Content uploaded by Yuko Okamura-Oho
Author content
All content in this area was uploaded by Yuko Okamura-Oho on Mar 18, 2014
Content may be subject to copyright.
© 1999 Oxford University Press Human Molecular Genetics, 1999, Vol. 8, No. 6 947–957
ARTICLE
Dentatorubral-pallidoluysian atrophy protein interacts
through a proline-rich region near polyglutamine with
the SH3 domain of an insulin receptor tyrosine kinase
substrate
Yuko Okamura-Oho, Toshiyuki Miyashita, Kazuhiro Ohmi
1
and Masao Yamada
+
Department of Genetics and
1
Department of Pathology, National Children’s Medical Research Center, 3-35-31
Taishido, Setagaya, Tokyo 154-8509, Japan
Received December 22, 1998; Revised and Accepted February 28, 1999 DDBJ/EMBL/GenBank accession nos AB017119 and AB017120
Dentatorubral-pallidoluysian atrophy (DRPLA) is an autosomal dominant neurodegenerative disorder associ-
ated with CAG/glutamine repeat expansion. While the
DRPLA
gene is ubiquitously expressed, neuron death
occurs in specific anatomical areas of the brain. This predicts that the DRPLA protein interacts with other pro-
teins and that these interactions may play a role in pathogenesis. Here, we describe a protein that binds to the
DRPLA
product. One of the clones isolated with a yeast two-hybrid system was identified as a human
homolog of the insulin receptor tyrosine kinase substrate protein of 53 kDa (IRSp53). The gene produced two
mRNA forms by differential splicing and encoded 552 and 521 amino acids, respectively. The longer form was
mainly expressed in the brain and the shorter one in other tissues. The products were phosphorylated upon
stimulation of cultured cells with insulin or insulin-like growth factor 1. Binding of the DRPLA protein to
IRSp53 was ascertained by co-immunoprecipitation with antibodies and also by co-localization in perinuclear
oval dots in cells expressing engineered constructs. A proline-rich region near the polyglutamine tract of the
DRPLA protein and the SH3 domain of IRSp53 were involved in the binding. An extended polyglutamine tract
significantly reduced binding ability in yeast cells, but not in
in vitro
binding assays. The identification of
IRSp53 and other proteins detected by the yeast hybrid system predicts that DRPLA functions in a signal
transduction pathway coupled with insulin/IGF-1.
INTRODUCTION
Dentatorubral-pallidoluysian atrophy (DRPLA) is an auto-
somal dominant, progressive neurodegenerative disorder char-
acterized by selective neuron death in the dentatofugal and
pallidofugal pathways (1). It has been shown that DRPLA is
caused by CAG repeat expansion (2,3). The number of CAG
repeats in the DRPLA gene is highly polymorphic in the nor-
mal population with a range of 7–23, increasing to 49–88 in
patients (2,4). The gene is localized to chromosome band
12p13.31 and encodes a protein of an apparent molecular size
of 160 kDa (the DRPLA protein) in which the CAG repeat is
translated into a polyglutamine tract (5–8). The gene is ubiqui-
tously expressed and the levels of mRNA and protein expres-
sion are not affected by the repeat expansion (5,6,9).
In addition to DRPLA, a CAG/glutamine repeat expansion
has been detected in seven other inherited neurodegenerative
disorders including spinal and bulbar muscular atrophy
(SBMA), Huntington’s disease (HD) and spinocerebellar
ataxia (SCA) types 1–3, 6 and 7 (10–16). These disorders
exhibit characteristic phenotypes resulting from dysfunction
and neuron death in respective regions in the central nervous
system, but each responsible gene is fundamentally expressed
in ubiquitous tissues. A recent study has demonstrated that
overexpression of a relatively short peptide containing an
extended polyglutamine tract induces apoptosis in cultured
cells, while overexpression of an intact gene product carrying
the same extended polyglutamine tract does not (17). As pre-
dicted based on the anti-parallel β-sheet structure of synthetic
polyglutamine tracts (18), short peptides with an extended
polyglutamine tract form large, poorly soluble aggregates after
expression in cultured cells. Nuclear inclusion bodies found in
affected brains of patients and mice transgenic for HD, SCA1,
SCA3 and DRPLA seem to be consistent with aggregate for-
mation in such experimental systems (19–21). Most of the
products for CAG repeat diseases have been shown to be
cleaved by caspases during various apoptotic processes
(8,22,23) and cleaved products tend to form aggregates if they
+
To whom correspondence should be addressed. Tel: +81 3 3414 8121; Fax: +81 3 3412 2259; Email: myamada@nch.go.jp
948 Human Molecular Genetics, 1999, Vol. 8, No. 6
carry an extended polyglutamine tract. These results may
account for neuron death by an extended polyglutamine tract;
however, little is known about the regional specificity of neu-
ron death. Studies on associated proteins are useful to elucidate
the pathogenesis of selective neuron death as well as the nor-
mal functions of the responsible genes since none has been
identified except for the androgen receptor for SBMA (10) and
a calcium channel subunit for SCA6 (15).
