Chapter #

Spinocerebellar ataxia type 2.

Daniel R. Scoles & Stefan M. Pulst

Department of Neurology, University of Utah, 175 North Medical Drive East, 5th Floor,
Salt Lake City, UT 84132, USA.
Abstract
Spinocerebellar ataxia type 2 (SCA2) is autosomal dominantly inherited and caused by CAG
repeat expansion in the ATXN2 gene. Because the CAG repeat expansion is localized to an
encoded region of ATXN2, the result is in an expanded polyglutamine (polyQ) tract in the ATXN2
protein. SCA2 is characterized by progressive ataxia, and slow saccades. No treatment for SCA2
exists. ATXN2 mutation causes gains of new or toxic functions for the ATXN2 protein, resulting in
abnormally slow Purkinje cell (PC) firing frequency and ultimately PC loss. This chapter describes
the characteristics of SCA2 patients briefly, and reviews ATXN2 molecular features and progress
toward the identification of a treatment for SCA2.

#.1 SCA2 Clinical characteristics

While patients with SCA2 possess many of the core clinical characteristics that define the SCAs
as a group of neurodegenerative disorders, SCA2 is a clinically distinct. Considered a hallmark
characteristic of any SCA, the most noticeable symptom of onset is gait ataxia. In SCA2, onset
also frequently although not always coincides with muscle cramping. Ataxia onset is then followed
by multiple other symptoms characteristic of cerebellar degeneration. For SCA2 these symptoms
include appendicular ataxia with instability of stance, dysarthria, and ocular signs including
nystagmus, and ocular dysmetria. The signs and symptoms of SCA2 are almost entirely of
cerebellar origin, with clearly defined involvement of cerebellar regions and associated cerebellar
circuits. However, one predominant ocular feature typical of SCA2, slow or absent saccades,
arises from degeneration of neurons of the oculomotor brainstem. Dystonia and myoclonus are
also frequent in patients with SCA2, as well as neuropathy, muscle spasticity, and frontal-
executive dysfunction [1-3].

SCA2 phenotype is one characterized by gait ataxia in most SCA2 patients, however variant
phenotypes have been defined. These variant phenotypes reside outside of the cerebellar
spectrum and include L-DOPA responsive parkinsonism and amyotrophic lateral sclerosis (ALS)
[3,4]. Patients with these variant phenotypes present with idiopathic forms of parkinsonism or
ALS. While ATXN2-associated parkinsonism and ALS presents with no symptoms of cerebellar
ataxia, more imaging data are necessary to define whether these variant phenotypes are
accompanied with cerebellar atrophy.

SCA2 and ALS

ATXN2 CAG repeat expansions are also associated with ALS-like motor phenotypes. For ATXN2
CAG repeats in the normal range for SCA2, between 27-33 repeats in length, a statistically
significant increased risk for ALS has been defined [4,5]. A meta-analysis of ATXN2 alleles
drawing on worldwide reporting of ALS, however, showed that ALS-risk only increased
significantly for CAG repeats >31. In these patients, the phenotype is indistinguishable from other
idiopathic forms of ALS. The causes of the ALS-like phenotypes in patients with ATXN2
expansions are not well described but may associate with the ATXN2 interacting proteins TDP-43
and FUS [4,6], since mutations in the genes encoding these proteins can cause ALS. ALS and
intermediately expanded ATXN2 connects functionally to the action of C9ORF72, since
aggregates partially depleted of C9ORF72 including intermediately expanded ATXN2 were
neurotoxic due to impaired autophagy [7]. However, it remains to be determined why intermediate
expanded ATXN2 increases risk of ALS in the absence of C9ORF72 mutations. In our meta-
analysis an 11-fold increased risk was observed for ATXN2 repeats of 32 [5]. Thus the rarest
ATXN2 alleles represent the highest risk for ALS. However, SCA2 patients with longer ATXN2
mutations can also present with ALS-like phenotypes [5,8].

#.2 Discovery of the ATXN2 gene

SCA2 was first described in India with the discovery of nine patient families [9]. Nearly two
decades later a large population of SCA2 patients was discovered in eastern Cuba [9]. Both
discoveries noted that the affected families were characterized by ataxia and other cerebellar
signs as well as slow saccades; clinical features that are now known to typify patients with SCA2.
In Cuba it was later noted that the prevalence of SCA2 was especially high compared to other
regions in the world, attributed to a founder population in the eastern part of the island where 4 out
of every 10,000 inhabitants of Holguin province has SCA2 [10,11]. The existence of the large
SCA2 populations in Cuba and elsewhere has aided the identification of the ATXN2 gene. Varying
age of onset (AO) in SCA2 pedigrees helped to establish anticipation in SCA2 of 14.4 ± 7.9 years
per generation strongly hinting that the ATXN2 mutation was likely a repeat expansion mutation
[12]. Mapping studies ultimately localized ATXN2 to Chr 12q24.12 following initial mapping to
chromosome 12 [13], and fine mapping to 12q24 [12]. The ATXN2 gene was identified in 1996
demonstrating the causative mutation as a CAG repeat expansion in the coding region of ATXN2
resulting in a polyglutamine expansion in the ATXN2 protein [14-16].

#.3 Molecular genetics of SCA2

Most commonly, the ATXN2 gene has 22 CAG repeats while ≥33 CAG repeats causes SCA2
[17]. SCA2 is characterized by anticipation with strong correlation between age of onset and CAG
repeat length (Fig. 1). The CAG repeat expansion is dynamic and unstable during meiosis with a
strong propensity for expansion.

The ATXN2 gene consists of 25 exons and spans a total of 147 megabase pairs (147,463 bp).
The ATXN2 transcript is 4,699 bp long with relatively small untranslated regions (162 bp 5’-UTR,
601 bp 3’-UTR). There are two in-frame start codons at the 5’-end of the sequence with the
second one located four codons upstream of the CAG repeat. Transcriptional studies have only
partly described which of these are utilized in translation. The predicted molecular weight for
ATXN2 when made from the first start codon is 144 kDa and ATXN2 made from the second start
codon is 17 kDa smaller. While western blot analyses typically produce a single ATXN2 protein
band consistent for use of the further-most upstream ATG, artificial luciferase tagged ATXN2
promoter constructs lacking the second ATG fail to express proteins [18]. Note that a smaller
approximately 42 kDa fragment of ATXN2 was observed in brain extracts from SCA2 patients and
SCA2 mice [19-21]. Huynh et al., 2000 identified a consensus aa sequence for caspase-3
cleavage at ATXN2 aa 396-399 that could explain the origin of this band [20]. ATXN2 is a
cytoplasmic protein that also localizes to the trans-Golgi network [22,23], and is a phosphorylated
protein with half-life of ≥21 hours [22]. ATXN2 transcription is also regulated by the ETS1
transcription factor [18], and might be altered by CAG repeat expansion since the ATXN2 CAG
repeat is located inside a CpG island [24]. The molecular features of the ATXN2 gene and the
encoded protein are summarized in Table 1.  

Macromolecules interacting ATXN2

The ATXN2 interacting proteins provided clues on the functions that ATXN2 controls. ATXN2
interacts with multiple RNA binding proteins (RBPs) suggesting ATXN2 has a role in RNA
metabolism. ATXN2 also interacts with staufen which controls stress granule formation and itself
interacts with RBPs. ATXN2 interactions with IP3R and with the RGS8 mRNA transcript
supporting ATXN2 roles in calcium homeostasis. ATXN2 also interacts with endophilins and
CIN85 indicating a function for ATXN2 in synaptic vesicle endocytosis. The ATXN2 interacting
proteins are summarized in Table 1, and a graphic representation of the binding sites within
ATXN2 is presented in Figure 2.

