Transcriptome-Wide Profiling of Protein-RNA Interactions by Cross-Linking and Immunoprecipitation Mediated by FLAG-Biotin Tandem Purification

Summary

Here we present a modified CLIP-seq protocol called ^FbioCLIP-seq with FLAG-biotin tandem purification to determine the RNA targets of RNA-binding proteins (RBPs) in mammalian cells.

Abstract

RNA and RNA-binding proteins (RBPs) control multiple biological processes. The spatial and temporal arrangement of RNAs and RBPs underlies the delicate regulation of these processes. A strategy called CLIP-seq (cross-linking and immunoprecipitation) has been developed to capture endogenous protein-RNA interactions with UV cross-linking followed by immunoprecipitation. Despite the wide use of conventional CLIP-seq method in RBP study, the CLIP method is limited by the availability of high-quality antibodies, potential contaminants from the copurified RBPs, requirement of isotope manipulation, and potential loss of information during a tedious experimental procedure. Here we describe a modified CLIP-seq method called ^FbioCLIP-seq using the FLAG-biotin tag tandem purification. Through tandem purification and stringent wash conditions, almost all the interacting RNA-binding proteins are removed. Thus, the RNAs interacting indirectly mediated by these copurified RBPs are also reduced. Our ^FbioCLIP-seq method allows efficient detection of direct protein-bound RNAs without SDS-PAGE and membrane transfer procedures in an isotope-free and protein-specific antibody-free manner.

Introduction

RNAs and RNA-binding proteins (RBPs) control diverse cellular processes including splicing, translation, ribosome biogenesis, epigenetic regulation, and cell fate transition^1,2,3,4,5,6. The delicate mechanisms of these processes depend on the unique spatial and temporal arrangement of RNAs and RBPs. Therefore, an important step towards understanding RNA regulation at the molecular level is to reveal the positional information about the binding sites of RBPs.

A strategy referred to as cross-linking and immunoprecipitation (CLIP-seq) has been developed to capture protein-RNA interactions with UV cross-linking followed by immunoprecipitation of the protein of interest⁷. The key feature of the methodology is the induction of covalent cross-links between an RNA-binding protein and its directly bound RNA molecules (within ~1 Å) by UV irradiation⁸. The RBP footprints can be determined by CLIP tag clustering and peak calling, which usually have a resolution of 30−60 nt. Alternatively, the reverse transcription step of CLIP can lead to indels (insertions or deletions) or substitutions to the cross-linking sites, which allows identification of protein binding sites on the RNAs at a single-nucleotide resolution. Pipelines like Novoalign and CIMS have been developed for the analysis of the high-throughput sequencing results of CLIP-seq⁸. Several modified CLIP-seq methods have also been proposed, including individual-nucleotide resolution cross-linking and immunoprecipitation (iCLIP), enhanced CLIP (eCLIP), irCLIP, and photoactivatable ribonucleoside-enhanced cross-linking and immunoprecipitation (PAR-CLIP)^9,10,11,12.

Despite the wide use of traditional CLIP-seq methods in the study of RBPs, the CLIP methods have several drawbacks. First, the tedious denatured gel electrophoresis and membrane transfer procedure may lead to loss of information, and cause limited sequence complexity. Second, the protein specific antibody-based CLIP method may pull down a protein complex instead of a single target protein, which may lead to false positive protein-RNA interactions from the copurified RBPs. Third, the antibody-based strategy requires a large amount of high-quality antibodies, which makes the application of these methods inadequate for the study of RBPs without high-quality antibodies available. Fourth, the traditional CLIP method requires radiolabeled ATP to label the protein-bound RNAs.

The high affinity of streptavidin to biotinylated proteins makes it a very powerful approach to purify specific proteins or protein complexes. The efficient biotinylation of proteins carrying an artificial peptide sequence by ectopically expressed bacterial BirA biotin ligase in mammalian cells makes it an efficient strategy to perform biotin purification in vivo¹³. We developed a modified CLIP-seq method called ^FbioCLIP-seq (FLAG-Biotin-mediated Cross-linking and Immunoprecipitation followed by high-throughput sequencing) using FLAG-biotin tag tandem purification¹⁴ (Figure 1). Through tandem purification and stringent wash conditions, almost all the interacting RBPs are removed (Figure 2). The stringent wash conditions also allow circumventing the SDS-PAGE and membrane transfer, which is labor intensive and technically challenging. And similar to eCLIP and irCLIP, the ^FbioCLIP-seq method is isotope-free. Skipping the gel running and transfer steps avoids the loss of information, keeps authentic protein-RNA interactions intact, and increases the library complexity. Moreover, the high efficiency of the tagging system makes it a good choice for RBPs without high-quality antibodies available.

