ArticlePDF Available

Dendritic Homeostasis Disruption in a Novel Frontotemporal Dementia Mouse Model Expressing Cytoplasmic Fused in Sarcoma

Authors:

Abstract and Figures

Highlights • ALS model mice expressing FUS with nuclear localization signal deletion showed behavioral and cognitive FTD phenotypes. • Histological analyses revealed decreased dendritic spine and synaptic density before neuronal loss. • Mutant FUS aggregates trap robust mRNA and RNA transporters, leading to dendritic homeostasis disruption. • Strategies targeting dendritic homeostasis might support future approaches to the development of therapies against ALS-FTD. Cytoplasmic aggregation of FUS is a pathological hallmark in brain regions affected by FUS-linked ALS-FTD. To understand the pathomechanism of ALS-FTD, this study sought to evaluate FTD pathophysiology in a novel ALS mouse model expressing FUS with a deleted nuclear localization signal. Mutant mice showed hyperactivity, social interactional deficits, and impaired memory retrieval, which are compatible with FTD phenotypes. Histological analyses showed decreased dendritic spine and synaptic density before neuronal loss. Examination of cultured cells confirmed that mutant but not wild-type FUS aggregates trap robust mRNA and RNA transporters and decrease dendritic growth and protein synthesis, resulting in ALS-FTD phenotypes.
Content may be subject to copyright.
Research Paper
Dendritic Homeostasis Disruption in a Novel Frontotemporal Dementia
Mouse Model Expressing Cytoplasmic Fused in Sarcoma
Gen Shiihashi
a
,DaisukeIto
a,
, Itaru Arai
b
,YukiKobayashi
c
, Kanehiro Hayashi
d
, Shintaro Otsuka
b
,
Kazunori Nakajima
d
, Michisuke Yuzaki
b
, Shigeyoshi Itohara
c
,NorihiroSuzuki
a
a
Department of Neurology, Keio University School of Medicine, Shinjuku-ku, Tokyo 160-8582, Japan.
b
Department of Physiology, Keio University School of Medicine, Shinjuku-ku, Tokyo 160-8582, Japan
c
Laboratory for Behavioral Genetics, RIKEN Brain Science Institute, Wako, Saitama 351-0198, Japan
d
Department of Anatomy, Keio University School of Medicine, Shinjuku-ku, Tokyo 160-8582, Japan
abstractarticle info
Article history:
Received 4 June 2017
Received in revised form 17 August 2017
Accepted 5 September 2017
Available online 9 September 2017
Cytoplasmic aggregation of fused in sarcoma (FUS) is detected in brain regions affected by amyotrophic lateral
sclerosis (ALS) and frontotemporal dementia (FTD), which compose the disease spectrum, FUS proteinopathy.
To understand the pathomechanism of ALS-FTD-associated FUS, we examined the behavior and cellular proper-
ties of an ALS mouse model overexpressing FUS with nuclear localization signal deletion. Mutant FUS transgenic
mice showed hyperactivity, social interactional decits, and impaired fear memory retrieval, all of which are
compatible with FTD phenotypes. Histological analyses showed decreased dendritic spine and synaptic density
in the frontal cortex before neuronal loss. Examination of cultured cells conrmed that mutant but not wild-
type FUS was associated with decreased dendritic growth, mRNA levels, and protein synthesis in dendrites.
These data suggest that cytoplasmic FUS aggregates impair dendritic mRNA trafcking and translation, in turn
leading to dendritic homeostasis disruption and the development of FTD phenotypes.
© 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
1. Introduction
The identication of DNA- and RNA-binding protein TARDNA-bind-
ing protein (TDP-43)as a component of inclusions in amyotrophic later-
al sclerosis (ALS) and frontotemporal dementia (FTD) has led to an ALS
researchbreakthrough in 2006 (Arai et al., 2006; Neumannet al., 2006).
Since then,it has been proposed that FTD and ALSconstitute a novel dis-
ease spectrum sharing a common molecular basis (Ito and Suzuki,
2011). In 2009, mutations in the gene encoding the RNA-binding pro-
tein fused in sarcoma (FUS) (also known as translocated in liposarcoma
[TLS]) were identied as the cause of familial ALS (ALS6) (Kwiatkowski
et al., 2009; Vance et al., 2009). Subsequently, several investigators re-
ported observing FUS-positive inclusions in FTD that was classied as
basophilic inclusion body disease (BIBD) and neuronal intermediate l-
ament inclusion disease (NIFID) (Munoz et al., 2009; Neumann et al.,
2009b). Thus, it has been proposed that mutant-FUSlinked ALS, atypi-
cal frontotemporal lobar degeneration with ubiquitinated inclusions,
BIBD, and NIFID also comprise an ALS-FTD disease spectrum: FUS
proteinopathy. Thereafter, a growing number of studies identied
many causative genes of ALS-FTD, such as C9orf72 G4C2 repeat
expansion and hnRNPA2B1, emphasizing the importance of this novel
disease spectrum (Kim et al., 2013; Majounie et al., 2012).
In FUS, which plays roles in DNA repair, transcription, alternative
splicing, translation, and RNA transport (Ito and Suzuki, 2011), ALS-
linked mutations are enriched at the nuclear localization signal at the
C-terminus. Cellular analysis reveals that the increased mislocalisation
of mutant protein into the cytoplasm is negatively correlated with the
age of disease onset (Dormann et al., 2010; Ito et al., 2010), which is
similar to polyglutamine diseases. The impairment of the nucleartrans-
port of FUS is therefore directly associated with neurodegeneration in
ALS and FTD. Furthermore, a mutation in a truncated protein lacking a
nuclear localization signal was identied to cause juvenile ALS with
rapid disease progression and cognitive impairment (Zou et al., 2013),
suggesting that impaired nucleocytoplasmic trafcking of FUS alone
can lead to the ALS phenotype (Waibel et al., 2010).
A major unresolved question is whether FUS-mediated neurodegen-
eration is caused by a toxic gain of function by cytoplasmic FUS aggre-
gates or a loss of normal FUS function in the nucleus. Recently, we
generated a transgenic mouse line overexpressing exogenous FUS
with nuclear localization signal deletion (ΔNLS-FUS), reecting juvenile
ALS (Shiihashiet al., 2016). This FUS transgenic (tg) mouse, in which ex-
ogenous FUS protein is located strictly in cytoplasm at moderate level
(~80% of the endogenous FUS level), showed signicant progressive
motor impairment from 20 weeks of age. Pathological analysis revealed
EBioMedicine 24 (2017) 102115
Corresponding a uthor at: Departm ent of Neurology, K eio University School of
Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan.
E-mail address: d-ito@jk9.so-net.ne.jp (D. Ito).
http://dx.doi.org/10.1016/j.ebiom.2017.09.005
2352-3964/© 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Contents lists available at ScienceDirect
EBioMedicine
journal homepage: www.ebiomedicine.com
ubiquitin- and p62-positive cytoplasmic ΔNLS-FUS aggregates and also
astrocytosis, microgliosis, and neuronal loss in the brains of FUS tg mice
at 1 year, indicating the recapitulation of some aspects of pathological
ALS. However, the expression level, nuclear distribution, and function
of RNA splicing of the endogenous FUS were not changed in FUS tg
mice. We therefore concluded that the toxicgain of function of cytoplas-
mic FUS aggregates was sufcient to lead to neurodegeneration
(Shiihashi et al., 2016).
To understand the pathomechanism of cytoplasmic FUS aggregates
in ALS-FTD, it is necessary to establish a novel FTD model of FUS
proteinopathy. Several groups have reported that transgenic mice
with mutant FUS showed ALS-like motor decits (Qiuetal.,2014;
Scekic-Zahirovic et al., 2017; Scekic-Zahirovic et al., 2016; Sharma et
al., 2016); however, thus far none has evaluated FTD phenotypes. In
this study, we examined the behavior and cognitive function of ΔNLS-
FUS tg mice before the appearance of motor decits. We also performed
anatomical and cellular examinations in order to understand the molec-
ular basis of neurodegeneration by cytoplasmic FUS aggregates. During
behavioral analysis, FUS tg mice showed hyperactivity, social interac-
tional decits, and impaired fear memory retrieval, all of which are
compatible with FTD phenotypes. Regarding the pathological mecha-
nism of cytoplasmic aggregation of FUS, histological analysis revealed
that cytoplasmic FUS aggregates recruited robust mRNA and RNA trans-
porters such as Staufen and fragile X mental retardation protein
(FMRP). Golgi staining and electrophysiology showed decreased den-
dritic spine density in the frontal cortex and hippocampus of FUS tg
mice before neuronal loss. Moreover, examination of primary cortical
cultured neurons conrmed that mutant but not wild-type FUS de-
creased dendritic growth, mRNA levels, and protein synthesis in den-
drites. These data suggest that cytoplasmic FUS aggregates trap mRNA
and its transporters, impairing dendritic mRNA trafcking and transla-
tion, in turn leading to the disruption of dendritic homeostasis and the
development of FTD phenotypes.
2. Materials and Methods
2.1. Generation of ΔNLS-FUS Tg Mice
This procedure is described in full in the article of our previous study
(Shiihashi et al., 2016); briey, a human FUS cDNA with deleted NLS
was inserted into the pTSC21k vector encoding the murine Thy1.2 ex-
pression cassette. Tg lines were bred and backcrossed more than ve
generations with C57Bl6/J mice. Mice were housed in standardized ven-
tilated microisolation caging (four animals per cage). Genotyping from
tail DNA was performed using the following primer pairs: 5-
AAGAAGACCTGGCCTCAAACG-3and 5-TATCCCTGGGGAGTTGACTG-
3. Animal experiments were approved by the Committee on Animal
Care and Use, Keio University (09217- [3]) and conducted according
to the Animal Experimentation Guidelines of Keio University School of
Medicine.
2.2. Behavioral Tests
Male non-tg (n= 15) and FUS tg (n= 14) mice between 9 and
13 weeks of age were used for the following behavioral tests, which
were performed sequentially.
Home cage test: Each mouse (9 weeks old) was placed alone in a
testing cage (18.8 cm [W] × 28.8 cm [L] × 13.7 mm [H]) under a
12-h light-dark cycle (light on at 08:00) and with free access to
both food and water. Spontaneous activity in the cage was assessed
for 3 continuous days by counting the number of infrared beam
crossings (Scanet; Melquest).
Open-eld test: Each mouse (11 weeks old) was placed in an open-
eld arena (50 cm [W] × 50 cm [L] × 40 cm [H]; O'Hara & Co., Ltd.)
made of white polyvinyl chloride. The distance travelled in the
open eld was recorded for 30 min using a video-imaging system
(Image OF9; O'Hara & Co., Ltd.). The central area was denedas
the central 18 cm × 18 cm region of the arena. The mice were tested
on two consecutive days.
Y-maze test: The Y-maze test was performedat 11 weeks of age after
the open-eld test. The apparatus consisted of a plastic maze with
three V-shaped arms (40 cm [L] × 3 cm [W] at the bottom and
10 cm [W] at thetop opening× 12 cm [H], 120 degrees' separation).
Mice were allowed to move freely through the maze for 10 min. The
sequence and total number of arm entries were recorded by a video
camera. A novel arm selectionwas counted when a mouse succes-
sively entered three different arms. The recorded data were ana-
lyzed ofine with EthoVision XT software (Noldus Information
Technology).
Three-chamber test: We evaluated social recognition and
response to social novelty using the three-chamber test at
12 weeks of age. The apparatus consisted of a white plastic box
(42 cm [W] × 61.5 cm [L] × 22 cm [H]) with partitions dividing the
box into three equal-sized chambers, with 10-cm openings between
chambers. The side chamber has a plastic triangular prism cage at
the distal corner for constraining a stranger mouse. Target subjects
(stranger one and stranger two) were age-matched males. This
task was carried out in three phases of 10 min each. In phase one,
the test mouse was placed in the middle chamber and left to explore
the area containing the empty cages for 10 min. In phase two, the
mouse was placed in the middle chamber, but an unfamiliar
mouse (stranger one) was placed into a cage in one of the side cham-
bers. In phase three, a novel stranger mouse (stranger two) was
placed in the previously empty cage and again the test mouse was
left to explore for 10 min. We measured the time the test mouse
spent in each chamber for each phase.
