CLIP Method to Study Protein-RNA Interactions in Intact Cells and Tissues

CLIP method to study protein-RNA interactions in intact cells and tissues.

James Tollervey1, Jernej Ule1*

Address:

1MRC-Laboratory of Molecular Biology, Hills Road, Cambridge, CB2 0QH, United Kingdom.

Keywords: CLIP; UV cross-linking; Immunoprecipitation; RNA-protein interaction

Abstract

In order to understand the mechanisms by which RNA-binding proteins carry out their functions, it is important to identify where they bind their targets. To facilitate this, the UV-crosslinking and Immunoprecipitation (CLIP) method was developed which allows for in vivo identification of protein-RNA interactions. To identify the sequence of CLIP RNAs, these need to be ligated to adapters and amplified to a cDNA library, which can be an inefficient process that had been improved over the last years. In this chapter, we present the general CLIP protocol and describe how the individual steps in the protocol can be optimised depending on the protein studied and the cell type or tissue used.

Background

Bioinformatic and experimental studies found that the position on nascent transcripts to which a protein binds often relates to its function in regulating alternative splicing (Dredge et al., 2005; Hui et al., 2005; Kanopka et al., 1996; Ule and Darnell, 2006). Therefore, to understand splicing regulation, it is important to profile the RNA sites bound by individual proteins on nascent transcripts with high positional resolution. Since intronic regions of nascent transcripts are rapidly degraded and therefore present in cells at very low abundance, it is important to crosslink the protein-RNA interactions in intact cells before purifying the RNA. This is achieved by in vivo UV cross-linking and immunoprecipitation (CLIP) (Ule et al., 2003), which purifies the RNA sites that crosslink to a particular RNA-binding protein.

CLIP employs immunoprecipitation and SDS-PAGE analysis of the protein-RNA complex in a way that was originally developed for studies of IRE1-RNA complex in the mammalian unfolded protein response (Bertolotti and Ron, 2001). The RNA is then isolated from the protein-RNA complex and ligated to RNA adapters, similar to the way originally developed for cloning of small interfering RNAs (Elbashir et al., 2001). The originalCLIP protocol ligatesthe 5` RNA adapter to phosphorylated CLIP RNA (Ule et al., 2003); this is potentially problematic, since itallows self-circularisation of the CLIP RNA, which tends to be more efficient than ligation to the RNA adapter.Therefore, a later protocol first ligated the 3` RNA adapter to dephosphorylated CLIP RNA (Ule et al., 2005). Since the original protocol ligated the RNA adapters sequentially to the isolated CLIP RNA, it often resulted in amplification of bacterial and yeast RNAs, or RNA adapter concatamer sequences. The solution to this problem was to perform the first RNA adapter ligation on-bead during immunoprecipitation(Ule et al., 2005).

Ligation of 3` RNA adapter to CLIP RNA was inefficient in our hands when performed on-bead, therefore we have resorted to ligation of 5` RNA adapter on-bead (Wang et al., 2009). This was efficient when using 3' phosphatase minus T4 Polynucleotide Kinase to phosphorylate the CLIP RNA, since the 2`-3` phosphate that remains on the CLIP RNA after RNase cleavage blocks self-circularisation. Another approach was described in a study that used a pre-adenylated 3` DNA adapter (Granneman et al., 2009). Here we describe an alternative solution that increases the efficiency of 3` RNA adapteron-bead ligation by using PEG 400 in the ligation reaction.

The CLIP protocol is comprised of seven basic steps. In this chapter, we will describe the protocol for each step, comment on the problems that can be encountered, and describe some of the potential solutions.

Material and reagents

1) UV cross-linking.

Cold PBS, Stratalinker (such as model 2400), 10 cm tissue culture dishes (Falcon).

2) Immunoprecipitation.

a) Lysis Buffer: 50 mM Tris-HCl, pH 7.4; 100 mM NaCl; 1 mM MgCl2; 0.1 mM CaCl2; 1 % NP-40; 0.5 % sodium deoxycholate; 0.1 % SDS; protease inhibitor and ANTI-RNase added fresh.

b) High-salt wash: 50 mM Tris-HCl, pH 7.4; 1 M NaCl; 1 mM EDTA; 1 % NP-40; 0.5 % sodium deoxycholate; 0.1 % SDS.

c) PNK wash: 20 mM Tris-HCl, pH 7.4; 10 mM MaCl2; 0.2 % Tween-20.

