J Neuropathol Exp Neurol
Copyright Ó 2014 by the American Association of Neuropathologists, Inc.
Vol. 73, No. 6
June 2014
pp. 519Y535
ORIGINAL ARTICLE
Mechanisms Involved in Spinal Cord Central Synapse Loss
in a Mouse Model of Spinal Muscular Atrophy
Olga Tarabal, PhD, Vı́ctor Caraballo-Miralles, PhD, Andrea Cardona-Rossinyol, MSc,
Francisco J. Correa, MSc, Gabriel Olmos, PhD, Jerònia Lladó, PhD, Josep E. Esquerda, MD, PhD,
and Jordi Calderó, MD, PhD
From the Unitat de Neurobiologia CelIlular, Departament de Medicina Experimental, Facultat de Medicina, Universitat de Lleida and Institut de
Recerca Biomèdica de Lleida, Lleida, Catalonia (OT, FJC, JEE, JC); and
Grup de Neurobiologia CelIlular, Institut Universitari d’Investigacions en
Ciències de la Salut and Departament de Biologia, Universitat de les Illes
Balears, Palma de Mallorca (VC-M, AC-R, GO, JL), Spain.
Send correspondence and reprint requests to: Jordi Calderó, MD, PhD,
Unitat de Neurobiologia CelIlular, Departament de Medicina Experimental, Facultat de Medicina, Universitat de Lleida and Institut de Recerca
Biomèdica de Lleida, Av Rovira Roure 80, Lleida 25198, Catalonia, Spain;
E-mail: [email protected]
This work was supported by grants from Ministerio de Ciencia y Tecnologı́a
and Ministerio de Economı́a y Competitividad and was cofinanced by Le
Fondo Europeo de Desarrollo Regional, (Grant Nos. SAF2011-22908 and
SAF2012-31831) and Fundación Genoma España/Fundación Atrofia Muscular Espinal.
The authors declare that no conflict of interest exists.
cord and causes muscular weakness and atrophy of limb and
trunk muscles. With an incidence of 1 in 6,000 to 10,000 live
births and a carrier frequency of 1 in 35 to 50, this neurodegenerative disease is the leading genetic cause of early
childhood lethality (1, 2). Spinal muscular atrophy is caused
by a homozygous deletion or specific mutations in the SMN1
(survival of motor neuron-1) gene, which is localized in the
telomeric region of chromosome 5 (3). SMN1 codes for survival motor neuron (SMN) protein, which plays an essential
role in the assembly of small nuclear ribonucleoproteins
necessary for the preYmessenger RNA splicing machinery (4).
In humans, an additional centromeric gene, SMN2, also produces low levels of functional SMN. Whereas SMN1 generates full-length SMN, a single nucleotide polymorphism in
SMN2 results in the exclusion of exon 7 from most transcripts, producing a high amount of unstable truncated SMN
(SMN$7) protein, which is rapidly degraded (5). The low
levels (È10%Y20%) of functional full-length SMN generated
by SMN2 are insufficient to compensate for the lack of SMN1
function. Thus, in the absence of a functional SMN1 gene, the
severity of the disease correlates with the copy numbers of the
SMN2 gene (6). In contrast with humans, mice have a single
SMN (Smn) gene, and its inactivation leads to embryonic lethality (7). One of the most widely used models of SMA is
a mouse model lacking the endogenous Smn gene but with
2 copies of a transgenic human SMN2 and an additional
transgene expressing high levels of the human SMN transcript
lacking exon 7 (SMN$7). Mice with this genotype (Smnj/j;
SMN2+/+;SMN$7+/+)V‘‘SMN$7’’ miceVexhibit a severe postnatal SMA phenotype with a median life span of approximately 2 weeks (6, 8).
Although SMN is a ubiquitous protein, its low expression levels predominantly lead to damage of lower MNs (9).
Interestingly, recent reports indicate that SMN is involved in
U12-dependent splicing events, which are important for MN
function (10). However, only a modest loss of MNs has been
found in SMA mouse models at end stages of diseases and in
patients with SMA (6, 11Y16). This suggests that MN death is
a late event in the pathogenesis of this disorder (17).
Astrocytes and microglia respond profoundly to neuronal
injury and undergo a series of metabolic and morphologic
changes known as ‘‘reactive gliosis’’ (18); glial activation often
occurs in parallel with either nonspecific (innate) or specific
(adaptive) immune responses within the CNS in a process
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
519
Abstract
Motoneuron (MN) cell death is the histopathologic hallmark of
spinal muscular atrophy (SMA), although MN loss seems to be a late
event. Conversely, disruption of afferent synapses on MNs has been
shown to occur early in SMA. Using a mouse model of severe SMA
(SMN$7), we examined the mechanisms involved in impairment of
central synapses. We found that MNs underwent progressive degeneration in the course of SMA, with MN loss still occurring at late
stages. Loss of afferent inputs to SMA MNs was detected at embryonic stages, long before MN death. Reactive microgliosis and
astrogliosis were present in the spinal cord of diseased animals after
the onset of MN loss. Ultrastructural observations indicate that dendrites and microglia phagocytose adjacent degenerating presynaptic
terminals. Neuronal nitric oxide synthase was upregulated in
SMN$7 MNs, and there was an increase in phosphorylated myosin
light chain expression in synaptic afferents on MNs; these observations implicate nitric oxide in MN deafferentation and suggest that
the RhoA/ROCK pathway is activated. Together, our observations
suggest that the earliest change occurring in SMN$7 mice is the loss
of excitatory glutamatergic synaptic inputs to MNs; reduced excitability may enhance their vulnerability to degeneration and death.
Key Words: Glia, Motoneuron, Nitric oxide, RhoA/ROCK pathway,
SMN$7 mouse, Spinal muscular atrophy, Synaptic afferents.
INTRODUCTION
Spinal muscular atrophy (SMA) is an autosomal recessive disease that affects > motoneurons (MNs) in the spinal
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
Tarabal et al
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
termed ‘‘neuroinflammation.’’ In recent years, neuroinflammation has been thought to play an active role in the pathology
of different neurodegenerative diseases, including Alzheimer
disease, Huntington disease, Parkinson disease, multiple sclerosis, and amyotrophic lateral sclerosis (ALS) (19Y24). In addition to MN loss, astroglial or microglial activation is observed
early in the spinal cord of patients with SMA and may be another important pathologic aspect of this disease (12, 25Y27).
Astrocytes and microglia may propagate injury and, under
certain circumstances, may even initiate neurodegeneration,
but the pathologic roles of glial cells in SMA have not been
studied in depth.
Glial cells play important roles in the formation, function, and elimination of synapses under normal and pathologic
conditions (28). Different studies performed on SMA mouse
models have reported synaptic defects at the neuromuscular
junction, including impaired maturation, structural disruption
and denervation, deficient neurotransmitter release, and neurofilament accumulation (9, 12, 15, 17, 25, 29Y32). Additional
studies indicate that defects in central synapses are also involved in SMA pathogenesis (13, 25, 33, 34). The molecular
bases underlying these synaptic alterations are poorly defined.
It has been proposed that nitric oxide (NO) plays a role in
synaptic remodeling (35) and that expression of neuronal
NO synthase (nNOS) induces synapse loss on adult and neonatal MNs. Nitric oxideYdirected synapse elimination is mediated by the activation of the small Rho GTPase RhoA and its
major effector Rho kinase ROCK by inducing neurite retraction (36). The role of nNOS in SMA synaptic defects has not
been investigated.
In the present study, we used the SMN$7 mouse model
to examine the contribution of astroglial and microglial cells
to SMA spinal cord pathology, including degenerative changes
in MN synaptic inputs. In addition, we show that nNOS is
upregulated and the RhoA/ROCK pathway is activated during
central synaptic remodeling in SMA.
Histology and Cell Count
MATERIALS AND METHODS
Animals
Mice were purchased from The Jackson Laboratory
(Sacramento, CA). Experimental mice were obtained by breeding pairs of SMA carrier mice (Smn+/j;SMN2+/+;SMN$7+/+) on
a FVB/N background. Identification of wild-type (WT; Smn+/+;
SMN2+/+;SMN$7+/+) and mutant (Smnj/j;SMN2+/+;SMN$7+/
+
[SMN$7]) mice was carried out by polymerase chain reaction
genotyping of DNA extracted from the tail, as previously described (6). Age-matched WT littermates of mutant animals
were used as controls.
Mice were anesthesized with an intraperitoneal injection of 2% pentobarbital (1 mL/10 g body weight) and, except
for electron microscopy studies, killed by quick decapitation.
The WT and SMN$7 animals analyzed were pooled in 6 different groups as follows: embryonic day (E) 18, postnatal
day (P) 0Y1, P4Y5, P7Y8, P10Y11, P14Y15. All experiments
were performed according to the guidelines of the European
Council Directive for the Care of Laboratory Animals.
520
For MN count, the spinal cord of mice was quickly
dissected and fixed in Carnoy or Bouin solution (for mice
older than P5) and processed for paraffin embedding. Serial
transverse sections (8 Km thick for E18; 12 Km thick for
P0Y1, P4Y5, and P7Y8; and 14 Km thick for P14Y15) were
obtained through the entire lumbar segment of the spinal cord
and stained with either Cresyl violet (E18Y5) or hematoxylin
and eosin (P7Y15). Motoneurons were identified by their size
and shape. The numbers of apparently healthy MNs and
degenerating (pyknotic) cells in the ventral horn region were
counted blindly on a side of every 10th section according to
established procedures (37, 38). For healthy MNs, only cells
with a large nucleus and a visible clump of nucleolar material,
along with a substantial cytoplasm, were counted. For pyknotic
neurons, intact cells with condensation of nuclear chromatin,
broken spherical profiles apparently engulfed by phagocytes,
and rounded hyaline masses were counted. With these criteria, it was not necessary to use a correction factor to avoid
double counting (38). The numbers of MNs and pyknotic
neurons counted were multiplied by 10 to estimate the total
number of these cells per ventral horn.