Here we report that the DRPLA protein interacts with a sub-
strate of insulin receptor tyrosine kinase. We obtained several
clones after screening libraries with the DRPLA protein as bait
in a yeast two-hybrid system and one of the clones was identi-
fied to be a human homolog of a 53 kDa substrate of the insulin
receptor tyrosine kinase (IRSp53). Its hamster homolog has
been detected by screening for substrates of the insulin and
insulin-like growth factor 1 (IGF-1) receptors with a mono-
clonal antibody (24). We characterized human IRSp53 and
studied its interaction with DRPLA protein with respect to
binding regions and effects of an extended polyglutamine tract.
Since IGF-1 is known to be a trophic or anti-apoptotic factor in
the brain (25–26), our finding is also relevant to the elucidation
ofIGF-1signalingpathwaysinthebrain.
RESULTS
Isolation of DRPLA-interacting proteins using the yeast
two-hybrid system
Almost the entire coding region, but without its N-terminal
portion, of the normal DRPLA protein was used in screening
with the yeast two-hybrid system. A cDNA sequence carrying
10 CAG repeats was fused in-frame with the GAL4 DNA-
binding domain to generate pBDDR335–1185, which carried
amino acid residues 335–1185 of the DRPLA protein includ-
ing a polyglutamine tract (Fig. 1A). We screened ahuman fetal
brain Matchmaker cDNA library, which consisted of clones
carrying cDNA fused in-frame with the GAL4 activation
domain (AD). We finally isolated 30 His
+
/LacZ
+
clones out of
1.2 ×10
7
Trp
+
/Leu
+
auxotrophic transformants. The insert size
of these isolated clones ranged from 0.5 to 3.5 kb. When the 30
clones were sequenced, seven were found to overlap and
turned out to be a human homolog of a 53 kDa substrate of
insulin receptor tyrosine kinase (IRSp53) by sequence homol-
ogy to the previously reported IRSp53 of Chinese hamster
(24). Others included proteins functioning in the signal trans-
Figure 1. Schematic illustrations of fused proteins and in vitro translation products used in this study. Amino acid residues of each protein are indicated. The
DRPLA regions, the polyglutamine tract, IRSp53 and the SH3 domain are indicated by gray, filled, open and hatched boxes, respectively. (A) ‘Bait’ protein used
in the yeast two-hybrid library screening and in the binding assay in yeast. A portion of DRPLA cDNA was inserted downstream of the BD and the translational
reading frame was adjusted. (B) GFP-tagged protein constructs used in expression experiments in HeLa cells. The indicated portions of DRPLA or IRSp53 cDNA
of the longer form were inserted in-frame downstream of GFP. The position of an NLS in the DRPLA protein is indicated by a broad line. (C) A series of deletion
constructs producing a truncated form of the DRPLA protein. The C-terminal end of the DRPLA amino acid sequence is indicated by the number after DR. A series
of constructs having extended polyglutamine tracts are indicated by Q+. (D) Three constructs to generate GST–IRSp53 fusion proteins.
Human Molecular Genetics, 1999, Vol. 8, No. 6 949
duction pathway from insulin and Wnt, namely hDVL1, δ-
catenin and the 14-3-3 protein homolog (27–30), and also
nuclear proteins containing a zinc finger motif. Here we further
characterize human IRSp53.
Characterization of the human IRSp53 gene
The longest insert of the isolated IRSp53 clones (clone 274)
consisted of 2877 nt, with the exception of the AD portion and
a poly(A) tail. By single path sequencing, two other isolated
clones seemed to have the identical sequence and one carried
nucleotides 1–2831. In contrast, the remaining three clones
carried 2033 nt, besides the AD and poly(A) tail, in which the
nucleotide sequence of nt 1–1628 was completely identical to
that of clone 274. The open reading frame in the longer form
started with ATG at nt 94, ended at nt 1749 and encoded 552
amino acid residues, while that in the shorter form started at the
same position, ended at nt 1656 and encoded 521 amino acids
(Fig. 2A). When the amino acid sequences were aligned with
the reported hamster IRSp53 sequence, the shorter form resem-
bled it in size and sequence; 94% of residues were identical
after two spaces were inserted. Three putative tyrosine phos-
phorylation sites appeared in the N-terminal portion and an
SH3 domain similar to those of human proteins VAV2 and
yeast BOB1 (24) localized to positions 407–453 of the amino
acid sequence (Fig. 2B).
