Examining SRP pathway function in mRNA localization to the endoplasmic reticulum

CRISPR/Cas9 knockout cell lines and allele-specific genome sequencing

HeLa and 293T cells were transduced with a Cas9-containing lentivirus encoding blasticidin resistance. Transduced cells were selected with 10 µg/mL blasticidin and tested for Cas9 editing efficiency using a positive control sgRNA. High-efficiency editing clones were selected by mutational screening using the Surveyor Mutational Detection Kit (IDT), and performed as described by the manufacturer. High-efficiency editing clones stably expressing Cas9 (HeLa Cas9 WT and 293T Cas9 WT) were expanded and transduced with lentivirus containing sgRNA against SR subunit β (SRB, AGCCCTACCTAGACACCTTG) and including puromycin resistance. Transduced cells were selected with 2 µg/mL puromycin and 10 µg/mL blasticidin to ensure sgRNA and Cas9 expression, respectively, and single-cell cloned. Clones were screened for mutations via the Surveyor Mutational Detection Kit (IDT) following amplification of the SRB genomic DNA region spanning the sgRNA target site, performed with Takara Hot Start DNA Polymerase (Takara Bio USA, Inc.) and the primers: 5′-TGTGTGTCCCTCGACAAATAAC-3′ (forward), 5′-AGTTGGTGTCAAGACCCTGAGT-3′ (reverse). Clones showing mismatch mutations were screened for SRB protein expression via immunoblot, as described below. Clones without detectable SRB protein expression were selected and genomic SRB PCR products sequenced following Zero Blunt TOPO cloning (Thermo Fisher). Bacterial colonies were screened until each independent sequence was identified at least three times for all SRB KO clonal cell lines, to ensure all alleles were detected. Nonsense mutations in the mutated alleles were confirmed by reading the frame analysis of sequenced PCR products using Expasy Translate (Swiss Institute of Bioinformatics). HeLa and 293T clones with mutations were detected by Surveyor Mutational Detection, and lacking protein expression via immunoblot, and with nonsense mutations in all detected alleles, were then expanded (HeLa Cas9 SRB KO1, HeLa Cas9 SRB KO2, and 293T Cas9 SRB KO3).

Sequential detergent fractionation

All samples and reagents were RNase-free and kept on ice unless otherwise specified. Cell monolayers were washed twice with PBS, recovered by scraping, incubated on ice for 10 min to depolymerize the microtubule network, and pelleted by centrifugation (2500 rpm, 4°C, 5 min). Cell pellets were resuspended in ice-cold digitonin cytosol extraction buffer (0.03% digitonin, 110 mM KCl, 25 mM HEPES, pH 7.2, 15 mM MgOAc₂, 4 mM CaCl₂, 1× Roche cOmplete protease inhibitor, 40 U RNaseOut) and incubated on ice for 10 min. Samples were centrifuged at 6000 rpm, 5 min, 4°C to recover the cytosol-extracted permeabilized cells. Supernatants were collected as cytosol fractions and the permeabilized cell pellets were suspended in 100 µL digitonin wash buffer (0.0015% digitonin, 110 mM KCl, 25 mM HEPES, pH 7.2, 15 mM MgOAc₂, 4 mM CaCl₂, 1× Roche cOmplete protease inhibitor, 40 U RNaseOut). Samples were centrifuged (6000 rpm, 5 min, 4°C) to pellet the cytosol-extracted cells. Supernatant fractions were combined as the cytosol fraction and membrane pellets were resuspended in ice-cold n-dodecylmaltoside (DDM) solubilization buffer (2% DDM, 200 mM KCl, 25 mM HEPES, pH 7.4, 15 mM MgOAc₂, 4 mM CaCl₂, 1× Roche cOmplete protease inhibitor, 40 U RNaseOut) and incubated on ice for 10 min. Samples were centrifuged (12,000 rpm, 5 min, 4°C) to clear the samples of nuclei and cell debris. Supernatants were collected as membrane (ER) fractions. Cytosol and ER fractions were processed for protein precipitation or RNA extraction, as described below.

Protein precipitation

Following sequential detergent fractionation or whole-cell lysis with RIPA buffer for 10 min on ice, TCA was added to fractions or lysates to a final concentration of 10% wt/vol. Samples were incubated on ice at 4°C and centrifuged (14,000 rpm, 10 min, 4°C) to pellet TCA-precipitated protein. Supernatants were aspirated and protein pellets washed twice in ice-cold 100% acetone. Washed TCA pellets were dried for 10 min at 95°C and stored at −20°C.

Immunoblotting

TCA-precipitated protein pellets were resuspended in 0.5 M Tris, pH 11, 5% SDS at 95°C for 10 min. Protein concentrations were determined via BCA assay (Pierce) or Nanodrop spectrophotometer (Thermo Scientific). Equivalent protein quantities were separated by SDS–PAGE, transferred to nitrocellulose membranes, blocked in 5% nonfat dry milk/Tris-buffered saline, and processed for immunoblotting. Antibodies used include SRA (affinity purified rabbit polyclonal antisera, 0.8 µg/mL), SRB (Santa Cruz, sc-376723, 1:100 dilution), SRP54 (Santa Cruz, sc-393855, 1:50 dilution), SRP72 (Thermo Scientific, PA5-49601, 1:500 dilution), GAPDH (DSHB, DSHB-hGAPDH-2G7, 1:500 dilution), Tubulin (DSHB, E-7, 1:1000 dilution), GRP94 (rabbit polyclonal antisera, 1:2000 dilution), and TRAPα (Migliaccio et al. 1992; 1:1000 dilution).

