Compromised paraspeckle formation as a pathogenic factor in FUSopathies - PubMed (original) (raw)

. 2014 May 1;23(9):2298-312.

doi: 10.1093/hmg/ddt622. Epub 2013 Dec 11.

Affiliations

Compromised paraspeckle formation as a pathogenic factor in FUSopathies

Tatyana A Shelkovnikova et al. Hum Mol Genet. 2014.

Abstract

Paraspeckles are nuclear bodies formed by a set of specialized proteins assembled on the long non-coding RNA NEAT1; they have a role in nuclear retention of hyperedited transcripts and are associated with response to cellular stress. Fused in sarcoma (FUS) protein, linked to a number of neurodegenerative disorders, is an essential paraspeckle component. We have shown that its recruitment to these nuclear structures is mediated by the N-terminal region and requires prion-like activity. FUS interacts with p54nrb/NONO, a major constituent of paraspeckles, in an RNA-dependent manner and responds in the same way as other paraspeckle proteins to alterations in cellular homeostasis such as changes in transcription rates or levels of protein methylation. FUS also regulates NEAT1 levels and paraspeckle formation in cultured cells, and FUS deficiency leads to loss of paraspeckles. Pathological gain-of-function FUS mutations might be expected to affect paraspeckle function in human diseases because mislocalized amyotrophic lateral sclerosis (ALS)-linked FUS variants sequester other paraspeckle proteins into aggregates formed in cultured cells and into neuronal inclusions in a transgenic mouse model of FUSopathy. Furthermore, we detected abundant p54nrb/NONO-positive inclusions in motor neurons of patients with familial forms of ALS caused by FUS mutations, but not in other ALS cases. Our results suggest that both loss and gain of FUS function can trigger disruption of paraspeckle assembly, which may impair protective responses in neurons and thereby contribute to the pathogenesis of FUSopathies.

PubMed Disclaimer

Figures

Figure 1.

Figure 1.

N-terminal domains of FUS are required for targeting the protein to paraspeckles. (A) Both endogenous and overexpressed FUS are excluded from nucleolar regions (arrows) and enriched in multiple small puncta (arrowheads) in interphase nuclei of neuroblastoma SH-SY5Y or COS7 cells. (B) FUS-containing nuclear dots overlap with paraspeckles (arrowheads) visualized with an antibody against PSP1 or p54nrb/NONO. (C) Domain organization of human FUS protein and schematic representation of constructs used in the study. (D) FUS lacking NLS and significantly redistributed to the cytoplasm is still enriched in paraspeckles, as is N-terminal fragment of FUS (NT), while C-terminal fragment of the protein (CT) fails to localize to paraspeckles. (E) Prion-like activity of FUS N-terminal domains is required for paraspeckle recruitment. Schematic map of a chimeric protein with N-terminal part of FUS replaced by the prion domain from a yeast protein Sup35 (aminoacids 1–125). Sup35–FUS localizes predominantly to the nucleus where it is excluded from nucleolar regions (arrows) and found in small puncta that overlap with paraspeckle marker PSP1 (arrowheads). Scale bars, 10 µm.

Figure 2.

Figure 2.

FUS is a component of nucleolar caps completely overlapping with those formed by paraspeckle proteins in neuroblastoma SH-SY5Y cells. (A) FUS becomes redistributed to the perinucleolar region upon treatment with DNA polymerase II inhibitor DRB but does not form nucleolar caps, in contrast to actinomycin D treatment which induces FUS recruitment to classical crescent shaped caps. (B and C) N-terminal fragment of FUS (B) and chimeric protein Sup35–FUS (C) efficiently localize to nucleolar caps upon exposure to actinomycin D. (D–F) FUS is not a component of coilin p80 caps (D), but FUS caps partially colocalize with RNA helicase p68 caps (E) and completely overlap with caps formed by PSP1 in actinomycin D-treated cells (F). (G) FUS is not essential for redistribution of other paraspeckle proteins to nucleolar caps, since PSP1-positive caps were observed in actinomycin D-treated cells depleted of FUS by RNA interference. Actinomycin D or DRB were added to the cells for 1.5 h prior to fixation in all experiments. Scale bars, 10 µm.

Figure 3.

Figure 3.