A2BP1/RBFOX1

A2BP1/RBFOX1 is a regulator of RNA alternative splicing. We discovered A2BP1 as an ATXN2

interacting protein by yeast two-hybrid screening. Secondary two-hybrid assays using protein
fragments determined that A2BP1 interacted with the C-terminal half of ATXN2 residues 760-
1313 and that the full-length A2BP1 protein interacted stronger than did the C-terminal A2BP1
fragment while an N-terminal A2BP1 fragment did not interact with ATXN2 [25]. A2BP1 labeled in
granules present in SCA2 patient dentate neurons and Purkinje neurons. A2BP1 functions in RNA
splicing suggested that ATXN2 may regulate alternative splicing in a tissue-specific manner or of
a subset of RNAs. A2BP1 was the first RNA binding protein discovered to interact with ATXN2
and in simultaneous work in our research group this led to our first attention on the Lsm and
LsmAD domains in located in the N-terminal region of ATXN2 [19] that are common in
spliceosomal small nuclear ribonucleoproteins (snRNPs) and function in RNA binding and protein-
protein interactions [26].

PABP & DDX6

Poly(A)-binding protein (PABP) interacts with the polyA end of mRNAs in the initiation of protein
translation [27]. Other interactions made by PABP are facilitated by the 12 amino acid PABP-
interacting motif 2 (PAM2) domain. A survey of multiple PAM2 proteins demonstrated that a
PAM2 domain in ATXN2 and a high level of conservation of the PAM2 domain among the proteins
[28]. A physical interaction between PABP and ATXN2 was demonstrated by yeast two hybrid
testing and co-immunoprecipitation [27]. PABP is a component of mammalian stress granules,
and ATXN2 and PABP colocalized in stress granules in heat-shock treated COS1 cells [27]. The
study was the first to demonstrate the localization of ATXN2 to stress granules. The same
research group further investigated ATXN2 in stress granules by characterizing its interaction with
the DEAD/H-box RNA helicase (DDX6) [29]. DDX6 is a stress granule protein that like PABP is
localized to stress granules as well as processing bodies (p-bodies). ATXN2 was shown to
directly interact with DDX6 by way of the Lsm and LsmAD domains in ATXN2 by yeast two-hybrid
interaction testing [30]. Upon identifying DDX6 as an ATXN2 interacting protein the investigators
further demonstrated that ATXN2 localized to both stress granules and processing bodies (p-
bodies). ATXN2 also interacts with polyribosomes which are also known to be regulated by RNA
granule formation [29].

Interaction between ATXN2 and PABP appears to connect functionally to the control of translation
by ATXN2 involving mTOR signaling. Increased ATXN2 mRNA was observed in SH-SY5Y cells
stressed by serum starvation [31]. The authors also showed sequestration of PABP and proteins
of the cap-binding complex with ATXN2 in stress granules in mouse embryo fibroblasts (MEFs)
stressed with arsenite [31]. A connection between ATXN2 and mTOR signaling was further
confirmed by demonstrating increased phosphorylation of S6 and 4EBP1 in MEFs null for ATXN2,
as well as elevated ATXN2 mRNA abundance in SH-SY5Y cells treated with the mTOR inhibitor
rapamycin but not the PI3-kinase inhibitor LY294002 [31], suggesting the presence of a
compensatory feed-back mechanism activating ATXN2 when mTOR is inhibited.

TDP-43 & FUS

Both TDP-43 and FUS are RBDs that are mutated in amyotrophic lateral sclerosis (ALS). The
identification of ATXN2 as an interactor with TDP-43 was presented along with the discovery that
moderate expansions in the ATXN2 gene CAG repeat are associated with increased risk for ALS
[4]. More on ATXN2 and ALS is discussed below. The interaction between TDP-43 and ATXN2
was demonstrated by yeast two-hybrid interaction testing, and in HEK293 cells by
coimmunoprecipitation (co-IP) of overexpressed GFP-TDP-43 fragments with endogenous
ATXN2 and by immunofluorescent colocalization. In co-IP tests including RNAse or including
TDP-43 proteins mutated to abolish RNA binding the TDP-43-ATXN2 interaction was abolished,
demonstrating that the interaction is RNA dependent. FUS was also demonstrated to interact with
ATXN2 [6]. Both TDP-43 and FUS have been characterized in RNA granules containing ATXN2
[32] suggesting a pathogenic connection for ATXN2 in increased ALS risk is associated with
abnormal stress granule function.

Parkin

To investigate a functional connection that might explain why Parkinsonism is sometimes

observed in SCA2, Huynh et al. (2007) tested ATXN2 as an interacting protein with Parkin [33].
Parkin directly interacted with the ATXN2 N-terminal domain (residues 1-396) when hemagglutinin
(HA)-tagged Parkin was pulled down with an anti-HA antibody in HEK293 cells expressing GFP
fused to full-length ATXN2 or N- or C-terminal fragments of ATXN2 [33]. The interaction was
verified for ATXN2 proteins with Q22 as well as expanded polyglutamine tracts of Q58 and Q104.
Parkin, an E3 ubiquitin ligase, ubiquitinated full-length ATXN2 more efficiently than ATXN2 N-
terminal fragments. ATXN2 ubiquitination by Parkin was more pronounced when ATXN2 was
polyglutamine expanded but less efficient with Parkin was C289G mutated. The induced
overexpression of Parkin in tetracycline inducible PC12 cells was associated with increased
turnover of the ATXN2 protein. ATXN2 and Parkin colocalized in cytoplasmic structures of
Purkinje cells from normal (non-SCA2) individuals [33]. The interaction between Parkin and
ATXN2 was independently confirmed by coimmunoprecipition of polyglutamine expanded ATXN2
with Parkin from the cerebella of ATXN2-CAG42 knock-in mice [34]. Note that the latter study
also demonstrated that the E3 ubiquitin ligase Fbxw8 also coimmunoprecipitated with ATXN2.

Staufen
Recently we demonstrated that ATXN2 interacts with Staufen1. Staufen is a key regulator of
stress granule formation. We demonstrated that staufen expression is increased in SCA2 derived
patient fibroblasts, lymphoblasts, iPSCs, and in the cerebella of our ATXN2-Q127 transgenic and
our ATXN2-Q72 BAC mice. The result of elevated staufen expression in these systems is
constitutively present stress granules. The identification of staufen as an interacting protein with
ATXN2 whose expression is elevated upon ATXN2 mutation demonstrates a functional role for
ATXN2 in either staufen mediated decay, stress granule mediated mRNA processing, or stress
granule mediated dendritic mRNA trafficking for localized expression control.

