De novo design of RNA-binding proteins with a prion-like domain related to ALS/FTD proteinopathies

Abstract

Aberrant RNA-binding proteins form the core of the neurodegeneration cascade in spectrums of disease, such as amyotrophic lateral sclerosis (ALS)/frontotemporal dementia (FTD). Six ALS-related molecules, TDP-43, FUS, TAF15, EWSR1, heterogeneous nuclear (hn)RNPA1 and hnRNPA2 are RNA-binding proteins containing candidate mutations identified in ALS patients and those share several common features, including harboring an aggregation-prone prion-like domain (PrLD) containing a glycine/serine-tyrosine-glycine/serine (G/S-Y-G/S)-motif-enriched low-complexity sequence and rich in glutamine and/or asparagine. Additinally, these six molecules are components of RNA granules involved in RNA quality control and become mislocated from the nucleus to form cytoplasmic inclusion bodies (IBs) in the ALS/FTD-affected brain. To reveal the essential mechanisms involved in ALS/FTD-related cytotoxicity associated with RNA-binding proteins containing PrLDs, we designed artificial RNA-binding proteins harboring G/S-Y-G/S-motif repeats with and without enriched glutamine residues and nuclear-import/export-signal sequences and examined their cytotoxicity in vitro. These proteins recapitulated features of ALS-linked molecules, including insoluble aggregation, formation of cytoplasmic IBs and components of RNA granules, and cytotoxicity instigation. These findings indicated that these artificial RNA-binding proteins mimicked features of ALS-linked molecules and allowed the study of mechanisms associated with gain of toxic functions related to ALS/FTD pathogenesis.

Figure 1

Expression in cultured mammalian cells of artificial RNA-binding proteins containing PrLDs. (a) Schematic diagram showing artificial RNA-binding proteins containing PrLDs, as well as the deletion constructs. All constructs contained a GFP tag at the C-terminus. sPFD: synthetic prion-forming domain; cPFD: control prion-forming domain [exhibiting low degrees of prion propensity]. (b) Immunoblot analysis of SYG/SYGQ-NES expression in transfected HeLa and Neuro2a cells. Lysates of cells transiently transfected with plasmids were fractionated by SDS-PAGE and analyzed by immunoblot using a mouse monoclonal anti-GFP antibody. α-tubulin was used as an internal control. Asterisks (*) indicate aggregation resolved using stacking gels. Arrowheads indicate full-length SYG/SYGQ-NES. Green arrows indicate predicted size. (c) Immunoblot analysis of deletion constructs expressed in 293 T cells. Protein (12 μg/lane) was loaded on a 4% to 20% Tris-glycine gradient gel and detected using an anti-GFP antibody for western blot analysis. Asterisks (*) indicate aggregation resolved using stacking gels. (d) Immunoblot analysis of SYG, SYGQ, and SYGQ/N-NES expression in transfected 293 T cells. α-tubulin was used as an internal control. Asterisks (*) indicate aggregation, and arrowheads indicate full-length proteins. (e) Neuro2a cell lysates expressed of SYG/SYGQ-NES were separated into RIPA-soluble and -insoluble urea fractions. Full-length proteins (indicated by arrowheads) were detected in only the insoluble fractions.

Figure 2

SYGQ-NES proteins form IBs in cultured cells. (a) At 48-h post-transfection, 293 T cells were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (blue). (b) Percentage of cells harboring IBs in ~300 transfected cells from three independent experiments (mean ± standard deviation). Plots indicate mean values from each experiment (n = 3). Scale bar: 10 μm. Asterisks (*) indicate significant differences between SYGQ-NES-GFP vs. GFP (control). *P < 0.0001, vs. GFP (One-way ANOVA). (c) Cytoplasmic IBs of SYGQ-NES are p62-positive. Scale bars: 10 μm.

