Protocol to study in vivo circRNA interactions in the mouse cortex using an RNA pull-down approach

Summary

Here, we present a protocol for identifying interactors of circular RNAs (circRNAs), specifically circDlc1(2), in the mouse cortex. We outline steps for tissue dissociation and UV crosslinking to maintain native interactions, followed by an RNA pull-down to isolate the circRNA and its associated molecules. The protocol has been optimized to detect potential protein interactors and can be adapted for use in other regions of the mouse nervous system.

For complete details on the use and execution of this protocol, please refer to Silenzi et al.¹

Graphical abstract

Highlights

• •Steps for the dissociation of cortical tissue followed by UV crosslinking
• •Instructions for the subsequent circRNA-pull-down experiment
• •Sample preparation for both RNA and protein analyses

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

Here, we present a protocol for identifying interactors of circular RNAs (circRNAs), specifically circDlc1(2), in the mouse cortex. We outline steps for tissue dissociation and UV crosslinking to maintain native interactions, followed by an RNA pull-down to isolate the circRNA and its associated molecules. The protocol has been optimized to detect potential protein interactors and can be adapted for use in other regions of the mouse nervous system.

Before you begin

Few data currently exist of the function of recently re-discovered circRNAs in vivo.^2,3 CircRNAs are highly expressed in the mammalian brain,^4,5,6 where they have been seen to localize in neurites and synapses.^1,7,8 Identifying interactors for these types of molecules may provide useful hints concerning their functions and role in regulating gene expression, for example in the nervous district where they are most expressed. This protocol focuses on the enrichment of endogenous circDlc1(2), a circRNA whose role in the striatum and at the glutamatergic synapse has been previously described in Silenzi et al., 2024.¹ In this manuscript, the protocol has been adapted for the study of the circRNA in the cortex. Additional steps have been added concerning the identification of putative protein interactors.

Note: The protocol does not include instructions for the dissection of the brain regions; instead it begins with the dissociation of pre-dissected cortical tissue, followed by the RNA pull-down procedure. We recommend performing tissue dissociation and the circRNA pull-down procedure on separate days due to the overall length of the protocol.

Innovation

The RNA pull-down approach itself results in the isolation of a RNA molecule of interest, together with its binding partners (proteins and/or other RNAs). A number of protocols currently exist for the study of circRNA interactions in vitro, however the study of their interactions occurring in a more physiological in vivo setting has not yet been described as thoroughly. Performing the experiment using brain tissue can be advantageous, as native interactions which occur in a more physiological setting are preserved. These same interactions may not occur or could be altered in a more isolated environment such as in vitro cell cultures. One of the present challenges is the efficient crosslinking of proteins and RNAs. Our protocol comprises the dissociation of the cortical tissue into a cell suspension prior to the crosslinking procedure, so as to increase its efficiency and the likelihood of protein detection in subsequent downstream experiments whilst maintaining cell integrity.

Institutional permissions

The murine cortical samples used for this protocol were collected according to the regulatory standards and have been approved by the Ministry of Health with the authorization n° 1078/2020-PR. Institutional approval was not required for experiments involving ex vivo mouse tissue.

Probe design

Timing: variable

• 1.
  Design 20 nucleotides (nt) long DNA oligonucleotide probes that have a biotin-TEG modification at their 3′ end. This allows the probes to remain tethered at one end to streptavidin-coated magnetic beads, minimizing steric hindrance near the site of hybridization; with the biotin-streptavidin interaction being one of the strongest non-covalent bonds occurring in nature (see Figure 1A). The probes must pair in an antisense manner to the circRNA of interest. A GC content of 45%–50% is preferred,⁹ and off-target effects should be checked for every probe that is designed.

  Note: Online tools (e.g. the Stellaris Probe Designer tool by LGC Biosearch Technologies or the OligoAnalyzer tool by IDT) can be used to facilitate the design.

