Analysis of striatal transcriptome in mice overexpressing human wild-type alpha-synuclein supports synaptic dysfunction and suggests mechanisms of neuroprotection for striatal neurons.

Abstract

Alpha synuclein (SNCA) has been linked to neurodegenerative diseases (synucleinopathies) that include Parkinson's disease (PD). Although the primary neurodegeneration in PD involves nigrostriatal dopaminergic neurons, more extensive yet regionally selective neurodegeneration is observed in other synucleinopathies. Furthermore, SNCA is ubiquitously expressed in neurons and numerous neuronal systems are dysfunctional in PD. Therefore it is of interest to understand how overexpression of SNCA affects neuronal function in regions not directly targeted for neurodegeneration in PD.
The present study investigated the consequences of SNCA overexpression on cellular processes and functions in the striatum of mice overexpressing wild-type, human SNCA under the Thy1 promoter (Thy1-aSyn mice) by transcriptome analysis. The analysis revealed alterations in multiple biological processes in the striatum of Thy1-aSyn mice, including synaptic plasticity, signaling, transcription, apoptosis, and neurogenesis.
The results support a key role for SNCA in synaptic function and revealed an apoptotic signature in Thy1-aSyn mice, which together with specific alterations of neuroprotective genes suggest the activation of adaptive compensatory mechanisms that may protect striatal neurons in conditions of neuronal overexpression of SNCA.

Full text

Page 1

RESEARCH ARTICLEOpen Access
Analysis of striatal transcriptome in mice
overexpressing human wild-type alpha-synuclein
supports synaptic dysfunction and suggests
mechanisms of neuroprotection for striatal
neurons
Yofre Cabeza-Arvelaiz1*, Sheila M Fleming2, Franziska Richter2, Eliezer Masliah3, Marie-Francoise Chesselet2and
Robert H Schiestl1
Abstract
Background: Alpha synuclein (SNCA) has been linked to neurodegenerative diseases (synucleinopathies) that
include Parkinson’s disease (PD). Although the primary neurodegeneration in PD involves nigrostriatal
dopaminergic neurons, more extensive yet regionally selective neurodegeneration is observed in other
synucleinopathies. Furthermore, SNCA is ubiquitously expressed in neurons and numerous neuronal systems are
dysfunctional in PD. Therefore it is of interest to understand how overexpression of SNCA affects neuronal function
in regions not directly targeted for neurodegeneration in PD.
Results: The present study investigated the consequences of SNCA overexpression on cellular processes and
functions in the striatum of mice overexpressing wild-type, human SNCA under the Thy1 promoter (Thy1-aSyn
mice) by transcriptome analysis. The analysis revealed alterations in multiple biological processes in the striatum of
Thy1-aSyn mice, including synaptic plasticity, signaling, transcription, apoptosis, and neurogenesis.
Conclusion: The results support a key role for SNCA in synaptic function and revealed an apoptotic signature in Thy1-
aSyn mice, which together with specific alterations of neuroprotective genes suggest the activation of adaptive
compensatory mechanisms that may protect striatal neurons in conditions of neuronal overexpression of SNCA.
Keywords: α-synuclein, apoptosis, neuroprotection, Parkinson’s disease, Alzheimer’s disease, synaptic plasticity, vesi-
cle release, diabetes
Background
Abnormal accumulation of the pre-synaptic protein a-
synuclein (SNCA) is a hallmark of several neurodegen-
erative disorders including the second most frequent
neurodegenerative disease Parkinson’s disease (PD) [1].
Neurodegeneration in PD is predominant in the sub-
stantia nigra pars compacta (SNc), but cell loss and
Lewy Body (LB) formation also occur in other brain and
peripheral tissues [2]. Familial forms of PD have been
linked to mutations in SNCA, and also to multiplications
of the locus encompassing the SNCA gene, which lead
to increased levels of SNCA expression indicating that
the wild-type (wt) protein can be pathogenic if produced
in excess [3]. Furthermore, Genome Wide Association
Studies (GWAS) have consistently identified the SNCA
gene as most associated with PD risk [4]. In most synu-
cleinopathies, SNCA aggregates form in neurons [5].
Transgenic (tg) murine models expressing human SNCA
under neuronal promoters reproduce some phenotypic
features of PD such as inclusion formation, motor and
* Correspondence: yofrecabeza@gmail.com
1Department of Pathology and Environmental Health Sciences, The Geffen
School of Medicine and School of Public Health, University of California, Los
Angeles, 650 Charles E. Young Dr. S, CHS 71-295; Los Angeles, CA 90095,
USA
Full list of author information is available at the end of the article
Cabeza-Arvelaiz et al. Molecular Neurodegeneration 2011, 6:83
http://www.molecularneurodegeneration.com/content/6/1/83
© 2011 Cabeza-Arvelaiz et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.