Reagents: Protein A Dynabeads (Dynal, 100.02), protease inhibitor cocktail (Calbiochem, 535140), ANTI-RNase (Ambion, AM2692), RNase I (Ambion, AM2295), Turbo DNase (Ambion, AM2239), Magnetic stand (Invitrogen), Thermomixer R (Eppendorf)

3) 3` RNA adapter ligation.

Shrimp alkaline phosphatase (Promega, M820A), Shrimp alkaline phosphatase buffer (Promega, M821A), RNA ligase (NEB, M0204L), 10x RNA ligation buffer (NEB, B0204S), Polyethylene glycol (PEG) 400 (Sigma, 81170), RNasin Plus(Promega, N2611), 4x Nupage loading buffer (Invitrogen, NP0007), L3 adapters (Dharmacon):

L31: P-UGAGAUCGGAAGAGCGGUUCAG-Puro

L32: P-GAAGAUCGGAAGAGCGGUUCAG-Puro

L33: P-CCAGAUCGGAAGAGCGGUUCAG-Puro

L34: P-AUAGAUCGGAAGAGCGGUUCAG-Puro

5` RNA phosphorylation.

10x PNK buffer and T4 Polynucleotide Kinase (3` phosphatase minus) (NEB, M0236L), 32P-γ-ATP, 10 mM ATP

4) RNA Purification.

SDS-PAGE electrophoresis and nitrocellulose transfer.

Novex NuPAGE 4-12 % Bis-Tris gel (Invitrogen, NP0321), 20x MOPS Novex NuPAGE running buffer (Invitrogen, NP0001), nitrocellulose membrane (Whatman, Protran), Novex wet transfer apparatus (Invitrogen), 20x Transfer buffer (Invitrogen, NP0006-1), methanol, BioMax XAR film (Kodak, 853 2665).

RNA isolation and precipitation.

a) PK buffer: 100 mM Tris-HCl, pH 7.4; 50 mM NaCl; 10 mM EDTA.

b) PK+urea buffer: 100 mM Tris-HCl, pH 7.4; 50 mM NaCl; 10 mM EDTA, 7 M urea.

Proteinase K (Roche, 1373196), RNA phenol/chloroform (Ambion, 9722), glycoblue (Ambion, 9510), 3 M sodium acetate pH 5.5 (Ambion, AM9740), absolute ethanol, Phase lock gel heavy 2ml (5 Prime, 2302830)

5) 5` adapter ligation.

a) TE-buffer: 10mM Tris-HCL, pH 7.4; 1 mM EDTA. 80% ethanol, L5 adapters (Dharmacon):

L5: ACACGACGCUCUUCCGAUCU

Gel purification of cDNA.

2x TBE-urea loading buffer (Invitrogen, LC6876), 15 well 10% TBE-urea gel (Invitrogen, EC68755BOX), 1 ml syringe plunger, Costar SpinX column (Corning, 8161), 1 cm glass pre-filter (Whatman, 1823010).

6) Reverse transcription.

dNTP set (GE healthcare, 27 2035 01), Superscript III, RT buffer and 0.1 M DTT (Invitrogen, 18080-093), PCR cycler (Biorad i-cycler).

RT primer:CTGAACCGC

7) PCR amplification.

2x Phusion flash high-fidelity PCR (Fermentas, F548), SYBR green II (Invitrogen, S7564), UV imaging facility.

PE PCR primers

5`: AATGATACGGCGACCACCGAGATCTACA

CTCTTTCCCTACACGACGCTCTTCCGATCT

3`:CAAGCAGAAGACGGCATACGAGATCGGTC

TCGGCATTCCTGCTGAACCGCTCTTCCGATCT

Oligonucleotide sequences © 2006 and 2008 Illumina, Inc. All rights reserved.

Comment

RNA adapters are designed for pair-end sequencing using Illumina Genome Analyser. The 3` adapter contains 5` phosphate and 3` puromycin to allow ligation to CLIP RNA, and prevent self-circularisation. The sequence of the 3` RNA adapter is reverse complementary to the last 20 nucleotides of the 3` sequencing primer. To allow multiplexing during sequencing, each adapter contains two additional nucleotides at the end. The different 3` adapters can be used for different experiments as well as for a control, allowing these to be sequenced in the same lane of a flowcell. The barcoding later allows identification of sequences specific for each experiment.The sequence of the 5` RNA adapter is identical to the last 20 nucleotides of the 5` sequencing primer. PE PCR primer sequences were as provided by Illumina for preparation of the Solexa paired-end libraries.

