Characterization of RVFV nucleocapsid protein binding sites on RNA by iCLIP-seq

Abstract

The nucleocapsid protein (N) in Rift Valley fever virus is an RNA-binding protein that functions in viral transcription, replication, and packaging. In this chapter, the method for studying protein-RNA interactions in context of viral infection using individual nucleotide resolution, crosslinking, immunoprecipitation, and sequencing (iCLIP-seq) is explained. The method is useful for identifying the interactions between both host and viral RNAs with N, and can identify RNA motifs that interact with the protein of interest.

1. Introduction

The nucleocapsid protein (N) of Rift Valley fever virus (RVFV) encapsidates its single-stranded RNA genome and is involved in transcription and replication of viral RNA. Because of the multi-functionality of the protein, understanding the interactions between N and RNAs are of interest in developing antiviral strategies. In this chapter, a high-throughput sequencing method that can be used to characterize in-cell RNA-protein interactions is discussed.

The individual nucleotide-resolution crosslinking, immunoprecipitation, and sequencing (iCLIP-seq) was first described in study of pre-mRNA splicing regulation [1,2]. The method uses the next generation sequencing to elucidate RNA binding partners for the protein of interest and targeted sequence motifs. Similar to studying other RNA binding proteins (RBPs) using this method, iCLIP-seq follows six basic steps: (1) ultraviolet (UV) crosslinking of RNA and proteins in cell culture, (2) lysis and RNase treatment, (3) immunoprecipitation (IP) of the RNA-protein complex, (4) size selection, (5) preparation of complementary DNA (cDNA) from immunoprecipitated RNA using reverse transcriptase, and (6) sequencing library preparation by PCR (Figure 1). In the first step, the protein of interest and the bound RNA are covalently linked by irradiation with UV light. The cells are lysed and treated with RNase, so that each ribonucleoprotein complex (RNP) of interest includes RNAs with optimal length for subsequent library preparation. During the third IP step, the crosslinked and partially digested RNPs are enriched by using antibody-conjugated magnetic beads. The greatest benefit in crosslinking RNA and protein is that it enables more stringent wash conditions (i.e. high salt, with detergent/denaturant, etc.). Moreover, particularly in iCLIP and related individual nucleotide resolution CLIP-seq methods, the crosslinked site (= protein binding site) is deduced later during data analysis.

Figure 1: Workflow for iCLIP-seq of viral protein.

To prepare iCLIP sequencing library, cells expressing the viral protein of interest are irradiated by UV to form covalent linkages between the protein and RNA. Following lysis and RNase digestion, RNPs are immunoprecipitated (IP) using antibody-conjugated magnetic beads. Then, RNA fragments are ligated with the 3’-adaptor and end-labeled while still bound to the beads. The labeled RNP is fractionated by denaturing gel electrophoresis. After transferring RNP to the nitrocellulose membrane, the region just above the over-digested band (marked with dashed rectangle) is excised together with the size-matched (SM) control. Proteinase K digestion releases RNA-peptide conjugate from the membrane. Using a primer containing the sequence complementary to the adaptor, the reverse transcriptase synthesizes cDNA. The extension of the primer, however, is terminated at the crosslink site (shown as a pin) due to the peptide hindering the activity of the reverse transcriptase. cDNA is then circularized and annealed to a short oligonucleotide, creating a double-stranded site that is a substrate for the endonuclease, BamHI. Digestion by BamHI results in linearized cDNA that is then PCR amplified and analyzed by Illumina sequencing. (Figure created with BioRender)

Many variations of CLIP-seq (iCLIP-seq being one of them) have been elaborated since the development of original CLIP-seq [3–5]. Most methods are comparable up to the IP step. After IP, however, library preparation methods vary from protocol to protocol. In this chapter, the widely used, radioactivity-based visualization of RNPs, use of size-matched (SM) control, and circularized cDNA preparation for subsequent PCR are explained. Because availability of equipment and regulatory restrictions vary lab-to-lab, it should be noted that modifications can be made to this protocol. For example, the use of radioactive materials can be avoided by using a fluorescently labeled adaptor [2,6,7]. Although less flexible, sequencing library preparation kits (in this case, use of a directional kit that can distinguish sense and anti-sense transcript is recommended) are commercially available. The iCLIP-seq library preparation procedure can be modified to better suit the need.

