Biophysical characterization of the phase separation of TDP-43 devoid of the C-terminal domain

Abstract

Background: Frontotemporal lobar degeneration with ubiquitin-positive inclusions (FTLD-TDP), amyotrophic lateral sclerosis (ALS) and limbic-predominant age-related TDP-43 encephalopathy (LATE) are associated with deposition of cytoplasmic inclusions of TAR DNA-binding protein 43 (TDP-43) in neurons. One complexity of this process lies in the ability of TDP-43 to form liquid-phase membraneless organelles in cells. Previous work has shown that the recombinant, purified, prion-like domain (PrLD) forms liquid droplets in vitro, but the behaviour of the complementary fragment is uncertain.

Methods: We have purified such a construct without the PrLD (PrLD-less TDP-43) and have induced its phase separation using a solution-jump method and an array of biophysical techniques to study the morphology, state of matter and structure of the TDP-43 assemblies.

Results: The fluorescent TMR-labelled protein construct, imaged using confocal fluorescence, formed rapidly (< 1 min) round, homogeneous and 0.5-1.0 µm wide assemblies which then coalesced into larger, yet round, species. When labelled with AlexaFluor488, they initially exhibited fluorescence recovery after photobleaching (FRAP), showing a liquid behaviour distinct from full-length TDP-43 and similar to PrLD. The protein molecules did not undergo major structural changes, as determined with circular dichroism and intrinsic fluorescence spectroscopies. This process had a pH and salt dependence distinct from those of full-length TDP-43 and its PrLD, which can be rationalized on the grounds of electrostatic forces.

Conclusions: Similarly to PrLD, PrLD-less TDP-43 forms liquid droplets in vitro through liquid-liquid phase separation (LLPS), unlike the full-length protein that rather undergoes liquid-solid phase separation (LSPS). These results offer a rationale of the complex electrostatic forces governing phase separation of full-length TDP-43 and its fragments. On the one hand, PrLD-less TDP-43 has a low pI and oppositively charged domains, and LLPS is inhibited by salts, which attenuate inter-domain electrostatic attractions. On the other hand, PrLD is positively charged due to a high isoionic point (pI) and LLPS is therefore promoted by salts and pH increases as they both reduce electrostatic repulsions. By contrast, full-length TDP-43 undergoes LSPS most favourably at its pI, with positive and negative salt dependences at lower and higher pH, respectively, depending on whether repulsive or attractive forces dominate, respectively.

Keywords: Electrostatics; LLPS; Liquid–liquid phase separation; Liquid–solid phase separation; Motor neuron diseases; RNA-binding proteins; Self-assembly.

Fig. 1

Confocal fluorescence microscopy images of PrLD-less TDP-43 assemblies. Representative confocal fluorescence images of PrLD-less TDP-43-TMR during self-assembly in 0 mM NaCl (A) and 150 mM NaCl (B). In the second and third rows, higher magnifications of the confocal fluorescence images are shown in the coloured boxed areas. The isolated images on the left were acquired with native PrLD-less TDP-43-TMR under conditions that do not promote phase separation

Fig. 2

FRAP of PrLD-less TDP-43 assemblies. Representative confocal fluorescence images of 5 μM PrLD-less TDP-43 labelled with Alexa-488 during self-assembly in 150 mM NaCl, and recorded at time 0 min, as in Fig. 1. The figure shows, in particular, representative images from FRAP experiments on two distinct classes of assemblies (top and bottom panels, respectively). Relative fluorescence intensity (RFI) after photobleaching was plotted over time for 3 distinct assemblies (n = 3) of each class which were averaged (red and green curves, respectively). Fluorescence values were normalized to pre-bleach fluorescence values (taken as 1.0) and to post-bleach fluorescence values (taken as 0.0). Error bars: SEM. The red continuous line represents the best fit of experimental data to the double exponential equation (Eq. 1) described in the Materials and methods section

Fig. 3

Spectroscopic characterization of PrLD-less TDP-43 self-assembly in 0 mM NaCl. A Time course of PrLD-less TDP-43 self-assembly monitored with turbidimetry under native (red) and self-assembly promoting (black) conditions. The number of replicates was 3 (n = 3). Error bars: SEM. The dashed line indicates the threshold of turbidimetry above which the signal is significantly high to indicate visible phase separation. B Far-UV CD spectra of PrLD-less TDP-43 during self-assembly at the indicated time points. The spectrum of native PrLD-less TDP-43 is also shown. C Intrinsic Trp fluorescence spectra of PrLD-less TDP-43 during self-assembly at the indicated time points. The spectrum of native PrLD-less TDP-43 is also shown

Fig. 4

Multi-parametric dependence of PrLD-less TDP-43 self-assembly monitored by turbidimetry. A–D Fixed parameters were 5 μM PrLD-less TDP-43, pH 5.5, 1 mM TCEP, 25 °C. Final NaCl concentrations were 0 mM (A), 50 mM (B), 150 mM (C) and 300 mM (D), each containing 0% (w/v) PEG 8000 (black), 2.5% (w/v) PEG 8000 (red), 5% (w/v) PEG 8000 (blue) or 8% (w/v) PEG 8000 (pink), as indicated in each graph. The native protein was in 20 mM Tris–HCl, pH 8.0 (green). E–F Fixed parameters were 5 μM PrLD-less TDP-43, 5% (w/v) PEG 8000, 1 mM TCEP, 25 °C, at a final pH of 4.0 (E) or 7.0 (F). In all panels, final NaCl concentrations were 0 mM (solid line) or 150 mM (dash-dotted line), as indicated in each graph. The number of replicates was 3 (n = 3). Error bars: SEM. The dashed line indicates the threshold of turbidimetry above which the signal is significantly high to indicate visible phase separation

Fig. 5

Electrostatic forces governing phase separation of PrLD-less, PrLD and full-length TDP-43. A Domains composing full-length (centre), PrLD-less (left) and PrLD (right) TDP-43. Net charges of individual domains at neutral pH are negative if featuring low pI (red), weakly positive with moderately high pI (pale blue) and positive with very high pI (blue). pIs are indicated (pIs with a 6 His-tag in brackets). Numbers below domains indicate residue boundaries. B Types of phase separated assemblies observed early (top) and later (bottom) for the TDP-43 constructs. Only PrLD-less (left) and PrLD (right) TDP-43 form genuine liquid droplets. Full-length TDP-43 forms gel-like droplets with limited recovery of RFI after photobleaching that cluster further without fusion (middle). C Factors governing phase separation for the three constructs. D Schematics of electrostatic forces governing phase separation for the three constructs (colours as in A)