Crosslinking and Immunoprecipitation
Followed by MS (IP-MS) from XRNAX Extracts

For validating the protein-RNA interaction of one particular protein, conventionally a polynucleotide kinase (PNK) assay is performed. Therefore cells are transfected with a tagged version of this protein, UV-crosslinked, lysed, their RNA fragmented and immunoprecipitation performed towards the protein-of-interest. If this protein is in fact RNA-binding, some of it will carry a covalently attached RNA-fragment, which is co-immunoprecipitated and can be radiolabeled by PNK and visualized after SDS-PAGE [e.g. Baltz et al., Mol Cell, 2012].
The following protocol describes a procedure by which an RNA-binding protein is detected through MS with a covalent nucleotide-modification after UV-crosslinking, XRNAX and immunoprecipitation. Nucleotide-peptide adducts are direct evidence for protein-RNA interaction sites and can confirm findings of a PNK assay through orthogonal methodology. Furthermore, nucleotide-crosslinked peptides pinpoint the RNA interaction to a localized site in the protein sequence, thereby allowing for sub-domain resolution.

XRNAX
Depending on how well your cells can be transfected, how strong your protein-of-interest expresses and how well this protein crosslinks to RNA the input amount of cells can vary. We have successfully confirmed protein-RNA interactions from XRNAX extracts of one 15 cm dish MCF7 cells with transient expression from a pcDNA5 vector. The following protocol assumes this amount of input, however, smaller amounts of input material should work as well and could be processed with the identical protocol.
UV-crosslink and XRNAX extract cells as described in the ‘The Protocol’ section. Dissolve the entire yield of the XRNAX extraction in 200 µl tris-Cl 50 mM.

Required Material
Vector DNA with FLAG-tagged protein-of-interest 15 µg (e.g. pcDNA5 FRT TO FLAG-HA)
Lipofectamine 3000 (invitrogen, L3000008)
Opti-MEM (Gibco, 31985070)
Tris-Cl pH=7.5 1M
RNase A (Thermo, EN0531)
RNase I (Ambion, AM2295)
IP buffer 2 x (Tris-Cl 100 mM pH=7.5, 1 % NP40, 300 mM LiCl, 0.2 % LiDS)
Anti-FLAG M2 Magnetic Beads (Sigma, M8823)
Elution buffer (tris-Cl 50 mM, SDS 10 %)
SP3 Magentic beads (GE, 44152105050250)
Acetonitrile 100 %
EtOH 70 %
Trypsin/LysC (Promega, V5073)
TEAB 20 mM
Formic acid 10 %

Transfection
Depending on your cell type and your protein-of-interest transfection conditions may vary. For transfection of MCF7 cells with pcDNA5 constructs we use Lipofectamine 3000 and see good expression after 24-48 hours.

RNase Digestion
As an initial step RNA in XRNAX extracts is completely degraded in order to leave one or several few nucleotides on RNA-crosslinked proteins behind. Adduct masses of nucleotide-crosslinked peptides from such proteins are small enough to be picked up in a mass-tolerant petide search by MSFragger.
Add 2.5 µl of Rnase A and 2.5 µl of RNase I to the 200 µl XRNAX extract and allow digestion over night at 37 °C, 700 rpm shaking.

IP
Prepare 50 µl of anti-FLAG M2 beads per sample by washing them three times with 1 ml IP buffer 1 x. Finally resuspend the beads in 50 µl of IP buffer 1 x. Add 200 µl of IP buffer 2 x to the RNase digested sample, mix by pipetting and add the beads. Allow binding to occur for 4 hours at room temperature on a rotating wheel. Capture the beads on a magnetic rack, discard the supernatant and wash the beads three times with 500 µl IP buffer 1 x, each time resuspending them for 5 minutes on a rotating wheel. Collect residual wash buffer with a quick spin on a microcentrifuge and remove it before addition of 200 µl elution buffer. Elute for 30 minutes at 37 °C, 700 rpm shaking. Transfer the eluate to a fresh tube and repeat the elution with another 200 µl elution buffer. Combine eluates to 400 µl total volume.

SP3 Cleanup and Trypsin Digestion
Wash 40 µl SP3 beads three times with 1 ml MilliQ water and resuspend them in 100 µl MilliQ water. Add 10 µl of those beads to one sample, mix and combine with 500 µl ACN 100 %. Mix again and allow binding for 15 minutes at room temperature. Collect the beads on a magnetic rack for one minute, discard the supernatant and wash the beads three times with 2 ml EtOH 70 % while remaining collected on the magnet. Collect the residual EtOH with a quick spin on a microcentrifuge and remove it. Add 100 ng trypsin/LysC in 20 µl TEAB and digest over night at 37 °C, 700 rpm shaking (use 2 ml tubes for better mixing of beads). Collect the beads on a magnetic rack for one minute and transfer the supernatant to a fresh tube. Add 2 µl formic acid 10 % and centrifuge the sample for 5 minutes with maximum speed at room temperature. Carefully transfer 20 µl of the sample to a fresh tube without disturbing a pellet. Analyze by HPLC-MS using the orbitrap at a resolution of 120000 or higher.

Data Analysis
An unbiased identification of peptide adducts with unknown masses can be accomplished by MSFragger [Kong et al., Nat Biotech 2017]. If it is unclear which amino acid in your protein-of-interest is crosslinked to which ribonucleotide, conventional searches taking variable modifications into account become impossible, due to the enourmous multiplicity of at least four modifications (4 ribonucleotides, cyclic, non-cyclic, etc.) possibly crosslinking to all 20 amino acid.
Search your data with MSFragger in open-search mode with a mass-tolerance of +/- 1000 Da, which allows for the discovery of up to three nucleotides crosslinked to a peptide (e.g. the RNA sequence UpUpUp crosslinked to a peptide would result in an adduct mass of 936.085 Da).
Alternatively, as uridine and cyclic uridine are the most common crosslinked nucleotides identified, search your data with a conventional search engine such as MaxQuant [Cox and Mann, Nat Biotech 2008] and specify these two as variable modification.