C9orf72-derived arginine-rich poly-dipeptides impede phase modifiers

Abstract

Nuclear import receptors (NIRs) not only transport RNA-binding proteins (RBPs) but also modify phase transitions of RBPs by recognizing nuclear localization signals (NLSs). Toxic arginine-rich poly-dipeptides from C9orf72 interact with NIRs and cause nucleocytoplasmic transport deficit. However, the molecular basis for the toxicity of arginine-rich poly-dipeptides toward NIRs function as phase modifiers of RBPs remains unidentified. Here we show that arginine-rich poly-dipeptides impede the ability of NIRs to modify phase transitions of RBPs. Isothermal titration calorimetry and size-exclusion chromatography revealed that proline:arginine (PR) poly-dipeptides tightly bind karyopherin-β2 (Kapβ2) at 1:1 ratio. The nuclear magnetic resonances of Kapβ2 perturbed by PR poly-dipeptides partially overlapped with those perturbed by the designed NLS peptide, suggesting that PR poly-dipeptides target the NLS binding site of Kapβ2. The findings offer mechanistic insights into how phase transitions of RBPs are disabled in C9orf72-related neurodegeneration.

Fig. 1. Arginine-rich poly-dipeptides impede the Kapβ2 function of modifying FUS phase transitions.

a Domain architecture of FUS, LC-domain of FUS (FUS-LC), and FUS-LC fusion NLS (FUS-LC:NLS). b Graphical representation of FUS droplet formation for microscope observation and turbidity assay. c Turbidity of 8 μM MBP:FUS in the presence of buffer, ±8 μM Kapβ2, and ±8 μM PR20/GR20/PA20/GP20/GA20. OD 395 nm is normalized to measurement of MBP:FUS + buffer + Tev. Mean of three technical replicates, ±SD. d Microscopic images of FUS droplets in the absence and presence of Kapβ2 and/or PR20 show that FUS droplets were dissolved by Kapβ2 and not melted in the presence of PR20. Mixture of 7.6 μM MBP:FUS and 0.4 μM MBP:FUS:EGFP were treated with TEV for an hour in the presence or absence of 16 μM Kapβ2, 16 μM Kapβ2–M9M complex, and 50 μM PR20. The experiment was independently repeated three times with similar results. Here, 10 μm scale bars are shown in the image. e Graphical description of hydrogel binding assay. Hydrogels of mCh:LC-domain accumulate GFP:LC-domain as it co-polymerizes. f Hydrogel binding assay for FUS-LC in the absence and presence of Kapβ2. mCh:FUS-LC (lower images) were incubated with 1.0 µM of GFP (left panel) or 1.0 µM of GFP:FUS-LC:NLS (right panel) in the presence of different concentrations of Kapβ2 (left to right: 0.1, 0.3, and 1.0 µM). g Quantitative analysis of Fig. 1f. Relative intensity of GFP signals is shown as the mean of three independent experiments ± SD, analyzed by one-way ANOVA followed by Dunnett’s multiple comparison test (**P < 0.01, ***P < 0.001). h Hydrogel binding assay for FUS-LC in the absence and presence of Kapβ2 and/or PR20. Hydrogel droplets of mCh:FUS-LC (lower images) were incubated with 1.0 µM of GFP (left panel) or GFP:FUS-LC:NLS (right panel). GFP:FUS-LC:NLS containing Kapβ2 (1.0 µM) was challenged for homotypic polymer extension in the absence or presence of different concentration of PR20. i Quantitative analysis of Fig. 1h. Relative intensity of GFP signals is shown as the mean of three independent experiments ± SD, analyzed by one-way ANOVA followed by Dunnett’s multiple comparison test (**P < 0.01). Source data are provided as a Source Data file.

Fig. 2. Interaction between Kapβ2 and PR poly-dipeptides.

a Immunoprecipitation showing interaction between endogenous Kapβ2 and HA:SBP:GFP:PR20:HA expressed in HeLa cells. b Pull-down binding assay showing interaction between GST:Kapβ2 and MBP:PR18/MBP:PR8. c Dissociation constant (K_d) measured by ITC of Kapβ2 (ΔLoop) binding to MBP:PR18. d SEC-MALS of Kapβ2 in the absence and presence of MBP:PR18, showing a 1 : 1 complex formation between Kapβ2 and MBP:PR18. The experiments (a, b) were independently repeated three times or more with similar results. Source data are provided as a Source Data file.