Here we provide a step-by-step description of the ^FbioCLIP-seq protocol for mammalian cells. Briefly, cells are cross-linked by 254 nm UV, followed by cell lysis and FLAG immunoprecipitation (FLAG-IP). Next, the protein-RNA complexes are further purified by biotin affinity capture and RNAs are fragmented by partial digestion with MNase. Then, the protein-bound RNA is dephosphorylated and ligated with a 3’ linker. A 5’ RNA linker is added after the RNA is phosphorylated with PNK and eluted by proteinase K digestion. After reverse transcription, the protein-bound RNA signals are amplified by PCR and purified by agarose gel purification. Two RBPs were chosen to exemplify the ^FbioCLIP-seq result. LIN28 is a well-characterized RNA-binding protein involved in microRNA maturation, protein translation, and cell reprogramming^15,16,17. WDR43 is a WD40 domain-containing protein thought to coordinate ribosome biogenesis, eukaryotic transcription, and embryonic stem cell pluripotency control^14,18. Consistent with previously reported results for LIN28 with CLIP-seq, ^FbioCLIP-seq reveals binding sites of LIN28 on “GGAG” motifs in the microRNA mir-let7g and mRNAs^16,19 (Figure 3). WDR43 ^FbioCLIP-seq also identified the binding preference of WDR43 with 5' external transcribed spacers (5'-ETS) of pre-rRNAs²⁰ (Figure 4). These results validate the reliability of the ^FbioCLIP-seq method.

Protocol

1. Cell line construction

1. Clone the gene of interest into a PiggyBac vector pPiggyBac-FLAG-bio-[cDNA of interest]-(Hygromycin-resistant) plasmid that carries a FLAG-biotin epitope^13,21 to express a FLAG-biotin tag fused RBP (^FBRBP).
2. Cotransfect the ^FBRBP expressing vector with the pBase vector into a cell line expressing the BirA enzyme²².
   NOTE: In this study, the ^FBRBP gene was transfected into mouse embryonic stem cells (mESCs) carrying an integrated BirA expression vector²¹. Alternatively, the ^FBRBP expressing vector can be cotransfected into cells with pBase and pPiggyBac-BirA-V5 together.
3. Grow the cells in mESC on gelatinized dishes and maintain in mES culture medium (Table of Materials) in 5% CO₂ at 37 °C. Select the transfected cells with hygromycin b (100 µg/mL) for 1 week.
4. After drug selection, harvest cells by SDS loading buffer (Table 1) and use a Western blot²³ to confirm the tagging and expression of the gene with a FLAG antibody and streptavidin-HRP, respectively.

2. Cross-linking

1. Plate ~3 x 10⁶ mES cells in a 10 cm plate 1 day before the experiment so that the cells will grow to 70−90% confluence when harvested (~5−10 million cells per sample).
2. Remove the medium with a vacuum. Treat the cells with 2 mL of 0.25% trypsin-EDTA for 2 min at room temperature (RT) and quench the trypsin by 4 mL of fresh mESC medium. Transfer the cells to a 15 mL centrifuge tube. Spin down the cells by centrifugation at 300 x g for 3 min at 4 °C and remove the supernatant.
3. Suspend the cells with 4 mL of cold PBS and plate the cells back to the 10 cm plate for cross-linking. Cross-link the cells by radiation with 254 nm UV light (set the UV cross-linker with a parameter of 400 mJ/cm²).
   NOTE: For cell monolayers, the cells can be cross-linked with UV light directly on the plate before trypsin treatment.
4. Transfer the cross-linked cells to a 15 mL tube. Spin down by centrifugation at 300 x g for 3 min at 4 °C and remove the supernatant.
   NOTE: The cross-linked cell pellet can be stored at -80 °C until use.

3. Cell lysate preparation

1. Lyse the cells with 500 µL of wash buffer A (Table 1) freshly supplemented with 1 mM DTT, 1 mM PMSF, 1/500 protease inhibitor cocktail, and 400 U/mL RNase inhibitor. Transfer the cells to an RNase-free 1.5 mL tube. Incubate the cells for 30−60 min on a rotor with gentle rotation at 4 °C.
2. Treat the lysate with 30 µL of DNase I at 37 °C for 10 min. Stop the reaction by adding 4 µL of 0.5 M EDTA. Spin down the insoluble pellet by 12,000 x g for 20 min at 4 °C and transfer the supernatant to a new prechilled 1.5 mL tube.
   NOTE: For immediate use, keep on ice. If the supernatant will not be used immediately, store it at -80 °C until use.