Proteasome inhibition

Where indicated, cells were treated with 10 µM MG132 for 16 h in complete culture media. Cells were then lysed, proteins TCA precipitated, and samples analyzed via immunoblot as described above.

SRB complementation

Cells were transfected with 0.2 µg/mL (1×) or 0.4 µg/mL (2×) SRB expression-ready cDNA plasmid [Transomics, pCS6(BC065299)] or pCMV-SPORT6 empty vector using Lipofectamine 2000, per the manufacturer's recommendation. Samples were processed 48 h post-transfection.

RNA extraction

All samples and reagents were RNase-free and kept on ice unless otherwise specified. RNA was extracted from subcellular fractions or total cell pellets by guanidinium thiocyanate (GT)-phenol–chloroform extraction as described in Chomczynski and Sacchi (2006). Briefly, 2:1 GT: phenol was added to fractions or cell pellets, inverted to mix, and incubated on ice for 15 min. Chloroform was then added to each sample, shaken by hand for 10 sec, and incubated on ice for 15 min. Samples were then phase separated via centrifugation, the aqueous phase was collected, combined with an equal volume of 100% isopropanol, and samples were inverted to mix. After incubation at −20°C, RNA was pelleted by centrifugation, reprecipitated with 1:1 GT: isopropanol, washed twice with 75% ethanol, dried at 65°C, resuspended in nuclease-free water, and RNA purity and concentration were assessed via Nanodrop Spectrophotometer (Thermo Scientific). RNA samples were then used for RNA-seq, RT-PCR, or RT-qPCR as described below.

Reverse transcription and qPCR

Reverse transcription was performed with equal quantities of input total RNA using the iScript cDNA Synthesis Kit (BioRad), per the manufacturer's protocol. cDNA products were diluted 1:10 with nuclease-free water, and quantitative PCR (qPCR) was performed using Luna Universal qPCR Master Mix (New England Biolabs) following the manufacturer's recommendation and a T100 PCR Thermal Cycler (BioRad, 1861096). qPCR data were analyzed with BioRad CFX Maestro software. Results were reported as relative expression between experimental and parental cell lines for a given cell type after normalization to GAPDH mRNA levels. In experiments where mRNA partitioning between the cytosol and ER compartments was examined by RT-qPCR, ΔCq values were determined, normalized to the total, and fractional mRNA distributions calculated. This analysis assumes equivalent PCR efficiencies for the paired cytosol and ER (membrane) fractions. Primer sequences include SRA, 5′-ACCTGTGAGGTCCATGATTGA-3′ (forward) and 5′-GGTTTGCTGGTAGCCAAAGGA-3′ (reverse); SRB, 5′-AGTAGTGGTGGCGGTTCTTG-3′ (forward) and 5′-ACGTTTTCCCGGAATCACAAAG-3′ (reverse); GAPDH, 5′-TCATCAGCAATGCCTCCTGC-3′ (forward) and 5′-GATGGCATGGACTGTGGTCA-3′ (reverse); GRP94, 5′-CTGGAAATGAGGAACTAACAGTCA-3′ (forward) and 5′-TCTTCTCTGGTCATTCCTACACC-3′ (reverse); GRP78, 5′-CAACCAACTGTTACAATCAAGGTC-3′ (forward) and 5′-CAAAGGTGACTTCAATCTGTGG-3′ (reverse); B2M, 5′-TTCTGGCCTGGAGGCTATC-3′ (forward) and 5′-TCAGGAAATTTGACTTTCCATTC-3′ (reverse). For SRB splicing variant analysis, the following primer sets were used: Set 1: (forward) 5′-AGTAGTGGTGGCGGTTCTTG-3′, (reverse) 5′-ACGTTTTCCCGGAATCACAAAG-3′; Set 2: (forward) 5′-AGTAGTGGTGGCGGTTCTTG-3′, (reverse) 5′-CAGCACAGCTGTCAGTAATGG-3′; Set 3: (forward) 5′-TCTGACCTTGATTGACCTTCC-3′, (reverse) 5′-CAGACCCATACTGTCAATGAGG-3′; Set 4: (forward) 5′-CAATGGCAAAAATCAGCAAAG-3′, (reverse) 5′-CAAGTCCTGGATGTCAGCAG (reverse); Set 5: (forward) 5′-TGCAATGGCAAAATCAGCAAAG-3′, (reverse) 5′-GAACTCCACTTTGAGGGGCA-3′.

For noncoding RNA studies, the following primer sets were used: NORAD (forward) 5′-CTC-TGC-TGT-GGC-TGC-CC-3′, (reverse) 5′-GGG-TGG-GAA-AGA-GAG-GTT-CG-3′; MALAT1 (forward) 5′-ACG-ATG-GTG-TCG-AGG-TCT-TT-3′, (reverse) 5′-TCC-CAC-CCA-GCA-TTA-CAG-TT-3′; NEAT1 (forward) 5′-GTG-GCT-GTT-GGA-GTC-GGT-AT-3′, (reverse) 5′-TAA-CAA-ACC-ACG-GTC-CAT-GA-3′.