Page 2

non-motor impairments, loss of striatal dopamine (DA),
and, in a few tg lines, nigrostriatal degeneration (For
review see: [6,7]). Mice expressing human wild-type
SNCA under the Thy1 promoter (Thy1-aSyn mice)
express high level of mRNA and protein in neurons
throughout the brain [8] and develop proteinase K-resis-
tant SNCA aggregates [9,10]. These mice show a 40%
loss of DA in the striatum by 14 months of age [11].
We have shown that these mice display early and pro-
gressive sensorimotor anomalies, abnormal response to
stimulants, olfactory deficits and digestive dysfunction
before the loss of striatal DA [6,7,10,12-14]. In addition,
they show profound anomalies of cortico-striatal trans-
mission [15,16], suggesting alterations within motor cor-
tico-subcortical loops. Whole transcriptome analysis
provides a valuable alternative approach for the detec-
tion of key changes that might not be practical to
attempt by directed single-gene/protein approaches. Pre-
vious studies have evaluated alterations in gene expres-
sion patterns in cells from SNc and striatal tissues from
transgenic mice overexpressing SNCA in the SNc under
different promoters [17,18]. However, little is known of
the effects of SNCA overexpression in the striatum
itself, the region that contains the axon terminals of
dopaminergic (DArgic) neurons and mediates the beha-
vioral effects of DA depletion in PD.
To gain a better understanding of the consequences of
excessive SNCA expression on basal ganglia function,
we performed transcriptome analysis of striatal tissue
from male Thy1-aSyn-mice and wt littermates. The data
further support a key role for SNCA in synaptic func-
tion and also reveal alterations in multiple biological
processes including signaling, apoptosis, and neurogen-
esis, which may be related to both functional deficits
resulting from SNCA overexpression and the sparing of
striatal neurons in most synucleinopathies.
Results and discussion
Differentially regulated genes and functional categories
To understand how increased SNCA expression causes
neuronal cellular dysfunction, we analyzed gene
expression in the striatum, the major target of the
nigrostriatal DArgic projections and a brain region that
plays a key role in the control of motor, affective and
cognitive functions but does not degenerate in PD [19].
We elected to examine changes in gene expression in
the striatum of 6-month-old mice because, as shown in
our previous publications [9-12,14], the Thy1-aSyn mice
begin the show progressive behavioral impairments from
2 months of age but do not show significant DArgic
loss in the striatum until 14 months of age [11]. There-
fore, the time point chosen for these studies corre-
sponds to a pre-manifest phase of PD, when neuronal
dysfunction due to SNCA pathology is present, but DA
is not yet lost.
In order to have sufficient mRNA to generate tran-
script concentrations in the range required to be
detected safely above background levels, total RNA from
groups of 6 tg Thy1-aSyn or 6 wt mice were pooled.
Pooling has the added advantage that it minimizes indi-
vidual variations as a source of gene-expression variance,
which impacts the identification of differentially
expressed genes by DNA microarrays [20]. Genes differ-
entially regulated in Thy1-aSyn mice compared to wt lit-
termates were identified by whole transcriptome
analysis. The relative fold increased expression of
human SNCA in the Thy1-aSyn mice used, as assessed
by qRT-PCR, is shown in Table 1. The SNCA primer
set used hybridized to human SNCA and mouse Snca
mRNAs, albeit less efficiently to the latter, therefore
showing relative differences in total SNCA mRNA
expression between wt and tg mice. However, the very
high level of expression of SNCA mRNA only results in
a 50% increased level of total SNCA protein in this
region in the Thy1-aSyn mice based on quantitative
immunohistochemical analysis (Franich, Richter, and
Chesselet unpublished observations). This suggests that
the model is suitable to evaluate the effects of excess
SNCA in the range observed in humans with SNCA
gene duplication leading to familial forms of PD [3].
Notably, our mouse microarray data rule out a loss of
transcriptional expression of endogenous Snca in Thy1-
Table 1 Array data quality control and differentially-regulated genes in ASO mice
Mice Genotype
SNCA expressionData quality control Altered genes
species
mSncaa
hSNCAb
fold change (mean ± sd) source
Allc
D.E.G.d
Allc
D.E.G.d
% calls (present)intensity (mean ± sd)updown total
ASO
0.9 ± 0.157.5 197 ± 445349484 833
155 ± 8584.1220 ± 271 96137233
wt hSNCA
1.1 ± 0.5
58.6194 ± 425000
mSnca
1.0 ± 0.0
91.3223 ± 295000
Abbreviations: ASO: a-Synuclein (SNCA) overexpressing; wt; wild type; D.E.G.: differentially expressed gene
Notes: a- mSnca detected by 3 probe sets in theMoe-430A chip; b- hSNCA measured by qRT-PCR in 3 wt and 3 ASO mice (Figure 2B); c- Include all changed
genes P < 0.005; d- Selection criteria for D.E.G. (1) gene is present in at least one of the mice groups examined and (2) Fold change > 1.52 (Signal log2 ratio >
0.6), p < 0.005.
Cabeza-Arvelaiz et al. Molecular Neurodegeneration 2011, 6:83
http://www.molecularneurodegeneration.com/content/6/1/83
Page 2 of 16

