doi:10.1093/brain/aws006
Brain 2012: 135; 819–832
| 819
BRAIN
A JOURNAL OF NEUROLOGY
Progressive neuronal inclusion formation and
axonal degeneration in CHMP2B mutant
transgenic mice
Shabnam Ghazi-Noori,1 Kristina E. Froud,1 Sarah Mizielinska,2 Caroline Powell,1
Michelle Smidak,1 Mar Fernandez de Marco,1 Catherine O’Malley,1 Michael Farmer,1
Nick Parkinson,1 Elizabeth M. C. Fisher,2 Emmanuel A. Asante,1 Sebastian Brandner,2
John Collinge1,2 and Adrian M. Isaacs2
1 MRC Prion Unit, UCL Institute of Neurology, London WC1N 3BG, UK
2 Department of Neurodegenerative Disease, UCL Institute of Neurology, London WC1N 3BG, UK
Correspondence to: Dr Adrian M. Isaacs,
Department of Neurodegenerative Disease,
UCL Institute of Neurology,
London WC1N 3BG, UK
E-mail: [email protected]
Mutations in the charged multivesicular body protein 2B (CHMP2B) gene cause frontotemporal lobar degeneration. The mutations lead to C-terminal truncation of the CHMP2B protein. We generated Chmp2b knockout mice and transgenic mice expressing either wild-type or C-terminally truncated mutant CHMP2B. The transgenic CHMP2B mutant mice have decreased
survival and show progressive neurodegenerative changes including gliosis and increasing accumulation of p62- and
ubiquitin-positive inclusions. The inclusions are negative for the TAR DNA binding protein 43 and fused in sarcoma proteins,
mimicking the inclusions observed in patients with CHMP2B mutation. Mice transgenic for mutant CHMP2B also develop an
early and progressive axonopathy characterized by numerous amyloid precursor protein-positive axonal swellings, implicating
altered axonal function in disease pathogenesis. These findings were not observed in Chmp2b knockout mice or in transgenic
mice expressing wild-type CHMP2B, indicating that CHMP2B mutations induce degenerative changes through a gain of function
mechanism. These data describe the first mouse model of dementia caused by CHMP2B mutation and provide new insights into
the mechanisms of CHMP2B-induced neurodegeneration.
Keywords: CHMP2B; frontotemporal dementia; transgenic mice; ESCRT
Abbreviations: CHMP2B = charged multivesicular body protein 2B; ESCRT = endosomal sorting complex required for transport;
FTLD = frontotemporal lobar degeneration; FUS = fused in sarcoma; GFAP = glial fibrillary acidic protein; Iba1 = ionized calcium
binding adaptor molecule 1; TDP-43 = TAR DNA binding protein 43
Introduction
Frontotemporal lobar degeneration (FTLD) is the second most
common form of young onset dementia after Alzheimer’s disease
(Ratnavalli et al., 2002; Harvey et al., 2003). It can present with
behavioural changes or with language impairment, dependent on
which parts of the frontal and temporal lobes are affected, and
in a subset of cases is accompanied by either parkinsonism or
Received November 17, 2011. Revised December 28, 2011. Accepted December 29, 2011
ß The Author (2012). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved.
For Permissions, please email: [email protected]
820
| Brain 2012: 135; 819–832
motor neuron disease (Neary et al., 1998; McKhann et al., 2001).
Mutations in the MAPT gene (Hutton et al., 1998; Poorkaj et al.,
1998; Spillantini et al., 1998) encoding the tau protein, in the
GRN gene (Baker et al., 2006; Cruts et al., 2006) encoding progranulin and in the C9ORF72 gene (Dejesus-Hernandez et al.,
2011; Renton et al., 2011) are the most common genetic causes
of FTLD. Rare mutations have been identified in the valosin-containing protein (VCP), TARDBP (which encodes TDP-43), FUS
(fused in sarcoma) and charged multivesicular body protein 2B
(CHMP2B) genes (Watts et al., 2004; Skibinski et al., 2005; van
der Zee et al., 2008; Benajiba et al., 2009; Borroni et al., 2009;
Kovacs et al., 2009; Van Langenhove et al., 2010). FTLD is characterized neuropathologically by the presence of either tau- or
ubiquitin-positive intraneuronal inclusions (Cairns et al., 2007).
The ubiquitin-positive inclusions generally contain the ubiquitin
binding protein p62 (Pikkarainen et al., 2008) and either TAR
DNA binding protein 43 (TDP-43, the most common subtype)
or FUS (Mackenzie et al., 2010). CHMP2B mutation cases
belong to a rare subtype that have ubiquitin- and p62-positive
inclusions that are negative for both TDP-43 and FUS (Holm
et al., 2007, 2009; Urwin et al., 2010b).
CHMP2B is part of ESCRT-III (endosomal sorting complex
required for transport-III), a multiprotein complex involved in
endosomal and autophagic trafficking of proteins to lysosomes
for degradation (Katzmann et al., 2002; Rusten and Stenmark,
2009). There is evidence that CHMP2B mutations can affect
endosomal and autophagic trafficking (Filimonenko et al., 2007;
Lee et al., 2007; Metcalf and Isaacs, 2010; Urwin et al., 2010a),
but the mechanisms that lead to neuronal cell death are unknown
(Lee et al., 2009; Urwin et al., 2009).
The first CHMP2B mutation was identified in a large Danish
FTLD family (Skibinski et al., 2005) and a second mutation was
subsequently identified in a Belgian patient with familial FTLD (van
der Zee et al., 2008). The Danish mutation occurs in the acceptor
splice site of the sixth and final CHMP2B exon, leading to the
formation of two aberrant transcripts termed CHMP2Bintron 5 and
CHMP2B10 which both replace the final 36 amino acids of
CHMP2B with novel C-termini due to mis-splicing (Skibinski
et al., 2005). The Belgian mutation is a truncation mutation that
removes the final 48 amino acids, suggesting that loss of the
C-terminus of CHMP2B may be a common mechanism by
which these mutations lead to disease. However, both mutations
also lead to the loss of one normal copy of CHMP2B, suggesting
that loss of function could also play a role in disease pathogenesis.
To distinguish between these possibilities, and to generate the first
mouse models of CHMP2B-induced FTLD, we generated Chmp2b
knockout mice and transgenic mice expressing either
CHMP2Bwild-type or CHMP2Bintron 5, which is the mutant transcript
expressed at the highest level in patient brains (Urwin et al.,
2010a). We provide evidence in favour of a gain of function
mechanism in CHMP2B-induced FTLD by showing that neuropathology mimicking human patients are observed only in
CHMP2Bintron 5 mice and not in CHMP2Bwild-type mice or
Chmp2b knockout mice. We also identify axonal changes as an
early degenerative feature in CHMP2Bintron 5 mice, suggesting a
potential role for altered axonal function in the disease process.