Northern blotting using a probe common to the longer and
shorter forms of human IRSp53 cDNA visualized a 3.0 kb
transcript expressed in most tissues examined as well as
another 3.8 kb transcript mainly expressed in the brain (Fig. 2A
and C). Another probe, which was uniqueto the longer form of
cDNA, predominantly hybridized with the 3.8 kb transcript
(Fig. 2A and D). The size of the two transcripts differed by
0.8 kb, which accords with the size difference of the 3'-portion
of the two cDNA forms. Thus, the two forms seem to be pro-
duced by alternative splicing. This was confirmed by RT–PCR
with a primer set of a forward primer situated in the common
region and a reverse primer situated in the respective unique 3'-
regions (data not shown).
Figure 2. Structure and expression of human IRSp53. (A) Schematic illustration of two forms of human IRSp53. The two are identical up to position 1628 and diffe
r
in the 3'-portion. The probes used in northern blotting (probes1 and 2, 1.0 kb NcoI–NcoI and 0.7 kb SseI–SseI fragments, respectively) are illustrated. (B) Amino acid
sequences of human IRSp53 aligned with hamster IRSp53. IRSp53(S) and IRSp53(L) are the shorter and longer forms, respectively, and only the last 73 amino acids
are shown for IRSp53(L). Identical and similar amino acid residues are indicated by asterisks and dots, respectively, between the sequence lines. The alternative
splicing junction is indicated by arrows, putative tyrosine phosphorylation sites are boxed and the SH3 domain is underlined. (C–E) Northern blots showing the
expression pattern for human IRSp53. Probe 1, derived from the common region of both the shorter andlonger forms, visualized 3.0 and 3.8 kb transcripts(C). Probe
2, unique to the longer form, hybridized with only the 3.8 kb transcript, which is mainly expressed in the brain (D). β-Actin was used as a control for the amount o
f
applied RNA (E). The nucleotide sequences of IRSp53(S) and IRSp53(L) will appear in the DDBJ, GSDB, EMBL and NCBI databases with accession nos AB017120
and AB017119.
950 Human Molecular Genetics, 1999, Vol. 8, No. 6
Characterization of the human IRSp53 protein
When in vitro translation products from each of the shorter and
longer forms of human IRSp53 cDNA were analyzed by SDS–
PAGE, multiple protein bands were detected (data not shown).
To further examine the products, we raised a rabbit polyclonal
antibody against a C-terminal portion of IRSp53 by the use of
a glutathione S-transferase (GST) fusion construct (GST–
IRSp53c; Fig. 1D). The antibody also detected multiple prod-
ucts in western blotting of the in vitro translation products (Fig.
3A). The largest productsgenerated with the shorter and longer
forms migrated at 58 and 63 kDa, respectively, which corre-
sponds with the expected size calculated from the first methio-
nine residue (Fig. 3A, lanes S and L). The second largest
products in the respective lanes were 53 and 58 kDa, appar-
ently consistent with the size calculated from the fifth methio-
nine residue at position 59. The antibody also recognized
endogenous products in PC12 cells (Fig. 3A, lane U). Since the
53 kDa species was the major product of in vitro translation
with the shorter form of human IRSp53, Met59 seemed prefer-
able as an initiation site for translation. Alternatively, multiple
species may be generated in a post-translational process.
Although IRSp53 is phosphorylated, the multiple species
detected here cannot solely be explained by phosphorylation,
as described below.
Phosphorylation of IRSp53
The total amount of IRSp53 in PC12 cells was not much
changed upon stimulation with insulin (Fig. 3A, lanes 0, 5 and
120 min). Phosphorylated IRSp53 was detected by immuno-
precipitation with the anti-IRSp53 antibody followed by west-
ern blotting with an anti-phosphotyrosine antibody (Fig. 3B).
Endogenous IRSp53 in PC12 cells was phosphorylated with as
low as 10 nM insulin or IGF-1 (Fig. 3C). The phosphorylated
form increased to a detectable level 5 min after insulin stimu-
lation and continued to accumulate up to 120 min in culture
with sodium vanadate, a phosphatase inhibitor. These results
are consistent with previous observations in Chinese hamster
ovary cells with overexpressed insulin receptor (24). The phos-
phorylated form migrated at the same distance as the largest
form of endogenous IRSp53 in PC12 cells. Since the amount
of the 58 kDa species detectable on western blotting with the
anti-IRSp53 antibody was not changed by insulin stimulation,
only a fraction of that migrating at the 58 kDa position was
phosphorylated. Thus, phosphorylation alone does not explain
the multiple forms of IRSp53. A GFP–IRSp53 fusion
expressed with cloned cDNA was also phosphorylated under
insulin stimulation (data not shown). Through these studies,
we confirmed the isolated clone in this study to be phosphory-
lated upon insulin stimulation.
