Supplementary Materials and Methods

Cell treatment and sequencing protocols. 293S and HeLa cells were grown in DMEM + 10% FBS + penicillin/streptomycin, in adherent monolayer, to ~70% confluency and treated with 1 mM sodium arsenite for 30 minutes, or 1 μM hippuristanol for 30 minutes, or 20 μg/mL emetine for 60 minutes. For immunofluorescence imaging, cells were grown on glass coverslips in 6-well plates, and stained by the Kedersha and Anderson protocol1, except blocking and incubation was done in 0.5% BSA in PBS. Anti-eIF3η (N-20, Santa Cruz Antibodies) and anti-Ago2 (clone 11A9, Sigma) antibodies were used. For nascent protein synthesis determination by 35S labeling, 5 μL of EasyTag EXPRESS35S Protein Labeling Mix (Perkin Elmer) were added to treated cells in 6-well plates for an additional 10 minutes, the cells were washed once in PBS, harvested and analyzed by 10% SDS-PAGE. For CLIP-seq, treated samples, along with untreated controls, were processed using a modification of the Darnell procedure2 (see detailed protocol at the end of Supplementary Materials and Methods). For each replicate/condition, 2-5 15 cm plates were used. For arsenite treatment, 4 biological replicates were used with the Abnova Ago2 antibody (clone 2E12-1C9). For hippuristanol treatment, two biological replicates were split and used with the Abnova antibody, a Sigma Ago2 antibody (clone 11A9), and a Santa Cruz Ago2 antibody (clone 4F9). For emetine treatment, one biological replicate was split and used with the above three antibodies. During the isolation of Ago2-RNA crosslinks from the protein membrane in the emetine experiment, two adjacent regions of the membrane containing the Ago2-mRNA "HI" smear, from 130kD and up, were excised and sequenced separately, but showed similar sequence content and were analyzed together. For RNA-seq, total RNA from aliquots of cells harvested for CLIP was extracted with Trizol. The aqueous phase was further extracted twice with acid phenol and once with chloroform, ethanol-precipitated and washed twice with 70% ethanol. Sequencing libraries were prepared from 1 μg of total RNA using a modified NSR primer protocol3. Briefly, total RNA was annealed with 2 μL of 100 μM 1st strand NSR primer pool in a 10 μL volume at 65 °C for 5 minutes. The 1st strand reverse transcription reaction mix (4 μL 5x reaction buffer, 1 μL 0.1 M DTT, 4 μL 10 mM dNTPs, 1 μL Superscript III (Invitrogen)) was added. The reaction was incubated at 40 °C for 90 minutes (in some replicates, the reaction was incubated for 30 minutes, without apparent differences in the result), followed by 15 minutes at 70 °C. RNase H (1 μL, Invitrogen) was added for a 20 minute incubation at 37 °C, followed by 15 minutes at 75 °C. The DNA was purified with the Qiaquick PCR purification kit, eluted in 30 μL EB buffer, and subjected to second strand synthesis: 25 μL of cDNA, 10 μL 10X NEBuffer 2, 5 μL 10mM dNTPs, 4 μL of 5 U/μL exo- Klenow (NEB), 10 μL of 100 μM 2nd strand NSR primer pool, and 46 μL H2O. The reaction was incubated at 37 °C for 30 minutes and purified over Qiaquick columns as above. Next, the PCR reaction was set up as follows: 25 μL purified DNA, 20 μL 5x buffer 2, 10 μL 25 mM MgCl2, 5 μL 10mM dNTPs, 10 μL 10 μM P5-SBS3T-NSR primer (see oligonucleotide sequence table below), 10 μL 10 μM P7-SBS8-NSR primer, 1 μL Expand High Fidelity Plus Polymerase (Roche), and 19 μL H2O. The PCR reaction was cycled with the following program: 94 °C for 2 minutes; 2 cycles of 94 °C for 10 seconds, 40 °C for 2 minutes, and 72 °C for 1 minute; 25 cycles of 94 °C for 15 seconds, 60 °C for 30 seconds, and 72 °C for 1 minute; 72 °C for 5 minutes. The PCR products were run on a 2% agarose gel, excised in the 200-500bp range and gel-purified. The above libraries were sequenced on one Illumina GAIIx lane each.