Polysome gradients and RT-PCR

Polyribosome profiling was performed on cell lysates prepared in DDM lysis buffer (200 mM KCl, 25 mM K-HEPES, pH 7.2, 10 mM MgOAc₂, 2 mM DTT, 100 µg/mL cycloheximide, 1× Roche cOmplete protease inhibitor, 40 U/mL RNaseOUT, and 2% DDM) and resolved on 15%–50% sucrose gradients. Sucrose gradients were fractionated using a Teledyne/Isco gradient fractionator as described previously (Stephens and Nicchitta 2007; Stephens et al. 2008). Total RNA was extracted from collected gradient fractions and reverse transcribed as described above using equal gradient fraction volumes. Equal volumes of cDNA reaction products were used for PCR analysis, performed using Taq DNA polymerase and 10× ThermoPol buffer (NEB), per the manufacturer's recommendation. PCR products were resolved on a 2% agarose gel containing SYBR Safe DNA stain (Thermo Fisher) and visualized using an Amersham Imager 600 (GE Healthcare).

siRNA transfections

HeLa Cas9 WT and HeLa Cas9 SRB KO2 cells were transfected at 60%–80% confluency with nontargeting (Silencer Select Negative Control No. 1 siRNA, Thermo Fisher) or SRA (Silencer Select SRPRA siRNA ID s13458, Thermo Fisher) siRNA using Lipofectamine RNAiMax (Thermo Fisher) following the manufacturer's protocol. One-half of the recommended transfection reagent concentration was used to reduce post-transfection toxicity. Cells were grown for 24–96 h, as indicated, and assayed for protein or RNA content as described above. In experiments examining the effects of combined SRA/SRB knockdown on mRNA localization to the ER, HeLa Cas9 WT cells were transfected with either nontargeting siRNA control, as noted above, or with pooled SRA and SRB siRNA (25 pmol Silencer Select SRPRA siRNA ID s13458, 25 pmol Silencer Select SRPRB siRNA ID s33819). After a 96 h culture period, cells were fractionated into cytosol and ER (membrane) fractions by sequential detergent extraction, as described above, and mRNA levels in the two fractions determined as described above for reverse transcription and qPCR.

Immunofluorescence

For optical imaging experiments, HeLa cells were seeded onto 12 mm round #1 glass coverslips; for 293T cells, coverslips were precoated with poly-l-lysine. Cells were washed with PBS and fixed with 4% paraformaldehyde in PBS for 15 min at room temperature. Cells were then washed with PBS and permeabilized with 0.1% Triton X-100 for 5 min on ice. Cells were blocked in 1% BSA in PBS, stained with primary antibody (polyclonal TRAPα antisera) rabbit, 1:100 dilution and monoclonal tubulin hybridoma antisera, DSHB, E7, mouse, 1:200 dilution), and counterstained with secondary antibody (goat anti-rabbit AlexaFluor 488, Invitrogen, 1:2000 dilution and goat anti-mouse AlexaFluor 647, Invitrogen, 1:2000 dilution) and 1 µg/mL DAPI (Sigma, MBD0015). Stained coverslips were mounted with FluorSave Mounting Reagent (EMD Millipore). Images were obtained on a DeltaVision deconvolution microscope (Applied Precision) with 60× NA 1.4 oil immersion objective (UPlanSApo 100XO; Olympus) and a high-resolution CCD camera (CoolSNAP HQ2; Photometrics). Z-stacks were acquired at 0.2 µm steps using identical exposure conditions across samples for a given protein. Data were deconvolved using the SoftWoRx program (Applied Precision) and further processed using ImageJ/Fiji software to merge channels, linearly adjust contrast, and pseudo color images. All changes were applied uniformly across all images for a given experiment.

Single-molecule fluorescence in situ hybridization

SRB KO2 cells transfected with either nontargeting siRNA or SRA-targeting siRNA were seeded to 60% confluency at the time of transfection and assayed 96 h post-transfection (∼90% confluency). HeLa Cas9 WT and SRB KO2 were seeded to 90% confluency and assayed in parallel with transfected cells. Cells were seeded onto 12 mm round #2 glass coverslips in 24-well plates. Media was removed from wells, cells were washed with DEPC-treated PBS, and fixed with 3.7% paraformaldehyde for 15 min at room temperature. Cells were washed with DEPC-PBS and permeabilized with 0.1% Triton X-100 for 5 min on ice. Cells were incubated in Stellaris Wash Buffer for 5 min at room temperature. A humidified chamber was assembled, and coverslips were incubated in 30 µL Stellaris Hybridization buffer plus 0.3 µL Stellaris respective probe(s) at 37°C overnight in the dark. Coverslips were returned to a 24-well plate, cell-side up, and washed with Stellaris Wash Buffer A for 5 min. DAPI stain comprised 30 µL Stellaris Wash Buffer A + 0.05 µg/mL DAPI. Cells were incubated with 30 µL DAPI stain for 15–30 min in the dark at 37°C. Coverslips were then incubated with Stellaris Wash Buffer B for 5 min at room temperature. Images were obtained on a DeltaVision deconvolution microscope with 60× NA 1.4 oil immersion objective and high-resolution CCD camera, as noted above. For a given transcript, sample exposure conditions were identical, and Z-stacks were obtained in 0.2 µm steps. Deconvolution was performed on SoftWoRx software and image processing performed using ImageJ/Fiji software to merge channels, produce maximum intensity projections, linearly adjust contrast, and pseudo color images. Cell outlines were manually mapped from paired DIC micrographs.