Post-IP, RNAs are tagged with adaptor and labeled with radioactive phosphate for visualization for the subsequent size selection step. The 3’ adaptor includes a primer binding site for cDNA synthesis and is ligated to the partially digested RNA by T4 RNA ligase. Then, the 5’ end of RNA is labeled with radioactive phosphate, which is used to visualize RNP on the membrane.

The fourth step is the size selection. This step is important to select the optimally sized RNPs as well as to prepare the SM control. Although different sequencing controls can be used in CLIP-seq experiments, the SM control is easy to prepare and accounts for false positive signals that arise from non-specific binding of RNAs and RNPs to the beads [8,9]. To prepare SM control, a small portion of pre-IP lysate is set aside and run in a gel lane next to the experimental sample. Then, the region that corresponds to the excised membrane in the experimental lane, is also excised in the SM control lane. The control follows the same steps as the experimental sample to be sequenced at the end. The sequencing reads from SM control library are subtracted at the stage of bioinformatic analysis. It is recommended that minimally three replicate libraries and one SM library are prepared for each experimental condition.

In the fifth step, cDNA that contains information on the protein binding site is prepared. After excising the membrane containing size selected RNPs, the proteinase K treatment release RNAs from the membrane, leaving a peptide still attached to the RNA fragment. The reverse transcriptase extends the primer to synthesize cDNA. The residual peptide on the template RNA, however, causes reverse transcriptase to halt the cDNA synthesis, creating a mixture of fully extended product and truncated product. The 3’ end created in the truncated product serves as a footprint, where peptide was originally conjugated (= the protein-RNA interaction site). This information is conserved by attaching 5’ adaptor through the circularization of cDNA, and will be analyzed later.

The final step is to prepare double-stranded DNA (dsDNA; i.e. sequencing library) to be analyzed by next generation sequencer. After circularization of cDNA, the cut oligo is annealed between the 5’ and 3’ adaptor sequence to create partially dsDNA that is digested by addition of an endonuclease. The linearized template is then subjected to two rounds of PCR. During the pilot PCR, PCR products are taken out after every 5 cycles. These samples are then analyzed by gel electrophoresis to determine the optimal number of PCR cycles that produces enough material without overamplification of the product (Figure 2). Given the result of pilot PCR, the template is now amplified to produce a sequencing library. Finally, the library is purified, then concentration and size are analyzed by using Qubit assay and TapeStation, respectively. The resulting library can be analyzed by Illumina sequencing in-house or by using commercially available services.

Figure 2: PCR amplification of linearized cDNA.

During the pilot PCR step, the optimal number of PCR cycles is determined. Insufficiently amplified sample does not contain enough DNA to be sequenced (lane 1). On the other hand, overamplified sample exhibits high molecular weight bands (lanes 3 and 4). To maintain library complexity, overamplification of the sample should be avoided.

iCLIP-seq is a powerful tool to analyze native RNA-viral protein interactions in cell culture. This chapter focuses on preparation of next generation sequencing library with iCLIP for viral RNA-binding protein. Various pipelines are available to analyze and visualize iCLIP sequencing data [10–14]. Combining high quality sample preparation and through data analysis, iCLIP-seq will provide a broad view of RNA-viral protein interactions at single nucleotide resolution.

2. Materials

Bake all glassware at 225°C for 48 hours to prevent RNase contamination. Prepare all solutions using nuclease-free ultrapure water. Filter all buffers using 0.22 μm membrane if the buffer is to be used in incubation with RNAs and DNAs.

2.1. Cells and UV crosslinking

Reagents described below are tailored for culturing HEK293 cells. Adjust growth and infection media appropriately to suit cells being used. Although it is not a focus of this chapter, modified nucleotides, such as 4-thiouridine, can be used in the cell culture to enable the use of >310 nm UV for crosslinking. The procedure, known as photoactivated ribonucleotide-enhanced CLIP (PAR-CLIP) [15], allows mapping of crosslinking sites at single nucleotide resolution like iCLIP, but follows different procedures, including during crosslinking and data analysis. Thus, this section focuses on iCLIP that does not require addition of modified nucleotides in the cell culture media.