4. FLAG beads preparation

1. Add 40 µL of slurry FLAG beads per sample to a fresh 1.5 mL tube.
2. Quickly wash the beads with 0.5 mL of ice-cold wash buffer A and spin down by 3,000 x g for 2 min at 4 °C.
3. Repeat step 4.2 once. Spin down the beads with 3,000 x g for 2 min at 4 °C. Carefully remove the buffer with a narrow-end pipette tip. Avoid removing the beads.

5. Immunoprecipitation

1. Transfer the cell lysate from step 3.2 to the pre-equilibrated FLAG beads. Rotate all the samples on a roller shaker at 4 °C gently. Incubate the FLAG beads with lysate for 2−4 h or overnight.
2. Centrifuge the beads for 2 min at 3,000 x g and remove the supernatant.
3. Wash the beads with 0.5 mL of prechilled wash buffer A by incubation for 5 min in a rotator at 4 °C, spin down the beads with 3,000 x g for 2 min and remove the supernatant. Repeat the wash 1x.
4. Repeat the wash 2x as described in step 5.3 with 0.5 mL of prechilled wash buffer B (Table 1).

6. Elution with 3x FLAG peptide

1. Prepare 3x FLAG elution solution by dissolving 1 mg of 3x FLAG peptide with 5 mL of wash buffer A to a final concentration of 200 ng/mL.
2. Spin down the FLAG beads from step 5.4 with 3,000 x g for 2 min. Remove the supernatant carefully. Make sure most of the supernatant is removed by using a narrow end pipette tip.
3. Add 200 µL of 3x FLAG elution solution to each sample. Incubate the samples with gentle rotation for 30 min at 4 °C. Spin down the FLAG beads for 2 min at 3,000 x g. Transfer the supernatants to fresh tubes.
4. Repeat the elution 2x as described in steps 6.2 and 6.3 and pool the eluents together. Save 5% of the eluents for Western blot analysis or silver staining to check the efficiency of FLAG-IP.

7. Streptavidin beads preparation

1. Prepare 50 µL of a streptavidin bead slurry for each sample. Collect the beads with a magnetic stand for 30 s and remove the supernatant with a pipette tip.
2. Quickly wash the beads with 0.5 mL of ice-cold wash buffer A and collect the beads with a magnetic stand for 30 s.
3. Repeat the wash 1x. Remove the supernatant after the wash.

8. Biotin affinity purification

1. Transfer the pooled eluents from step 6.4 to the streptavidin beads from step 7.3 and gently rotate the samples at 4 °C for 1−3 h or overnight.
2. Collect the beads with a magnetic stand for 30 s and remove the supernatant. Wash the beads 2x with 500 µL of wash buffer C (Table 1) by gentle rotation at RT for 5 min.
3. Collect the beads with the magnetic stand and remove the supernatant. Wash the beads 2x with 500 µL of wash buffer D (Table 1) by rotating at RT for 5 min.
4. Collect the beads with the magnetic stand and remove the supernatant. Quickly wash the beads 2x with 500 µL of ice-cold PNK buffer (Table 1).

9. Partial RNA digestion

1. Make a 10⁵-fold dilution of MNase with MNase reaction buffer (Table 1). Add 100 µL of MNase solution to each sample. Vortex the beads at 37 °C with a thermal mixer set at 1,200 rpm (5 s run, 30 s stop) for 10 min, collect the beads with a magnetic stand for 30 s and remove the supernatant.
   NOTE: The concentration of MNase should be optimized for different RNA-binding proteins. The MNase concentration that can digest RNAs into 30−50 nt fragments (determined by the size of insert of the final library) is optimal.
2. Quickly wash the beads 2x with 500 µL of ice-cold PNK+EGTA buffer (Table 1). Collect the beads with a magnetic stand and remove the supernatant between each wash step.
3. Wash the beads 2x with 500 µL of wash buffer C by gentle rotation for 5 min at RT. Then quickly wash the beads 2x with 500 µL of ice-cold PNK buffer.

10. Dephosphorylation of RNA

1. Make a calf intestine phosphatase (CIP) reaction mix with 8 µL of 10x CIP buffer (Table of Materials), 3 µL of CIP enzyme, and 69 µL of water. The total volume is 80 µL per reaction.
2. Collect the beads with a magnetic stand and remove the buffer. Add the CIP reaction mix to the beads. Vortex the beads at 37 °C with a thermal mixer set at 1,200 rpm (5 s run, 30 s stop) for 10 min.
3. Quickly wash the beads 2x with 500 µL of ice-cold PNK+EGTA buffer. Quickly wash the beads 2x with 500 µL of ice-cold PNK buffer.