Cell proliferation

Cell proliferation was assessed by labeling cell monolayers with 1 µM CFSE (Thermo Fisher, C34554) for HeLa cell lines or 0.5 µM CFSE for 293T cell lines, per the manufacturer's recommendation. Cells were collected, fixed in 1% formaldehyde, and assayed on a CANTO II Flow Cytometer (BD Biosciences) immediately following fixation. Data were visualized using FlowJo software (BD Biosciences).

[³⁵S] Methionine/cysteine incorporation

Cells were methionine starved for 15 min in methionine- and cysteine-free media at 37°C and then labeled with 50 µCi/mL [³⁵S] methionine/cysteine in methionine- and cysteine-free media for 7.5 min. After the labeling period, cells were washed twice with media containing 100 µg/mL cycloheximide, incubated in the second wash for 10 min at 37°C, washed twice in PBS containing 100 µg/mL cycloheximide, and then lysed for 10 min at room temperature in RIPA buffer containing 100 µg/mL cycloheximide and 1 mM PMSF. Cycloheximide controls were prepared in parallel, with cycloheximide addition prior to [³⁵S] methionine/cysteine. Lysates were collected and cleared by centrifugation (12,000 rpm, 5 min and 4°C). Proteins were precipitated from cleared lysates with 10% final TCA concentration (wt/vol), protein pellets were washed four times with ice-cold 10% TCA, and then washed three times with ice-cold acetone. Washed pellets were dried at 95°C and resuspended in 0.5 M Tris, pH 11, 5% SDS. Radioactivity was assessed by liquid scintillation counting, and protein concentration was assessed by BCA assay (Pierce). Radioactivity data are reported as counts per minute (CPM) per milligram (mg) of protein for experimental samples minus average CPM/mg of cycloheximide controls.

Cell fractionation for quantitative proteomics

HeLa Cas9 WT cells were grown in SILAC media (DMEM, Athens ES, 0420, supplemented with l-glutamine, sodium pyruvate, pen/strep/antifungal, 10% dialyzed FBS, methionine, leucine, and proline) with light arginine and lysine, and HeLa Cas9 SRB KO cells were grown in SILAC media with heavy arginine and lysine (¹³C₆ ¹⁵N₄-l-Arg and ¹³C₆ ¹⁵N₂-l-Lys) for 7 d. On the third day of labeling, SRB KO cells were transfected with SRA siRNA and wild-type cells were transfected with nontargeting siRNA, as described above, in duplicate. Transfected cells were cultured in the appropriate SILAC media for an additional 96 h. At this time point, an equal number of cells from matched replicates (i.e., wild-type, light-labeled replicate and SR-KO, heavy-labeled replicate) were combined, washed twice in HBS, swelled in 15 mM KCl, 1.5 mM MgOAc₂, 1 mM DTT, 10 mM HEPES-KOH, pH 7.5, with protease inhibitor cocktail on ice for 10 min, and homogenized with 25 strokes of a Dounce homogenizer with a tight pestle. Osmotic balancer (375 mM KCl, 22.5 mM MgOAc₂, 1 mM DTT, 220 mM HEPES-KOH, pH 7.5, 1× protease inhibitor; 1/5 volume), was added to the homogenate. Total membranes were isolated from homogenate via ultracentrifugation in a TLA 100.3 rotor at 45,000 rpm for 40 min at 4°C. Supernatants from total membrane centrifugation were collected, centrifuged in a TLA 100.3 rotor at 50,000 rpm for 60 min at 4°C to clear, and TCA precipitated as the cytosol samples. Total membrane pellets were resuspended in 100 mM sodium carbonate, pH 11, plus 4 M urea, homogenized with 100 passes of a Dounce homogenizer with a tight pestle, incubated on ice for 30 min, and centrifuged in a TLA 100.3 rotor at 90,000 rpm for 30 min at 4°C. Carbonate/urea supernatants were collected, and TCA precipitated as the peripheral membrane fraction. Carbonate/urea pellets were gently washed with water and stored as the integral membrane fraction.