1. Complete media: Dulbecco’s modified Eagles media (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS) and antibiotics.

2. Infection media: DMEM supplemented with 2% (v/v) FBS. This medium does not contain antibiotics.

3. Phosphate buffered saline (PBS).

4. UV light source for crosslinking.

2.2. Lysate preparation

1. Lysis buffer: PBS (or tris buffered saline; TBS) containing 0.1% (w/v) sodium dodecyl sulphate (SDS), 0.5% (w/v) sodium deoxycholate, 0.5% (v/v) IGEPAL, and mammalian protease inhibitor cocktail. Prepare protease inhibitors on the day of the experiment and chill to 4°C before use.

2. DNase RQ1 (Promega).

3. RNase A, T1, or A/T1 mix.

4. 4x Gel loading buffer: 988 mM Tris-HCl pH 8.5, 8% (w/v) lithium dodecyl sulfate, 40% (v/v) glycerol, 2 mM EDTA, 0.88% (w/v) Coomassie Brilliant Blue G250, 0.7 mM phenol red.

2.3. Immunoprecipitation

Salt and detergent concentrations for buffers below may require adjustments depending on the antibody being used for the immunoprecipitation (IP). If IP results in high background, increase salt and/or detergent concentration.

1. Protein A/G magnetic beads (see Note 1).

2. Magnetic stand.

3. Anti-RVFV nucleocapsid protein antibody (see Note 2).

4. TBS (for the Wash buffer below): 50 mM Tris-HCl pH 7.5 and 150 mM NaCl.

5. Wash buffer: TBS with 0.05% (v/v) Tween-20.

6. High-salt wash buffer: 50 mM Tris-HCl pH 7.4, 1 M NaCl, 1 mM EDTA, 1% (v/v) IGEPAL, 0.5% (w/v) sodium deoxycholate, 0.1% (w/v) sodium dodecyl sulfate.

7. Antarctic phosphatase (or equivalent with 3’-dephospholyase activity, such as polynucleotide kinase).

8. Dephosphorylation buffer: preparation of the buffer depends on the phosphatase being used in the experiment. Here, Antarctic phosphatase from New England Biolabs was used in 1x Antarctic phosphatase buffer supplied by the manufacturer.

2.4. Preparation of 3’ adaptor

Prior to cDNA synthesis, an adaptor needs to be ligated to the 3’-end of RNA crosslinked to the protein on interest. The adaptor should be pre-adenylated at the 5’-end to enable ligation reaction, and its 3’-end must be protected to prevent self-circularization. This adaptor can be purchased pre-adenylated or the adenylation reaction can be done in-house using a 5’-phosphorylated synthetic oligonucleotide and enzyme, such as Mth RNA ligase.

1. L3 adaptor: purchase 5’-(rApp)AGATCGGAAGAGCGGTTCAG(ddC)-3’, or purchase 5’-(Phos)AGATCGGAAGAGCGGTTCAG(ddC)-3’ and adenylate in house.

2. Adenosine 5’-triphosphate and Mth RNA ligase, which can be purchased as a kit (5’ DNA Adenylation Kit; New England BioLabs).

3. DNA Clean & Concentrator – 5 (Zymo Research).

2.5. Adaptor ligation

1. T4 RNA ligase 1.

2. Ligase buffer: 1x buffer provided with T4 RNA ligase 1 supplemented with RNase inhibitor.

3. Pre-adenylated L3 adaptor (see section 2.4).

4. Polyethylene glycol 400 (PEG400).

2.6. 5’ end labeling

1. T4 polynucleotide kinase.

2. Gamma-³²P-ATP (radioactive ATP is not necessary if fluorescently labeled 3’ adaptor is used; see Note 3).

3. PNK buffer: 50 mM Tris-HCl pH 7.5, and 10 mM MgCl₂.

4. 1x Gel loading buffer: dilute 4x gel loading buffer 4-fold in water.

2.7. Size selection of immunoprecipitated N-RNA complexes

1. 10% acrylamide bis-tris gel (see Note 4).