RGS8 mRNA & IP3R

ATXN2 interacts with Rgs8 mRNA and IP3R supporting roles for ATXN2 in calcium homeostatsis.
We determined that Rgs8 expression is reduced in the cerebella of SCA2 mice by transcriptome
analysis, and verified RGS8 reduction in SCA2 patient lymphoblast cells. We also demonstrated
that Rgs8 translation is reduced in the presence of mutant ATXN2 using rabbit reticulocyte in vitro
translation assays. Thus, reduced Rgs8 could be the result of mRNA degradation, as well as
RGS8 mRNA translation inhibition perhaps mediated by sequestration in stress granules. RGS8
inhibition could impact calcium levels in Purkinje cells since RGS8 is believed to be an inhibitor of
mGluR1. The role of mGlur1 in the normal functioning of Purkinje neuron and motor coordination
is well described in a review by Hartmann et al. (2011) [35]. In Purkinje cells, mGluRs produce two
distinct signals including a local dendritic Ca2+ signal and a slow excitatory postsynaptic potential:
The dendritic Ca2+ signal originates through Ca2+ release from the ER mediated by the inositol-
triphosphate receptor type 1 (IP3R). The slow excitatory postsynaptic potentials are mediated by
Ca2+ influx via the transient receptor potential cation channel TRPC3 that is gated by
diacylglycerol (DAG) and IP3R [35]. IP3R is abnormally activated upon interaction by mutant
ATXN2, resulting in abnormal release of Ca2+ from intracellular stores [36]. The Bezprozvanny
group verified that IP3R specifically interacts with the polyglutamine expanded ATXN2 protein but
not the normal ATXN2 protein [36]. The interaction was demonstrated between endogenous
ATXN2 and overexpressed GST-IP3R in COS7 cells using a pull-down assay. A second assay
using cerebellar homogenates from Pcp2-ATXN2-Q58 mice demonstrated that polyglutamine
expanded ATXN2 co-precipitated with greater abundance of radiolabeled IP3 than did wildtype
ATXN2, consistent with an interaction between mutant ATXN2 and IP3R [36]. Further
experiments on modulating IP3R function for modifying SCA2 mouse phenotypes are described
below.

Endophilins and CIN85

Endophilin and CIN85 are proteins that function along with Cbl in endocytosis of cell surface
receptor tyrosine kinases [37]. The Huntington disease protein huntingtin, another polyglutamine
disease protein, interacts with endophilin A3 resulting in abnormal sequestration of proteins of
endocytic vesicle systems [38,39]. Therefore, to test whether ATXN2 could interact with
endophilins, Ralser et al. (2005) performed two-hybrid interaction tests that demonstrated ATXN2
direct binding with both endophilin A1 and endophilin A3 [40]. The interaction was mediated by the
SH3 domain binding motif 2 (SBM2) in ATXN2, and the investigators also showed that ATXN2
failed to interact with endophilin A2. The study also demonstrated competitive binding between
ATXN2 and huntingtin for endophilin A3 in the yeast two-hybrid system. Another study
investigated ATXN2 interactions with endophilin A1 and endophilin A3 by GST pull-down tests
showing that the endophilins interacted with other ATXN2 protein regions not including the SBM2
domain [41]. Extensive cytoplasmic colocalization of ATXN2 with endophilins A1 and A3 was also
demonstrated by immunofluorescent labeling of HEK293 and SH-SY5Y cells [40,41]. Co-
immunoprecipitation assays demonstrated ATXN2 exists in complexes containing endophilin A3,
CIN85, Cbl, and EGF receptor (EGFR) [41]. Overexpression of ATXN2 in CHO cells inhibited
EGF-stimulated EGFR internalization demonstrating a functional role of ATXN2 in endocytosis
[41]. Endocytosis controlled by the Endophilin-CIN85-Cbl complex is mediated by clathrin-coated
pits. A putative site for clathrin binding in ATXN2 was described by Huynh et al., 2000 at aa 414-
416 [20]. However Turnbull et al. (2004) could demonstrate no co-localization between ATXN2
and clathrin-coated pits or vesicles [22].

Other ATXN2 interaction studies

Various studies have demonstrated proteins with which ATXN2 coimmunoprecipitated without
formally testing direct interactions. Discussed briefly in the PABP paragraph above, Lastres-
Becker et al. (2016) demonstrated that ATXN2 coimmuniprecipitated with TIA1, eIF3B, eIF4G,
eIF4A1 and S6 from HEK293 cells treated with or without arsenite [31]. Blokhuis et al., 2016
characterized the ATXN2 interactome in Neuro2A cells using mass spectrometry with validations
performed by coimmunoprecipitation [42]. Key interacting proteins verified in
coimmunoprecipitation experiments included Fmrp, Upf1, Caprin1, HuD, Pabpc4, and Dhx9. The
investigators also produced interactomes for Fus and Tdp43 and presented interactions shared
among these proteins and ATXN2 [42].

Studies of the ATXN2 yeast homolog Pbp1 suggest other proteins likely to interact with the
ATXN2 protein. The PAS kinase 1 (Psk1) was shown to interact with the C-terminal half of Pbp1
resulting in Pbp1 phosphorylation proximal to the interaction [43]. Another study of Pbp1
demonstrated interactions with Lsm12 and Pbp4 in addition to the yeast homologs of PABP and
DDX6 [44].

#.4 SCA2 mouse models

We and others have produced multiple SCA2 mouse models, including transgenic and knockout
models [20,45-51]. In this section we describe these mice as well as the SCA2 mice that have
been developed by other groups. Note that a recent review comprehensively describes each of
these mice [52].

Pcp2-ATXN2 transgenic mice

We have made two types of Pcp2-ATXN2 transgenic mice, including one with ATXN2-CAG58
(Q58) and another with ATXN2-CAG127 (Q127). Both of these mice have ATXN2 expressed
under the control of the Purkinje cell protein 2 (Pcp2)/L7 promoter. These mice are
characterized by age-dependent molecular, motor, and electrophysiological phenotypes.
Rotorod testing demonstrated ATXN2-dose dependent motor phenotype for ATXN2-Q58 mice
first observed at six months of age, and Purkinje cells in these mice contained cytoplasmic but
not nuclear inclusion bodies [20]. The ATXN2-Q58 mouse was also used in studies
demonstrating dantrolene treatment could restore ATXN2 mouse motor phenotypes [36]. This is
discussed in further detail in the section on calcium homeostasis below. ATXN2-Q127 mice also
has Purkinje cells with cytoplasmic inclusion bodies but with the longer repeat length we have
observed the motor phenotype as early as eight weeks of age [45]. The Auburger group has
also created an ATXN2-CAG42 knock-in mouse by replacing the single CAG in the mouse
Atxn2 gene with an expanded CAG42 repeat [50]. The ATXN2-CAG42 mouse was
characterized for how mutant ATXN2 alters PABPC1 solubility and availability for functions in
RNA metabolism.

SCA2 BAC transgenic mice

SCA2 BAC mice possess the entire 176 kb ATXN2 gene region including 16 kb upstream
sequence and 2.5 kb downstream sequence. Presently we have two SCA2 BAC lines including
alleles expressing ATXN2-Q22 normal length and ATXN2-Q72 expanded [46]. The Q22 line has
no motor, transcriptomic or neurophysiological phenotype. However, the Q72 line has a
progressive onset of its motor phenotype determined using the accelerating rotarod that is
mimicked by progressive reduction of the expression of various neuronal and Purkinje cell
specific genes beginning at 8 wks of age [46]. More recently we have identified changes in
Purkinje cell firing frequencies in the SCA2-Q72 BAC mice compared to wildtype littermates
present for mice age six and 12 months but not mice four months of age (unpublished
observation). This demonstrates that the neurophysiological phenotype of the BAC-Q72 mice
appears later than for the ATXN2-Q127 mice mirroring the later motor phenotype observed in
the BAC-Q72 mice. The delayed onset of SCA2 phenotypes in the BAC-Q72 mice is due to the
lower expression from the native ATXN2 promoter as well as the shorter Q72 repeat compared
to the ATXN2-Q127 mouse. Cerebellar molecular phenotype changes determined by qPCR and
RNA-seq were largely similar to those observed in ATXN2-Q127 mice [45,46].

