Molecular Cell, Volume 46

Supplemental Information

Reduced Expression of Ribosomal Proteins

Relieves MicroRNA-Mediated Repression Maja M. Janas, Eric Wang, Tara Love, Abigail S. Harris, Kristen Stevenson, Karlheinz Semmelmann, Jonathan M. Shaffer, Po-Hao Chen, John G. Doench, Subrahmanyam V.B.K. Yerramilli, Donna S. Neuberg, Dimitrios Iliopoulos David E. Housman, Christopher B. Burge, and Carl D. Novina Supplemental Experimental Procedures Transfections and cell culture HeLa and A549 cells were cultured in DMEM supplemented with 10% adult bovine serum. Cells were typically transfected with 20 nM siRNAs for 72 hrs or 5 nM LNA antagomir (control antagomir: 5’-T*+C*A*A*+G*A*T*+C*A*G*+T*C*T*+C*A*T*+A*A*G*G*T-3’; miR-21 antagomir: 5’-T*+C*A*A*+C*A*T*+C*A*G*+T*C*T*+G*A*T*+A*A*G*C*T-3’, where “+N” indicates an LNA base, and the “*” indicates a phosphorothioate linkage) for 24 hrs using Hiperfect (Qiagen) according to manufacturer’s instructions. Plasmid DNA was transfected with Lipofectamine2000 (Invitrogen) for 24 or 48 hrs according to manufacturer’s instructions. Generation of the D8 cell line HeLa cells were transfected with FL containing six imperfectly-complementary sites for miR-21 (in pcDNA3.1), cells were selected with neomycin (G418) and single-cell cloned. Subsequently, RL with no miRNA sites (in pcDNA 3.1) was transfected, cells were selected with Hygromycin B and single-cell cloned. Plasmids The following plasmids were used in this study: pGL3-KRAS (Addgene 13919); HMGA2 3’UTR wt luciferase (Addgene 14785); N-terminally-tagged (HA) RPGs cloned into pcDNA3.1 between BamHI and XhoI; N-terminally-tagged (FLAG) Ago1 and Ago2 cloned into pcDNA3.1 between EcoRI and NotI. Antibodies The following antibodies were used: Drosha 07-717 (Upstate), Actin 13E5 #4970 (Cell Signaling), Ago1 07-599 (Millipore), Ago2 04-642 (Millipore), eIF6 611120 (BD), FLAG 2368 (Cell Signaling), RPS6 A00465 (GenScript), RPL7a A300-749A (Bethyl), GAPDH 2118 (Cell Signaling), p53 (DO-1) sc-126 (Santa Cruz). siRNAs The siRNAs were obtained from Qiagen and had the following target sequences: CAGGCTGTGTTCTCAGGATGA (RPS5), CAGACTGAGCGTGCCTACCAA (RPS11), TGGAGGTGTAATGGACGTTAA (RPS12), CACCTACAAGCCCGTAAAGCA (RPS15), TGGAACGTGTGATCACCATTA (RPS18), AAAGCTCATGCTGCTATACGA (RPL5), CTGGTTCCAGCAGAAGTATGA (RPL11), CCTGATCATCAAAGCCCTCAA (RPL12), TACGCCCGAGATGAAACAGAA (RPL35A), CAGCGTGGGTATCGAGGCGGA (RPLP2), CAGCTACAACTTAGATCCCTA (Ago1), CAGGCGTTACACGATGCACTT (Ago2), CCGAGCUUCGUUCGAGAACAACU (eIF6), AACGAGUAGGCUUCGUGACUU (Drosha) UUCUCCGAACGUGUCACGUdTdT (Scr).

