid

metabolism and cancer. BMC Genomics 10, 442.

Bachand, F., Lackner, D.H., Bahler, J., and Silver, P.A. (2006). Autoregulation

of ribosome biosynthesis by a translational response in fission yeast. Mol. Cell.

Biol. 26, 1731–1742.

Bagga, S., Bracht, J., Hunter, S., Massirer, K., Holtz, J., Eachus, R., and

Pasquinelli, A.E. (2005). Regulation by let-7 and lin-4 miRNAs results in target

mRNA degradation. Cell 122, 553–563.

Basu, U., Si, K., Warner, J.R., and Maitra, U. (2001). The Saccharomyces

cerevisiae TIF6 gene encoding translation initiation factor 6 is required for

60S ribosomal subunit biogenesis. Mol. Cell. Biol. 21, 1453–1462.

Bazzini, A.A., Lee, M.T., and Giraldez, A.J. (2012). Ribosome Profiling Shows

That miR-430 Reduces Translation Before Causing mRNA Decay in

Zebrafish. Science, in press. Published online March 15, 2012.

Behm-Ansmant, I., Rehwinkel, J., Doerks, T., Stark, A., Bork, P., and

Izaurralde, E. (2006). mRNA degradation by miRNAs and GW182 requires

both CCR4:NOT deadenylase and DCP1:DCP2 decapping complexes.

Genes Dev. 20, 1885–1898.

Bhattacharyya, S.N., Habermacher, R., Martine, U., Closs, E.I., and Filipowicz,

W. (2006a). Relief of microRNA-mediated translational repression in human

cells subjected to stress. Cell 125, 1111–1124.

Bhattacharyya, S.N., Habermacher, R., Martine, U., Closs, E.I., and Filipowicz,

W. (2006b). Stress-induced reversal of microRNA repression and mRNA

P-body localization in human cells. Cold Spring Harb. Symp. Quant. Biol. 71,

513–521.

Boyerinas, B., Park, S.M., Shomron, N., Hedegaard, M.M., Vinther, J.,

Andersen, J.S., Feig, C., Xu, J., Burge, C.B., and Peter, M.E. (2008).

Identification of let-7-regulated oncofetal genes. Cancer Res. 68, 2587–

2591.

Campagnoli, M.F., Ramenghi, U., Armiraglio, M., Quarello, P., Garelli, E.,

Carando, A., Avondo, F., Pavesi, E., Fribourg, S., Gleizes, P.E., et al. (2008).

RPS19 mutations in patients with Diamond-Blackfan anemia. Hum. Mutat.

29, 911–920.

Ceci, M., Gaviraghi, C., Gorrini, C., Sala, L.A., Offenhauser, N., Marchisio, P.C.,

and Biffo, S. (2003). Release of eIF6 (p27BBP) from the 60S subunit allows 80S

ribosome assembly. Nature 426, 579–584.

Chendrimada, T.P., Finn, K.J., Ji, X., Baillat, D., Gregory, R.I., Liebhaber, S.A.,

Pasquinelli, A.E., and Shiekhattar, R. (2007). MicroRNA silencing through

RISC recruitment of eIF6. Nature 447, 823–828.

David-Pfeuty, T., Nouvian-Dooghe, Y., Sirri, V., Roussel, P., and Hernandez-

Verdun, D. (2001). Common and reversible regulation of wild-type p53 function

and of ribosomal biogenesis by protein kinases in human cells. Oncogene 20,

5951–5963.

Deisenroth, C., and Zhang, Y. (2010). Ribosome biogenesis surveillance:

probing the ribosomal protein-Mdm2-p53 pathway. Oncogene 29, 4253–

4260.

Dorner, S., Lum, L., Kim, M., Paro, R., Beachy, P.A., and Green, R. (2006). A

genomewide screen for components of the RNAi pathway in Drosophila

cultured cells. Proc. Natl. Acad. Sci. USA 103, 11880–11885.