Interaction of IRSp53 with the DRPLA protein in yeast
We verified the interaction of IRSp53 with the DRPLA protein
in yeast. As in the screening process, when yeast was co-trans-
formed with a plasmid carrying full-length IRSp53 to produce
a fusion protein with the AD (ADIRSp53) and pBDDR335–
1185, both the HIS3 and LacZ reporter genes were activated
(Fig. 4). The reporter genes were not activated when trans-
formed with pADIRSp53 alone or together with either a plas-
midcarryingonlytheGAL4DNA-bindingdomain(BD)
portion or a control plasmid like BD–murine p53 (data not
shown). Similarly, transformation with pBDDR335–1185
alone or together with a plasmid carrying only the AD portion
caused no activation (data not shown). Thus, activation of the
reporter genes depended on interaction of the two proteins
IRSp53 and DRPLA and did not result from self-activation nor
from a single protein product. We then examined which por-
tion of the DRPLA protein was involved in binding by the use
of a series of deletion plasmids. pBDDR335–786, which
expressed a fusion protein of the BD with DRPLA amino acids
335–786, activated the reporter genes after co-transformation
of yeast with ADIRSp53 and produced a higher level of β-
galactosidase activity (Fig. 4). In contrast, BDDR676–973 and
BDDR964–1185 showed almost no β-galactosidase activity in
co-transformed yeast. Thus, the interacting portion of the
DRPLA protein seemed to be localized in the amino acid 335–
675 region. A construct with an extended polyglutamine tract
(BDDR335–786Q+) still maintained a positive interaction
with IRSp53, but β-galactosidase activities were reduced to
one third of those for a similar construct with a normal range of
the CAG repeat (Fig. 4C).
Co-immunoprecipitation of the DRPLA protein and
IRSp53
Interaction between IRSp53 and the DRPLA protein was con-
firmed by immunoprecipitation. GFP-tagged IRSp53 produced
by transfection in HeLa cells was detectable at the expected
position by western blotting with an anti-GFP antibody. The
GFP-tagged IRSp53 protein, but not GFP itself, was recovered
in precipitates with the anti-DRPLA antibody (Fig. 5A). Con-
versely, endogenous DRPLA protein in the cells was immuno-
Figure 3. Characterization of IRSp53 protein. Western blots showing phos-
phorylation of endogenous IRSp53 in PC12 cells upon insulin or IGF-1 stimu-
lation. (A) Proteins isolated from untreated PC12 cells (lane U) were
visualized by western blotting with the anti-IRSp53 antibody with parallel runs
of in vitro translation products of the shorter and longer forms of human
IRSp53 cDNA (lanes S and L) as a control for size. PC12 cells were serum
starvedfor3hfollowedbyincubationwith10µM insulin and 500 µMsodium
vanadate for 0, 5 and 120 min. Proteins isolated from the treated cells were vis-
ualized with the anti-IRSp53 antibody (lanes 0, 5 and 120 min). (B)Thesame
protein fractions were subjected to immunoprecipitation with the anti-IRSp53
antibody and detected by western blotting with the anti-phosphotyrosine anti-
body. (C) PC12 cells were treated with the indicated concentrations of insulin
or IGF-1 together with 500 µM sodium vanadate for 120 min, then the proteins
were analyzed as in (B).
Human Molecular Genetics, 1999, Vol. 8, No. 6 951
precipitated with the anti-GFP antibody (data not shown).
Moreover, the endogenous IRSp53 and DRPLA proteins in
PC12 cells were co-immunoprecipitated with the anti-IRSp53
antibody (Fig. 5B). These results clearly indicate that IRSp53
binds to the DRPLA protein.When PC12 cells were stimulated
with insulin, the amount of precipitated DRPLA protein was
not significantly changed. Upon insulin stimulation, phosphor-
ylated IRSp53 levels increased markedly but the total amount
of IRSp53 was not changed, as described above. Thus, phos-
phorylation of IRSp53 seemed not to enhance binding to the
DRLPA protein.
Co-localization of the DRPLA protein and IRSp53
Subcellular localization of the DRPLA protein and IRSp53 in
HeLa cells was studied under a confocal microscope. Endog-
enous IRSp53 detected with the anti-IPSp53 antibody was dif-
fusely distributed in the cytoplasm (Fig. 6, panel 2).
