Mapping phase diagram of tau-RNA LLPS under live cell coculturing conditions

Abstract

Protein Liquid liquid phase separation (LLPS) has been associated with biological functions and pathological aggregation. Mapping the phase separation conditions is the first step to identify and quantify the driving forces of LLPS. Here we describe the protocols to draw the phase diagram of tau-RNA LLPS and use the mapped diagram to guide experimental conditions for LLPS-cell coculturing, electron resonance spectroscopy in particular Double Electron Electron Resonance spectroscopy, crosslinking immunoprecipitation, as well as isothermal titration calorimetry.

1. Introduction

Besides soluble state and solid-like aggregates, proteins have been observed to undergo liquid liquid phase separation (LLPS) in vivo and in vitro. LLPS yields a protein-dilute phase separated from a protein-rich phase, which is often observed in the form of liquid droplets. Some droplets have been linked to membraneless organelles that facilitates essential biological functions [1]–[3], while others are known to initiate and/or promote amyloid aggregation that is associated with pathologies [2], [4, p.], [5], [6]. In recent years, our group and others found tau undergoes LLPS and form droplets in vitro and in vivo [7]–[11]. Specifically, tau forms electrostatically driven complex coacervates upon association with polyanions, such as RNA, or hydrophobically driven LLPS upon entropy-driven dehydration. Given the functional and pathological significance of protein droplets, it is therefore important to understand the physical nature and the driving forces of tau LLPS. Here we present detailed protocols for the electrostatically driven (ED) LLPS of tau, referred to as LLPS-ED, as well as key characterization methods.

Identifying and quantifying the driving forces of protein LLPS both rely on mapping the conditions where the phase separation occurs. Tau-RNA LLPS can be treated as a three-component two-phase thermodynamic equilibrium. The three components are tau-RNA, salt ions and water. According to the Gibbs phase rule, there are three degrees of freedom to modulate tau-RNA complex coacervation, belonging to LLPS-ED. We choose temperature, T, salt concentration, [salt], and tau-RNA concentration, [tau-RNA], as the three independent experimentally readily accessible variables. By keeping one of the three variables constant, while varying the remaining two, we can obtain a map of a two-dimensional plane, showing the coexisting conditions, i.e. the phase diagram.

Note that we constrain one degree of freedom: the ratio between tau and RNA. Under certain circumstances when the tau:RNA ratio is of interest to observe and control, the tau-RNA complex coacervation system can be treated as a four-component two phase system, and the tau:RNA ratio will be the additional one degree of freedom. The experiment protocol here is still applicable for mapping such system.

2. Materials

1. Recombinant tau proteins can be expressed and purified following published methods in [12]. For demonstration of the methods in this chapter, we used tau187 (residues 255–441 of 2N4R). To avoid interference of disulfide bonding, we introduced C291S and C322S mutations to the protein, resulting in a cystein-free construct.

2. For demonstration, we used polyuridylic acid potassium salt (PolyU, Sigma P9528) for preparing tau-RNA complex coacervates. Other single stranded nucleic acids can be used, including polyadenylic acid potassium salt (PolyA, Sigma P9403) as well as tRNA (Sigma). All have been shown to form tau-RNA complex coacervates [13], [14].

3. For demonstration of coacervate-cell coculturing, we used H4 Neuroglioma cells (ATCC^® HTB-148), DMEM media supplemented with 10% FBS, 100μg/ml penicillin/streptomycin, Ibidi u-slide 8 well glass bottom (80827), as well as Alexa Fluor^™ 488 5-SDP ester.

4. For spin labeling and EPR spectroscopy, we used MTSL ((1-Acetoxy-2,2,5,5-tetramethyl-δ-3-pyrroline-3-methyl) Methanethiosulfonate; CAS 81213-52-7; Toronto research chemicals) and PD-10 desalting column (GE).

5. For PAR-iCLIP, we used tau antibody HJ 8.5 (gifts from Dr. Dave Holtzman, Washington University); GFP rabbit antibody (5 mg/ml, Abcam); anti-CDK5 (0.2 mg/ml, Santa Cruz).