  CRITICAL: One of the probes must be designed to be complementary to the backsplicing junction (BSJ) of the circRNA of interest, as this region is exclusive to the circular molecule and is not present within the linear cognate. The BSJ should preferably be targeted by the middle of the antisense probe (equal number of nts before and after the BSJ sequence) in order to minimize the probability of pairing with the linear counterpart. Shifting the sequence by 1-3 nt with the respect to the BSJ is also acceptable.

  • a.
    Design probes against a negative control, e.g. LacZ mRNA, that is not expressed in the murine system.

Figure 1.

Schematic representation of probe hybridization

Antisense DNA oligonucleotide probes targeting a circRNA of interest (A) and pairing to the linear counterpart (B). Related to Step 1 of ‘Probe Design’.

Preparation of reagents and equipment on the day of dissociation

Timing: 30 min

• 2.Wipe bench surface and relevant equipment with alcohol or RNase cleaning agents (e.g. RNaseZap). Pipettes should also be sterilized.
• 3.Switch on a microcentrifuge and set it to 4 degrees Celsius (°C).
• 4.
  Prepare the reagents.

  • a.
    Resuspend the DNAse I in 500 μL of Earle’s Balanced Salt Solution (EBSS) to a final concentration of 10 mg/mL. Once resuspended, store single use aliquots at −20°C. Only freeze and thaw once: aliquots are stable at 4°C for several weeks.

    Note: DNAse I is used to digest DNA from damaged cells, reducing the viscosity of the samples.

  • b.
    Resuspend the papain in 5 mL EBSS (20U/mL) and incubate it at 37°C for at least 30 min prior to its use, to fully activate the enzyme. Then, add 12.5 μL DNAse I (final concentration of 25 μg/mL) to this solution.
  • c.
    Resuspend the ovomucoid protease inhibitor in 32 mL EBSS (10 mg/mL of ovomucoid inhibitor and 10 mg of albumin/mL) and incubate at 37°C prior to use.

    Note: Papain and ovomucoid protease inhibitors can be stored at 2°C–8°C after use for up to 6 months.

• 5.Fill an ice bucket with ice and layer it with aluminum foil. Use this to pre-chill a culture dish, a scalpel blade, a new set of 2 mL centrifuge tubes, DPBS, and DPBS w/o MgCl₂ and CaCl₂.
• 6.Place harvested tissue samples on ice.

Note: Working quickly will help minimize degradation.

• 7.Set a thermomixer at 37°C.

Preparation of reagents and equipment for the circRNA pull-down

Timing: 45 min

• 8.
  Prepare buffers as indicated in the ‘materials and equipment’ section.

  • a.
    Pre-cool Lysis Buffer (LB), Hybridization Buffer (HB) and RNA Elution Buffer on ice before use (these can be stored at 4°C for up to 6 months).

Note: We recommend performing step 8 the day before the pull-down experiment for time purposes.

CRITICAL: Keep Proteinase K Buffer 2X (PKB) and Benzonase Elution Buffer at 20°C–25°C (RT). These can be stored at RT for up to 6 months.

• 9.Thaw or resuspend the antisense biotinylated DNA probes to a concentration of 100 μM (100 pmol/μL). Once resuspended, the probes should be stored at −20°C.

Note: If the probes have been previously resuspended and are to be thawed, we recommend doing this as soon as possible as they thaw very slowly.

