Molecular Dynamics Simulations of Phosphorylated Intrinsically Disordered Proteins: A Force Field Comparison

Abstract

Phosphorylation is a common post-translational modification among intrinsically disordered proteins and regions, which helps regulate function by changing the protein conformations, dynamics, and interactions with binding partners. To fully comprehend the effects of phosphorylation, computer simulations are a helpful tool, although they are dependent on the accuracy of the force field used. Here, we compared the conformational ensembles produced by Amber ff99SB-ILDN+TIP4P-D and CHARMM36m, for four phosphorylated disordered peptides ranging in length from 14–43 residues. CHARMM36m consistently produced more compact conformations with a higher content of bends, mainly due to more stable salt bridges. Based on comparisons with experimental size estimates for the shortest and longest peptide, CHARMM36m appeared to overestimate the compactness. The difference between the force fields was largest for the peptide showing the greatest separation between positively charged and phosphorylated residues, in line with the importance of charge distribution. For this peptide, the conformational ensemble did not change significantly upon increasing the ionic strength from 0 mM to 150 mM, despite a reduction of the salt-bridging probability in the CHARMM36m simulations, implying that salt concentration has negligible effects in this study.

The Effect of Multisite Phosphorylation on the Conformational Properties of Intrinsically Disordered Proteins

“…The SN15 data are obtained from Ref. [20] (2020 American Chemical Society), and data for the phosphorylated variants of Tau2, bCPP, and Stath from [31].…”

“…Average radius of gyration (R g ) and end-to-end distance (R ee ) of the non-phosphorylated (non-phos) and phosphorylated (phos) variants. Data for SN15 are obtained from [20] and for the phosphorylated peptides of Tau2, bCPP, and Stath from [31]. The shape factor, presented in Figure 2, can be used as an estimate of the shape of the peptide.…”

“…Notice also that, while it is hard to make quantitative comparisons with CD data, our study on SN15 suggested that the simulations underestimate the structural content [20], which is the same as observed for Tau2. The data for the phosphorylated peptide are previously published in [31]. The positions of phosphorylated and positively charged residues are highlighted in yellow and blue, respectively; (g) change in contact probability upon phosphorylation of Tau2; (h,i) energy landscapes and conformations in minima of non-phosphorylated and phosphorylated Tau2.…”

“…Less attention has been given to charge-charge interactions, although it has been determined that many standard force fields have a tendency to overestimate salt bridges [29,30]. More recently, it has been shown that, for phosphorylated peptides, this can cause serious discrepancies between simulations and experiments [10,20,31].…”

“…In our most recent work involving all-atom molecular dynamics simulations of phosphorylated disordered peptides, Amber ff99SB-ILDN in combination with the TIP4P-D water model showed promising results in describing the conformational ensemble of short disordered peptides [20,31]. Here, we have extended the work with simulations of the non-phosphorylated variants of the four peptides in [31], using the aforementioned force field, and additional analyses of a fifth peptide published in [20], to study the conformational and structural effects upon phosphorylation, with the aim of gaining better insight into the controlling factors. By experimental comparison, we also detect limitations of the force field.…”