Proteomics sample preparation and proteolytic digestion

Cell homogenate-derived protein fractions were resuspended in denaturing buffer (9 M urea, 10 mM HEPES, pH 8.5, 15 mM NaCl) and protein concentrations were measured via BCA assay (Thermo Fisher). Fifty micrograms of each sample was brought to the same volume (150 µL) with lysis buffer. In parallel, 8 µg of each sample (48 µg total) were combined and the volume brought to 150 µL with lysis buffer to serve as a control sample. 1,4-dithiothreitol (DTT) was added to each sample to 1 mM final concentration and incubated at 60°C for 15 min (Eppendorf, Thermomixer C). Iodoacetamide was then added to 1.6 mM per sample and incubated at room temperature for 20 min in the dark. All reactions were quenched with an additional 1 mM DTT. Three hundred and thirty microliters of H₂O and 40 µL 100 mM HEPES (pH 8.5) were added to each sample for enzymatic digestion. Fifteen microliters of Tryp/LysC (0.1 µg/µL, Thermo Fisher Scientific) was added to each sample and shaken at 37°C for 12 h (Eppendorf, Thermomixer C). An additional 5 µL of Tryp/LysC was added to each sample after the 12 h incubation and incubated an additional 4 h at 37°C. Sixty microliters of 10% trifluoroacetic acid was added to each sample to quench the trypsin reaction after the 16 h total incubation. Each sample was then passed through a SEK PAK column (SOLA HRP SPE, Cat. No. 60109-001), and eluates from each sample were dried via Speedvac concentrator (Thermo Fisher Scientific) for TMT labeling.

TMT labeling and TMT label assessment

Dried samples were resuspended in 50 µL 100 mM HEPES (pH 8.5) and gently vortexed. TMT labels (as indicated in Fig. 4A) were resuspended in acetonitrile (Optima, LC/MS grade, Fisher Chemical), and 20 µL of TMT reagent was added to the appropriate samples. All samples were incubated at room temperature for 1 h. An amount of 4 µL of each TMT labeled sample was removed and combined to perform a label quality assessment (Nusinow et al. 2020), which also served as a normalization marker for downstream analysis. The TMT label assessment mixture was desalted with C18 stage tips (PhyNexus Inc.) and combined with 80 µL reconstitution buffer (0.1% formic acid). LC/MS analysis of the label assessment sample identified 9235 TMT labeled peptides in 9280 total peptides. An amount of 35 µL of 5% hydroxylamine solution was added to each TMT labeled sample to quench the TMT reaction. Equal masses of TMT labeled samples were mixed according to the label assessment LC/MS data and separated via high pH reverse phase high-performance liquid chromatography (HPLC), as described below.

High pH reverse phase high-performance liquid chromatography fractionation

The TMT labeled mixture was fractionated by high pH reverse phase HPLC (Vanquish HPLC, Thermo Fisher Scientific). A total of 96 fractions were collected over 75 min at a flow rate of 0.5 mL/min with mobile phase A as 20 mM formic acetate, pH 9.3 in deionized water and mobile phase B as acetonitrile (Optima, LC/MS grade, Fisher Chemical) with 20 mM formic acetate, pH 9.3. Fractions were pooled into 12 total samples (i.e., eight fractions per sample) and were dried and desalted using C18 stage tips (PhyNexus Inc.) before LC–MS–MS.

Liquid chromatography—tandem mass spectrometry (LC–MS–MS)

Fractionated samples were analyzed by nano flow HPLC (Ultimate 3000, Thermo Fisher Scientific), followed by Thermo Orbitrap Mass Spectrometer (Q-Exactive HF-X), and a Nanospray Flex Ion Source (Thermo Fisher Scientific) equipped with a Column Oven (PRSO-V2, Sonation) to heat the nano column (PicoFrit, 100 µm × 250 mm × 15 µm tip, New Objective) for peptide separation. The nano LC method was water/acetonitrile-based and 210 min long with a 0.25 µL/min flow rate. For each TMT fraction, all TMT labeled peptides were first captured on a trap column (No. 160454, Thermo Fisher) and were then delivered to the separation nano column by the mobile phase. A TMT-specific MS2-based mass spectrometry method on QE HF-X was used to sequence TMT peptides that were eluted from the nano column. For the full MS, 120,000 resolutions were used with a 3E6 AGC target or 50 msec IT, with a scan range of 300–1500 m/z. For the dd-MS(MS/MS), 60,000 resolutions were used with a 1E5 AGC target or 100 msec IT. An isolation window of 0.7 Da with a fixed first mass of 110.0 Da was used. Normalized collision energy (NCE) was set to 32 with a 15-cycle loop.

Quantitative proteomic analysis

Collected LC–MS data were analyzed by Proteome Discoverer 2.4 (Thermo Fisher). Because all peptides were labeled with SILAC and TMT tags, multiple quantitative proteomics searches were needed to deconvolve the complex data set. Therefore, a TMT database search was performed separately from a SILAC database search, providing TMT quantitation separate from SILAC protein abundance (Welle et al. 2016). All searches were performed in the Sequest HT node with mass tolerance of 20 ppm MS1 and 0.05 Da for MS2. Homo sapiens database (UP000005640) from Swiss-Prot was used. Percolator node was used for peptide FDR filtering (strict: 0.01, relaxed: 0.05). TMT search: TMT quantitative search was performed for heavy peptides and light peptides. For heavy-labeled TMT search, TMT-based modification (+237.177 Da) was added to accommodate the fact that all TMT labeled lysines were heavy lysine (+8.014 Da). Both heavy TMT modification (+237.177 Da) and heavy Arginine (+10.008 Da) were set as static modification. A standard TMT quantitative proteomics search was performed for the light TMT peptides. SILAC search: SILAC search was performed to parse out heavy-labeled peptides from the light peptides. Standard TMT tag mass (+229.163 Da) was used as variable modification and amino-terminal modification for the SILAC quantification search. For analysis, proteins that were detected in only one cell line were assigned zero abundance values for the paired cell line. All proteomic data have been deposited at the MassIVE repository as dataset MSV000087992.