RT-qPCR RNA was extracted with Trizol (Invitrogen). Reverse transcription was performed using random hexamer primers with SuperScript III (Invitrogen) after DNase I treatment according to manufacturer’s instructions. Platinum SYBR Green qPCR SuperMix-UDG with ROX (Invitrogen) was used for qPCR detection. The following primer pairs were used: FL (Fwd: CGTTGACCGCCTGAAGTCTCTGATTA, Rev: GGGTGTTGGAGCAAGATGGAT), RL (Fwd: AATGGCTCATATCGCCTCCTGGAT, Rev: TGGACGATGGCCTTGATCTTGTCT), eIF6 (Fwd: TGGTGAATGACTGGTGTGCCTTCT, Rev: AGGTGAGGCTGTCAATGAGGGAAT), RPS5 (Fwd: GTGGAGCGCCTCACTAACTC, Rev: CCCTGTGAGCAGGTGTATGA, RPS11 (Fwd: TTGACAAGAAATGCCCCTTC, Rev: GTGACCTTGAGCACGTTGAA), RPS12 (Fwd: TTGCTGCTGGAGGTGTAATG, Rev: AACCACTTTACGGGGTTTCC), RPS15 (Fwd: CTTCCGCAAGTTCACCTACC, Rev: TCTCCACCTGGTTGAAGGTC), RPS18 (Fwd: GAGGATGAGGTGGAACGTGT, Rev: GGACCTGGCTGTATTTTCCA), RPL5 (Fwd: ATTATGCTCGGAAACGCTTG, Rev: TTCATCACCAGTCACCTCCA), RPL11 (Fwd: CATCCGCAAACTCTGTCTCA, Rev: ATGTGTTCCTGGATCCCAAA), RPL12 (Fwd: GGGCCTGAGGATTACAGTGA, Rev: CGATGATGTCATGAGGATGG), RPL35A (Fwd: CCAGAGTCATCTGGGGAAAA, Rev: TCCTTGAGGGGTACAGCATC), RPLP2 (Fwd: AGCTTGCCAGTGTACCTGCT, Rev: AAAAAGGCCAAATCCCATGT), GAPDH (Fwd: GAAGGTGAAGGTCGGAGT, Rev: GAAGATGGTGATGGGATTTC). Luciferase assay HeLa cells were lysed in 1X Passive Lysis Buffer (Promega), incubated on a shaker for 15 min at room temperature, and transferred to white 96-well plates. FL and RL were measured using the Dual-Glo Luciferase Assay System (Promega) on Victor 3V Multilabel Plate Reader (Perkin Elmer). Northern blotting To detect mRNAs, RNA was resolved on 1% agarose in 1X MOPS and 1.85% formaldehyde, gels were equilibrated in 0.5X TBE and transferred onto Hybond membrane (GE Healthcare) using a semi-dry transfer cell (BioRad). RNA was UV-crosslinked to the membrane, pre-hybridized at 42°C in 5X SSPE, 2X Denhardt’s solution, and 0.1% SDS containing denatured salmon sperm DNA. Membranes were incubated overnight at 42°C with probes prepared using DECAprime II kit (Ambion). Membranes were washed for 10 min at room temperature with 2X SSC, 0.1% SDS and for 10 min with 0.2X SSC, 0.1% SDS. To detect small RNAs, RNA was resolved on 12% urea-PAGE and probed with 5' end-labeled probes. Pulse metabolic labeling HeLa cells were washed three times with 1X PBS and incubated for 20 min at 37°C with labeling media (RPMI 1640 without L-glutamine, L-cystine, L-methionine, supplemented with 10% dialyzed FBS and 25 mM HEPES pH 7.4), followed by incubation with 10 uCi/ml EXPRESS 35S Protein Labeling Mix (Perkin Elmer) for 20 min at 37°C. Cells were washed three times with 1X PBS and lysed in RIPA lysis buffer. Immunoprecipitation HeLa cells were lysed in lysis buffer containing 50 mM Tris pH 7.4, 100 mM NaCl, 0.5 mM EDTA, 1% Triton X-100, 0.4 U/uL RNase inhibitors, and protease inhibitor cocktail tablet, and centrifuged at 12,000g for 10 min at 4°C. Mouse IgG agarose (Sigma) and anti-FLAG M2 agarose (Sigma) were washed in TBS (50 mM Tris pH 7.4, 100 mM NaCl, 0.5 mM EDTA, 1% Triton X-100). After pre-clearing lysates for 1 hr at 4°C with mouse IgG agarose, IP was performed for 2 hrs at 4°C with anti-FLAG agarose pre-blocked with bovine serum albumin (BSA) and tRNA. After washing the beads five times with TBS, complexes were eluted with 150 ug/ml FLAG peptide in lysis buffer by shaking for 30 min at 4°C.