Fig. 3. Solution NMR of Kapβ2 and interaction between Kapβ2 and PR20.

a The crystal structure of Kapβ2 in complex with NLS of FUS. NLS of FUS is presented as a green cylinder and Kapβ2 is shown as a gray ribbon, with spheres representing methyl groups of Ala (pink), Ile (green), Leu (blue), Val (yellow), and Met (orange). Kapβ2 is enriched in methyl-baring residues, indicating that sufficient information can be collected from the ¹H-¹³C-correlated methyl NMR spectra. b ¹H-¹³C-correlated methyl NMR spectra of [U-²H; Ile-δ1-¹³CH₃; Leu, Val-¹³CH₃/¹²CH₃]-labeled Kapβ2. The spectral regions for the resonances of Ile, Met, Val, and Leu methyl groups are labeled. c Interaction between Kapβ2 and PR poly-dipeptide investigated by solution NMR. ¹H-¹³C-correlated methyl NMR spectra of [U-²H; Ile-δ1-¹³CH₃; Leu, Val-¹³CH₃/¹²CH₃]-labeled Kapβ2 in the absence (gray) and presence (blue) of PR20. Significant perturbations were observed for several resonances, indicating that PR poly-dipeptides bind to specific region(s) of Kapβ2. The perturbed resonances are indicated by peak numbers. The peak numbering corresponds to those in Supplementary Fig. 3.

Fig. 4. Interaction between Kapβ2 and PR20/M9M/FUS-ΔNLS investigated by solution NMR.

¹H-¹³C-correlated methyl NMR spectra of [U-²H; Ile-δ1-¹³CH₃; Leu,Val-¹³CH₃/C²H₃]-labeled Kapβ2 in the absence (gray) and presence (blue) of PR20 (a), in the absence (gray) and presence (green) of M9M (b), or in the presence of M9M (green) and the presence of M9M and FUS-ΔNLS (magenta) (c). Significant representative perturbations are indicated by arrow heads. Perturbations only seen for PR20 are indicated by red arrow heads. Perturbations only seen for M9M are indicated by green arrow heads. Perturbations common to PR20 and M9M are indicated by purple arrow heads. For clarity, only the region of the Ile methyl resonances is shown. The full-range spectra are shown in Supplementary Fig. 4 (for a) and Supplementary Fig. 5 (for b and c). Chemical shift differences (d–f) and intensity ratios (g–i) of the methyl resonances of Kapβ2 by the interaction with PR20 (d, g), M9M (e, h), and FUS-ΔNLS (f, i) are shown. In d–f, resonances that disappeared with the addition of a binding partner are indicated by asterisks. In f and i, resonances that disappeared with the complex formation with M9M are indicated by gray bars. In order to compare the perturbations, peaks on the spectra are labeled in peak numbers as shown in Supplementary Fig. 3. Less significant perturbations with the addition of PR poly-dipeptides than the addition of M9M can be explained by lower binding affinity of PR poly-dipeptides for Kapβ2 (Supplementary Table 1). j Selected views of the interaction between Kapβ2 and PR20/M9M/FUS-ΔNLS. Selected views of the representative resonance from ¹H-¹³C-correlated methyl NMR spectra of [U-²H; Ile-δ1-¹³CH₃; Leu, Val-¹³CH₃/C²H₃]-labeled Kapβ2 in the absence (gray) and presence of PR20 (blue), M9M (green), and M9M and FUS-ΔNLS (magenta). Graphical representation corresponds to the interaction between Kapβ2 and PR20/M9M/FUS-ΔNLS. k An expanded view of the electrostatic surface potential of Kapβ2. Positive and negative surface potentials are drawn in blue and red, respectively. NLS of FUS, which is compatible with M9M, is represented as a stick model, colored pale green.

Fig. 5. The model of interaction between PR poly-dipeptides and Kapβ2.

a Electrostatic potential of Kapβ2 and MD calculated model of PR poly-dipeptide inside of Kapβ2 cavity are shown. Blue and red colors show the positively charged and negatively charged regions, respectively. PR poly-dipeptide is colored blue. b Close-up view of Kapβ2 cavity. Methyl groups located close to the Kapβ2 cavity, isoleucine, leucine, valine, and methionine, are represented as spheres colored green, blue, yellow, and orange, respectively. Amino acids within 5 Å around the PR poly-dipeptide are colored cyan. c Pull-down binding assay showing the competition of PR20 and NLS of FUS to GST:Kapβ2. The experiment was independently repeated twice with similar results. Source data are provided as a Source Data file.

Fig. 6. Graphical abstract for the model of interaction between PR poly-dipeptides and Kapβ2.

FUS is prone to self-associate (left). Kapβ2 modifies a phase transition of FUS by recognizing NLS (middle). PR poly-dipeptides partially bind to the NLS-binding site of Kapβ2 and impede its ability (right).