• 10.Switch on a microcentrifuge and set it to 4°C.
• 11.Set a thermomixer to 80°C.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Biological samples
Mouse C57BL/6J WT cortical tissue	This paper	N/A
Mouse C57BL/6J, circDlc1(2)−/− cortical tissue	This paper	N/A
Chemicals, peptides, and recombinant proteins
Papain	Worthington Biochemical Corporation	cat#LK003176
Ovomucoid protease inhibitor with bovine serum albumin	Worthington Biochemical Corporation	cat#LK003182
Sterile Earle’s balanced salt solution (EBSS)	Sigma-Aldrich	cat#E7510
Deoxyribonuclease I from bovine pancreas (DNase I)	Sigma-Aldrich	cat#DN25
Formamide	Sigma-Aldrich	cat#47671
Dulbecco’s phosphate-buffered saline w/o MgCl2 and CaCl2 (DPBS w/o)	Sigma-Aldrich	cat#D8537
Dulbecco’s phosphate-buffered saline with MgCl2 and CaCl2 (DPBS)	Sigma-Aldrich	cat#D8662
RiboLock RNase inhibitor	Thermo Fisher Scientific	cat#EO0384
cOmplete, EDTA-free protease inhibitor cocktail (PIC)	Roche	cat#11873580001
Proteinase K, recombinant PCR Grade	Roche	cat#03115828001
Benzonase nuclease	EMD-Millipore	cat#70746-3
QIAzol lysis reagent	QIAGEN	cat#79306
NP-40	Sigma-Aldrich	cat#85124
HEPES sodium salt	ICN Biomedicals Inc	cat#105598
NaOH	Sigma-Aldrich	cat#S5881
KCl	Sigma-Aldrich	cat#P9541
MgCl2	Sigma-Aldrich	cat#63069
EDTA	Sigma-Aldrich	cat#E5134
Glycerol	Sigma-Aldrich	cat#G7893
Dithiothreitol (DTT)	Sigma-Aldrich	cat#29309098
NaCl	Sigma-Aldrich	cat#S9888
Trizma	Roche	cat#11814273001
HCl	ROMIL-SA	cat#A9396
Sodium dodecyl sulfate (SDS)	PanReac AppliChem	cat#A0675
N-lauroyl sarcosine, sodium salt (NLS)	Sigma-Aldrich	cat#L7414
Acetone	Sigma-Aldrich	cat#32201
Trichloroacetic acid (TCA) solution	Sigma-Aldrich	cat#T0699
Streptavidin MagneSphere paramagnetic particles	Promega	cat#PR-Z5482
RNaseZAP	Sigma-Aldrich	cat#R2020
Bio-Rad protein assay dye reagent concentrate	Bio-Rad	cat#5000006
Critical commercial assays
SuperScript VILO cDNA synthesis kit	Invitrogen	cat#11754050
miRNeasy mini kit	QIAGEN	cat#217004
PowerUp SYBR Green master mix	Life Technologies	cat#A25742
Oligonucleotides
DNA oligonucleotide probes are listed in Table S1.	Silenzi et al.1	N/A
Oligonucleotides for qPCR experiments in this work are listed in Table S2.	Silenzi et al.1	N/A
Software and algorithms
Prism 9	GraphPad by Dotmatics	https://www.graphpad.com/scientific-software/prism/
QuantStudio 3 and 5 real-time PCR system software	Thermo Fisher Scientific	RRID:SCR_020238https://www.thermofisher.com/it/en/home/global/forms/life-science/quantstudio-3-5-software.html
Other
Magna GrIP rack	Sigma-Aldrich	cat#20-400
Spectrolinker XL-1000 UV crosslinker	Spectro-UV	N/A
Mini LabRoller rotator	Labnet	cat#H5500-230V-EU

Materials and equipment

LB

Reagent	Final concentration
HEPES-NaOH pH 7.5	10 mM
KCl	20 mM
MgCl2	1.5 mM
EDTA	0.5 mM
Glycerol	20%
DTT	1 mM
NP-40	1%
Nuclease-free H2O	up to final volume

Filter buffer (0.22 μm) before the addition of NP-40.

HB

Reagent	Final concentration
NaCl	500 mM
Tris-HCl pH 7.5	100 mM
SDS	0.2%
EDTA	10 mM
Formamide	15%
DTT	1 mM
NP-40	1%
Nuclease-free H2O	up to final volume

Filter buffer (0.22 μm) before the addition of NP-40.

CRITICAL: Add formamide using a fume hood.

