Using Solution NMR to Characterize Biomolecular Condensates Under Biphasic Conditions

We developed NMR methods to study biomolecular condensates. These are key for cellular organization, but their ever information is implied in multiple disorders. Using our methods, we investigate condensate material properties, analyze the mechanism of their formation, and monitor long-term changes such as condensate maturation or pathological fibrillation of biomolecules.

To begin, prepare the final biphasic sample stabilized by agarose hydrogel by mixing the protein stocks, RNA stocks, or both with the respective warm agarose buffer in a 1.5 milliliter micro centrifuge tube. The final protein, or RNA concentration is typically 50 to 500 micromolar for NMR detection. Insert the prepared biphasic sample into a spectrometer, then lock, shim, match, and tune the proton channel as usual.

Calibrate the 90 degree pulse length on the proton channel using the general calibration procedure. Then record a one-dimensional water suppressed spectrum and check if the spectrum contains strong biomolecular signals. If no or very little signal is observed, do not proceed with REDIFINE, but use the CONDENSE-MT instead.

Record a diffusion ordered spectroscopy experiment to array the diffusion gradient strength from 2 to 95%in 16 linear steps. To continue acquiring data for the full REDIFINE set, create nine new step PPP GPQ S19 experiments by copying the parameters from the initial experiment that was just run. In each consecutive experiment, change only the D20 diffusion delay in variable steps to span the range from 100 to 1000 milliseconds.

Run each of these experiments with the suitable command. Use the last N to preserve the receiver gain throughout the 10 experiments. Set Q Sign with SSB2 as the window function.

Then extract the fifth slice using the appropriate command. Phase the one dimensional spectrum and save the phasing parameters to the two-dimensional dataset. Process the two-dimensional dataset with the XF2 and perform baseline correction with ABS2.

After that, extract the 16 slices with split 2D. Repeat the similar procedure for the remaining nine experiments. Check at least the first four to five slices to ensure correct phasing and baseline correction.

If needed, correct the phasing and baseline manually. Then use the suitable MATLAB code for the analysis. In the code, modify the diffusion times, set the PG values according to the probe gradient strengths and assign experiments one through 10 based on the total number of experiments.

Then fit the data according to the REDIFINE model using this code. Record the 1D water suppressed spectrum. If very little signal is observed, proceed with CONDENSE-MT experiment.

Determine the R1 rate of water by T1 inversion recovery for quantitative CONDENSE-MT. Copy the provided pulse sequence and VD list to the respective folders. After calibrating the proton 90 degree pulse, set NS to 2, DS to 0, D1 to 20 seconds, SW to 10 parts per million.

TD2 to 16K and TD1 to 24. For the CONDENSE-MT experiment, copy the provided Bruker parameter set, sequence, and frequency list to the Bruker specific folders. Then adjust the frequencies and power according to the spectrometer.

Set P1 to the calibrated 90 degree pulse length and P4 to one microsecond for the excitation pulse to avoid radiation damping. Set NS to 2, DS to 16, D1 to four seconds, SW2 to 20 parts per million, TD2 to 8k, and TD1 to the length of the frequency list. Set SPNAM10 to Squa100, 000, and set P10 to five seconds.

Find the SP10 power that yields 100 hertz irradiations. Adjust the SP10 power of the irradiation pulse accordingly. Run the pseudo two dimensional experiment with the appropriate command and set up a series of these pseudo 2D experiments, varying SP10 irradiation powers.

After adjusting all the parameters, run the code to read through the NMR spectra, which integrates the water signal and save it into a new variable. In the code inversion recovery, adjust the path to the appropriate data and check the water integration region. Then adjust the experiment number containing the inversion recovery data.

Set all experimental parameters and enter the agarose empty pool values previously obtained by fitting the agarose dataset in a similar manner. Run the first section of the condensed empty code to obtain the fit for water longitudinal relaxation. Then run the condense empty fitting section to initiate the live figure showing the minimization of the fitting function.

Run the next section of the code to determine the uncertainties and plot the data. The REDIFINE fit determined the diffusion coefficients in the dilute and condensed phases, the population of protein in the condensed phase, the average droplet size and the interphase permeability for the FUS N-terminal domain sample. The NMR signal of the FUS N-terminal domain remained detectable in the condensed phase.

No difference in the magnetization transfer profile was observed between the agarose reference and the FUS N-terminal domain by phasic sample. For CAG repeat RNA condensates, no NMR signal is observed. Accordingly, the magnetization transfer profile is significantly different from the agarose reference.

The CONDENSE-MT experiment provided dynamic characterization of repeat expansion RNA condensates. The CONDENSE-MT analysis yielded the proton exchange kinetics between the condensate and bulk water. The CONDENSE-MT analysis also yielded the transverse relaxation time of the RNA in the condensate and the apparent transverse relaxation rate of bulk water.

In this protocol, we can characterize the condensate property in a biphasic state without the need of a chemical tag. The droplets are coexisting with a dilute environment, allowing exchange between the two phase mimicking the cellular environment. Our methods focused on condensate characterization can be complemented by conventional NMR methods to study molecular structure interactions and dynamics across phases.

In future studies, we aim to develop more complex condensate models, adapting the methodology to more complex systems, and eventually study condensate directly in their in-cellular environments.