3. Methods

3.1. Preparation of Tau-RNA complex coacervate

1. Prepare working buffer (WBuffer): 20 mM HEPES, pH 7.0, filtered with 0.2 μm filter, and stored in 4 °C.

2. Prepare tau stock. For tau187: 1 mM in WBuffer. For 2N4R: 500 μM in WBuffer with additional 100 mM NaCl (See Notes 1). Tau stocks should be stored in -80 °C for less than 1 year. Same protocol applies to the following tau variants: wild type, C291S, C291SC322S, P301L and C291SP301L.

3. Prepare RNA stock. Dissolve 10 mg/mL RNA in nuclease-free water. Aliquot and store in -80 °C for less than 1 year. When thawed, consume within 1 hour. Do not re-freeze. (See Notes 2). Same protocol applies to PolyU RNA, PolyA RNA and tRNA.

4. Estimate the charge of tau protein at pH 7.0 using Innovagen’s Peptide Property Calculator (http://pepcalc.com/).

5. Mix tau and RNA. Freshly thaw both tau stock and RNA stock to room temperature. For 1 μM tau protein, use $x$ μg/mL RNA, where

   $$x=\text{tau}\text{protein}\text{net}\text{charge}\times \text{RNA}\text{monomer}\text{MW}\text{in}\text{kD}$$

   In an Eppendorf tube, mix desired volume of RNA stock, NaCl stock and WBuffer. As the last step, add desired volume of tau stock and mix again. When using 1~100 μM tau and corresponding concentration of RNA in WBuffer at room temperature, samples should become cloudy immediately after the last step of mixing. (See Figure 1)

6. Alternatively, tau-RNA coacervate samples can be prepared in a multiwell plate for high throughput analysis. Compatible plates should have round well to facilitate mixing, flat and clear bottom required for analysis, as well as non-ionic hydrophilic surface coating for chemical compatibility (e.g. Corning 3844). For measurements that lasts longer than 10 minutes, seal the plate using a transparent plate seal to avoid evaporation (For Corning 3844, use Microseal MSB1001). Same protocols above apply.

Figure 1. Mixing tau and RNA to prepare complex coacervates.

This figure and its caption are reprinted from ref [12], authored by Y. Lin et al, and licensed under the terms of the Creative Commons Attribution 4.0 International (CC BY 4.0).

3.2. Bright-field and Confocal microscopy

1. Prepare microscope slides. On a rinsed glass microscope slide, use two layers of double-side sticky tape separated by ~1 cm in parallel, forming a chamber.

2. Immediately following step 3.1, pipette 1 μL of the mixture in between the sticky tapes. Close the microscope slide with a cover slide gapped by the sticky tapes to generate a liquid sample region with consistent thickness.

3. Keep the microscope slide at room temperature for 10 minutes with the cover slide facing down. During this time the tau-RNA coacervates within the liquid sample region settled down onto the surface of the cover slide.

4. For bright-field microscope imaging, use an inverted compound microscope, e.g. Olympus IX70. Image the prepared microscope slide with the cover slide facing down. Adjust the microscope lens following Köhler illumination to ensure homogeneous brightness. Adjust the magnification to 10X to 40X. Bring the focal plane close to the surface of the cover slide, so that the edges of the droplets appear dark and the surroundings appear bright (Figure 2A). Take images and save in tiff format for analysis using ImageJ or MATLAB.

5. For confocal microscope imaging, adjust the focus to the surface of the cover slide and take images (Figure 2B).

6. Alternatively, both bright field and confocal microscope images can be taken directly from a multi-well plate immediately following step 3.1. A multi-well plate-compatible microscope is required. (see. Notes 3)

Figure 2. Representative bright field microscope images of tau-RNA complex coacervates.

(A) Bright field snapshots of droplets from 400 μM tau187 and 800 μg/ml polyA RNA showing 2 droplets seamlessly fusing (highlighted with red circle). (B) Confocal microscopy image of tau187 labeled with Alexa-488 of the droplets from the same sample as in (A). The images and text of this figure and its caption are reprinted from ref [13] authored by X. Zhang et al, and licensed under the terms of the Creative Commons Attribution 4.0 International (CC BY 4.0).