Polysome profiling and RNA-seq. One 15 cm plate of 293S cells at ~70% confluency was used for each condition. Cells were washed twice with cold PBS + 100 mg/mL cycloheximide, harvested by scraping, centrifuged and resuspended in 850 mL of hypotonic lysis buffer (5 mM Tris pH 7.5, 2.5 mM MgCl2, 1.5 mM KCl). Next, 12 mL RNasin Plus (Promega), and cycloheximide and DTT to a final concentration of 100 mg/mL and 2 mM respectively, were added and mixed by vortexing. 25 mL of 10% Triton X-100 and 25 mL of 10% sodium deoxycholate were added, and the cells were vortexed and centrifuged for 2 minutes at 14,000 rpm in a tabletop centrifuge. The supernatants were loaded onto 11 mL 10-50% sucrose gradients (10 to 50 % sucrose, 20 mM Hepes pH 7.5, 100 mM KCl, 5 mM MgCl2, 1 mM DTT) and centrifuged at 38,000 rpm for 2 hours at 4 °C in an SW41 rotor. The gradients were fractionated into 8 1.5 mL fractions using a Teledyne Isco density gradient fractionator. From each fraction, 375 mL were spiked with 1 mL of a 1:10 dilution of ERCC RNA spike-ins (mix 1, Ambion), and total RNA was extracted by Trizol LS, with an additional acid phenol-chloroform and a chloroform extraction to exclude DNA contamination. RNA-seq libraries were constructed from 230 ng of starting total RNA material from each fraction using the above NSR protocol, but with 3` (reverse) PCR primers suitable for multiplexing and sequencing on Illumina HiSeq (see oligonucleotide table in Supplementary Materials and Methods). Three to four individual libraries were multiplexed on a single HiSeq lane.

QPCR of fractionated RNA samples. ActB, ATF4 and ERCC00096 spike-in RNA levels were quantified by Taqman assays (see oligonucleotide table in Supplementary Materials and Methods for primer sequences) in triplicates of the PCR reaction. The detection was multiplexed by dye (ATF4 – FAM, ActB – HEX, ERCC00096 – Cy5). For PKM, NUP37, FKBP11 and SEC31B mRNAs, FAM-labeled Taqman probe sets were purchased from IDT, and assayed without multiplexing, along with ERCC00130 spike-in RNAs. mRNA abundance was normalized to spike-in levels and further normalized to a sum of 1 across all fractions in a given condition. For fraction 3 of the +arsenite condition, amplification of the spike-in controls failed, and a Ct value equal to the average of all of the other fractions was assumed for that fraction. Processing of the data without normalization yielded very similar profiles.

Processing of RNA-seq data associated with CLIP samples. Reads were trimmed to nucleotides 9-36 and reverse-complemented to obtain sense strand sequences, and mapped to the hg19 human genome assembly using Bowtie version 0.12.8 iteratively with 0, 1, and 2 mismatches, retaining the mapped reads from each stage and passing the non-mappers to the next stage, only allowing unique mappers. Reads that did not map to the genome were processed with Tophat4 to identify reads mapping to RefSeq RNA and lincRNA5 splice junctions. Combined reads were annotated for RefSeq, lincRNA, microRNA and RepeatMasker categories, and analyzed by DESeq6.

Processing of CLIP libraries for Ago2 binding site identification. Sequenced reads were trimmed of the fixed 3` adaptor and 5` adaptor sequences. The 5` RNA adaptor contains a 4-nucleotide random region (a "counting" barcode) that serves to count distinct 5` ligation events in the library preparation. The count of such unique barcodes associated with each unique CLIPped RNA fragment (the insert sequence without adaptors) was used as its readcount, instead of the total read multiplicity. This avoids PCR amplification bias in the quantitation. The insert sequences without adaptors were mapped to the hg19 human genome assembly using Bowtie7 version 0.12.8 iteratively with 0, 1, and 2 mismatches, retaining the mapped reads from each stage and passing the non-mappers to the next stage, only allowing unique mappers. Reads that did not map to the genome were processed with Tophat4 to identify reads mapping to RefSeq RNA and lincRNA5 splice junctions. Reads mapping to identical genomic positions were collapsed and summed. Combined reads were annotated for RefSeq, lincRNA, microRNA and RepeatMasker categories. To identify peaks, reads from all replicates for each condition were pooled and used as input to Findpeaks8 version 4.0. Parameters were set to denote peaks separated by valleys of 50% height or less, and to trim the sides of peaks dropping below 10% of peak height. To identify peaks from two conditions (eg., +/- arsenite) that correspond to the same Ago2 binding site, while allowing for some differences (shifting) in their end coordinates, peaks from the two conditions that overlapped more than 50% of their width were designated as the same peak/binding location, and the union of their coordinates was used as the genomic coordinates for this "peak pair" or "combined peak". This included pseudo-pairs, with a peak present in one condition and completely absent in the other. Annotations for peaks were obtained as above, and the coordinates of combined peaks were used to re-extract readcounts from individual replicates under the peak. Readcounts for the corresponding mRNAs from RNA-seq replicates were appended to the dataset, and peaks were filtered for the presence of greater than 10 readcounts across all CLIP replicates in an experiment (treatment and control), as well as greater than 10 readcounts of the corresponding mRNA. CLIP readcounts were then normalized to the miRNA fraction (separately for each replicate), and RNA-seq readcounts were normalized across replicates by applying DESeq-derived sizing factors. Normalized replicate counts were padded with a value of 0.25, and the CLIP signal for each peak was then divided by the corresponding mRNA abundance in the replicate.