S. Ghazi-Noori et al.
Materials and methods
Generation of CHMP2B transgenic mice
CHMP2Bwild-type and CHMP2Bintron 5 were polymerase chain reaction
amplified from previously generated expression plasmids (Skibinski
et al., 2005) using primers that added a SalI restriction site to the 50
end of the gene and an XhoI site to the 30 end of the gene. The
polymerase chain reaction products were TOPOÕ cloned into pCR
2.1 (Invitrogen) and sequence verified. The CHMP2Bwild-type and
CHMP2Bintron 5 complementary DNA fragments were then isolated
with SalI and XhoI. Subsequent subcloning into the cosmid vector
SHaCosTt (Scott et al., 1989) and preparation of high quality DNA
of the insert was as previously reported (Asante et al., 2002). The
purified transgenes were microinjected into single cell eggs of
(CBA C57BL/6) mice. The injected eggs were cultured to the
2-cell stage and then surgically transferred to F1 (CBA C57BL/6)
recipient females. Tail biopsies were taken from putative transgenics
and screened by polymerase chain reaction using human CHMP2B
specific primers 50 -gtcgacaccatggcgtccctcttcaagaag-30 and 50 -ctcgagc
taatctactcctaaagccttgag-30 . Founders were bred to non-transgenic
C57BL/6 mice and transgene expression assessed by real
time-polymerase chain reaction and confirmed by immunoblotting.
Two CHMP2Bintron 5 lines were generated, Tg153 and Tg156. The
Tg153 line was interbred to generate homozygous mice and was
maintained as a homozygous line. Homozygous Tg153 mice were
used for all experiments described unless otherwise stated. The
Tg156 line would not breed as a homozygous line so was maintained
as a hemizygous line by interbreeding and was used for confirmation
of pathological findings (Supplementary Fig. 1).
Two CHMP2Bwild-type lines were generated, Tg167 and Tg168, and
were maintained as hemizygous lines by interbreeding. Non-transgenic
littermates from both Tg167 and Tg168 matings were used as controls. The lower expressing line Tg168 was used for all experiments
described unless otherwise stated and Tg167 used for confirmatory
experiments.
Generation of Chmp2b knockout mice
An in silico search of the BayGenomics genetrap resource identified
that embryonic stem cell line XL952 (genetic background 129P2/
OlaHsd) had a genetrap insertion (vector pGTOLxf) in intron 2 of
Chmp2b (between exons 2 and 3). XL952 embryonic stem cells
were injected into host blastocysts to generate chimeras. Two male
chimeras predicted to be 490% chimeric based on coat colour were
bred to C57BL/6J females and germline transmission confirmed. Sperm
from the resulting F1 males was recovered and in vitro fertilization
with C57BL/6J eggs was used to produce 2-cell embryos for shipment
to the UK. Embryos were implanted into pseudopregnant females and
the resulting mice interbred to generate homozygous knockouts.
Sequencing identified the insertion site of the genetrap cassette to
be 685 bp upstream of the beginning of intron 2.
Survival analysis
We analysed survival of our Chmp2b knockout colony (Chmp2b / n = 125, Chmp2b + / + n = 117), and CHMP2B transgenic lines Tg153
(n = 18), Tg168 (n = 16), and non-transgenics (n = 21). Kaplan–Meier
survival analysis was performed using SPSS 15.0, and a Log Rank test
used to determine significance.
Neurodegenerative changes in CHMP2B mutant mice
Generation of CHMP2B antibodies
Ten rabbit polyclonal antisera against full length recombinant CHMP2B
were generated by Biogenes and tested for immunoblotting and
immunohistochemistry. The optimal antiserum for western blotting
and immunohistochemistry were designated 3335 and 3371,
respectively.
Brain 2012: 135; 819–832
| 821
and representative pictures taken. Antibodies used were: anti-amyloid
precursor protein (MAB348, Millipore) 1:2000; anti-CHMP2B 3371 (in
house rabbit polyclonal) 1:400; anti-FUS (NB100-565, Novus) 1:100;
anti-glial fibrillary acidic protein (GFAP) (Z0334, Dako) 1:1000;
anti-Iba1 (019-19741, Wako) 1:250; anti-p62 (GP62-C, Progen Biotechnik) 1:100; anti-synaptophysin (080130, prediluted from Invitrogen); anti-TDP-43 (10782-2-AP, ProteinTech Group), 1:10 000;
anti-ubiquitin (SC-8017, Santa Cruz Biotechnology) 1:10 000.
Quantification of RNA transcripts
RNA was extracted from frozen brain using TRIzol (Invitrogen), treated
with TURBO DNase (Ambion) and purified using an RNeasy Mini
Column (Qiagen). Complementary DNA was generated using the
Omniscript Reverse Transcription Kit (Qiagen). CHMP2B RNA was
quantified with real-time polymerase chain reaction using primers
and a TaqManÕ probe that detected both human and mouse
CHMP2B (see Supplementary Table 1 for primer sequences). Total
CHMP2B RNA levels for three mice per genotype at 6 months of
age were compared after normalization to GAPDH.
Immunoblotting
Brain homogenates (10% w/v) were prepared in Dulbecco’s phosphate-buffered saline lacking Ca2 + or Mg2 + ions (D-PBS) using a
RiboLyser homogenizer (Hybaid), boiled in the appropriate volume
of 2 Laemmli sample buffer and run on 4–20% or 12% sodium
dodecyl sulphate-polyacrylamide gel electrophoresis gels (Invitrogen),
transferred onto polyvinylidene fluoride or nitrocellulose and probed
with primary antibodies as follows: anti-CHMP2B 3335 (in-house
rabbit polyclonal used for the detection of CHMP2B in the transgenic
mouse lines) 1:2000; anti-CHMP2B C-terminus (ab33174, Abcam),
1:1000; anti-p62 (H00008878-MO1, Abnova) 1:40 000. Detection
was performed with horseradish peroxidase-conjugated secondary
antibodies and SuperSignal West Pico Chemiluminescent Substrate
(Thermo scientific). Quantification was performed using Volocity software (Perkin Elmer), bands were normalized to actin and averages
taken of three mice per genotype. For fractionation, 10% (w/v)
brain homogenates were prepared in D-PBS and combined 1:1 with
2% sarkosyl (N-lauroylsarcosine) in D-PBS. Benzonase (50 U/ml;
Novagen) were added to remove DNA and the homogenates incubated with constant agitation at 37 C for 30 min, followed by ultracentrifugation at 100 000 g for 30 min at 4 C. The supernatant was
removed and saved as the soluble fraction. Pellets were washed in 1%
sarkosyl in D-PBS and spun at 100 000 g for 30 min. The supernatant
was discarded and the pellet solubilized in 4% sodium dodecyl sulphate, followed by the addition of 2 Laemmli sample buffer and
heating at 100 C for 10 min prior to sodium dodecyl
sulphate-polyacrylamide gel electrophoresis.