Page 3

aSyn mice (Table 1). The quality of the data from the
microarray was assessed (see Table 1 and Additional file
1 Figure S1) by inspecting: (1) the percentage of genes
called present (include all changed genes p < 0.005),
which was comparable and above 50% for both mice
groups (wt and tg), and (2) the mean signal intensity,
which was quite similar for both mice groups. After
pairwise comparison, a list of 833 genes altered in
Thy1-aSyn mice was generated. Further filtering to
remove genes with fold change below 1.5, or that were
called absent in both groups reduced the list to 263
genes, which was reduced further to the final 233 differ-
entially expressed genes after removal of redundant
genes. The number of upregulated genes was 96 genes
whereas 137 genes were downregulated (Table 1), indi-
cating that SNCA overexpression effects are mediated
more prominently through negative regulation of tran-
scription. The complete list of differentially expressed
genes with respective test data is available (see Addi-
tional file 2 Table S1). To identify functional categories
overrepresented in the lists of genes altered in Thy1-
aSyn mice we used DAVID [21], which implements the
Gene Ontology (GO) terms in three structured ontolo-
gies that relate gene products on the basis of their asso-
ciated biological processes (BPs), cellular components
(CC), and molecular functions (MFs); it also incorpo-
rates the Genetic Association Database (GAD) of
human diseases, which allows for the identification of
genes associated with human diseases [22]. The signifi-
cant overrepresentation in GO and GAD of the 224
genes that had meaningful annotations pointed to
alterations in gene expression between Thy1-aSyn and
wt mice. These 224 genes were categorized into BPs,
CCs and MFs and multiple overlapping GO terms
(referred to as functional categories hereafter) and dis-
eases were identified; 28 representative and significantly
represented groups (p < 0.05) were selected and are
listed in Table 2; these groups were then organized into
six ad hoc function-related groups, with considerable
overlap between them, and are shown in Table 2. Each
of these groups will be discussed separately below.
Validation of Microarray results by alternative methods
Microarray data was validated by qRT-PCR analysis of a
total of 25 selected genes. 18 selected genes were vali-
dated using pooled total RNAs from Thy1-aSyn and wt
mice (Figure 1A). Similarly, 11 of these 18 genes plus 7
other different genes were also validated using total
RNA from a separate cohort of individual mice (n = 3)
to assess biological variability in gene expression (Figure
1B). In addition, the microarray and qRT-PCR results
for the highly expressed transthyretin (Ttr) gene were
validated using ELISA [23] to measure Ttr protein levels
in the same cohort of individual mice (Figure 2A) and
compared to the levels of SNCA gene expression (Figure
