C9orf72 polyPR interaction with the nuclear pore complex

Abstract

The C9orf72 gene associated with amyotrophic lateral sclerosis/frontotemporal dementia is translated to five dipeptide repeat proteins, among which poly-proline-arginine (PR) is the most toxic in cell and animal models, contributing to a variety of cellular defects. It has been proposed that polyPR disrupts nucleocytoplasmic transport (NCT) through several mechanisms including accumulation in the nuclear pore complex (NPC), accumulation in the nucleolus, and direct interactions with transport receptors. The NPC, which is the key regulator of transport between the cytoplasm and nucleus, plays a central role in these suggested mechanisms. Exploring polyPR interaction with the NPC provides valuable insight into the molecular details of polyPR-mediated NCT defects. To address this, we use coarse-grained molecular dynamics models of polyPR and the yeast NPC lined with intrinsically disordered FG-nucleoporins (FG-Nups). Our findings indicate no aggregation of polyPR within the NPC or permanent binding to FG-Nups. Instead, polyPR translocates through the NPC, following a trajectory through the central low-density region of the pore. In the case of longer polyPRs, we observe a higher energy barrier for translocation and a narrower translocation channel. Our study shows that polyPR and FG-Nups are mainly engaged in steric interactions inside the NPC with only a small contribution of specific cation-pi, hydrophobic, and electrostatic interactions, allowing polyPR to overcome the entropic barrier of the NPC in a size-dependent manner.

Figure 1

Coarse-grained models of yeast NPC with FG-Nups and polyPR. (a) Snapshot of the NPC, showing the FG-Nups attached to the inner pore region. Each FG-Nup type is color coded, and the NPC scaffold is represented in gray. (b) Side-view snapshot of the simulation box cross section, featuring FG-Nups in yellow, polyPR in red, and the NPC scaffold in gray. To enhance visualization, the bead size for polyPR chains has been increased.

Figure 2

Simulation results for PR20 translocation through the NPC. (a) Snapshots taken between 0 and 1.5 μs (from left to right) showing PR20 translocation through the NPC. (b) Top panel: the time evolution of PR20 counts within the cytoplasmic side, NPC, and nucleoplasmic side based on the center of mass positions of PR20 chains. Bottom panel: the vertical positions of the centers of mass of 25 copies of PR20 during the simulation. The NPC region ($\left|z\right|<15.4$ nm) is highlighted with a pink box. The color bar represents the radial positions of the PR20 chains.

Figure 3

Density distribution of FG-Nups and polyPR, and energy barrier for polyPR translocation. (a) Time-averaged 2D density map of the FG-Nups. The scaffold is depicted in white. (b) Left panels: time-averaged normalized 2D density maps for PR20 and PR50. Right panels: time-averaged normalized density of PR20 and PR50 in the $z$ direction $n\left(z\right)$ (in black), and the corresponding energy barrier $e\left(z\right)$ calculated from Boltzmann inversion using $r_0=$ 2 nm (in red), see methods. Polynomials of 12th order are used for curve fitting. (c) The energy barrier in the $z$ direction for different polyPR lengths as a function of $r_0$.

Figure 4

PR20 contact with FG-Nups and individual residues. (a) Time-averaged number of contacts between PR20 and different types of FG-Nups within the pore region ($\left|z\right|<15.4$ nm). For each case, the number of contacts is normalized by the length of the FG-Nup, the number of FG-Nup copies, and the polyPR length. (b) Time-averaged number of contacts between PR20 and different residue types within the nuclear pore region ($\left|z\right|<15.4$ nm). The number of contacts is normalized by the polyPR length, see methods for more details. The error bars show the standard deviation from block averaging using five blocks at equilibrium.