Trace fear conditioning test: This test was carried out in mice aged
13 weeks according to the procedure described previously, with
minor modications (Suh et al., 2011). On day one, the mouse was
placed in a conditioning chamber. Each mouse was exposed to
tone-foot shock pairing (tone = 20 s, 65 dB white noise; footshock
= 2 s, 0.75 mA, 20 s after termination of tone). The tone-footshock
pairing was performed three times (beginning at 240, 400, and
560 s). Mice were returned to their home cages 98 s after the last
footshock. Fear memory of the tone was investigated on day two
by placing each mouse in a novel chamber. After 4 min of free explo-
ration, each mouse was exposed to three tones (60 s, 65 dB white
noise, separated by a 3-min interval). Animal images were captured
twice per second using a camera, and the area (in pixels) in which
the mouse moved was measured. Freezing was dened as a lack of
movement such that fewer than 20 pixels changed between succes-
sive video frames for at least 2 s. The electrical intensity required to
make the mouse jump was measured.
2.3. Immunoblot Analysis
Brain tissues were homogenized in cold lysis buffer containing
50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5% NP-40, 0.5% sodium
deoxycholate, 0.25% sodium dodecyl sulfate, 5 mM EDTA, and protease
inhibitor cocktail (Sigma). Total protein concentrations in the superna-
tants were determined using a Bio-Rad protein assay kit (Hercules). The
proteins were then analyzed by immunoblotting as follows. Protein
samples were separated by reducing sodium dodecyl sulfate polyacryl-
amide gel electrophoresis (SDS-PAGE) on 10% or 14% Tris-glycine
103G. Shiihashi et al. / EBioMedicine 24 ( 2017) 102115
gradient gels and then transferred to polyvinylidene diuoride
membranes (MilliporeA). The membranes were incubated with prima-
ry antibodies and subsequently with horseradish peroxidase (HRP)-
conjugated secondary antibodies. Detection was performed using en-
hanced chemiluminescence reagents according to the manufacturer's
instructions (PerkinElmer Life Sciences, Norwalk, CT). The primary anti-
bodies used in this study were anti-c-Myc 9E10 (Santa Cruz, sc-40,
1:1000), anti-FUS (Thermo, PA5-23696, 1:1000; Bethul, A304-819A-
M, 1:1000), anti-GAPDH (Cell Signaling, 14C10, 1:1000), anti-Lamin
A/C (Cell Signaling, #2032, 1:1000), anti-puromycin (Millipore,
MABE343, 1:1000), anti-V5 (Invitrogen, 46-0705, 1:1000), and αtubu-
lin (Cell Signaling,#2125, 1:1000) as an internal loading control. Protein
levels were determined by densitometry using an Epson ES-2000 scan-
ner and ImageJ (National Institutes of Health).
2.4. Cytoplasmic and Nuclear Fractionation
Frozen brain was homogenized in 10 ml/g buffer containing 10 mM
HEPES,10mMNaCl,1mMKH
2
PO
4
, 5 mM NaHCO
3
, 5 mM EDTA, 1 mM
CaCl
2
, 0.5 mM MgCl
2
. After 10 min on ice, 0.5 ml/g 2.5 M sucrose was
added. Next, tissue was homogenized and centrifuged at 6300 ×gfor
10 min. The supernatant was collected as cytoplasmic fraction. The pel-
let was resuspended in RIPA buffer (50 mM Tris, 150 mM NaCl, 1% NP-
40, 5 mM EDTA, and 0.5% sodium deoxycholate) with 2% SDS as nuclear
fraction.
2.5. Immunohistochemical Staining
Mice were perfused intracardially with ice-cold PBS and then 4%
paraformaldehyde under deeply anesthesia. Brains were extracted,
postxed in 4% paraformaldehyde overnight at 4 °C, transferred to a
30% sucrose solution for cryoprotection, frozen, and stored at 80 °C.
We sectioned brain samples with a cryostat (Leica). The sections were
then incubated with 0.2% triton-X for 15 min at room temperature in
preparation for immunohistochemistry. Immunohistochemistry was
performed using anti-NeuN (Chemicon, MAB377, 1:1000), anti-MAP2
(Merck, AB5622, 1:1000), anti-c-Myc 9E10 (1:1000), anti-Myc (Cell Sig-
naling, #06-549, 1:1000), anti-FMRP (Millipore, MAB2160, 1:1000),
anti-stau1 (Rockland, 600-401-EV4, 1:1000), anti-PSD 95 (NeuroMab,
75-028, 1:500), anti-vesicular glutamate transporter 1 (VGLUT1) (Syn-
aptic System, 135 303, 1:1000), and anti-puromycin (Millipore,
MABE343, 1:1000). Fluorescence intensity was determined by densi-
tometry on sections of the frontal cortex and hippocampus (n= 3 per
genotype) using ImageJ.
2.6. Golgi Staining
The Golgi-Cox impregnation stains were performed using the FD
Rapid GolgiStain Kit according to the manufacturer's instructions
(MTR Scientic Inc.). In brief, brains from 15-week-old FUS and non-tg
mice were immersed in impregnation solution for 2 weeks, transferred
to Solution Cfor 2 days, and cut into 100-μm sections using a cryostat
(CM3050S; Leica). Sections were mounted on APS-coated slides
(Matsunami) and allowed to dry for 2 days before staining with silver
nitrate solution (Solution D and E) and dehydrated through descend-
ing alcohol series before mounting with Permount (Falma). For dendrit-
ic spine analysis, z-stack images of the Golgi-stained pyramidal neurons
from Layer V in the frontal cortex and CA1 radiatum region in hippo-
campus were obtained using a 60× objective (BZ-9000; Keyence). We
analyzed the spines from 15 neurons (n = 3 mice per genotype). We
calculated the density by dividing the number of spines per 100 μmof
dendrite starting at a distance N30 μm away from the soma along the
dendrite over a distance 120 μm. The head width of dendritic spines
was measured using the line tool in ImageJ. The protrusion whose
width was b0.3 μm was labeled lopodia and distinguished from the
spine. A spine was labeled mushroom (mature spine) when its maxi-
mum head diameter was N0.6 μm and at least twice the neck diameter.
2.7. Mouse Primary Cultured Cortical Neurons and Cell Line Culture
The cerebral cortex obtained from E16.5 ICR mice (Japan SLC) or
E16.5 C57Bl6/J mice (Japan SLC) mated with FUS tg mice was excised,
incubated with 5 U/ml of papain (Nacalai) in BCG solution (0.02% bo-
vine albumin [Sigma], 0.02% L-cysteine [Sigma], 0.5% glucose [Wako])
at 37 °C, dissociated, and washed with neurobasal medium (Gibco).
The dissociated cells were cultured on the glass area of poly-L-lysine
coated glass-bottom dishes (Matsunami) in neurobasal medium sup-
plemented with B-27 (Life technologies). Neuro2a neuroblastoma
cells were cultured in αMEM (Gibco) supplemented with 10% fetal bo-
vine serum. Transfection was performed using Lipofectamine 2000 (Life
Technologies Corporation, Carlsbad, CA, USA) according to the
manufacturer's instructions at day 2 in vitro.
2.8. Reverse Transcriptase-polymerase Chain Reaction Analysis
The cDNA was synthesized from 1 μg total RNA using the
SuperscriptIII First-Strand Synthesis System (Life Technologies). Ex-
pression levels of glutamate receptor subunit were quantied using
the DyNAmo ColorFlash SYBR Green qPCR Kit (Thermo Fisher Scientic)
on a PikoReal 96 Real-Time PCR System (Thermo Fisher Scientic). Each
sample wasmeasured in triplicate and normalized to β-actin levels. The
primer sets are listed in Supplementary Table 1.
2.9. Electrophysiology
Coronal slices containing prefrontal cortex were prepared from 15-
week-old FUS-tag and non-tg mice (Gascon et al., 2014). Briey, after
decapitation, the brain was rapidly removed and placed in ice-cold slic-
ing solution containing 87 mM NaCl, 25 mM NaHCO
3
, 2.5 mM KCl,
1.25 mM NaH
2
PO
4
,10mMD-glucose, 75 mM sucrose, 0.5 mM CaCl
2
,
and 7 mM MgCl
2
, (pH 7.4 in 95% O
2
/5% CO
2
, 325 mOsm). Coronal corti-
cal slices containing prefrontal cortex (300 μm thick) were cut using
DOSAKA linear slicer Pro 7. After a 20-min incubation at 34°C, the slices
were stored at room temperature. Experiments were performed at
room temperature. During experiments, slices were superfused with
an extracellular solution containing: 125 mM NaCl, 2.5 mM KCl,
25 mM NaHCO
3
,1.25mMNaH
2
PO
4
,25mMD-glucose, 2 mM CaCl
2
,
and 1 mM MgCl
2
(pH 7.4 in 95% O
2
/5% CO
2
, ~325 mOsm). Whole-cell
voltage-clamp recordings (holding potential of 70 mV) from visually
identied layer V pyramidal neurons in the prefrontal cortex were per-
formed to record mEPSC. Recording pipettes were fabricated from thin
borosilicate glass tubing (World Precision Instruments). Pipette resis-
tance was ~ 3 MΩwhen lled with intracellular solution containing
150 mM Cs-gluconate, 10 mM HEPES, 4 mM MgCl
2
,4mMNa
2
ATP,
1mMNa
2
GTP, 0.4 mM EGTA, and 5 mM QX-314 (pH adjusted to 7.3
with CsOH). A 5-mV hyperpolarizing test pulse was applied every 10 s
to monitor series resistance (R
s
). Data were discarded if R
s
changed
N20% or leak current was N250 pA. 1 μM TTX, 50 μM D-AP5, and 100
μM Picrotoxin were added to extracellular solution to pharmacological-
ly isolate the mEPSCs. In subsets of experiments, 10 μMNBQXwere
added at the end of recording to conrm that the recorded synaptic
events were mediated by the AMPA receptor. Total 9 neurons from 3
non-tg mice and 10 neurons from 4 FUS tg mice were analyzed. Data
were acquired with EPC-9 amplier (HEKA), low-pass ltered at
2.9 kHz, and sampled at 20 kHz. Liquid junction potential was not
corrected. mEPSCs were detected using Igor pro 6.3.7 (Wavemetrics)
by template matching algorithm and visually veried (Chen et al.,
2017; Pernia-Andrade et al., 2012).
104 G. Shiihashi et al. / EBioMedicine 24 (2017) 102115
2.10. Co-RNA FISH and Immunouorescence Staining
The brain frozen sections or cultured neurons xed in 4% parafor-
maldehyde for 15 min were incubated with pre-hybridization buffer
(50% formamide [Wako], 5× saline-sodium citrate buffer [SSC], 0.05%
heparin sodium [Wako], 0.02% ribonucleic acid from torula yeast
[Sigma])for 60 min at 65 °C and then hybridized with a uorescently la-
beled locked nucleic acid (LNA) probes, PolyT(25)Vn 5-uorescein
(Exiqon 300510-04), Scramble-ISH 5-uorescein (Exiqon 300,514
04), and 5-FAM-AGGCCAGGTCTTCTTCAGAAATCA-3(for exogenous
FUS mRNA) (Exiqon) diluted to nal concentration of 80 nM, for 24 h
at 65 °C in a dark, humidied chamber. Next, sections were washed
with 2× SSC/formamide for 1 h at 65 °C and washed with TTBS (0.5 M
NaCl, 0.02 M Tris HCl [pH 8.0], 0.1% tween 20) three times. Then immu-
nohistochemistry was performed using anti-c-Myc 9E10 (1:1000), and
anti-V5 (Cell Signaling, #06-549, 1:1000). Fluorescence intensity was
determined by densitometry using ImageJ.