FUS associates with p54nrb/NONO in vivo via RNA and this interaction is regulated by ongoing transcription. (A and B) Immunoprecipitation revealed an RNase sensitive interaction of p54nrb/NONO with full-length FUS protein (FUS WT) but not with N-terminal part of FUS (FUS NT) (A) or full-length TDP-43 (B). Full-length FUS, N-terminal FUS fragment (aminoacids 1–359, NT), p54nrb/NONO or TDP-43 expressing plasmids were transfected into SH-SY5Y cells and 24 h after transfection immunoprecipitated on anti-GFP antibody coated beads. To test the role of RNA in FUS-p54nrb/NONO interaction the lysate of FUS-GFP WT transfected cells was treated with RNase A for 30 min at RT prior to incubation with beads. Asterisks mark non-specific bands. (C) Interaction of endogenous FUS and p54nrb/NONO proteins in COS7 cells. Co-immunoprecipitations of FUS with anti-p54nrb/NONO antibody from cell lysates. A part of very intense 50 kDa immunoglobulin heavy chain band is seen just under the p54nrb/NONO band because the same antibodies were used for immunoprecipitation and western blotting. (D) FUS remains associated with p54nrb/NONO in actinomycin D but not DRB treated cells. Protein complexes of full-length GFP-tagged FUS were immunoprecipitated from lysates of SH-SY5Y cells untreated or treated with inhibitors of transcription for 1.5 h.

Figure 4.

Figure 4.

Protein methylation regulates distribution of paraspeckle proteins in the nucleus. (A) Paraspeckle proteins p54nrb/NONO, PSP1 and FUS accumulate in the perinucleolar region in the majority of MCF7 breast cancer cells under basal conditions. (B) Ectopic expression of Flag-tagged MTAP protein in MCF7 cells restores paraspeckle distribution of endogenous p54nrb/NONO protein. Cells were fixed and processed for staining 24 h post-transfection. Arrows show p54nrb/NONO-positive paraspeckles in MTAP-expressing cells. (C) Prolonged (24 h) treatment of neuroblastoma SH-SY5Y cells with MTA, a global inhibitor of protein methyltransferases, induces redistribution of p54nrb/NONO and FUS into perinucleolar region in a fraction of cells (arrows). Paraspeckles are still preserved in cells where such redistribution did not occur (arrowheads). (D) Perinucleolar localization of PSP1 in MCF7 cells is abolished by siRNA knockdown of FUS expression. Scale bars, 10 µm.

Figure 5.

Figure 5.

FUS is important for the integrity of paraspeckles and regulates NEAT1 levels. (A–C) siRNA knockdown of FUS causes loss of paraspeckles in cultured cells as visualized by immunostaining for core paraspeckle proteins PSP1 and p54nrb/NONO (A and B) without alterations in total levels of these proteins (C). Representative images for PSP1 and p54nrb/NONO distribution in COS7 and SH-SY5Y cells and quantification for COS7 cells are shown. Anti-PSP1 staining was used to visualize paraspeckles for counts. Arrows indicate paraspeckles preserved in cells with normal FUS levels. (D and E) FUS protein levels regulate abundance of long non-coding RNA NEAT1. Downregulation of FUS expression by siRNA knockdown significantly decreases NEAT1 levels (D), while FUS overexpression results in elevated NEAT1 (E) in MCF7 cells as measured by qPCR with primers specific for both short (NEAT1_1) and long (NEAT1_2) isoforms of NEAT1. Cells were transfected with either empty pEGFP-C1 vector or GFP-FUS and analysed 24 h post-transfection. Western blotting with anti-FUS antibody shows approximately equal levels of FUS-GFP and endogenous FUS in total cell culture lysates, considering that efficiency of transfection of MCF7 cells was ∼25%, transfected cells expressed approximately 4 times more ectopic than endogenous FUS protein. (F and G) p54nrb/NONO substitutes for loss of FUS function required for paraspeckle formation. GFP fused p54nrb/NONO expressed in FUS-depleted COS7 cells formed multiple paraspeckle-like structures in dose-dependent manner (F), and these were positive for PSP1 (G, arrows**).** In all experiments cells were transfected with either a pool of siRNA specifically targeting FUS protein (FUS siRNA) or scrambled siRNA (scrmb siRNA) and analysed 72 h post-transfection. *P < 0.05 and ***P < 0.001 (Mann–Whitney _U_-test). Error bars represent SEM. Scale bars, 10 µm.