Transmission electron microscopy

HeLa Cas9 parental cells and HeLa Cas9 SRB KO cells were cultured in standard conditions (DMEM, 10% FCS, 37°C, 5% CO₂). Cells were transfected with SRA-targeting siRNA or a nontargeting siRNA, and 24 h after transfection, cells were split onto #1 15 mm glass coverslips and cultured in 12-well culture plates. At 96 h post-transfection, the media was removed and replaced with 2.5% glutaraldehyde in 0.1 M sodium cacodylate, pH 7.4 and fixed for 1 h at RT. Following the glutaraldehyde fixation period, the 2.5% glutaraldehyde in 0.1 M sodium cacodylate, pH 7.4 fixation buffer was removed and replaced with 2% paraformaldehyde in 0.1 M sodium cacodylate, pH 7.4, and the tissue culture plates stored overnight at 4°C. Subsequently, the 2% paraformaldehyde in 0.1 M sodium cacodylate, pH 7.4 fixation buffer was removed and replaced with 0.1 M sodium phosphate, pH 7.4 and shipped on ice to the Center for Cellular and Molecular Imaging, Electron Microscopy Facility at Yale Medical School. Cell samples were post-fixed in 1% osmium tetroxide for 1 h and stained en-bloc in aqueous 2% uranyl acetate for 1 h, rinsed in distilled water, dehydrated in an ethanol series and resin infiltrated with Embed 812 (Electron Microscopy Sciences). Samples were placed in silicone molds and cured at 60°C for 24 h. Resin blocks were sectioned on a Leica UltraCut UC7. Sixty-nanometer sections were collected on formvar-coated nickel grids and stained using 2% uranyl acetate and lead citrate. Electron micrographs were obtained on a FEI Tencai Biotwin TEM at 80Kv using a Morada CCD and iTEM (Olympus) cellSens Dimension software. Micrographs were blinded and mitochondrial surface areas determined using Image J, by the method of Lam et al. (2021). An amount of 120–150 individual mitochondria were quantified for each condition.

RNA gel electrophoresis

Cell equivalent volumes of RNA extracted from cytosol and ER fractions were diluted twofold with RNA-loading dye (80% formamide, 1 mg/mL xylenol blue, 1mg/mL bromophenol blue, 10 mM EDTA), heated to 65°C for 10 min to denature RNA secondary structure, and separated via electrophoresis in 1% agarose containing SYBR Safe (Thermo Fisher). RNA was visualized via green fluorescence on an Amersham Imager 600 (GE Healthcare).

4-Thiouridine pulse-chase labeling, RNA biotinylation, and detection

Cell cultures were supplemented with 200 µM 4SU (Sigma) for 15 min in DMEM, 10% FBS (pulse). Labeled cells were washed with 1 mM uridine in PBS and then incubated in 1 mM uridine in DMEM, 10% FBS for 0, 1.5, 3, or 6 h (chase). At the indicated time points, cells were washed twice with PBS, incubated on ice, and processed for sequential detergent fractionation as described. Total RNA from the resulting cytosol and ER fractions was extracted by GT-phenol–chloroform as described above. Control samples were processed identically with the exception that 4SU was omitted (mock pulse).

An amount of 100 µg of 4SU-labeled or control RNA was collected from cytosol and ER fractions, or pooled polysome gradient fractions, and treated with DNase I (New England Biolabs) for 10 min at 37°C. DNase I was then inactivated with 5 mM EDTA at 75°C for 10 min. DNA-free RNA was purified using GT-phenol–chloroform extraction as above. An amount of 50 µg RNA per sample was incubated with 5 µg MTSEA-biotin (Biotium) in 10 mM HEPES, pH 7.5, 1 mM EDTA, 20% dimethylformamide for 2 h at RT in light-protected sample tubes. Total biotinylated RNA (including biotinylated, 4SU-labeled RNA and nonlabeled RNA synthesized before or after the 4SU pulse) was extracted with phenol/chloroform/isoamyl alcohol (25:24:1). For assessment of 4SU-labeled rRNA, biotinylated RNA was separated on a 0.8% agarose gel as described above. Electrophoresed RNA was transferred to a nitrocellulose membrane overnight with 10× saline sodium citrate (SSC). RNA was cross-linked to the membrane using a UV crosslinker and stained with methylene blue solution (0.02% wt/vol in 0.3 M sodium acetate pH 5.5, 3–5 min). The membrane was imaged on an Amersham Imager 600 (GE Healthcare) as total, input RNA. The membrane was then destained with 0.2× SSC, 1% wt/vol SDS for 15 min and rinsed with nuclease-free water. To assess 4SU-labled RNA on the destained membrane, the biotinylated RNA was detected using a streptavidin-coupled antibody. The membrane was first equilibrated in 10% SDS, 1 mM EDTA in PBS with gentle shaking for 1.5 h. The membrane was then stained with IRDye 800CW Streptavidin (LI-COR Inc.) at a 1:10,000 dilution in 10% SDS, 1 mM EDTA in PBS for 30 min at room temperature with gentle shaking, protected from light. The membrane was then washed in sequential washes of 10% SDS in PBS, 1% SDS in PBS, and 0.1% SDS in PBS for 8 min each. Finally, the membrane was washed in PBS for 5 min, and IRDye was imaged on an Odyssey Clx (LI-COR Inc.) to detect 4SU-labeled, biotinylated RNA. IRDye signal intensity was quantitated using Empiria Studio Software (LI-COR Inc.) where indicated.