2B). Genes were selected for PCR validation mainly for
the following reasons: they encompass low, moderate
and high intensity signal genes and they form part of
one or more of the functional categories (listed in Table
2) that were significantly over-represented in the list of
differentially-expressed genes. Although not exhaustive,
we consider this list of genes representative of the total
gene expression profile in affected pathways. In particu-
lar, it encompasses at least 3 genes from those overlap-
ping pathways that have been associated with
neurodegeneration and the PD phenotype, summarized
in Figure 3. The qRT-PCR-analyzed genes are distribu-
ted among the functional categories as follows: A- Vesi-
cles and Synapse (Srebf2, Bdnf, Stx1A, Adora2A), B-
Synaptic function (Bdnf, Stx1A, Ywhag, Drd2, Adora2A,
Pde7b), C- Apoptosis (Bdnf, Nr4a2, Mef2c, Adora2A,
Dhcr24, Ttr), D- Behavior (Nr4a2, Rasd2, Drd2,
Adora2A, Ttr, Trh, Bdnf, Cckbr), and E- Vasculature
and lipid metabolism (Mef2c, Srebf2, Dhcr24); and a few
other categories, which for reasons of space we chose
not to show in Figure 3: Neurogenesis (Bdnf, Nr4a2,
Phgdh), Cell cycle (Pttg1, Ptprk, Dhcr24, Ywhag), Signal-
ing (Ptprk, Pde7b, Drd2, Adora2A, Trh, Ywhag), Cell
growth/proliferation (Bdnf, Meg3, Nov, Ptprk), Tran-
scription (Med1, Mef2c, Nr4a2, Srebf2), Response to oxi-
dative stress (Gpx3, Dhcr24, Nr4a2). In addition, we
checked 4 “orphan” genes, not associated with any sig-
nificant category, but which could still be relevant to PD
pathogenesis including, Ckmt1 (Creatine metabolism
and energy generation), Tnnt1 (Muscle contraction),
Psb6 and Psmc4 (Proteasome and cell cycle check
point). Validation in individual mice was done on blocks
of striatal tissue carefully dissected out from tissue slices
to eliminate potential contamination from adjacent tis-
sue, including the subventricular zone and choroid
plexus. The expression of 3 of these genes in Thy1-
aSyn; namely, Psmc4 (Proteasome (prosome, macropain)
26S subunit, ATPase, 4) in pooled samples (Figure 1A)
plus Adora2a (Adenosine A2a receptor) and Med1
(mediator complex subunit) in non-pooled samples (Fig-
ure 1B) did not differ significantly from wt mice by
qRT-PCR and therefore disagree with the microarray
results. In addition, Meg3 (Maternally expressed 3,
imprinted maternally expressed untranslated mRNA)
and Mef2c (myocyte enhancer factor 2c) changed in
opposite directions by the two methods. Nevertheless,
the expression of most genes in the Thy1-aSyn mice
group with both pooled and non-pooled RNA was gen-
erally similar between the microarray and qRT-PCR
analyses, as indicated by the assessment of correlation
using the Pearson’s test which found strong and signifi-
cant correlation between microarray and qRT-PCR mea-
sured expression values in A (r2= 0.7288, p < 0.0001)
Cabeza-Arvelaiz et al. Molecular Neurodegeneration 2011, 6:83
http://www.molecularneurodegeneration.com/content/6/1/83
Page 3 of 16