Endogenous DRPLA protein detected with the anti-DRPLA
antibody was mainly localized in nuclei (Fig. 6, panel 1),
although another anti-DRPLA antibody revealed cytoplasmic
localization in brain tissues (6; see Discussion). Most of the
GFP-tagged full-length DRPLA protein (GFP–DRQ14) was
also localized in the nuclei when it carried a normal range of
polyglutamines (Fig. 6, panel 3). To enrich the cytoplasmic
fraction of the DRPLA protein, we made GFP-tagged con-
structs without the nuclear localization signal (NLS) at the N-
terminal end of the DRPLA protein (GFP–DRQ14∆Nand
GFP–DRQ71∆N; Fig. 1B). Such an engineered DRPLA pro-
tein with a normal range of polyglutamines was localized in the
cytoplasm as small oval dots, although a fraction of the protein
still entered the nuclei (Fig. 6, panel 9). Although most endog-
enous IRSp53 maintained its diffuse cytoplasmic distribution
in transfected cells (Fig. 6, panel 10), co-localization of GFP–
DRQ14∆N with IRSp53 was noted as yellow dots in the peri-
nuclear region (Fig. 6, panel 11). The GFP-tagged DRPLA
protein with an extended polyglutamine tract (GFP–DRQ71)
formed nuclear aggregates as previously described (Fig. 6,
panel 6) (31). In this condition, endogenous IRSp53 seemed
not to participate in aggregate formation (Fig. 6, panels 6–8).
However, the GFP-tagged DRPLA protein without NLS and
carrying an extended polyglutamine tract (GFP–DRQ71∆N)
composed perinuclear and intranuclear oval dots with a diffuse
distribution in the nuclei and IRSp53 was involved in aggre-
gates in the perinuclear dots.
An SH3 domain interacts with a proline-rich region near
the polyglutamine tract
To identify the region ofthe DRPLA protein involved in bind-
ingtoIRSp53,in vitro translation products from the DRPLA
constructs were subjected to an in vitro binding assay using a
GST column. GST fusion proteins applied to the column were
produced in Escherichia coli with constructs carrying full-
length IRSp53 (GST–IRSp53f), its N-terminal portion (GST–
IRSp53n) and its C-terminal portion (GST–IRSp53c), respec-
tively, in addition to GST alone and an unrelated fusion protein
(GST–Bcl2) as a control (Figs 1D and 7A). A full-sized
DRPLA protein made by in vitro translation bound to GST–
IRSp53f and GST–IRSp53c, but not to GST–IRSp53n and
control columns (Fig. 7B–D). Although not studied exten-
sively, an SH3 domain in the C-terminal portion of IRSp53
Figure 4. Interaction of IRSp53 with DRPLA protein fragments in the yeast assay. The series of plasmids indicated by BDDR are the constructs used to produce
a portion of the DRPLA protein with the normal or expanded polyglutamine tract fused with the BD as illustrated in Figure 1A. Yeast cells were co-transformed
with one of the BDDR constructs and AD–IRSp53. Interaction between the two products was analyzed by growth on a plate containing selection medium without
Trp, Leu and His (A), by color development in a β-galactosidase filter assay (B) and by quantitative measurements of β-galactosidase activity in disrupted cells
(C). The activity was measured in six independent colonies of each transformant and is illustrated as the average ± SD. ND, below background level. The β-
galactosidase activity measured with yeast carrying murine p53 and SV40 large T-antigen on the BD and AD constructs is included as a positive control.
952 Human Molecular Genetics, 1999, Vol. 8, No. 6
was suggested to be involved in the binding. We then made a
series of DRPLA products by in vitro translation of DRPLA
constructs which had various sized deletions from the 3'-end
(Fig. 1C) and applied them to the GST–IRSp53c column.
Truncated DRPLA proteins produced from DR567, DR596
and DR749 constructs bound to the GST column but DR516
did not (Fig. 8A). These results suggest that the amino acid
517–567 region in the DRPLA protein is involved in binding.
Another truncated construct with an internal deletion,
DR749del, which had lost amino acid residues 428–601, did
not bind to GST–IRSp53c. These results and the binding assay
in yeast cells support the proposed binding region. The RPYP-
PGP sequence is a consensus proline-rich motif for binding to
the SH3 domain (32,33). There is one completely matched
sequence in the proposed binding region, which appears 38
amino acids downstream of the polyglutamine tract, while the
proposed binding region contained a few other similar
sequences (Fig. 8C).
An extended polyglutamine tract modulates binding
Contrary to the binding assay in yeast, an extended poly-
glutamine tract in DRPLA protein seemed not to significantly
affect binding in the GST column assay. Regardless of the size
of the polyglutamine tract, DRPLA proteins of full length
bound to GST–IRSp53f (Fig. 7). More careful quantitative
assays with truncated proteins showed that products having an
extended polyglutamine tract (DR749Q+ and DR696Q+)
bound to GST–IRSp53c to a similar extent as their counter-
parts having a normal sized polyglutamine tract (Fig. 8B).