CLIP Protocol

1) UV cross-linking.

For tissue:

Instructions here are given for 50-100 mg of starting amount of tissue.

a) Add 10 volumes of cold PBS and partially dissociate the tissue by triturating using a 5 or 10 ml pipette. Add a 200 μl pipette tip to the end of 5 or 10 ml pipette, and further dissociate the tissue by passing through the tips several times. It is not necessary to disrupt the tissue into single cells for tissue cross-linking, as UV light can penetrate a few cell layers.

b) Transfer to 10 cm tissue culture dish, place on a tray with ice and irradiate suspension 3 times on ice for 100 mJ/cm2 in Stratalinker. Mix suspension between each irradiation.

c) Split cell suspension into 2 ml tubes, and centrifuge at maximum speed in a tabletop centrifuge for 3 minutes.

d) Remove supernatant, snap freeze pellets on dry ice, and store at -80°C until further use.

For cell culture:

a) For adherent cells, grow cells on 10 cm tissue culture dishes until 80% confluent.

b) Remove media, add 6 ml of cold PBS and place in Stratalinker on ice with the lid off. Irradiate once for 150 mJ/cm2.

c) Scrape off the cells and split into three 2 ml tubes. Centrifuge at maximum speed in a tabletop centrifuge for 3 minutes.

d) Remove supernatant, snap freeze pellets on dry ice, and store at -80°C until further use.

Comment

On a western blot of crosslinked cells, we have so far only been able to detect the protein migrating at its normal molecular weight, and not the crosslinked protein, suggesting that UV crosslinking between RNA and proteins is inefficient. Other methods have used formaldehyde to crosslink RNA to proteins with higher efficiency (Niranjanakumari et al., 2002), but this has the disadvantage of being less specific for the direct protein-RNA interactions and also requiring an incubation of the cells with a potentially toxic reagent.

2) Immunoprecipitation

Prepare antibody-conjugated Dynabeads:

a) For each experiment use 100 μl of protein A Dynabeads. Resuspend the beads and transfer 100 μl to a non-sticky 1.5 ml tube.

b) Wash beads 3 times with lysis buffer.

c) Resuspend in 200 μl of lysis buffer and add 1-5 μg of antibody depending on the efficiency of the antibody.

d) Rotate at room temperature for 30-60 minutes.

e) Wash 3 times with lysis buffer and leave beads in last wash until you are ready to add the lysates.

Prepare RNase-treated cell extract:

f) Resuspend each pellet of cross-linked material in 1 ml lysis buffer. It is optional to add 10 μl protease inhibitor and 1 μl ANTI-RNase (The ANTI-RNase inhibits RNase A, the predominant RNase in mammalian tissues, but does not inhibit RNase I. This allows for more standardised RNase conditions to be applied to diverse biologic source materials).

g) Sonicate on ice until cells or tissue are dissociated. Avoid foaming of the lysates.

h) Centrifuge at maximum speed at 4oC for 3 minutes.

i) Carefully collect the supernatant and transfer it to a new 2 ml tube.

j) Optional: make a dilution of RNase I at 1:50 in lysis buffer (high-RNase).

k) Make a dilution of RNase I at 1:1000 in lysis buffer (low-RNase).

l) Add 10 μl RNase dilution and 5 μl Turbo DNase per 1 ml of lysates (Turbo DNase is used because it is active in 100 mM NaCl). Incubate at 37°C for 3 minutes, then place on ice for 3 minutes.

Immunoprecipitate:

m) Add the solution to the antibody-conjugated Dynabeads.

n) Rotate at 4°C for 2 hours or overnight.

o) Discard the supernatant. Wash beads 2 times with high-salt buffer and 2 times with PNK buffer.

Comment

RNase I is advantageous relative to other RNAses since it has no nucleotide bias. This avoids sequence biases that could be introduced by other RNases. For instance, micrococcal nuclease preferentially cleaves 5` of A or T, RNAse A cleaves 3` of C or U and RNAse T1 3` of Gs. The protocol also uses Turbo DNase instead of standard DNase I, since Turbo DNase is active in conditions of up to 200 mM NaCl.