ATXN2-Q75 transgenic mice

ATXN2-Q75 mice are transgenic for the ATXN2 cDNA under transcriptional control by the native
ATXN2 promoter, and include ATXN2 with 75 CAG repeats [51]. Ubiquitous transgene
expression was observed, and hemizygous mice were ataxic by 12 weeks of age in rotarod
tests, corresponding with abnormal Purkinje cell morphology.

Atxn2 knockout mice

Our group and the Auburger group have both produced ATXN2 knockout mice and we have
demonstrated key characteristics that are common to both, which include viable mice with
marked obesity and the lack of any significantly debilitating neuropathology [48,49]. We have
also demonstrated normal Purkinje cell physiology in ATXN2 knockout mice (unpublished
observation) but these mice are also characterized by abnormal fear-related behavior [47]. The
Auburger group demonstrated that ATXN2 knockout mice had abnormally low insulin receptor
expression in both the cerebellar and the liver and concluded that these molecular changes
associated with the onset of obesity [49]. These investigators further evaluated their ATXN2
knockout mouse employing microarray analysis revealing increased expression of transcription
factors but overall lower translation [53]. The lack of neuropathology in these mice supports the
concept for developing SCA2 therapeutics that target the total expression of ATXN2 such as we
are with ATXN2 compounds and ATXN2 antisense oligonucleotides (ASOs), described below.

#.5 Transcriptome analyses

Multiple studies on the cerebellar transcriptomes have been conducted using SCA2 mice. In our
initial study we used cerebellar RNAs isolated from wildtype and BAC-ATXN2-Q72 mice at ages 1
day, 3 weeks and 6 weeks. In day 1 mice we observed ~200 transcripts that were significantly
dysregulated, and more transcripts were altered in the older animals [46]. Most transcripts were
reduced in abundance in SCA2 mice vs willdtype mice. We also performed transcriptome analysis
using Pcp2-ATXN2-Q127 mice for the purpose of identifying pathways commonly modified in
these mice and BAC-ATXN2-Q127 mice [46]. It was these studies that resulted in the
identification of significant reductions in the RGS8 mRNA as described above. Other genes that
were significantly reduced in both of these SCA2 mouse models included Pcp2, Fam107b, others.
Pathways that we identified that are altered in both Pcp2-ATXN2-Q127 and BAC-ATXN2-Q127
mice include glutamate signaling, calcium signaling, others [46]. We have also begun to compare
the transcriptomes of these mice with that of age-matched knockout mice. Unlike the
transcriptomes in the SCA2 mice we observed that few transcripts were altered in the ATXN2
knockout mice (unpublished observations). Similarly, the Auburger group compared
transcriptomes of ATXN2-CAG42 transgenic mice with ATXN2 knockout mice using a
microarray analysis approach, demonstrating overlapping abnormalities in calcium homeostasis
pathways [54].

#.6 SCA2 therapeutics

Despite presence of many ATXN2 interactors we still lack information on which ATXN2 functions
are targetable to prevent SCA2 pathogenesis or to delay SCA2 progression. The concept of two
different approaches for developing SCA2 therapeutics discussed here closely follows that
developed in a recent review on precision medicine for the spinocerebellar ataxias [55]. One
approach is to develop the known functions for ATXN2 as therapeutic targets for SCA2, including
glutamate signaling, calcium homeostatsis, and RNA metabolism (Fig. 3). Another promising
approach is to target ATXN2 expression directly, because like for other SCAs caused by
mutations leading to polyglutamine expansions SCA2 is characterized predominantly by a toxic
gain of function. We are making efforts to develop SCA2 therapeutics that target ATXN2
functionally and we are also developing antisense oligonucleotides that lower overall ATXN2
expression.

Therapies targeting ATXN2-related SCA2 pathways

The approach that we have taken for developing therapeutics that target ATXN2-related SCA2
pathways is to target pathways leading to abnormal rise of cytoplasmic Ca2+. These efforts were
initiated upon the identification that the mutant but not the wildtype ATXN2 protein interacted with
the inositol-triphosphate receptor type 1 (ITPR1), also described above [36]. Targeting the
Ca2+pathway in SCA2 is in line with the notion that defective Ca2+signaling underlies most
neurodegenerative diseases [56]. ITPR1 mutations or haploinsufficiency are also causative for
SCA15/16 [57,58]. ITPR1 is a calcium channel located on the endoplasmic reticulum membrane
controlling the release of intracellular Ca2+ stores, and is expressed highly in PCs. Mice harboring
ATXN2-Q58 were characterized with increased Ca2+ release from the endoplasmic reticulum
associated with molecular layer thinning and Purkinje cell loss [20,36]. Blocking of the functionally
coupled ryanodine receptor with dantrolene reduced abnormal calcium release and cell death in
culture [36]. Liu et al. (2009) also demonstrated that SCA2 motor phenotypes of ATXN2-Q58 mice
as well as improved Purkinje cell survival was delayed by oral treatment of ATXN2-Q58 mice with
dantrolene. Tests included the beam walk and accelerating rotarod [36].

We have also begun to investigate mGlur1 as a therapeutic target for SCA2. As described above
we demonstrated that RGS8 expression is reduced in an age-dependent manner in SCA2 mice.
RGS8 is a putative regulator of mGlur1 and its reduced expression is predicted to deregulate
mGlur1 [46]. We have now used our Pcp2-ATXN2-Q127 model to replicate the in vitro findings
and show that the mGlur1 agonist DHPG enhances firing frequency of Pcp2-ATXN2-Q127 mouse
PCs accompanied by abnormally elevated intracellular Ca2+ at specific PC firing rates (Pulst and
Otis, unpublished). ATXN2 expression itself may be regulated by intracellular calcium and
mGluR1. These data suggest that mGlur1 antagonists could be therapeutic for SCA2.

ATXN2 ASO therapeutics

Antisense oligonucleotides (ASOs) represent a promising approach for treating SCA2. SCA2 is
characterized by a gain of toxic function thus we hypothesize that lowering ATXN2 expression
would be therapeutic for SCA2. Reducing total expression as a therapeutic approach is
supported for polyQ diseases by several observations. In SCA2 patients two copies of the
mutant ATXN2 allele can be accompanied by earlier age of onset and more rapid disease
progression [59]. The importance of gene dosage is further supported by studies on mice.
Doubling of gene dosage in transgenic ATXN2-Q58 mice led to earlier onset of abnormal
rotarod performance [20]. Studies using tetracycline-regulated promoters in HD, SCA1 or SCA3
mice have demonstrated reversibility of phenotypes even after disease onset [60-63]. Another
study showed that intracerebellar injection of shRNA AAV virus reduced transgene expression
improved motor coordination, restored cerebellar morphology and resolved ataxin-1 inclusions
in Purkinje neurons (PNs) of SCA1 mice [64]. Recently, ASOs have proven useful for the
treatment of spinal muscular atrophy and SOD-ALS [65,66], and newer phase 1 clinical trials
have been initiated using ASOs for the treatment of myotonic dystrophy (DM1) and Huntington
disease.