Sucrose cushion HeLa cells were incubated with 100 ug/ml cycloheximide (CHX) for 5 min at 37°C and washed twice on ice with 5 mL cold PBS containing 100 ug/ml CHX. Cells were scraped in 300 µL lysis buffer (0.1% Triton X-100, 50 mM HEPES pH 7.6, 100 mM KCl, 2 mM MgCl2, 100 ug/mL CHX, 0.4 U/uL RNase inhibitors, protease inhibitor cocktail tablet), incubated on ice for 15 min and centrifuged at 3,500g for 15 min at 4°C. Supernatants were layered onto 4.5 mL 0.8 M sucrose in the lysis buffer and centrifuged at 55,000 rpm in SW55Ti rotor for 2 hrs at 4°C. Pellets were rinsed with PBS and resuspended in the lysis buffer. RNA-Seq sample preparation RNA was diluted to 50 uL with water, heated at 65°C for 3 min, placed on ice and mixed for 5 min at room temperature with 100 uL dynal oligo(dT) beads (Invitrogen) that were washed twice with 100 uL binding buffer (20 mM Tris pH 7.5. 1 M LiCl, 2 mM EDTA) and resuspended in 50 uL binding buffer. After poly(A) mRNA binding, the beads were washed twice with 100 uL washing buffer (10 mM Tris-HCl pH 7.5, 0.15 M LiCl, 1 mM EDTA). RNA was eluted in 20 uL of 10 mM Tris pH 7.5 by heating at 80°C for 2 min. Beads were washed and purification was repeated. Eluted RNA was incubated at 70°C for 5 min in 1X fragmentation buffer (Ambion) and the reaction was stopped by addition of stop solution (Ambion) on ice. RNA was precipitated and reverse transcribed using random hexamer primers with SuperScript III (Invitrogen). Second strand cDNA was synthesized at 16°C for 2.5 hrs in 1X second strand buffer (Invitrogen) containing dNTPs, 2 U RNaseH (NEB), 50 U DNA Polymerase I (NEB), . DNA was purified with a QIAquick PCR purification kit (Qiagen). End repair reaction was performed at 20°C for 30 min in 1X T4 DNA ligase buffer (NEB) containing dNTPs, 15 U T4 DNA Polymerase (NEB), 5 U Klenow DNA polymerase (NEB), and 50 U T4 PNK (NEB). DNA was purified with QIAquick PCR purification kit. Single ‘A’ base was added at 37°C for 30 min in NEB2 buffer (NEB) containing 1 mM ATP and 15 U Klenow 3’ to 5’ exo. DNA was purified with QIAquick minElute kit (Qiagen). Adaptors were ligated using Quick Ligase (NEB), DNA was purified with QIAquick minElute kit (Qiagen), and 300 bp band was gel-purified from 2% agarose gel using QIAquick gel extraction kit (Qiagen). DNA was amplified (15-21 cycles) using Phu DNA polymerase (Finnzymes), and 300 bp band was gel-purified from 2% agarose gel for Illumina RNA-Seq. Overlap of gene regulation Fold-change values for input, polysome, and input/polysome ratios between knockdowns and controls were computed, and the numbers of genes up- or down-regulated in each knockdown relative to controls were counted. The numbers of genes up- or down-regulated in pairs, triples, sets of 4, … all 8 knockdowns were also counted to obtain observed values for the number of genes regulated in common among different numbers of knockdown experiments. To obtain the expected number of genes regulated in common, the same analysis was performed using genes shuffled in a manner that preserves the distribution of gene expression values (or input/polysome ratio for the input/polysome ratio analysis) in each control sample. Expected values were obtained by taking the mean of 1000 shuffling trials. The extent to which genes were commonly regulated when using k knockdowns, where k=2, 3, … 8, was computed by calculating the mean observed ratio and mean expected ratio among all combinations of k knockdowns. miRNA enrichment in commonly regulated genes Conserved 8-mer miRNA sites with Pct > 0.5 were obtained from TargetScan version 5.1. The numbers of miRNA sites in 3’UTRs of commonly regulated genes among k knockdowns, where k = 2, 3, … 8, were counted for all combinations of k knockdowns, for miRNAs expressed in HeLa cells. The expected number of miRNA sites was obtained by repeating the analysis with a set of randomly selected genes sharing the same gene expression in control samples and 7-mer conservation in 3’UTRs. One thousand control sets were chosen, and the fraction of genes containing miRNA sites in each trial was compared to the fraction found in the true set of commonly regulated genes. The same procedure was applied for various subsets of k