2. Gel running buffer: 50 mM MOPS, 50 mM tris base, 0.1% (w/v) SDS, and 1 mM EDTA. Do not adjust pH.

3. Nitrocellulose membrane (see Note 5).

4. Transfer buffer: 25 mM bicine, 25 mM bis-tris (free base), 1 mM EDTA, and 10% methanol. Do not adjust pH.

5. Phosphorimager.

6. Proteinase K.

7. PK buffer: 200 mM Tris-HCl pH 7.5, 100 mM NaCl, 20 mM EDTA, and 2% (w/v) SDS (see Note 6).

8. Phenol:chloroform:isoamyl alcohol (25:24:1, v/v).

9. 200 proof ethanol.

10. GlycoBlue.

11. 3 M sodium acetate pH 5.5.

12. Ice-cold 80% ethanol.

13. Qubit microRNA assay (optional; Invitrogen).

2.8. Reverse transcription

1. RT primer: 5’-(Phos)WWWNNNATCACGNNNNNAGATCGGAAGAGCGTCGTggatccTGAACCGC, where bold letters represent the TruSeq barcode sequence, and lowercase letters indicate the BamHI site (see Note 7).

2. 10 mM dNTPs.

3. Superscript IV (Invitrogen; see Note 8).

4. 100 mM DTT.

5. RNase inhibitor.

6. 1 M NaOH.

7. 1 M HEPES-NaOH pH 7.3.

8. AMPure XP reagent (Beckman Coulter; see Note 9)

2.9. 5’ adaptor ligation

1. Circligase II (Biosearch Technologies).

2. 10x Circligase II reaction buffer (Biosearch Technologies).

3. 50 mM MnCl₂.

4. Cut oligo: 5’-GTTCAggatccACGACGCTCTTCAAAA, where lowercase letters indicate the BamHI site.

5. BamHI-HF (New England BioLabs).

6. 10x rCutsmart Buffer (New England BioLabs).

2.10. Pilot PCR

1. P5 primer: 5’-AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT.

2. P3 primer: 5’-CAAGCAGAAGACGGCATACGAGATCGGTCTCGGCATTCCTGCTGAACCGCTCTTCCGATCT.

3. Q5 Hot Start High-Fidelity 2x Master Mix (New England BioLabs)

4. 1x TBE gel running buffer: dilute from 10x TBE (1 M tris-base, 1 M boric acid, and 20 mM EDTA).

5. 6% acrylamide TBE gel (see Note 10).

6. 6x DNA loading buffer: 6x TBE, 30% (v/v) glycerol, 0.25% (w/v) bromophenol blue, 0.25% (w/v) xylene cyanol.

7. SYBR Green I.

2.11. Preparative PCR and quality control

1. P5 and P3 primer.

2. Q5 Hot Start High-Fidelity 2x Master Mix (New England BioLabs)

3. Qubit assay dsDNA HS assay kit (Invitrogen)

4. Qubit assay tubes (Invitrogen)

5. Qubit fluorometer (Invitrogen)

6. TapeStation (Agilent)

7. TapeStation D1000 screen tape and reagents (Agilent)

3. Methods

3.1. Cells and UV crosslinking

1. Prepare approximately 10⁷ infected cells (see Note 11) in 150-mm dish and wash twice with 10 mL cold PBS.

2. Add 5 mL cold PBS onto cells. With lid of the dish open and cells placed on ice, UV crosslink protein and RNA for a total energy of 0.1 to 0.4J/cm² (see Note 12).

3. Scrape cells with cell scraper and transfer to a clean tube.

4. Pellet cells by centrifugation. Spin 10 min at 1000 g in cold centrifuge.

5. Discard supernatant and freeze pellet with liquid nitrogen. The pellet can be stored frozen at −80°C.

3.2. Lysate preparation

1. Thaw cell pellet on ice and resuspend with 350 μL lysis buffer by pipetting up and down.

2. Incubate on ice for 10 minutes to lyse cells.

3. Add 10 μL DNase RQ1 and 3.6 μL diluted RNase, then incubate 5 minutes at 37°C. The optimal dilution (generally 1:100 ~ 1:10,000) of the RNase needs to be determined experimentally (see Note 12).

4. Centrifuge lysate for 20 minutes at top speed (~20,000 g with a typical tabletop centrifuge), 4°C.

5. Transfer supernatant to a clean tube. Aliquot ~5% (~20 uL) lysate into a separate tube, which will be used as size-matched control (see Note 13). Add 7 μL 4x gel loading buffer to 20 μL lysate (this sample can be stored frozen at −80°C).