Our ASO approach to therapeutics is being conducted collaboratively with Ionis Pharmaceuticals
utilizing modified 2’-MOE-gapmer ASOs. The 2’-MOE-gapmer are 20 bp in length, are
phosphorothioate throughout, and have a 2’-O-methoxyethyl group (MOE) on the terminal 5 bps
at each end of the oligonucleotide [67]. These modifications prevent degradation by nucleases
and the MOE chemical modifications also aid in increasing specificity of target mRNA interaction,
supporting target degradation by RNase-H [68].

The SCA2 ASOs that we are developing are for non-allele-specific targeting of ATXN2 unlike the
approach undertaken for Huntington’s disease. In Huntington’s disease ASOs are made to target
SNPs in linkage with CAG repeat expanded alleles [69,70] because HTT knockout in mice
disrupts neuronal development [71]. In our approach we permit the ASO to target the mutant
ATXN2 allele as well as the non-mutant allele because knockout of the ATXN2 gene in mice well
is tolerated and associated with no neurodegeneration [47,48]. However, progress on developing
non-allele-specific RNAi therapeutics for HD have had favorable outcomes in mice and non-
human primates [72,73].

In collaboration with Ionis Pharmaceuticals we have produced ATXN2 ASOs that lower human
ATXN2 expression in both our Pcp2-ATXN2-Q127 mouse model as well as our BAC-ATXN2-Q72
mouse model. We have observed as much as 80% ATXN2 reduction when delivered to mice by
introcerebroventricular (ICV) injection to the right lateral ventricle, and we have observed no
cytotoxicity indicated by following AIF1 and GFAP expression post injection. We have employed
our most promising lead ASO, designated ASO7, in a blinded preclinical treatment trial using
Pcp2-ATXN2-Q127 mice (Fig. 4). Mice treated at 8 weeks of age were tested at different
treatment timepoints on the accelerating rotarod demonstrating delayed progression of the SCA2
motor phenotypes in both mouse models. At the endpoint we determined the cerebellar
expression of ATXN2, Rgs8 and Pcp2 by qPCR and Western blotting demonstrating prolonged
ATXN2 reduction and increases in Rgs8 and Pcp2 expression (not shown). Finally, subsets of
mice were tested to determine the effect of the ASO7 treatments on Purkinje cell physiology. We
observed that ASO7 treatment could restore the mean PC firing frequency to that unlike observed
in age matched mice. Finally, we also have ongoing studies to characterize the transcriptomes of
SCA2 mice treated with or without ATXN2 ASO7. Information resulting from this work might
indicate new pathways that can be exploited for the treatment of ALS associated with moderate
expansions in the ATXN2 gene.

#.7 Conclusions
SCA2 is a debilitating disorder for which there is no treatment. Research by numerous teams on
SCA2 has resulted in the identification of multiple interacting proteins that have given rise to clues
about ATXN2 function. Additionally, multiple SCA2 transgenic mouse models and ATXN2
knockout models have been studied giving rise to understanding on pathways dysregulated in
SCA2 mice. Collectively, these studies have aided understanding on ATXN2 function and have
indicated possible pathways that can be targeted in order to delay SCA2 progression. Progress
toward developing drugs targeting calcium signaling and related pathways is accumulating. But
even with this increased knowledge of ATXN2 function, the opportunity to target ATXN2 directly
using antisense oligonucleotides remains a primary goal of our research group for treating SCA2,
garnered by positive preliminary data in SCA2 mice and recent data demonstrating that ASOs are
tolerated in humans and effective for SMA and SOD ALS. Preliminary data demonstrating that
SCA2 ASOs lower ATXN2 expression in mouse spinal cord also lend hope toward developing
ATXN2 ASOs for SCA2 ALS and perhaps ALS in a more generalized manner.

FUNDING
This work was supported by the Carmen and Louis Warschaw Endowment Fund, the National
Ataxia Foundation, grants R01NS33123, R56NS33123 and R37NS033123 from the National
Institutes of Neurological Disorders and Stroke to SMP, the Noorda foundation to SMP, and
grants RC4NS073009 and R21NS081182 to DRS and SMP. SMP received grant support from
the Target ALS Foundation.

CONFLICT OF INTEREST STATEMENT

SMP is a consultant for Progenitor Life Sciences.

ACKNOWLEDGEMENTS
We thank Duong P. Huynh and Sharan Paul for editing the chapter. We thank Darren Ames and
Lance Pflieger for assistance with transcriptome analysis computations.
REFERENCES
1. Geschwind DH, Perlman S, Figueroa CP, Treiman LJ, Pulst SM (1997) The prevalence and
wide clinical spectrum of the spinocerebellar ataxia type 2 trinucleotide repeat in patients with
autosomal dominant cerebellar ataxia. Am J Hum Genet 60: 842-850.

2. Ashizawa T, Figueroa KP, Perlman SL, Gomez CM, Wilmot GR, Schmahmann JD, et al.
(2013) Clinical characteristics of patients with spinocerebellar ataxias 1, 2, 3 and 6 in the US; a
prospective observational study. Orphanet J Rare Dis 8: 177.

3. Gwinn-Hardy K, Chen JY, Liu HC, Liu TY, Boss M, Seltzer W, et al. (2000) Spinocerebellar
ataxia type 2 with parkinsonism in ethnic Chinese. Neurology 55: 800-805.

4. Elden AC, Kim HJ, Hart MP, Chen-Plotkin AS, Johnson BS, Fang X, et al. (2010) Ataxin-2
intermediate-length polyglutamine expansions are associated with increased risk for ALS.
Nature 466: 1069-1075.

5. Neuenschwander AG, Thai KK, Figueroa KP, Pulst SM (2014) Amyotrophic lateral sclerosis
risk for spinocerebellar ataxia type 2 ATXN2 CAG repeat alleles: a meta-analysis. JAMA Neurol
71: 1529-1534.

6. Farg MA, Soo KY, Warraich ST, Sundaramoorthy V, Blair IP, Atkin JD (2013) Ataxin-2
interacts with FUS and intermediate-length polyglutamine expansions enhance FUS-related
pathology in amyotrophic lateral sclerosis. Hum Mol Genet 22: 717-728.

7. Sellier C, Campanari ML, Julie Corbier C, Gaucherot A, Kolb-Cheynel I, Oulad-Abdelghani M,

et al. (2016) Loss of C9ORF72 impairs autophagy and synergizes with polyQ Ataxin-2 to induce
motor neuron dysfunction and cell death. EMBO J.

8. Fischbeck KH, Pulst SM (2011) Amyotrophic lateral sclerosis and spinocerebellar ataxia 2.
Neurology.

9. Wadia NH, Swami RK (1971) A new form of heredo-familial spinocerebellar degeneration

with slow eye movements (nine families). Brain 94: 359-374.
10. Pulst SM, Santos N, Wang D, Yang H, Huynh D, Velazquez L, et al. (2005) Spinocerebellar
ataxia type 2: polyQ repeat variation in the CACNA1A calcium channel modifies age of onset.
Brain 128: 2297-2303.

11. Velazquez-Perez L, Rodriguez-Labrada R, Canales-Ochoa N, Montero JM, Sanchez-Cruz

G, Aguilera-Rodriguez R, et al. (2014) Progression of early features of spinocerebellar ataxia
type 2 in individuals at risk: a longitudinal study. Lancet Neurol 13: 482-489.

