The insulin-like growth factor pathway is altered
in spinocerebellar ataxia type 1 and type 7
Jennifer R. Gatchel*, Kei Watase†, Christina Thaller‡, James P. Carson‡§, Paymaan Jafar-Nejad¶, Chad Shaw¶, Tao Zu?,
Harry T. Orr?, and Huda Y. Zoghbi*,**††
Departments of *Neuroscience,‡Biochemistry and Molecular Biology, and¶Molecular and Human Genetics and **Howard Hughes Medical Institute, Baylor
College of Medicine, Houston, TX 77030;†Tokyo Medical and Dental University, Tokyo 113-8519, Japan;§Pacific Northwest National Laboratory, Richland,
WA 99354; and?Institute of Human Genetics, University of Minnesota, Minneapolis, MN 55455
Contributed by Huda Y. Zoghbi, November 30, 2007 (sent for review October 3, 2007)
Polyglutamine diseases are inherited neurodegenerative disorders
caused by expansion of CAG repeats encoding a glutamine tract in
the disease-causing proteins. There are nine disorders, each having
distinct features but also clinical and pathological similarities. In
particular, spinocerebellar ataxia type 1 and 7 (SCA1 and SCA7)
patients manifest cerebellar ataxia with degeneration of Purkinje
cells. To determine whether the disorders share molecular patho-
express the glutamine-expanded protein from the respective en-
dogenous loci. We found common transcriptional changes, with
down-regulation of insulin-like growth factor binding protein 5
(Igfbp5) representing one of the most robust changes. Igfbp5
down-regulation occurred in granule neurons through a non-cell-
autonomous mechanism and was concomitant with activation of
the insulin-like growth factor (IGF) pathway and the type I IGF
receptor on Purkinje cells. These data define one common patho-
genic response in SCA1 and SCA7 and reveal the importance of
intercellular mechanisms in their pathogenesis.
cerebellum ? neurodegeneration ? polyglutamine ? Purkinje cell ? granule
types 1, 2, 3, 6, 7, and 17 as well as Huntington disease (HD),
dentatorubral–pallidoluysian atrophy (DRPLA), and spinal bul-
bar muscular atrophy (SBMA). They are caused by expansion of
CAG repeats encoding a polyglutamine tract in different pro-
teins (1). Evidence suggests that the polyglutamine expansion
confers toxicity predominantly through a gain-of-function mech-
anism, although loss of function may also play a role (1, 2).
Despite widespread expression of the disease proteins in the
central nervous system (CNS), there is selective vulnerability of
specific neurons in each disease (1). Such selective neuronal
vulnerability gives rise to the specific features of each disease.
SCA1 and SCA7 have distinct features while also sharing some
key similarities. In SCA1, primary symptoms include ataxia,
dysarthria, and bulbar dysfunction, whereas cognitive impair-
ment is more varied (1). This presentation is associated with
loss of granule neurons (1). In SCA7, progressive visual loss,
cerebellar ataxia, and dysarthria are the most common clinical
features (3). In addition to retinal degeneration, cerebellar PCs
and neurons of the dentate nucleus and inferior olive are among
the earliest to degenerate (3).
Given that SCA1 and SCA7 share a cerebellar degenerative
phenotype, we proposed that some shared molecular changes
might occur in both diseases, and that common molecular
alterations could pinpoint pathways that could be targeted to
modulate or monitor the pathogenesis of more than one disease.
We focused on transcriptional changes because both ATXN1
and ATXN7 play roles in transcriptional regulation (4–8), and
transcriptional defects can be detected in early-symptomatic
stages of both SCA1 and SCA7 mouse models (9, 10).
olyglutamine diseases are a group of inherited neurodegen-
erative disorders that include spinocerebellar ataxia (SCA)
To test our hypothesis, we examined cerebellar gene expres-
sion patterns in SCA1 and SCA7 knockin (KI) models—
Sca1154Q/2Qmice and Sca7266/5Qmice. These mice express glu-
tamine-expanded ATXN1 and ATXN7, respectively, in the
endogenous pattern, and they accurately replicate disease fea-
tures (10, 11). Here, we describe the identification of a signal
transduction pathway—the insulin-like growth factor (IGF)
pathway—that is affected in both diseases. The alterations we
describe provide insight into cerebellar pathogenesis in SCA1
and SCA7 and might prove helpful in monitoring disease
Sca1154Q/2Qand Sca7266Q/5QMice Share Cerebellar Gene Expression
Changes. We surveyed global transcriptional changes in
Sca1154Q/2Qand Sca7266Q/5Qcerebellum using microarray tech-
nology. To capture early changes, we carried out experiments at
an early symptomatic disease stage (Sca1154Q/2Qmice: 4, 9, and
12 weeks; Sca7266Q/5Qmice: 5 wks) (10, 11). [The timing of
microarray and all subsequent experiments in the context of
Sca1154Q/2Qand Sca7266Q/5Qdisease progression is provided in
supporting information (SI) Fig. 7 a and b.]