knockdowns, and the mean fraction of genes containing miRNA sites for all combinations of k knockdowns were plotted, for the true (observed) and control (expected) subsets of genes. Global miRNA expression profiling Following the manufacturer’s instructions, reverse transcription reactions containing a fixed volume of RNA were assembled using 5X miScript HiSpec Buffer. The reactions were incubated for 60 min at 37°C followed by a heat inactivation for 5 min at 95°C. cDNA pre-amplification was then performed using the suggested cycling program for whole miRNome profiling, which consists of an initial hold of 95°C for 15 min, 2 cycles of 94°C for 30 sec, 55°C for 1 min, 70°C for 1 min, 10 cycles of 94°C for 30 sec, and 60°C for 3 min. Real-time PCR was performed using the miScript SYBR Green PCR Kit and the Human miRNome miScript miRNA PCR Array (assays current through miRBase v16). Real-time PCR was performed on an ABI-7900HT (Applied Biosystems) using the miScript cycling program, which consists of an initial hold of 95°C for 15 min followed by 40 cycles of 94°C for 15 sec, 55°C for 30 sec, and 70°C for 30 sec. A miRNA was deemed to be expressed if its Ct value was less than 32, its dissociation curve was a single-sharp melt peak, and its dissociation curve temperature was indicative of specific miRNA amplification.

Figure S1. D8 Reports on miRNA Biogenesis and mRNA Degradation, Related to Figure 1 (A) Inhibiting miR-21 with an antagomir relieves repression of FL translation. The miR-21-specific antagomir was transfected into D8 every 48 hrs, and cells were harvested every 24 hrs to assess FL and RL protein levels by the dual luciferase assay, and FL and RL mRNA levels by RT-qPCR. (B) Knockdown of miRNA biogenesis factors (Drosha and DGCR8), miRNA effector factor (eIF6), and mRNA degradation factors (PARN and DCP2) leads to siRNA-dose-dependent derepression of FL. D8 was transfected with indicated siRNAs at increasing concentrations and FL and RL levels were assessed by the dual luciferase assay after 72 hrs. Bar graphs are the mean ±SD from three independent experiments.

Figure S2. Depletion of RPGs or Ribosome Biogenesis Factors Relieves miRNA-mediated Translational Repression but Not siRNA-Mediated mRNA Cleavage, Related to Figure 2 (A) Depletion of ribosome biogenesis factors increases expression of miRNA-targeted but not untargeted mRNAs. D8 or HeLa was transfected with siRNAs against factors involved in either 40S (Bms1 and Tsr1) or 60S (Bop1 and Nip7) subunit biogenesis, and after 48 hrs, HeLa was transfected with vectors expressing either RL-HMGA2 3’UTR, RL with six imperfect binding sites for the artificial CXCR4 miRNA (RL6X), or RL-KRAS 3’UTR reporters, and FL lacking miRNA binding sites (control) . After 24 hrs, FL and RL levels were assessed by the dual luciferase assay. (B) Double knockdown of a 40S and a 60S RPG leads to greater derepression of FL compared to a single knockdown. D8 was transfected with the indicated siRNA pairs, and FL and RL levels were assessed by the dual luciferase assay after 72 hrs. (C) Confirmation of RPG knockdown efficiencies by RT-qPCR. D8 or HeLa was transfected with siRNAs against indicated genes, and RPG mRNA levels were assessed by RT-qPCR and normalized to GAPDH mRNA levels after 72 hrs. (D) Knockdowns of small and large RPGs in D8 derepress FL on the protein