Immunocytochemistry and Image Analysis
Lumbar spinal cords were fixed by immersion in 4%
paraformaldehyde in 0.1 mol/L phosphate buffer (PB; pH 7.4)
for 24 hours, cryoprotected with 30% sucrose in 0.1 mol/L
PB, embedded in Tissue Freezing Medium (Triangle Biomedical Sciences, Durham, NC), and frozen. Transverse serial cryostat sections (10Y16 Km thick) were obtained and
stored at j80-C.
Sections were sequentially rinsed in PBS containing 0.1%
Triton X-100 for 30 minutes, blocked in 10% normal goat serum, and incubated with the primary antibody overnight at 4-C.
The following primary antibodies were used: mouse monoclonal antiYphosphorylated neurofilament heavy chain (SMI-31;
1:1000; Abcam, Cambridge, United Kingdom), mouse monoclonal antiYnonphosphorylated neurofilament heavy chain (SMI-32;
1:1000; Abcam), rabbit polyclonal antiYvesicular glutamate
transporter 1 (VGLUT1; 1:1000; Synaptic Systems, Göttingen,
Germany), rabbit polyclonal antiYvesicular GABA transporter
(VGAT; 1:500; Millipore, Temecula, CA), rabbit polyclonal
anti-ionized calcium-binding adaptor molecule 1 (Iba1; 1:500;
Wako Pure Chemical Industries Ltd, Osaka, Japan), rat monoclonal anti-mouse CD68 (1:100; AbD Serotec, Oxford, United
Kingdom), rabbit polyclonal antiYglial fibrillary acidic protein
(GFAP; 1:1000; Dako Cytomation, Glostrup, Denmark), chicken
polyclonal anti-GFAP (1:1000; Abcam), mouse anti-synaptophysin
(Syphys; 1:200; Millipore), and rabbit antiYphosphorylated myosin light chain (P-MLC; 1:250; Cell Signaling Technologies,
Danvers, MA). The 2 antibodies to GFAP gave similar results
and are referred to as ‘‘anti-GFAP.’’ After several washes, sections were incubated at room temperature for 1 hour with the
appropriate secondary fluorescent antibodies: Alexa Fluor 488
goat anti-rabbit IgG (1:500; Molecular Probes, Eugene, OR),
Alexa Fluor 555 goat anti-mouse IgG (1:500; Molecular Probes),
Dylight 549 donkey anti-chicken IgG (1:500; Jackson ImmunoResearch Laboratories, West Grove, PA), Dylight 488 donkey
anti-rat IgG (1:500; Jackson ImmunoResearch Laboratories),
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
Mechanisms of Motoneuron Deafferentation in SMA
Dylight 649 donkey anti-rabbit IgG (1:500; Jackson ImmunoResearch Laboratories), and Dylight 405 donkey anti-rat IgG
(1:100; Jackson ImmunoResearch Laboratories). Sections were
also labeled with 4¶,6-diamidino-2-phenylindole dihydrochloride
(50 ng/mL; Molecular Probes) for DNA staining and counterstained with Neuro-Trace 500/525 or 530/615 fluorescent Nissl
stain (Molecular Probes). Sections were then washed and mounted
using Fluorescent Mounting Medium (Dako Cytomation) or
Vectashield (Vector Laboratories, Burlingame, CA). Immuno-
histochemical controls, obtained by omitting the primary antibody, resulted in the absence of immunostaining.
Mounted slices were examined and imaged with an
Olympus BX51 epifluorescence microscope (Olympus, Hamburg,
Germany) equipped with a DP30BW camera or a Leica DMR
epifluorescence microscope (Leica Microsystems, Wetzlar,
Germany) equipped with a Leica DC 300 camera. A FluoView
500 Olympus and a Leica TCS SP2 confocal laser-scanning
microscope were also used. Confocal micrographs of every
FIGURE 1. Motoneuron pathology in SMN$7 mice. (A) Quantification of apparently healthy MNs in the ventral horn of the lumbar
spinal cord of WT and SMN$7 mice at the indicated ages (n = 5Y6 animals per age and experimental condition). (B, C) Representative images of MNs present in the lateral motor column of the lumbar spinal cord of 2 mice (WT, B; SMN$7, C) used for cell
counting; micrographs of paraffin sections stained with hematoxylin and eosin were taken. (D) Quantification of degenerating
(pyknotic) cells in the ventral horn of the lumbar spinal cord of P4Y5 and P14Y15 WT and SMN$7 animals (n = 6 mice per age and
experimental condition). (E, F) Representative images of hematoxylin and eosinYstained sections of P5 (E) and P14 (F) SMN$7
lumbar spinal cords showing the appearance of pyknotic (degenerating) cells (arrows), presumably MNs, in the ventral horn. (G)
Area (in square micrometers) of MN soma in the ventral horn of the lumbar spinal cord of WT and SMN$7 mice at different ages
(140Y160 MNs from 5Y6 mice per age and experimental condition were measured); (C) in comparison with P15 WT animals (B),
the spinal MNs of P15 SMN$7 mice display a reduction in the somatic area. (H) Percentage of SMI-31 (phosphorylated heavychain neurofilament)Ypositive MNs in P14Y15 WT and SMN$7 mice (50 MNs from 4Y5 mice per group were examined). (IYL)
Representative images of lumbar spinal cord showing SMI-31Yimmunostained MNs (green) after fluorescent Neuro-Trace Nissl
staining (red) in P14 WT (I, J) and SMN$7 (K, L) mice. Data on all graphs are expressed as mean T SE. * p G 0.05, ** p G 0.001, and
*** p G 0.001 versus age-matched WT animals (one-way ANOVA followed by Bonferroni test in A, B, and E, or Student t-test in
F). Scale bars = (B; valid for C) 50 Km; (F; valid for E) 10 Km; (I; valid for JYL) 10 Km.
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
521
Tarabal et al
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
30th section of the entire lumbar spinal cord were taken. The
same scanning parameters were used for the acquisition of
confocal images corresponding to the different experimen-
tal groups. Slides from the different animals were processed
in parallel for immunocytochemistry and subsequent confocal imaging.
FIGURE 2. Electron micrographs showing degenerative alterations in ventral horn MNs and microglial cells in SMN$7 mice. (AYE)
Images taken on P14 (AYC) and P7 (D, E). (A) Motoneuron soma from a WT animal displaying a typical regular, round nucleus
(blue) filled with fine clumped chromatin; the cytoplasm is shown in yellow. (B) Atrophic MN soma (yellow) from an SMN$7
mouse displaying a highly convoluted nucleus (blue) and a dark cytoplasm (yellow). This altered MN maintains some axosomatic
synaptic profiles (* in inset); some swollen dendritic profiles are seen adjacent to MN soma (arrows). (C) A microglial cell (green)
interposed between 2 apparently healthy MN somata. MN cytoplasm and adjacent dendrites are shown in yellow; nuclei of MNs
and the microglial cell are shown in blue. (D) A microglial cell (green) in the ventral horn containing large and electron-dense
cytoplasmic inclusions (arrows), reflecting its phagocytic activity. (E) A dendritic profile (yellow), in which normal axodendritic
synaptic terminals are lacking, is abnormally surrounded by a microglial cell (green). One remaining abnormal axon terminal (red)
has been captured embedded in the dendritic cytoplasm being phagocytosed (see Fig. 3 for details). Scale bars = (AYC) 2 Km;
(inset to B) 0.5 Km; (D) 2.5 Km; (E) 1 Km.
522
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
Mechanisms of Motoneuron Deafferentation in SMA
Levels of GFAP and Iba1 immunoreactivity were quantified by digital image analysis using Visilog 6.3 software
(Noesis, Orsay, France). The numbers of VGLUT1-, VGAT-,
and P-MLCYimmunoreactive synaptic boutons on MN soma
were manually counted on the screen. Only boutons in close
contact with MNs showing a large nucleus, a visible nucleolus,
and a healthy appearance were included in the counts, which
were then normalized to the perimeter of MN soma. The areas
of VGLUT1- and VGAT-positive boutons and the length of
Syphys-immunoreactive terminals were also measured.
and postsynaptic densities were classified as symmetric; and
synapses containing a subsynaptic cistern were classified as
C-terminal. A proportion of afferent boutons were scored
as ‘‘nonclassified.’’
Statistical Analysis
All quantitative data are expressed as mean T SE. Statistical analysis was performed with either Student t-test or
one-way analysis of variance (ANOVA) followed by post hoc
Bonferroni test. The level of significance was set at p G 0.05.
Electron Microscopy
Deeply anesthesized mice were transcardially perfused
with cold saline serum followed by a freshly made solution
containing 1% glutaraldehyde and 1% paraformaldehyde in
0.1 mol/L PB (pH 7.4). The spinal cord was dissected by
dorsal laminectomy, and the lumbar segment was removed
and placed in the same fixative overnight at 4-C. After washes
with PB, samples were sectioned to obtain transverse slices
that were postfixed with 1% OsO4 for 2 hours, dehydrated,
and embedded in EMbed 812 (Electron Microscopy Sciences,
Fort Washington, PA). Ultrathin sections from selected
areas, including ventral horns, were counterstained with uranyl
acetate and lead citrate and observed with a Zeiss EM 910
(Zeiss, Oberkochen, Germany) electron microscope.