We found significant concordance between Sca1154Q/2Qand
Sca7266Q/5Qcerebellar expression patterns (P ? 0.00040), with 31
probe sets commonly misregulated (SI Tables 1 and 2). Of note
is that 26 of the 31 differentially regulated probe sets in both
Sca1154Q/2Qand Sca7266Q/5Qcerebellum were down-regulated in
predominance of down-regulation is itself significant (P ?
Ontology analysis revealed that genes involved in growth
factor-binding protein complexes and inositol kinase and phos-
phatase activities were among those enriched in this overlap list
(SI Figs. 8–13). Despite the significant overlap between
Sca1154Q/2Qand Sca7266Q/5Qdatasets, many differentially ex-
pressed genes were specific to each model [NCBI Gene Expres-
sion Omnibus (GEO) accession no. GSE9914]. These included
genes involved in mRNA processing and chromatin (Sca1154Q/2Q)
and those involving transcription factor-binding and MHC class
I proteins (Sca7266Q/5Q) (SI Figs. 14–19).
Author contributions: J.R.G., K.W., H.T.O., and H.Y.Z. designed research; J.R.G., K.W., C.T.,
P.J.-N., and T.Z. performed research; J.P.C. contributed new reagents/analytic tools; J.R.G.,
C.S., and T.Z. analyzed data; and J.R.G. and H.Y.Z. wrote the paper.
The authors declare no conflict of interest.
Freely available online through the PNAS open access option.
Data deposition: The data reported in this paper have been deposited in the Gene
Expression Omnibus (GEO) database, www.ncbi.nlm.nih.gov/geo (accession no. GSE9914).
††To whom correspondence should be addressed. E-mail: email@example.com.
This article contains supporting information online at www.pnas.org/cgi/content/full/
© 2008 by The National Academy of Sciences of the USA
January 29, 2008 ?
vol. 105 ?
no. 4 ?
Igfbp5 Is Down-Regulated in Sca1154Q/2Qand Sca7266Q/5QCerebellum.
The most significant common gene expression change based on
fold change was the down-regulation of Igfbp5 (SI Table 1), a key
component of the IGF axis (12). Northern blot analysis con-
firmed the down-regulation of Igfbp5 at an early disease stage in
both Sca1154Q/2Qand Sca7266Q/5Qcerebellum (Fig. 1 a and b).
Furthermore, this down-regulation was progressive, such that
RNA levels at the late disease stage (40 weeks in Sca1 KI and 14
weeks in Sca7 KI) were ?60% of those of WT animals (Fig. 1
because we did not find similar changes in cortex, a tissue
relatively unaffected in pathogenesis (SI Fig. 20).
Igfbp5 Is Down-Regulated in a Non-Cell-Autonomous Manner in the
Granule Layer of the Cerebellar Cortex. Because Igfbp5 changed
progressively in both Sca1154Q/2Qand Sca7266Q/5Qcerebellum, we
were interested in determining the cell types in which the
expression changes were occurring. Because PCs are among the
most severely affected neurons in SCA1 and SCA7, we hypoth-
esized that Igfbp5 might be expressed and down-regulated in
these neurons. To test this hypothesis, we performed in situ
hybridization (ISH) experiments and found that within the
cerebellum, Igfbp5 was predominantly expressed in the granule
layer (Fig. 2 and SI Fig. 21b). The size, number, and location of
the cells in which this intense signal was found were all consistent
with granule neurons. Expression was also found in the deep
cerebellar nuclei, whereas signal in other layers of the cerebellar
cortex was extremely sparse. In addition, we found a sharp
anterior–posterior gradient of Igfbp5 expression in the cerebellar
cortex. Expression was highest in the anterior compartment
(lobules I–VI) and was marked by a boundary in lobule VI (Fig.
2 a and b). Importantly, this expression pattern was observed
independently by using a different probe and methodology (13).