12. Pulst SM, Nechiporuk A, Starkman S (1993) Anticipation in spinocerebellar ataxia type 2.
Nat Genet 5: 8-10.

13. Gispert S, Twells R, Orozco G, Brice A, Weber J, Heredero L, et al. (1993) Chromosomal
assignment of the second locus for autosomal dominant cerebellar ataxia (SCA2) to
chromosome 12q23-24.1. Nat Genet 4: 295-299.

14. Pulst SM, Nechiporuk A, Nechiporuk T, Gispert S, Chen XN, Lopes-Cendes I, et al. (1996)
Moderate expansion of a normally biallelic trinucleotide repeat in spinocerebellar ataxia type 2.
Nat Genet 14: 269-276.

15. Sanpei K, Takano H, Igarashi S, Sato T, Oyake M, Sasaki H, et al. (1996) Identification of
the spinocerebellar ataxia type 2 gene using a direct identification of repeat expansion and
cloning technique, DIRECT. Nat Genet 14: 277-284.

16. Imbert G, Saudou F, Yvert G, Devys D, Trottier Y, Garnier JM, et al. (1996) Cloning of the
gene for spinocerebellar ataxia 2 reveals a locus with high sensitivity to expanded
CAG/glutamine repeats. Nat Genet 14: 285-291.

17. Fernandez M, McClain ME, Martinez RA, Snow K, Lipe H, Ravits J, et al. (2000) Late-onset
SCA2: 33 CAG repeats are sufficient to cause disease. Neurology 55: 569-572.

18. Scoles DR, Pflieger LT, Thai KK, Hansen ST, Dansithong W, Pulst SM (2012) ETS1
regulates the expression of ATXN2. Hum Mol Genet 21: 5048-5065.

19. Huynh DP, Del Bigio MR, Ho DH, Pulst SM (1999) Expression of ataxin-2 in brains from
normal individuals and patients with Alzheimer's disease and spinocerebellar ataxia 2. Ann
Neurol 45: 232-241.
20. Huynh DP, Figueroa K, Hoang N, Pulst SM (2000) Nuclear localization or inclusion body
formation of ataxin-2 are not necessary for SCA2 pathogenesis in mouse or human. Nat Genet
26: 44-50.

21. Koyano S, Uchihara T, Fujigasaki H, Nakamura A, Yagishita S, Iwabuchi K (1999) Neuronal

intranuclear inclusions in spinocerebellar ataxia type 2: triple-labeling immunofluorescent study.
Neurosci Lett 273: 117-120.

22. Turnbull VJ, Storey E, Tarlac V, Walsh R, Stefani D, Clark R, et al. (2004) Different ataxin-2
antibodies display different immunoreactive profiles. Brain Res 1027: 103-116.

23. Huynh DP, Yang HT, Vakharia H, Nguyen D, Pulst SM (2003) Expansion of the polyQ
repeat in ataxin-2 alters its Golgi localization, disrupts the Golgi complex and causes cell death.
Hum Mol Genet 12: 1485-1496.

24. Aguiar J, Santurlidis S, Nowok J, Alexander C, Rudnicki D, Gispert S, et al. (1999)

Identification of the physiological promoter for spinocerebellar ataxia 2 gene reveals a CpG
island for promoter activity situated into the exon 1 of this gene and provides data about the
origin of the nonmethylated state of these types of islands. Biochem Biophys Res Commun 254:
315-318.

25. Shibata H, Huynh DP, Pulst SM (2000) A novel protein with RNA-binding motifs interacts
with ataxin-2. Hum Mol Genet 9: 1303-1313.

26. Hermann H, Fabrizio P, Raker VA, Foulaki K, Hornig H, Brahms H, et al. (1995) snRNP Sm
proteins share two evolutionarily conserved sequence motifs which are involved in Sm protein-
protein interactions. EMBO J 14: 2076-2088.

27. Ralser M, Albrecht M, Nonhoff U, Lengauer T, Lehrach H, Krobitsch S (2005) An integrative

approach to gain insights into the cellular function of human ataxin-2. J Mol Biol 346: 203-214.

28. Albrecht M, Lengauer T (2004) Survey on the PABC recognition motif PAM2. Biochem
Biophys Res Commun 316: 129-138.

29. Satterfield TF, Pallanck LJ (2006) Ataxin-2 and its Drosophila homolog, ATX2, physically
assemble with polyribosomes. Hum Mol Genet 15: 2523-2532.
30. Nonhoff U, Ralser M, Welzel F, Piccini I, Balzereit D, Yaspo ML, et al. (2007) Ataxin-2
interacts with the DEAD/H-box RNA helicase DDX6 and interferes with P-bodies and stress
granules. Mol Biol Cell 18: 1385-1396.

31. Lastres-Becker I, Nonis D, Eich F, Klinkenberg M, Gorospe M, Kotter P, et al. (2016)

Mammalian ataxin-2 modulates translation control at the pre-initiation complex via PI3K/mTOR
and is induced by starvation. Biochim Biophys Acta 1862: 1558-1569.

32. Nihei Y, Ito D, Suzuki N (2012) Roles of ataxin-2 in pathological cascades mediated by TAR
DNA-binding protein 43 (TDP-43) and Fused in Sarcoma (FUS). J Biol Chem 287: 41310-
41323.

33. Huynh DP, Nguyen DT, Pulst-Korenberg JB, Brice A, Pulst SM (2007) Parkin is an E3
ubiquitin-ligase for normal and mutant ataxin-2 and prevents ataxin-2-induced cell death. Exp
Neurol 203: 531-541.

34. Halbach MV, Stehning T, Damrath E, Jendrach M, Sen NE, Basak AN, et al. (2015) Both
ubiquitin ligases FBXW8 and PARK2 are sequestrated into insolubility by ATXN2 PolyQ
expansions, but only FBXW8 expression is dysregulated. PLoS One 10: e0121089.

35. Hartmann J, Henning HA, Konnerth A (2011) mGluR1/TRPC3-mediated Synaptic

Transmission and Calcium Signaling in Mammalian Central Neurons. Cold Spring Harb
Perspect Biol 3.

36. Liu J, Tang TS, Tu H, Nelson O, Herndon E, Huynh DP, et al. (2009) Deranged calcium
signaling and neurodegeneration in spinocerebellar ataxia type 2. J Neurosci 29: 9148-9162.

37. Soubeyran P, Kowanetz K, Szymkiewicz I, Langdon WY, Dikic I (2002) Cbl-CIN85-

endophilin complex mediates ligand-induced downregulation of EGF receptors. Nature 416:
183-187.

38. Sittler A, Walter S, Wedemeyer N, Hasenbank R, Scherzinger E, Eickhoff H, et al. (1998)

SH3GL3 associates with the Huntingtin exon 1 protein and promotes the formation of polygln-
containing protein aggregates. Mol Cell 2: 427-436.
39. Qin ZH, Wang Y, Sapp E, Cuiffo B, Wanker E, Hayden MR, et al. (2004) Huntingtin bodies
sequester vesicle-associated proteins by a polyproline-dependent interaction. J Neurosci 24:
269-281.

40. Ralser M, Nonhoff U, Albrecht M, Lengauer T, Wanker EE, Lehrach H, et al. (2005) Ataxin-2
and huntingtin interact with endophilin-A complexes to function in plastin-associated pathways.
Hum Mol Genet 14: 2893-2909.