Benzonase elution buffer

Reagent	Final concentration
Tris-HCl pH 8.0	20 mM
NLS	0.05%
MgCl2	2 mM
DTT	0.5 mM
Nuclease-free H2O	up to final volume

Filter buffer (0.22 μm).

PKB 2X

Reagent	Final concentration
Tris-HCl pH 7.5	200 mM
NaCl	300 mM
EDTA	25 mM
SDS	2%
Nuclease-free H2O	up to final volume

Filter buffer (0.22 μm).

RNA elution buffer

Reagent	Final concentration
Tris-HCl pH 8	20 mM
EDTA	10 mM
NLS	2%
DTT	2.5 mM
Nuclease-free H2O	up to final volume

Filter buffer (0.22 μm).

Step-by-step method details

Brain tissue dissociation and UV crosslinking

Timing: 3–4 h, as needed

This procedure comprises the dissociation of the cortex (or desired brain region) into a cell suspension to increase the surface area available and thus facilitating the penetration of UV radiation. UV-crosslinking allows endogenous RNA-protein interactions occurring within a cell to be ‘frozen’ at the time of irradiation.

CRITICAL: All steps from this time onward must be performed on ice unless noted otherwise.

Note: For this protocol, cortices were harvested from adult mice (3–5 months old). On average, a mouse cortex weighs ∼100 mg and yields 1.5–3.5 mg of total extract after dissociation and crosslinking.

Note: If possible, harvest the tissue and dissociate it straight away.

• 1.Collect the mouse cortex in 2 mL microcentrifuge tubes with pre-chilled Dulbecco’s Phosphate Buffered Saline with MgCl₂ and CaCl₂ (DPBS). Place the tubes on ice while harvesting the remaining tissues.
• 2.Rinse the tissue with fresh DPBS to remove left-over traces of blood.
• 3.Transfer each cortex to a pre-chilled 60 mm petri dish and chop into small pieces using a sterile scalpel. These should be 1-2 mm thick.
• 4.Transfer the minced tissue back to the tubes kept on ice using a P1000 pipette tip. We recommend cutting the pipette tip to facilitate tissue transfer to the tubes.
• 5.Spin the tissue at 300 x g for 5 min at 4°C and discard the supernatant.
• 6.
  Enzymatically digest the cortical tissue using papain.

  • a.
    Add the pre-warmed papain solution (20U/cortex) to the tissue pellet and flick the sample.

    Note: Using the pipette to resuspend the pellet may cause it to stick to the wall of the pipette tip.

  • b.
    Incubate for 30 min at 37°C and 750 rpm using a thermomixer. Adjust rpm based on volume of papain and tissue. The tissue pieces should be in a homogenous suspension throughout the incubation period.

    Note: Slower agitation will cause the tissue pellet to stick together and form a large clump.

  • c.
    Inactivate the papain using an equal volume of ovomucoid inhibitor and triturate using a P1000 tip.

    CRITICAL: Do not introduce bubbles while triturating as this will damage cell membranes, leading to increased cell death and debris in the suspension.

• 7.Spin the tissue suspension at 300 x g for 5 min at 4°C and discard the supernatant. During this centrifugation step, chill a 10 cm petri dish on a tray containing a bed of ice slush which will be required for the following steps.
• 8.
  Prepare samples for UV-crosslinking.

  • a.
    Resuspend the pellet in 1 mL of DPBS w/o and place on ice.
  • b.
    Add 2 mL DPBS w/o to the pre-chilled 10 cm petri dish followed by the 1 mL of resuspended cell suspension.
  • c.
    Wash the centrifuge tube with an additional 1 mL of DPBS w/o and add this to the 10 cm petri dish (final volume of 4 mL) in order to cover the whole surface of the dish. Make sure the sample is evenly distributed across the dish.
• 9.Place the dish, along with the ice in a UV crosslinker (e.g. Spectrolinker) without the lid and crosslink once at 400 mJ/cm² energy (254 nm UV light).