3) 3` RNA adapter ligation

3` RNA dephosphorylation and adapter ligation (low RNase experiment only, high RNase experiment go to step 3g)

a) Remove the PNK wash buffer and resuspend the beads in 10 μl of the following:

• 1 μl Shrimp alkaline phosphatase

• 1 μl 10x Shrimp alkaline phosphatase buffer

• 8 μl dH2O

b) Incubate in Thermomixer R at 37°C for 20 min, with intermittent shaking on 1000 rpm.

c) Discard the supernatant. Wash beads 2 times with high-salt buffer and 2 times with PNK buffer. Rotate the second high-salt wash at 4oC for 1 minute.

d) Remove supernatant from remaining beads and resuspend in the following mix:

• 3.5 μl dH2O

• 0.75 μl 10x RNA ligation buffer

• 2.5 μl PEG 400

• 0.25 μl RNA ligase

• 0.1 μl RNasin plus

• 3 μl L3 adapter (20 μM)

e) Incubate in cooled Thermomixer R at 16°C for 16 hours with intermittent shaking.

f) Wash beads once with PNK buffer.

5' RNA phosphorylation

g) Remove wash buffer and add 5 μl of hot PNK mix to the beads:

• 0.5 μl T4 Polynucleotide Kinase (3' phosphatase minus)

• 0.5 μl 32P-γ-ATP

• 0.5 μl 10x PNK buffer

• 3.5 μl dH2O

Incubate in a Thermomixer R at 37°C for 5 minutes at 1000 rpm.

h) To low RNase experiment add 5 μl of cold PNK mix to the beads:

• 1 μl 10mM ATP

• 0.5 μl 10x PNK buffer

• 3.5 μl dH2O

i) Incubate in a Thermomixer R at 37°C for 10 minutes at 1000 rpm.

j) Remove supernatant and discard as radioactive waste appropriately. Add 20 μl 1x Nupage loading buffer.

k) Incubate in a Thermomixer R at 70°C for 10 minutes at 1000 rpm, place on magnet and collect the supernatant.

Comment

PEG 400 is used in the ligation mixture because we found it increases ligation efficiency dramatically (up to a hundred fold). If paired-end sequencing will be carried out, different L3 adapters can be used for each experiment. Each adapter contains a different barcode sequence, allowing for easy identification of experiments after sequencing.

Instead of labelling the CLIP RNA using T4 Polynucleotide Kinase, and alternative is to radiolabel the L3 RNA adapter. This might be necessary in some cases to confirm that RNA is being visualised, especially if this can not be tested by comparing the high and low RNAse conditions. For instance, when analysing the micro-RNA bound to Argonoute proteins, the protein-RNA complex migrates as a sharp band regardless of the RNase concentration, therefore using radiolabelled L3 RNA adapter was advantageous (Chi et al., 2009).

4) RNA purification

SDS-PAGE electrophoresis and nitrocellulose transfer.

a) Load a 9 or 10 well Novex NuPAGE 4-12% Bis-Tris gel. Use 500 ml 1x MOPS running buffer.

b) Run gel at 200 V, and afterwards transfer to nitrocellulose membrane using Novex wet transfer apparatus.

c) Rinse the membrane in 1xPBS, and gently blot dry. Wrap in plastic wrap and expose to autoradiogram overnight. If the signal is overexposed, then also expose for one hour or less.

RNA isolation

d) Analyse the autoradiogram, and use the high-RNase sample to determine the specificity of the RNA-protein complexes. See the comment below to decide where to cut out a band corresponding to the protein-RNA complex.

e) Use a clean scalpel blade to excise a piece of membrane into a 1.5 ml tube.

f) Make a 2 mg/ml proteinase K solution in PK buffer and incubate at 37°C for 2 min to digest any contaminating RNAse.

g) Add 200 μl of proteinase K solution to each tube of isolated membrane and incubate at 37°C for 20 min.

h) Add 200 μl of PK/7M urea buffer and incubate for a further 20 minutes at 55°C.

i) Transfer the solution to a 2 ml Phase lock gel tube and add 400 μl RNA phenol/chloroform. Incubate in Thermomixer R at 37°C for 5 min at 1000 rpm.

j) Centrifuge for 5 min at maximum speed in a tabletop centrifuge at room temperature and pour the aqueous phase into a clean 1.5 ml tube.

k) Add 1 μl of glycoblue and 50 μl of 3 M sodium acetate, pH 5.5 and mix well. Add 1 ml of absolute ethanol, mix well and precipitate at -20°C for 1 hour or overnight. Centrifuge precipitating RNA samples for 10 minutes at full speed in a 4oC tabletop centrifuge. Remove supernatant and carefully was pellet twice with 80% ethanol. Leave pellet to dry at room temperature for 3 minutes

Comment

Analyse the autoradiogram to determine if the signal corresponds to a specific protein-RNA complex, and if the RNA is of the desired size distribution. Control experiments that lack the protein-RNA complex are crucial in initial optimisations; these include experiments that lack the antibody during immunoprecipitation, cell extract that lacks UV crosslinking, and cell extracts that lacks the RNA-binding protein (alternatively, knockdown cells, or cells overexpressing a tagged version of the protein can be used). In addition, comparison of the high and low RNase can test if the band representing bound RNA shifts up following low-RNase treatment.