level and, to a lesser extent, on the mRNA level. D8 was transfected with Scr siRNA or siRNAs against indicated RPGs, FL and RL protein levels were assessed by dual luciferase assay, and FL and RL mRNA levels were assessed by RT-qPCR after 72 hrs. (E) Depletion of small and large RPGs in D8 increases FL protein levels while decreasing cell number, as reflected in decreased RL protein levels. D8 was transfected with Scr siRNA or siRNAs against indicated RPGs, and FL and RL protein levels were assessed by the dual luciferase assay after 72 hrs. (F) FL protein levels increase while RL protein levels remain unchanged after RPG knockdowns. D8 was transfected with Scr or RPG-specific siRNAs, and Western blotting was performed after 72 hrs. The amounts of lysates were adjusted for decreased cell number in RPG knockdowns. (G) Reconstitution of siRNA-resistant RPGs reverses the derepression of FL caused by RPG knockdowns in D8. Silent mutations conferring siRNA resistance were introduced into indicated RPG expression constructs. D8 was transfected first with indicated siRNAs, and after 24 hrs, with vectors expressing siRNA-resistant RPG cDNAs. Fold derepression relative to control transfection was assessed by the dual luciferase assay 48 hrs after vector transfection (left). Expression of siRNA-resistant HA-tagged RPG constructs was confirmed by Western blotting with an anti-HA antibody (right). (H) FL with six imperfect binding sites for the artificial CXCR4 miRNA (FL6X) is derepressed after RPG knockdowns. RPGs were knocked down in HeLa, FL6X reporter (test) and RL reporter lacking CXCR4 miRNA binding sites (control) were transfected with CXCR4 miRNA after 48 hrs, and dual luciferase assays were performed after 24 hrs. (I) RL with six imperfect binding sites for CXCR4 miRNA (RL6X) is derepressed after RPG knockdowns only in the presence of the targeting CXCR4 miRNA. RPGs were knocked down in HeLa, RL6X reporter (test) and FL reporter lacking CXCR4 miRNA binding sites (control) were transfected with or without CXCR4 miRNA after 48 hrs, and dual luciferase assays were performed after 24 hrs. (J) An FL reporter containing one perfectly-complementary site for an artificial CXCR4 miRNA (FL1P) is not significantly derepressed after RPG depletions. RPGs were knocked down in HeLa, FL1P reporter (test) and RL reporter lacking CXCR4 miRNA binding sites (control) were transfected with CXCR4 miRNA after 48 hrs, and dual luciferase assays were performed after 24 hrs. As expected, Ago2 knockdown significantly derepressed FL1P, while Ago1 knockdown increased mRNA cleavage, possibly by increasing the pool of free CXCR4 miRNA available for Ago2 binding. (K) Knockdown of five representative small and large RPGs derepresses transiently-transfected RL reporter containing HMGA2 3’UTR with seven sites imperfectly complementary to endogenous let-7. HeLa was transfected with siRNAs against indicated genes, and after 48 hrs, with vectors expressing RL-HMGA2 3’UTR reporter (test) and FL reporter with no miRNA sites (control). After 24 hrs, FL and RL levels were assessed by the dual luciferase assay. Ago1 and eIF6 knockdowns served as positive controls. (L) Knockdown of small and large RPGs derepresses RL reporter containing KRAS 3’UTR with seven sites imperfectly complementary to endogenous let-7. HeLa was transfected with siRNAs against indicated genes, and after 48 hrs, with vectors expressing RL-KRAS 3’UTR reporter (test) and FL reporter with no miRNA sites (control). After 24 hrs, FL and RL levels were assessed by the dual luciferase assay. Bar graphs are the mean ±SD from three independent experiments.