CRITICAL: The dish must be kept on ice during crosslinking in order to limit RNA degradation. The dish should also be positioned on the ice as flat as possible to ensure an even exposure of the sample to UV radiation.

• 10.
  Collect the cells using a cell scraper and transfer them to a conical tube placed on ice.

  Optional: If a refrigerated bench-top centrifuge is not available, divide the collected cells between two 2 mL microcentrifuge tubes and use a refrigerated microcentrifuge.

  • a.
    Spin the samples at 300 x g at 4°C for 5 min and discard the supernatant.
  • b.
    Snap freeze the pellets.

    Pause point: The UV-crosslinked cell pellets can be stored at −80°C until the day of the pull-down experiment.

Preparation of cortical tissue lysate

Timing: 45 min

The purpose of this step is to prepare a cortical tissue lysate ensuring the removal of cell membranes and debris (insoluble cellular components).

Note: We recommend preparing the tissue lysate on the same day of the circRNA pull-down experiment in order to minimize sample degradation due to additional freeze-thaw cycles. Nevertheless, left over lysate can be snap-frozen in liquid nitrogen and stored at −80°C.

CRITICAL: All steps must be performed on ice unless noted otherwise.

• 11.Thaw dissociated tissue pellets on ice and complement LB with protease inhibitors (1:100) and RNase inhibitors (1:200).

CRITICAL: Brain tissue features high RNase activity, so inhibitors must be added to each working buffer.

• 12.Add 500 μL LB to each pellet and resuspend using a P1000 pipette tip first, followed by a P200.
• 13.Place samples on a tube rotator (e.g. Mini LabRoller) at 4°C for 15 mins.
• 14.Resuspend lysate once again with a P200 tip set to 190 μL and place on ice for 3–5 mins.
• 15.Spin the samples at 4°C and 16,000 x g for 15 min to pellet the insoluble cellular components. Transfer the supernatant to a fresh 1.5 mL centrifuge tube.
• 16.Quantify the amount of extract using the Bradford assay (Bovine Serum Albumin protein is used as standard).

CRITICAL: At least 1 mg of total protein extract is required per probe set (individual RNA pull-down). Lower amounts may result in poor detection of the circRNA and interacting partners.

Optional: You may also choose to perform steps 11–15 in a refrigerated room.

Pre-cleaning of extract

Timing: 40 min

The purpose of this step is to remove any non-specific interactions occurring between the magnetic streptavidin beads and the extract.

• 17.Wash 25 μL of streptavidin paramagnetic beads per mg of extract twice with 250 μL of LB.
• 18.Resuspend the beads in 50 μL complemented LB.
• 19.Transfer the required volume of total protein extract (quantified in step 16) for all probe sets to a new tube and add the beads.
• 20.Incubate extract with the beads on a tube rotator for 30 min at RT.
• 21.Discard the beads using a magnetic rack (e.g. Magna GrIP rack) to transfer the pre-cleaned extract to a fresh tube.
• 22.Transfer 10% of the volume to a new tube. This will be the input control sample.

Note: From here onwards, the input sample should undergo the same incubation steps as the respective pull-down samples to ensure that the resulting enrichment values for the circRNA of interest are reliable.

CRITICAL: Do not place beads on ice to avoid clumping which interferes with their efficient resuspension.

Probe hybridization

Timing: 4 h 30 min

• 23.
  Prepare the probe mixes. We recommend making aliquots of these and store them at −20°C to limit the number of freeze-thaw cycles.

  Note: For the purpose of time, this step can be performed during step 20 of the pre-cleaning.

  • a.
    Transfer 100 pmol probe mix per mg of extract to a new 1.5 mL microcentrifuge tube and bring to a final volume of 10–20 μL with RNase-free water. Do this for both circRNA-specific probes and controls.

    Note: The same volume should be transferred for each individual probe comprising the probe mix.