We used custom-designed imaging software (14) to pseudo-
color and quantify Igfbp5 expressing cells from tissue sections
spanning entire cerebellar hemispheres (Fig. 2 b, d, and f). This
anterior granule compartment (Fig. 2b). Expression was dra-
matically decreased in Sca1154Q/2Qand Sca7266Q/5Qcerebellum
(Fig. 2 c–f and i). Expression of Fgf-1, an anterior granule
compartment marker (15), and Igfbpl1, an additional growth
factor pathway component, were not altered (SI Fig. 21 c–f),
indicating that the decrease in Igfbp5 was specific and not due to
a general expression deficiency of all anterior compartment
genes or growth factors.
To gain insight into the nature of Igfbp5 down-regulation in
granule neurons and determine whether it was or was not cell
autonomous, we used an SCA1 transgenic (Tg) mouse model
(SCA1[82Q]Tg) that expresses a mutant SCA1 allele encoding
ATXN1 with 82 glutamines only in PCs (16) and develops a
progressive cerebellar degenerative phenotype (SI Fig. 7c). ISH
experiments and quantification revealed a decrease in Igfbp5
expression in the cerebella of these mice (Fig. 2 g, h, and j). This
suggested that the down-regulation of Igfbp5 in the cerebellar
granule layer occurred by a non-cell autonomous mechanism,
because it occurred when the mutant protein was expressed only
Igfbp5 Down-Regulation Is Coupled with Increased Activation of the
IGF1R. To probe the underlying cause of Igfbp5 down-regulation,
we examined other IGF axis components. Given the role of
and b) Northern blots of cerebellar RNA at early-symptomatic (Sca1154Q/2Q, 4
and 12 weeks; Sca7266Q/5Q, 5 weeks) and late-symptomatic (Sca1154Q/2Q, 40
weeks; Sca7266Q/5Q, 14 weeks) disease stages show progressive down-
regulation of Igfbp5 RNA. (c and d) Quantification of Igfbp5 RNA relative to
Gapdh using multiple animal pairs (n ? 3 pairs of Sca1154Q/2Qand WT controls;
n ? 5 and 4 of Sca7266Q/5Qand WT mice, respectively) shows significant
down-regulation of Igfbp5 in late-symptomatic cerebellum. Values represent
mean ? SEM. The asterisks indicate significant genotype differences (*, P ?
Igfbp5 RNA is decreased in Sca1154Q/2Qand Sca7266Q/5Qcerebellum. (a
for Igfbp5 (a, c, e, and g) and pseudocolored sagittal sections (b, d, f, and h)
with Igfbp5-expressing cells colored according to their expression level (red,
strong; blue, moderate; yellow, weak). (i) Quantification of strongly and
moderately expressing cells reveals a decrease in Sca1154Q/2Q(18–19 weeks)
and Sca7266Q/5Q(12–13 weeks) cerebellum compared with WT. A decrease is
also observed when mutant ATXN1 is only expressed in PC, in SCA1[82Q]Tg/?
(22–23 weeks) cerebellum. n ? 2 KI and 2 WT animals for both Sca1154Q/2Qand
Sca7266Q/5Qand 2 Tg/? and 2 WT littermates for SCA1[82Q], with 40–47
sections quantified per animal. KI and Tg/? expression values are shown as a
percentage of WT. Values represent mean ? SEM.
Igfbp5 is decreased in the cerebellar granule layer of Sca1 and Sca7
www.pnas.org?cgi?doi?10.1073?pnas.0711257105 Gatchel et al.
Igfbp5 in influencing IGF1 bioavailability (12), we determined
whether the biological effects of IGF1, which are mainly medi-
ated through the IGF1R, were altered. On binding to IGF1, the
cytoplasmic portion of the IGF1R is phosphorylated, creating a
binding site for IRS-1 and phosphatidyl-inositol-3-kinase (PI3-
K). These substrates subsequently play a role in IGF1R-
mediated protection and proliferation by recruiting downstream
phosphorylation events, including activation of Akt and Ras/
Given the potential of Igfbp5 to block the availability of IGF1
for its receptor (18, 19), we hypothesized that in Sca1 and Sca7
KI cerebellum, decreased Igfbp5 expression would lead to an
increased amount of IGF1 available to activate the IGF1R and
downstream signaling. To test this hypothesis, we examined
IGF1R phosphorylation in cerebellar homogenates from Sca1
and Sca7 KI mice. We found a significant increase in IGF1R
phosphorylation in KI cerebellum (Fig. 3 a–d). To determine
whether expression of mutant polyglutamine protein exclusively
in PCs was sufficient to confer this increase, we collected
homogenates from WT, SCA1[82Q]Tg/?, and SCA1[82Q]Tg/
Tg, mice. We again found a significant increase in IGF1R
activation (Fig. 3 e and f). Thus, the significant decrease in Igfbp5
expression was accompanied by an increase in IGF1R activation.