Eulalio, A., Rehwinkel, J., Stricker, M., Huntzinger, E., Yang, S.F., Doerks, T.,

Dorner, S., Bork, P., Boutros, M., and Izaurralde, E. (2007). Target-specific

requirements for enhancers of decapping in miRNA-mediated gene silencing.

Genes Dev. 21, 2558–2570.

Filipowicz, W., Bhattacharyya, S.N., and Sonenberg, N. (2008). Mechanisms of

post-transcriptional regulation by microRNAs: are the answers in sight? Nat.

Rev. Genet. 9, 102–114.

Flygare, J., Kiefer, T., Miyake, K., Utsugisawa, T., Hamaguchi, I., Da Costa, L.,

Richter, J., Davey, E.J., Matsson, H., Dahl, N., et al. (2005). Deficiency of

Molecular Cell



http://dx.doi.org/doi:10.1016/j.molcel.2012.04.008

ribosomal protein S19 in CD34+ cells generated by siRNA blocks erythroid

development and mimics defects seen in Diamond-Blackfan anemia. Blood

105, 4627–4634.

Friedman, R.C., Farh, K.K., Burge, C.B., and Bartel, D.P. (2009). Most

mammalian mRNAs are conserved targets of microRNAs. Genome Res. 19,

92–105.

Gandin, V., Miluzio, A., Barbieri, A.M., Beugnet, A., Kiyokawa, H., Marchisio,

P.C., and Biffo, S. (2008). Eukaryotic initiation factor 6 is rate-limiting in

translation, growth and transformation. Nature 455, 684–688.

Gazda, H.T., Sheen, M.R., Vlachos, A., Choesmel, V., O’Donohue, M.F.,

Schneider, H., Darras, N., Hasman, C., Sieff, C.A., Newburger, P.E., et al.

(2008). Ribosomal protein L5 and L11 mutations are associated with cleft

palate and abnormal thumbs in Diamond-Blackfan anemia patients. Am. J.

Hum. Genet. 83, 769–780.

Guo, H., Ingolia, N.T., Weissman, J.S., and Bartel, D.P. (2010). Mammalian

microRNAs predominantly act to decrease target mRNA levels. Nature 466,

835–840.

Holzel, M., Orban, M., Hochstatter, J., Rohrmoser, M., Harasim, T.,

Malamoussi, A., Kremmer, E., Langst, G., and Eick, D. (2010). Defects in

18 S or 28 S rRNA processing activate the p53 pathway. J. Biol. Chem. 285,

6364–6370.

Johnson, S.M., Grosshans, H., Shingara, J., Byrom, M., Jarvis, R., Cheng, A.,

Labourier, E., Reinert, K.L., Brown, D., and Slack, F.J. (2005). RAS is regulated

by the let-7 microRNA family. Cell 120, 635–647.

Kim, J., Krichevsky, A., Grad, Y., Hayes, G.D., Kosik, K.S., Church, G.M., and

Ruvkun, G. (2004). Identification of many microRNAs that copurify with

polyribosomes in mammalian neurons. Proc. Natl. Acad. Sci. USA 101,

360–365.

Kim, J.K., Gabel, H.W., Kamath, R.S., Tewari, M., Pasquinelli, A., Rual, J.F.,

Kennedy, S., Dybbs, M., Bertin, N., Kaplan, J.M., et al. (2005). Functional

genomic analysis of RNA interference in C. elegans. Science 308, 1164–

1167.

Kirn-Safran, C.B., Oristian, D.S., Focht, R.J., Parker, S.G., Vivian, J.L., and

Carson, D.D. (2007). Global growth deficiencies in mice lacking the ribosomal

protein HIP/RPL29. Dev. Dyn. 236, 447–460.

Lai, K., Amsterdam, A., Farrington, S., Bronson, R.T., Hopkins, N., and Lees,

J.A. (2009). Many ribosomal protein mutations are associated with growth

impairment and tumor predisposition in zebrafish. Dev. Dyn. 238, 76–85.

Langmead, B., Trapnell, C., Pop, M., and Salzberg, S.L. (2009). Ultrafast and

memory-efficient alignment of short DNA sequences to the human genome.