Immunohistochemistry
Mouse brain and spinal cord were immersion fixed in 10% buffered
formaldehyde-saline, embedded in paraffin wax and sections cut at
3–4 mm nominal thickness. Antigen retrieval was performed with
citrate buffer or proteases depending on the antibody. Immunohistochemical staining was carried out on an automated immunostaining
machine (Ventana Medical Systems Inc.). Sections were developed
with 30 3 diaminobenzidine tetrachloride as the chromogen and counterstained with haematoxylin. Bright field photographs were taken on
an ImageView II 3.5 Mpix digital camera (www.soft-imaging.de)
mounted on a ZEISS Axioplan microscope and composed with
Adobe Photoshop. At least three mice were analysed per genotype
Quantification of pathological
findings
Quantification was performed blinded on four mice per genotype
(CHMP2Bwild-type, CHMP2Bintron 5 and non-transgenics) at 6, 12 and
18 months of age using brightfield images taken as described earlier.
For p62 inclusion quantification in brain, three 40 images
(21 069 mm2 per image) were taken in the thalamus and cortex and
six in the corpus callosum and inclusions counted using the touch
count tool in analySIS imaging software (Olympus). For p62 inclusion
quantification in spinal cord, motor neurons were identified by size and
location and scored as either inclusion positive or negative. At least 50
motor neurons were counted in at least 3 lumbar spinal cord sections
per mouse. For quantification of GFAP and Iba1 staining, 6–10 10
images (335 628 mm2 per image) were taken in the cortex and corpus
callopsum and one in the thalamus. The same threshold for GFAP
staining was applied to all images using VolocityÕ image analysis software (Perkin Elmer) and the area covered by GFAP immunoreactivity
quantified. Activation of Iba1-positive microglia was performed by
scoring each image from 1–4 (1 = ramified, 2 = reactive, 3 = amoeboid
and 4 = phagocytic) (Supplementary Fig. 3), as previously described
(Ahmed et al., 2010), and taking the average of all images per
region. Statistical testing was performed with Graphpad Prism 5 software using a two-way ANOVA and post hoc Bonferroni test analysing
selected pairs of columns.
Electron microscopy
Mice were perfused with 3% glutaraldehyde in phosphate buffer and
brains removed and post-fixed overnight in the same fixative. Brains
were treated with 1% osmium tetroxide for 3 h at 4 C and embedded
in AralditeÕ CY212 resin. Ultrathin sections (70 mm) were stained with
lead citrate and uranyl acetate and digital images taken on a Phillips
CM10 electron microscope with a Megaview III digital camera
(Olympus).
Results
Generation of CHMP2B transgenic mice
and Chmp2b knockout mice
We generated two independent lines expressing human
CHMP2Bintron 5 (designated Tg153 and Tg156) and two lines expressing human CHMP2Bwild-type (Tg167 and Tg168) under the
control of the hamster prion protein gene promoter (Fig. 1A). To
determine protein expression levels of each line, we generated
polyclonal antibodies directed against full-length recombinant
CHMP2B that were able to detect both CHMP2Bwild-type and
CHMP2Bintron 5. Both CHMP2Bintron 5 lines expressed the protein
822
| Brain 2012: 135; 819–832
S. Ghazi-Noori et al.
Figure 1 Generation and expression of CHMP2B transgenic and Chmp2b / mice. (A) Cloning strategy for the generation of CHMP2B
transgenic mice. CHMP2Bwild-type and CHMP2Bintron 5 complementary DNAs were cloned into the SalI sites of cosmid vector SHaCosTt
(Scott et al., 1989), which contains the hamster (Ham) prion protein (PrP) gene promoter, non-coding exon 1, intron 1 and 30 untranslated
region (UTR). Black double-headed arrows indicate the open reading frame (ORF). The CHMP2Bintron 5 complementary DNA reads
directly into intron 5 due to the causative mutation in the acceptor splice site of exon 6 (Skibinski et al., 2005). The second codon of intron
5 is a stop codon. (B) Schematic of Chmp2b gene-trap insertion. The gene-trap vector is located in intron 2 of Chmp2b and is composed of
a splice acceptor site (SA), a beta-galactosidase-neomycin fusion protein (b-geo) and a polyA tail (pA). (C) CHMP2B immunoblot of a
non-transgenic (non-Tg) mouse, human CHMP2Bintron 5 lines Tg153 and Tg156 and human CHMP2Bwild-type lines Tg167 and Tg168. A
non-specific band is present in between CHMP2Bwild-type and CHMP2Bintron 5. (D) CHMP2B immunoblot of Chmp2b / and Chmp2b + / +
mice. (E) Semi-quantitative real time-polymerase chain reaction shows human (h) CHMP2B RNA levels relative to mouse Chmp2b in
non-transgenics are 3.26 0.37 (SEM) for Tg153 and 3.20 0.04 (SEM) for Tg168. The average of three mice per line is shown. (F)
Protein levels of transgenic human CHMP2B relative to mouse CHMP2B in non-transgenics are 0.9 0.1 (SEM) for Tg153 and 6.6 0.4
(SEM) for Tg168. The average of three mice per line is shown. (G) Quantification shows Chmp2b / mice have 0.16 0.02 (SEM) the
level of CHMP2B protein in Chmp2b + / + mice. The average of three mice per line is shown.
at levels similar to endogenous mouse CHMP2B, while both
CHMP2Bwild-type lines expressed at levels higher than endogenous
mouse CHMP2B (Fig. 1C). We focused our analyses on Tg153 and
Tg168 for which transgenic human CHMP2B protein levels were
0.9 0.1 (SEM) and 6.6 0.4 (SEM), respectively relative to
mouse CHMP2B (Fig. 1F). Transgene RNA levels were 3-fold
higher than endogenous mouse Chmp2b RNA levels for both
Tg153 and Tg168 (Fig. 1E).
We generated homozygous Chmp2b knockout mice
(Chmp2b / ) from a BayGenomics embryonic stem cell line
with a genetrap insertion in intron 2 of Chmp2b (Fig. 1B).
Analysis of protein levels showed that the knockout is incomplete
Neurodegenerative changes in CHMP2B mutant mice
Brain 2012: 135; 819–832
| 823
Figure 2 Survival analysis of CHMP2B transgenic and Chmp2b / mouse lines. (A) Decreased survival in CHMP2Bintron 5 mice,
P = 0.001, Kaplan–Meier analysis and Log Rank test. (B) Survival of Chmp2b / mice is not significantly different to controls.
Figure 3 p62-positive inclusions in brain and spinal cord of CHMP2Bintron 5 mice. p62 inclusions were identified in several brain regions
including cortex, thalamus, brainstem and spinal cord motor neurons of CHMP2Bintron 5 mice but not CHMP2Bwild-type, non-transgenic
or Chmp2b / mice. Chmp2b / mice are 24 months of age, all other mice are 18 months of age. Dot and thread-like inclusions
were observed as well as intraneuronal inclusions, e.g. arrows in D and T. Scale bars: A–P = 80 mm; Q–T = 50 mm.
824
| Brain 2012: 135; 819–832
S. Ghazi-Noori et al.
(Fig. 1D), leading to an 84% reduction in CHMP2B levels
(Fig. 1G).
Reduced survival in CHMP2Bintron 5 mice
Kaplan–Meier survival analysis showed that CHMP2Bintron 5 mice
had decreased survival when compared with CHMP2Bwild-type
mice and non-transgenic controls (P = 0.001) (Fig. 2A).
Chmp2b / mice did not have significantly different survival to
controls (Fig. 2B).
Progressive inclusion formation
in CHMP2Bintron 5 mice
CHMP2Bintron 5 mice developed an age-dependent accumulation
of p62-positive inclusions that were abundant by 18 months of
age and were observed in a range of brain regions including frontal cortex, brainstem, thalamus and corpus callosum, as well as in
motor neurons in the spinal cord (Fig. 3). This pattern of pathology is likely to reflect the distribution of transgene expression
driven by the hamster prion protein gene promoter. Large neuronal inclusions (examples are indicated by black arrows in Figs 3
and 5) as well as dot and thread-like inclusions in the neuropil
(examples are indicated by white arrows in Fig. 5) were observed.
We compared 18-month-old transgenic mice to 24-month-old
Chmp2b / mice to ensure that we were not missing age-related
pathology in the knockouts. Inclusions were not observed in any
brain region in Chmp2b / mice, CHMP2Bwild-type transgenics or
in non-transgenics (Fig. 3), but were observed in our second
CHMP2Bintron 5 line (Supplementary Fig. 1). This suggests that inclusion formation is a specific effect of mutant CHMP2B.
We next defined the extent and localization of inclusion pathology in 18-month-old CHMP2Bintron 5 brain and spinal cord (Fig.
4). The spinal cord showed widespread inclusions throughout the
grey matter, with inclusions generally less frequent in the white
matter other than a concentration in the dorsal corticospinal tract
(Fig. 4B). The brain also showed widespread inclusion pathology
which we scored as mild, moderate or severe (Fig. 4A). We then
chose the cortex, corpus callosum and thalamus, which showed
mild, moderate and severe inclusion pathology respectively, for
quantification of inclusion number at 6, 12 and 18 months of
age. Inclusions were observed in each area at 6 months and
increased with age, with inclusions more frequent in the thalamus
and corpus callosum than the cortex (Figs 5 and 6). Quantification
of the percentage of motor neurons containing inclusions at 6, 12
and 18 months of age showed 5% of motor neurons contained
inclusions at 6 months of age, rising to 30% at 18 months of age
(Fig. 6).
The formation of insoluble inclusions containing p62 was confirmed using immunoblotting of sarkosyl insoluble fractions from
the transgenic mouse brains. The level of p62 in the insoluble
fraction was increased in CHMP2Bintron 5 mice when compared
to CHMP2Bwild-type mice and non-transgenics (Fig. 7).
Further immunohistochemical analysis of the inclusions showed
that they were also positive for ubiquitin but negative for TDP-43
and FUS (Fig. 8), mimicking the composition of inclusions in
CHMP2B mutation patient brains (Holm et al., 2007, 2009).
Figure 4 Distribution of p62 inclusion pathology in
18-month-old CHMP2Bintron 5 mouse brain and spinal cord.
(A) Inclusion frequency was qualitatively assessed in four
18-month-old CHMP2Bintron 5 mouse brains and then scored as
mild (yellow), moderate (orange), or severe (red). Mild inclusion
pathology is observed throughout the cortex, in the white
matter of the cerebellum and in the superior colliculus, other
than the superficial layer which has severe pathology. Moderate
pathology was observed in the corpus callosum and brainstem,
other than occasional axonal tracts in the brainstem which had
severe pathology. Severe inclusion pathology was observed in
the thalamus. (B) Every inclusion in a representative lumbar
spinal cord section was traced onto a schematic with large and
small dots indicating the size of the inclusions. Inclusions are
abundant throughout the grey matter and less common in the
white matter, other than the dorsal corticospinal tract.
To determine whether CHMP2B was also present in the inclusions
we generated antibodies against full-length recombinant
CHMP2B. CHMP2B was present in a minority of the inclusions
(Fig. 8), specifically large inclusions rather than the dot and
thread-like inclusions.
Age-related gliosis in CHMP2Bintron 5
mice
Immunohistochemical analysis of astrocytes and microglia was performed at 6, 12 and 18 months of age in Chmp2b / ,
CHMP2Bwild-type, CHMP2Bintron 5 and non-transgenic mice.
Progressive activation of both astrocytes and microglia was
observed specifically in CHMP2Bintron 5 mouse brains (Fig. 9 and
Supplementary Fig. 2). Quantification of GFAP immunoreactivity
in the cortex, thalamus and corpus callosum of CHMP2Bintron 5
mouse brain showed no evidence of astrogliosis at 6 months of
age, suggesting that astrogliosis occurs after inclusion formation;
Neurodegenerative changes in CHMP2B mutant mice
Brain 2012: 135; 819–832
| 825
Figure 5 p62-positive inclusions increase with age in CHMP2Bintron 5 mouse brain. Representative images of the cortex, corpus callosum,
thalamus and ventral horns of the spinal cord in 6-, 12- and 18-month-old CHMP2Bintron 5 mice. Examples of inclusion types are shown in
the cortex panel: larger neuronal inclusions (black arrows), and dot and thread-like inclusions (white arrows). Scale bar = 80 mm.
however, by 12 months of age both the cortex and thalamus
showed proliferation of GFAP-positive astrocytes (Figs 9, 10A
and B). The astrogliosis was more marked in the thalamus (Fig.
10A and B), mirroring the severity of inclusion formation.
Astrocytes were abundant in the corpus callosum in mice of all
genotypes so no differences were observed (data not shown).
We also quantified the activation of Iba1-positive microglia, classifying them in increasing order of activation as either ramified,
reactive, amoeboid or phagocytic (Supplementary Fig. 3) as previously described (Ahmed et al., 2010). Reactive microglia were
observed in the thalamus of CHMP2Bintron 5 mice at 6 months of
age, and progressed to amoeboid by 12 months and phagocytic
by 18 months of age (Figs 9 and 10D). Microgliosis was also
observed in the cortex and corpus callosum, but was less pronounced, with some reactive microglia observed at 6 and 12
months, but not consistently until 18 months of age (Fig. 10C
and E). The appearance of reactive microglia at 6 months of age
in the thalamus indicates that microgliosis occurs prior to astrogliosis in this brain region. The presence of marked gliosis shows
that CHMP2Bintron 5 expression can lead to severe pathological
changes reminiscent of those in human patients.
Progressive axonopathy in
CHMP2Bintron 5 mice
Due to the presence of inclusions in the corpus callosum, we further investigated axonal pathology using amyloid precursor protein
staining, which is a sensitive marker of axonal damage (Coleman,
2005). Amyloid precursor protein staining revealed axonal swellings in the corpus callosum of CHMP2Bintron 5 mice but not in
CHMP2Bwild-type, Chmp2b / knockout or non-transgenic mice
(Fig. 11). The swellings were apparent at 6 months and increased
with age. Electron microscopy showed that the swellings contained an accumulation of mitochondria and vesicles from endolysosomal and autophagy pathways (Fig. 12). We also observed
accumulation of the presynaptic marker synaptophysin in
18-month-old CHMP2Bintron 5 thalamus (Supplementary Fig. 4),
further suggestive of axonal dysfunction. These findings suggest
826
| Brain 2012: 135; 819–832
S. Ghazi-Noori et al.
Figure 6 Quantification of p62 inclusions in CHMP2Bintron 5, CHMP2Bwild-type and non-transgenic brain and spinal cord. Blind counts
of p62 inclusion number were performed at 6, 12 and 18 months of age in the cortex, corpus callosum, thalamus and spinal cord
of non-transgenic (non-Tg), CHMP2Bintron 5 and CHMP2Bwild-type mice (n = 4 per genotype). Inclusions are significantly increased
in CHMP2Bintron 5 mice and increase with age. *P 5 0.05, **P 5 0.01, ***P 5 0.001, ****P 5 0.0001, two-way ANOVA with
Bonferroni post-test.
Figure 7 Accumulation of p62 in the insoluble fraction of
CHMP2Bintron 5 mouse brain. Brain homogenates from
18-month-old CHMP2Bintron 5, CHMP2Bwild-type and
non-transgenic mice were extracted in 1% sarkosyl and the
soluble and insoluble fractions immunoblotted for p62. The
amount of p62 in the insoluble fraction is increased in
CHMP2Bintron 5 brain when compared to CHMP2Bwild-type
and non-transgenic brain.
that early axonal impairment could contribute to neuronal dysfunction in FTLD caused by CHMP2B mutations.
Discussion
We have generated the first mouse model of CHMP2B-induced
neurodegeneration and show that expression of mutant
CHMP2Bintron 5 leads to striking neurodegenerative changes that
are not observed in CHMP2Bwild-type transgenics or Chmp2b
knockout mice. Importantly, the neurodegenerative changes
were very similar to those observed in patients with CHMP2B
mutation, particularly the formation of p62- and ubiquitin-positive,
but TDP-43- and FUS-negative neuronal inclusions. These data
also show that the overt neuropathology observed in patients
with CHMP2B mutation is caused by toxic gain of function of
CHMP2Bintron 5 and not loss of function of CHMP2B or a
dominant negative effect on CHMP2B function. As 15% of
mouse CHMP2B protein levels remain in our Chmp2b / line,
we cannot completely rule out the possibility that loss of function
contributes to neuropathological changes, however, this seems
unlikely as the protein level is well below the 50% reduction
observed in patients and no neuropathological changes were
observed up to 2 years of age, which is towards the maximal lifespan of a laboratory mouse (only 30% of both controls and
knockouts survived to 2 years of age). Another possibility is that
loss of function of CHMP2B contributes to other aspects of the
disease phenotype. Such a phenomenon has been described for
spinocerebellar ataxia type 1 (SCA1), which is caused by an expanded CAG repeat in the ATXN1 gene. Loss of Atxn1 in mice
does not lead to neurodegenerative changes (Matilla et al., 1998)
while expression of the expanded repeat does (Watase et al.,
2002). However, it has now been demonstrated that partial loss
of function of Atxn1 contributes to other features of the disease
including behavioural changes and transcriptional dysregulation
(Lim et al., 2008; Crespo-Barreto et al., 2010).
Neurodegenerative changes in CHMP2B mutant mice
Brain 2012: 135; 819–832
| 827
Figure 8 Inclusions in CHMP2Bintron 5 mice are ubiquitin-positive and TDP-43- and FUS-negative. Ubiquitin staining in 18-month-old
mice reveals numerous inclusions that are not detected with TDP-43 and FUS staining. CHMP2B staining identifies a subset of generally
large inclusions (arrows). Scale bars: 80 mm for CHMP2B and ubiquitin panels and 50 mm for TDP-43 and FUS panels.
Figure 9 Progressive gliosis in CHMP2Bintron 5 mouse brain. GFAP and Iba1 staining in the thalamus of 6-, 12- and 18-month-old
CHMP2Bintron 5 and non-transgenic mice. Proliferation of GFAP-positive astrocytes and Iba1-positive microglia is specific to
CHMP2Bintron 5 mice. Scale bar = 160 mm.
828
| Brain 2012: 135; 819–832
S. Ghazi-Noori et al.
Figure 10 Quantification of gliosis in CHMP2Bintron 5, CHMP2Bwild-type and non-transgenic mouse brains. Quantification of astrogliosis
(area covered by GFAP immuonreactivity) and microgliosis (activation of Iba1-positive microglia) was performed blind in the cortex,
corpus callosum and thalamus of non-transgenic, CHMP2Bintron 5 and CHMP2Bwild-type mice (n = 4 per genotype). CHMP2Bintron 5 mice
showed increased astrogliosis at 12 and 18 months of age in the cortex and thalamus, but were not significantly different to controls at 6
months of age. No difference in astrogliosis was observed at any age in the corpus callosum as astrocytes were abundant in all genotypes
(data not shown). CHMP2Bintron 5 mice showed significantly increased microgliosis at 6, 12 and 18 months in the thalamus and at 18
months only in the cortex and corpus callosum. *P 5 0.05, **P 5 0.01, ***P 5 0.001, ****P 5 0.0001, two-way ANOVA with
Bonferroni post-test.
We showed that inclusion and glial pathology is widespread and
progressive in CHMP2Bintron 5 brain and spinal cord. This distribution of pathology will be influenced by the pattern of gene expression driven by the hamster prion protein gene promoter so is
unlikely to be identical to that seen in patients. We observed cortical pathology, as well as abundant pathology in the thalamus,
which is consistent with thalamic expression in other transgenics
using the hamster prion protein gene promoter (Asante et al.,
2002, 2006). The other striking pathology in CHMP2Bintron 5
mice was the presence of axonal swellings, which combined
with gliosis and inclusion formation indicate that a neurodegenerative process is occurring. Future investigations to quantify
neuron and axon numbers with stereological techniques will be
important to address the extent of axon and cell loss.
Similar pathological changes have been observed in progranulin
knockout mice. Loss of function mutations in progranulin are a
relatively common cause of FTLD suggesting knockout mice may
be a useful model to investigate disease pathogenesis. Progranulin
knockout mice exhibit progressive accumulation of ubiquitinpositive inclusions and progressive activation of microglia and
astrocytes (Ahmed et al., 2010; Yin et al., 2010a, b; Petkau
et al., 2011; Ghoshal et al., 2012). This pathology was observed
in the cortex, but to a greater extent in the hippocampus, brainstem and thalamus (Ahmed et al., 2010). This pattern was suggested to reflect brain areas with high levels of progranulin
expression that became particularly affected after progranulin deletion (Ahmed et al., 2010). The neuronal ubiquitin-positive inclusions were described as granular and cytoplasmic (Ahmed et al.,
2010; Petkau et al., 2011; Ghoshal et al., 2012) so appear morphologically distinct from the large dense neuronal aggregates that
we observed in CHMP2Bintron 5 mice. However, as both CHMP2B
and progranulin have been implicated in lysosomal degradation
Neurodegenerative changes in CHMP2B mutant mice
Brain 2012: 135; 819–832
| 829
Figure 11 Progressive axonal damage in CHMP2Bintron 5 mice. Amyloid precursor protein staining revealed numerous axonal swellings
within the corpus callosum at both 6 and 12 months of age that were not present in CHMP2Bwild-type, non-transgenic or Chmp2b /
mice. Scale bar = 160 mm.
pathways (Filimonenko et al., 2007; Hu et al., 2010; Urwin et al.,
2010a), it will be interesting to determine whether they share
overlapping disease mechanisms.
The generation of the first CHMP2Bintron 5 mouse model provides new insight into the mechanisms by which mutant CHMP2B
leads to neuronal dysfunction in vivo. CHMP2B is part of the
ESCRT-III complex, which has been implicated in autophagic and
endosomal trafficking to the lysosome (Katzmann et al., 2002;
Rusten and Stenmark, 2009). It has been shown that expression
of CHMP2Bintron 5 in cell culture can lead to the formation of
ubiquitin- and p62-positive inclusions (Filimonenko et al., 2007;
Lee et al., 2007) mimicking the data described here in transgenic
mice. The formation of inclusions is likely to be due to impaired
lysosomal degradation through the autophagy and endosomelysosome pathways (Filimonenko et al., 2007; Lee et al., 2007;
Metcalf and Isaacs, 2010; Urwin et al., 2010a). How impairment
of such central degradative processes leads to selective neuronal
loss remains unresolved. One possibility is that CHMP2B is
required for the degradation of specific substrates that are
vital for neuronal function. For example, it has been shown
that levels of the Toll receptor are increased in Drosophila
when CHMP2Bintron 5 is overexpressed leading to the suggestion
830
| Brain 2012: 135; 819–832
S. Ghazi-Noori et al.
Figure 12 Ultrastructure of axonal swellings in 12-month-old CHMP2Bintron 5 mice. (A) Electron microscopy in 12-month-old
CHMP2Bintron 5 corpus callosum showing an axonal swelling containing mitochondria and numerous vesicles of the endolysosomal
and autophagy pathways. (B) A similar swelling in 12-month-old CHMP2Bintron 5 thalamus. Scale bar = 2 mm.
that altered Toll-like receptor signalling may play a role in
CHMP2B-induced neurodegeneration (Ahmad et al., 2009). An
impairment of neuron-specific endosomal trafficking functions
could also provide cell-type specificity. Recent data show that
CHMP2Bintron 5 can affect dendritic spine morphology in primary
neurons (Belly et al., 2010) and combined with evidence that
other ESCRT proteins can promote neurotransmitter receptor trafficking to lysosomes (Kantamneni et al., 2009), it is possible that
dendritic neurotransmitter levels could be altered by mutant
CHMP2B and contribute to neuronal dysfunction. Our identification of early and progressive axonal swellings in CHMP2Bintron 5
mice raises the possibility that axonal transport could also be affected. Recruitment of Rab7 was shown to be an important step
for rendering an endosome capable of long range retrograde
axonal trafficking (Deinhardt et al., 2006). We have previously
shown that Rab7 recruitment to endosomes is impaired by
CHMP2Bintron 5 (Urwin et al., 2010a), suggesting a potential
mechanism for impaired axonal transport. Altered axonal transport
has been implicated in a range of neurodegenerative diseases and
further deciphering the mechanisms and specific cargoes affected
is an important unanswered question (Wu et al., 2009; Perlson
et al., 2010). Electron microscopy showed that at 12 months of
age axonal swellings contained mitochondria and vesicles from the
endo-lysosomal and autophagy pathways. Axonal accumulations
of such vesicles have also been observed in other neurodegenerative conditions including Lurcher mice, which suffer an excitotoxic
Purkinje cell loss (Wang et al., 2006), and in dystrophic neurites in
Alzheimer’s disease (Nixon et al., 2005). Axonal swellings in motor
neuron disease have been described to contain densely packed
neurofilaments (Delisle and Carpenter, 1984; Tu et al., 1996;
Laird et al., 2008), while both neurofilament and vesicle-laden
axonal swellings have been observed in mutant tau mouse
models (Spittaels et al., 1999; Probst et al., 2000; Lin et al.,
2003). Further work will be required to determine whether the
contents of the swellings change over time and if this is related
to the neurodegenerative mechanism.
The exact role of mutant CHMP2B in inclusion formation will
also require further investigation. Our observation that it is only
present in larger inclusions suggests that another protein is accumulating as a result of CHMP2B mutation. Identification of the
accumulating protein will be an important step towards defining
the full complement of aggregating proteins in FTLD, and the
CHMP2Bintron 5 mice will be a valuable tool for this endeavour.
In this context, it is intriguing that the recently discovered FTLDand motor neuron disease-causing repeat expansion in C9ORF72
(Dejesus-Hernandez et al., 2011; Renton et al., 2011) leads to
abundant p62- and ubiquitin-positive inclusions that are also
negative for TDP-43 and FUS (Al Sarraj et al., 2011; Murray
et al., 2011). It will be important to determine whether the
same protein is accumulating in C9ORF72 and CHMP2B mutation
cases. In conclusion, the development of a mouse model that
faithfully recapitulates aspects of the pathology observed in
CHMP2B mutation patients has provided new insight into the
mechanism of CHMP2B-induced neuronal dysfunction in vivo,
and is an important new model for further dissecting this degenerative pathway.
Acknowledgements
We thank the Frontotemporal Dementia Research in Jutland
Association (FReJA) for ongoing collaboration and critical review
of the manuscript: Peter Johannsen, Jorgen Nielsen, Ida Holm,
Troels Tolstrup Nielsen, Anders Gade, Tove Thusgaard, Susanne
Gydesen, Elisabet Englund, Jerry Brown and Martin Rossor. We
thank Kerrie Venner and Mike Groves for help with electron
microscopy and Ray Young for help with the preparation of
figures.
Funding
UK Medical Research Council.
Neurodegenerative changes in CHMP2B mutant mice
Supplementary material
Supplementary material is available at Brain online.
References
Ahmad ST, Sweeney ST, Lee JA, Sweeney NT, Gao FB. Genetic screen
identifies serpin5 as a regulator of the toll pathway and CHMP2B
toxicity associated with frontotemporal dementia. Proc Natl Acad Sci
USA 2009; 106: 12168–73.
Ahmed Z, Sheng H, Xu YF, Lin WL, Innes AE, Gass J, et al. Accelerated
lipofuscinosis and ubiquitination in granulin knockout mice suggest a
role for progranulin in successful aging. Am J Pathol 2010; 177:
311–24.
Al Sarraj S, King A, Troakes C, Smith B, Maekawa S, Bodi I, et al. p62
positive, TDP-43 negative, neuronal cytoplasmic and intranuclear inclusions in the cerebellum and hippocampus define the pathology of
C9orf72-linked FTLD and MND/ALS. Acta Neuropathol 2011; 122:
691–702.
Asante EA, Linehan JM, Desbruslais M, Joiner S, Gowland I, Wood AL,
et al. BSE prions propagate as either variant CJD-like or sporadic
CJD-like prion strains in transgenic mice expressing human prion protein. EMBO J 2002; 21: 6358–66.
Asante EA, Linehan JM, Gowland I, Joiner S, Fox K, Cooper S, et al.
Dissociation of pathological and molecular phenotype of variant
Creutzfeldt-Jakob disease in transgenic human prion protein 129 heterozygous mice. Proc Natl Acad Sci USA 2006; 103: 10759–64.
Baker M, Mackenzie IR, Pickering-Brown SM, Gass J, Rademakers R,
Lindholm C, et al. Mutations in progranulin cause tau-negative frontotemporal dementia linked to chromosome 17. Nature 2006; 442:
916–9.
Belly A, Bodon G, Blot B, Bouron A, Sadoul R, Goldberg Y. CHMP2B
mutants linked to frontotemporal dementia impair maturation of dendritic spines. J Cell Sci 2010; 123: 2943–54.
Benajiba L, Le BI, Camuzat A, Lacoste M, Thomas-Anterion C,
Couratier P, et al. TARDBP mutations in motoneuron disease with
frontotemporal lobar degeneration. Ann Neurol 2009; 65: 470–3.
Borroni B, Bonvicini C, Alberici A, Buratti E, Agosti C, Archetti S, et al.
Mutation within TARDBP leads to frontotemporal dementia without
motor neuron disease. Hum Mutat 2009; 30: E974–83.
Cairns NJ, Bigio EH, Mackenzie IR, Neumann M, Lee VM, Hatanpaa KJ,
et al. Neuropathologic diagnostic and nosologic criteria for frontotemporal lobar degeneration: consensus of the Consortium for
Frontotemporal Lobar Degeneration. Acta Neuropathol 2007; 114:
5–22.
Coleman M. Axon degeneration mechanisms: commonality amid diversity. Nat Rev Neurosci 2005; 6: 889–98.
Crespo-Barreto J, Fryer JD, Shaw CA, Orr HT, Zoghbi HY. Partial loss of
ataxin-1 function contributes to transcriptional dysregulation in spinocerebellar ataxia type 1 pathogenesis. PLoS Genet 2010; 6: e1001021.
Cruts M, Gijselinck I, van der ZJ, Engelborghs S, Wils H, Pirici D, et al.
Null mutations in progranulin cause ubiquitin-positive frontotemporal
dementia linked to chromosome 17q21. Nature 2006; 442: 920–4.
Deinhardt K, Salinas S, Verastegui C, Watson R, Worth D, Hanrahan S,
et al. Rab5 and Rab7 control endocytic sorting along the axonal retrograde transport pathway. Neuron 2006; 52: 293–305.
Dejesus-Hernandez M, Mackenzie IR, Boeve BF, Boxer AL, Baker M,
Rutherford NJ, et al. Expanded GGGGCC hexanucleotide repeat in
noncoding region of C9ORF72 causes chromosome 9p-linked FTD
and ALS. Neuron 2011; 72: 245–56.
Delisle MB, Carpenter S. Neurofibrillary axonal swellings and amyotrophic lateral sclerosis. J Neurol Sci 1984; 63: 241–50.
Filimonenko M, Stuffers S, Raiborg C, Yamamoto A, Malerod L,
Fisher EM, et al. Functional multivesicular bodies are required for
autophagic clearance of protein aggregates associated with neurodegenerative disease. J Cell Biol 2007; 179: 485–500.
Brain 2012: 135; 819–832
| 831
Ghoshal N, Dearborn JT, Wozniak DF, Cairns NJ. Core features of frontotemporal dementia recapitulated in progranulin knockout mice.
Neurobiol Dis 2012; 45: 395–408.
Harvey RJ, Skelton-Robinson M, Rossor MN. The prevalence and causes
of dementia in people under the age of 65 years. J Neurol Neurosurg
Psychiatry 2003; 74: 1206–9.
Holm IE, Englund E, Mackenzie IR, Johannsen P, Isaacs AM. A
Reassessment of the Neuropathology of Frontotemporal Dementia
Linked to Chromosome 3. J Neuropathol Exp Neurol 2007; 66:
884–91.
Holm IE, Isaacs AM, Mackenzie IR. Absence of FUS-immunoreactive
pathology in frontotemporal dementia linked to chromosome 3
(FTD-3) caused by mutation in the CHMP2B gene. Acta
Neuropathol 2009; 118: 719–20.
Hu F, Padukkavidana T, Vaegter CB, Brady OA, Zheng Y, Mackenzie IR,
et al. Sortilin-mediated endocytosis determines levels of the frontotemporal dementia protein, progranulin. Neuron 2010; 68: 654–67.
Hutton M, Lendon CL, Rizzu P, Baker M, Froelich S, Houlden H, et al.
Association of missense and 50 -splice-site mutations in tau with the
inherited dementia FTDP-17. Nature 1998; 393: 702–5.
Kantamneni S, Holman D, Wilkinson KA, Nishimune A, Henley JM. GISP
increases neurotransmitter receptor stability by down-regulating
ESCRT-mediated lysosomal degradation. Neurosci Lett 2009; 452:
106–10.
Katzmann DJ, Odorizzi G, Emr SD. Receptor downregulation and
multivesicular-body sorting. Nat Rev Mol Cell Biol 2002; 3: 893–905.
Kovacs GG, Murrell JR, Horvath S, Haraszti L, Majtenyi K, Molnar MJ,
et al. TARDBP variation associated with frontotemporal dementia, supranuclear gaze palsy, and chorea. Mov Disord 2009; 24:
1843–7.
Laird FM, Farah MH, Ackerley S, Hoke A, Maragakis N, Rothstein JD,
et al. Motor neuron disease occurring in a mutant dynactin mouse
model is characterized by defects in vesicular trafficking. J Neurosci
2008; 28: 1997–2005.
Lee JA, Beigneux A, Ahmad ST, Young SG, Gao FB. ESCRT-III dysfunction causes autophagosome accumulation and neurodegeneration.
Curr Biol 2007; 17: 1561–7.
Lee JA, Liu L, Gao FB. Autophagy defects contribute to neurodegeneration induced by dysfunctional ESCRT-III. Autophagy 2009; 5: 1070–2.
Lim J, Crespo-Barreto J, Jafar-Nejad P, Bowman AB, Richman R, Hill DE,
et al. Opposing effects of polyglutamine expansion on native protein
complexes contribute to SCA1. Nature 2008; 452: 713–8.
Lin WL, Lewis J, Yen SH, Hutton M, Dickson DW. Ultrastructural neuronal pathology in transgenic mice expressing mutant (P301L) human
tau. J Neurocytol 2003; 32: 1091–105.
Mackenzie IR, Neumann M, Bigio EH, Cairns NJ, Alafuzoff I, Kril J, et al.
Nomenclature and nosology for neuropathologic subtypes of frontotemporal lobar degeneration: an update. Acta Neuropathol 2010; 119:
1–4.
Matilla A, Roberson ED, Banfi S, Morales J, Armstrong DL, Burright EN,
et al. Mice lacking ataxin-1 display learning deficits and decreased
hippocampal paired-pulse facilitation. J Neurosci 1998; 18: 5508–16.
McKhann GM, Albert MS, Grossman M, Miller B, Dickson D,
Trojanowski JQ. Clinical and pathological diagnosis of frontotemporal
dementia: report of the Work Group on Frontotemporal Dementia and
Pick’s Disease. Arch Neurol 2001; 58: 1803–9.
Metcalf D, Isaacs AM. The role of ESCRT proteins in fusion events
involving lysosomes, endosomes and autophagosomes. Biochem Soc
Trans 2010; 38: 1469–73.
Murray ME, Dejesus-Hernandez M, Rutherford NJ, Baker M, Duara R,
Graff-Radford NR, et al. Clinical and neuropathologic heterogeneity of
c9FTD/ALS associated with hexanucleotide repeat expansion in
C9ORF72. Acta Neuropathol 2011; 122: 673–90.
Neary D, Snowden JS, Gustafson L, Passant U, Stuss D, Black S, et al.
Frontotemporal lobar degeneration: a consensus on clinical diagnostic
criteria. Neurology 1998; 51: 1546–54.
Nixon RA, Wegiel J, Kumar A, Yu WH, Peterhoff C, Cataldo A, et al.
Extensive involvement of autophagy in Alzheimer disease: an
832
| Brain 2012: 135; 819–832
immuno-electron microscopy study. J Neuropathol Exp Neurol 2005;
64: 113–22.
Perlson E, Maday S, Fu MM, Moughamian AJ, Holzbaur EL. Retrograde
axonal transport: pathways to cell death? Trends Neurosci 2010; 33:
335–44.
Petkau TL, Neal SJ, Milnerwood A, Mew A, Hill AM, Orban P, et al.
Synaptic dysfunction in progranulin-deficient mice. Neurobiol Dis
2011.
Pikkarainen M, Hartikainen P, Alafuzoff I. Neuropathologic features of
frontotemporal lobar degeneration with ubiquitin-positive inclusions
visualized with ubiquitin-binding protein p62 immunohistochemistry.
J Neuropathol Exp Neurol 2008; 67: 280–98.
Poorkaj P, Bird TD, Wijsman E, Nemens E, Garruto RM, Anderson L,
et al. Tau is a candidate gene for chromosome 17 frontotemporal
dementia. Ann Neurol 1998; 43: 815–25.
Probst A, Gotz J, Wiederhold KH, Tolnay M, Mistl C, Jaton AL, et al.
Axonopathy and amyotrophy in mice transgenic for human
four-repeat tau protein. Acta Neuropathol 2000; 99: 469–81.
Ratnavalli E, Brayne C, Dawson K, Hodges JR. The prevalence of frontotemporal dementia. Neurology 2002; 58: 1615–21.
Renton AE, Majounie E, Waite A, Simon-Sanchez J, Rollinson S,
Gibbs JR, et al. A hexanucleotide repeat expansion in C9ORF72 is
the cause of chromosome 9p21-linked ALS-FTD. Neuron 2011; 72:
257–68.
Rusten TE, Stenmark H. How do ESCRT proteins control autophagy?
J Cell Sci 2009; 122: 2179–83.
Scott M, Foster D, Mirenda C, Serban D, Coufal F, Walchli M, et al.
Transgenic mice expressing hamster prion protein produce
species-specific scrapie infectivity and amyloid plaques. Cell 1989;
59: 847–57.
Skibinski G, Parkinson NJ, Brown JM, Chakrabarti L, Lloyd SL,
Hummerich H, et al. Mutations in the endosomal ESCRTIII-complex
subunit CHMP2B in frontotemporal dementia. Nat Genet 2005; 37:
806–8.
Spillantini MG, Murrell JR, Goedert M, Farlow MR, Klug A,
Ghetti B. Mutation in the tau gene in familial multiple system
tauopathy with presenile dementia. Proc Natl Acad Sci USA
1998; 95: 7737–41.
Spittaels K, Van den HC, Van Dorpe J, Bruynseels K, Vandezande K,
Laenen I, et al. Prominent axonopathy in the brain and spinal cord
of transgenic mice overexpressing four-repeat human tau protein. Am
J Pathol 1999; 155: 2153–65.
S. Ghazi-Noori et al.
Tu PH, Raju P, Robinson KA, Gurney ME, Trojanowski JQ, Lee VM.
Transgenic mice carrying a human mutant superoxide dismutase transgene develop neuronal cytoskeletal pathology resembling human
amyotrophic lateral sclerosis lesions. Proc Natl Acad Sci USA 1996;
93: 3155–60.
Urwin H, Authier A, Nielsen JE, Metcalf D, Powell C, Froud K, et al.
Disruption of endocytic trafficking in frontotemporal dementia with
CHMP2B mutations. Hum Mol Genet 2010a; 19: 2228–38.
Urwin H, Josephs KA, Rohrer JD, Mackenzie IR, Neumann M, Authier A,
et al. FUS pathology defines the majority of tau- and TDP-43-negative
frontotemporal lobar degeneration. Acta Neuropathol 2010b; 120:
33–41.
Urwin H, Ghazi-Noori S, Collinge J, Isaacs A. The role of CHMP2B in
frontotemporal dementia. Biochem Soc Trans 2009; 37: 208–12.
van der Zee J, Urwin H, Engelborghs S, Bruyland M, Vandenberghe R,
Dermaut B, et al. CHMP2B C-truncating mutations in frontotemporal
lobar degeneration are associated with an aberrant endosomal phenotype in vitro. Hum Mol Genet 2008; 17: 313–22.
Van Langenhove T, van der ZJ, Sleegers K, Engelborghs S,
Vandenberghe R, Gijselinck I, et al. Genetic contribution of FUS to
frontotemporal lobar degeneration. Neurology 2010; 74: 366–71.
Wang QJ, Ding Y, Kohtz DS, Mizushima N, Cristea IM, Rout MP, et al.
Induction of autophagy in axonal dystrophy and degeneration.
J Neurosci 2006; 26: 8057–68.
Watase K, Weeber EJ, Xu B, Antalffy B, Yuva-Paylor L, Hashimoto K,
et al. A long CAG repeat in the mouse Sca1 locus replicates SCA1
features and reveals the impact of protein solubility on selective neurodegeneration. Neuron 2002; 34: 905–19.
Watts GD, Wymer J, Kovach MJ, Mehta SG, Mumm S, Darvish D, et al.
Inclusion body myopathy associated with Paget disease of bone and
frontotemporal dementia is caused by mutant valosin-containing protein. Nat Genet 2004; 36: 377–81.
Wu C, Cui B, He L, Chen L, Mobley WC. The coming of age of axonal
neurotrophin signaling endosomes. J Proteomics 2009; 72: 46–55.
Yin F, Banerjee R, Thomas B, Zhou P, Qian L, Jia T, et al. Exaggerated
inflammation, impaired host defense, and neuropathology in
progranulin-deficient mice. J Exp Med 2010a; 207: 117–28.
Yin F, Dumont M, Banerjee R, Ma Y, Li H, Lin MT, et al. Behavioral
deficits and progressive neuropathology in progranulin-deficient mice:
a mouse model of frontotemporal dementia. FASEB J 2010b; 24:
4639–47.