For further IGF pathway studies, we focused on SCA1 pa-
thology, given the availability of knockin, PC-specific, and
conditional models of disease (11, 16, 20). Within the cerebellar
cortex, the IGF1R is expressed on PC soma and dendrites, in
presynaptic axon terminals in the molecular layer, and in mossy-
fiber rosettes and granule-cell soma; expression in glia and in
endothelial cells has also been observed (21). We carried out
immunostaining experiments and found increased receptor
phosphorylation in these populations in SCA1[82Q]Tg/Tg cer-
ebellum compared with WT, particularly in the PC and molec-
ular layers, as well as intense staining in the granule layer (Fig.
4a). A similar pattern was observed in Sca1154Q/2Qcerebellar
cortex (SI Fig. 22).
The increased IGF1R activation occurred in conjunction with
increased activation of the downstream effectors, phosphory-
lated-Akt and -Erk, in SCA1[82Q]Tg/Tg cerebellum (SI Figs. 23
and 24a). Interestingly, relatively less granule layer phospho-
IGF1R immunoreactivity and phospho-ERK activation was
found in lobule 10 of SCA1[82Q] cerebellum compared with
anterior lobules (Fig. 4 a and b and SI Fig. 24a); lobule 10 was
also relatively spared in disease compared with anterior lobules
based on coimmunofluorescence with calbindin, a protein that
labels PC dendrites (Fig. 4 and SI Fig. 24). This suggested that
activation in certain cell types occurred most prominently in
regions of enhanced neuronal vulnerability.
The Down-Regulation of Igfbp5 Is Reversible Upon Modulation of
Mutant ATXN1 Expression. The pathology induced by polyglu-
tamine-expanded proteins can be partially reversed if mutant
protein expression is halted early enough in disease (22). Spe-
cifically, in a conditional SCA1 mouse model, the PC specific
expression of mutant ATXN1 with 82 glutamines can be turned
off with the administration of doxycyline, resulting in rescue of
both clinical features and cellular pathology (20).
To determine whether Igfbp5 down-regulation was responsive
to the disease state, we collected cerebellar RNAs from WT
animals and from four groups of conditional Pcp2-tTA(Tg/?)/
TRE-SCA1[82Q](Tg/Tg) littermate animals: age-matched 12-
week-old mice in which the SCA1 transgene had never been
turned off and mice in which the SCA1 transgene had been
expressed for 12 weeks and then turned off for either 4, 8, or 12
weeks. Reduction of mutant ATXN1 expression resulted in a
gradual, significant increase in Igfbp5 levels to near WT (Fig. 5).
SCA1[82Q] Tg cerebellum. (a, c, and e) Western blot analyses using cerebellar
(active) form of the IGF1R as well as to Gapdh. (b, d, and f) Quantification of
phospho-IGF1R levels relative to Gapdh using multiple pairs of KI, Tg, and WT
compared with WT (f). Values represent mean ? SEM. The asterisks indicate
significant genotype differences (*, P ? 0.05;**, P ? 0.01).
IGF1R phosphorylation is increased in Sca1154Q/2Q, Sca7266Q/5Q, and
ogy in SCA1[82Q]Tg/Tg cerebellum. (a) Representative confocal images from
11-week-old SCA1[82Q]Tg/Tg and WT mice showing calbindin and phospho-
IGF1R immunostaining in sections from anterior cerebellar lobules and from
lobule 10. (b) Calbindin immunostaining is decreased, and phospho-IGF1R
levels are increased in the PC, molecular, and granule layers in Tg/Tg cerebel-
lum compared with WT. Immunoreactivity for phosphorylated ERK, which is
in anterior lobules compared with lobule 10. Similar results were obtained
from three pairs of SCA1[82Q]Tg/Tg and WT mice (three to four sections per
animal) in four independent experiments. (Scale bars, 100 ?m.)
Patterns of IGF1R and ERK1/2 phosphorylation relative to PC pathol-
Gatchel et al.
January 29, 2008 ?
vol. 105 ?
no. 4 ?
Thus, the changes in Igfbp5 closely correlate with the disease
state in SCA1.
Sca1154Q/2Qand Sca7266Q/5QCerebellum. The overlapping cerebellar
phenotypes in SCA1 and SCA7 raised the possibility that these
two disorders share some molecular pathogenic changes.
