Scripts for simulation and plotting of thermodynamic model for cooperative RNA-binding of multidomain proteins

Publication

Stitzinger S.H., Sohrabi-Jahromi S., and Söding J. (2023) Cooperativity boosts affinity and specificity of proteins with multiple RNA-binding domains. NAR Genomics & Bioinformatics.

Abstract

Numerous cellular processes rely on the binding of proteins with high affinity to specific sets of RNAs. Yet most RNA-binding domains display low specificity and affinity in comparison to DNA-binding domains. The best binding motif is typically only enriched by less than a factor 10 in high-throughput RNA SELEX or RNA bind-n-seq measurements. Here, we provide insight into how cooperative binding of multiple domains in RNA-binding proteins (RBPs) can boost their effective affinity and specificity orders of magnitude higher than their individual domains. We present a thermodynamic model to calculate the effective binding affinity (avidity) for idealized, sequence-specific RBPs with any number of RBDs given the affinities of their isolated domains. For seven proteins in which affinities for individual domains have been measured, the model predictions are in good agreement with measurements. The model also explains how a two-fold difference in binding site density on RNA can increase protein occupancy ten-fold. It is therefore rationalized that local clusters of binding motifs are the physiological binding targets of multi-domain RBPs.

Simulation script

rbp_model.py implements the setup of the simulations and the calculations based on our analytical approach. The parameters of the proteins are loaded from text files in the folder examples.

Usage with command-line interface

The prediction of the total K_d can be started from the command line, using the script get_total_kd.py.

Usage:
  python3 get_total_kd.py [-h] [--simulate] [-n REPLICATES] [--plot] parameter_path

positional arguments:
  parameter_path	path to a file that stores parameters of the protein

optional arguments:
  -h, --help            
  -s, --simulate	run gillespie simulations to predict the Kd;
  			without this option, only the calculations are used
  -n REPLICATES, --replicates REPLICATES
			number of replicate simulations, over which to
			average, default: 1
  -p, --plot		plot the trajectory of the simulation
  			and save the plot to PDF, next to the parameter file

For example, to run a simulation of the protein hnRNP A1, averaged over 5 trajectories, and to show a plot of the trajectory after the simulation we can use

python3 get_total_kd.py ../examples/hnrnp_a1.csv -s -n 5 -p

Plotting script

To plot the figures in our paper the following functions from the script visualizations.py were used.

Figure 1 Many RBPs have more than one domain

rbp_distribution()
oligomer_distribution()

Figure 3C Effective concentration c₁₂ of domain 2 at RNA site 1, when at least one RNA site is already bound

ceff_L()

Figure 4 Comparison of experimental K_ds to predictions from our model are in good agreement

example_overview()

Figure 5A Dissociation constants decrease exponentially with the number of binding domains

kd_tot_N()

Figure 5B Individual domains contribute to the total K_d up to a threshold in the individual K_d

kd_tot_k3()

Figure 5C Dissociation constants decrease with the binding site density on the RNA

kd_motif_density()

Figure 5D Relative occupancy as a function of binding site density on the RNA

occupancy_linker_length()

Dependencies

The scripts require python3 with the following libraries:

• numpy
• scipy
• matplotlib
• gillespy2