Analysis of noncoding RNA subcellular distributions

HeLa cells at ∼80% confluence were fractionated as described in the Sequential Detergent Fractionation section above, with the exception that fractionations were conducted on the adherent cells. Following solubilization with the DDM-supplemented membrane solubilization buffer, which leaves nuclei intact, the nuclear fraction was obtained by addition of 1.0 mL of a TRIzol/10 cm dish. Cytosol and ER fractions were extracted with TRIzol according to the manufacturer's recommendations for aqueous samples. Total RNA was recovered by phase separation of the TRIzol extracts with quantification by UV spectrometry.

RNA-seq library construction

An amount of 5 µg of RNA extracted from the cytosol, ER, or total cell fractions was DNase treated using TURBO DNase (Thermo) for 30 min at 37°C per the manufacturer's recommendation. DNase was deactivated by adding 0.1× volume 25 mM EDTA to each reaction and heating to 65°C for 10 min. DNase-treated RNA was precipitated by adding 2.5× volume 100% ethanol and 0.1× 3 M NaOAc, pH 5.2, mixing by inversion, and incubating at −20°C for 30 min. Samples were then centrifuged (14,000 rpm, 30 min, at 4°C). Precipitated pellets were washed twice by adding 500 µL 75% ethanol, incubating at room temperature for 10 min, centrifuging at 14,000 rpm, 10 min, 4°C, and repeating. Washed pellets were dried for 10 min at 65°C and 20 µL DEPC-treated water was added to each dried pellet. Samples were heated to 65°C for 10 min to aid resuspension, and RNA purity and concentration were assessed via Nanodrop Spectrophotometer (Thermo Fisher). rRNA depletion was performed on 1 µg DNase-treated RNA per sample using either the NEBNext rRNA Depletion Kit (human/mouse/rat, NEB #E6310) or the KAPA RiboErase Kit (H/M/R, KAPA Biosystems), following the manufacturer's recommendation, except the HighPrep PCR Clean-up System (MagBio) was used for RNA and cDNA cleanup steps in place of the suggested magnetic bead systems. Briefly, for the NEBNext rRNA Depletion Kit, DNase-treated RNA was diluted to 1 µg in 12 total microliters with nuclease-free water. RNA was combined with rRNA depletion solution and hybridization buffer and annealed by heating to 95°C and then ramping down to 22°C at 0.1°C/sec. Annealed RNA was combined with RNase H and buffer and heated to 37°C for 30 min. Digested RNA was then combined with DNase and buffer and heated to 37°C for 30 min. rRNA-depleted RNA was then purified using HighPrep beads. For KAPA RiboErase, DNase-treated RNA was diluted to 1 µg in 10 total µL with nuclease-free water. RNA was combined with hybridization buffer and hybridization oligos and annealed by heating to 95°C and then ramped down to 45°C at 0.1°C/sec. RNase H and buffer were added to annealed RNA and heated to 37°C for 30 min. Digested RNA was isolated using HighPrep beads as described. Beads containing cleaned and digested RNA were then resuspended in DNase and buffer to elute. Once separated from the beads, eluted RNA in DNase and buffer were heated to 37°C for 30 min. rRNA-depleted RNA was then purified using HighPrep beads.

rRNA-depleted RNA was used in RNA-seq library construction with the NEBNext Ultra II RNA Library Prep Kit for Illumina (NEB #E7770), per the manufacturer's recommendation, except that SPRIselect beads were replaced with the HighPrep PCR Clean-up System (MagBio). Briefly, RNA was fragmented using the suggested incubation for intact RNA (RIN > 7), samples were combined with first-strand buffer and random primers, heated to 94°C for 15 min, and immediately placed on ice. First-strand synthesis was then completed, immediately followed by second-strand synthesis. cDNA was purified using HighPrep beads and was eluted in 0.1× TE Buffer. cDNA ends were repaired using NEB End Prep Enzyme Mix and Buffer. Adaptors were ligated to the cDNA using 1:5 diluted adaptors, ligation enhancer, and ligation master mix. Adaptor-ligated cDNA was purified using HighPrep beads. Purified cDNA was then amplified via PCR using NEBNext Index Primers for Illumina 1-48 (NEB #7335, #E7500, #E7710, and #E7730) and seven cycles of extension, as recommended for 1 µg RNA input. PCR amplified cDNA was purified using HighPrep beads and eluted in 20 µL 0.1× TE Buffer. cDNA Libraries were then sequenced on the NovaSeq 6000 platform by the Duke Center for Genomic and Computational Biology.