However, DR598Q+, the smallest construct still retaining the
proposed SH3 binding sequence and an extended poly-
glutamine tract, displayed a significantly reduced affinity for
IRSp53c compared with that of DR596 (Fig. 8). Thus, an
extended polyglutamine tract in a certain form seemed to
obstruct binding to the proposed target region. However, the
proline-rich region flanking the consensus binding sequence
may serve as an auxiliary binding site, on which an extended
polyglutamine tract has little influence.
DISCUSSION
A growing number of neurodegenerative disorders have been
shown to be caused by CAG/glutamine repeat expansion.
Despite ubiquitous expression of each CAG repeat disease
gene, neuron death occurs in distinctive anatomical areas of the
brain, indicating that additional regional factors are involved in
Figure 5. Interaction between the DRPLA protein and IRSp53 verified by
immunoprecipitation. (A) Interaction between GFP–IRSp53 and the DRPLA
protein. We made a construct to produce a fusion protein of GFP with the longer
form of IRSp53. Protein fractions were isolated from HeLa cells transfected with
the GFP–IRSp53 construct or with a construct carrying GFP alone. A portion o
f
the fractions was subjected to immunoprecipitation with the anti-DRPLA anti-
body and analyzed by western blotting using the anti-GFP antibody with parallel
runs of fractions without immunoprecipitation. The single and double filled
arrows represent GFP–IRSp53 and GFP, respectively. (B) Interaction between
endogenous DRPLA and IRSp53 proteins. Protein fractions were isolated from
PC12 cells cultured with or without 10 µM insulin and 500 µM sodium vanadate
for 120 min. An aliquot of the fraction was subjected to immunoprecipitation
with the anti-IRSp53 antibody or preimmune sera and analyzed by western blot-
ting with the anti-DRPLA antibody. The filled arrow represents the size of the
endogenous DRPLA protein. Detected bands indicated by open arrows in both
(A) and (B) are the IgG heavy chain. Figure 6. Subcellular co-localization of IRSp53 with the DRPLA protein.
Endogenous DRPLA protein was mainly localized in the nucleus while
IRSp53 was localized in the cytoplasm, as shown with specific antibodies
stained with a TRITC-conjugated secondary antibody (red) (panels 1 and 2).
Co-localization of IRSp53 with DRPLA was ascertained using GFP-tagged
DRPLA protein. We transfected HeLa cells with each of the four DRPLA con-
structs indicated (Fig. 1) and visualized the DRPLA with GFP (green) and
IRSp53 with the antibody (red) at 24 h after transfection. Panels in the right
column (panels 5, 8, 11 and 14) were obtained after merging the green and red
pictures of the same cell, where yellow shows co-localization of the two mole-
cules. A fraction of IRSp53 was co-localized with DRPLA protein as far as it
localized in the cytoplasm after removing the nuclear localization signal
(GFP–DRQ14∆N and GFP–DRQ71∆N).
Human Molecular Genetics, 1999, Vol. 8, No. 6 953
the site-specific pattern of neurodegeneration. Several candi-
dates having a binding capacity for Huntingtin, the gene prod-
uct of the HD gene, have been detected to date and they
provide a clue to understanding the normal and pathological
functions of the HD gene (34). Some of the Huntingtin-binding
proteins show a cytoskeletal localization, indicating that Hunt-
ingtin may play a role in vesicle trafficking within cells (35). A
Grb2-like protein and members of the signaling complex for
the epidermal growth factor receptor have been shown to bind
through their SH3 domains with Huntingtin. These studies
imply that Huntingtin is involved in signal transduction
(36,37). A few proteins interacting with DRPLA protein have
been identified, including WW domain-containing proteins
which seem to function in the cytoskeleton (38).
In this report, we have identified an insulin receptor substrate
of 53 kDa (IRSp53) as one of the DRPLA-interacting proteins.
Hamster IRSp53 was previously detected during screening for
substrates of the tyrosine kinase of the insulin or IGF-1 recep-
tor by use of a monoclonal antibody and its gene was cloned
(24). Our identification of human IRSp53 is grounded not only
in sequence homology to the previously characterized hamster
IRSp53 but also in our experiments on phosphorylation. Sev-
eral insulin receptor substrates (IRSs) have been identified,
including IRS-1, IRS-2, IRS-3 and IRS-4, of which IRS-1 has
been most extensively characterized in the signal transduction
pathway from the insulin and IGF-1 receptors (39). Several
IRSs form pre-assembled complexes with other members
involved in the signal transduction pathway and the complexes
may be associated with the actin cytoskeleton (40). In contrast,
although hamster IRSp53 is clearly demonstrated to be phos-
phorylated upon insulin stimulation in experimental systems,
IRSp53 has not been fully characterized molecularly and in
relation to the signal pathway (41). Data presented in this
report on the human homolog characterize more fully the
IRSp53 gene and product.
The human IRSp53 gene generated two transcripts by differ-
ential splicing and produced multiple forms of protein distin-
guishable by SDS–PAGE not only in cells but also by in vitro
translation from each cDNA construct. Although a fraction of
IRSp53 is phosphorylated, the multiple forms are not
accounted for by phosphorylation alone. We assume that some
of the species are generated by utilization of some other
methionine as an initiation site, although any methionine in the
first 70 amino acid residues is not necessarily situated in a suit-
able context for initiation of translation (42). The previous
study on hamster IRSp53 also reported two forms, 53 and
58 kDa, detectable with the specific antibody and concluded
that the 53 kDa form was present in the brain (24). Since they
isolated only one form of cDNA, corresponding to the shorter
form of human cDNA, it is unknown whether differential
splicing takes places in rodents. Our study shows that the
longer transcript is mostly expressed in the brain and poten-
tially encodes a larger protein. Nevertheless, our preliminary
study detected a smaller form comparable with the 53 kDa
product in rat brain. Thus, more studies will be required to
define protein species, especially the brain form, of IRSp53.
UsinganantibodyraisedagainsttheC-terminalportionof
DRPLA, we previously detected DRPLA protein mostly in the
cytoplasm in brain tissues (6). In contrast, another antiserum
raised against the GST–DRPLA fusion protein used in this
study as well as in a previous study revealed a primarily
nuclear localization (31). This seems to be accounted for by
preferential recognition by the former antibody of phosphor-
ylated species (unpublished data). Other laboratories have also
reported both cytoplasmic and nuclear localizations of DRPLA
protein (20). We propose that the DRPLA protein is a shuttle
plying across the nuclear membrane. Since IRSp53 is localized
Figure 7. GST column assay showing interaction of the DRPLA protein and IRSp53. (A) GST–IRSp53 fusion proteins were expressed in E.coli, analyzed by SDS–
PAGE and stained with Coomassie Brilliant Blue to verify the molecular size and amount. The apparent size of each product accorded well with the calculated size
indicated on the left. (B) SDS–PAGE verifying the full-length DRPLA products with normal and extended polyglutamine tracts. DRPLA constructs DR1185 and
DR1185Q+, carrying normal range and expanded CAG repeats, respectively, were in vitro translated with a radioisotope. Products were analyzed by SDS–PAGE
to measure the radioactivity. (Cand D) GST column assay showing the interaction between the DRPLA protein and IRSp53. Radiolabeled products with a defined
radioactivity were applied to the GST columns and retained protein was analyzed by SDS–PAGE. Each column was pretreated with a GST fusion protein as indi-
cated at the bottom. Constructs for the GST fusion protein and in vitro translation products are seen in Figure 1C and D.
954 Human Molecular Genetics, 1999, Vol. 8, No. 6
in the cytoplasm, we engineered a DRPLA protein without an
NLS in order to detect in vivo interaction of the two molecules.
The DRPLA protein without an NLS was distributed more
abundantly in the cytoplasm and co-localization with IRSp53
was ascertained.
An extended polyglutamine tract in the DRPLA protein con-
siderably reduced binding to IRSp53 in the yeast assay system,
but did not significantly affect subcellular co-localization or in
vitro binding. The consensus structure for binding to the SH3
domain has been identified to be a PP II helix composed of
PXXPXΦor ΦXXPXXP, where Φis a basic amino acid resi-
due, preferably arginine (32,33). The RPYPPGP sequence at
amino acids 535–541 of the DRPLA protein exactly matches
the consensus sequence and this is within the essential region
for binding in the in vitro binding assay. The binding ability of
the smallest construct having the consensus sequence was
much influenced by an extended polyglutamine tract, but that
for larger constructs having the additional proline-rich
sequence was not significantly affected. Thus, the effects of an
extended polyglutamine tract seem to depend on the position,
the conformational structure or the flanking sequence of the
target. Since proteins are correctly folded in the in vivo situa-
tion, overall affinity may be maintained, as observed under the
microscope, but partially reduced in yeast.
IGF-1 is known to exhibit metabolic and trophic actions in
the brain. IGF-1 is transiently expressed in rat brain during
maturation and expression is evident in a specific group of
functionally related cerebellar projection neurons, including
the deep cerebellar and red nuclei systems (43), which is com-
parable with the affected areas in DRPLA. Knockout mice for
the IGF-1 gene show defects in the brain with a reduced size,
hypomyelination and loss of neurons in particular areas (44).