Processing of CLIP libraries to assess miRNA abundance. Four biological replicate "LO" libraries for the –arsenite and +arsenite conditions were processed separately. Reads were trimmed of 3` adaptor and 5` adaptor sequences, including the random barcodes (thus using the total read multiplicity as the readcount), and mapped to the hg19 human genome assembly using Bowtie7 iteratively with 0, 1, and 2 mismatches, retaining the mapped reads from each stage and passing the non-mappers to the next stage. Multiple mappers were allowed for this analysis. Reads mapping to identical genomic positions were collapsed and summed, and annotated with miRBase9 version 18 microRNA genomic positions. Reads mapping to different genomic loci of the same annotated miRNA (eg, let-7a-1 and let-7a-2) were further collapsed and summed. The resulting miRNA counts were filtered for more than 10 reads across all replicates/conditions, and analysed for differential expression using the DESeq6 R package. To obtain the top 50 miRNA seed families, miRNA counts were collapsed and summed for each miRNA family that shares a common 7mer seed (nucleotides 2-8 of mature -3p or -5p miRNAs), and processed by DESeq to yield normalized mean counts across replicates/conditions. The top 50 miRNA families, spanning ~ 380-fold in abundance and accounting for 97.6% of all miRNA counts, were selected.

Processing of polysome profile RNA-seq libraries. Reads were processed as for RNA-seq above, with an added initial mapping step to the ERCC spike-in sequences. ERCC scaling factors were calculated by linear regression of length-corrected counts for each ERCC RNA between samples. Non-ERCC reads were then mapped further as above. Exonic read counts were collapsed by gene, filtered for genes with 10 or more counts across all fractions for each condition (+/- arsenite), normalized to ERCC scaling factors, and further normalized to a total abundance of 1 across all fractions within a condition.

DNA oligonucleotides used:

PCR primers for NSR sequenced on Illumina GAIIx:
P5-SBS3T-NSR / AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCTCT
P7-SBS8-NSR / CAAGCAGAAGACGGCATACGAGATCGGTCTCGGCATTCCTGCTGAACCGCTCTTCCGATCTGA
3` PCR primers for NSR sequenced on Illumina HiSeq (used instead of P7-SBS8-NSR):
Mplex rev NSR PCR, index 1 / CAAGCAGAAGACGGCATACGAGATCGTGATGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
Mplex rev NSR PCR, index 2 / CAAGCAGAAGACGGCATACGAGATACATCGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
Mplex rev NSR PCR, index 3 / CAAGCAGAAGACGGCATACGAGATGCCTAAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
Mplex rev NSR PCR, index 4 / CAAGCAGAAGACGGCATACGAGATTGGTCAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
Mplex rev NSR PCR, index 5 / CAAGCAGAAGACGGCATACGAGATCACTGTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
Mplex rev NSR PCR, index 6 / CAAGCAGAAGACGGCATACGAGATATTGGCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
Mplex rev NSR PCR, index 7 / CAAGCAGAAGACGGCATACGAGATGATCTGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
Mplex rev NSR PCR, index 8 / CAAGCAGAAGACGGCATACGAGATTCAAGTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
Mplex rev NSR PCR, index 9 / CAAGCAGAAGACGGCATACGAGATCTGATCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
Mplex rev NSR PCR, index 10 / CAAGCAGAAGACGGCATACGAGATAAGCTAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
Mplex rev NSR PCR, index 11 / CAAGCAGAAGACGGCATACGAGATGTAGCCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
Mplex rev NSR PCR, index 12 / CAAGCAGAAGACGGCATACGAGATTACAAGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGA
qPCR primers
ATF4_5`PCR6 / GCCATGGCGCTTCTCACG
ATF4_3`PCR5 / CTT TGC TGG AAT CGA GGA ATG TGC
ATF4_probe1 / /56-FAM/AG CAG CGT TGC TGT AAC CGA CAA AGA /3IABkFQ/
ERCC00096 Set 1 Forward Primer / TCG CAG ACG GTA TCA ACA GGA ACA
ERCC00096 Set 1 Reverse Primer / TCA TGA CTG GAC ACT GCA TCG GAA
ERCC00096 Set 1 Probe / /5Cy5/AT ATT GAG GCT GCT TCG TGT CGG CA/3IAbRQSp/
ACTB Set 1 Forward Primer / TCA GAA GGA TTC CTA TGT GGG CGA
ACTB Set 1 Reverse Primer / TTT CTC CAT GTC GTC CCA GTT GGT
ACTB Set 1 Probe / /5HEX/AA GAG AGG CAT CCT CAC CCT GAA GTA /3IABkFQ/
ERCC00130 Set 1 Forward Primer / ATG ATA TCC CGA TGC TGA CGG CTT
ERCC00130 Set 1 Reverse Primer / TCT TGG CGC AAA TAG CGC TGA ATC
ERCC00130 Set 1 Probe / /5Cy5/AA TTT GAG GAC GAC GCT GCA GCC TTT /3IAbRQSp/

Argonaute2 CLIP protocol