3.3. Turbidimetry

1. Before turbidity measurement, microscope images of the samples should be examined following step 3.2, to ensure turbidity signal comes from liquid droplets.

2. For room temperature turbidity reading, immediately following step 3.1, pipette 120 μL of the mixture in a 100 μL cuvette (Starna Scientific Ltd) and measure optical density at 500 nm wavelength (OD₅₀₀). It is important to use large sample volume to completely fill the measurement window, to avoid interference to the light path from bubbles or the water-air surface. Optical density can be measured using a photometer, such as Shimadzu UV-1601 spectrophotometer (Shimadzu Inc.). Turbidity is represented by OD_500.

3. For temperature variated turbidity reading, pre-chill tau stock, RNA stock, WBuffer and all other solutions on ice. Prepare tau-RNA mixture following step 3.1. Pipette 120 μL of the mixture in a 100 μL cuvette following the step 2 above. Seal the cuvette with a cuvette lid or scotch tape. Kept in 4 °C for 5 min before heating. Use a photometer equipped with a temperature controller. Set the heating and cooling temperature ramping rate to 1 °C/min.

4. Alternatively, turbidity can be read from a multi-well plate, immediately following sample preparation at step 3.1.6. A plate reader equipped with a temperature controller is needed.

3.4. Mapping phase diagram

1. For mapping salt concentration of the phase diagram boundary at room temperature, fix the tau-RNA concentration and gradually increase [NaCl] until turbidity vanishes. Here we use 10 μM tau187 and corresponding amount of polyU RNA for instance. Prepare tau-RNA mixture following step 3.1. Estimate the maximum [NaCl] for titration (in this example 100 mM). Prepare titrant of 10 times concentrated NaCl in WBuffer (in this example 1.0 M). Immediately after preparing tau-RNA mixture, record the turbidity following step 3.3.2. Add 1.0 μL titrant to the cuvette, mix well with a pipette and immediately record turbidity. Repeat 10~15 times until the turbidity reach baseline. (Note: when mixing sample using a pipette, avoid creating bubbles as they will block the measurement window of the cuvette and create artifacts in the turbidity readings).

2. Plot the [NaCl] as x-axis and turbidity as y-axis. Fit the data with

   $$y=\text{max}(0,ax+b)$$
   The fitting can be done in R using nlsLM function from minpack.lm package (Figure 3). The [NaCl] of the phase diagram boundary will be

   $${\left[\text{NaCl}\right]}_{\text{max}}=-\widehat{b}/\widehat{a}$$

   where $\widehat{a}$ and $\widehat{b}$ are results from the fitting.

3. Change tau-RNA concentration by changing [tau] while keeping tau:RNA ratio the same. Repeat step 1 and 2 to obtain ${\left[\text{NaCl}\right]}_{\text{max}}$ at varying [tau] concentration. Plot [tau] as x-axis and ${\left[\text{NaCl}\right]}_{\text{max}}$ as y-axis. A phase diagram at room temperature is complete. For statistical power estimation, repeat each experiment independently for no less than 3 times. (Note: Titrating tau-RNA is not recommended, as titrant of highly concentrated tau-RNA mixture might lead to uncontrollable artifacts including protein aggregation and RNA degradation.)

4. For determining temperature of the phase diagram boundary, prepare designed concentrations of tau-RNA and salt following step 3.1, record temperature-variated turbidity following step 3.3.3. Plot T as x-axis, and turbidity as y-axis. Fit the data with

   $$\frac{\text{y}-\text{min}(y)}{\text{max}(y)-\text{min}(y)}=\frac{1}{1+exp\left(-k\times \left(x-T_{cp}\right)\right)}$$

   The fitting can be done in R using nlsLM function from minpack.lm package. The temperature of the phase diagram boundary, or cloud point, will be the fitted value of $T_{cp}$ (Figure 4).

Figure 3. Mapping phase separation salt concentration using least square regression.