Through analysis of Sca1154Q/2Qand Sca7266Q/5Qmice, we show
that there are shared transcriptional changes that happen early
in disease and, at least for Igfbp5, are specific to pathogenesis.
ATXN1 and ATXN7 each affect transcription in a context-
specific fashion (4–8), and many of the categories of gene
changes in our microarray studies were Sca1154Q/2Q- or
Sca7266Q/5Q-specific (SI Figs. 14–19). This indicates that dis-
tinct processes are likely central to the pathogenesis of each
disorder, which may partly reflect the role ATXN1 in RNA
binding and ATXN7 in a histone acetyltransferase (HAT)
complex (7, 8, 23). In contrast, many of the common gene
changes, including Igfbp5, are not likely direct targets of
ATXN1- and ATXN7-mediated pathogenesis, but, rather,
represent shared early events in a cerebellar response. Igfbp5
down-regulation may further be part of a more general re-
sponse to neuronal injury. We describe this down-regulation at
early disease stages in response to ATXN1- and ATXN7-
induced injury, but changes in Igfbp5 levels have also been
reported at symptomatic/late-symptomatic disease stages in
models based on overexpression of mutant polyglutamine
proteins (24). It would be interesting to determine whether
these reported changes also occur at earlier disease stages as
they might hint that several degenerative disorders share some
pathogenic responses. Thus, we describe a common transcrip-
tional read-out of the early disease state that could potentially
be used to pinpoint ways to modulate and monitor pathogen-
esis in more than one disorder.
Non-Cell-Autonomous Down-Regulation of Igfbp5: A Role for PC-
Driven Neuron–Neuron Interactions in SCA1 and SCA7 Pathogenesis.
Despite the primary PC pathology in SCA1 and SCA7, expres-
sion and down-regulation of Igfbp5 occurred in the cerebellar
granular layer. Even more remarkably, we show that Igfbp5
down-regulation takes place through a non-cell-autonomous
mechanism. Previous work supporting the importance of inter-
cellular interactions in polyglutamine disorders has relied on the
use of a truncated fragment of huntingtin or the overexpression
of the mutant ATXN7 protein in cells that provide support to
PCs (25–27). To date, however, this issue has not been investi-
gated in KI models, whereby the mutant protein is expressed in
the endogenous manner such that intercellular interactions and
their influence on selective vulnerability can be investigated in
the endogenous disease context. In this study, we show that
whether PC injury arises from endogenous or cell-specific ex-
pression of full-length mutant polyglutamine protein, a similar
transcriptional change occurs in granule neurons, indicating that
this change is driven at least in part from a pathogenic response
initiated in affected PCs.
It is interesting to consider how PC injury might be conveyed
to the granule neuron to cause the decrease in Igfbp5. Expression
of mutant ATXN1 in SCA1 mouse models leads to reduction in
PC dendritic arborization (11, 16). Thus, one possibility is that
changes in PC dendrites—the postsynaptic targets of granule cell
parallel fibers—result in altered PC-granule neuron synaptic
contacts that could trigger gene expression changes in granule
neurons. Another possibility is that mutant ATXN1 or ATXN7
induces decreased expression of IGF1 in PCs, which in turn
could drive this change, given that IGF1 levels can influence
Igfbp5 expression (28, 29) (Fig. 6).
No matter what the mechanism, concomitant changes in
pathogenesis of these disorders in the cerebellum (Fig. 6). The
findings from these two disease models thus underscore the role
of Igfbp5 and the IGF pathway as a means of PC–granule neuron
communication, particularly under conditions of stress.
Igfbp5 Down-Regulation Occurs in the Context of Additional IGF
Pathway Misregulation. The finding of additional IGF pathway
misregulation provides insight into how Igfbp5 down-regulation
weeks. (b) Igfbp5 RNA levels relative to Gapdh in SCA1[82Q] cerebellum were
normalized to WT, with three animals in each group. Data are expressed as
mean ? SEM; (*, P ? 0.05).
Igfbp5 down-regulation is reversible in an SCA1[82Q] conditional
Sca7266Q/5Q, and SCA1[82Q] (B) cerebellum. Igfbp5 is expressed in granule
neurons, whereas the IGF1R is more broadly expressed, suggesting possible
cross-talk between PCs and granule neurons. In normal cerebellum, IGF1 can
ATXN1- and ATXN7-induced PC injury causes down-regulation of Igfbp5 in
granule neurons by a non-cell-autonomous mechanism (B). This mechanism
may involve IGF1, expressed in part by PCs or, alternatively, could result from
granule neurons sensing altered synaptic activity at granule–PC–parallel fiber
connections. Given the increased IGF1R activation in Sca1154Q/2Q, Sca7266Q/5Q,
Igfbp5 in granule neurons influences the local actions of IGF1, in part by
increasing the availability of IGF1 for its receptor on PCs, which can lead to
activation of downstream effectors such as Akt.