2.11. RNA Labeling With SYTO RNASelect
The brain frozen sections or cultured neurons xed with 20 °C
methanol for 10 min were incubated with 500 nM SYTO RNASelect
(Life technologies) for 20 min at room temperature as per the
manufacturer's instruction. Then, immunohistochemistry was per-
formed using anti-c-Myc 9E10 (1:1000) and anti-V5 (Cell Signaling,
#06-549, 1:1000).
2.12. Puromycin Labeling
Cultured neurons and N2a cells were treated with 1.8 μMpuromycin
(Nacalai) for 15 min. Then immunohistochemistry or immune blotting
with anti-puromycin (Millipore, MABE343, 1:1000) was performed.
Fluorescence intensity was determined by densitometry using ImageJ.
2.13. Statistical Analysis
Results from non-tg and tg mice were compared using the Student's
t-test. The JMP 11 software package (SAS Institute, Inc.) was used for all
analyses.
3. Results
3.1. Expression and Distribution of Exogenous FUS (ΔNLS-FUS)
FUS proteinopathy denes the majority of tau- and TDP-43-negative
cases of frontotemporal dementia (Urwin et al., 2010). Several FUS mu-
tations have been reported to be linked to familial ALS with features of
frontotemporal dementia (Yan et al., 2010), and juvenile-onset ALS
with cognitive impairment (Hirayanagi et al., 2016). To explore the mo-
lecular mechanism underlying FUS proteinopathy in FTD, we examined
tg mice overexpressing Myc-tagged exogenous ΔNLS-FUS under Thy-1
promoter, which developed progressive motor weakness from
20 weeks of age and formed ubiquitin/p62-positive cytoplasmic aggre-
gates from 6 months of age (Shiihashi et al., 2016). Immunohistochem-
istry and western blot showed that the accumulation of exogenous FUS
protein was located prominently in the frontal cortex, hippocampus,
and entorhinal cortex, indicating that the accumulation of FUS was re-
stricted to specic cell types (Fig. 1a and b, Fig. S1a), consistent with
ndings from the brains of patients with FTD (Neumann et al., 2009a;
Neumann et al., 2009b). Although the expressions of exogenous FUS
RNA at 15 weeks of age were signicantly lower in the olfactory bulb
and striatum, these in the frontal cortex, hippocampus, occipital cortex,
and cerebellum were similar, indicating no correlation between the
levels of ΔNLS-FUS protein accumulation and exogenous FUS RNA ex-
pression (Fig. 1candd).Weconrmed that the expression l evel of exog-
enous FUS was ~80% that of the expression level of endogenous FUS,
and exogenous FUS (ΔNLS-FUS) was strictly localized to the cytoplasm
in tg mice at 15 weeks of age; nuclear localization of endogenous mouse
FUS was not affected by ΔNLS-FUS expression (Shiihashi et al., 2016)
(Fig. S1bd).
3.2. Hyperactivity of ΔNLS-FUS Tg Mice
To examine FTD phenotypes in our tg mice, we performed behavior-
al analyses from 9 to 13 weeks of age, before the appearance of motor
impairment (Shiihashi et al., 2016). First, we performed the home
cage test and open eld test to assess locomotor activity from 9 to
11 weeks of age. In the home cage test, evaluating basal motor activity
in a familiar environment, there was no signicant difference between
the total locomotor activity of FUS tg and non-tg mice over the 3 days
of observation (Fig. S2a). However, at the rst adaptation period, FUS
tg mice showed signicantly higher activity levels compared to non-tg
mice (Fig. 2a), indicating the hyperactivity of FUS tg mice in the novel
environment. In the open eld test, FUS tg mice also showed greater
movement distances, durations, and speed compared to non-tg mice
(Fig. 2b). These results also indicate the hyperactivity of FUS tg mice.
However, there was no signicant difference in the duration spent in
the centre region between FUS and non-tg mice (Fig. 2b), suggesting
that anxiety-related behavior was not altered in FUS tg mice.
Further, we performed the Y-maze test at 11 weeks of age. The num-
ber of novel arm selections did not differ between FUS tg and non-tg mice
(Fig. S2b), indicating that the spatial working memory of FUS tg mice was
not impaired. However, the total number of arm selections was higher in
FUS tg mice compared to non-tg mice, and FUS tg mice also travelled lon-
ger distances and at greater speeds than non-tg mice, indicating hyperac-
tivity (Fig. S2b). Taken together, these results indicate that ΔNLS-FUS
overexpression induces novelty-induced hyperactivity.
3.3. Social Interaction Decit of FUS Tg Mice
To determinewhether the mice had disease-relevant behavioral def-
icits, we performed the three-chamber test when the mice were
12 weeks of age. In the second phase, both FUS tg and non-tg mice
spent more time in the chamber with the novel mouse than in the
chamber with the empty cage (Fig. 2c), indicating that FUS tg mice
could recognize its conspecic. However, in the third phase of the test,
while non-tg mice spent more time in the chamber with the novel
mouse than in that which contained the familiar mouse, FUS tg mice
spent almost the same amount of time in the chamber with the novel
mouse as with that of the familiar mouse (Fig. 2c). Although perfor-
mance in this test depends critically upon proper olfaction, the olfactory
bulb of FUS tg mice showed no exogenous FUS expression (Fig. 1b) and
no morphological changes (Fig. S2c), suggesting olfactory decit was
unlikely in FUS tg mice. We concluded FUS tg mice failed to distinguish
novel and familiar mice, suggesting a decit in social memory.
3.4. Memory Decit of FUS Tg Mice
When the mice were 13 weeks of age, we conducted the trace fear-
conditioning test to evaluate long-term (24-h) memory formation. In
the paradigm, conditioned stimulus (CS) and unconditioned stimulus
(US) are separated by an interval period, which engages the hippocam-
pus and prefrontal cortex (Gilmartin and Helmstetter, 2010; Suh et al.,
2011). No signicant difference in shock sensitivity was detected be-
tween FUS tg and non-tg mice (Fig. S2d). In the cue-retention condition,
FUS tg mice spent signicantly less time freezing than non-tg controls
when they received the rst and second tone (Fig. 2d). Notably, FUS tg
mice showed normal levels of freezing upon thethird tone (Fig. 2d), in-
dicating that fear memory was retained, and retrieved following repeti-
tion of the cue. These ndings suggest that the impairment of 24-h
memory function is due toimpairment in the retrieval of stored memo-
ry rather than disrupted encoding or consolidation of the information
105G. Shiihashi et al. / EBioMedicine 24 ( 2017) 102115
(Roy et al., 2016). In summary, in behavioral analyses, FUS tg mice
showed hyperactivity, social interactional decits, and memory dys-
function, recapitulating most FTD phenotypes.
3.5. Decreased Spine Density in the Frontal Cortex and Hippocampus of FUS
Tg Mice
As shown in Fig. S3, and our previous study(Shiihashi et al., 2016),
there was no neuronal loss in the frontal cortex or hippocampus at
15 weeks of age even though behavioral dysfunction was observed. To
reveal the anatomical basis of the cognitive decit, synaptic connections
were quantied by counting the total number of spines, and those of
mushroom-like morphology alone, in the dendritic region of the frontal
cortex and hippocampus. FUS tg mice showed decreased total spine
density in the frontal cortex and hippocampus relative to non-tg mice
(Fig. 3a). The number of mushroom-shaped mature spines was also
found to be decreased in both regions in FUS tg mice, but the ratio of
mature spines to all spines did not differ between FUS tg mice and
Fig. 1. Expression of exogenous FUS in transgenic mice at 15 weeksof age. (a)Immunohistochemistry with anti-Myc (red) antibody in each part of the brain. Sections counterstainedwith
4,6-diamidino-2-phenylindole (DAPI; blue). Scale bar, 10 μm. (b) Western blot analysis of brain lysates using anti-Myc antibody. Expression of FUS with nuclear localization signal
deletion (ΔNLS-FUS) protein appears restricted to the frontal cortex and hippocampus. α-tubulin was used as an internal control. (c) Quantitative reverse transcriptase polymerase
chain reaction ( qRT-PCR) analysis for exogenous FUS mRNA in the brain of tg mice. (n= 3 per genotype; *Pb0.05 vs. frontal, occi pital cortex and cerebellum; #P b0.05 vs.
hippocampus by Student's t-test). (d) In situ hybridization showing the expression of theexogenous FUS mRNA (green) in the brain of tg mice. Scale bar, 10 μm.
106 G. Shiihashi et al. / EBioMedicine 24 (2017) 102115
non-tg mice, indicating that spine formation, but not maturation, is im-
paired by ΔNLS-FUS expression. Next, immunohistochemical evaluation
of postsynaptic and presynaptic markers, PSD95 and VGLUT1, in the
frontal cortex and hippocampus revealed decreased synaptic density
in FUS tg mice relative to non-tg mice (Fig. 3b and c).
To explore thefunctional consequences of reduction of synaptic den-
sity, we performed whole-cell patch-clamp recording from layerV pyra-
midal neurons in brain slices acutely prepared from the frontal cortices
of non-tg and FUS tg mice at 15 weeks of age. Passive membrane prop-
erties were similar between non-tg [input resistance (R
input
), 453.1 ±
91.1 MΩ, membrane capacitance (C
m
), 149.7 ± 13.6 pF] and FUS tg
(R
input
, 318.2 ± 28.6 MΩ,C
m
, 221.4 ± 31.4 pF)] pyramidal neurons.
However, the frequency of miniature excitatory postsynaptic currents
(mEPSCs) was signicantly reduced in FUS tg compared to non-tg pyra-
midal neurons (Fig. 3d-f). By contrast, the amplitude of mEPSCs did not
differ between non-tg and FUS tg pyramidal neurons (Fig. 3df). These
results indicate that although the function of each excitatory synaptic
input is spared, the total number of excitatory synaptic inputsis reduced
in FUS tg mice.
To further examine synaptic defects, we performed a quantitative
RT-PCR analysis of the expression levels of different glutamate receptor
subunits. Most N-methyl-D-aspartate (NMDA), α-amino-3-hydroxy-5-
Fig. 2. MutantFUS transgenic mice showed behavioral and cognitive impairment. Fourteen mutant FUSand fteen non-tg mice wereevaluated on all batteries. (a) Home cage locomotor
activity during the adaptation period. (b) The total distance moved, total movement duration, average speed, and time spent in the center region in the open-eld test. Transgenic mice
exhibited hyperactivity in their homecages and the open-eld test. (c) Three-chamber testto assess sociability. The paradigm is brieyillustrated. Graphsshow the time spent by the test
mouse in thechambers with and without a novel mouse (phase two: left panel), or the novel mouse and familiar mouse chambers (phasethree: right panel). (d) Trace fearconditioning
test to evaluate long-term memory function. The temporal relationshipsof conditioned stimulus (CS) and unconditioned stimulus (US) are schematically shown. Blue and red numbers
represent the onsets of CS andUS. Twenty-four hoursafter conditioning, freezing ratios were assessed along with theauditory cues. Asterisks (*) indicate signicantdifferences vs. non-tg
mice (P b0.05 by Student's t-test). n.s., not signicant. a.u., arbitrary units. All error bars represent standard deviation of the mean.
107G. Shiihashi et al. / EBioMedicine 24 ( 2017) 102115
methyl-4-isoxazolepropionic acid (AMPA), and kainate receptor sub-
units showed a tendency to decrease in FUS tg mice. mRNA encoding
AMPA receptor subunits Gria3 and Gria4 especially were signicantly
decreased in FUS tg mice at 15 weeks of age (Fig. S4).