6. Adjust volume of the remaining lysate to 1 mL with lysis buffer and proceed to immunoprecipitation.

3.3. Immunoprecipitation

1. Vortex protein A/G magnetic beads to fully resuspend and aliquot 50 μL into a microcentrifuge tube.

2. Add 350 μL wash buffer to beads and vortex to mix.

3. Collect magnetic beads by placing tube on a magnetic stand. Carefully discard buffer by pipetting.

4. Add 1 mL wash buffer, vortex beads to resuspend, then rotate 1 minutes at room temperature.

5. Collect beads by placing tube on a magnetic stand and discard supernatant.

6. Dilute 20 μg of anti-N antibody to total volume of 400 μL in wash buffer. Add diluted antibody to pre-washed beads (see Note 14).

7. Let the antibody binds to beads by incubating 1 hour at room temperature. During incubation, rotate the tube end-over-end to ensure sufficient mixing.

8. Collect beads with a magnetic stand and remove supernatant.

9. Wash beads twice with 500 μL wash buffer. Gently resuspend beads by flicking tube each time, collect beads, and discard supernatant.

10. Add 1 mL lysate to antibody-bound beads and incubate overnight at 4°C. During incubation, rotate the tube end-over-end to ensure sufficient mixing (see Note 15).

11. Wash beads twice with 1 mL high-salt wash buffer. Gently resuspend beads by flicking tube each time, collect beads, and discard supernatant.

12. Wash beads once with 500 μL dephosphorylation buffer.

13. Add 500 μL dephosphorylation buffer to beads together with 5 μL Antarctic phosphatase, then incubate 30 min at 37°C with agitation.

14. Wash beads with 500 μL dephosphorylation buffer and high-salt wash buffer.

3.4. Preparation of 3’ adaptor

This step is not required if commercially synthesized, 5’ adenylated L3 adaptor is purchased. The protocol below explains steps to adenylate 5’ phosphorylated L3 adaptor.

1. Mix 1 nmol 5’-phosphorylated L3 adaptor with 20 μL 1 mM ATP, 20 μL 10x adenylation buffer, and 20 μL Mth RNA ligase to make total volume of 200 μL.

2. Incubate 1 hour at 37°C.

3. Purify adenylated adaptor using DNA Clean & Concentrator kit and elute in 20 μL water.

4. Determine adaptor concentration by UV-vis spectrophotometry and store aliquots at −80°C.

3.5. 3’ Adaptor ligation

1. Collect beads by placing tube on a magnetic stand and discard supernatant.

2. Resuspend beads with 30 μL ligase buffer and add 1 μL T4 RNA ligase 1, 1 μL 1 μM pre-adenylated L3 adaptor, and 6 μL PEG400.

3. Incubate reaction overnight at 16°C with agitation on a rocker or rotisserie.

4. Add 500 μL dephosphorylation buffer and collect beads on a magnetic stand.

5. Discard supernatant and wash twice with 500 μL high-salt wash buffer.

3.6. 5’ end labeling

Please note that sections below include handling of radioactive materials. Discard solid and liquid waste appropriately, and use of filtered pipette tips is recommended.

1. Collect beads and discard supernatant.

2. Resuspend beads in 100 μL PNK buffer and add 2 μL gamma-³²P-ATP and 2 μL T4 polynucleotide kinase.

3. Incubate 30 min at 37°C with agitation.

4. Wash beads twice with 500 μL high-salt wash buffer.

5. Discard supernatant and resuspend beads with 25 μL 1x gel loading buffer.

6. Store eluate at −80°C or directly proceed to the size selection step.

3.7. Size selection of immunoprecipitated N-RNA complexes

1. Denature immunoprecipitated (IP) sample and corresponding size-matched sample by heating 10 min at 70°C.

2. Load 25 μL denatured sample per lane of 10% acrylamide bis-tris gel. Electrophorese in gel running buffer at 200 V until the dye reaches the bottom of the gel.

3. Electro-transfer N-RNA complexes to a nitrocellulose membrane in transfer buffer for 1 hour at 30 V, room temperature. Use a minigel wet-transfer apparatus.