Genome Biol. 10, R25.

Lee, Y.S., and Dutta, A. (2007). The tumor suppressor microRNA let-7

represses the HMGA2 oncogene. Genes Dev. 21, 1025–1030.

Lempiainen, H., and Shore, D. (2009). Growth control and ribosome biogen-

esis. Curr. Opin. Cell Biol. 21, 855–863.

Lewis, B.P., Shih, I.H., Jones-Rhoades, M.W., Bartel, D.P., and Burge, C.B.

(2003). Prediction of mammalian microRNA targets. Cell 115, 787–798.

Lim, L.P., Lau, N.C., Garrett-Engele, P., Grimson, A., Schelter, J.M., Castle, J.,

Bartel, D.P., Linsley, P.S., and Johnson, J.M. (2005). Microarray analysis

shows that some microRNAs downregulate large numbers of target mRNAs.

Nature 433, 769–773.

MacInnes, A.W., Amsterdam, A., Whittaker, C.A., Hopkins, N., and Lees, J.A.

(2008). Loss of p53 synthesis in zebrafish tumors with ribosomal protein gene

mutations. Proc. Natl. Acad. Sci. USA 105, 10408–10413.

Maroney, P.A., Yu, Y., Fisher, J., and Nilsen, T.W. (2006). Evidence that

microRNAs are associated with translating messenger RNAs in human cells.

Nat. Struct. Mol. Biol. 13, 1102–1107.

Mayr, C., and Bartel, D.P. (2009). Widespread shortening of 30UTRs by

alternative cleavage and polyadenylation activates oncogenes in cancer cells.

Cell 138, 673–684.

Miller, S.A., Brown, A.J., Farach-Carson, M.C., and Kirn-Safran, C.B. (2003).

HIP/RPL29 down-regulation accompanies terminal chondrocyte differentia-

tion. Differentiation 71, 322–336.

Nelson, P.T., Hatzigeorgiou, A.G., and Mourelatos, Z. (2004). miRNP:mRNA

association in polyribosomes in a human neuronal cell line. RNA 10,

387–394.

Nottrott, S., Simard, M.J., and Richter, J.D. (2006). Human let-7a miRNA

blocks protein production on actively translating polyribosomes. Nat. Struct.

Mol. Biol. 13, 1108–1114.

Oliver, E.R., Saunders, T.L., Tarle, S.A., and Glaser, T. (2004). Ribosomal

protein L24 defect in belly spot and tail (Bst), a mouse Minute. Development

131, 3907–3920.

Olsen, P.H., and Ambros, V. (1999). The lin-4 regulatory RNA controls develop-

mental timing in Caenorhabditis elegans by blocking LIN-14 protein synthesis

after the initiation of translation. Dev. Biol. 216, 671–680.

Oristian, D.S., Sloofman, L.G., Zhou, X., Wang, L., Farach-Carson, M.C., and

Kirn-Safran, C.B. (2009). Ribosomal protein L29/HIP deficiency delays

osteogenesis and increases fragility of adult bone in mice. J. Orthop. Res.

27, 28–35.

Pani�c, L., Montagne, J., Cokari�c, M., and Volarevi�c, S. (2007). S6-haploinsuf-

ficiency activates the p53 tumor suppressor. Cell Cycle 6, 20–24.

Parry, D.H., Xu, J., and Ruvkun, G. (2007). A whole-genome RNAi Screen for

C. elegans miRNA pathway genes. Curr. Biol. 17, 2013–2022.

Petersen, C.P., Bordeleau, M.E., Pelletier, J., and Sharp, P.A. (2006). Short

RNAs repress translation after initiation in mammalian cells. Mol. Cell 21,

533–542.

Pillai, R.S., Bhattacharyya, S.N., Artus, C.G., Zoller, T., Cougot, N., Basyuk, E.,

Bertrand, E., and Filipowicz, W. (2005). Inhibition of translational initiation by

Let-7 MicroRNA in human cells. Science 309, 1573–1576.