3.6. mRNA Transport was Impaired by the Cytoplasmic FUS Aggregate
We next explored the molecular basis of the dendritic spine decits
observed in FUS tg mice. The RNA binding protein FUS contributes to
mRNA transport and regulates local translation in the dendrites (Fujii
et al., 2005; Yasuda et al., 2013). We asked whether cytoplasmic FUS ag-
gregates disturb mRNA transport in the dendrites. RNA distribution was
examined using SYTO RNASelect, which selectively stains RNA
(Knowles et al., 1996), in the frontal cortex and hippocampus of FUS
tg mice at 15 weeks of age. The SYTO signal was colocalized with the cy-
toplasmic FUS aggregates and increased with age (Fig. 4a). We further
performed uorescence in situ hybridization (FISH) for the poly-A
tails of mRNA in the frontal cortex and hippocampus. The signal of
FISH in FUS aggregates also increased with age (Fig. 4b). These results
indicate that cytoplasmic FUS aggregates sequester RNA.
To evaluate mRNA transport in the dendrites, primary cultured crit-
ical neurons from the wild-type mouse embryos were transiently
transfected with mCherry and empty vector, wild-type FUS, or ΔNLS-
FUS at 2 days in vitro (DIV) before being stained by SYTO RNASelect at
Fig. 3. FUS transgenic mice showed synapticimpairment. (a) Representative gure of Golgi staining in the frontal cortex and hippocampus of FUS transgenic (n= 3) and non-tg (n= 3)
mice. Dendritic spine number was quantied along a 100-μmsegment from the origin of theprimary apical dendritic branches of thepyramidal neurons.Asterisks (*) indicate signicant
differences vs. non-tg mice (Pb0.05 by Student's t-test). (b, c) Immunohistochemistry for quantication of postsynaptic density 95 (PSD95) protein (b) and vesicular glutamate
transporter 1 (VGLUT1) (c) in the frontal cortex and hippocampus of FUS transgenic and non-tg mice (n = 3 per genotype, P b0.05 by Student's t-test). Scale bar, 20 μm. (d) Whole-
cell patch-clamp recordings of mEPSCs in pyramidal neurons of the frontal cortex. The amplitude and frequency of these miniature events were compared between FUS transgenic and
non tg mice (P= 0.58 and 0.015 by Wilcoxon rank test for amplitude and frequency, respectively). (e, f) Cumulative distributions of mEPSC frequency and amplitude in the frontal
cortex of FUS transgenic and non tg mice (Pb0.01 and 0.16 by KolmogorovSmirnov test for frequency and amplitude, respectively).
108 G. Shiihashi et al. / EBioMedicine 24 (2017) 102115
5 DIV. Nuclear localization of endogenous mouse FUS was not affected
by ΔNLS-FUS expression (Fig. S5). The intensity of SYTO decreased in
the dendrites of the cultured neurons expressing ΔNLS-FUS (Fig. 5a).
We also performed the FISH for poly-A tails in these cultured neurons.
The signal of FISH was clearly decreased in the dendrites of the cultured
neurons expressing ΔNLS-FUS compared to those expressing empty
Fig. 4. The signal of messenger ribonucleic acid (mRNA) was colocalized withthe cytoplasmic FUSaggregates. (a) Frontal cortexand hippocampus of mice at 15 weeks of age was stained
with anti-Myc (red) and SYTORNA select (green). (b) Double-uorescence insitu hybridization staining for the poly-A tailof mRNA (green) and anti-Myc (red) inthe frontal cortex and
hippocampus of FUS transgenic and non-tg mice. The lower panel shows a scrambled probe staining serving as a negative control. Scale bar, 10 μm.
109G. Shiihashi et al. / EBioMedicine 24 ( 2017) 102115
vector and wild-type FUS (Fig. 5b and c). We then validated these nd-
ings in primary cultured neurons from FUS tg mice. We performed RNA
staining using SYTO RNASelect or FISH for poly-A tails in the cultured
neurons from FUS and non-tg mice embryos at 28 DIV, when exogenous
FUS under post mitotic neurons promoter Thy-1 was substantially
expressed as described below in detail (Fig. 7h). The RNA signal
was also decreased in the dendrites of the cultured neurons from
FUS tg mice (Fig. 5df). Collectively,these results indicate that cytoplas-
mic FUS aggregates sequester the mRNA and impair its dendritic
transport.
Fig. 5. The messenger ribonucleic acid(mRNA) signal was decreased in thedendrites of the cultured neuronsexpressing FUS withnuclear localization signal deletion (ΔNLS-FUS), but not
those expressing wild-type FUS. (a) Cultured neurons transfected with mCherry and empty vector, wild-type FUS, or ΔNLS-FUS tagged V5, were stained with SYTO RNASelect (green).
Exogenous FUS expression was conrmed using anti-V5 tag antibody (red; right panels). The SYTO RNASelect signal was decreased in the dendrites of neurons expressing ΔNLS-FUS.
Scale bar, 10 μm. (b) Fluorescence in situ hybridization (FISH) for the poly-A tail of mRNA (green) was performed for the cultured neurons transfected with mCherry and empty
vector, wild-type FUS, or ΔNLS-FUS. (d) Cultured neurons form mutant FUS and non-tg mice were stained with SYTO RNASelect (green). Scale bar, 10 μm. (e) FISH was performed on
the cultured neurons from transgenic mice. Scale bar, 10 μm. (c and f) Quantication of dendritic mRNA as achieved by FISH (c from b, f from e). The mRNA signal in each group was
normalized to the mean mCherry signal in each dendritic segment, intensities relative to the empty vector group were then calculated in (c). The mRNA signal was normalized to the
mean microtubul e associated protein 2 (MAP2) signal in eac h dendrite, and then to the mRNA signal of the non-tg group in (f). A total of 15 neurons from three independent
experiments were analyzed. *Pb0.05 by Student's t-test.
110 G. Shiihashi et al. / EBioMedicine 24 (2017) 102115
3.7. Cytoplasmic FUS Aggregates Recruit RNA Transporters
FUS binds and retains other RNA-binding proteins via a prion-like
domain in vitro (Blokhuis et al., 2016; Kato et al., 2012; Murakami et
al., 2015). Inclusions in FTD-FUSalso co-accumulated with RNA-binding
proteinsin patients' brains (Neumann et al., 2012), and a previous study
demonstrates that ΔNLS-FUS inclusion sequesters the RNA granules
markers SMN and G3BP (Shiihashi et al., 2016). We therefore examined
Fig. 6. Ribonucleic acid transporters werecolocalized with the cytoplasmic FUS aggregates. (a, b) Frontal cortex and hippocampus fromFUS transgenic and non-tg mice at each age were
stainedwith anti-Myc (red)and FMRP (green;a) or Staufen (green;b). Sections werecounterstainedwith DAPI. Notethat cytoplasmicFUS aggregates recruit both FMRPand Staufen in an
age-dependent manner. Scale bar, 10 μm.
111G. Shiihashi et al. / EBioMedicine 24 ( 2017) 102115
Fig. 7. Fused in sarcoma (FUS) with nuclear localization signal deletion (ΔNLS-FUS) but not wild-type FUS impaired protein synthesis. (a) Protein synthesis was detected by anti-
puromycin antibody (red) in the primary cultured neurons transfected with GFP and empty vector, wild-type FUS, or ΔNLS-FUS after pulse-treatment with puromycin. Dendrites were
identied by GFP expression. Scale bar, 20 μm. (b) Quantication of dendritic protein synthesis as assessed by puromycylation in Fig. 7a. The uorescence intensities of each group
were normalized to the mean GFP signal in each dendritic segment. The graphs show the percentage of the mean puromycin signal (mean ± standard error) in each transfected cell to
the emptyvector group in eachdendritic segment.Fifteen transfected cellsfrom three independent experiments were analyzed. (c) Quantitativedata on the numberof dendritic branches
in the primary cultured neurons expressing wild-type FUS or ΔNLS-FUS. n= 15 per group. (d) Neuro2acells were transfected with empty vector, wild-type FUS, or ΔNLS-FUS, and then
treatedwith puromycin. Celllysates were analyzedby western blottingusing an anti-puromycin antibody. (e) Quantication of dendritic proteinsynthesis as assessedby puromycylation
in Fig. 7d. The graphs show the intensities of puromycylationin each cell relative to thatof the empty vector (mean ± standard error; n = 3). *P b0.05vs. empty vector by Student'st-test.
(f) Primary cultured neuronsfrom FUS and non-tg mice were treated with puromycin and immunostaining by anti-MAP2 (green) andpuromycin (red) antibodies was performed. Scale
bar, 20 μm. (g) Quanticationof dendritic proteinsynthesis as assessed by puromycylation in Fig.7f. The uorescence intensitiesof each group were normalized to themean MAP2 signal.
Fifteen neurons from three independent experiments were analyzed. (h) Cell lysates of primary neurons from the FUS and non-tg mouse embryos were treated with puromycin and
assessed by western blotting. (i) Intensiti es of puromycylation i n the neurons at 28 DIV from transg enic FUS mice relative to those of non-tg mice. *P b0.05 vs. non-tg mi ce by
Student's t-test.
112 G. Shiihashi et al. / EBioMedicine 24 (2017) 102115
the distribution of other RNA-binding proteins contributing to mRNA
transport (Bardoni et al., 2006; Zhang et al., 1995). We performed im-
munohistochemical tests for FMRP and Staufen (St Johnston et al.,
1991) in the frontal cortex and hippocampus. Both RNA transporters
were colocalized with the cytoplasmic FUS aggregates, and the size of
these aggregates increased with age (Fig. 6a and b). Collectively, cyto-
plasmic FUS aggregates trap RNA transporters, possibly leading pertur-
bation of dendritic transport of their mRNA targets.
3.8. Protein Synthesis was Impaired by ΔNLS-FUS
We next asked whether ΔNLS-FUS expression affected translation in
the dendrites. In primary cultured neurons transiently transfected with
green uorescence protein (GFP) and empty vector, wild-type FUS, or
ΔNLS-FUS, and pulse-treated bypuromycin, which labels newly synthe-
sized proteins (Schmidt et al., 2009). The intensity of puromycin was
decreased in the dendrites of the cultured neurons expressing ΔNLS-
FUS, but not those expressing wild-type FUS (Fig. 7a and b). Moreover,
the number of longdendrites (N60 μm) was decreased in the case of cul-
tured neurons expressing ΔNLS-FUS (Fig. 7c).
Further, mouse neuroblastoma Neuro2a cells were transiently
transfected with empty vector, wild-type FUS, or ΔNLS-FUS, and the in-
tensity of puromycin was assessed by western blot analysis. Total new
protein synthesis was signicantly reduced in cells expressing ΔNLS-
FUS, but rather increased in cells expressing wild-type FUS for some un-
known reason, indicating that ΔNLS-FUS expression impaired protein
synthesis (Fig. 7d and e). Furthermore, we validated these ndings in
primary cultured neurons from FUS tg mice. Because Thy-1 expression
is initiated in post mitotic neurons in the perinatal period, substantial
expression of exogenous FUS was detected at 28 DIV, but not at 14
DIV (Fig. 7h). As shown in Fig. 7fi, neurons from ΔNLS-FUS tg mice
showed decreased protein synthesis at 28 DIV. Taken together, these
data indicate that cytoplasmic FUS impairs protein synthesis in den-
drites, leading to the perturbation of dendritic homeostasis (Fig. 8).
4. Discussion
The present study provides several novel insights into the molecular
basis of FUS-linked ALS-FTD's pathology. First, ΔNLS-FUS tg mice,
previously shown to exhibit the toxic gain of function ALS phenotype
(Shiihashi et al., 2016), also recapitulate several aspects of FTD pheno-
types, hyperactivity, social interactional decits, and impaired fear
memory retrieval before the appearance of progressive motor decits
(Fig. 2), establishing them as a novel ALS-FTD mouse model. Our previ-
ous study showed that endogenous FUS expression level, nuclear local-
ization, and splicing activity were not altered in this FUS tg mouse line
(Shiihashi et al., 2016). Therefore, the present study indicates that cyto-
plasmic FUS is also sufcient for an FTD phenotype via a dominant toxic
effect without loss of FUS function. Histological and electrophysiological
analysis revealed decreased dendritic spine and synaptic density in the
frontal cortex and hippocampus of FUS tg mice at 15 weeks (Fig. 3), in
which neuronal loss was not observed (Shiihashi et al., 2016) (Fig.