For analysis of dendrites, images of the spinal cord
ventral horn at the neuropil area interposed between MN
somata were randomly taken. Axosomatic synapses were
studied after scoring systematically taken images of the periphery of MN cell bodies. Morphometry was performed
using Visilog 6.3 software (Noesis). In transversally sectioned
large dendrites, the following parameters were scored: occurrence of autophagosomes within dendrites; density, size,
and type of axodendritic synapses; and frequency of autophagosomes in presynaptic terminals. In the sectioned soma
of large MNs, the following parameters were analyzed: density, size, and type of axosomatic synapses; incidence of
autophagosomes in presynaptic nerve terminals; and abundance of mitochondria. Axodendritic and axosomatic synapses were identified based on classical descriptions (39) but
were slightly modified for immature mice spinal cord ultrastructure. Thus, MN afferent boutons were categorized as
follows: Synapses with a thick postsynaptic density were
classified as asymmetric; synapses with similar presynaptic
RESULTS
MN Loss in the Spinal Cord of SMN$7 Mice
A moderate but significant reduction in the number of
MNs was observed in the lumbar spinal cord of SMN$7 mice.
Motoneuron loss started between P0Y1 and P4Y5 (È20% of
WT mice) and was slightly more prominent between P7Y8
and P14YP15 (Figs. 1AYC). Thus, there was an approximately
25% decrease in MN number on P14Y15 in SMN$7 versus
WT animals. Counts of pyknotic (degenerating) cells, presumably MNs, in the ventral horn of P4Y5 and P14Y15 mice
revealed significantly increased numbers in SMN$7 animals
versus those in WT littermates (Figs. 1DYF). In addition to the
decrease in MN survival in the spinal cord of P7Y8 and P14Y15
mutants, SMN$7 MNs also showed a significant size reduction starting on P7Y8, becoming more marked on P14Y15
(Fig. 1G, compare Figs. 1B and C). Another histopathologic
hallmark of MN diseases is the presence of abnormally phosphorylated neurofilament (40). There was a significant increase (È58%) in the number of ventral horn MNs that are
immunoreactive to SMI-31 antibody against phosphorylated
neurofilament in SMN$7 mice on P14Y15 versus WT littermates (Figs. 1HYL). Overall, these results suggest that spinal
cord MNs progressively degenerate postnatally in mutant mice
and that MN loss persists in the end stages.
Ultrastructural Pathology of MN Degeneration
in SMN$7 Mice
Ultrastructural analysis of the spinal cord ventral horn
of SMN$7 mice showed MN degeneration on P7Y8, which
was more extensive at end stage (P14Y15). Most degenerating
MNs did not have a typical apoptotic ultrastructure, such as the
formation of circumferential electron-dense spheroids of highly
TABLE. Morphometric Analysis of Electron Micrographs From MN Somata and Dendrites in the Spinal Cord of P7 to P8 WT
and SMN$7 Mice
Synaptic occupancy in the membrane of large ventral horn dendritic profiles, %
Large ventral horn dendritic profiles lacking axodendritic synapses, %
Length of axodendritic synaptic contacts, Km
Length of axosomatic synaptic contacts, Km
Synaptic occupancy in the somatic membrane of ventral horn MNs, %
Dendrites containing autophagosomes, %
WT
SMN$7
27.77 T 4.18 (n = 33)
0 (n = 33)
0.96 T 0.06 (n = 84)
0.97 T 0.14 (n = 28)
27.19 T 6.51 (n = 13)
29.57 T 7.3 (n = 33)
22.04 T 2.09 (n = 48)
11.14 T 4.01 (n = 75)*
0.75 T 0.04 (n = 116)†
1.02 T 0.07 (n = 78)
26.52 T 3.71 (n = 33)
33.01 T 1.35 (n = 48)
Numbers represent the mean T SE of data obtained from 3 to 5 animals.
* p G 0.05 versus WT (Student t-test).
† p G 0.01 versus WT (Student t-test).
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
523
Tarabal et al
524
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
Mechanisms of Motoneuron Deafferentation in SMA
condensed chromatin and apoptotic bodies (41). Instead, they
showed wide and marked cytoplasmic condensation and occasionally severe mitochondrial vacuolization. The nuclei of
SMN$7 MNs displayed irregularVsometimes hyperconvoluted
Vcontours (Figs. 2B vs A) and contained irregular clumps of
condensed chromatin (data not shown). Abnormally invaginated or multilobular nuclei have been reported in fetal MNs
from human cases of SMA (42); they were described as a new
phenotype (Type 4) of MN death (43). Dendritic profiles immediately adjacent to these degenerating MNs seemed to undergo necrotic-like (possibly glutamate-mediated excitotoxicity
-induced) changes because they appeared swollen and completely
depleted of cellular organelles (Fig. 2B), comparable to the
description by Olney and Ishimaru (41).
These obvious and terminal degenerative changes seemed
to be preceded by other more subtle alterations revealed by
morphometric analysis (Table), including mitochondrial depletion in apparently healthy SMN$7 MNs. Other involutional
changes may herald terminal MN death in SMA. Indeed, there
were increased numbers of autophagic vacuoles within MN
afferent nerve terminals apposed to the soma and presumptive
proximal dendrites (Figs. 3AYJ). Similar autophagic vacuoles
were also observed on P7Y8 in WT synapses, but significantly
less than in mutants (Figs. 3B, I, J). Further degenerative
changes, such as altered synaptic vesicle morphology, were also
observed in some SMN$7 axodendritic synapses (Fig. 3C).
Although nonsignificant, apparently increased numbers of autophagic organelles were also seen inside the dendritic profiles
of mutant animals (Table). A more detailed examination of
different forms of autophagic vacuoles within dendrites suggested that they arise from the engulfment of degenerating presynaptic terminals, which are found inside autophagosomes
(Figs. 3D, F). Thus, at least in part, autophagic vacuoles within
dendrites come from phagocytosis of adjacent degenerating
presynaptic structures; therefore, their interpretation as heterophagic organelles should be considered (Fig. 3K). All of these
alterations indicate overall synaptic loss (Table). Microglial
profiles often enwrapped dendrites containing degenerating
synapses, suggesting that they will be degraded further by
microglial phagocytosis (Figs. 2E; 3E, F). In fact, infiltrating
microglial cells, some of which contained abundant large heterophagic inclusions, were observed in the ventral horn neuropil located between MN somata (Figs. 2C, D).
Astroglial and Microglial Activation in the
Ventral Horn of the Lumbar Spinal Cord of
SMN$7 Mice
For analysis of astroglia and microglia, immunofluorescence studies were performed on sections of the lumbar
spinal cord at different time points ranging from the newborn
stage to the presymptomatic stage (P4Y5), early symptomatic
stage (P7YP8), and end stage (P14Y15) of the disease. There
was a significant increase in GFAP-positive astrocyte profiles
surrounding MNs in the ventral horn of P4Y5 SMN$7 mice
(Fig. 4A). This astrocyte reaction persisted until the end stages
and was more marked on P10Y11 and P14Y15 (Figs. 4AYC).
To determine whether there were differences in astroglial activation between the dorsal horn and the ventral horn, we
performed immunofluorescence analysis with anti-GFAP antibody independently in the dorsal and ventral areas of the
spinal cord of WT and SMN$7 animals. In P4Y5 mutant mice,
astroglial activation was predominantly observed in the ventral
horn, with the dorsal horn displaying immunoreactivity to GFAP
similar to that observed in WT littermates (Figs. 4DYG, L). In
contrast, on P10Y11, notable astroglial activation was observed
in both the dorsal horn and the ventral horn of SMN$7 mice
(Figs. 4HYL).
Immunofluorescence with anti-Iba1 antibody revealed
an overt increase in the density of microglial cells in the
lumbar spinal cord of SMN$7 versus WT animals. Diseased
animals showed an increase (though not statistically significant) in Iba1-immunoreactive profiles around MNs on P4Y5.
This increase was significant on P7Y8 and persisted until the
late stage of the disease (Fig. 5A). Interestingly, a high concentration of Iba1-positive profiles was observed adjacent
FIGURE 3. Autophagosome-like structures are increased in MN synaptic afferents and dendrites in the spinal cord of P7 SMN$7
mice. (AYH) Dendritic profiles (AYF) and MN somata (G, H) are shown in yellow. Axodendritic (AYF) and axosomatic (G, H)
nerve terminals are shown in red; glial profiles are shown in green. (A, B, G) From P7 WT animals. (CYF, H) From P7 SMN$7 mice.
(A) Unaltered dendritic profile from a WT that is extensively covered by afferent nerve terminals. (B) Autophagic vacuoles are
occasionally seen in WT dendrites. (CYF) A presumptive sequence of changes in axodendritic synapse degeneration in SMA. These
changes include initial alteration of degenerating presynaptic boutons consisting of altered synaptic vesicle morphology (C), a
damaged nerve terminal (arrow) beginning to sink into dendritic cytoplasm, a degenerating axon terminal containing several
autophagosome-like vacuoles (arrow) deeply embedded in dendritic cytoplasm (D), a degenerating nerve terminal (arrow)
containing autophagosome-like vacuoles, and altered synaptic vesicles enwrapped by dendritic membranes in a phagocytosis-like
process (E). Overall, this structural assembly is in turn engulfed by a microglial cell (*). (F) A dendritic profile containing a late large
phagosome-like vacuole (arrow) presumably derived from nerve terminal phagocytosis by a dendrite (similar to E). This
degenerating dendrite appears engulfed by microglial processes (*). (G, H) Representative images of axosomatic nerve terminals in
WT (G) and SMN$7 (H) mice displaying accumulation of autophagosome-like vacuoles in SMA degenerating nerve terminals
(arrows). (I, J) Quantification of autophagosome-like vacuoles in axodendritic (I) and axosomatic (J) terminal boutons from WT
and SMN$7 mice. (I) Eighty-four (WT; n = 3 animals) and 116 (SMN$7; n = 2 animals) synapses were scored. (J) Twenty-eight
(WT; n = 2 animals) and 80 (SMN$7; n = 2 animals) synapses were scored. (K) Diagram depicting the suggested mechanism of
synaptic bouton elimination in SMA. A degenerating afferent nerve terminal (DNT) containing autophagic vacuoles is engulfed by a
dendrite (D), later evolving to phagocytic inclusions within the dendrite. Overall, this complex is further eliminated by the recruitment of microglial cells (M). A, astrocytic processes; NT, normal nerve terminals. * p G 0.05 and ** p G 0.01 versus WT (Student
t-test). Scale bars = (A, B, G, H) 500 nm; (CYF) 100 nm.