  • b.
    Heat the probes at 80°C for 2 min to ensure removal of secondary structures and quickly place on ice.
• 24.Divide the extract into a different tube for every probe set.
• 25.Add two volumes of HB complemented with protease and RNase inhibitors as above. This should also be done for the input control samples.
• 26.Add the probes to each corresponding tube and incubate for 4 h on a tube rotator at RT.

Note: The concentration of the extract will determine the type of microcentrifuge tube to be used for steps 24–31.

CRITICAL: Do not add the probes to the input control sample.

circRNA pull-down

Timing: 1 h

This step leads to the enrichment of the circRNA of interest, along with its interacting protein and RNA partners. The subsequent wash steps enable the removal of any non-crosslinked, non-specific interactions occurring with the target circRNA.

• 27.
  Prepare the beads for the pull-down.

  • a.
    Wash 100 μL of streptavidin paramagnetic beads per mg of extract twice with 500 μL of HB.
  • b.
    Resuspend the beads in 100 μL HB complemented with protease and RNase inhibitors as above.
• 28.Add the same volume of beads to every pull-down sample.

CRITICAL: Do not add the beads to the input sample.

• 29.Incubate samples for 30 min on a tube rotator at RT.
• 30.Collect the beads bound by the circRNA target with a magnetic rack and discard the supernatant.

Note: The supernatant can be stored as the ‘unbound’ fraction which can be used to assess the efficiency of the pull-down procedure.

• 31.
  Wash the beads.

  Optional: You can choose to perform this step in a refrigerated room.

  • a.
    Add 500 μL of ice-cold HB and gently resuspend the beads.

    CRITICAL: Keep the HB used for the washes on ice prior to use.

    Note: Complementing HB for the washes with protease and RNase inhibitors as above is not required.

  • b.
    Incubate using a tube rotator for 3 min.
  • c.
    Spin the tubes, then place on a magnetic rack for 1 min. Carefully discard the supernatant.
  • d.
    Repeat step 31 (a-c) for a total of 5 washes.

    Optional: Proceed to step 44 if proteins are not required. If both RNA and protein interactors are of interest, separate the sample into two tubes before collecting the beads following the fifth wash. One fraction (50 μL, typically 10% of the volume) will be destined for RNA extraction (step 44–51) and the other (450 μL - 90%) for the collection of proteins (steps 32–43). Separate the input control sample following the same ratio.

Protein elution and purification

Timing: 3.5 h on hands and 16 h (overnight) incubation

These steps are performed on the protein fraction collected in step 31 and are required for the removal of the antisense biotinylated DNA probes as well as other nucleic acids to increase the purity of the protein sample.

CRITICAL: To maintain a keratin-free environment, use sterile pipette tips with filters and change gloves regularly.

• 32.Discard the supernatant from the beads using a magnetic rack.
• 33.Resuspend the beads in 1 mL Benzonase elution buffer supplemented with protease inhibitors (1:100), containing 80 U of benzonase non-specific nuclease.
• 34.Incubate the tubes at 37°C for 2 h in a thermomixer. Set this for 30 s on and 30 s off intermittent mixing at 1100 rpm. During this incubation period label six 1.5 mL microcentrifuge tubes for each pull-down condition that will be required for bead removal in steps 35 and 36.
• 35.Place the tubes on a magnetic rack for 1 min to allow bead collection. Transfer the supernatant to a fresh tube.
• 36.Repeat step 35 five times.

Note: This step ensures the complete removal of streptavidin beads from the sample.

• 37.Using a fume hood, add a 10% final concentration of ice-cold trichloroacetic acid (TCA) to each sample, including the input control sample. The volume of the latter can be increased by adding benzonase buffer to match the volume of the pull-down samples.
• 38.Incubate the samples overnight at −20°C.

Note: Increasing the volume of the 10% input control sample may facilitate the precipitation of proteins. In addition, we recommend an overnight incubation with TCA, although a 30 min incubation on ice should be sufficient.

• 39.Centrifuge the samples at 16,000 x g for 30 min at 4°C. A faint pellet should be visible at the bottom of the tube by the end of this centrifugation step.
• 40.Carefully discard the supernatant without dislodging the pellet and replace with 1 mL ice cold 100% acetone.
• 41.Briefly vortex the tube and centrifuge at 16,000 x g for 15 min at 4°C. Carefully discard the supernatant. A white pellet should be visible at the bottom of the tube. You can mark this pellet with a dot in order to easily identify it after it has dried.
• 42.Leave the tubes open and allow the pellet to air-dry for 5-15 min. Lay the tube sideways to prevent air-drop contamination, in a fume hood. Do not let the pellets dry for too long to avoid difficulties in resuspension.
• 43.Store the protein pellets at −80°C.

Note: The protein yield at the end of the pull-down procedure will be very low. We therefore recommend analyzing the samples through mass spectrometry. The buffer composition for the resuspension may vary depending on the specific requirements of the chosen instrument.

RNA elution

Timing: 1 h 15 min

These steps are performed on the RNA fraction collected in step 31 and entail the removal of proteins from the pull-down samples through treatment with proteinase K. This is important as it inactivates RNases and increases the efficiency of RNA reverse transcription in subsequent steps, that is required to assess the outcome of the pull-down.

Note: RNA elution from the dedicated fraction (step 31, optional) can be performed in parallel to protein elution during the 2 h incubation of step 34.

• 44.Discard the supernatant from the beads using a magnetic rack.
• 45.Resuspend the beads in a volume of RNA elution buffer.
• 46.Add proteinase K to each sample at a final concentration of 1 mg/mL in PKB.

CRITICAL: Do not place the PKB on ice as this will cause the SDS to precipitate. Likewise, do not place the samples on ice after adding the PKB.

• 47.Incubate samples at 52°C for 1 h at 300 rpm using a thermomixer.
• 48.Briefly spin down the tubes and add 3-5 volumes of QIAzol Lysis Reagent or Trizol equivalent to each sample.
• 49.Vortex the samples for 5 min to detach the circRNA target from the beads.
• 50.Spin the samples and place them on a magnetic rack for at least 3 min in order to collect the beads.
• 51.Carefully transfer the supernatants to fresh tubes.

Pause point: Samples can be frozen and stored at −80°C before proceeding with RNA extraction following the manufacturer’s instructions (available at: https://www.qiagen.com/us/resources/resourcedetail?id=f646813a-efbb-4672-9ae3-e665b3045b2b&lang=en).

• 52.Check circRNA enrichment via RT-qPCR.

Note: We suggest performing a DNase treatment during RNA extraction for the removal of genomic DNA as well as the DNA probes.

Note: Samples can be sent for RNA-sequencing analyses for the identification of RNA interactors for the circRNA of interest.

Expected outcomes

A single cortex, after dissociation and UV-crosslinking should yield around 1.5–3.5 mg of total extract. This may vary depending on the age and gender of the animal, as well as the efficiency of tissue lysis.

The correct enrichment of the circRNA can be confirmed by performing qRT-PCR. The enrichment of the linear counterpart should always be assessed in parallel to that of the circRNA. Indeed, the enrichment of the circRNA should be significantly greater than that of the linear counterpart when compared to a negative control (see Figure 2).

Figure 2.

circRNA enrichment

UV-crosslinked RNA pull-down (PD) assay performed on total cortex extracts. Real-time qPCR quantification is shown for the enrichment of circDlc1(2), lin-Dlc1, and negative control Atp5o in PD-circDlc1(2) and PD-LacZ samples relative to the input (n=6). Data are represented as mean ± SEM. See also Table S2.

The circRNA pull-down can be followed by RNA sequencing and/or mass spectrometry analyses in order to identify RNA and protein interactors for the circRNA of interest. Make sure the target circRNA has successfully been enriched before proceeding with these analyses.

Limitations

One limitation of this protocol is the inevitable enrichment of the linear counterpart (lin-Dlc1), although to a lesser extent with respect to the circRNA. This is because the regions being targeted by the majority of the antisense probes are in common with both isoforms (see Figure 1B). Nevertheless, the ratio between the number of probes and the size of the molecule being pulled down (coverage), together with the use of a BSJ-specific probe favors the pull-down of the circular isoform and are likely responsible for its greater enrichment. In addition, circRNAs are likely more accessible to probe binding, compared to for example linear mRNA that is bound by other proteins/complexes such as ribosomes.

The interactors (RNA or protein) identified via the circRNA pull-down may therefore not be specific for the circular isoform, and their interaction with the linear counterpart cannot entirely be excluded. To ensure specificity of interaction with the circular isoform only, a pull-down with probes targeting the linear counterpart exclusively can be performed in parallel as an additional control.

Troubleshooting

Problem 1

High non-specific binding to the circ-specific probes and/or negative control probes.

Potential solution

Increase number and/or duration of washes after bead binding (related to step 31). The salt and/or detergent concentration can also be increased in the HB used for the washes.

Problem 2

Poor enrichment of the circRNA of interest.

Potential solution

The poor enrichment of the circRNA could be due to a number of variables, including a low expression of the target circRNA in the chosen district. Increasing the amount of starting material for each pull-down could prove to be useful in this case and/or choosing a brain region where the expression of the circRNA is higher. Sample degradation could also be responsible for a poor enrichment. Degradation is inevitable, even more so when using RNase-rich tissues as starting material but can be minimized by following specific precautions such as working quickly, keeping buffers and samples on ice when required, and finally using the appropriate RNase and protease inhibitors. The input can be quantified after RNA extraction using a nanodrop in order to determine the amount of material to assess via qRT-PCR. Sample ‘input’ aliquots taken at different time points during the procedure could help identify when the majority of degradation takes place. In addition, inefficient probe hybridization to the circRNA can reduce overall pull-down efficiency. Use online prediction tools such as RBPMap¹⁰ for RNA binding proteins and IntaRNA¹¹ for RNA-RNA interactions to identify putative regions of the circRNA involved in interactions with other molecules. These should be excluded when designing the probes, as probe binding to these regions may result in loss of interactions between the circRNA and other molecules. Indeed, in step 30 the supernatant representing the ‘unbound’ fraction can be collected to assess RNA pull-down efficiency.

Problem 3

Protein pellet not seen.

Potential solution

If this occurs after the TCA centrifugation (step 39), remove the supernatant very carefully. You may leave a small volume behind, as long as the majority is replaced by ice cold acetone. The pellet should become easier to see after the acetone centrifugation (step 41). In some cases, the pellet may break up into smaller pieces or appear as a film on the side of the tube so take extra care when removing the supernatant. A small volume can be left to air-dry. The orientation of the microcentrifuge tube can also be marked to help find the pellet.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Mariangela Morlando (mariangela.morlando@uniroma1.it).

Technical contact

Technical questions on executing this protocol should be directed to and will be answered by the technical contact, Valentina Silenzi (valentina.silenzi@uniroma1.it).

Materials availability

This study did not generate new unique reagents.

Data and code availability

This study did not generate or analyze datasets or code.

Acknowledgments

We would like to thank Carmine Nicoletti for technical help with sample collection and Manuela Caruso for assistance. This work was partially supported by grants from Fondi di Ateneo Ricerca di base Università di Perugia to M.M. and ERC-2019-SyG 855923-ASTRA; AIRC IG 2019 Id. 23053; and “National Center for Gene Therapy and Drugs based on RNA Technology” (CN00000041), NextGeneration EU PNRR MUR, to I.B.

Author contributions

Conceptualization: V.S. and M.M. Methodology: V.S. Investigation: V.S. and N.S. Writing – original draft: V.S. and N.S. Writing – review and editing: V.S., I.B., and M.M. Funding acquisition: I.B. and M.M.

Declaration of interests

The authors declare no competing interests.