Figure S3. RPG Knockdowns Increase the Loading of Active Ribosomes onto miR-21-Targeted FL mRNA but Do Not Affect miRNA Biogenesis, Related to Figure 3 (A) RPG depletions do not alter the levels of primary miRNA transcripts. HeLa was transfected with indicated siRNAs, and the levels of pri-miR-21 were assessed by RT-qPCR and normalized to GAPDH mRNA levels after 72 hrs. (B) RPG knockdowns do not alter levels of mature or pre-miRNAs. HeLa with RPG knockdowns were Northern blotted for miR-17, miR-21, let-7a, pre-miR-17, pre-miR-21, pre-let-7a, and tRNA (as a loading control). Notably, Ago2 knockdown decreased mature miRNA levels, possibly due to altered miRNP stability. (C) RPG knockdowns do not affect the levels of Ago proteins. Western blots for Ago1, Ago2, and GAPDH after RPG knockdowns in D8 are shown. (D) Rapidly-sedimenting complexes formed on miR-21-targeted FL mRNA after RPG knockdowns are actively-translating polysomes. Polysome profiles demonstrate that knockdowns of S15 or L12 shift the miRNA-targeted FL mRNA but not untargeted RL mRNA to actively-translating polysomes. D8 was transfected with Scr, S15-, or L12-specific siRNAs, and after 72 hrs, cells were treated with or without puromycin for 5 min to dissociate only actively-translating ribosomes, and lysates were resolved on sucrose gradients. As expected, without puromycin treatment, FL mRNA was detected in rapidly-sedimenting fractions only after RPG knockdowns. However, after puromycin treatment, rapidly-sedimenting complexes formed on FL mRNA after RPG knockdowns shifted toward slowly-sedimenting fractions. Therefore, rapidly-sedimenting complexes formed on FL mRNAs are actively-translating polysomes. RL mRNA also shifted to slowly-sedimenting fractions after puromycin treatment confirming that RL mRNAs are associated with actively-translating polysomes both before and after RPG depletion. (E) Quantification of Northern blots using ImageQuant. FL and RL mRNA detected in each fraction is represented as % of the total mRNA detected in all fractions across the gradient. Bar graphs are the mean ±SD from three independent experiments.

Figure S4. RPG Knockdowns Perturb Ribosomal Subunit Stoichiometry and miRNP Association with Target mRNAs, Related to Figures 3 and 4 (A) A254 absorbance traces of lysates fractionated on continuous sucrose gradients after depletion of RPGs demonstrate that knockdown of RPGs alters the stoichiometry between 40S, 60S, and 80S. HeLa or D8 were transfected with indicated siRNAs, and lysates were fractionated on sucrose gradients after 72 hrs. Knockdown of small RPGs consistently leads to a reduction in free 40S, an increase in free 60S, a reduction in 80S, and no change in polysomes. Knockdown of large RPGs consistently leads to an increase or no detectable change in free

40S, a reduction in free 60S, a reduction in 80S, and no change in polysomes. These A254 absorbance traces confirm that the knockdowns of RPGs were efficient and functional without the need for Western blotting. Each column represents an independent group of experiments. (B) rRNA levels are reduced after RPG knockdowns. Ethidium bromide staining of total RNA from RPG knockdown HeLa shows that knockdowns of 40S RPGs decrease the levels of 18S rRNA (component of the 40S ribosomal subunit) without affecting 28 rRNA levels (component of the 60S ribosomal subunit). Knockdowns of 60S RPGs decrease the levels of both 18S rRNA and 28 rRNA. As expected, eIF6 knockdown has the same phenotype as 60S RPG knockdowns while Ago1 knockdown has no effect on rRNA levels. 28S/18S ratio based on band intensity was quantified using ImageQuant and is shown below the gel. (C) The distribution of FL and RL mRNAs in polysome gradients is consistent regardless of polysome profiling method used. In other methods that do not resolve 40S, 60S, and 80S, 10-50% gradients are centrifuged for 2 hrs at 35,000 rpm. In the method used throughout this study, 4.5-45% gradients were centrifuged for 2.5 hrs at 39,000 rpm to resolve 40S, 60S, and 80S. When using the former method, FL mRNA still demonstrates polysomal shift after RPG knockdowns, while RL mRNA remains largely polysomal. (D) Knockdown of S15 or L12 decreases association of Ago1 and Ago2 with endogenous miRNA-targeted mRNAs leading to their stabilization. HeLa was transfected with Scr, S15- or L12-specific siRNAs, and after 24 hrs, with vectors expressing FLAG-tagged Ago proteins. After 48 hrs, KRAS and PTEN mRNAs were detected in IPs and inputs by RT-qPCR and normalized to GAPDH mRNA. IP efficiency was assessed by Western blotting for FLAG-Ago. Bar graphs are the mean ±SD from three independent experiments.