RNA-seq data analysis

RNA-seq reads were analyzed using Kallisto (Bray et al. 2016). A Kallisto index was built from all RefSeq genes (downloaded from https://ftp.ncbi.nlm.nih.gov/refseq/H_sapiens/mRNA_Prot/ on 04.06.2018). Illumina read files were mapped to the RefSeq index at default paired end settings. Mapped reads were transformed to transcript per million (TPM), including only annotated mRNAs (NM prefix) and summed, to generate a TPM/gene metric.

4SU-seq library preparation

Cells were pulse-labeled with 4SU, incubated in 1 mM uridine for 15, 30, 45, or 60 min (chase), fractionated, and RNA extracted as described above. Where indicated, 1 mM DTT or 2 µg/mL harringtonine was added at the start of the 4SU pulse and included in the chase buffers. An amount of 100 µg of 4SU-labeled or control RNA was collected from cytosol and ER fractions and experiments were performed in biological triplicate. An amount of 10 ng of a Bacillus subtilis in vitro transcribed thiolated mRNA (spike-in₇₀₀) was added to each sample to correct for fractional losses and recovery during subsequent biotin labeling and isolation procedures (Neymotin et al. 2014). In vitro transcriptions were performed with 4-thioUTP/UTP admixtures using a B. subtilis sequence cloned into pSP64 (Polyansky et al. 2013) as the DNA template. The B. subtilis pSP64 (spike-in₇₀₀) construct was the kind gift of Dr. David Gresham, New York University. RNA samples were treated with DNase I, purified using GT-phenol–chloroform extraction, biotinylated, and extracted by phenol/chloroform/isoamyl alcohol, as described above.

Biotinylated RNAs were recovered by affinity purification using NanoLink streptavidin magnetic beads (Vector Laboratories), prepared as follows. Streptavidin magnetic beads (30 µL slurry per sample) were washed with 0.1 M NaCl, followed by three washes in 1× binding buffer (5 mM Tris, pH 7.4, 0.5 mM EDTA, 1 M NaCl). Beads were then resuspended in 2× binding buffer (10 mM Tris, pH 7.4, 1 mM EDTA, 2 M NaCl) and 1 µg tRNA (Ambion) per microliter of beads and incubated overnight at 4°C to reduce nonspecific RNA binding in subsequent steps. tRNA-blocked beads were washed three times with wash buffer (100 mM Tris, pH 7.4, 10 mM EDTA, 1 M NaCl, 0.1% Tween 20), once with 0.1 M NaCl, three times with 1× binding buffer, and resuspended in 2× binding buffer. Equal volumes of biotinylated RNA samples were added to tRNA-blocked beads (30 µL RNA sample per 30 µL blocked bead slurry in 2× binding buffer) and incubated at room temperature for 60 min protected from light. Magnetic beads, now bound with biotinylated RNA, were washed once with wash buffer preheated to 48°C and then washed six times with wash buffer at room temperature. RNA was eluted from beads by the addition of 0.1 M DTT. Biotinylated RNA elution was repeated, and eluates were combined. RNA was recovered with RNeasy MiniElute Spin columns (Qiagen) per the manufacturer's recommendation. Eluted 4SU-RNA was prepared for deep sequencing using the NEBNext Ultra II RNA Library Prep Kit for Illumina (New England Biolabs, E7770) with the NEBNext Poly(A) mRNA Magnetic Isolation Module (New England Biolabs, E7490) to de-enrich for ribosomal RNAs prior to library construction, following the manufacturer's protocol with multiplexing. Prepared cDNA libraries were sequenced using the Illumina NovaSeq6000 platform at the Duke Center for Genomic and Computational Biology Facility. As noted above, all 4SU pulse-chase studies were performed in biological triplicate. RT-qPCR analyses were performed on all replicates for marker gene distributions (GRP94, GAPDH) to validate reproducibility. Deep sequencing was performed on two of the three biological replicates.

4SU-seq data analysis

4SU-seq reads were analyzed using Kallisto (Bray et al. 2016). A Kallisto index was built from all RefSeq genes (downloaded from https://ftp.ncbi.nlm.nih.gov/refseq/H_sapiens/mRNA_Prot/ on 04.06.2018). B. subtilis spike-in₇₀₀ sequence was obtained from Dr. David Gresham, New York University and is described in Neymotin et al. (2014). Illumina read files were mapped to the RefSeq index and B. subtilis spike-in sequences at default settings (single-end mapping; parameters of -1 300 -s20). Mapped reads were transformed to TPM, including only annotated mRNAs (NM prefix) and B. subtilis spike-in₇₀₀ sequence and summed, to generate a TPM/gene metric. TPM values were then normalized to B. subtilis spike-in₇₀₀ abundance. Background corrections, representing non-4SU labeled poly(A) mRNA recovered in paired (control) pulse-chase studies where 4SU was omitted, were performed by subtracting TPM read metrics generated from the paired (−) 4SU background control libraries, processed as above and B. subtilis spike-in₇₀₀ normalized. All code used to analyze the 4SU-seq data sets is available on request. RNA-seq and 4SU-seq raw and processed read files are publicly available on GEO at record GSE164330.

Quantification and statistical analysis

Quantification of streptavidin signals from 4SU/biotin-labeled RNA gels was performed with Fiji/ImageJ (Schindelin et al. 2012). Statistical analyses, that is, mean, median, standard error, and standard deviation, were performed using Microsoft Excel.