Sample was prepared using 100 μM 2N4R and 160 μg/mL polyU RNA following step 3.1. Data was collected following step 3.4.1. Solid line shows the fit results following step 3.4.2. The maximum NaCl concentration is ~75 μM.

Figure 4. Phase diagram of tau-RNA complex coacervate.

(A). Turbidity of tau187-polyU RNA complex coacervate upon heating (Expt.1 ([tau], [NaCl]) = (50 μM, 120 mM), green dots; Expt.2 ([tau], [NaCl]) = (40 μM, 30 mM), purple dots). Optical density at λ = 500 nm were normalized and used as turbidity value. Turbidity-temperature data of the heating curves were fitted with a sigmoidal function (solid line) as described in step 3.4.5, and cloud point, T_cp was extracted from the fitting representing the temperature at which normalized turbidity reaches 0.5. (B)-(C). Experimental phase diagram (points) showing [tau] vs. T_cp and [NaCl] vs. T_cp along with the binodal curve generated from fitting the data to Flory Huggins-derived model. The images and text of this figure and its caption are reprinted from ref [12], authored by Y. Lin et al, and licensed under the terms of the Creative Commons Attribution 4.0 International (CC BY 4.0).

3.5. Droplet-cell co-culturing

1. Culture H4 neuroglioma cells in DMEM supplemented with 10% FBS, 100μg/ml penicillin/streptomycin (“DMEM-10”) in a humidified atmosphere of 5% CO2 at 37°C. Other cell types can also be used (i.e. HEK293T).

2. Sub-culture H4 cells into an 8-well glass bottom chamber slide, plating ~25,000 cells/well and allow cells to adhere to surface overnight.

3. Label tau protein (Tau187, K18 or 2N4R) with fluorescent tag for visualization (ex. Alexa Fluor^™ 488 5-SDP ester) following manufacturer’s instructions (See Notes 4).

4. Prepare Tau-RNA complex coacervates as described in step 3.1. Replace WBuffer with DMEM-10 and to aid in visualization use a mix of labeled to unlabeled (1:20) tau protein.

5. Once coacervate mixture from step 4 has been prepared, aspirate media from cells and replace with 50–100μL of coacervate mixture. To maintain health of the cells, this step should be done quickly and the coacervate volume should be sufficient to fully cover the cells.

6. Image cells immediately using a fluorescent microscope, ideally using a temperature-controlled stage to maintain desired temperature throughout the experiment.

3.6. Spin labeling for electron paramagnetic resonance (EPR) spectroscopy

Tau proteins were engineered to have cysteines on desired sites, either through site-directed mutagenesis or DNA synthesis. After protein purification, spin labels are tethered to the cysteines in order to perform EPR spectroscopy. This can be achieved with the following steps:

1. Reduce cysteine with a reducing agent. TCEP is preferred to DTT since the latter attacks and reduces spin labels [15]. Incubate tau with 5mM TCEP at room temperature for at least 2h.

2. Remove reducing agent using size exclusion method. A PD-10 desalting column can be typically used, although other columns can be used depending on the sample volume and the protein size. Two desalting steps can be used in certain conditions (see Notes 5).

3. Add ~10 times molar excess of MTSL, compared to concentration of cysteine. Incubate the mixture overnight at 4 °C, or for at least 2 h at room temperature. (Note: addition of MTSL needs to be done promptly after removal of TCEP in order to limit re-oxidation of the cysteines.)

4. Remove excess of free MTSL using a size exclusion method. PD-10 desalting columns are typically used.

5. Measure spin labelling efficiency by measuring independently the protein concentration, with UV spectrometry for instance, and the spin label concentration, with quantitative EPR. The spin labelling efficiency is calculated by

   $$\frac{\text{molar}\text{concentration}\text{of}\text{spin}\text{labels}}{\text{number}\text{of}\text{cysteines}\text{per}\text{protein}\times \text{protein}\text{molar}\text{concentration}}$$