Figure 6.

Figure 6.

Cytosolic FUS aggregates sequester paraspeckle proteins. p54nrb/NONO protein is consistently present in aggregates formed by cytoplasmically mislocalized FUS bearing an ALS-linked R522G substitution (A, arrows), and in some cells p54nrb/NONO is cleared from the nucleus and accumulates in FUS aggregates (B, arrowheads). Core paraspeckle proteins PSP1 and PSF are recruited into FUS aggregates in neuroblastoma SH-SY5Y (C and E, arrows) and COS7 (D, arrows) cells. Cells were analysed 24 h post-transfection. Scale bar, 10 µm.

Figure 7.

Figure 7.

p54nrb/NONO is sequestered into cytoplasmic and nuclear inclusions formed by truncated FUS in a transgenic mouse model of FUSopathy. (A) Immunohistochemical staining with antibody against p54nrb/NONO shows that the protein accumulates in the cytoplasm of large spinal motor neurons but in the cytoplasm of small neurons and glial cells of non-transgenic mice. (B–G) p54nrb/NONO is detected in virtually all nuclear FUS aggregates by double immunofluorescence (arrowheads in higher magnification image shown in G). Occasionally p54nrb/NONO is also detected by immunohistochemistry in cytoplasmic inclusions (C and D, arrows) formed by truncated FUS protein in spinal motor neurons of symptomatic FUS-TG mice. Both truncated and endogenous FUS were visualized by an antibody recognizing an N-terminal FUS epitope (N-term FUS) present in both proteins. Scale bars, A–D: 15 µm; E and F: 35 µm; G: 10 µm.

Figure 8.

Figure 8.

p54nrb/NONO is a constituent of cytoplasmic and nuclear inclusions in human familial ALS-FUS. (A) p54nrb/NONO is confined to the nucleus in the majority of glial cells and small neurons in the spinal cord of non-ALS individuals (control#1). However, in spinal motor neurons p54nrb/NONO is present at considerable levels in the cytoplasm. Representative images of spinal motor neurons from two healthy individuals and one MS case stained with anti-p54nrb/NONO antibody are shown. (B) Multiple nuclear and cytoplasmic p54nrb/NONO-positive inclusions are detected in two familial ALS cases with FUS mutations (ALS-FUS). (C) p54nrb/NONO is diffusely distributed in the nucleus and cytoplasm of sporadic ALS (sALS) cases, including a case with confirmed TDP-43 inclusions (sALS-TDP), as well as in a case of familial ALS with SOD1 mutation (ALS-SOD1). Scale bars, 30 µm.

Figure 9.

Figure 9.

A hypothetical scheme describing how mislocalization of FUS protein typical for human FUSopathies may disrupt early response of neurons to stressful conditions because of compromised paraspeckle formation.

References

    1. Fiesel F.C., Kahle P.J. TDP-43 and FUS/TLS: cellular functions and implications for neurodegeneration. FEBS J. 2011;278:3550–3568. - PubMed
    1. Crozat A., Aman P., Mandahl N., Ron D. Fusion of CHOP to a novel RNA-binding protein in human myxoid liposarcoma. Nature. 1993;363:640–644. - PubMed
    1. Da Cruz S., Cleveland D.W. Understanding the role of TDP-43 and FUS/TLS in ALS and beyond. Curr. Opin. Neurobiol. 2011;21:904–919. - PMC - PubMed
    1. Kwiatkowski T.J., Jr, Bosco D.A., Leclerc A.L., Tamrazian E., Vanderburg C.R., Russ C., Davis A., Gilchrist J., Kasarskis E.J., Munsat T., et al. Mutations in the FUS/TLS gene on chromosome 16 cause familial amyotrophic lateral sclerosis. Science. 2009;323:1205–1208. - PubMed
    1. Neumann M., Rademakers R., Roeber S., Baker M., Kretzschmar H.A., Mackenzie I.R. A new subtype of frontotemporal lobar degeneration with FUS pathology. Brain. 2009;132:2922–2931. - PMC - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources