Als2-deficient mice exhibit disturbances
in endosome trafficking associated with
motor behavioral abnormalities
R. S. Devon*†‡, P. C. Orban*†, K. Gerrow§, M. A. Barbieri¶, C. Schwab*, L. P. Cao*, J. R. Helm*, N. Bissada*,
R. Cruz-Aguado*, T.-L. Davidson*, J. Witmer*, M. Metzler*, C. K. Lam?, W. Tetzlaff?, E. M. Simpson*,
J. M. McCaffery**, A. E. El-Husseini§, B. R. Leavitt*, and M. R. Hayden*††
*Centre for Molecular Medicine and Therapeutics, Department of Medical Genetics, University of British Columbia, and Child & Family Research Institute,
980 West 28th Avenue, Vancouver, BC, Canada V5Z 4H4;§Department of Psychiatry, Brain Research Centre, University of British Columbia, Vancouver, BC,
Canada V6T 2A1;¶Department of Biological Sciences, Florida International University, Miami, FL 33199; **Integrated Imaging Center, Department of
Biology, The Johns Hopkins University, Baltimore, MD 21218; and?International Collaboration on Repair Discoveries and Department of Zoology,
University of British Columbia, Vancouver, BC, Canada V6T 1Z4
Edited by Huda Y. Zoghbi, Baylor College of Medicine, Houston, TX, and approved April 27, 2006 (received for review November 24, 2005)
ALS2 is an autosomal recessive form of spastic paraparesis (motor
neuron disease) with juvenile onset and slow progression caused
by loss of function of alsin, an activator of Rac1 and Rab5 small
GTPases. To establish an animal model of ALS2 and derive insights
into the pathogenesis of this illness, we have generated alsin-null
mice. Cytosol from brains of Als2?/?mice shows marked diminu-
tion of Rab5-dependent endosome fusion activity. Furthermore,
primary neurons from Als2?/?mice show a disturbance in endo-
somal transport of insulin-like growth factor 1 (IGF1) and BDNF
receptors, whereas neuronal viability and endocytosis of trans-
ferrin and dextran seem unaltered. There is a significant decrease
in the size of cortical motor neurons, and Als2?/?mice are mildly
hypoactive. Altered trophic receptor trafficking in neurons of
Als2?/?mice may underlie the histopathological and behavioral
changes observed and the pathogenesis of ALS2.
ALS ? alsin ? knockout mouse ? motor neuron ? Rab5
the related conditions infantile onset ascending hereditary spas-
tic paraplegia (IAHSP) and juvenile onset primary lateral scle-
rosis (JPLS) (1–5). IAHSP and JPLS are characterized by
degeneration of upper motor neurons of the corticospinal and
spinocerebellar tracts, whereas a diagnosis of ALS also requires
lower motor neuron involvement. The expression pattern of Als2
is consistent with this pattern of neurodegeneration, because the
gene is expressed primarily in neurons of the CNS, including
cortical motor neurons and the large alpha motor neurons of the
spinal cord. Interestingly, however, it is also found at particularly
high levels in the granule cell neurons of the cerebellum, a region
not previously implicated in ALS (6).
Alsin, the protein encoded by ALS2, has been characterized as
a novel guanine nucleotide exchange factor (GEF) for Rab5 and
Rac1 (7–9). Rab5 is known to be an important factor in many
stages of endocytosis and the early trafficking of signaling
molecules (10–13). In neurons, Rab5-mediated endocytic traf-
ficking is essential for presynaptic transmission at the Drosophila
neuromuscular junction (14), and Rab5 is required for the
formation of endocytic pits for internalization of ?-amino-3-
hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors
after the induction of long-term depression (LTD) (15).
To reconcile the divergent activities of Rab5, it has been
proposed that Rab5 is compartmentalized and activated for
specific functions through interaction with particular GEFs and
effector proteins (16). Its interaction with alsin has been sug-
gested to modulate the signaling of neurotrophic factors (7).
All described mutations in the ALS2 gene are predicted to
cause premature protein termination and loss of function of the
utations in the ALS2 gene cause autosomal recessive
juvenile onset amyotrophic lateral sclerosis (ALS2) and
alsin protein (5). Therefore, we sought to investigate the effect
of alsin deficiency in mice, to further understand human ALS2,
and with regard to the role of alsin in the trafficking of specific
Als2?/?Mice Are Viable and Fertile. Als2-deficient mice were
generated by using promoter trap gene targeting in ES cells to
replace exons 3 and 4 of the Als2 gene with an SA-IRES-
?geo-pA [splice acceptor-internal ribosome entry site-?-
galactosidase-neomycin–poly(A) addition site] cassette encod-
ing a bifunctional lacZ–neomycin fusion protein (17). The
resulting protein encoded from the Als2 locus would comprise
only the five N-terminal amino acids of alsin fused to lacZ-
neomycin, thereby inactivating both the full-length and the
proposed short form of alsin (1, 2). Five homologous recombi-
nant ES cell clones were confirmed to be correctly targeted, of
which two were used to generate mouse lines for analysis (Fig.
1 A and B). Absence of alsin in Als2?/?mice was confirmed by
Western blotting (Fig. 1C). The genotypes of progeny of het-
erozygous intercrosses showed a normal Mendelian distribution
(genotypes from line 1: Als2?/?, n ? 124 (22.9%); Als2?/?, n ?
278 (51.3%); Als2?/?, n ? 140 (24.8%)). Als2?/?mice were
indistinguishable from their WT littermates in size, appearance,
and observed behavior in the home cage. Comparison of the
number of progeny from homozygous Als2?/?mating pairs
versus WT mating pairs over an 8-month period indicated that
alsin deficiency did not affect fertility or fecundity (data not
The Nervous System of Als2?/?Mice Exhibits Significant but Subtle
Neuropathological Changes. Because ALS in humans is character-
ized by neurodegeneration in the brain and spinal cord, we
assessed Als2?/?mice histologically for evidence of this. We
performed TUNEL (Roche, Indianapolis), Fluorojade-B
(Chemicon, Temecula, CA) staining, and immunostaining for
cleaved caspase-3 of brain sections of WT and Als2?/?mice at
Conflict of interest statement: No conflicts declared.
This paper was submitted directly (Track II) to the PNAS office.
Abbreviations: ALS2, autosomal recessive juvenile onset ALS; GEF, guanine nucleotide
exchange factor; CGN, cerebellar granule cell neuron; IGF1, insulin-like growth factor 1;
IGF1R, IGF1 receptor; EEA1, early endosome antigen 1.
†R.S.D. and P.C.O. contributed equally to this work.
‡Present Address: Medical Genetics Section, University of Edinburgh Molecular Medicine
Centre, Western General Hospital, Crewe Road, Edinburgh EH4 2XU, United Kingdom.
††To whom correspondence should be addressed. E-mail: email@example.com.
© 2006 by The National Academy of Sciences of the USA
June 20, 2006 ?
vol. 103 ?
no. 25 ?
3, 6, and 12 months. We did not find any evidence of increased
apoptosis or necrosis.
To assess other possible neuropathological changes in these
mice, we performed unbiased stereological assessments of brain,
spinal cord, and sciatic nerve at 12 and 22 months of age. At 12
months, cell bodies of pyramidal motor neurons in layer V of the
motor cortex of Als2?/?mice were significantly smaller than
those of their WT counterparts (7% reduction in mean cell area;
P ? 0.006; Table 1, which is published as supporting information
on the PNAS web site). All other measures of neuronal size
axons of diminished caliber compared with WT, but no other
measurement reached statistical significance (Table 1). The
trend toward reduced size in Als2?/?mice was also apparent for
cerebellar weight and total brain weight (Table 1). No further
decrease in cortical neuron size was found in mice aged 22
Conventional thin section electron microscopy analysis was
the brain, coronal sections containing regions of the cerebrum,
hippocampus, cerebellum, and brainstem were analyzed,
whereas transverse sections from both the thoracic and lumbar
spinal cord were examined. In each case, no discernible differ-
ence in ultrastructure was observed between WT and Als2?/?
mice (data not shown).
To assess whether neuropathology typically found in human
ALS and mutant superoxide dismutase 1 (SOD1) transgenic
mice (18) was present in aged Als2?/?mice, we stained motor
cortex, spinal cord, and cerebellum sections of 22-month-old
animals with antibodies against ubiquitin, phosphorylated neu-
rofilament (SMI31) and glial markers [glial fibrillary acidic
protein (GFAP) for astroglia and Iba1 for microglia]. In addi-
tion, we stained with NeuN, calbindin antibodies, and cresyl
and no inclusions or abnormal neuronal processes (axonal
spheroids, swellings) were found in sections stained with ubiq-
uitin or SMI31. No astroglial or microglial activation could be
demonstrated in cortex, hippocampus, or cerebellum, nor did we
see evidence of glial activation in motor neuron areas or fiber
tracts of the spinal cord.
Fusion of Early Endosomes. It is well established that alsin interacts
with and activates Rab5 (7–9). To assess early endosome fusion
activity [known to be Rab5-dependent (12)], we used an in vitro
assay of fusion activity, in which cytosolic test extracts are added
to suitably labeled early endosomes, and the degree of resultant
fusion assessed by colorimetry (19). Addition of cytosol from
Als2?/?mouse brain resulted in up to 72% less fusion activity
than cytosol from WT mouse brain (Fig. 2). WT brain cytosol
supported endosome fusion in a concentration-dependent man-
ner (Fig. 2A).
Exogenous alsin rescued the deficit in Als2?/?mouse brain
cytosol. Lysate from SF9 cells expressing alsin using a baculo-
virus vector (7), but not control or heat-inactivated SF9 lysate,
probe is shown as a hatched box. (B) Southern blot of MscI-digested WT ES cell DNA (WT) and two targeted ES cell clones (lines 1 and 2). WT (16 kb) and targeted
(10 kb) alleles are present as expected. (C) Western blot analysis of cortex, cerebellum, and testis from WT (???), Als2-heterozygous (???), and Als2-null (???)
mice showing absence of alsin in homozygous targeted mice and reduced expression in heterozygotes.
Generation of Als2?/?mice. (A) The 5? end of the WT Als2 locus and the targeted allele, in which exons 3 and 4 were replaced. The Southern blotting
activity. (A) Fusion activity of indicated concentrations of WT (open squares)
or Als2?/?(filled squares) brain cytosol (means ? SD of triplicates). Fusion
activity was significantly different (*, P ? 0.001) between Als2?/?and WT. (B)
brains. Brain cytosol (0.75 mg?ml) was supplemented with lysates of SF9 cells
that had been transfected with vector alone (?), vector-expressing alsin (?),
or heat-inactivated SF9 alsin?lysate (?§) (means ? SD from three experi-
activity for WT (open bars) and Als2?/?(filled bars) mice (*, P ? 0.001); this
effect was abolished if heat-inactivated alsin?lysate was used (*, P ? 0.001).
Brain cytosol of Als2?/?mice is deficient in early endosome fusion
www.pnas.org?cgi?doi?10.1073?pnas.0510197103Devon et al.
restored endosome fusion activity (P ? 0.001) (Fig. 2B). Inter-
estingly, the addition of alsin also enhanced the endosome fusion
activity of WT brain cytosol (Fig. 2B).
Neurons from Als2?/?Mice Exhibit Disturbances of Receptor Traffick-
ing. We postulated that an impediment in endosomal trafficking
of neurotrophin receptors, and resultant diminution in neuro-
trophic support, might be responsible for the reduction in
neuronal size in Als2?/?mice. We analyzed cortical neurons and
cerebellar granule cell neurons (CGNs) in culture, because both
these cell types express alsin at high levels (6) and cortical
neurons in particular are affected in ALS.
Neuronal morphology and survival in WT and Als2?/?neuron
cultures were examined by DiI staining, fluorescence, and
differential interference contrast (DIC) microscopy. No differ-
ences in morphology, size, or survival were noted between WT
both genotypes (data not shown). There was also no difference
between WT and Als2?/?CGNs in response to induction of
apoptosis using low potassium culture conditions (5 mM as
opposed to 25 mM), as assessed by a substrate-based assay for
caspase 3 activity, immunoblotting of active caspase 3, or
TUNEL staining (data not shown).
We initially examined whether a reduction in Rab5-GEF
activity caused by the absence of alsin may result in abnormal
trafficking of neurotrophic factor receptors. After 3 h of trophic
factor deprivation, cortical or CGNs at 8–12 days in culture were
treated by bath application of 50 ng?ml BDNF for between 5 and
180 min, and the positions of the BDNF receptor, TrkB, were
assessed by indirect immunofluorescence. In both cell types,
there was a marked difference between WT and Als2?/?neurons
in the accumulation of fluorescence in the cell bodies (Fig. 3A,
from cortical neurons). WT but not Als2?/?neurons showed a
significant increase in perinuclear anti-TrkB staining in the first
60 min of BDNF stimulation (Fig. 3B).
There was no general disturbance of endocytosis, because
uptake of Alexa Fluor 549-conjugated transferrin (Fig. 3 C and
D) or FITC-conjugated dextran (not shown) showed no differ-
ence between WT and Als2?/?cultures.
We next repeated the BDNF stimulation experiment using
insulin-like growth factor 1 (IGF1) and antibody to its receptor,
IGF1R. After 30 min or more of IGF1 stimulation, puncta of at
least 3.5 ?m2area (calculated after flattening the confocal
images) were apparent in the dendrites of some (between 1?200
and 1?2,000) Als2?/?CGNs (Fig. 4A). All positive cells con-
tained at least 20 large puncta throughout several dendrites.
Als2?/?cells were, on average, 1,000 times more likely to display
this phenotype than WT cells (four separate cultures examined
at the 60-min stimulation time point). To define the nature of the
IGF1R puncta, the cells were costained with markers of endo-
cytic compartments. As shown in Fig. 4 C and D, 32% of the
puncta colocalized with early endosome antigen 1 (EEA1), and
the majority (86%) colocalized with Rab5, both markers of early
endosomes, and 17% colocalized with Rab11, a marker of
recycling endosomes (Fig. 4E). Similar results were obtained
with cortical neurons, but a smaller proportion of the Als2?/?
cells showed the phenotype.
There was no significant change in the total amount of IGF1
or BDNF receptor present during the course of the stimulation,
nor was there any apparent difference between the amount of
receptor present in WT and Als2?/?neurons in culture (Fig. 6,
which is published as supporting information on the PNAS web
site). Control experiments confirming that these results were not
due to receptor synthesis during the experiment and that the
changes were stimulus-specific are described in Supporting Ma-
terials and Methods (Receptor Trafficking Controls), which is
published as supporting information on the PNAS web site. In
a sciatic nerve crush (ligation) experiment to assess motor
neuron transport in vivo, IGF1R-positive puncta were visible in
motor axons of Als2?/?but not WT mice (Fig. 7, which is
published as supporting information on the PNAS web site),
consistent with endosome trafficking defects. However, we
found no genotype-linked difference in rapid axonal transport
(as evidenced by 8 h accumulation at a ligature) of IGF1R, TrkB,
or choline acetyl transferase (ChAT) (three WT and three
To identify intracellular signaling changes that might accom-
pany the apparent trafficking disturbances, we stimulated
starved neurons with IGF1 or BDNF and assessed phosphory-
lation of Akt and extracellular signal-regulated kinases (ERKs).
Although primary fibroblasts from Als2?/?mice did show a
significant decrease in Akt phosphorylation after stimulation
with IGF1 (n ? 2, P ? 0.033), we did not find any significant
differences in these indices of signal transduction between WT
and Als2?/?neurons (Fig. 8, which is published as supporting
information on the PNAS web site).
Western blot analyses showed no apparent differences in the
abundance of 13 intracellular trafficking pathway components,
including other Rab GTPases and synaptic vesicle proteins, in
cerebral cortical extracts and cerebellar extracts of 18-month-old
WT and Als2?/?mice (Fig. 9, which is published as supporting
information on the PNAS web site).
Als2?/?Mice Are Hypoactive.Acohortof40mice(20eachWTand
Als2?/?) was followed until 15 months of age for assessment of
activity and motor skills. During this time, four WT and one
Als2?/?mouse died or had to be euthanized. These observations
fluorescence in the cell-body region after stimulation with BDNF than WT
tified in cortical neurons randomly selected by using differential interference
contrast (DIC) microscopy (**, P ? 0.01; 14–20 cells for each time point and
genotype). (C) There is no difference in transferrin uptake between WT and
Als2?/?neurons: after 30 min, both show intense transferrin fluorescence in
their cell bodies. (D) Cell-body region fluorescence in a representative trans-
ferrin uptake time course, quantified in neurons randomly selected by using
DIC microscopy. (Scale bars: 10 ?m.)
Neurons from Als2?/?mice show disturbances of BDNF receptor
Devon et al.PNAS ?
June 20, 2006 ?
vol. 103 ?
no. 25 ?
indicate that absence of alsin does not adversely affect lifespan,
at least until 15 months of age. Measurements of body weight
revealed no difference between male or female Als2?/?and WT
mice, although there was a trend toward increased body weight,
particularly for male Als2?/?mice (two-way ANOVA, P ? 0.167
for males and P ? 0.426 for females) (Fig. 5 A and B).
In a 3-min open-field trial to measure exploratory activity,
Als2?/?homozygotes were significantly hypoactive compared
with WT mice. Two-way repeated-measures ANOVA revealed
a significant effect of genotype across the time course of the
time ambulating (P ? 0.027, Fig. 5D), number of ambulatory
episodes (P ? 0.027, Fig. 5E), and time resting (P ? 0.034, Fig.
5F). This phenotype was evident in even the youngest mice (3
months of age) and showed little if any worsening over time.
Apart from hypoactivity, Als2?/?mice showed no other
deficits in motor skills and coordination. Grip strength and
latency to fall from an accelerating rotarod were both indistin-
guishable from WT. Als2?/?mice did not show evidence of
deficit in the leg extension reflex (hind-limb ‘‘clasping’’), when
suspended by the tail. Additionally, there was no difference
between WT and Als2?/?mice in the proportion of time spent
in the perimeter versus central areas of the arena.
The absence of alsin in mice results in a pronounced cellular
phenotype, and subtle neuropathology and behavioral changes
that are consistent with motor neuron disease. Brain cytosol
from Als2?/?mice showed a marked decrease in ability to
support Rab5-dependent endosome fusion in vitro. Further-
in the endosomal transport of receptors for the neurotrophic
factors BDNF and IGF1. Als2?/?mice exhibit a decrease in
locomotor activity and a small but significant decrease in the size
of cortical motor neurons.
8), we focused on examining cellular phenotypes related to Rab5
activity, namely, early endosome trafficking and fusion. We
observed dramatic reduction in Rab5-dependent endosome
fusion activity in Als2-null brain cytosol. This finding suggests
that alsin is an important component of early endosome fusion
activity in adult mouse brain.
An abnormal accumulation of large, IGF1R-positive puncta
was seen in Als2?/?neurons after IGF1 stimulation and in sciatic
nerve motor axons. The puncta in cultured neurons primarily
colocalized with markers of early endosomes (Rab5 and EEA1),
suggesting that they represent aberrant early endosome compo-
nents of the IGF1R endocytosis pathway, reflecting a failure of
normal maturation of these endosomes. This finding may imply
a role for alsin in conversion from early to late endosomes, a
process that has been shown to require the activity of a Rab7
GEF, the class C VPS?HOPS complex (20). Interestingly, IGF1
has long been flagged as a potential therapeutic agent for ALS
because of its ability to promote motor neuron survival. Re-
cently, it has been shown to confer a modest benefit in human
adult onset ALS (21) and to delay disease onset and ameliorate
(A) CGNs were stimulated with IGF1 for 0, 30, 60, and 360 min before staining
with antibodies to the IGF1R alpha chain. Note the large puncta seen at later
time points in Als2?/?neurons. (B) A cell with puncta is shown at lower
and antibodies to EEA1 (C), Rab5 (D), or Rab11 (E). A large majority of
IGF1R-positive puncta costain with anti-Rab5, a smaller proportion with anti-
EEA1, and few with anti-Rab11. (Scale bars: 10 ?m.)
Neurons from Als2?/?mice show disturbances of IGF1R endocytosis.
hypoactive. Body weight for male (A) and female (B) Als2?/?mice (filled
squares) shows a trend but no significantly increased weight gain compared
with WT littermates (open squares). (C–F) Exploratory open-field activity for
Als2?/?mice (filled squares) versus WT mice (open squares) shows that Als2-
null mice are hypoactive throughout their life (two-way ANOVA, P ? 0.05 for
each measure). Values shown are means ? SEM of n ? 19–20 mice (n ? 9–10
each males and females).
Als2?/?mice show a tendency to increased body weight and are
www.pnas.org?cgi?doi?10.1073?pnas.0510197103Devon et al.
progression in mutant superoxide dismutase 1 (SOD1) trans-
genic mice (21, 22).
Neurons from Als2-null mice also exhibited a defect in traf-
ficking of TrkB when stimulated with its ligand, BDNF, whereas
other forms of receptor-mediated and fluid-phase endosomal
trafficking were unaffected. It is likely that aberrant growth
factor receptor endocytosis in Als2-null mice results in a defi-
ciency in trophic factor support of neurons lacking alsin. This
trophic support is thought to be dependent on the transport of
endosomes bearing activated receptors from the neurite to
nucleus, resulting in transcriptional changes (23, 24). In exper-
iments in which we assessed phosphorylation of Akt and extra-
cellular signal-regulated kinases (ERKs), prominent compo-
nents of the intracellular signaling pathways of IGF1 and BDNF,
we found significant differences between WT and Als2-null
fibroblasts but not neurons. It is possible that these assays are too
crude to reveal subtle signaling defects in neurons or that
alsin-null neurons have undergone compensatory changes in the
regulation of signal transduction.
Als2-null mice exhibited subtle pathological changes in the
CNS. At 12 months of age, cortical motor neurons of Als2?/?
mice were slightly smaller than their WT counterparts, and there
was a trend toward reduction in neuronal size and axonal caliber
in all cell types observed. That this trend extends to lower motor
neurons is consistent with the observation that ALS2 mutation
shrinkage is strongly associated with motor neuron disease,
particularly primary lateral sclerosis (PLS) (25). Additionally, in
ALS, the largest alpha motor neurons are the most vulnerable to
degeneration, although the role of axonal caliber in the etiology
of motor neuron disease is as yet unclear (26). Because there
have as yet been no published descriptions of pathological
investigations in ALS2 patients, the precise expected findings in
mice are not obvious.
In the context of ALS pathogenesis, it is interesting to note
that it is only the cortical motor neurons of Als2?/?mice that
exhibit both functional and histopathological abnormalities,
even though CGNs express the highest levels of alsin in the
mouse brain (6). The abnormal IGF1R puncta were seen less
frequently in cortical neuron cultures than in CGNs, but this
result is likely to reflect the fact that sensory neurons within
cortical cultures do not express alsin. The selectivity toward
motor neuron pathology may be due to the larger size of motor
neurons, or peculiarities of calcium homeostasis, which may
make them particularly vulnerable to decreases in trophic sup-
port (reviewed in ref. 27).
In humans, ALS2-related disease causes severe spastic para-
paresis with onset in childhood and slow progression, and its
autosomal recessive pattern of inheritance suggests complete
loss of alsin function. The early onset of hypoactivity in Als2-null
mice and the fact that it did not worsen over time are consistent
with this finding. Nevertheless, a more obvious phenotype in
Als2?/?mice might have been predicted. This incongruity is
unlikely to be due to simple redundancy of protein function,
because there is no other protein in the mouse (or human) that
shares the same complement of domains as alsin. The closest
related protein, ALS2 C-terminal-like (ALS2CL) lacks alsin’s
N-terminal regulator of chromatin condensation (RCC1) do-
main, and is poorly expressed in the brain (6, 28). We have not
found significant differences in the expression of ALS2CL
between WT and Als2?/?motor cortex or cerebellum by quan-
titative RT-PCR (data not shown). We have also found no
difference between WT and Als2?/?motor cortex or cerebellum
in the abundance of 13 different neuronal trafficking-associated
proteins. The absence of a dramatic phenotype may instead
activity, rather than mere abundance, may be altered. Alterna-
tively, it may be related to the physical length of neurons in
mouse versus human, whereby human motor neurons would
have a much lower tolerance of defects in neurotrophic factor
signaling and transport.
Our findings are in partial agreement with two recent descrip-
tions of Als2-null mice (29, 30). All studies agree that Als2?/?
mice exhibit no gross phenotype other than increased (or a trend
toward increased) body weight. All three studies also demon-
strate some impairment in motor function: early-onset hypoac-
tivity (this study), or a significant, or trend toward, late-onset
rotarod deficit (refs. 29 and 30, respectively). Histologically, Cai
et al. (29) demonstrated no abnormalities, whereas Hadano et al.
(30) showed astrocytosis and activation of microglia and a
late-onset loss of Purkinje cells and lower motor neuron axons.
The reason for these discrepancies is unclear; however, they are
possibly due to differences in the ES cells used, leading to
differences in the background strain of experimental mice, or to
differences in design of the targeting vectors. In culture, Hadano
et al. (30) showed subtle defects in early trafficking of EGF in
Als2-deficient fibroblasts, whereas Cai et al. (29) demonstrated
increased susceptibility to oxidative stress in cortical neurons.
Our description of defects in early endosomal trafficking and
fusion in Als2-null mice may provide a mechanism for the
increased susceptibility to oxidative stress seen (29). The Rab5-
associated endocytic activity of cells has been shown to increase
in response to oxidative stress via the mitogen-activated protein
kinase (MAPK) p38 (31). Cells lacking alsin may be more
vulnerable to oxidative stress than WT cells, as a direct result of
their defect in this adaptive Rab5-associated endocytic response.
In conclusion, we have demonstrated that Als2-deficient mice
have subtle motor neuron pathology and motor behavior ab-
normalities and that neurons from these mice show marked
defects in specific endosomal trafficking pathways with a severe
deficit in early endosome fusion stimulating activity in vitro.
These abnormalities are likely to be related to the role of alsin
as a GEF for Rab5 and may provide a cellular mechanism
associated with the pathogenesis of ALS2.
Materials and Methods
Further details are described in Supporting Materials and Methods.
Als2 Gene Targeting. A targeting construct was generated in the
an SA-IRES-?geo-PA [splice acceptor-internal ribosome entry
site-?-galactosidase-neomycin–poly(A) addition site] cassette
(Fig. 1A). The construct was electroporated into 2 ? 107
mEMS128 ES cells. Five correctly targeted clones were con-
firmed by PCR and Southern blotting.
Generation of Als2?/?Mice. ES cells were microinjected into
C57BL?6J blastocysts and implanted into pseudopregnant fe-
males to produce chimeras. Male chimeras from two clones were
bred with C57BL?6J females, and F1Als2?/?heterozygotes were
intercrossed to generate WT, heterozygous, and homozygous
mice. The lines were maintained by backcrossing of heterozy-
gotes with C57BL?6J mice. Absence of alsin in Als2?/?mice was
confirmed by Western blotting using the N-alsin-24 monoclonal
antibody (6) relative to a GAPDH control. All experiments were
performed on two lines, with identical results, except for behav-
ioral analysis, which was performed on line 1 only.
Histological Measurements. Tissues were prepared from 12- and
22-month-old WT and Als2?/?mice as described in Devon et al.
(6). Motor neuron size and density were determined in 25-?m-
thick cryostat sections. Measurements of cell size and axon
diameter were performed in 5-?m-thick paraffin-embedded
sections, or 500-nm semithin sections, respectively, stained with
cresyl violet or toluidine blue and measured by using STEREO-
INVESTIGATOR software (Microbrightfield, Williston, VT). Data
Devon et al.PNAS ?
June 20, 2006 ?
vol. 103 ?
no. 25 ?
were analyzed by using an independent-samples t test using SPSS
software. For conventional electron microscopy, specimens were
were collected on a Philips (Eindhoven, The Netherlands) EM
410 transmission electron microscope (TEM) equipped with a
Olympus Soft Imaging Solutions (Lakewood, CO) Megaview III
In Vitro Endosome Fusion Assay. Assessment of fusion of early
endosomes loaded with either dinitrophenylated ?-glucuroni-
dase or mannosylated anti-dinitrophenyl IgG was performed as
analyzed by using an unpaired two-tailed Student’s t test.
3- to 4-month-old male and female Als2?/?and WT littermate
mice were generated by homogenization of snap-frozen tissues
followed by centrifugation to remove particulates. Alsin protein
was supplied as supernatant of SF9 cells transfected with a
baculovirus vector encoding full-length mouse alsin (7).
Neuronal Culture. Dissociated primary neuronal cultures were
prepared from cerebellum of postnatal day 7–8 mice or cortex
of day 16.5 embryos. IGF1 and BDNF stimulations were per-
formed after starving cells for 3 h (in medium lacking serum for
CGNs, or B27 supplement for cortical neurons). IGF1 or BDNF
(50 ng?ml final concentrations) were added to the starvation
medium and incubated for the indicated times before fixation.
Imaging and Analysis. Images were acquired on a Zeiss Axiovert
M200 motorized microscope with a 63 ? 1.4 NA ACROMAT oil
immersion lens and a monochrome 14-bit Zeiss Axiocam HR
charged-coupled camera with 1,300 ? 1,030 pixels. Analysis was
performed by using IMAGEJ (NIH) and NORTHERN ECLIPSE
software (Empix Imaging, Mississauga, ON, Canada).
Behavioral Assessment. Forty F2mice (20 Als2?/?homozygotes
and 20 WT littermate controls, each 10 females and 10 males)
were tested. Locomotor activity was measured in an automated
open-field apparatus (Medical Associates, St. Albans, VT).
Forelimb grip strength was assessed by using a Chatillon Digital
Force Gauge DFIS-2 (Columbus Instruments, Columbus, OH).
Motor coordination and balance were assessed with an acceler-
ating rotarod (0 to 45 rpm in 4 min). Comparisons were made by
using unpaired two-tailed Student’s t tests (95% confidence
intervals) using PRISM 3.02 software (GraphPad, San Diego) and
repeated measures two-way ANOVA using SPSS software.
Note Added in Proof. A recent paper (33) has described a JPLS patient
with a missense mutation in ALS2.
We thank Dr. Andrew MacLeod for assistance with statistics, and Justin
Topp, Sandy Severson, and Bruce Horazdovsky of the Mayo Foundation,
who have generously shared their expertise and reagents with us. This
work was supported by the Canadian Institutes for Health Research
(separate grants to M.R.H., A.E.E.-H., W.T., and B.R.L.), the ALS
Association (M.R.H.), the Spastic Paraplegia Foundation, Inc. (SPF)
(M.R.H. and B.R.L.), the Jose Carreras International Leukemia Foun-
dation (E. D. Thomas Program) at Florida International University
(M.A.B.), the Canada Foundation for Innovation and the British Co-
lumbia Knowledge Development Fund (E.M.S.), the Canadian Genetic
Disease Network (B.R.L.), the Michael Smith Foundation (A.E.E.-H.),
and the EJLB Foundation (A.E.E.-H.). R.S.D. was the recipient of a
Wellcome Trust International Prize Travelling Research Fellowship.
E.M.S. holds a Canada Research Chair in Genetics and Behavior.
A.E.E.-H. is a Canadian Institutes of Health Research new investigator
and a Michael Smith Foundation for Health Research scholar. M.R.H.
holds a Canada Research Chair in Human Genetics.
1. Hadano, S., Hand, C. K., Osuga, H., Yanagisawa, Y., Otomo, A., Devon, R. S.,
Genet. 29, 166–173.
2. Yang, Y., Hentati, A., Deng, H. X., Dabbagh, O., Sasaki, T., Hirano, M., Hung,
W. Y., Ouahchi, K., Yan, J., Azim, A. C., et al. (2001) Nat. Genet. 29, 160–165.
M. H., Devon, R. S., Hayden, M. R., Andermann, F., Andermann, E. &
Rouleau, G. A. (2003) Ann. Neurol. 53, 144–145.
E. & Boespflug-Tanguy, O. (2002) Am. J. Hum. Genet. 71, 518–527.
5. Devon, R. S., Helm, J. R., Rouleau, G. A., Leitner, Y., Lerman-Sagie, T., Lev,
D. & Hayden, M. R. (2003) Clin. Genet. 64, 210–215.
6. Devon, R. S., Schwab, C., Topp, J. D., Orban, P. C., Yang, Y. Z., Pape, T. D.,
Helm, J. R., Davidson, T. L., Rogers, D. A., Gros-Louis, F., et al. (2005)
Neurobiol. Dis. 18, 243–257.
7. Topp, J. D., Gray, N. W., Gerard, R. D. & Horazdovsky, B. F. (2004) J. Biol.
Chem. 279, 24612–24623.
8. Otomo, A., Hadano, S., Okada, T., Mizumura, H., Kunita, R., Nishijima, H.,
Showguchi-Miyata, J., Yanagisawa, Y., Kohiki, E., Suga, E., et al. (2003) Hum.
Mol. Genet. 12, 1671–1687.
9. Kanekura, K., Hashimoto, Y., Kita, Y., Sasabe, J., Aiso, S., Nishimoto, I. &
Matsuoka, M. (2005) J. Biol. Chem. 280, 4532–4543.
10. McLauchlan, H., Newell, J., Morrice, N., Osborne, A., West, M. & Smythe, E.
(1998) Curr. Biol. 8, 34–45.
11. Sato, M., Sato, K., Fonarev, P., Huang, C. J., Liou, W. & Grant, B. D. (2005)
Nat. Cell Biol. 7, 559–569.
12. Gorvel, J. P., Chavrier, P., Zerial, M. & Gruenberg, J. (1991) Cell 64, 915–925.
13. Nielsen, E., Severin, F., Backer, J. M., Hyman, A. A. & Zerial, M. (1999) Nat.
Cell Biol. 1, 376–382.
14. Wucherpfennig, T., Wilsch-Brauninger, M. & Gonzalez-Gaitan, M. (2003) J.
Cell Biol. 161, 609–624.
15. Brown, T. C., Tran, I. C., Backos, D. S. & Esteban, J. A. (2005) Neuron 45,
16. Zerial, M. & McBride, H. (2001) Nat. Rev. Mol. Cell Biol. 2, 107–117.
(1998) Nat. Genet. 20, 163–169.
18. Watanabe, M., Dykes-Hoberg, M., Culotta, V. C., Price, D. L., Wong, P. C. &
Rothstein, J. D. (2001) Neurobiol. Dis. 8, 933–941.
19. Barbieri, M. A., Li, G., Colombo, M. I. & Stahl, P. D. (1994) J. Biol. Chem. 269,
20. Rink, J., Ghigo, E., Kalaidzidis, Y. & Zerial, M. (2005) Cell 122, 735–749.
21. Nagano, I., Ilieva, H., Shiote, M., Murakami, T., Yokoyama, M., Shoji, M. &
Abe, K. (2005) J. Neurol. Sci. 235, 61–68.
22. Kaspar, B. K., Llado, J., Sherkat, N., Rothstein, J. D. & Gage, F. H. (2003)
Science 301, 839–842.
23. Miaczynska, M., Pelkmans, L. & Zerial, M. (2004) Curr. Opin. Cell Biol. 16,
24. Howe, C. L. & Mobley, W. C. (2005) Curr. Opin. Neurobiol. 15, 40–48.
25. Hudson, A. J., Kiernan, J. A., Munoz, D. G., Pringle, C. E., Brown, W. F. &
Ebers, G. C. (1993) Brain Res. Bull. 30, 359–364.
26. Nguyen, M. D., Lariviere, R. C. & Julien, J. P. (2000) Proc. Natl. Acad. Sci. USA
27. von Lewinski, F. & Keller, B. U. (2005) Neurosci. Lett. 380, 203–208.
28. Hadano, S., Otomo, A., Suzuki-Utsunomiya, K., Kunita, R., Yanagisawa, Y.,
Showguchi-Miyata, J., Mizumura, H. & Ikeda, J. E. (2004) FEBS Lett. 575,
29. Cai, H., Lin, X., Xie, C., Laird, F. M., Lai, C., Wen, H., Chiang, H. C., Shim,
H., Farah, M. H., Hoke, A., et al. (2005) J. Neurosci. 25, 7567–7574.
30. Hadano, S., Benn, S. C., Kakuta, S., Otomo, A., Sudo, K., Kunita, R.,
Suzuki-Utsunomiya, K., Mizumura, H., Shefner, J. M., Cox, G. A., et al. (2006)
Hum. Mol. Genet. 15, 233–250.
31. Cavalli, V., Vilbois, F., Corti, M., Marcote, M. J., Tamura, K., Karin, M.,
Arkinstall, S. & Gruenberg, J. (2001) Mol. Cell 7, 421–432.
32. McCaffery, J. M. & Farquhar, M. G. (1995) Methods Enzymol. 257,
33. Panzeri, C., De Palma, C., Martinuzzi, A., Daga, A., De Polo, G., Bresolin, N.,
Miller, C. C., Tudor, E. L., Clementi, E. & Bassi, M. T. (May 2, 2006) Brain,
www.pnas.org?cgi?doi?10.1073?pnas.0510197103Devon et al.