Model of Igfbp5 and IGF1 signaling in normal (A) and in Sca1154Q/2Q,
www.pnas.org?cgi?doi?10.1073?pnas.0711257105 Gatchel et al.
might influence pathogenesis in SCA1 and SCA7 models. Igfbp5
the potential to inhibit IGF1 action when present in excess in an
in vitro cerebellar model (19). Furthermore, when cerebellar
granule neurons in vitro undergo KCl deprivation, Igfbp5 is
down-regulated, whereas the IGF1R is transiently increased,
which may reflect an attempt to maximize IGF1 signaling (31).
The changes we observe in Sca1154Q/2Q, Sca7266Q/5Q, and
SCA1[82Q] cerebellum may similarly represent an adaptive
and ATXN7 stress.
IGF1 signaling can have neuroprotective effects in the context
of specific mutant polyglutamine proteins (32, 33). Indeed, IGF1
atrophy, vascular dysfunction, and accumulation of deleterious
products often associated with the aging brain (34). As such,
compensatory responses to maximize IGF1 signaling in
Sca1154Q/2Qand Sca7266Q/5Qmice could serve a protective role.
However, prolonged activation of neurotrophic signaling may
of IGF1 downstream signaling results in lifespan extension in
Caenorhabditis elegans, Drosophila, and mice (36), suggesting that
IGF pathway activation may become deleterious in the context of
mutant ATXN1, given that phosphorylation of ATXN1 at Ser-776,
a putative Akt site, promotes stabilization and accumulation of
a Drosophila SCA1 model (37, 38). It is of interest that we found
relative preservation of PCs in cerebellar lobule 10 of SCA1[82Q]
lobule. Thus, although IGF pathway changes could be adaptive
responses to pathology, our data suggest that chronic overstimula-
tion of this pathway may adversely modulate pathogenesis in these
Igfbp5 Expression Is Tightly Correlated with the Disease State. The
change in Igfbp5 we describe occurs very early in two disease
expression to the disease state are key properties of a biomarker
could be used in this capacity in these or other disorders. In sum,
the common changes we describe, in addition to providing
insight into the role of intercellular signaling in the cerebellar
cortex as part of the pathogenic response in SCA1 and SCA7,
may provide future opportunities to modulate and monitor
disease course in these devastating diseases.
Materials and Methods
Animals. Sca1154Q/2Q, Sca7266Q/5Q, SCA1[82Q], and SCA1[82Q] conditional Tg
mice have been described (10, 11, 16, 20). Additional details are available in SI
Text. Experiments were in accordance with protocols approved by the Insti-
tutional Animal Care and Use Committees (IACUC) of Baylor College of
Medicine and the University of Minnesota.
Microarray Studies and Other Statistical Analysis. Cerebella were dissected on a
clean-up over RNeasy mini columns (Qiagen) according to manufacturers’ pro-
tocols. For Sca1 KI, n ? 3 per genotype at both 4 weeks and at 9–12 weeks. For
hybridization onto oligonucleotide arrays (Mouse Expression Array 430A (Sca1
SI Text. An ANOVA and linear contrasts to estimate KI vs. WT differences were
computed for Sca1 and Sca7 datasets. To determine the concordance between
scores and a Kendall’s ? analysis of the ‘‘up-regulated’’ vs. ‘‘down-regulated’’
genes lists for the Sca1 and Sca7 experiments were calculated. T tests at the P ?
0.01 level identified 667 down- vs. 349 up-regulated probe sets (Sca1), and 157
down- vs. 196 up-regulated probe sets (Sca7) (NCBI GEO database, accession
GSE9914). A binomial test of proportions against the null hypothesis of equal
proportion was computed to analyze the asymmetric distribution of up- and
down-regulated probe sets in the commonly misregulated list.
For Northern blot experiments, KI, Tg, and WT differences were evaluated
by using independent sample t tests. Western blot data were analyzed by
using an independent sample t test for KI vs. WT differences, or a one-way
ANOVA for SCA1[82Q]Tg data.