11. 3’ linker ligation

1. Make a 3’ linker mix with 4 µL of 20 µM 3’ linker, 4 µL of 10x T4 RNA ligase buffer, 4 µL of 50% PEG8000, 2 µL of T4 RNA ligase 2 (truncated), and 26 µL of water. The total volume is 40 µL per reaction.
2. Collect the beads from step 10.3 with a magnetic stand and remove the buffer. Add 40 µL of the 3’ linker ligation mix to the beads. Vortex the beads at 16 °C with a thermal mixer set at 1,200 rpm (5 s run, 30 s stop) for 3 h or overnight.
3. Quickly wash the beads 2x with 500 µL of ice-cold PNK+EGTA buffer. Quickly wash the beads 2x with 500 µL of ice-cold PNK buffer.

12. PNK treatment

1. Make a PNK mix with 4 µL of 10x PNK buffer, 2 µL of T4 PNK enzyme, 1 µL of 10 mM ATP, and 33 µL of water. The total volume is 40 µL per reaction.
2. Collect the beads with a magnetic stand and remove the buffer. Add 40 µL of PNK mix to the beads. Vortex the beads gently at 37 °C with a thermal mixer set at 1,200 rpm (5 s run, 30 s stop) for 10 min.
3. Quickly wash the beads with 500 µL of ice-cold PNK+EGTA buffer.
4. Quickly wash the beads with 500 µL of ice-cold PNK buffer. Save 5% of the beads for Western or silver staining analysis to confirm the efficiency of immunoprecipitation.

13. RNA isolation

1. Collect the beads from step 12.4 with a magnetic stand and remove the buffer. Add 200 µL of proteinase K digestion buffer (Table 1). Vortex the beads for 30 min at 37 °C with a thermal mixer set at 1,200 rpm (5 s run, 30 s stop). Magnetically isolate the beads and take the supernatant for the experiment.
2. Add 400 µL of RNA isolation reagent (Table of Materials) to the supernatant from step 13.1, mix thoroughly, and incubate on ice for 5 min. Add 80 µL of chloroform and mix thoroughly and then incubate on ice for 5 min. Spin down with 12,000 x g for 10 min 4 °C.
   NOTE: This step should be performed in a chemical hood.
3. Take the supernatant (~500 µL) and add two volumes of isopropanol:ethanol mix (1:1), 1/10 volume of 3 M sodium acetate (pH = 5.5), and 1 µL of glycogen. Freeze at -20 °C for 2 h. Centrifuge at 4 °C with 12,000 x g for 20 min.
4. Wash the pellet 2x with ice-cold 70% ethanol. Spin down the pellet at 4 °C with 12,000 x g for 5 min. Remove the supernatant and dry the pellet. Dissolve the RNAs in 5 µL of RNase-free H₂O.

14. 5’ RNA linker ligation

1. Make a 5’ RNA ligation mix with 1 µL of 10x T4 RNA ligase buffer, 0.5 µL of BSA (1 µg/µL), 1 µL of 10 mM ATP, 1 µL of RNase inhibitor, and 0.5 µL of T4 RNA ligase. The total volume is 4 µL per reaction.
2. Add 1 µL of 5’ 20 µM RNA linker to the RNA from step 13.4, denature the RNA by heating at 70 °C for 2 min, and chill on ice immediately.
3. Add the 5’ RNA ligation mix to the RNA from step 14.2. Incubate at 16 °C overnight.

15. Reverse transcription

1. Purify the RNAs as described in section 13 and dissolve the RNA with 11.5 µL of RNase-free water.
2. Add 1 µL of 10 µM reverse transcription primer to the purified RNA, heat at 70 °C for 2 min, and chill on ice immediately.
3. Make a reverse transcription mix with 4 µL of 5x reverse transcription buffer, 1 µL of 10 mM dNTPs, 1 µL of 0.1 M DTT, 1 µL of RNase inhibitor, and 0.5 µL of reverse transcriptase (Table of Materials). The total volume is 7.5 µL per reaction.
4. Add 7.5 µL of reverse transcription mix to the RNA, incubate at 50 °C for 30 min, and stop the reaction by incubation at 70 °C for 10 min. Leave at 4 °C.