41. Nonis D, Schmidt MH, van de Loo S, Eich F, Dikic I, Nowock J, et al. (2008) Ataxin-2
associates with the endocytosis complex and affects EGF receptor trafficking. Cell Signal 20:
1725-1739.

42. Blokhuis AM, Koppers M, Groen EJ, van den Heuvel DM, Dini Modigliani S, Anink JJ, et al.
(2016) Comparative interactomics analysis of different ALS-associated proteins identifies
converging molecular pathways. Acta Neuropathol 132: 175-196.

43. DeMille D, Badal BD, Evans JB, Mathis AD, Anderson JF, Grose JH (2015) PAS kinase is
activated by direct SNF1-dependent phosphorylation and mediates inhibition of TORC1 through
the phosphorylation and activation of Pbp1. Mol Biol Cell 26: 569-582.

44. Swisher KD, Parker R (2010) Localization to, and effects of Pbp1, Pbp4, Lsm12, Dhh1, and
Pab1 on stress granules in Saccharomyces cerevisiae. PLoS One 5: e10006.

45. Hansen ST, Meera P, Otis TS, Pulst SM (2013) Changes in Purkinje cell firing and gene
expression precede behavioral pathology in a mouse model of SCA2. Hum Mol Genet 22: 271-
283.

46. Dansithong W, Paul S, Figueroa KP, Rinehart MD, Wiest S, Pflieger LT, et al. (2015) Ataxin-
2 regulates RGS8 translation in a new BAC-SCA2 transgenic mouse model. PLoS Genet 11:
e1005182.

47. Huynh DP, Maalouf M, Silva AJ, Schweizer FE, Pulst SM (2009) Dissociated fear and
spatial learning in mice with deficiency of ataxin-2. PLoS One 4: e6235.

48. Kiehl TR, Nechiporuk A, Figueroa KP, Keating MT, Huynh DP, Pulst SM (2006) Generation
and characterization of Sca2 (ataxin-2) knockout mice. Biochem Biophys Res Commun 339: 17-
24.
49. Lastres-Becker I, Brodesser S, Lutjohann D, Azizov M, Buchmann J, Hintermann E, et al.
(2008) Insulin receptor and lipid metabolism pathology in ataxin-2 knock-out mice. Hum Mol
Genet 17: 1465-1481.

50. Damrath E, Heck MV, Gispert S, Azizov M, Nowock J, Seifried C, et al. (2012) ATXN2-
CAG42 sequesters PABPC1 into insolubility and induces FBXW8 in cerebellum of old ataxic
knock-in mice. PLoS Genet 8: e1002920.

51. Aguiar J, Fernandez J, Aguilar A, Mendoza Y, Vazquez M, Suarez J, et al. (2006)

Ubiquitous expression of human SCA2 gene under the regulation of the SCA2 self promoter
cause specific Purkinje cell degeneration in transgenic mice. Neurosci Lett 392: 202-206.

52. Alves-Cruzeiro JM, Mendonca L, Pereira de Almeida L, Nobrega C (2016) Motor

Dysfunctions and Neuropathology in Mouse Models of Spinocerebellar Ataxia Type 2: A
Comprehensive Review. Front Neurosci 10: 572.

53. Fittschen M, Lastres-Becker I, Halbach MV, Damrath E, Gispert S, Azizov M, et al. (2015)
Genetic ablation of ataxin-2 increases several global translation factors in their transcript
abundance but decreases translation rate. Neurogenetics 16: 181-192.

54. Halbach MV, Gispert S, Stehning T, Damrath E, Walter M, Auburger G (2016) Atxn2
Knockout and CAG42-Knock-in Cerebellum Shows Similarly Dysregulated Expression in
Calcium Homeostasis Pathway. Cerebellum.

55. Bushart DD, Murphy GG, Shakkottai VG (2016) Precision medicine in spinocerebellar
ataxias: treatment based on common mechanisms of disease. Ann Transl Med 4: 25.

56. Bezprozvanny IB (2010) Calcium signaling and neurodegeneration. Acta Naturae 2: 72-82.

57. van de Leemput J, Chandran J, Knight MA, Holtzclaw LA, Scholz S, Cookson MR, et al.
(2007) Deletion at ITPR1 underlies ataxia in mice and spinocerebellar ataxia 15 in humans.
PLoS Genet 3: e108.

58. Iwaki A, Kawano Y, Miura S, Shibata H, Matsuse D, Li W, et al. (2008) Heterozygous

deletion of ITPR1, but not SUMF1, in spinocerebellar ataxia type 16. J Med Genet 45: 32-35.

59. Ragothaman M, Muthane U (2008) Homozygous SCA 2 mutations changes phenotype and
hastens progression. Mov Disord 23: 770-771.
60. Yamamoto A, Lucas JJ, Hen R (2000) Reversal of neuropathology and motor dysfunction in
a conditional model of Huntington's disease. Cell 101: 57-66.

61. Boy J, Schmidt T, Wolburg H, Mack A, Nuber S, Bottcher M, et al. (2009) Reversibility of
symptoms in a conditional mouse model of spinocerebellar ataxia type 3. Hum Mol Genet 18:
4282-4295.

62. Zu T, Duvick LA, Kaytor MD, Berlinger MS, Zoghbi HY, Clark HB, et al. (2004) Recovery
from polyglutamine-induced neurodegeneration in conditional SCA1 transgenic mice. J Neurosci
24: 8853-8861.

63. Oz G, Vollmers ML, Nelson CD, Shanley R, Eberly LE, Orr HT, et al. (2011) In vivo
monitoring of recovery from neurodegeneration in conditional transgenic SCA1 mice.
Experimental Neurology 232: 290-298.

64. Xia H, Mao Q, Eliason SL, Harper SQ, Martins IH, Orr HT, et al. (2004) RNAi suppresses
polyglutamine-induced neurodegeneration in a model of spinocerebellar ataxia. Nat Med 10:
816-820.

65. Hache M, Swoboda KJ, Sethna N, Farrow-Gillespie A, Khandji A, Xia S, et al. (2016)
Intrathecal Injections in Children With Spinal Muscular Atrophy: Nusinersen Clinical Trial
Experience. J Child Neurol.

66. Miller TM, Pestronk A, David W, Rothstein J, Simpson E, Appel SH, et al. (2013) An
antisense oligonucleotide against SOD1 delivered intrathecally for patients with SOD1 familial
amyotrophic lateral sclerosis: a phase 1, randomised, first-in-man study. Lancet Neurol 12: 435-
442.

67. Rigo F, Seth PP, Bennett CF (2014) Antisense oligonucleotide-based therapies for diseases
caused by pre-mRNA processing defects. Adv Exp Med Biol 825: 303-352.

68. Bennett CF, Swayze EE (2010) RNA targeting therapeutics: molecular mechanisms of
antisense oligonucleotides as a therapeutic platform. Annu Rev Pharmacol Toxicol 50: 259-293.

69. Skotte NH, Southwell AL, Ostergaard ME, Carroll JB, Warby SC, Doty CN, et al. (2014)
Allele-specific suppression of mutant huntingtin using antisense oligonucleotides: providing a
therapeutic option for all Huntington disease patients. PLoS One 9: e107434.
70. Ostergaard ME, Southwell AL, Kordasiewicz H, Watt AT, Skotte NH, Doty CN, et al. (2013)
Rational design of antisense oligonucleotides targeting single nucleotide polymorphisms for
potent and allele selective suppression of mutant Huntingtin in the CNS. Nucleic Acids Res 41:
9634-9650.