6. Samples can be flash-frozen and stored at high concentration in a freezer.

3.7. Double Electron Electron Resonance (DEER).

Once the protein is spin labelled, one can apply EPR and in particular double electron electron resonance spectroscopy (DEER). DEER allows to capture protein structural features by measuring intramolecular distance distributions [14], [16], [17]. For these measurements, DEER is applied on protein mutants containing two cysteines, i.e. two spin labels per molecule. DEER is performed with following steps:

1. Exchange the buffer with a D₂O-based buffer. This step can be combined with step 3.6.4.

2. Non labelled tau protein (noSL-prot; for instance a mutant containing no cysteine) is mixed with a spin labelled protein (SL-prot) at a molar ratio [SL-prot]:[noSL-prot] of about 1:10 to 1:20 (see Notes 6).

3. Prepare tau-RNA complex coacervates following step 3.1. Add 30% volume of D₈-glycerol for cryoprotection. Mix the sample thoroughly after preparation. Typical sample volumes are 28 μL of tau/tau-RNA coacervate sample and 12 μL of D₈-glycerol. (See Notes 7). It is important to verify that the addition of 30% cryoprotectant (step 3.7.3) does not disrupt phase separation. Turbidly measurements and bright filed microscopy should be applied before transferring the sample to the EPR tube.

4. Transfer 30–50 μL of sample into a quartz capillary tube of 3 mm OD and 2 mm ID, before flash freezing using liquid nitrogen.

5. After freezing, insert the sample into the sample holder of the DEER spectrometer.

6. Acquire DEER data. See Notes 8 for example pulse sequence.

7. Analyze DEER data to extract probability distribution of inter-spin distance. See Notes 9 for recommendations.

3.8. PhotoActivatable Ribonucleoside-enhanced Individual-nucleotide-resolution CrossLinking and ImmunoPrecipitation (PAR-iCLIP)

To verify direct interaction of tau with intrinsic RNA in cell, PAR-iCLIP can be performed.

1. Plate HEK 293T cells as monolayers in DMEM medium with 10% FBS and 100 μg/ml penicillin/streptomycin at the density of 1x 10⁶ cells per 10 cm dish, cultured in a humidified atmosphere of 5% CO2 at 37°C. To study intrinsic tau, neurons derived from human induced pluripotent stem cells (hiPSCs) or SH-SY5Y Neuroblastoma cells can be applied with proper media and culture condition.

2. Transfect human embryonic kidney (HEK 293T) cells with tau 2N4R-GFP construct using Lipofectamine 2000 (ThermoFisher) according to the manufacturer’s instruction.

3. UV crosslinking. Treat the cells with 4-thiouridine (4SU) (Sigma-Aldrich) for 1h at a final concentration of 500 μM at 37°C. Rinse with ice-cold 1x PBS and irradiate one time with 400 mJ/cm² of 365 nm UV light on ice. Centrifuge the cells and store the pellet at -80°C. Cells for control experiments should be prepared in parallel in the same manner (See Note 10).

4. Prepare beads for immunoprecipitation. Add 40 μL protein A/G plus beads (Thermo Scientific) to a spin column (Thermo Scientific) and wash 4 times with 500 μL conjugation buffer (1x PSB made from 10x buffer using RNase-free water). Add anti-tau antibody (HJ 8.5, 20 μL, 4.5 mg/ml) or anti-GFP (abcam, 1.5 μL, 5 mg/ml) or anti-CDK5 (Santa Cruz, 40 μL, 0.2 mg/ml) to the beads in 800 μL of conjugation buffer and rotate for 1 hour at room temperature. Filter the supernatant and wash the beads twice with 800 μL conjugation buffer. Crosslink the antibody to the beads using 800 μL of 450 μM DSS (Thermo Scientific) in conjugation buffer and rotate for 1 hour at room temperature. Discard the supernatant and quench the reaction with 800 μL 0.2 M Tris buffer pH 7.4 for 15 minutes at room temperature. Wash the beads 3 times with 600 μL IP buffer (50 mM Tris pH 7.4, 150 mM NaCl, 1 % NP40) and resuspend in 0.8 ml lysis buffer (50 mM Tris pH 7.4, 150 mM NaCl, 1% NP40 and 1x protease inhibitor from Roche).