Based on the control experiments, calculate the distance of the specific band from the closest contaminating band. If your protein band migrates less than 10 kDa away from other contaminating protein band, then it will be difficult to isolate specific RNA targets – in this case, it is worth spending more effort in optimising immunoprecipitation (increasing the stringency by raising the amount of salt and detergents during the washes). If no contaminating bands are present on the gel, or if they are far from the specific band, then a band can be excised from the gel of 20 to 40 kDa above the MW of the protein of interest. In the RNA-protein complexes less than 20 kDa above the MW of the protein, the insert RNA might be too short to map to the genome, but these can still be excised for comparative purposes.

The phase lock gel tubes are advantageous in separating RNA from the protein-containing phenol phase (step 4i). These tubes ensure that no phenol is carried over, which could inhibit the following RNA ligation reaction.

5) 5` RNA adapter ligation

RNA Ligation

a) Resuspend RNA pellet in the following mix:

• 5 μl dH2O

• 0.75 μl 10x RNA ligation buffer

• 2.5 μl PEG 400

• 0.25 μl T4 RNA ligase

• 0.25 μl RNasin plus

• 1.5 μl L5 adapter (20 μM)

b) Incubate in cooled PCR cycler at 16°C for 16 hours.

c) Add TE buffer to a volume of 350 μl and precipitate as in step 4k.

Gel purification of cDNA

a) Resuspend RNA in 8 μl of dH2O and 8 μl of 2x TBE-urea loading buffer.

b) Heat samples in Thermomixer R at 70oC for 5 minutes.

c) Load samples in a 10 or 15 well precast 10% TBE-urea gel in 1x TBE running buffer.

d) Run gel at 180 V for 40 minutes.

e) Cut out 2 bands per lane corresponding to 50-90 nucleotide and 90-150 nucleotide products (see comment below) and transfer each to a 1.5 ml tube.

f) Add 400 µl of TE-buffer to each gel piece and crush with a 1 ml syringe plunger.

g) Incubate in a Thermomixer R at 37°C for 2 hours at 1100 rpm.

h) Centrifuge tubes at full speed in a tabletop centrifuge for 2 minutes at room temperature.

i) Transfer the supernatant onto a Costar SpinX colum to which you have added two 1 cm glass pre-filters and spin at full speed for 1 minute in a tabletop centrifuge.

j) Precipitate as in step 4k.

Comment

The upper (lighter blue) dye runs at 120-150 nucleotides, and the lower (darker blue) dye runs at 20 nucleotides, and these can be used to guide excision: cut the lower band (50-90 nucleotides) from 2 cm below the bottom of the upper dye to 1 cm below, cut the upper band (90-150 nucleotides) from 1 cm below to the middle of the upper light blue dye. Cutting lower than this is not advised as free linker could be incorporated into the sample, which could potentially contaminate the further steps.

6) Reverse transcription

a) Resuspend RNA in the following mix:

• 6.25 μl dH2O

• 0.5 μl RT primer (0.5 μM)

• 0.5 μl dNTP mix (10 μM)

d) Transfer solution to a 0.2 ml PCR tube and incubate in a PCR cycler at 70oC for 5 minutes, then hold at 25oC whilst you add the following mix:

• 2 μl 5x RT buffer

• 0.5 μl DTT (0.1 M)

• 0.25 μl Superscript III RT enzyme (200 U/μl)

e) Incubate tubes for 5 minutes at 25oC, 20 minutes at 42oC, 40 minutes at 50oC and then hold at 4oC.

f) Mix the samples to be multiplexed (see comment below) and add TE-buffer to a total volume of 350 μl. Precipitate as in step 4k.

Comment

A few amino acids remain covalently attached to the RNA at the crosslink site after the proteinase K digestion. Primerextension assays have shown that the majority of cDNAs prematurely truncate immediately beforethe'crosslink nucleotide' (Urlaub et al., 2002). Such truncated cDNAs will not contain the sequence of the 5` adapter, therefore CLIP requires that reverse transcription passes over the residual amino acids.