Page 4

and B (r2= 0.7354, p < 0.0001). These data suggest that
the direction and magnitude of change of gene expres-
sion levels (i.e. either up or down regulation) is accu-
rately predicted by comparison of microarray expression
values.
Because the main purpose of this study is to assess
changes in transcriptome, individual changes were not
systematically validated at the protein level. However,
we did measure the protein levels of transthyretin
because upregulation of this gene has been associated
with neuroprotection in Alzheimer disease [24], and its
high induction in the striatum of Thy1-aSyn mice was
unexpected and of potential functional significance. The
results, shown in Figure 2A, indicate that Ttr protein
levels are significantly elevated in individual Thy1-aSyn
mice and appear to be associated with the levels of
SNCA mRNA as indicated by the assessment of correla-
tion using the Pearson’s test which found significant
correlation (r2= 0.6791, p < 0.043) between the levels of
Ttr protein (Figure 2A) and SNCA mRNA levels (Figure
2B).
SNCA overexpression affects signaling pathways
associated with the pathophysiology of
neurodegenerative diseases
The first group of functional categories influenced by
SNCA overexpression in Table 2 encompasses major
signal transduction pathways such as the mitogen-acti-
vated protein kinases (MAPK) and Ca2+ signaling cas-
cades that regulate multiple cellular processes [25-27].
Protein phosphorylation is a preponderant regulatory
mechanism of signal transduction cascades in eukaryotic
cells that is catalyzed by kinases and reversed by protein
phosphatases [26]. Not surprisingly, half of the genes
affected in Thy1-aSyn mice are phosphoproteins includ-
ing kinases, phosphatases and phosphodiesterases
Table 2 Enriched functional categories for genes differentially regulated by a-synuclein
Database identifierFunctional category (Enriched GO and GAD terms)Number of changed genes
updownpvalue
MF SP_PIR_KEY
BP GO:0007242
BP KEGG_PATH
BP KEGG_PATH
phosphoprotein
intracellular signaling cascade
mmu04010:MAPK signaling pathway
mmu04020:Calcium signaling pathway
49
11
5
3
64
8
4
5
8E-07
5E-02
3E-02
1E-02
CC GO:0045202
CC GO:0005856
CC GO:0031410
BP GO:0048167
BP GO:0007267
BP GO:0006897
synapse
cytoskeleton
710
10
7
3
5
3
3E-06
4E-02
1E-02
4E-04
3E-02
5E-02
12
8
3
4
4
cytoplasmic vesicle/vesicle-mediated transporta
regulation of synaptic plasticity
cell-cell signaling
endocytosis
BP GO:0045941
BP GO:0042127
MF GO:0005520
BP GO:0051259
BP GO:0042981
BP GO:0007610
positive regulation of transcription
regulation of cell proliferation
IGF binding/growth factor bindingb
protein oligomerization
regulation of apoptosis
behavior
5
8
1
2
6
10
10
9
4
3
11
13
3E-03
2E-03
7E-03
1E-02
2E-03
7E-04
BP GO:0007626
BP GO:0008344
BP GO:0007631
BP GO:0001975
BP GO:0030146
MF GO:0005179
locomotory behavior
adult locomotory behavior
feeding behavior
response to amphetamine
diuresis/excretionc
hormone activity
sterol metabolic process/cholesterol metabolismd
neuron differentiation
neuron projection development
vasculature development
5
1
5
2
3
1
8
4
7
1
1
5
4E-03
8E-03
5E-04
2E-02
2E-03
2E-02
BP GO:0016125
BP GO:0030182
BP GO:0031175
BP GO:0001944
542E-02
9E-07
2E-05
4E-02
10
7
2
10
6
6
D G_A_DB_D
D G_A_DB_D
neurological diseases
diabetes, types 1 and 2
4
3
5
6
5E-02
5E-02
Abbreviations: D: disease; G_A_DB_D:Genetic_Association_DataBase_D; GO: Gene Ontology; BP: Biological process; MF: Molecular function; CC: Cell. component;
KEGG_PATH: KEGG_PATHWAY; SP_PIR_KEY: SP_PIR_KEYWORDS;
Notes: a- GO:0016192; b- GO:0019838; c- GO:0007588; d- SP_PIR_KEYWORDS: cholesterol metabolism.
Cabeza-Arvelaiz et al. Molecular Neurodegeneration 2011, 6:83
http://www.molecularneurodegeneration.com/content/6/1/83
Page 4 of 16