5. Cell lysis and partial RNA digestion. Add lysis buffer (500 μL) to each cell pellet and incubate for 10 minutes on ice. Add Turbo DNase (2 μL, Thermos Fisher) per 500 μL lysate. Immediately place the samples at 37°C for exactly 3 minutes. Afterwards centrifuge the samples at 20,000 x g for 10 minutes at 4°C. Collect the supernatant without touching the pellet. Determine the protein concentrations of the supernatants using BCA protein assay kit (Thermo Scientific). Adjust the final concentration of protein with lysis buffer to 0.4 -1 mg in 500 μL. Input of the control is kept at the same amount for direct comparison.

6. Immunoprecipitation. Add lysate (500 μL) to the beads and rotate overnight at 4°C. The next day discard the supernatant and wash the beads twice with 600 μL cold high salt buffer (50 mM Tris pH 7.4, 1 M NaCl, 1% NP-40, 1 mM EDTA), each for 5 minutes, then twice with 600 μL wash buffer (20 mM Tris pH 7.4, 10 mM MgCl₂, 0.2% Tween-20) and leave the beads in the last wash. See Note 10 if DNA library preparation is intended.

7. For RNA 5’ end labeling, remove the supernatant and add 36 μL of hot PNK mix to the beads: 1.8 μL PNK (NEB); 3.6 μL PNK 10x buffer (NEB); 29.4 μL H₂O; 1.2 μL ATP [γ-³²P]. Incubate the mixture at 37°C for 5 minutes. Remove and discard the radioactive solution in the radioactive waste container.

8. Run 4–12% NuPAGE gel and transfer to nitrocellulose membrane. Mix the samples with loading buffer and heat for 5 minutes at 70°C. Run the NuPAGE gel at 180 volts for about 35–50 minutes with 20x MOPS-SDS running buffer (Thermo Fisher) and transfer the gel for 90 minutes at 30 volt with 20x transfer buffer (Thermo Fisher) to a nitrocellulose membrane (Protran BA85, VWR). Rinse the membrane twice with 1x PBS and blot briefly with Kimwipe at the back side, then wrap with parafilm. Mix blue marker dye with old radioactive ATP [γ-³²P] and unsymmetrically label a piece of paper board. This will be used as reference post phosphor imager scan. Place the blot on top of the paper board and wrap the packet in parafilm. Expose the packet to phosphor imaging screen for 1 hour. Scan the screen and analyze using Quantity One software (Bio-Rad).

3.9. RNA gel mobility shift assay

RNA gel shift assay can be applied to verify the association between tau and RNA.

1. Incubate full length tau with tRNA at 37 °C for 10 minutes in the presence of 0.5 mM EDTA, 0.5 mM MgCl_2, 2 U SUPERase• In^™ RNase Inhibitor (Thermo Fisher), 0.01% IGEPAL CA-630 (Sigma-Aldrich). Apply the mixture to a TBE 8% Polyacrylamide Gel (Thermo Fisher).

2. After gel separation, stain tRNA with SYBR Gold II (Thermo Fisher).

3. For quantitative analysis, the fraction of free and bound tRNA can be quantified in ImageJ (National Institute of Health).

3.10. Isothermal Titration Calorimetry (ITC)

ITC experiments can be carried out to quantify the affinity between tau and RNA. Here we full length tau 2N4R and tRNA for demonstration.

1. Dialyze tau overnight into a specified buffer for ITC (e.g. 20 mM ammonium acetate, pH 7). Resuspend tRNA (from Baker’s yeast, Sigma-Aldrich) in the ITC buffer and determine concentration using a Nanodrop 1000 (Thermo Scientific).

2. Run experiments on a Nano ITC (TA Instruments), in which 300 μM tRNA is titrated (5 μl injections) into a 1 ml protein solution of 30 μM Tau.

3. Analyze data using the NanoAnalyze v3.6 software (TA Instruments). After subtracting the heat generated by tRNA titration into empty buffer, fit the experimental data with an Independent binding model.