S3). Glutamate receptor expression analysis reveals that the expression
of AMPA receptor subunits Gria3 and Gria4 was signicantly decreased
in FUS tg mice (Fig. S4), consistent with previousstudies in which alter-
ations in AMPA receptors contribute to behavioral decits mimicking
FTD (Adamczyk et al., 2012; Gascon et al., 2014; Udagawa et al.,
2015). Taken together, the ndings of the present study indicate that
synaptic dysfunction is primary and critical for the development of
FTD phenotypes.
Recently, a single-copy mouse model of mutant cytoplasmic FUS has
been reported to display mild motor neuron phenotypes, but no ubiqui-
tin/p62-positive FUS inclusions (Scekic-Zahirovic et al., 2017). In con-
trast, our mice form ubiquitin- and p62-positive cytoplasmic ΔNLS-
FUS aggregates in affected neuron, indicating the recapitulation of as-
pect of pathological ALS (Shiihashi et al., 2016). Another nding of the
present study is that age-dependent progressive cytoplasmic FUS aggre-
gates recruit robust mRNA and RNA transporters such as Staufen and
FMRP (Figs. 4 and 6). Examination of cultured cells conrms that mu-
tant but not wild-type FUS decreases dendritic growth, mRNA levels,
and protein synthesis in dendrites (Figs. 5 and 7). Collectively, RNA dys-
regulation by cytoplasmic FUS aggregates leads to dendritic
dyshomeostasis, resulting in ALS-FTD-linked neuronal dysfunctions.
Notably, FUS aggregates capture various RNA-binding proteins, includ-
ing RNA transporters and stress granule components, and trap a broad
selection of mRNA molecules. Therefore, we propose that the correction
of specic RNAs regulated by FUS is insufcient as a therapeutic mech-
anism. Rather, the correction of widespread RNA dysregulation is
Fig. 8. The possible pathological cascadeof fused in sarcoma(FUS) proteinopathy.ALS-linkedmutants and/or certainpathologicalstresses lead the mislocalization of FUS to the cytoplasm.
ExcessivecytoplasmicFUS assembles as RNAgranules, and thelocal concentration of proteinincreases, resulting in theformation of toxicand tight aggregates/inclusion bodies.Aggregates
trap robust messenger RNA and RNA transporters, and impair dendritic messenger RNA transporters. Disruption of dendritic homeostasis results in the development of neurological
phenotypes, nally causing neuronalloss.
113G. Shiihashi et al. / EBioMedicine 24 ( 2017) 102115
requiredto ameliorate FUS proteinopathy. A morereasonable molecular
approach for the development of therapeutics should be the clearance
and/or inhibition of the aggregation of FUS itself.
Several limitations and alternative interpretations should be consid-
ered alongside the ndings of this study. First, although the frontal cor-
tex and hippocampus showed a dense accumulation of mutant FUS
(Fig. 1), spatial working memory measured by the Y maze, which de-
pends upon hippocampal-prefrontal-cortical circuits, was not impaired
in FUS tg mice until at least 15 weeks of age (Fig. S2b). Behavioral stud-
ies of other neurodegenerative mouse models overexpressing amyloid
precursor protein and Tau, were reported spatial working memory
was preserved in younger mice(Chen et al., 2000; Morgan et al., 2000;
Pennanen and Gotz, 2005). It is possible that the Y maze is not suf-
ciently sensitive to detect the subtle spatial working memory decits
of younger mice. Future studies ought to address whether these spatial
working memory impairments develop with age and/or whether other
preserved neuronal circuits compensate for the defective hippocampal-
frontal circuits of FUS tg mice.
In our trace fear conditioning test (Fig. 2d), the third toneelicited the
retrieval of stored memory information, suggesting a retrieval problem,
rather than a storage impairment, in the mutant mice. We speculate
that a reduction in dendritic spine density in the hippocampus and pre-
frontal cortex leads to the impairment of memory retrieval, while mem-
ory engramcells survive until 15 weeks of age. Similarly, it was recently
reported that optogenetic activation of hippocampal memory engram
cells results in memory retrieval in an Alzheimer's disease mouse
model, which also showed a progressive reduction in dendritic spine
density in the hippocampus (Roy et al., 2016). Thus, selective rescue
of dendritic spine density on surviving cells may represent an effective
strategy for treating cognitive dysfunction, and possibly alsomotor def-
icits, in the early stages of ALS-FTD.
Finally,FUS knockout and knockdown did not leadto the ALS pheno-
types, but instead resulted in FTD-like behavioral phenotypes, such as
hyperactivity and reduced anxiety-related behavior (Kino et al., 2015;
Udagawa et al., 2015), indicating that the loss of FUS function also con-
tributes to the development of FTD phenotypes. Although in our previ-
ous study we could not detect evidence of loss of FUS function in ΔNLS-
FUS tg mice (Shiihashi et al., 2016), we cannot rule out the possibility
that the function of cytoplasmic FUS, such as in mRNA trafcking, was
partially lost as a result of its aggregation in restricted cell types, and
that this is associated with the observed cognitive dysfunction.
In summary, our data demonstrate that cytoplasmic FUS aggregates
trap robust mRNA and RNA transporters and lead to synaptic and den-
dritic spinedysfunction resulting in ALS-FTD phenotypes before the ap-
pearance of neuronal loss. Thus, we propose that strategies targeting
dendritic homeostasis might support future approaches to treating the
motor/cognitive decits of ALS-FTD. The mouse model here reported
on will contribute to a more detailed understanding of ALS-FTD patho-
genesis and to the development of new therapeutic strategies.
Author Contributions
G.S. and D.I. designed the research. G.S. performed experiments and
analyzed the data. I.A., Y.K., K.H.and S.O. performed electrophysiological
experiments, behavior test, primary culture, and Golgi staining, respec-
tively. G.S., D.I., K.N., M. Y., S. I. and N. S. wrote the manuscript.
Conicts of Interest
The authors declare no conict of interest.
Funding Sources
This work was supported by Eisai Co. Ltd., KANAE Foundation for the
Promotion of Medical Science, Nakabayashi Trust For ALS Research,
Japan Society for the Promotion of Science (No. 16 J05812) and Life
Science Foundation of Japan and the Ministry of Education, Culture,
Sports, Science and Technology of Japan (No. 15 K09323).
Acknowledgement
We are grateful to Dr. Herman van der Putten (Novartis Institutes for
Biomedical Research, Basel, Switzerland) for providing pTSC21k and to
Naomi Kogo, Chie Sano, Atsuko Ohba (Laboratory for Behavioral Genet-
ics, RIKEN Brain Science Institute), and Tamao Kitamura (Department of
Neurology, Keio University School of Medicine) for technical assistance.
We thank the Collaborative Research Resources, Keio University School
of Medicine, for technical assistance.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.ebiom.2017.09.005.
References
Adamczyk, A., Mejias, R., Takamiya, K., Yocum, J., Krasnova, I.N., Calderon, J., Cadet, J.L.,
Huganir, R.L., Pletnikov, M.V., Wang, T., 2012. GluA3-deciency in mice is associated
with increased social and aggressive behavior and elevated dopamine in striatum.
Behav. Brain Res. 229, 265272.
Arai, T., Hasegawa, M., Akiyama, H., Ikeda, K., Nonaka, T., Mori, H., Mann, D., Tsuchiya, K.,
Yoshida, M., Hashizume, Y., et al., 2006. TDP-43 is a component of ubiquitin-positive
tau-negative inclusions in frontotemporal lobar degeneration and amyotrophic later-
al sclerosis. Biochem. Biophys. Res. Commun. 351, 602611.
Bardoni, B., Davidovic, L., Bensaid, M., Khandjian, E.W., 2006. The fragile X syndrome: ex-
ploring its molecular basis and seeking a treatment. Expert Rev. Mol. Med. 8, 116.
Blokhuis, A.M., Koppers, M., Groen, E.J., van den Heuvel, D.M., Dini Modigliani, S., Anink,
J.J., Fumoto, K., van Digg elen, F., Snelting, A., Sodaar, P., et al., 2016. Com parative
interactomics analysis of different ALS-associated proteins identies converging mo-
lecular pathways. Acta Neuropathol. 132, 175196.
Chen, G., Chen, K.S., Knox, J., Inglis, J., Bernard, A., Martin, S.J., Justice, A., McConlogue, L.,
Games, D., Freedman, S.B., et al., 2000. Alearningdecit related to age and beta-am-
yloid plaques in a mouse model of Alzheimer's disease. Nature 408, 975979.
Chen, C., Arai, I., Sattereld, R., Young Jr., S.M., Jonas, P., 2017. Synaptotagmin 2 is the fast
Ca2+ sensor at a central inhibitory synapse. Cell Rep. 18, 723736.
Dormann, D., Rodde, R., Edbauer, D., Bentmann, E., Fis cher, I., Hruscha, A., Than, M.E.,
Mackenzie, I.R., Capell, A., Schmid, B., et al., 2010. ALS-associated fused in sarcoma
(FUS) mutations d isrupt Transpor tin-mediated nuclear import. EMBO J. 29,
28412857.
Fujii, R., Okabe, S., Urushido, T., Inoue, K., Yoshimura, A., Tachibana, T., Nishikawa, T.,
Hicks, G.G., Takumi, T., 2005. The RNA bindingprotein TLS is translocated to dendritic
spines by mGluR5 activation and regulates spinemorphology. Curr.Biol. 15, 587593.
Gascon, E., Lynch, K., Ruan, H., Almeida, S., Verheyden, J.M., Seeley, W.W., Dickson, D.W.,
Petrucelli, L., Sun, D., Jiao, J., et al., 2014. Alterations in microRNA-124 and AMPA re-
ceptors contribute to social behavioral decits in frontotemporal dementia. Nat. Med.
20, 14441451.
Gilmartin, M.R., Helmstetter, F.J., 2010. Trace and contextual fear conditioning require
neural activity and NMDA receptor-dependent transmission in the medial prefrontal
cortex. Learn. Mem. 17, 289296.
Hirayanagi, K., Sato, M., Furuta, N., Makioka, K., Ikeda, Y., 2016. Juvenile-onset sporadic
amyotrophic lateral sclerosis with a frameshift FUS gene mutation presenting unique
neuroradiological ndings and cognitive impairment. Intern. Med. 55, 689693.
Ito, D., Suzuki, N., 2011. Conjoint pathologic cascades mediated by ALS/FTLD-U linked
RNA-binding proteins TDP-43 and FUS. Neurology 77, 16361643.
Ito, D., Seki, M., Tsunoda, Y., Uchiyama, H., Suzuki,N., 2010. Nuclear transport impairment
of ALS-linked mutations in FUS/TLS. Ann. Neurol. 69, 152162.
Kato, M., Han, T.W., Xie,S., Shi, K., Du, X., Wu,L.C., Mirzaei, H.,Goldsmith, E.J., Longgood, J.,
Pei, J., et al., 2012. Cell-free formation of RNA granules: low complexity sequence do-
mains form dynamic bers within hydrogels. Cell 149, 753767.
Kim, H.J., Kim, N.C ., Wang, Y.D., Scarborough, E.A., Moore, J., Diaz, Z., Ma cLea, K.S.,
Freibaum, B., Li , S., Molliex, A., et al., 2013. Mutations in pr ion-like domains in
hnRNPA2B1 and hnRNPA1 cause multisystem proteinopathy and ALS. Nature 495,
467473.