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
525
Tarabal et al
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
FIGURE 4. Astroglial activation in SMN$7 mice. (A) Quantification of GFAP-positive astroglia in the ventral horns of the spinal cord
of WT and SMN$7 mice at different ages. Bars represent the percentage of the area occupied by GFAP-positive profiles in serial
sections of the lumbar spinal cord of 3 to 5 animals per experimental condition. (B, C) Confocal micrographs of lumbar spinal cord
cryostat sections from WT (B) and SMN$7 (C) mice on P15 showing GFAP-immunostained profiles (red) adjacent to MN cell
bodies, which are visualized after fluorescent Neuro-Trace Nissl staining (green). Note the increase in GFAP immunoreactivity in (C)
compared with (B). (DYK) Representative immunofluorescence images after GFAP immunocytochemistry (green) in sections of
the lumbar spinal cord of P5 (DYG) and P11 (HYK) WT (D, F, H, J) and SMN$7 (E, G, I, K) mice showing selective astroglial
activation in the ventral horn on P5 (compare E with D) and in both dorsal and ventral horns on P11 (compare I with H, and K with
J). Note that the increase in GFAP immunoreactivity in SMN$7 animals is much more intense and extended on P11 than on P5. (L)
Quantification of GFAP-positive astroglia in the spinal cord of P5 and P11 WT and mutant animals. Bars represent the percentage of
the area occupied by GFAP in SMN$7 mice with respect to age-matched WT animals; 8 sections per animal and 4 different fields
per section of spinal cord from at least 4 mice per experimental condition were analyzed. Data in graphs are expressed as mean T
SE. ** p G 0.01 and *** p G 0.001 versus WT animals from each age (one-way ANOVA and Bonferroni test). Scale bars = (C; valid for
B) 50 Km; (K; valid for DYJ) 50 Km.
526
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
to MNs in the ventral horn of SMN$7 animals (compare
Figs. 5B and C). Colocalization studies with anti-Iba1 and antiCD68 antibodies, as a marker for activated microglia (44),
were performed to assess the extent of microglial activation.
We found a significant increase (È2.3-fold; p G 0.05) in the
number of profiles showing both Iba1- and CD-68Ypositive
immunoreactivity in the spinal cord of P7Y8 SMN$7 mice
compared with their WT littermates, and this increase was even
higher (È3-fold; p G 0.001) on P14Y15 (Figs. 5D, E). This indicates that the increase in microglial cells observed in SMN$7
mice mostly corresponds to activated microglia.
Mechanisms of Motoneuron Deafferentation in SMA
Changes in Afferent Inputs to Spinal Cord MNs
in SMN$7 Mice
Alterations in sensory-motor connectivity and glutamatergic synaptic afferents on MNs have been reported in distinct
mouse models of SMA (25, 33, 34). To explore whether there
was a correlation between changes in glial reaction and the loss
of MN afferent synapses in SMN$7 mice, we quantified the
number of glutamatergic excitatory and GABAergic inhibitory
synapses after immunofluorescence using anti-VGLUT1 and
anti-VGAT antibodies, respectively (45), in combination with
fluorescent Nissl staining for MN identification. There was a
significant decrease in VGLUT1 and VGAT synaptic boutons
contacting MN somata from P0YP1 to P14YP15 in WT mice
(Figs. 6A, D). This decrease in bouton density cannot be attributed to the increase in MN size occurring during normal
development because comparable results were obtained when
the counts were referred to individual MN somata (data not
shown). The gradual loss of central synapses on MNs at postnatal ages could in part be related to the programmed cell death
of spinal interneurons in normal mice, which has been reported
to occur postnatally in the rat (46). VGLUT1 terminals on MNs
of SMN$7 mice were found to be significantly reduced compared with WT littermates: These differences were already seen
at embryonic stages (E18; È25% decrease in mutant mice in
relation to WT animals; p G 0.05) and persisted at the postnatal
stages of the disease, with a maximal loss (È37%; p G 0.001)
on P0 to P1 (Figs. 6AYC). Although an increase in the density
of VGLUT1-positive boutons was found on SMN$7 MNs
between E18 and P0YP1, these differences were not statistically
significant; no significant changes in somatic VGLUT1 bouton
density were found either when we compared mutant animals
at the different postnatal ages. A significant reduction in the
FIGURE 5. Microglial activation in SMN$7 mice. (A) Quantification of Iba1-positive microglial cells in the ventral horn of
the lumbar spinal cord of WT and SMN$7 mice at the indicated ages. Bars represent the percentage of the area occupied
by Iba1-positive profiles in serial sections of the spinal cord of 3
to 5 animals per experimental condition. (B, C) Representative
confocal micrographs of lumbar spinal cord cryostat sections
from WT (B) and SMN$7 (C) mice on P7 showing Iba1 immunostaining (green) and Neuro-Trace Nissl staining (red) for
visualization of MN cell bodies. Note the marked increase in
Iba1 immunoreactivity observed in (C) compared with (B).
(D) Percentage of Iba1-positive profiles displaying CD68 immunostaining in the ventral horn of the lumbar spinal cord of
WT and SMN$7 mice at the indicated ages. Counts were
performed in serial transverse cryostat sections of the spinal
cord of 4 animals per age and experimental condition; sections
were processed for double immunofluorescence with anti-Iba1
and anti-CD68 antibodies and stained with 4¶,6-diamidino-2phenylindole dihydrochloride for nuclei visualization. (E) Representative confocal micrograph showing colocalization of
Iba1-positive profiles (red) with CD68-positive profiles (green)
in the ventral horn of the lumbar spinal cord of an SMN$7
mouse on P7. Fluorescent Nissl staining for MN identification
is shown in blue. Data in graphs are expressed as mean T SE. *
p G 0.05 and *** p G 0.001 versus WT animals from each age
(Student t-test). Scale bars = (C; valid for B) 50 Km; (E) 5 Km.
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
527
Tarabal et al
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
FIGURE 6. SMN$7 mice show decreased numbers of excitatory glutamatergic and inhibitory GABAergic synapses on MN somata.
(A) Numbers of VGLUT1-positive boutons on MNs (per 100-Km soma perimeter) in the lumbar spinal cord of WT and SMN$7
mice at the indicated ages. (B, C) Representative confocal micrographs of VGLUT1-immunoreactive synaptic boutons (green) in
the spinal cord ventral horn of P15 WT (B) and SMN$7 (C) mice. MN somata are visualized after fluorescent Nissl staining (red).
(D) Numbers of VGAT-positive boutons on MNs (per 100-Km soma perimeter) in the lumbar spinal cord of WT and SMN$7 mice
at different ages. (E, F) Representative confocal micrographs showing VGAT-immunoreactive boutons (green) in the spinal cord
ventral horn of P8 WT (E) and SMN$7 (F) mice. Fluorescent Nissl staining (red) was used for MN visualization. (GYJ) Representative micrographs showing the colocalization of the microglia/lysosome marker CD68 (green in G) with a phagocytosed VGLUT1positive profile (red in H). Merged images are shown in (I) and (J) combined with either fluorescent Nissl (blue in I; *adjacent MN
profile) or 4¶,6-diamidino-2-phenylindole dihydrochloride staining (blue in J; *MN nucleus). The microglial profile is delimited with
a dashed line. Values shown in graphs are expressed as mean T SE; 60 to 80 MNs from 3 to 4 different animals per age and
experimental condition were analyzed. * p G 0.05, ** p G 0.01, and *** p G 0.001 (one-way ANOVA and Bonferroni test). Experimental conditions compared are indicated with either solid light-blue or dashed dark-blue lines (1-way ANOVA and Bonferroni
test). Scale bars = (B, E; valid for C, F) 20 Km; (G; valid for HYJ) 10 Km.
528
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
number of VGAT synaptic afferents on SMN$7 MNs was observed on P7Y8; however, in agreement with a previous report
(25), no differences were seen on P14Y15 (Figs. 6DYF). Similar
results were obtained when the numbers of VGLUT1- and
VGAT-immunoreactive boutons on SMN$7 MNs were counted
per individual MN soma (data not shown), indicating that the
observed changes in bouton density were not caused by variations in MN size.
The area of presynaptic boutons was measured between
P0Y1 and P14Y15. Although we found no changes in the size of
VGAT-positive terminals in mutant mice compared with WT
mice, there was a modest but significant reduction in the area
of VGLUT1-containing boutons in P14 to P15 SMN$7 spinal
cords (WT, 2.6 T 0.1 Km2; SMN$7, 2.2 T 0.1 Km2; p G 0.05,
Student t-test, n = 75 boutons per condition).