4. Notes

1. For 2N4R (and likely other long tau fragments), if the net charge of the tau protein is close to neutral, coacervation will happen at high protein concentration and low ionic strength. Therefore, it is important to store such protein stock with additional NaCl in the buffer.

2. RNA is sensitive to RNase in the environment. Consider following RNase control protocols.

3. For imaging using multiwell plate, use the wells that are near the center of the plate for better background homogeneity, and protect the microscope objective lens from being blocked by the edge frame of the plate.

4. We normally label tau protein with 5-SDP ester fluorescent dyes that react with primary amine residues on tau. We have found that overnight labeling at a 5x molar ratio of dye is sufficient for ~1 molar incorporation of label after clean-up.

5. PD-10 columns desalt at >98% efficiency according to manufacturer’s specifications. If 5 mM TCEP was present before desalting, one expects to obtain a residual concentration of up to 100 μM TCEP after desalting. This residual amount of reducing agent needs to be at least 5 times lower than the concentration of free spin labels added in the subsequent step of spin labelling. If this criterion is not matched, one need to apply a second desalting step in order to further lower the residual concentration of reducing agent.

6. Incorporation of non-spin labelled protein in the sample aims at preventing intermolecular interactions in the case where proteins are crowded (e.g. in tau-RNA complex coacervate). The higher the local protein concentration, the more noSL-prot needs to be used.

7. Sample volumes and concentration greatly depends on the instrument and the sensitivity achieved. Using Q-band Bruker E580 Elexsys spectrometer with a 300 W TWT amplifier, we routinely use 20–40 μM of spin label proteins in 40 μL volume, but one can go as low as 5 μM and still get a good signal-noise ratio.

8. There is a variety of available EPR pulse sequence. With our instrumentation (Q-band Bruker E580 Elexsys spectrometer with a 300 W TWT amplifier), the following parameters gave optimal signal/noise ratio. Four-pulse DEER experiments were carried out at 50 K, with the following pulse sequence: π_obs/2 – τ₁ – π_obs – (t-π_pump) – (τ₂-t) – π_obs – τ₂ – echo. Rectangular observe pulses were used with lengths set to π_obs/2 = 10–12 ns and π_obs = 20–24 ns. A chirp π_pump pulse was applied with a length of 20–24 ns and a frequency width of 60 MHz. The observe frequency was 90 MHz higher than the center of the pump frequency range. τ₁ was 180 ns and τ₂ was typically set to 2.4 μs

9. In order to analyze the data, one needs to remove the background signal originating from intermolecular interactions. In order to do so, it is recommended to measure DEER on singly labelled protein (reflecting only intermolecular interactions) and doubly labelled protein (containing both inter and intramolecular interactions) in the exact same conditions. The signal from the first sample can be used to correct the signal from the second sample in order to interpret only intramolecular distances. Another approach is to fit the signal from the doubly labelled protein at long time (last 1/3 of the signal) with the following equation:

   $$B\left(t\right)=V_0\times e^{-\alpha \times t^{\left(\frac{D}{3}\right)}}$$

   where D is typically set to 3, and V₀ and α are fitted. B(t) is then divided from the total signal in order to remove the contribution from intermolecular interactions. Interpretation of DEER signal is a complex and active field of research and one can follow different procedures to extract distance distributions. See for instance a genral review on DEER [18] and a detail protocol for the tau protein [Fichou et al, 2017 method in cell biology]. Interpretation of DEER signal is typically done using popular software such as DEER analysis [19] or LongDistance (https://sites.google.com/site/altenbach/labview-programs/epr-programs/long-distances).

10. Control CLIP experiments with GFP antibody for the lysate expressing GFP protein, or with HJ 8.5 antibody for lysates without expressing tau can be done in parallel to rule out false positive binding caused by the beads. If DNA library preparation from the crosslinked RNA is intended, RNA 3’ end need to be dephosphorylated and linked to DNA adapter before step 6 is carried out for radiolabeling RNA 5’ end. Check the original publication ref [13] for further details of DNA sequencing and analysis.