16. PCR amplification

1. Make a PCR amplification mix with 15 µL of 2x PCR master mix (Table of Materials), 5 µL of cDNA from step 15.4, 0.6 µL of 10 µM forward primer 1 (Table 2), 0.6 µL of 10 µM reverse primer 1 (Table 2), and 8.8 µL of H₂O. The total volume is 30 µL.
2. Amplify the cDNA with the following settings: 98 °C 30 s, 98 °C 10 s, 60 °C 30 s, 72 °C 30 s (repeat steps 2−4 for 15−20 cycles), 72 °C 2 min, and hold at 4 °C. Run the PCR product on a 2−3% agarose gel. Cut out the DNA of ~100−150 bp, extract DNA with gel extraction kit (Table of Materials), and elute the DNA with 20 µL of water.
3. Take 3 µL of the purified PCR product, amplify with forward primer 2 (Table 2) and reverse primer 2 (index primer; Table 2) as described in step 16.2 for five cycles to introduce index sequences. Purify the PCR product as described in step 16.2.
4. Sequence the library with a high-throughput sequencing platform.

17. Bioinformatics analysis

1. Perform data analysis using commercial programs as previously described⁸.

Results

The schematic representation of the ^FbioCLIP-seq procedure is shown in Figure 1. Compared with FLAG-mediated or streptavidin-mediated one-step affinity purification, FLAG-biotin tandem purification removed almost all the copurified proteins, avoiding the contamination of indirect protein-RNA interactions (Figure 2). Representative results for ^FbioCLIP-seq for LIN28 and WDR43 are depicted in Figure 3 and

Discussion

Here we introduce a modified CLIP-seq method called ^FbioCLIP-seq, taking advantage of the FLAG-biotin double tagging system to perform tandem purification of protein-RNA complexes. The FLAG-biotin double tagging system has been shown to be powerful in identifying protein-protein and protein-DNA interactions^13,21. Here we demonstrate the high specificity and convenience of this system in identifying the RNAs interacting with proteins. Through tandem pur...

Materials

Name	Company	Catalog Number	Comments
Equipment
UV crosslinker	UVP	HL-2000 HybrilLinker
Affinity Purification Beads
ANTI-FLAG beads	Sigma-Aldrich	A2220
Streptavidin beads	Invitrogen	112.06D
Reagents
10x PBS	Gibco	70013032
3 M NaOAc	Ambion	AM9740
3 x FLAG peptide	Sigma-Aldrich	F4799
ATP	Sigma-Aldrich	A6559
Calcium chloride (CaCl2)	Sigma-Aldrich	C1016
CIP	NEB	M0290S	CIP buffer is in the same package.
DTT	Sigma-Aldrich	D0632
EDTA	Sigma-Aldrich	E9884
EGTA	Sigma-Aldrich	E3889
Gel purification kit	QIAGEN	28704
Glycogen	Ambion	AM9510
Magnesium chloride (MgCl2)	Sigma-Aldrich	449172
MNase	NEB	M0247S
NP-40	Amresco	M158-500ML
PMSF	Sigma-Aldrich	10837091001
Porteinase K	TAKARA	9033
Protease inhibitor cocktail	Sigma-Aldrich	P8340
Q5 High-Fidelity 2X Master Mix	NEB	0492S
reverse trancriptase (SupperScriptIII)	Invitrogen	18080093
RNA isolation reagent (Trizol)	Invitrogen	15596018
RNase Inhibitor	ThermoFisher	EO0381
RNaseOUT	Invitrogen	10777019
RQ1 Dnase	Promega	M6101
SDS	Sigma-Aldrich	1614363
Sodium chloride	Sigma-Aldrich	S9888
Sodium deoxycholate	Sigma-Aldrich	D6750
T4 PNK	NEB	M0201S	PNK buffer is in the same package.
T4 RNA ligaes	ThermoFisher	EL0021	T4 RNA ligase buffer and BSA are in the same package.
T4 RNA ligase2, truncated	NEB	M0242S	T4 RNA ligase buffer and 50% PEG are in the same package.
Trypsin-EDTA	ThermoFisher	25200072
Urea	Sigma-Aldrich	208884
mESC culture medium
DMEM (80%)	Gibco	11965126
2-Mercaptoethanol	Gibco	21985023
FCS (15%)	Hyclone
Glutamax (1%)	Gibco	35050061
LIF			purified recombinant protein; 10,000 fold dilution
NEAA (1%)	Gibco	11140050
Nucleoside mix (1%)	Millipore	ES-008-D
Penicillin-Streptomycin (1%)	Gibco	15140122
Kit
DNA gel extraction kit	QIAGEN	28704