The involvement of microglia in afferent MN synaptic
pruning seen in electron microscopy studies was also assessed
by multiple immunofluorescence imaging. Immunostaining
of CD68 can be used to highlight active phagocytosis in
microglia (47). When anti-CD68 antibody was combined
with anti-VGLUT1 antibody, some presumably degenerating
glutamatergic nerve terminals were visualized engulfed
inside microglial cells in the spinal cord of SMN$7 mice
(Figs. 6GYJ).
Changes in MN synapses in SMN$7 mice were also
quantified according to their ultrastructural features (39)
(Figs. 7AYE). The analysis of the different types of synapses
corroborated the loss of symmetric terminals, which are presumably inhibitory (48). Although we also saw an apparent
reduction in asymmetric boutons apposed to SMN$7 MNs,
this change did not reach statistical significance, possibly
because of sampling variability. Interestingly, we observed a
significant increase in C-type boutons on MNs of mutant mice
(Fig. 7E). These are cholinergic afferent inputs specific for
>-MNs originating from a recently defined interneuronal group
near the central canal (49). The increase in C-type boutons on
MNs may be the consequence of the sprouting of this particular
class of nerve terminals to compensate for the loss of other
afferent synaptic inputs. A similar phenomenon has been described in a mouse model of ALS (50Y53).
Upregulation of nNOS in MNs and Increased
Phosphorylation of Myosin Light Chain in
Afferent Synapses on MNs of SMN$7 Mice
Nitric oxide plays a role in synaptic remodeling in adult
MNs by mediating synaptic loss from pathologic neurons
(35, 54). Thus, we investigated whether there was induction
of nNOS in MNs as a possible cause of the synaptic remodeling in SMN$7 mice.
Neuronal NO synthase was increased in MNs (identified with SMI-32 antibody) of SMN$7 mice from P4Y5 to
P14Y15 compared with age-matched WT animals (Figs. 8AYG).
Because NO activates the RhoA/ROCK pathway, which is
essential for neurite retraction (36, 55, 56), we hypothesized
that this pathway participates in the loss of synapses on spinal
cord MN somata in SMA. The final effector to induce actomyosin contraction and neurite retraction is P-MLC, a substrate
of ROCK (57). Therefore, we studied P-MLCYimmunoreactive
Mechanisms of Motoneuron Deafferentation in SMA
FIGURE 7. Changes in the ultrastructurally defined types of
afferent synaptic boutons apposed to the spinal cord MNs of
P7 SMN$7 mice. (AYD) Electron micrographs showing the
different morphologic types of synapses scored in this analysis.
Presynaptic structures are shown in red, whereas postsynaptic
areas are highlighted in yellow. (A) A typical asymmetric synaptic bouton with a clearly defined postsynaptic density (*).
(B) In symmetric terminals, there are no clear differences between presynaptic and postsynaptic densities. (C) A C-type
bouton defined by the presence of a subsynaptic cistern
(arrowed, blue) closely apposed to a postsynaptic membrane.
(D) Detail of a subsynaptic cistern (arrowed, blue) of a C-type
afferent synapse on an MN. (E) Quantification of the density of
axosomatic synaptic MN afferents into groups defined as follows: asymmetric (Asym), symmetric (Sym), C-terminals (Cterm), and unclassifiable (NC). Data are expressed as mean T SE
after scoring of the synaptic terminals (n = 157 in 33 MNs of
WT mice and n = 110 in 41 MNs of SMN$7 mice taken from
5 and 3 animals, respectively). * p G 0.05 and ** p G 0.01 versus
WT (Student t-test). Scale bars = (AYC) 200 nm; (D) 100 nm.
boutons colocalizing with the presynaptic marker Syphys
around MNs on P11 and found a decrease (È58%) in the linear
density of Syphys-positive boutons on MNs in parallel with a
dramatic increase (È71%) in the percentage of these boutons
showing immunoreactivity to P-MLC (Figs. 8HYQ).
DISCUSSION
In the present study, we report that MNs of SMN$7
mice undergo early and gradual degeneration that correlates
with the progression of the clinical disease phenotype. Although MN death is only moderate, surviving MNs display
cytologic changes indicative of cellular dysfunction, which
starts long before overt MN loss. Alterations in MNs include
significant defects in glutamatergic and GABAergic synaptic
afferents accompanied by marked astroglial and microglial
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
529
Tarabal et al
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
FIGURE 8. Neuronal NO synthase and P-MLC are upregulated in the lumbar spinal cord MNs of SMN$7 mice. (A) Percentage
increases in nNOS-immunoreactive MNs after double immunofluorescence with antiYSMI-32 and anti-nNOS antibodies in the
lumbar spinal cord sections of SMN$7 mice versus age-matched WT animals; 8 sections per animal and 4 fields per section of at
least 4 mice per each group were analyzed. (BYG) Representative immunofluorescence images of the lumbar spinal cord of P11 WT
(BYD) and SMN$7 (EYG) mice showing an increase in the number of nNOS (green)Ypositive MNs (immunostained with SMI-32
antibody; red) in mutant animals (compare C with F). (HYO) Representative immunofluorescence images of the lumbar spinal cord
of P11 WT (HYK) and SMN$7 (LYO) mice showing colocalization of Syphys (red) and P-MLC (green) around MNs (stained with
fluorescent Nissl stain; blue). (P) Length (in micrometers) of Syphys-positive boutons per 100 Km of MN soma membrane in the
lumbar spinal cord of P11 WT and SMN$7 mice. (Q) Percentage of Syphys-immunoreactive synaptic boutons apposed to MN
somata showing positive P-MLC immunolabeling in the lumbar spinal cord of P11 WT and SMN$7 animals. Data in all graphs are
expressed as mean T SE. ** p G 0.01 and *** p G 0.001 versus age-matched WT mice (one-way ANOVA and Bonferroni test). Fifty
MNs from 4 to 5 mice per group were examined. Scale bars = (G; valid for BYF) 20 Km; (O; valid for HYN) 10 Km.
530
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
Mechanisms of Motoneuron Deafferentation in SMA
reactions in the lumbar spinal ventral horn. Dendrites and
microglial cells were found to participate in the phagocytosis
and clearance of degenerating afferent boutons. Concomitant
with these alterations, nNOS was upregulated in MNs together
with the activated RhoA/ROCK pathway, as deduced from the
increase in P-MLC expression in synaptic afferents on MNs.
These results suggest the involvement of nNOS and the RhoA/
ROCK pathway in the central synaptic loss occurring in SMA.
Overall, our observations suggest that the earliest
(E18YP1) change occurring in the degenerative process in
SMA is the loss of excitatory glutamatergic synaptic inputs
to MNs. Reduced excitability may sensitize MNs to undergo
degeneration and death, as has been demonstrated in the
context of both ALS (51) and SMA (58). Motoneuron death
starts between P0 and P4, concomitant with an astroglial
reaction and increased MN NO expression. Later, as MN
death progresses, there occurs a microglial reaction that is
involved in both the clearance of dying MNs and the removal
of degenerating synapses.
suggest that even surviving SMN-deficient MNs are dysfunctional. This could explain, in part, the substantial degree of
neuromuscular dysfunction found in diseased animals, especially in advanced disease stages, despite modest cell death
(20%Y25% of initial MN populations in the lumbar spinal
cord). In agreement with our results, the extent of MN loss has
been shown to be limited in other mouse models of SMA (16,
63). The independence of clinical signs from the activation of
MN cell death pathways has also been reported in the ALS
mouse model (52, 53, 64). Our observation of increased numbers of pyknotic cells (presumptive MNs) at end stages in
SMN$7 mice suggests that MN loss is still an ongoing process
when animals die as a consequence of the disease.
Progressive MN Degeneration in the
Spinal Cord of SMN$7 Mice
Progressive MN death has been considered a pathologic
hallmark of human SMA. In agreement with previous studies
in which different mouse models of SMA have been used
(6, 12, 13, 16, 59Y61), we also found only moderate (although
significant) MN loss in the spinal cord of SMN$7 mice. The
reduction in the number of apparently healthy MNs in mutant
animals begins between P0Y1 and P4Y5, coinciding with the
appearance of the initial phenotypic changes (13), and is more
conspicuous at the end stage of the disease (P14YP15; Fig. 1A).
This time course of MN degeneration in SMN$7 mice contrasts with that reported by others. Thus, Le et al (6) showed
some loss of MNs in the lumbar spinal cord on P9, but not on
P4. Kariya et al (15) reported a significant reduction in the
number of axons in the phrenic nerves and cervical ventral
roots of SMN$7 animals, but loss of neither MNs nor motor
axons in ventral roots was observed in the lumbar spinal cord
even at the end stage (P14) of the disease. A possible explanation for these discrepancies could be the differences in the
spinal cord segments used for MN count: Whereas the whole
lumbar spinal cord was included in cell counting in our study,
the abovementioned reports analyzed the loss of MNs and
motor axons in the ventral roots of restricted spinal cord segments (i.e. L4YL5). Indeed, we found a substantial number of
degenerating MNs in the most rostral segments of the lumbar
spinal cord of P4Y5 and P14Y15 SMN$7 mice (unpublished
observations). Indeed, a rostrocaudal gradient of MN loss has
been described in the spinal cord of mice with SMA, with MNs
located at the rostral levels being more susceptible to degeneration (15, 33). We also observed that MN degeneration in
affected animals was accompanied by accumulation of phosphorylated neurofilament (evidenced with SMI-31 antibody)
and reduction in the soma size of apparently healthy cells. Interestingly, an increase in phosphorylated neurofilament in
MNs has been shown in several neurodegenerative disorders,
including the adult-onset MN disease ALS (40, 62). Overall,
the histopathologic alterations that we found in mutant mice
Astroglial and Microglial Activation in the
Spinal Cord of SMN$7 Mice
Neuroinflammation is a common process in a variety
of neurodegenerative disorders that involves the activation
of microglia, astrocytes, and, in some cases, T lymphocytes.
