ReSMAP: Web Server for Predicting Residue-Specific Membrane-Association Propensities of Intrinsically Disordered Proteins

Abstract

The functional processes of many proteins involve the association of their intrinsically disordered regions (IDRs) with acidic membranes. We have identified the membrane-association characteristics of IDRs using extensive molecular dynamics (MD) simulations and validated them with NMR spectroscopy. These studies have led to not only deep insight into functional mechanisms of IDRs but also to intimate knowledge regarding the sequence determinants of membrane-association propensities. Here we turned this knowledge into a web server called ReSMAP, for predicting the residue-specific membrane-association propensities from IDR sequences. The membrane-association propensities are calculated from a sequence-based partition function, trained on the MD simulation results of seven IDRs. Robustness of the prediction is demonstrated by leaving one IDR out of the training set. We anticipate there will be many applications for the ReSMAP web server, including rapid screening of IDR sequences for membrane association.

Keywords: amphipathic helix; intrinsically disordered proteins; intrinsically disordered regions; membrane binding; membrane-association propensity.

Figure 1

MD data and ReSMAP predictions for membrane-association propensities. (a) A snapshot of the ChiZ IDR associated to an acidic membrane, reprinted from Ref. [5]. The lipid headgroups in the leaflet in contact with the IDR are shown as surface; Arg side chains are shown as stick. (b–d) Comparison of MD membrane-contact probabilities (gray bars) and predicted membrane-association propensities (red curves) for ChiZ, N-WASP, and FtsQ, respectively. The sequence of each IDR is listed, with positively and negatively charged residues colored blue and red, respectively.

Figure 2

Comparison of ReSMAP-predicted membrane-association propensities (red curves) with experimental data (gray bars). (a,b) Prolactin receptor and growth-hormone receptor disordered intracellular domains. The experimental data are from Haxholm et al. [6], displaying $\Delta {\delta }_{\mathrm{NH}}=\sqrt{{\left(\Delta {\delta }_{\mathrm{H}}\right)}^2+0.154{\left(\Delta {\delta }_{\mathrm{N}}\right)}^2}$, where $\Delta {\delta }_{\mathrm{N}}$ and $\Delta {\delta }_{\mathrm{H}}$ are changes in backbone amide chemical shifts when the IDRs were moved from solution to vesicles formed by POPC/POPS lipids at a 3:1 ratio. $\Delta {\delta }_{\mathrm{NH}}$ values < 0.005 ppm were assumed to be within experimental error and set to 0; missing resonances were assigned a $\Delta {\delta }_{\mathrm{NH}}$ value of 0.25 ppm. The protein sequences are from Uniprot (https://www.uniprot.org/ (accessed on 28 July 2022) entries P16471 and P10912, with numbering shortened by 24 and 18 residues (i.e., without the n-terminal signal peptides), respectively. (c) Hck kinase disordered N-terminal region (residues G2-E79). The experimental data are from Pond et al. [11], displaying $\Delta {\delta }_{\mathrm{NH}}=\sqrt{{\left(\Delta {\delta }_{\mathrm{H}}\right)}^2+0.2{\left(\Delta {\delta }_{\mathrm{N}}\right)}^2}$ obtained when the IDR was moved from solution to bicelles formed with DMPC:DMPA lipids at a 4:1 ratio. (d) Synaptobrevin-2 residues M1-M96. For the experimental data [12], I₀ and I were measured, respectively, in solution and in the presence of liposomes formed by DOPC/DOPS/DOPE/Cholesterol at 5:2:2:1. Resonances missing in the presence of liposomes were assigned a zero value for I. (e) T-cell receptor ζ chain disordered intracellular domain. The protein sequence is from Uniprot entry P24161, with numbering shortened by 21 residues. The tall bars indicate basic-rich stretches where alanine mutations of basic residues abolished plasma membrane binding, as assayed by in-cell FRET [7]. The short bars indicate tyrosine-containing motifs where phenylalanine mutations of tyrosine residues had no effect on plasma membrane binding. (f) ADAM17 membrane-proximal domain (residues F581-E642). Bars indicate four residues (R625-K628) that experienced significant chemical shift perturbations in the presence of phosphorylserine [8].

Figure 3

Predicted membrane-association propensities of FtsB. (a) Predicted structure by AlphaFold (https://alphafold.ebi.ac.uk/ (accessed on 18 March 2022) [18]. The color spectrum displays confidence levels of prediction, ranging from red for high confidence to blue for low confidence. Three putative helices are indicated with start and end residues: amphipathic (am) helix, residues I60-K70; transmembrane (TM) helix, residues A83-A98; and coiled-coil, M111-Q131. (b) Predicted membrane-association propensities using the sequence of the first 82 residues, high for residues 9–48 but only modest for the putative amphipathic helix. The protein sequence is from Uniprot entry P96376.

Figure 4

Predicted membrane-association propensities of FtsL. (a) Predicted structure by AlphaFold (https://alphafold.ebi.ac.uk/ (accessed on 18 March 2022) [18]. The color spectrum displays confidence levels of prediction, ranging from red for high confidence to blue for low confidence. Three putative helices are indicated with start and end residues: amphipathic (am) helix, residues T78-A92; transmembrane (TM) helix, residues F124-T144; and coiled-coil, L153-D173. (b) Predicted membrane-association propensities using the sequence of the first 123 residues, high for residues 25–53 and the putative amphipathic helix. The protein sequence is from Uniprot entry O06213.