Kino, Y.,Washizu, C., Kurosawa,M., Yamada, M., Miyazaki, H., Akagi, T., Hashikawa, T., Doi,
H., Takumi,T., Hicks, G.G., et al.,2015. FUS/TLS deciency causes behavioral and path-
ological abnormalities distinct from amyotrophic lateral sclerosis. Acta Neuropathol.
Commun. 3, 24.
Knowles, R.B., Sabry, J.H., Martone, M.E., Deerinck, T.J., Ellisman, M.H., Bassell, G.J., Kosik,
K.S., 1996. Translocation of RNA granul es in living neurons. J. Neurosci. 16,
78127820.
Kwiatkowski Jr., T.J., Bosco, D.A., Leclerc, A.L., Tamrazian, E., Vanderburg, C.R., Russ, C.,
Davis, A., Gilchrist, J., Kasarskis, E.J., Munsat, T., et al., 2009. Mutations in the FUS/
TLS gene on chromosome 16 cause familial amyotrophic lateral sclerosis . Science
323, 12051208.
Majounie,E., Renton, A.E., Mok,K., Dopper, E.G., Waite, A., Rollinson, S., Chio, A., Restagno,
G., Nicolaou, N. , Simon-Sanchez, J., et al., 2012. Frequency of the C9orf72
114 G. Shiihashi et al. / EBioMedicine 24 (2017) 102115
hexanucleotide repeat expansion in patients with amyotrophic lateral sclerosis and
frontotemporal dementia: a cross-sectional study. Lancet Neurol. 11, 323330.
Morgan, D., Diamo nd, D.M., Gottschall, P.E., Ugen, K. E., Dickey, C., Hardy, J., Duff, K.,
Jantzen, P., DiCarlo, G., Wilcock, D., et al., 2000. A beta peptide vaccination prevents
memory loss in an animal model of Alzheimer's disease. Nature 408, 982985.
Munoz, D.G., Neumann, M. , Kusaka, H., Yokota, O., Ishihara, K., Terada, S., Ku roda, S.,
Mackenzie, I.R. , 2009. FUS pathology in basophilic inclusion body dis ease. Acta
Neuropathol. 118, 617627.
Murakami, T., Qamar, S., Lin, J.Q., Schierle, G.S., Rees, E., Miyashita, A., Costa, A.R., Dodd,
R.B., Chan, F.T., Michel, C.H., et al., 2015. ALS/FTD mutation-induced phase transition
of FUS liquid droplets and reversible hydrogels into irreversible hydrogels impairs
RNP granule function. Neuron 88, 678690.
Neumann, M., Sampathu, D.M., Kwong, L.K., Truax, A.C., Micsenyi, M.C., Chou, T.T., Bruce,
J., Schuck, T., G rossman, M., Clark, C.M., et al., 2006. Ubiquitinated TDP-43 in
frontotemporal lobar degeneration and amyotrophic lateral sclerosis. Science 314,
130133.
Neumann, M., Rademakers, R., Roeber, S., Baker, M., Kretzschmar, H.A., Mackenzie, I.R.,
2009a. A new subtype of frontotemporal lobar degenerat ion with FUS pathology.
Brain 132, 29222931.
Neumann, M., Roeber, S., Kretzschmar, H.A., Rademakers, R., Baker, M., Mackenzie, I.R.,
2009b. Abundant FUS-immunoreactive pathology in neuronal intermediate lament
inclusion disease. Acta Neuropathol. 118, 605616.
Neumann, M., Valori, C.F., Ansorge, O., Kretzschmar, H.A., Munoz, D.G.,Kusaka, H., Yokota,
O., Ishihara, K., Ang, L.C., Bilbao,J.M., et al., 2012. Transportin 1 accumulates specical-
ly with FET proteins but no other transportin cargos in FTLD-FUS and is absent in FUS
inclusions in ALS with FUS mutations. Acta Neuropathol. 124, 705716.
Pennanen,L., Gotz, J., 2005. Different tau epitopes dene Abeta42-mediatedtau insolubil-
ity. Biochem. Biophys. Res. Commun. 337, 10971101.
Pernia-Andrade, A.J., Goswami, S.P., Stickler, Y., Frobe, U., Schlogl, A., Jonas, P., 2012. A
deconvolution-based method with high sensitivity and temporal resolution for de-
tection of sponta neous synaptic currents in vitro and in vivo. Biophys. J. 103,
14291439.
Qiu, H., Lee, S., Shang, Y., Wang, W.Y., Au, K.F., Kamiya, S., Barmada, S.J., Finkbeiner, S., Lui,
H., Carlton, C.E., et al.,2014. ALS-associated mutationFUS-R521C causes DNA damage
and RNA splicing defects. J. Clin. Invest. 124, 981999.
Roy, D.S., Arons, A., Mitchell, T.I.,Pignatelli, M., Ryan, T.J., Tonegawa, S., 2016. Memory re-
trieval by activating engram cells in mouse models of early Alzheimer's disease. Na-
ture 531, 508512.
Scekic-Zahirovic, J., Sendscheid, O., El Oussini, H., Jambeau, M., Sun, Y., Mersmann, S.,
Wagner, M., Dieterle, S., Sinniger, J., Dirrig-Grosch, S., et al., 2016. Toxic gain of func-
tion from mutant FUS proteinis crucial to trigger cell autonomous motorneuron loss.
EMBO J. 35, 10771097.
Scekic-Zahiro vic, J., Oussini, H.E., Mersmann, S., Drenner, K., Wagner, M., Sun, Y. ,
Allmeroth, K., Dieterle, S., Sinniger, J., Dirrig-Grosch, S., et al., 2017. Motor neuron
intrinsic and extrinsic mechanisms contribute to the pathogenesis of FUS-associated
amyotrophic lateral sclerosis. Acta Neuropathol. 133, 887906.
Schmidt, E.K., Clavarino, G., Ceppi, M., Pierre, P., 2009. SUnSET, a nonradioactive method
to monitor protein synthesis. Nat. Methods 6, 275277.
Sharma, A., Lyas hchenko, A.K., Lu, L., Nasr abady, S.E., Elmaleh, M., Mendelsohn, M.,
Nemes, A., Tapia, J.C., Mentis, G.Z., Shneider, N.A., 2016. ALS-associated mutant FUS
induces selective motor neuron degeneration through toxic gain of function. Nat.
Commun. 7, 10465.
Shiihashi, G., Ito, D., Yagi, T., Nihei, Y., Ebine, T., Suzuki, N., 2016. Mislocated FUS is suf-
cient for gain-of-toxic-function amyotrophic lateral sclerosis phenotypes in mice.
Brain 139, 238 02394.
St Johnston, D., Beuchle,D., Nusslein-Volhard, C., 1991. Staufen,a gene required to localize
maternal RNAs in the Drosophila egg. Cell 66, 5163.
Suh, J., Rivest, A.J., Nakashiba, T., Tominaga, T., Tonegawa, S., 2011. Entorhinal cortexlayer
III input to the hippocampus is crucialfor temporal association memory.Science 334,
14151420.
Udagawa, T., Fujioka,Y., Tanaka, M., Honda, D., Yokoi, S., Riku, Y., Ibi, D., Nagai, T., Yamada,
K.,Watanabe,H.,etal.,2015.FUS regulates AMPA receptor function andFTLD/ALS-as-
sociated behaviour via GluA1 mRNA stabilization. Nat. Commun. 6, 7098.
Urwin, H., Josephs, K.A., Rohrer, J.D., Mackenzie, I.R., Neumann, M., Authier, A., Seelaar, H.,
Van Swieten, J.C., Brown, J.M., Johannsen, P., et al., 2010. FUS pathology denes the
majority of tau- and TDP-43-negati ve frontotempor al lobar degeneration. Acta
Neuropathol. 120, 3341.
Vance, C., Rogelj, B., Hortobagyi, T., De Vos, K.J., Nishimura, A.L., Sreedharan, J., Hu, X.,
Smith, B., Ruddy, D., Wright, P., et al., 2009. Mutations in FUS, an RNA processing pro-
tein, cause familial amyotrophic lateral sclerosis type 6. Science 323, 12081211.
Waibel, S., Neumann, M., Rabe, M., Meyer, T., Ludolph, A.C., 2010. Novel missense and
truncating mutations in FUS/TLS in familial ALS. Neurology 75, 815817.
Yan, J., Deng, H.X., Siddique, N., Fecto,F., Chen, W., Yang, Y., Liu, E., Donkervoort, S., Zheng,
J.G., Shi, Y., et al., 2010. Frameshift and novel mutations in FUS in familial amyotro-
phic lateral sclerosis and ALS/dementia. Neurology 75, 807814.
Yasuda, K.,Zhang, H., Loiselle, D., Haystead, T., Macara,I.G., Mili, S., 2013.The RNA-binding
protein Fus directs translation of localized mRNAs in APC-RNP granules. J. Cell Biol.
203, 737746.
Zhang, Y., O'Connor, J.P., Siomi, M.C., Srinivasan, S., Dutra, A., Nussbaum, R.L., Dreyfuss,G.,
1995. The fragile X mental retardation syndrome protein interacts with novel homo-
logs FXR1 and FXR2. EMBO J. 14, 53585366.
Zou, Z.Y., Cui, L.Y., Sun, Q., Li, X.G., Liu, M.S., Xu, Y., Zhou, Y., Yang, X.Z., 2013. De novo FUS
gene mutations are associated with juvenile-onset sporadic amyotrophic lateral scle-
rosis in China. Neurobiol. Aging 34 (1312), e13111318.
115G. Shiihashi et al. / EBioMedicine 24 ( 2017) 102115
... Neuron-specific expression of ALS-linked FUS mutants in models of disease show similar changes in neuromorphology and synaptic loss [61][62][63][64] that impact neuronal function [63][64][65]. These studies highlight the importance of FUS in regulating neuronal status and the potential impact of ALS-linked FUS mutants in the central nervous system. ...
... Neuron-specific expression of ALS-linked FUS mutants in models of disease show similar changes in neuromorphology and synaptic loss [61][62][63][64] that impact neuronal function [63][64][65]. These studies highlight the importance of FUS in regulating neuronal status and the potential impact of ALS-linked FUS mutants in the central nervous system. ...
... These findings indicate that neuron-specific expression of FUSR521G affects neuron populations involved in both cognitive and motor functions in hFUS R521G/Syn1 mice. ALS-FUS variants cause loss of dendritic branching of motor neurons and decreased synaptic densities in ALS/FTD models of disease [42,[61][62][63][64], but it is unclear when these changes occur relative to motor decline. To address this question, the dendritic structures of cortical and spinal motor neurons of hFUS R521G/Syn1 mice were analysed at different stages of motor decline. ...
Article
Full-text available
Amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) are related neurodegenerative diseases that belong to a common disease spectrum based on overlapping clinical, pathological and genetic evidence. Early pathological changes to the morphology and synapses of affected neuron populations in ALS/FTD suggest a common underlying mechanism of disease that requires further investigation. Fused in sarcoma (FUS) is a DNA/RNA-binding protein with known genetic and pathological links to ALS/FTD. Expression of ALS-linked FUS mutants in mice causes cognitive and motor defects, which correlate with loss of motor neuron dendritic branching and synapses, in addition to other pathological features of ALS/FTD. The role of ALS-linked FUS mutants in causing ALS/FTD-associated disease phenotypes is well established, but there are significant gaps in our understanding of the cell-autonomous role of FUS in promoting structural changes to motor neurons, and how these changes relate to disease progression. Here we generated a neuron-specific FUS-transgenic mouse model expressing the ALS-linked human FUSR521G variant, hFUSR521G/Syn1, to investigate the cell-autonomous role of FUSR521G in causing loss of dendritic branching and synapses of motor neurons, and to understand how these changes relate to ALS-associated phenotypes. Longitudinal analysis of mice revealed that cognitive impairments in juvenile hFUSR521G/Syn1 mice coincide with reduced dendritic branching of cortical motor neurons in the absence of motor impairments or changes in the neuromorphology of spinal motor neurons. Motor impairments and dendritic attrition of spinal motor neurons developed later in aged hFUSR521G/Syn1 mice, along with FUS cytoplasmic mislocalisation, mitochondrial abnormalities and glial activation. Neuroinflammation promotes neuronal dysfunction and drives disease progression in ALS/FTD. The therapeutic effects of inhibiting the pro-inflammatory nuclear factor kappa B (NF-κB) pathway with an analog of Withaferin A, IMS-088, were assessed in symptomatic hFUSR521G/Syn1 mice and were found to improve cognitive and motor function, increase dendritic branches and synapses of motor neurons, and attenuate other ALS/FTD-associated pathological features. Treatment of primary cortical neurons expressing FUSR521G with IMS-088 promoted the restoration of dendritic mitochondrial numbers and mitochondrial activity to wild-type levels, suggesting that inhibition of NF-κB permits the restoration of mitochondrial stasis in our models. Collectively, this work demonstrates that FUSR521G has a cell-autonomous role in causing early pathological changes to dendritic and synaptic structures of motor neurons, and that these changes precede motor defects and other well-known pathological features of ALS/FTD. Finally, these findings provide further support that modulation of the NF-κB pathway in ALS/FTD is an important therapeutic approach to attenuate disease. Supplementary Information The online version contains supplementary material available at 10.1186/s40478-023-01671-1.