RNA Studies. Total cerebellar or cortical RNA was extracted by using TRIzol
(Invitrogen), and Northern blot was performed as described (20). Igfbp5 RNA
levels were normalized to the amount of Gapdh per lane. For ISH, brains from
Sca1154Q/2Q, Sca7266Q/5Q, and SCA1[82Q]Tg/? mice were frozen in optimal
quantitative analysis of the ISH signal using the Celldetekt protocol were
carried out as described in ref. 14 and www.genepaint.org/RNA.htm. Addi-
tional details and probe descriptions are available in SI Text. Details of animal
age and numbers are in figure legends.
Western Blotting. Animalswerekilledbyrapidcervicaldislocationandcerebella
dissected on a chilled glass plate and homogenized in 1? RIPA Buffer [150 mM
NaCl, 0.1% SDS, 1% Triton X-100, 50 mM Tris?HCl (pH 7.6)] with inhibitors
[Complete Protease Inhibitors (Roche); phosphatase inhibitor cocktails I and II
(Sigma)]. Samples were sonicated for 1–2 seconds, followed by centrifugation.
Supernatants were flash-frozen in liquid nitrogen and stored at ?80°C.
Proteins were evaluated by Western blotting according to standard proce-
provided in SI Text. Analysis results are shown for three Sca1 KI and three WT
littermates at 49–52 weeks; three Sca7 KI and three WT littermates at 12–13
weeks; and three SCA1[82Q]Tg/Tg, three Tg/? and three age-, sex-, and strain-
matched WT mice at 8–10 weeks. Results were repeated with two additional
Sca7 KI) and with one additional cohort of SCA1[82Q]Tg/Tg, Tg/? and WT mice.
Immunofluorescence. Details of perfusion, fixation, and immunolabeling with
mouse anti-calbindin antibody (1:1,000; Sigma) and image capture are avail-
able in SI Text. Experiments were carried out by using three SCA1[82Q]Tg/Tg,
one additional cohort at 20 weeks.
Baylor College of Medicine Microarray Core; Y. Klisch for artwork; and mem-
National Institutes of Health Grants NS27699 (to H.Y.Z.) and NS22920 (to
H.T.O.), Department of Energy Laboratory Directed Research and Develop-
ment Grant DE-AC05-76RL01830 (to J.P.C.), National Research Service Award
Grant 1 F30 MH072117 (to J.R.G.), and Baylor College of Medicine, Mental
Retardation Developmental Disabilities Research Center Gene Expression
Core Grant HD024064.
1. Orr HT, Zoghbi HY (2007) Trinucleotide repeat disorders. Annu Rev Neurosci 30:575–
2. Borrell-Pages M, Zala D, Humbert S, Saudou F (2006) Huntington’s disease: From
huntingtin function and dysfunction to therapeutic strategies. Cell Mol Life Sci
3. Stevanin G, et al. (1998) Spinocerebellar ataxia type 7. Analysis of Triplet Repeat
Disorders, ed Hayden MR, Rubinsztein DC (BIOS Scientific, Oxford, UK), pp 155–
4. Tsai CC, et al. (2004) Ataxin 1, a SCA1 neurodegenerative disorder protein, is function-
ally linked to the silencing mediator of retinoid and thyroid hormone receptors. Proc
Natl Acad Sci USA 101:4047–4052.
5. Serra HG, et al. (2006) RORalpha-mediated Purkinje cell development determines
disease severity in adult SCA1 mice. Cell 127:697–708.
6. Lam YC, et al. (2006) ATAXIN-1 interacts with the repressor Capicua in its native
complex to cause SCA1 neuropathology. Cell 127:1335–1347.
7. Helmlinger D, et al. (2006) Glutamine-expanded ataxin-7 alters TFTC/STAGA recruit-
ment and chromatin structure leading to photoreceptor dysfunction. PLoS Biol 4:e67.
8. Palhan VB, et al. (2005) Polyglutamine-expanded ataxin-7 inhibits STAGA histone
acetyltransferase activity to produce retinal degeneration. Proc Natl Acad Sci USA
9. Serra HG, et al. (2004) Gene profiling links SCA1 pathophysiology to glutamate
signaling in Purkinje cells of transgenic mice. Hum Mol Genet 13:2535–
10. Yoo SY, et al. (2003) SCA7 knockin mice model human SCA7 and reveal gradual
Gatchel et al.
January 29, 2008 ?
vol. 105 ?
no. 4 ?
11. Watase K, et al. (2002) A long CAG repeat in the mouse ScaI locus replicates SCA1 Download full-text
features and reveals the impact of protein solubility on selective neurodegeneration.