Figure 3

Interactions between SYGQ-NES and other ALS-related RNA-binding proteins in IBs. (a) 293 T cells expressing SYGQ-NES (green) were labeled with antibodies against TDP-43, G3BP, HuR, Ataxin2, SMN, FMRP and Staufen (red). Note that endogenous TDP-43 is predominantly nuclear but partially cytoplasmic in 293 T cells expressing naïve GFP (green) as control. Scale bar: 10 μm. (b) Emetine disassembles SYGQ-NES IBs. Treatment with 10 μg/mL emetine for 1 h resulted in disassembly of cytoplasmic IB formation in 293 T cells expressing SYGQ-NES-GFP and mutant FUS-GFP, but not those expressing mutant SOD1-GFP. Plots indicate mean values from each experiment (n = 3). Scale bar: 10 μm. *P < 0.05, vs. no treatment (Student’s t test). Plots indicate mean values of each experiment.

Figure 4

GEM number following expression of artificial RNA-binding proteins containing PrLDs. HeLa cells, in which GEM is well-observed, transiently expressed SYG or SYGQ-NES (green). Numbers of GEM formation was assessed by immunostaining with anti-SMN (red). Graphs show the average number of GEMs formed in cells expressing SYG or SYGQ-NES (n = 3). *P < 0.002, vs. GFP (One-way ANOVA). Scale bar: 10 μm

Figure 5

SYG/SYGQ-NES is recruited to SGs in arsenite-treated HeLa cells. Following 1 h treatment with 0.5 μM arsenite, HeLa cells were fixed and immunostained for SG markers and with the anti-HuR antibody. Graphs indicate the ratio of the mean intensity of GFP in SGs to that observed in the cytoplasm. The fluorescent distributions in SGs from cells expressing SYG/SYGQ-NES and ΔRRM were higher than those observed in cells expressing GFP alone. *P < 0.003, vs. GFP; ^#P < 0.0002; vs. SYGQ-NES (one-way ANOVA). Scale bar: 10 μm.

Figure 6

SG formation involving artificial RNA-binding proteins containing PrLDs. HeLa cells expressing GFP, SYG/SYGQ-NES, TDP-43-GFP, or FUS-GFP were double labeled with monoclonal anti-GFP and -HuR antibodies following a 1 h treatment with arsenite (0.5 mM). Approximately 25 transfected cells from three independent experiments were used to quantify the number of SGs formed per cell. *Indicates statistically significant differences (*P < 0.04 vs. GFP, One-way ANOVA). Scale bar: 10 μm.

Figure 7

Neurites in differentiated Neuro2a cells expressing SYG/SYGQ-NES. Representative images of immunocytochemistry for SYG/SYGQ-NES-positive differentaited Neuro2a cells using the anti-βIII-tubulin antibody. Graphs indicate quantitative data from neurite-bearing cells. Neuro2A cells containing neurites exhibiting diameters of at least two cell-bodies in length were scored as neurite-bearing cells (n = 3 independent experiments; mean ± standard deviation). *P < 0.05, one-way ANOVA. Scale bar: 20 μm.

Figure 8

Cytotoxicity induced by SYG/SYGQ-NES. (a) Neurites in differentiated Neuro2a cells expressing deletion variants of SYG and SYGQ-NES. Representative images of βIII-tubulin immunostaining of differentiated Neuro2a cells. sPFD-NES and cPFD-NES were used as a negative control. Graphs indicate quantitative data of βIII-tubulin-positive neurite-bearing neurons (n = 3 independent experiments; mean ± standard deviation) *P < 0.03 (One-way ANOVA). Scale bar: 20 μm. (b) Dead cells to the GFP positive cells by fluorescence-cell sorting. Neuro2a cells were subjected to transfection with SYG, SYGQ-NES, or mutant FUS tagged with GFP, followed by PI staining after a 48 h incubation for the detection of dead (GFP+PI+) cells. cPFD-NES, which shows similar GFP intensity as SYG and SYGQ-NES, was used as a negative control. Representative fluorescence-activated cell sorting pictograms. The percentage of dead (GFP+PI+) cells are shown, and the graph indicates the mean ± standard deviation (n = 5). Asterisks *indicate a significant differences (*P < 0.05; Student’s t test).