... Several studies with FUS mutations have shown defects in dendritic spines formation with defects in dendrites morphology, arborization and functions Qiu et al., 2014;Udagawa et al., 2015;Shiihashi et al., 2017) that could be restored by the reintroduction of crucial elements for dendrites homeostasis such as BDNF, Ndl-1, Gria1 or SynGAP α2 that closely interact with FUS functions Qiu et al., 2014;Udagawa et al., 2015;Yokoi et al., 2017). Schoen and collaborators showed by super-resolution microscopy that FUS localization is pre-synaptic at glutamatergic synapses, near Bassoon proteins and very close to synaptic vesicles. ...
... Overall their data suggest that FUS could interfere with local mRNA translation and sequester synaptic proteins into aggregates (Deshpande et al., 2019). Other studies have corroborated that mutated FUS seems to sequester mRNAs and proteins, impeding proper proteins homeostasis at the synapses, where FUS is found to accumulate (Shiihashi et al., 2017;Lopez-Erauskin et al., 2018). Furthermore, there is increasing evidence that synaptic and dendrites deficits due to FUS mutations lead to FTD symptoms such as hyperactivity, disinhibition, impaired memory and social interactions deficits (Udagawa et al., 2015;Lopez-Erauskin et al., 2018). ...
... The presence of FUS aggregates was observed in frontal cortex and hippocampus, where mRNAs and RNA transporters were found to be sequestered and the transport of dendritic mRNAs transport was disrupted. Finally, they observed reduced proteins synthesis in dendrites in vitro (Shiihashi et al., 2017). Lopez-Erauskin and collaborators recently published the effects of FUS mutations in a novel ALS-FTD mouse model by introducing floxed human transgenes of FUS WT , FUS R521C , FUS R521H in different mouse lines. ...
Thesis
Amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) are two untreatable neurodegenerative diseases. Even thought they are two distinctive diseases, they share a series of clinical, genetic and histological hallmarks, thus defining the ALS-FTD continuum. Mutations in the FUS gene have been linked with ALS, whereas alterations in FUS proteins have been detected in FTD patients. Both diseases are characterized by the presence of cytosolic FUS aggregates.We have studied the autoregulation mechanisms of FUS in a mouse model via de activation of an alternative splicing pathway by the insertion of a human wild-type FUS transgene, which has allowed us to potentially elucidate new therapeutic approaches by gene therapy. Furthermore, our mice develop FTD-like symptoms. Our results suggest an alteration in cortical synapses which could originate the observed cognitive and behavioural deficits, accompanied by alterations in the cholinergic system.
... The expression of ALS-FUS variants in multiple models reduces neuronal arborization and structural complexity as well as synaptic connectivity and integrity [18,[20][21][22][23][24]. The data showing impairment in 10 or 20 mg/kg arimoclomol, or the low-dose (5 mg/kg arimoclomol þ 10 mg/kg RGFP963) and high-dose (20 mg/kg arimoclomol þ 10 mg/kg RGFP963) combinations showed a significant increase in latency time in entering the dark compartment after aversive shock preconditioning compared to vehicle-treated FUS R521G mice. ...
... Although FUS was consistently present in the MS pro les of both the Hspa8 3′ UTR and the LacZ control, it showed relatively higher binding to the 3′ UTR in pulldowns (Fig. 6b). FUS was of particular interest because it regulates several steps in mRNA maturation, including transport to dendrites, and FUS mutations lead to dendritic retraction in motor neurons, leading to ALS and frontotemporal dementia 27,[109][110][111][112][113][114] . ...
Preprint
Full-text available
Proteostasis is maintained through regulated protein synthesis and degradation and chaperone-assisted protein folding. However, this is challenging in neuronal projections because of their polarized morphology and constant synaptic proteome remodeling. Using high-resolution fluorescence microscopy, we discovered that neurons localize a subset of chaperone mRNAs to their dendrites and use microtubule-based transport to increase this asymmetric localization following proteotoxic stress. The most abundant dendritic chaperone mRNA encodes a constitutive heat shock protein 70 family member (HSPA8). Proteotoxic stress also enhanced HSPA8 mRNA translation efficiency in dendrites. Stress-mediated HSPA8 mRNA localization to the dendrites was impaired by depleting fused in sarcoma—an amyotrophic lateral sclerosis-related protein—in cultured mouse motor neurons and expressing a pathogenic variant of heterogenous nuclear ribonucleoprotein A2/B1 in neurons derived from human induced pluripotent stem cells. These results reveal a crucial and unexpected neuronal stress response in which RNA-binding proteins increase the dendritic localization of HSPA8 mRNA to maintain proteostasis and prevent neurodegeneration.
... Amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) are progressive and incurable neurodegenerative diseases with a clinical and pathological continuum involving motor, behavior, and cognitive dysfunction [1][2][3]. A contributing feature of these diseases is decreased synaptic density and loss of mature dendritic spines before the neuronal loss [4][5][6][7][8][9][10]. A common histopathological hallmark of ALS and FTD neurons is the cytoplasmic aggregation of fused-in-sarcoma (FUS) and TAR DNA binding protein 43 (TDP-43) [3,11], RNA-binding proteins (RBPs) normally located in the nucleus and play multiple functions in RNA metabolism [12,13]. ...
Article
Full-text available
In neurodegenerative diseases, the condensation of FUS and TDP-43 with RNA granules in neurons is linked to pathology, including synaptic disorders. However, the effects of FUS and TDP-43 on RNA granule factors remain unclear. Here, using primary cultured neurons from the mouse cerebral cortex, we show that excess cytoplasmic FUS and TDP-43 accumulated in dendritic RNA granules, where they increased the dynamics of a scaffold protein RNG105/caprin1 and dissociated it from the granules. This coincided with reduced levels of mRNA and translation around the granules and synaptic loss in dendrites. These defects were suppressed by non-dissociable RNG105, suggesting that RNG105 dissociation mediated the defects. In contrast to the model where FUS and TDP-43 co-aggregate with RNA granule factors to repress their activity, our findings provide a novel pathogenic mechanism whereby FUS and TDP-43 dissociate RNA scaffold proteins from RNA granules which are required for local translation that regulates synapse formation.
Article
Advanced pathological and genetic approaches have revealed that mutations in fused in sarcoma/translated in liposarcoma (FUS/TLS), which is pivotal for DNA repair, alternative splicing, translation and RNA transport, cause familial amyotrophic lateral sclerosis (ALS). The generation of suitable animal models for ALS is essential for understanding its pathogenesis and developing therapies. Therefore, we used CRISPR-Cas9 to generate FUS-ALS mutation in the non-classical nuclear localization signal (NLS), H517D (mouse position: H509D) and genome-edited mice. Fus WT/H509D mice showed progressive motor impairment (accelerating rotarod and DigiGait system) with age, which was associated with the loss of motor neurons and disruption of the nuclear lamina and nucleoporins and DNA damage in spinal cord motor neurons. We confirmed the validity of our model by showing that nuclear lamina and nucleoporin disruption were observed in lower motor neurons differentiated from patient-derived human induced pluripotent stem cells (hiPSC-LMNs) with FUS-H517D and in the post-mortem spinal cord of patients with ALS. RNA sequence analysis revealed that most nuclear lamina and nucleoporin-linking genes were significantly decreased in FUS-H517D hiPSC-LMNs. This evidence suggests that disruption of the nuclear lamina and nucleoporins is crucial for ALS pathomechanisms. Combined with patient-derived hiPSC-LMNs and autopsy samples, this mouse model might provide a more reliable understanding of ALS pathogenesis and might aid in the development of therapeutic strategies.
Preprint
Full-text available
Neurons are challenged to maintain proteostasis in neuronal projections, particularly with the physiological stress at synapses to support intercellular communication underlying important functions such as memory and movement control. Proteostasis is maintained through regulated protein synthesis and degradation and chaperone-assisted protein folding. Using high-resolution fluorescent microscopy, we discovered that neurons localize a subset of chaperone mRNAs to their dendrites, particularly more proximal regions, and increase this asymmetric localization following proteotoxic stress through microtubule-based transport from the soma. The most abundant chaperone mRNA in dendrites encodes the constitutive heat shock protein 70, HSPA8. Proteotoxic stress in cultured neurons, induced by inhibiting proteasome activity or inducing oxidative stress, enhanced transport of Hspa8 mRNAs to dendrites and the percentage of mRNAs engaged in translation on mono and polyribosomes. Knocking down the ALS-related protein Fused in Sarcoma (FUS) and a dominant mutation in the heterogenous nuclear ribonucleoprotein A2/B1 (HNRNPA2B1) impaired stress-mediated localization of Hspa8 mRNA to dendrites in cultured murine motor neurons and human iPSC-derived neurons, respectively, revealing the importance of these RNA-binding proteins in maintaining proteostasis. These results reveal the increased dendritic localization and translation of the constitutive HSP70 Hspa8 mRNA as a crucial neuronal stress response to uphold proteostasis and prevent neurodegeneration. SUMMARY Localizing chaperones’ mRNAs in neuronal dendrites is a novel on-demand system to uphold proteostasis upon stress.
Article
Full-text available
Action selection is a capital feature of cognition that guides behavior in processes that range from motor patterns to executive functions. Here, the ongoing actions need to be monitored and adjusted in response to sensory stimuli to increase the chances of reaching the goal. As higher hierarchical processes, these functions rely on complex neural circuits, and connective loops found within the brain and the spinal cord. Successful execution of motor behaviors depends, first, on proper selection of actions, and second, on implementation of motor commands. Thus, pathological conditions crucially affecting the integrity and preservation of these circuits and their connectivity will heavily impact goal-oriented motor behaviors. Amyotrophic Lateral Sclerosis (ALS) and Frontotemporal Dementia (FTD) are two neurodegenerative disorders known to share disease etiology and pathophysiology. New evidence in the field of ALS-FTD has shown degeneration of specific neural circuits and alterations in synaptic connectivity, contributing to neuronal degeneration, which leads to the impairment of motor commands and executive functions. This evidence is based on studies performed on animal models of disease, post-mortem tissue, and patient derived stem cells. In the present work, we review the existing evidence supporting pathological loss of connectivity and selective impairment of neural circuits in ALS and FTD, two diseases which share strong genetic causes and impairment in motor and executive functions.