Whereas neuroinflammation has been widely studied in ALS
(20, 65), data on neuroinflammation in SMA are very scarce
(27). Severe gliosis has been described in the anterior horns
and ventral nerve roots of human SMA spinal cord tissue
(66Y69), and we observed increased numbers of astroglial and
microglial cells surrounding MNs in the spinal cord of SMN$7
mice. It has traditionally been thought that neuroinflammation
is a secondary cellular event arising in response to neuronal
loss. Glial reactions may have both beneficial and deleterious
consequences on MNs, that is, gliosis sometimes acts to limit
cell damage and to promote repair but astrocytes and microglia
can initiate neurodegeneration under certain circumstances
(24, 27, 65, 70, 71). Our results are in agreement with those
reported by Ling et al (25), who showed microgliosis around
MNs in SMN$7 mice. We further show that MN loss in
diseased animals (which is already significant on P4YP5)
precedes an overt microglial reaction, suggesting that microglial cells are not active players in MN death induction. It
is more likely that microgliosis is a reactive process resulting
from peripheral neuromuscular disconnection (i.e. synaptic
stripping after peripheral nerve injury [72]) and/or other
forms of MN damage leading to cell death, similar to those
occurring in other models of MN degeneration (73). In fact,
we observe an absence of microglial reaction around spinal
cord MNs in Smnj/j;SMN2+/+ mice (a model of a more
severe form of SMA) on P5Y6, an age corresponding to the
end stage of the disease and in which MN loss is maximal
(12). In contrast, we have previously described astrogliosis
in early stages of SMA, before the occurrence of a significant
reduction in the number of MNs in Smnj/j;SMN2+/+ mice
(12). This observation is consistent with our present data on
SMN$7 mice. While the present article was in preparation,
astrocytic changes were reported in vivo in SMN$7 mice
and in vitro in SMA-induced pluripotent stem cells (26);
McGivern et al (26) also found that SMA-induced pluripotent stem cellYderived astrocytes display higher basal calcium levels and increased calcium response to ATP. Overall,
these data suggest that dysfunctional astrocytes could be
important contributors to SMA pathogenesis by inducing
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
531
Tarabal et al
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
and/or aggravating MN pathology. For example, we have
previously shown that reactive astrocytes in the spinal cord
of SMN$7 mice display an increased expression of the main
components of the Notch signaling pathway, including the
Notch ligand Jagged 1, leading to increased Notch signaling
on adjacent MNs (74). Interestingly, the expression of Jagged 1
on astrocytes in an inflammatory environment has been reported to be increased (75). Indeed, astrocytes seem to play an
important role in MN degeneration in ALS. Impaired functioning of astrocytic excitatory amino acid transporters in ALS
results in excitotoxicity and neuronal cell death (76). In addition to perturbation in glutamate handling, astrocytes in ALS
exert direct toxic effects on MNs through the release of still
unidentified factors (77Y79). Although these mechanisms remain to be explored in SMA, disruption of neurotrophic signaling mediated by the astrocytic secreted glial cell line
Yderived neurotrophic factor has been reported (26). Astrocytes are also endogenous regulators of basal transmission at
central synapses (80), through their ability to modulate intracellular calciumYdependent processes, and may play a role in
network changes by influencing the retraction of synaptic
boutons and by providing a proper perisynaptic environment
(28, 81). Astrocytes also play an active role in promoting
synaptic differentiation and activity (82). It will be interesting
to examine whether astrocytic dysfunction in SMA affects their
ability to promote synaptic activity, which seems to be crucial
to the regulation of synaptic stability/elimination (83).
size and messenger RNA content (91). However, the contribution of MN deafferentation to SMA pathogenesis remains controversial. In fact, different studies have reported that the defects
in spinal cord synaptic connectivity observed in SMA could be
considered a consequence of primary MN pathology (92Y94).
Via electron microscopy, we observed a reduction in
afferent synaptic boutons on SMA MNs. Except for C-type
boutons, the different types of synapses studied showed decreased numbers in relation to MN somatic membrane length;
however, because of sampling variability, the results were
statistically significant only for symmetric and C-type synapses. The C-type boutons are cholinergic synapses derived
from interneurons located near the central canal (49). Because
these interneurons have short axons, their terminals should be
less vulnerable to defects in axonal transport and dying-back
degeneration, possibly explaining why they are not disrupted
in SMA MNs. The loss of other types of synaptic boutons
may create a vacant territory on MNs that is occupied by the
sprouting of cholinergic premotor interneurons, leading to
increased numbers of C-terminals.
One of the most conspicuous features observed in degenerating terminal synaptic boutons and contacting dendrites
was the accumulation of autophagic vacuoles, as reported in
other forms of neuronal injury (95). Autophagic vacuoles were
also occasionally seen in WT dendrites, presumably reflecting
a postnatal remodeling of neural processes. Detailed examination of the ultrastructural morphology of synaptic degeneration
in SMA revealed a novel mechanism involving neuronal
(dendritic) phagocytosis of synapses, suggesting that degenerating afferent nerve terminals enriched with autophagosomes
are engulfed by adjacent dendrites, which in turn appear to be
incorporated into the MN cytoplasm. We observed a similar
process in WT animals but to a much lesser extent, suggesting
that synapse elimination in spinal cord MNs is an example of
neuronal phagocytosis during normal postnatal synapse elimination that is exacerbated in SMA. Although the phagocytic
capacity of neurons is not usually considered, it is a recognized
phenomenon (96), but it has not previously been related to
synaptic elimination. These findings do not preclude the relevance of microglial cells to the clearance of cellular debris
during synapse degeneration/elimination. Indeed, we detected
active (amoeboid) microglial cells enwrapping structural complexes that include degenerating presynaptic terminals and
postsynaptic dendrites (Fig. 2E).
Loss of Afferent Synapses on Spinal Cord
MNs in SMN$7 Mice
We observed a gradual reduction in the density of central
glutamatergic and GABAergic synapses on spinal cord MNs
during normal postnatal development that could reflect naturally occurring synaptic remodeling; this requires further detailed studies. In support of this regressive event, we observed
degenerating synaptic boutons in WT animals via electron microscopy. This phenomenon is exacerbated in SMA for both
glutamatergic and, to a lesser extent, GABAergic nerve afferents. The increased loss of glutamatergic terminals in SMN$7
spinal cords is in agreement with previous studies (13, 25,
33, 34). In contrast to some reports (25) but in accordance
with Mentis et al (33), we show that the loss of central synapses
in SMA precedes MN death. Indeed, we found that glutamatergic deafferentation started at prenatal stages (E18), long
before MN cell death was first detected (P4Y5). SMN-depleted
PC-12 cells exhibit impaired neuritogenesis after never growth
factor differentiation (84), suggesting that impaired axonal
growth cone migration may be involved in synaptic loss at
prenatal stages. The loss of central synapses is a common
process in neurodegenerative diseases (85) that occurs before
MN death in ALS (52, 53, 86Y88) and has also been reported in
patients with SMA (89, 90). In addition to distal peripheral
motor axons, it is conceivable that, after SMN deficiency, dying back may affect other neuronal types, leading to a broad
synaptopathy involving large axons projecting to MNs. Indeed,
reduced proprioceptive reflexes and synapses have been observed early in mice with SMA (33), and growth cones from
SMA dorsal root ganglion sensory neurons display reduced
532
Increased nNOS in MNs as a Trigger of RhoA
Pathway Activation and Synaptic
Bouton Retraction
In addition to exacerbated neuronal phagocytosis in
SMA, the loss of glutamatergic and GABAergic synapses on
MNs could be the result of the induction of nNOS in MNs. In
this regard, it has been reported that NO is ‘‘necessary’’ and
‘‘sufficient’’ to induce the detachment of synaptic afferents
from MNs in response to physical injury to motor nerves
and in ALS (54). Moreover, it has been shown that nNOS
upregulation after nerve injury triggers the loss of excitatory
synaptic inputs to injured hypoglossal MNs (35, 97). We
report here that nNOS is upregulated in SMN$7 MNs,
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
Mechanisms of Motoneuron Deafferentation in SMA
suggesting a role in the withdrawal of synaptic afferents found
in our model. It has been proposed that, after motor nerve
injury, NO produced via upregulation of nNOS would lead to
activation of the RhoA/ROCK pathway (36). Interestingly,
this pathway has been shown to be abnormally increased in
the SMA mouse model, and treating SMA mice with a ROCK
inhibitor led to improved maturation of the neuromuscular
junction and dramatically increased the life span of animals
(98). Phosphorylated myosin light chain is the final effector of
the RhoA/ROCK pathway and is known to trigger actomyosin
contraction and neurite retraction (55, 56), thus leading to
synaptic detachment/synaptic bouton retraction (36). Therefore, we propose that the NO released from SMA MNs could
trigger the observed increase in the phosphorylation of synaptic myosin light chain around these cells, contributing to
afferent synapse loss.
13. Dachs E, Piedrafita L, Hereu M, et al. Chronic treatment with lithium
does not improve neuromuscular phenotype in a mouse model of severe
spinal muscular atrophy. Neuroscience 2013;250:417Y33
14. Ito Y, Shibata N, Saito K, et al. New insights into the pathogenesis of
spinal muscular atrophy. Brain Dev 2011;33:321Y31
15. Kariya S, Park G-H, Maeno-Hikichi Y, et al. Reduced SMN protein
impairs maturation of the neuromuscular junctions in mouse models of
spinal muscular atrophy. Hum Mol Genet 2008;17:2552Y69
16. Monani UR, Sendtner M, Coovert DD, et al. The human centromeric
survival motor neuron gene (SMN2) rescues embryonic lethality in
Smn(j/j) mice and results in a mouse with spinal muscular atrophy.