71. White JK, Auerbach W, Duyao MP, Vonsattel JP, Gusella JF, Joyner AL, et al. (1997)
Huntingtin is required for neurogenesis and is not impaired by the Huntington's disease CAG
expansion. Nat Genet 17: 404-410.

72. McBride JL, Pitzer MR, Boudreau RL, Dufour B, Hobbs T, Ojeda SR, et al. (2011) Preclinical
safety of RNAi-mediated HTT suppression in the rhesus macaque as a potential therapy for
Huntington's disease. Mol Ther 19: 2152-2162.

73. Boudreau RL, McBride JL, Martins I, Shen S, Xing Y, Carter BJ, et al. (2009) Nonallele-
specific silencing of mutant and wild-type huntingtin demonstrates therapeutic efficacy in
Huntington's disease mice. Mol Ther 17: 1053-1063.
Figure Captions

Figure 1. Anticipation in SCA2. SCA2 age of onset is negatively correlated with ATXN2 CAG
repeat length. Note that the variability in AO for any CAG repeat length is partly associated with
CACNA1A repeat length (but no other polyglutamine disease genes) and likely also to other
genetic and environmental factors [10].

Figure 2. ATXN2 domain structure and sites of interactions. The diagram depicts the amino
acids of the ATXN2 protein with locations of known domains indicated. The domains and their
locations are as follows: Polyglutamine tract (PolyQ) (aa 166-187), SRC homology 3 (SH3)
domain binding motifs 1 (SBM1) (aa 117-126) and 2 (SBM2) (aa 587-596), Like sm domain
(Lsm) (aa 254-345), Lsm associated domain (LsmAD) (aa 353-475), PABP interacting motif 2
(PAM2) (aa 908-925). In addition, ATXN2 has an acidic domain (aa 256-405), a predicted
clathrin-mediated sorting signal (aa 414-416), and a predicted site for caspase cleavage (aa
396-399) [20].

Figure 3. Pathogenic outcomes of ATXN2 mutation. Polyglutamine expansion in ATXN2 results

in misfolding and increased staufen1 (STAU1) expression and ATXN2 aggregations leading to
stress granule formation and abnormal RNA processing. Misfolded ATXN2 interacts directly
with IP3R leading to abnormal IP3R channel activity followed by calcium release from internal
stores. ATXN2 also directly interacts with RGS8 mRNA resulting in reduced RGS8 protein
abundance. RGS8 reduction is attributed to decreased RGS8 translation which may be caused
by sequestration in stress granules, and to decreased RGS8 mRNA abundance possibly also
related to stress granule functions. The consequence of reduced RGS8 is overactive mGluR1
leading to increased cytoplasmic calcium. The result of is abnormally slow Purkinje cell firing.

Figure 4. Reduction of ATXN2 expression improved SCA2 mouse phenotypes. (A) Compared
with saline-treated animals, spinocerebellar ataxia type 2 mice (Pcp2-ATXN2-Q127) show
significantly reduced progression of motor disability. Mean latency to fall in 3 trials on day 3 of
rotarod testing is shown. (B) Cerebellar endogenous mouse and human transgenic ataxin-2
gene (ATXN2) are reduced by ASO treatment 14 weeks after intracerebroventricular (ICV)
injection compared with saline. NS, not significant.
Table  1.    Molecular  features  of  the  ATXN2  gene*
Chromosomal  Posi=on 12q24.12
Number  Exons 25
Gene  length 147,463  bp
Transcript  length 4,699  bp
3’-­‐UTR  length 601  bp
Puta=ve  Start  Codons 2
               ATG1  use:
                               5’-­‐UTR  length 162  bp
                               Protein  length 1,312  aa
                               Protein  size 144  kDa
               ATG2  use:
                               5’-­‐UTR  length 642  bp
                               Protein  length 1,152  aa
                               Protein  size 127  kDa
* Transcript ID ENST00000377617 in Ensembl version
ENSG00000204842.14
Table  2.  ATXN2  Interac=ng  Macromolecules
ATXN2  interactor ATXN2  Binding   ATXN2   Interacting Protein Citation
Region‡,§ Domain|| Function
A2BP1 / RBFOX1 760-1312 - RNA binding Shibata et al., 2000 [25]

PABP 816-1312 PAM2 RNA binding Ralser et al., 2005 [27]

DDX6 254-475 Lsm & LsmAD RNA binding Nonhoff et al., 2007 [30]

TDP-43 FL - RNA binding / ALS Elden et al., 2010 [4]

FUS† FL - RNA binding / ALS Farg et al., 2013 [6]

Parkin 1-396 - Ubiquitination Huynh et al., 2007 [33]

Staufen1†   FL - RNA binding Paul et al., submitted.

IP3R† FL - Ca2+ signaling Liu et al., 2009 [36]

RGS8 mRNA FL - Ca2+ signaling Dansithong et al., 2015 [46]

Endophilin-A1 481-815* SBM2 Vesicle endocytosis Ralser et al., 2005 [40]

Endophilin-A3 481-815* SBM2 Vesicle endocytosis Ralser et al., 2005 [40]

CIN85† FL - Vesicle endocytosis Nonis et al., 2008 [41]

† Direct interaction with ATXN2 not demonstrated.

‡ Smallest ATXN2 amino acid regions experimentally tested for interaction.
§ FL, Full-length ATXN2; narrower interacting regions in ATXN2 not determined.
|| Domain within the ATXN2 binding region required for the interaction.

* Nonis et al., 2008 identified additional flanking binding sites for the endophilins.
 Age  at  onset  (AO)  and  CAG  repeat  length  are  inversely  related  
80  

70  
Age of disease onset

60  

50  

40  

30  

20  

10  

10  
32   34   36   38   40   42   44   46   48   50   52   54   56   58   60   62   64   66   68   70   72   74   76  
CAG repeat length

Figure  1  
 ATXN2  Domain  Structure  and  Sites  of  InteracQons  

SBM1 PolyQ Lsm LsmAD SBM2 PAM2

aa:  

DDX6   A2BP1/RBFOX1  

Parkin   PABP  

Endophilin  A1  &  A3  

Staufen1          TDP-­‐43          FUS          IP3R          CIN85  

(narrower  regions  not  determined)  

Figure  2  
 Pathogenic  Outcomes  of  ATXN2  MutaQon  
ATXN2  /  STAU1  
Misfolded   Aggregates  
NaQve  ATXN2   Mutant  ATXN2  
PolyQ
Expansion

Staufen
Rises

Abnormal  RNA  
Processing  in  SGs  
IP3R     RGS8    
acQvated   reduced  

Cytoplasmic  Ca2+  Rises  

Slow  Purkinje  
Cell  Firing  

Figure  3  
 ASO  treatment  of  SCA2  mice  improved  motor  behavior  

210 µg ASO7
ICV@ 8wks

A   NS P<0.05 P<0.01
B   P<0.01
200   150  
125  

ATXN2  /    Ac=n  (%)  

Latency  to  fall  (s)  

150  
100  
ASO7  
100   75  
SALINE  
50  
50   N=13-15
25  
0   0  
5 wks 9 wks 13 wks Treatment time
Saline   ASO7  
13 wks 17 wks 21 wks Age

Figure  4