Article
Full-text available
Synaptic loss is a pathological feature of all neurodegenerative diseases including amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). ALS is a disease of the cortical and spinal motor neurons resulting in fatal paralysis due to denervation of muscles. FTD is a form of dementia that primarily affects brain regions controlling cognition, language and behavior. Once classified as two distinct diseases, ALS and FTD are now considered as part of a common disease spectrum based on overlapping clinical, pathological and genetic evidence. At the cellular level, aggregation of common proteins and overlapping gene susceptibilities are shared in both ALS and FTD. Despite the convergence of these two fields of research, the underlying disease mechanisms remain elusive. However, recent discovers from ALS and FTD patient studies and models of ALS/FTD strongly suggests that synaptic dysfunction is an early event in the disease process and a unifying hallmark of these diseases. This review provides a summary of the reported anatomical and cellular changes that occur in cortical and spinal motor neurons in ALS and FTD tissues and models of disease. We also highlight studies that identify changes in the proteome and transcriptome of ALS and FTD models and provide a conceptual overview of the processes that contribute to synaptic dysfunction in these diseases. Due to space limitations and the vast number of publications in the ALS and FTD fields, many articles have not been discussed in this review. As such, this review focuses on the three most common shared mutations in ALS and FTD, the hexanucleuotide repeat expansion within intron 1 of chromosome 9 open reading frame 72 (C9ORF72), transactive response DNA binding protein 43 (TARDBP or TDP-43) and fused in sarcoma (FUS), with the intention of highlighting common pathways that promote synaptic dysfunction in the ALS-FTD disease spectrum.
Thesis
Les démences frontotemporales (DFT) et la sclérose latérale amyotrophique (SLA) sont deux maladies neurodégénératives faisant partie d'un même spectre clinique, génétique et neuropathologique. Afin d’identifier une voie biologique commune aux DFTs et SLA, nous nous sommes intéressés aux mécanismes d’activation microgliale. Deux axes d’étude principaux ont composé cette thèse : 1/ La protéine TDP-43, qui forme les inclusions majoritaires des DFT-SLA, est-elle capable d’activer les cellules microgliales ? Par quels mécanismes ? 2/ La perte de fonction des protéines C9ORF72 et PGRN a-t-elle des conséquences sur la réactivité microgliale en présence de TDP-43? Par l’utilisation de modèles in vitro de cultures primaires microgliales nous montrons que la protéine TDP-43 active la voie non canonique microgliale de l'inflammasome NLRP3. Celle-ci se produit suite à l’interaction de TDP-43 avec les récepteurs TLR2/4, et à son internalisation dans les cellules microgliales. Les cellules microgliales murines déficientes C9orf72-/- et GrnR493X/R493X sont hyper-réactives vis-à-vis d’une stimulation par TDP-43. Cet effet est reproduit dans un modèle « MDMi » de cellules microgliales dérivées de monocytes de patients DFT porteurs de mutation GRN ou C9ORF72. Enfin, nous montrons que cet effet peut être lié aux fonctions des protéines PGRN et C9ORF72 nécessaires au bon fonctionnement de la boucle de rétrocontrôle de l’activité de l’inflammasome par la voie autophagique. Nos données mettent en évidence une voie commune de neuroinflammation dans les DFTs et la SLA, et ouvrent de nouvelles pistes sur les processus neurodégénératifs de ces maladies.
Article
Full-text available
Motor neuron-extrinsic mechanisms have been shown to participate in the pathogenesis of ALS-SOD1, one familial form of amyotrophic lateral sclerosis (ALS). It remains unclear whether such mechanisms contribute to other familial forms, such as TDP-43 and FUS-associated ALS. Here, we characterize a single-copy mouse model of ALS-FUS that conditionally expresses a disease-relevant truncating FUS mutant from the endogenous murine Fus gene. We show that these mice, but not mice heterozygous for a Fus null allele, develop similar pathology as ALS-FUS patients and a mild motor neuron phenotype. Most importantly, CRE-mediated rescue of the Fus mutation within motor neurons prevented degeneration of motor neuron cell bodies, but only delayed appearance of motor symptoms. Indeed, we observed downregulation of multiple myelin-related genes, and increased numbers of oligodendrocytes in the spinal cord supporting their contribution to behavioral deficits. In all, we show that mutant FUS triggers toxic events in both motor neurons and neighboring cells to elicit motor neuron disease. Electronic supplementary material The online version of this article (doi:10.1007/s00401-017-1687-9) contains supplementary material, which is available to authorized users.
Article
Full-text available
Graphical Abstract Highlights d Syt2 is the Ca 2+ sensor of fast transmitter release at a cerebellar GABAergic synapse d Syt2 triggers transmitter release with faster time course than Syt1 d Syt2 ensures faster replenishment of the readily releasable pool than Syt1 d Syt2 is essential for fast feedforward inhibition in cerebellar microcircuits In Brief Chen et al. identify synaptotagmin 2 as the major Ca 2+ sensor of transmitter release at cerebellar inhibitory synapses. Furthermore, they demonstrate that synaptotagmin 2 triggers faster release and ensures faster refilling of the vesicular pool than synaptotagmin 1.
Article
Full-text available
Amyotrophic lateral sclerosis (ALS) is a devastating neurological disease with no effective treatment available. An increasing number of genetic causes of ALS are being identified, but how these genetic defects lead to motor neuron degeneration and to which extent they affect common cellular pathways remains incompletely understood. To address these questions, we performed an interactomic analysis to identify binding partners of wild-type (WT) and ALS-associated mutant versions of ATXN2, C9orf72, FUS, OPTN, TDP-43 and UBQLN2 in neuronal cells. This analysis identified several known but also many novel binding partners of these proteins. Interactomes of WT and mutant ALS proteins were very similar except for OPTN and UBQLN2, in which mutations caused loss or gain of protein interactions. Several of the identified interactomes showed a high degree of overlap: shared binding partners of ATXN2, FUS and TDP-43 had roles in RNA metabolism; OPTN- and UBQLN2-interacting proteins were related to protein degradation and protein transport, and C9orf72 interactors function in mitochondria. To confirm that this overlap is important for ALS pathogenesis, we studied fragile X mental retardation protein (FMRP), one of the common interactors of ATXN2, FUS and TDP-43, in more detail in in vitro and in vivo model systems for FUS ALS. FMRP localized to mutant FUS-containing aggregates in spinal motor neurons and bound endogenous FUS in a direct and RNA-sensitive manner. Furthermore, defects in synaptic FMRP mRNA target expression, neuromuscular junction integrity, and motor behavior caused by mutant FUS in zebrafish embryos, could be rescued by exogenous FMRP expression. Together, these results show that interactomics analysis can provide crucial insight into ALS disease mechanisms and they link FMRP to motor neuron dysfunction caused by FUS mutations.
Article
Full-text available
Alzheimer's disease (AD) is a neurodegenerative disorder characterized by progressive memory decline and subsequent loss of broader cognitive functions. Memory decline in the early stages of AD is mostly limited to episodic memory, for which the hippocampus has a crucial role. However, it has been uncertain whether the observed amnesia in the early stages of AD is due to disrupted encoding and consolidation of episodic information, or an impairment in the retrieval of stored memory information. Here we show that in transgenic mouse models of early AD, direct optogenetic activation of hippocampal memory engram cells results in memory retrieval despite the fact that these mice are amnesic in long-term memory tests when natural recall cues are used, revealing a retrieval, rather than a storage impairment. Before amyloid plaque deposition, the amnesia in these mice is age-dependent, which correlates with a progressive reduction in spine density of hippocampal dentate gyrus engram cells. We show that optogenetic induction of long-term potentiation at perforant path synapses of dentate gyrus engram cells restores both spine density and long-term memory. We also demonstrate that an ablation of dentate gyrus engram cells containing restored spine density prevents the rescue of long-term memory. Thus, selective rescue of spine density in engram cells may lead to an effective strategy for treating memory loss in the early stages of AD.
Article
Full-text available
FUS is an RNA-binding protein involved in amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). Cytoplasmic FUS-containing aggregates are often associated with concomitant loss of nuclear FUS. Whether loss of nuclear FUS function, gain of a cytoplasmic function, or a combination of both lead to neurodegeneration remains elusive. To address this question, we generated knockin mice expressing mislocalized cytoplasmic FUS and complete FUS knockout mice. Both mouse models display similar perinatal lethality with respiratory insufficiency, reduced body weight and length, and largely similar alterations in gene expression and mRNA splicing patterns, indicating that mislocalized FUS results in loss of its normal function. However, FUS knockin mice, but not FUS knockout mice, display reduced motor neuron numbers at birth, associated with enhanced motor neuron apoptosis, which can be rescued by cell-specific CRE-mediated expression of wild-type FUS within motor neurons. Together, our findings indicate that cytoplasmic FUS mislocalization not only leads to nuclear loss of function, but also triggers motor neuron death through a toxic gain of function within motor neurons.
Article
Full-text available
Mutations in FUS cause amyotrophic lateral sclerosis (ALS), including some of the most aggressive, juvenile-onset forms of the disease. FUS loss-of-function and toxic gain-offunction mechanisms have been proposed to explain how mutant FUS leads to motor neuron degeneration, but neither has been firmly established in the pathogenesis of ALS. Here we characterize a series of transgenic FUS mouse lines that manifest progressive, mutant-dependent motor neuron degeneration preceded by early, structural and functional abnormalities at the neuromuscular junction. A novel, conditional FUS knockout mutant reveals that postnatal elimination of FUS has no effect on motor neuron survival or function. Moreover, endogenous FUS does not contribute to the onset of the ALS phenotype induced by mutant FUS. These findings demonstrate that FUS-dependent motor degeneration is not due to loss of FUS function, but to the gain of toxic properties conferred by ALS mutations.
Article
Full-text available
The mechanisms by which mutations in FUS and other RNA binding proteins cause ALS and FTD remain controversial. We propose a model in which low-complexity (LC) domains of FUS drive its physiologically reversible assembly into membrane-free, liquid droplet and hydrogel-like structures. ALS/FTD mutations in LC or non-LC domains induce further phase transition into poorly soluble fibrillar hydrogels distinct from conventional amyloids. These assemblies are necessary and sufficient for neurotoxicity in a C. elegans model of FUS-dependent neurodegeneration. They trap other ribonucleoprotein (RNP) granule components and disrupt RNP granule function. One consequence is impairment of new protein synthesis by cytoplasmic RNP granules in axon terminals, where RNP granules regulate local RNA metabolism and translation. Nuclear FUS granules may be similarly affected. Inhibiting formation of these fibrillar hydrogel assemblies mitigates neurotoxicity and suggests a potential therapeutic strategy that may also be applicable to ALS/FTD associated with mutations in other RNA binding proteins.
Article
View largeDownload slide ‘Fused in sarcoma’ ( FUS ) mutations are associated with ALS. Shiihashi et al . generate a mouse model that overexpresses FUS without a nuclear localization signal (ΔNLS-FUS) and shows progressive motor deficits. ΔNLS-FUS forms cytoplasmic aggregates, whereas endogenous FUS is unaffected. Mislocation of FUS is thus sufficient for a gain-of-function ALS phenotype. View largeDownload slide ‘Fused in sarcoma’ ( FUS ) mutations are associated with ALS. Shiihashi et al . generate a mouse model that overexpresses FUS without a nuclear localization signal (ΔNLS-FUS) and shows progressive motor deficits. ΔNLS-FUS forms cytoplasmic aggregates, whereas endogenous FUS is unaffected. Mislocation of FUS is thus sufficient for a gain-of-function ALS phenotype.
Article
A 24-year-old Japanese woman developed anterocollis, weakness of the proximal arms, and subsequent cognitive impairment. A neurological examination revealed amyotrophic lateral sclerosis (ALS) without a family history. Systemic muscle atrophy progressed rapidly. Cerebral MRI clearly exhibited high signal intensities along the bilateral pyramidal tracts. An analysis of the FUS gene revealed a heterozygous two-base pair deletion, c.1507-1508delAG (p.G504WfsX515). A subset of juvenile-onset familial/sporadic ALS cases with FUS gene mutations reportedly demonstrates mental retardation or learning difficulty. Our study emphasizes the importance of conducting a FUS gene analysis in juvenile-onset ALS cases, even when no family occurrence is confirmed.