12. Beattie J, Allan GJ, Lochrie JD, Flint DJ (2006) Insulin-like growth factor-binding
protein-5 (IGFBP-5): A critical member of the IGF axis. Biochem J 395:1–19.
13. Lein ES, et al. (2007) Genome-wide atlas of gene expression in the adult mouse brain.
14. Carson JP, Eichele G, Chiu W (2005) A method for automated detection of gene
expression required for the establishment of a digital transcriptome-wide gene ex-
pression atlas. J Microsc 217:275–281.
the granule cell layer of the murine cerebellar cortex. J Comp Neurol 391:444–455.
by an expanded CAG trinucleotide repeat. Cell 82:937–948.
factor I receptor, phosphatidylinositol 3-kinase, and Akt. Mol Cell Biol 17:1595–1606.
18. Kalus W, et al. (1998) Structure of the IGF-binding domain of the insulin-like growth
factor-binding protein-5 (IGFBP-5): Implications for IGF, IGF-I receptor interactions.
EMBO J 17:6558–6572.
19. Zhong J, Deng J, Ghetti B, Lee WH (2002) Inhibition of insulin-like growth factor I
mutant mice: in vitro analysis. J Neurosci Res 70:36–45.
20. Zu T, et al. (2004) Recovery from polyglutamine-induced neurodegeneration in con-
ditional SCA1 transgenic mice. J Neurosci 24:8853–8861.
microscopic study. J Neurocytol 26:479–490.
22. Yamamoto A, Lucas JJ, Hen R (2000) Reversal of neuropathology and motor dysfunc-
tion in a conditional model of Huntington’s disease. Cell 101:57–66.
23. Yue S, Serra HG, Zoghbi HY, Orr HT (2001) The spinocerebellar ataxia type 1 protein,
ataxin-1, has RNA-binding activity that is inversely affected by the length of its
polyglutamine tract. Hum Mol Genet 10:25–30.
effects. Hum Mol Genet 11:1927–1937.
25. Gu X, et al. (2005) Pathological cell–cell interactions elicited by a neuropathogenic
form of mutant Huntingtin contribute to cortical pathogenesis in HD mice. Neuron
26. Garden GA, et al. (2002) Polyglutamine-expanded ataxin-7 promotes non-cell-
autonomous Purkinje cell degeneration and displays proteolytic cleavage in ataxic
transgenic mice. J Neurosci 22:4897–4905.
produces neurodegeneration by impairing glutamate transport. Nat Neurosci 9:1302–
28. Garcia-Segura LM, Perez J, Pons S, Rejas MT, Torres-Aleman I (1991) Localization of
insulin-like growth factor I (IGF-I)-like immunoreactivity in the developing and adult
rat brain. Brain Res 560:167–174.
gene expression in the brain. Endocrinology 139:65–71.
in the IGF system and effector on its own. J Endocrinol 172:423–440.
31. Roschier M, et al. (2001) Insulin-like growth factor binding protein 5 and type-1
insulin-like growth factor receptor are differentially regulated during apoptosis in
cerebellar granule cells. J Neurochem 76:11–20.
32. Humbert S, et al. (2002) The IGF-1/Akt pathway is neuroprotective in Huntington’s
disease and involves Huntingtin phosphorylation by Akt. Dev Cell 2:831–837.
33. Palazzolo I, et al. (2007) Akt blocks ligand binding and protects against expanded
polyglutamine androgen receptor toxicity. Hum Mol Genet 16:1593–1603.
34. Trejo JL, Carro E, Lopez-Lopez C, Torres-Aleman I (2004) Role of serum insulin-like
growth factor I in mammalian brain aging. Growth Horm IGF Res 14 Suppl A:S39–S43.
35. Stenvers KL, Lund PK, Gallagher M (1996) Increased expression of type 1 insulin-like
growth factor receptor messenger RNA in rat hippocampal formation is associated
with aging and behavioral impairment. Neuroscience 72:505–518.
37. Emamian ES, et al. (2003) Serine 776 of ataxin-1 is critical for polyglutamine-induced
disease in SCA1 transgenic mice. Neuron 38:375–387.
38. Chen HK, et al. (2003) Interaction of Akt-phosphorylated ataxin-1 with 14–3-3 medi-
ates neurodegeneration in spinocerebellar ataxia type 1. Cell 113:457–468.
www.pnas.org?cgi?doi?10.1073?pnas.0711257105 Gatchel et al.