4. Carefully disassemble the transfer apparatus and wrap the nitrocellulose membrane in a plastic film. Place imaging plate onto a wrapped membrane and let the image develop for 10 ~ 30 minutes.

5. Visualize radiolabeled N-RNA complexes with phosphorimager. Print out the image in the actual size and cut out the target region. By superimposing the pre-cut printout onto the membrane, cut out the part of membrane that contains N-RNA complex (see Note 16). Chop membrane into 2 mm × 2 mm pieces and transfer into a microcentrifuge tube.

6. To elute N-RNA complexes, add 200 μL PK buffer containing 4 mg/mL proteinase K to the membrane pieces and incubate 30 minutes at 55°C with agitation.

7. Add an equal volume (200 μL) of phenol:chloroform:isoamyl alcohol to the proteinase K digested sample and vortex well. Then, centrifuge 20 minutes at 20,000g, 4°C.

8. Carefully transfer aqueous supernatant to a clean tube.

9. Add 0.5 μL GlycoBlue to aid precipitation and 30 μL 3 M sodium acetate pH 5.5.

10. Add 1 mL ethanol and chill sample at −80°C for 30 minutes or −20°C overnight.

11. Centrifuge at 20,000 g for 20 minutes at 4°C in a microcentrifuge.

12. Carefully remove supernatant with a pipette. The presence of RNA can be confirmed by testing radioactivity in the tube using a Geiger-Muller counter.

13. Wash pellet with 500 μL ice-cold 80% ethanol and centrifuge again.

14. Discard supernatant and air-dry pellet for 10 minutes by leaving the tube on a benchtop with lid open.

15. Resuspend sample in 10 μL water (optional: estimate RNA concentration by Qubit microRNA assay).

16. Directly proceed to library preparation or store RNA at −80°C.

3.8. Reverse transcription

1. In a PCR tube, mix 8 μL RNA, 1 μL 2 μM RT primer, and 1 μL each 10 mM dNTP.

2. Heat 10 min at 70°C, then cool to 4°C to anneal RNA and RT primer.

3. Add 7 μL RT premix (4 μL 5x SSIV buffer, 1 μL 100 mM DTT, 1 μL RNaseOUT, and 1 μL SuperScriptIV) and mix thoroughly.

4. In a thermocycler, incubate 1 hour at 55°C, then 10 min at 80°C.

5. Add 2 μL 1 M NaOH, mix, and incubate 20 min at 98°C.

6. Add 20 μL 1 M HEPES-NaOH pH 7.3 and mix.

7. Add 7 μL isopropanol and 72 μL AMPure XP beads.

8. Incubate 10 minutes at room temperature, then collect beads by placing tube on a magnetic stand.

9. Discard supernatant and wash beads with 100 μL 80% ethanol.

10. Air-dry beads for 5 minutes and resuspend in 16 μL water.

11. Transfer cDNA to a clean tube.

3.9. 5’ adaptor ligation

1. To the 16 μL cDNA, add 2 μL 10x Circligase II reaction buffer, 1 μL 50 mM MnCl₂, 1 μL Circligase II.

2. Incubate cDNA 90 min at 60°C.

3. Add 22 μL water, 5 μL 10x rCutSmart buffer, and 1 μL 10 μM cut oligo.

4. Anneal cut oligo to cDNA by incubating 2 min at 95°C, then slowly cooling down to room temperature.

5. Add 2 μL BamHI-HF and incubate 1 hour at 37°C, then heat-inactivate 20 min at 80°C.

6. Add 8 μL isopropanol and 86 μL AMPure XP beads (see Note 17).

7. Incubate 10 minutes at room temperature, then collect beads by placing tube on a magnetic stand.

8. Discard supernatant and wash beads with 100 μL 80% ethanol.

9. Air-dry beads for 5 minutes and resuspend in 20 μL water.

10. Transfer linearized cDNA to a clean tube (see Note 18).

3.10. Pilot PCR

1. Mix 10 μL linearized cDNA, 12.5 μL Q5 Hot start high-fidelity 2x master mix, and 1.25 μL each of 10 μL P5 and P3 primer.

2. Run PCR as follows: 98°C for 30 seconds, then 30 cycles of 98°C 10 seconds, 63°C 30 seconds, and 72°C 20 seconds. Take 5 μL samples every 5 cycles starting at cycle number 10 (i.e. total 5 samples taken at cycles 10, 15, 20, 25, and 30).

3. Add 1 μL 6x DNA loading buffer to each sample and analyze PCR product on 6% TBE gel.

4. Stain gel with SYBR Green I and determine the lowest number of PCR cycles needed to obtain product without appearance of over-amplified, high molecular weight bands (see Note 19).

3.11. Preparative PCR and quality control

The sequencing library, when prepared and subjected to quality control steps below, can be submitted to sequencing by Illumina platform. In this section, preparation of the library for single-ended 50 bp read (recommended) is explained, as it is most likely sufficient and the most cost-effective option. Some viruses, however, include highly repetitive sequence that may be difficult to be mapped if shorter read length (i.e. 50 bp) was used. If this is a concern, longer, paired-end sequencing can be done by modifying the protocol at size selection and purification steps (see Note 20).

1. Based on the number of PCR cycles determined in the section 3.10, perform PCR amplification again using the same conditions.

2. To the crude library, add 4 μL isopropanol and 45 μL AMPure XP beads.

3. Incubate 10 minutes at room temperature, then collect beads by placing tube on a magnetic stand.

4. Discard supernatant and wash beads with 100 μL 80% ethanol.

5. Air-dry beads for 5 minutes and resuspend in 20 μL water.

6. Determine library concentration by Qubit assay.

7. Confirm DNA size distribution by TapeStation.

8. At this point, the library is ready to be sequenced using the Illumina platform.

4. Notes

1. Magnetic beads covalently conjugated with protein A, G, and A/G are available from multiple manufacturers. Protein A and G exhibits different affinity towards antibodies raised in various host species [16]. For example, protein G binds better to mouse IgG than rabbit IgG. Immunoprecipitation is a critical step in iCLIP procedure and is important to have an antibody with high affinity and specificity. In this protocol, beads conjugated with both protein A and G are used since these beads enable screening of antibodies from different host species.

2. The quality of the sequencing library and the final sequencing results greatly depend on the immunoprecipitation, since the underlying assumption in this procedure is that the protein of interest is enriched in the pulldown sample with minimal contaminating proteins. To produce high quality library, testing antibody specificity prior to the experiment is crucial. If using commercially available antibody, sharing manufacturer name, catalog number, and lot number along with the sequencing result is helpful for future studies and for reproducibility.

3. Use of radioactive materials may be challenging in some labs due to safety and regulatory concerns. A few groups have developed modified CLIP-seq protocols to use fluorescence-based detection rather than radioactivity. In these protocols, the 3’-adaptor, instead of the 5’-end of the digested RNA in the radioactive protocol, is labeled with fluorophore (e.g. IRD800) [7]. The use of a fluorescent adaptor may cost more initially due to custom synthesis of the modified DNA as well as labor intensive purification. The fluorescence has longer shelf life than ³²P and may better suit some laboratory settings. Detailed protocols using fluorescently labeled adaptor are described by different groups [2,6,7].

4. Tris-glycine gel, often used to separate proteins in gel electrophoresis, operates at a higher pH that risks degradation of RNA by alkaline hydrolysis. On the other hand, bis-tris gel electrophoresis operates at neutral pH. Therefore, to maintain RNA integrity during electrophoresis, the use of bis-tris gel is recommended.

5. Transferring RNA-protein complexes onto a membrane is an important step in size selection. Use of a nitrocellulose membrane is recommended over PVDF during this procedure to help minimize binding of free RNA.

6. Proteinase K buffer often includes denaturant, such as urea and SDS, to increase the digestion efficiency. SDS, however, is more stable than urea and is recommended.

7. Incorporation of 6-nucleotide barcode sequence allows multiplexing of sequencing libraries. It is recommended to use a different barcode for each control and for replicates. Other TruSeq barcodes (CGATGT, TTAGGC, TGACCA, ACAGTG, GCCAAT, CAGATC, ACTTGA, GATCAG, TAGCTT, GGCTAC, CTTGTA, AGTCAA, AGTTCC, ATGTCA, CCGTCC, GTCCGC, GTGAAA, GTGGCC, GTTTCG, CGTACG, GAGTGG, ACTGAT, ATTCCT) can also be used.

8. Various types of reverse transcriptase are available from different manufacturers. Other group II intron-encoded reverse transcriptases, such as TGIRT-III (ingex) and Induro (New England BioLabs), can also be used. If switching the enzymes, adjust reaction buffer according to manufacturer’s instructions.

9. Using magnetic beads for primer/free-nucleotide removal results in higher product recovery, although other clean-up methods like spin column-based and gel purification can be used. If preparing libraries with longer read lengths, the bead volume can be adjusted.

10. The mini gel can be purchased or poured in-house by mixing 1 mL 1x TBE, 1.5 mL 40% acrylamide, 7.5 mL water, 100 μL 10% (w/v) ammonium persulfate (APS), and 10 μL tetramethylethylenediamine (TEMED).

11. Immunoprecipitating viral protein may pose challenges since the protein of interest may be expressed at different levels throughout the course of infection. To achieve optimal immunoprecipitation result, the expression levels for the protein of interest should be characterized prior to the experiment. This includes optimization in multiplicity of infection (MOI), timing for harvesting cells post-infection, and number of cells to be used in the pulldown protocol. Because UV crosslinking efficiency is low, it is expected that only a fraction of the protein of interest gets covalently linked to RNA. To obtain sufficient material for sequencing library preparation, the amount of input protein may need to be adjusted.

12. Commercially available UV crosslinkers have various settings and options for wavelength. In this protocol, 254 nm UV is used to crosslink RNA and protein, since no specialty photoactivatable modified nucleotide, like 4-thiouridine, is used. In-cell UV crosslinking efficiency is about 1~3% [17] and is affected by other factors, such as UV flux and distance from the lamp. Successful crosslinking coupled with optimal RNase digestion should produce a smeared signal when visualized on the membrane. Optimal UV exposure together with RNase dilution must be determined experimentally prior to the sequencing library preparation. Moreover, if preparing sequencing library with longer read length (e.g 150 bp paired-end), RNase concentration needs to be adjusted accordingly to produce longer fragments (also see section3.7 on size selection and 3.11 on preparative PCR).

13. The size-matched control is a control sequencing library. Here, the pre-IP input sample is aliquoted and saved so that it will be loaded onto the gel lane right next to the post-IP sample (Figure 1). The control library should be prepared parallel to the experimental library. When analyzing the sequencing data, peaks and hits generated from the control library will be subtracted from the experimental library.

14. The binding capacity of beads differ, thus refer to manufacturer’s instruction and use optimal antibody-to-bead ratio for the preparation.

15. Conditions for immunoprecipitation (e.g. buffer, temperature, and duration) must be optimized prior to the experiment to produce minimum possible background and sufficient enrichment of the RNA-protein complex of interest.

16. The protein of interest covalently linked to a longer RNA fragment tends to run at higher molecular weight compared to the one with shorter RNA fragment. If the library is to be sequenced as single-end 50 bp, the RNA fragment length can be 20~50 nucleotide long to be meaningful. If the library is for paired-end 150 bp sequencing, the RNA length of up to 300 nucleotides can be read. Optimize the section of the membrane to be excised to produce sequencing library with desired insert length.

17. 1.8x volume of AMPure beads is used to purify the PCR product in this section. The volume ratio of the bead can be modulated to selectively capture higher molecular weight dsDNA if needed.

18. At this point, each cDNA carries an index that is unique per library. Therefore, these cDNAs can be pooled together if desired, before proceeding to the PCR amplification.

19. Overamplification of the library creates PCR artifacts that are visible on a gel as higher molecular weight species. To avoid overamplification, the number of PCR cycles to be used in the library preparation should be optimized. The ideal size for the library would be the size of adaptor (136 bp) plus the length of RNA fragment.

20. Paired-end sequencing allows production of longer RNA fragments to be sequenced, which may help during bioinformatics analysis. For example, if the protein of interest is expected to bind near a splicing junction, using longer read length may be helpful. However, choosing paired-end sequencing may require another flow cell/lane so that sufficient sequencing depth can be achieved, which can increase the sequencing cost.