Hum Mol Genet 2000;9:333Y39
17. Kong L, Wang X, Choe DW, et al. Impaired synaptic vesicle release and
immaturity of neuromuscular junctions in spinal muscular atrophy mice.
J Neurosci 2009;29:842Y51
18. Pekny M, Nilsson M. Astrocyte activation and reactive gliosis. Glia
2005;50:427Y34
19. Eikelenboom P, Bate C, Van Gool WA, et al. Neuroinflammation in
Alzheimer’s disease and prion disease. Glia 2002;40:232Y39
20. Evans MC, Couch Y, Sibson N, et al. Inflammation and neurovascular
changes in amyotrophic lateral sclerosis. Mol Cell Neurosci 2013;53:
34Y41
21. Frank-Cannon TC, Alto LT, McAlpine FE, et al. Does neuroinflammation fan the flame in neurodegenerative diseases? Mol Neurodegener
2009;4:47
22. McGeer PL, McGeer EG. Inflammatory processes in amyotrophic lateral
sclerosis. Muscle Nerve 2002;26:459Y70
23. Moisse K, Strong MJ. Innate immunity in amyotrophic lateral sclerosis.
Biochim Biophys Acta 2006;1762:1083Y93
24. Sargsyan SA, Monk PN, Shaw PJ. Microglia as potential contributors
to motor neuron injury in amyotrophic lateral sclerosis. Glia 2005;51:
241Y53
25. Ling KKY, Lin M-Y, Zingg B, et al. Synaptic defects in the spinal and
neuromuscular circuitry in a mouse model of spinal muscular atrophy.
PLoS One 2010;5:e15457
26. McGivern JV, Patitucci TN, Nord JA, et al. Spinal muscular atrophy
astrocytes exhibit abnormal calcium regulation and reduced growth factor
production. Glia 2013;61:1418Y28
27. Papadimitriou D, Le Verche V, Jacquier A, et al. Inflammation in ALS
and SMA: Sorting out the good from the evil. Neurobiol Dis 2010;37:
493Y502
28. Eroglu C, Barres BA. Regulation of synaptic connectivity by glia. Nature
2010;468:223Y31
29. Cifuentes-Diaz C, Nicole S, Velasco ME, et al. Neurofilament accumulation at the motor endplate and lack of axonal sprouting in a spinal
muscular atrophy mouse model. Hum Mol Genet 2002;11:1439Y47
30. Ling KK, Gibbs RM, Feng Z, et al. Severe neuromuscular denervation of
clinically relevant muscles in a mouse model of spinal muscular atrophy.
Hum Mol Genet 2012;21:185Y95
31. Murray LM, Talbot K, Gillingwater TH. Review: Neuromuscular synaptic vulnerability in motor neurone disease: Amyotrophic lateral sclerosis and spinal muscular atrophy. Neuropathol Appl Neurobiol 2010;36:
133Y56
32. Torres-Benito L, Neher MF, Cano R, et al. SMN requirement for synaptic
vesicle, active zone and microtubule postnatal organization in motor
nerve terminals. PLoS One 2011;6:e26164
33. Mentis GZ, Blivis D, Liu W, et al. Early functional impairment of
sensory-motor connectivity in a mouse model of spinal muscular atrophy.
Neuron 2011;69:453Y67
34. Park G-H, Maeno-Hikichi Y, Awano T, et al. Reduced survival of motor
neuron (SMN) protein in motor neuronal progenitors functions cell autonomously to cause spinal muscular atrophy in model mice expressing
the human centromeric (SMN2) gene. J Neurosci 2010;30:12005Y19
35. Sunico CR, Portillo F, Gonzalez-Forero D, et al. Nitric-oxideYdirected synaptic remodeling in the adult mammal CNS. J Neurosci 2005;25:1448Y58
36. Sunico CR, Gonzalez-Forero D, Dominguez G, et al. Nitric oxide induces
pathological synapse loss by a protein kinase GY, Rho kinaseYdependent
mechanism preceded by myosin light chain phosphorylation. J Neurosci
2010;30:973Y84
37. Calderó J, Ciutat D, Lladó J, et al. Effects of excitatory amino acids on
neuromuscular development in the chick embryo. J Comp Neurol 1997;
387:73Y95
ACKNOWLEDGMENTS
We thank Dr Ronald W. Oppenheim and Dr Carol
Milligan for critical reading of the manuscript and for helpful
comments and suggestions. We also thank Dr Elisabet Dachs,
Dr Lucı́a Tabares, and Dr Laura Torres-Benito for their
support in some parts of this work; Dr Anna Casanovas for
useful discussions; Lı́dia Piedrafita, Marta Hereu, and Montse
Ortega for their technical assistance; and Neus Montull,
Clàudia Cerveró, and Maria Calderó for their help with
some experiments in this study.
REFERENCES
1. Crawford TO, Pardo CA. The neurobiology of childhood spinal muscular
atrophy. Neurobiol Dis 1996;3:97Y110
2. Emery AE. Population frequencies of inherited neuromuscular diseases
Va world survey. Neuromuscul Disord 1991;1:19Y29
3. Lefebvre S, Bürglen L, Reboullet S, et al. Identification and characterization of a spinal muscular atrophyYdetermining gene. Cell 1995;80:
155Y65
4. Pellizzoni L, Kataoka N, Charroux B, et al. A novel function for SMN,
the spinal muscular atrophy disease gene product, in pre-mRNA splicing.
Cell 1998;95:615Y24
5. Gennarelli M, Lucarelli M, Capon F, et al. Survival motor neuron gene
transcript analysis in muscles from spinal muscular atrophy patients.
Biochem Biophys Res Commun 1995;213:342Y48
6. Le TT, Pham LT, Butchbach MER, et al. SMNDelta7, the major product
of the centromeric survival motor neuron (SMN2) gene, extends survival
in mice with spinal muscular atrophy and associates with full-length
SMN. Hum Mol Genet 2005;14:845Y57
7. Schrank B, Gotz R, Gunnersen JM, et al. Inactivation of the survival
motor neuron gene, a candidate gene for human spinal muscular atrophy,
leads to massive cell death in early mouse embryos. Proc Natl Acad Sci U
S A 1997;94:9920Y25
8. Avila AM, Burnett BG, Taye AA, et al. Trichostatin A increases SMN
expression and survival in a mouse model of spinal muscular atrophy.
J Clin Invest 2007;117:659Y71
9. Murray LM, Comley LH, Thomson D, et al. Selective vulnerability of
motor neurons and dissociation of pre- and post-synaptic pathology at the
neuromuscular junction in mouse models of spinal muscular atrophy.
Hum Mol Genet 2008;17:949Y62
10. Lotti F, Imlach WL, Saieva L, et al. An SMN-dependent U12 splicing
event essential for motor circuit function. Cell 2012;151:440Y54
11. Byers RK, Banker BQ. Infantile muscular atrophy. Arch Neurol 1961;5:
140Y64
12. Dachs E, Hereu M, Piedrafita L, et al. Defective neuromuscular junction
organization and postnatal myogenesis in mice with severe spinal muscular atrophy. J Neuropathol Exp Neurol 2011;70:444Y61
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
533
Tarabal et al
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
38. Clarke PG, Oppenheim RW. Neuron death in vertebrate development: In
vitro methods. Methods Cell Biol 1995;46:277Y321
39. Conradi S, Kellerth JO, Berthold CH. Electron microscopic studies of
serially sectioned cat spinal alpha-motoneurons. II. A method for the
description of architecture and synaptology of the cell body and proximal
dendritic segments. J Comp Neurol 1979;184:741Y54
40. Ackerley S, Grierson AJ, Banner S, et al. p38alpha stress-activated protein kinase phosphorylates neurofilaments and is associated with neurofilament pathology in amyotrophic lateral sclerosis. Mol Cell Neurosci
2004;26:354Y64
41. Olney JW, Ishimaru MJ. Excitotoxic cell death. In: Koliatsos VE, Ratan
RR, eds. Cell Death and Diseases of the Nervous System. Totowa, NJ:
Humana Press Inc, 1999:197Y219
42. Fidzianska A, Rafalowska J. Motoneuron death in normal and spinal
muscular atrophyYaffected human fetuses. Acta Neuropathol 2002;104:
363Y68
43. Brunet N, Tarabal O, Portero-Otin M, et al. Survival and death of mature
avian motoneurons in organotypic slice culture: Trophic requirements for
survival and different types of degeneration. J Comp Neurol 2007;501:
669Y90
44. Chiu IM, Phatnani H, Kuligowski M, et al. Activation of innate and
humoral immunity in the peripheral nervous system of ALS transgenic
mice. Proc Natl Acad Sci U S A 2009;106:20960Y65
45. Oliveira AL, Hydling F, Olsson E, et al. Cellular localization of three
vesicular glutamate transporter mRNAs and proteins in rat spinal cord
and dorsal root ganglia. Synapse 2003;50:117Y29
46. Lawson SJ, Davies HJ, Bennett JP, et al. Evidence that spinal interneurons undergo programmed cell death postnatally in the rat. Eur J
Neurosci 1997;9:794Y99
47. Peviani M, Salvaneschi E, Bontempi L, et al. Neuroprotective effects of
the Sigma-1 receptor (S1R) agonist PRE-084, in a mouse model of motor
neuron disease not linked to SOD1 mutation. Neurobiol Dis 2014;62C:
218Y32
48. Bodian D. Electron microscopy: Two major synaptic types on spinal
motoneurons. Science 1966;151:1093Y94
49. Zagoraiou L, Akay T, Martin JF, et al. A cluster of cholinergic premotor
interneurons modulates mouse locomotor activity. Neuron 2009;64:
645Y62
50. Pullen AH, Athanasiou D. Increase in presynaptic territory of C-terminals
on lumbar motoneurons of G93A SOD1 mice during disease progression.
Eur J Neurosci 2009;29:551Y61
51. Saxena S, Roselli F, Singh K, et al. Neuroprotection through excitability
and mTOR required in ALS motoneurons to delay disease and extend
survival. Neuron 2013;80:80Y96
52. Vinsant S, Mansfield C, Jimenez-Moreno R, et al. Characterization of
early pathogenesis in the SOD1 G93A mouse model of ALS. Part I.
Background and methods. Brain Behav 2013;3:335Y50
53. Vinsant S, Mansfield C, Jimenez-Moreno R, et al. Characterization of
early pathogenesis in the SOD1 G93A mouse model of ALS. Part II.
Results and discussion. Brain Behav 2013;3:431Y57
54. Moreno-Lopez B, Sunico CR, Gonzalez-Forero D. NO orchestrates the
loss of synaptic boutons from adult ‘‘sick’’ motoneurons: Modeling a
molecular mechanism. Mol Neurobiol 2011;43:41Y66
55. Luo L. Rho GTPases in neuronal morphogenesis. Nat Rev Neurosci
2000;1:173Y80
56. Luo L. Actin cytoskeleton regulation in neuronal morphogenesis and
structural plasticity. Annu Rev Cell Dev Biol 2002;18:601Y35
57. Amano M, Ito M, Kimura K, et al. Phosphorylation and activation of
myosin by Rho-associated kinase (Rho-kinase). J Biol Chem 1996;271:
20246Y49
58. Imlach WL, Beck ES, Choi BJ, et al. SMN is required for sensory-motor
circuit function in Drosophila. Cell 2012;151:427Y39
59. Bowerman M, Anderson CL, Beauvais A, et al. SMN, profilin IIa and
plastin 3: A link between the deregulation of actin dynamics and SMA
pathogenesis. Mol Cell Neurosci 2009;42:66Y74
60. Grondard C, Biondi O, Armand AS, et al. Regular exercise prolongs
survival in a type 2 spinal muscular atrophy model mouse. J Neurosci
2005;25:7615Y22
61. Jablonka S, Schrank B, Kralewski M, et al. Reduced survival motor
neuron (Smn) gene dose in mice leads to motor neuron degeneration: An
animal model for spinal muscular atrophy type III. Hum Mol Genet 2000;
9:341Y46
62. Hirano A, Nakano I, Kurland LT, et al. Fine structural study of neurofibrillary changes in a family with amyotrophic lateral sclerosis.
J Neuropathol Exp Neurol 1984;43:471Y80
63. Frugier T, Tiziano FD, Cifuentes-Diaz C, et al. Nuclear targeting defect
of SMN lacking the C-terminus in a mouse model of spinal muscular
atrophy. Hum Mol Genet 2000;9:849Y58
64. Gould TW, Buss RR, Vinsant S, et al. Complete dissociation of motor
neuron death from motor dysfunction by Bax deletion in a mouse model
of ALS. J Neurosci 2006;26:8774Y86
65. Philips T, Robberecht W. Neuroinflammation in amyotrophic lateral
sclerosis: Role of glial activation in motor neuron disease. Lancet Neurol
2011;10:253Y63
66. Chou SM. Glial bundles of nerve roots in Werdnig-Hoffmann disease.
Ann Neurol 1980;8:79Y82
67. Ghatak NR. Glial bundles in spinal nerve roots: A form of isomorphic
gliosis at the junction of the central and peripheral nervous system.
Neuropathol Appl Neurobiol 1983;9:391Y401
68. Kumagai T, Hashizume Y. Morphological and morphometric studies
on the spinal cord lesion in Werdnig-Hoffmann disease. Brain Dev 1982;
4:87Y96
69. Kuru S, Sakai M, Konagaya M, et al. An autopsy case of spinal muscular
atrophy type III (Kugelberg-Welander disease). Neuropathology 2009;
29:63Y67
70. Kreutzberg GW. Microglia: A sensor for pathological events in the CNS.
Trends Neurosci 1996;19:312Y18
71. Streit WJ, Walter SA, Pennell NA. Reactive microgliosis. Prog Neurobiol
1999;57:563Y81
72. Streit WJ, Graeber MB, Kreutzberg GW. Functional plasticity of microglia: A review. Glia 1988;1:301Y7
73. Calderó J, Brunet N, Ciutat D, et al. Development of microglia in the
chick embryo spinal cord: Implications in the regulation of motoneuronal
survival and death. J Neurosci Res 2009;87:2447Y66
74. Caraballo-Miralles V, Cardona-Rossinyol A, Garcera A, et al. Notch
signaling pathway is activated in motoneurons of spinal muscular atrophy. Int J Mol Sci 2013;14:11424Y37
75. Morga E, Mouad-Amazzal L, Felten P, et al. Jagged1 regulates the activation of astrocytes via modulation of NFkappaB and JAK/STAT/SOCS
pathways. Glia 2009;57:1741Y53
76. Rothstein JD, Dykes-Hoberg M, Pardo CA, et al. Knockout of glutamate
transporters reveals a major role for astroglial transport in excitotoxicity
and clearance of glutamate. Neuron 1996;16:675Y86
77. Di Giorgio FP, Carrasco MA, Siao MC, et al. Non-cell autonomous effect
of glia on motor neurons in an embryonic stem cellYbased ALS model.
Nat Neurosci 2007;10:608Y14
78. Haidet-Phillips AM, Hester ME, Miranda CJ, et al. Astrocytes from
familial and sporadic ALS patients are toxic to motor neurons. Nat
Biotechnol 2011;29:824Y28
79. Nagai M, Re DB, Nagata T, et al. Astrocytes expressing ALS-linked
mutated SOD1 release factors selectively toxic to motor neurons. Nat
Neurosci 2007;10:615Y22
80. Panatier A, Vallee J, Haber M, et al. Astrocytes are endogenous regulators of basal transmission at central synapses. Cell 2011;146:785Y98
81. Emirandetti A, Graciele Zanon R, Sabha M Jr, et al. Astrocyte reactivity
influences the number of presynaptic terminals apposed to spinal motoneurons after axotomy. Brain Res 2006;1095:35Y42
82. Ullian EM, Sapperstein SK, Christopherson KS, et al. Control of synapse
number by glia. Science 2001;291:657Y61
83. Purves D, Lichtman JW. Elimination of synapses in the developing nervous system. Science 1980;210:153Y57
84. Bowerman M, Shafey D, Kothary R. Smn depletion alters profilin II
expression and leads to upregulation of the RhoA/ROCK pathway and
defects in neuronal integrity. J Mol Neurosci 2007;32:120Y31
85. Selkoe DJ. Alzheimer’s disease is a synaptic failure. Science 2002;298:
789Y91
86. Ince PG, Slade J, Chinnery RM, et al. Quantitative study of synaptophysin
immunoreactivity of cerebral cortex and spinal cord in motor neuron disease. J Neuropathol Exp Neurol 1995;54:673Y79
87. Sasaki S, Maruyama S. Synapse loss in anterior horn neurons in
amyotrophic lateral sclerosis. Acta Neuropathol 1994;88:222Y27
88. Wishart TM, Parson SH, Gillingwater TH. Synaptic vulnerability in
neurodegenerative disease. J Neuropathol Exp Neurol 2006;65:733Y39
534
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 6, June 2014
Mechanisms of Motoneuron Deafferentation in SMA
89. Ikemoto A, Hirano A, Matsumoto S, et al. Synaptophysin expression
in the anterior horn of Werdnig-Hoffmann disease. J Neurol Sci 1996;
136:94Y100
90. Yamanouchi Y, Yamanouchi H, Becker LE. Synaptic alterations of anterior
horn cells in Werdnig-Hoffmann disease. Pediatr Neurol 1996;15:32Y35
91. Jablonka S, Karle K, Sandner B, et al. Distinct and overlapping alterations in motor and sensory neurons in a mouse model of spinal muscular
atrophy. Hum Mol Genet 2006;15:511Y18
92. Gogliotti RG, Quinlan KA, Barlow CB, et al. Motor neuron rescue in
spinal muscular atrophy mice demonstrates that sensory-motor defects
are a consequence, not a cause, of motor neuron dysfunction. J Neurosci
2012;32:3818Y29
93. Martinez TL, Kong L, Wang X, et al. Survival motor neuron protein in
motor neurons determines synaptic integrity in spinal muscular atrophy.
J Neurosci 2012;32:8703Y15
94. Thirumalai V, Behrend RM, Birineni S, et al. Preservation of VGLUT1
synapses on ventral calbindin-immunoreactive interneurons and normal
locomotor function in a mouse model of spinal muscular atrophy.
J Neurophysiol 2013;109:702Y10
95. Ruan YW, Han XJ, Shi ZS, et al. Remodeling of synapses in the CA1
area of the hippocampus after transient global ischemia. Neuroscience
2012;218:268Y77
96. Bowen S, Ateh DD, Deinhardt K, et al. The phagocytic capacity of
neurones. Eur J Neurosci 2007;25:2947Y55
97. Sumner BE. A quantitative analysis of boutons with different types of
synapse in normal and injured hypoglossal nuclei. Exp Neurol 1975;49:
406Y17
98. Bowerman M, Beauvais A, Anderson CL, et al. Rho-kinase inactivation
prolongs survival of an intermediate SMA mouse model. Hum Mol Genet
2010;19:1468Y78
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
535