Rudra, D., and Warner, J.R. (2004). What better measure than ribosome

synthesis? Genes Dev. 18, 2431–2436.

Saleh, M.C., van Rij, R.P., Hekele, A., Gillis, A., Foley, E., O’Farrell, P.H., and

Andino, R. (2006). The endocytic pathway mediates cell entry of dsRNA to

induce RNAi silencing. Nat. Cell Biol. 8, 793–802.

Scheffner, M.,Werness, B.A., Huibregtse, J.M., Levine, A.J., and Howley, P.M.

(1990). The E6 oncoprotein encoded by human papillomavirus types 16 and

18 promotes the degradation of p53. Cell 63, 1129–1136.

Seggerson, K., Tang, L., and Moss, E.G. (2002). Two genetic circuits repress

the Caenorhabditis elegans heterochronic gene lin-28 after translation initia-

tion. Dev. Biol. 243, 215–225.

Stewart, M.J., and Denell, R. (1993). Mutations in the Drosophila gene encod-

ing ribosomal protein S6 cause tissue overgrowth. Mol. Cell Biol. 13, 2524–

2535.

Talotta, F., Cimmino, A., Matarazzo, M.R., Casalino, L., De Vita, G.,

D’Esposito, M., Di Lauro, R., and Verde, P. (2009). An autoregulatory loop

mediated by miR-21 and PDCD4 controls the AP-1 activity in RAS transforma-

tion. Oncogene 28, 73–84.

Ulvila, J., Parikka, M., Kleino, A., Sormunen, R., Ezekowitz, R.A., Kocks, C.,

and Ramet, M. (2006). Double-stranded RNA is internalized by scavenger

receptor-mediated endocytosis in Drosophila S2 cells. J. Biol. Chem. 281,

14370–14375.

van Riggelen, J., Yetil, A., and Felsher, D.W. (2010). MYC as a regulator of

ribosome biogenesis and protein synthesis. Nat. Rev. Cancer 10, 301–309.

Vasudevan, S., Tong, Y., and Steitz, J.A. (2007). Switching from repression

to activation: microRNAs can up-regulate translation. Science 318, 1931–

1934.

Volarevic, S., Stewart, M.J., Ledermann, B., Zilberman, F., Terracciano, L.,

Montini, E., Grompe, M., Kozma, S.C., and Thomas, G. (2000). Proliferation,

but not growth, blocked by conditional deletion of 40S ribosomal protein S6.

Science 288, 2045–2047.

Watson, K.L., Konrad, K.D., Woods, D.F., and Bryant, P.J. (1992). Drosophila

homolog of the human S6 ribosomal protein is required for tumor suppression

in the hematopoietic system. Proc. Natl. Acad. Sci. USA 89, 11302–11306.

Werness, B.A., Levine, A.J., and Howley, P.M. (1990). Association of human

papillomavirus types 16 and 18 E6 proteins with p53. Science 248, 76–79.

Molecular Cell



Wightman, B., Ha, I., and Ruvkun, G. (1993). Posttranscriptional regulation of

the heterochronic gene lin-14 by lin-4 mediates temporal pattern formation in

C. elegans. Cell 75, 855–862.

Wu, L., and Belasco, J.G. (2005). Micro-RNA regulation of the mammalian lin-

28 gene during neuronal differentiation of embryonal carcinoma cells. Mol. Cell

Biol. 25, 9198–9208.

Wu, L., Fan, J., and Belasco, J.G. (2006). MicroRNAs direct rapid deadenyla-

tion of mRNA. Proc. Natl. Acad. Sci. USA 103, 4034–4039.

Zhou, R., Hotta, I., Denli, A.M., Hong, P., Perrimon, N., and Hannon, G.J.

(2008). Comparative analysis of argonaute-dependent small RNA pathways

in Drosophila. Mol. Cell 32, 592–599.

Molecular Cell



Reduced Expression of Ribosomal Proteins Relieves MicroRNA ...

Documents

Transcript of Reduced Expression of Ribosomal Proteins Relieves MicroRNA ...