Figure S5. Stability of miRNA-Targeted mRNAs Is Not Globally Affected by RPG, Ago, and eIF6 Knockdowns, Related to Figure 5 (A) Genes whose mRNA levels change after knockdown of Ago, RPGs, or eIF6 are shared. The numbers of genes whose mRNA levels are increased or decreased by > two-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 (obtained using a permutation procedure) are shown in parentheses below the observed values. The observed/expected ratio for each comparison is displayed in color; all comparisons exhibit positive observed/expected ratios. (B) A common set of mRNAs increases and decreases in Ago, RPG, and eIF6 knockdowns The numbers of genes whose mRNA levels are increased or decreased by > two-fold in 2-way, 3-way, ... 8-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. 90% confidence intervals are shown in pale red, blue, and black. The observed/expected ratio (dotted line) increases as the number of knockdowns increases. (C) The up- and down-regulated mRNAs are not enriched for HeLa-expressed miRNA target sites. The fraction of genes with HeLa-expressed conserved 8mer miRNA sites was obtained for those genes whose mRNA levels are up- or down-regulated in 2, 3, ... 8 knockdowns and shown in red or blue, respectively. Control genes were selected to match the up- or down-regulated genes, and analogous values were obtained (black).

Figure S6. RPG Knockdowns Have No Global Effects on miRNA Levels or eIF2α Phosphorylation, Related to Figure 6 (A) Polysome-enriched miRNAs are insensitive to puromycin treatment and thus are not associated with actively-translating polysomes. HeLa were treated with cycloheximide (chx) to preserve ribosomes on mRNAs or with puromycin (puro) to dissociate actively-translating ribosomes from mRNAs. Polysomal RNA was subjected to miRNA expression profiling with primers against the 195 miRNAs preferentially associated with polysome fractions in Figure 6C. No significant difference was detected using adjusted p-values. (B) Unsupervised hierarchical clustering of 342 miRNAs detected in HeLa transfected with either Scr (control) or S15 siRNAs demonstrates no global changes in miRNA levels after S15 knockdown. The average fold change was 1.00 with a range of 0.94-1.11. Comparisons are based on Ct values. (C) RPG knockdowns do not increase eIF2α phosphorylation. HeLa was transfected with siRNAs against indicated genes, and after 72 hrs, phosphorylated eIF2α (eIF2α-P) and GAPDH protein levels were assessed by Western blotting.

Figure S7. p53 Pathways Are Activated in RPG Knockdowns, Related to Figure 7 (A) p53-related pathways are significantly enriched among genes commonly up- or down-regulated on the mRNA levels in three RPG knockdowns (S15, L11, and L12) as assessed by RNA-Seq (Supplemental Table S2A). The gene ontology analysis was performed using DAVID Bioinformatic Database (v6.7). The p-values were calculated with Fisher exact test analysis and corrected p-values were calculated using the Bonferroni method. (B) Representative p53-related networks enriched in RPG knockdowns (top). Because p53 is inactive in HeLa, p53 was added to the networks (bottom) after the gene ontology analysis. (C) RPG knockdowns up-regulate p53 on the protein level. A549 expressing wildtype p53 was transfected with either control (Scr) or RPG-specific siRNAs and with either Scr or p53-specific siRNAs. After 72 hrs, Western blotting for p53 and GAPDH were performed. (D) Chemical induction of nucleolar stress does not affect miRNA levels. A549 was treated with DMSO (control) or indicated concentrations (µg/mL) of Actinomycin D (ActD) or 5,6-dichloro-1-b-D-ribofuranosylbenzimidazole (DRB). After 24 hrs, Northern blotting for let-7a and miR-21 was performed and normalized to tRNA. (E) Chemical induction of nucleolar stress phenocopies RPG knockdowns as soon as 8 hrs post-treatment. A549 expressing FL6X reporter with six imperfect CXCR4 miRNA binding sites (test) and RL reporter with no miRNA sites (control) was treated with DMSO (control) or indicated concentrations (µg/mL) of ActD or DRB. After 8 hrs, the dual luciferase assay was performed.