Page 5

Figure 1 Corroboration of microarray results for differentially
expressed genes. Confirmation of the changes of selected
microarray differentially expressed genes by comparison to the
changes assessed by quantitative real-time polymerase chain
reaction (qRT-PCR) using: A- pooled RNA from wild-type (wt) and a-
synuclein (SNCA) overexpressing transgenic (tg) mice groups; B- RNA
from individual wt and tg mice. The qRT-PCR average fold changes
of selected genes in tg mice striata normalized to Gapdh
(Glyceraldehyde-3-phosphate dehydrogenase), and relative to
expression in wt mice (dotted line) are shown (mean ± SEM); ns
(not significantly different expression between tg and wt mice,
using a t-test, n = 3, p > 0.05); SEM values in (A) show qRT-PCR
technical variation in, and in (B) show variation between mice plus
qRT-PCR technical variation). Pearson’s test found significant
correlation between microarray and qRT-PCR results in A (r2=
0.7288, p < 0.0001) and B (r2= 0.7354, p < 0.0001). Ttr
(transthyretin), Phgdh (3-phosphoglycerate dehydrogenase), Pde7b
(phosphodiesterases 7b), Tnn1 (troponin t1), Adora2a (adenosine
A2a receptor), Meg3 (Maternally expressed 3), Drd2 (dopamine
receptor 2), Rasd2 (rasD family, member 2), Psmb6 (proteasome
(macropain) subunit, b 6), Med1 (mediator complex subunit), Ptprk
(protein tyrosine phosphatase, receptor, K), Pttg1 (pituitary tumor-
transforming 1), Nov (Nephroblastoma Overexpressed), Pparbp
(peroxisome proliferator activated receptor binding protein), Stx1a
(syntaxin 1a), Nr4a2, Psmc4 (proteasome (macropain) 26S subunit,
ATPase, 4), Mef2c (myocyte enhancer factor 2c), Cckbr
(cholecystokinin B receptor), Dhcr24 (24-dehydrocholesterol
reductase), Ckmt1 (creatine kinase, mitochondrial 1), Srebf2 (sterol
regulatory element-binding factor 2), Trh (thyrotropin-releasing
hormone), Bdnf (brain derived growth factor), Gpx3 (glutathione
peroxidase 3), Ywhag (Y-3-monooxygenase/W 5-monooxygenase
activation protein, g polypeptide).
Figure 2 Validation of microarray and qRT-PCR results for the
highly expressed transthyretin (Ttr) gene. The confirmation of
the differential expression of Ttr was achieved by measuring
striatum Ttr protein levels by ELISA in individual wt (wild-type) mice
and aso (a-synuclein (SNCA) overexpressing or Thy1-aSyn mice). A-
The average Ttr protein levels in the striata from Thy1-aSyn mice
and wt mice groups are shown (mean + SEM, **p < 0.01, non-
parametric Mann-Whitney t-test, aso vs wt, n = 3 per group); levels
of Ttr in each individual mouse are also shown to illustrate variation
between mice (error bars represent SEM showing technical
variations in triplicate ELISA measurements). B- The average human
SNCA mRNA levels in the striata from Thy1-aSyn mice and wt mice
groups determined by qRT-PCR, normalized to Gapdh, and relative
to the levels of endogenous Snca mRNA in the wt mice that
partially hybridize to the human SNCA primers are shown (mean +
SEM, ***p < 0.001, non-parametric Mann-Whitney t-test, aso vs wt, n
= 3 per group); levels of SNCA mRNA in each individual mouse are
also shown to illustrate variation between mice (error bars represent
SEM showing technical variation in triplicate qRT-PCR analyses.
Pearson’s test found significant correlation (r2= 0.6791, p < 0.043)
between the two measured expression values in A and B.
Cabeza-Arvelaiz et al. Molecular Neurodegeneration 2011, 6:83
http://www.molecularneurodegeneration.com/content/6/1/83
Page 5 of 16