Treatment with exogenous IGF-1 is reported to enhance the
survival of neurons in vitro and in vivo by activating several
substrates of the receptor (24,25,45). Since our yeast two-
hybrid screening reveals several other proteins potentially
involved in insulin/IGF-1 signal transduction, the DRPLA pro-
tein may play a role in the neuronal signaling pathway from
insulin/IGF-1, specifically presenting a docking site for forma-
tion of a multiprotein complex. An extended polyglutamine
tract may influence the whole conformation of the complex,
resulting in an impairment of the IGF-1 signaling pathway.
MATERIALS AND METHODS
DNA techniques
Plasmid isolation, DNA manipulation and northern blotting
were carried out following standard methods described previ-
Figure 8. GST column assay showing the effect of an extended polyglutamine tract on the DRPLA protein. (Aand B) Various DRPLA constructs (Fig. 1C) were
in vitro translated with radioisotope to produce truncated products. Products were analyzed by SDS–PAGE to measure the radioactivity (upper panels). Products
with a defined radioactivity were applied to columns pretreated with GST–IRSp53c and retained protein was analyzed by SDS–PAGE (lower panels). (C)The
DRPLA amino acid sequence near the target region. The end-point of each construct is indicated. Since DR567 retained the binding activity while DR516 lost it,
the 517–567 amino acid sequence (underlined) seemed to be a primary target region, where a consensus sequence for binding to the SH3 domain appeared (double
underlined).
Human Molecular Genetics, 1999, Vol. 8, No. 6 955
ously (5,46). DNA sequences were determined using an auto-
mated sequencer with primers situated in the vector portion
and also with M13 universal primers after subcloning. In RT–
PCR to confirm alternative splicing, human poly(A)
+
RNA
from fetal brain, adult brain, liver, kidney and lung (Clontech,
Palo Alto, CA) was reverse transcribed with Superscript II
(Gibco BRL, Rockville, MD) and amplified with a sense
primer, 5'-AAGAGCAGCAGCACGGG, and either of the
antisense primers, 5'-ACCAACCCAAGAACAAACCA for
the shorter form, or 5'-TTCTGGATGGGAGGTTGG for the
longer form.
Plasmid construction
All constructs used in this study are illustrated in Figure 1. The
original cDNA clones of the DRPLA gene used in this experi-
ment have been described previously (2,5). All the constructs
representing a normal range of repeats carry 10 CAG repeats
but encode 14 glutamine residues because of the (CAGCAA)
2
sequence ahead of the CAG repeat. To construct a series of
plasmids used as bait in the yeast two-hybrid system
(pBDDR335–1185, pBDDR335–786, pBDDR676–973 and
pBDDR964–1185), an NcoI–NcoI fragment (nt 1239–3814 in
the DRPLA cDNA sequence; GenBank accession no.
D31840), an NcoI–BalI (blunt-ended) fragment (1239–2598),
aPstI–PstI fragment (2259–3162) and a blunt-ended BamHI–
NcoI fragment (3125–3814) isolated from DRPLA cDNA
clones were fused in-frame at the NcoI, NcoI/SmaI, PstIand
SmaI sites, respectively, in the GAL4 BD of a yeast two-hybrid
vector, pAS2-1 (Clontech). A series of plasmids having a trun-
cated C-terminus (pDR749, pDR696, pDR596, pDR567 and
pDR516) were reported previously (8), in which the cDNA
portion including the first methionine residue was located
downstream of the T3 promoter in a pBluescript SK– vector
(Stratagene, La Jolla, CA). To construct a plasmid without the
polyglutamine tract and flanking regions, pDR749 plasmid
DNA was digested with AccIandBstEII, blunt-ended and then
self-ligated (pDR749del). To construct plasmids carrying
extended CAG repeats such as pBDDR335–786Q+,
pDR749Q+ and pDR696Q+, the AccI–AgeI fragment was
replaced with the corresponding fragment of pMY1247, which
is a cDNA construct carrying 71 CAG repeats. pDR598Q+ was
constructed by digestion of pDR749Q+ with AgeIandNruI,
followed by filling-in with the Klenow fragment and self-
ligation. Plasmids for GFP–DRPLA fusion protein (pGFP–
Q14), extended DRPLA protein (pGFP–Q71) and DRPLA
proteins without an NLS (pGFP–Q14∆N and pGFP–Q71∆N)
have been described previously (31). To generate a plasmid
producing a fusion protein of IRSp53 with GFP (pGFP–
IRSp53), a SalI–ApaI fragment (1–1846) of the longer form of
IRSp53 cDNA was ligated in-frame with the XhoI–ApaI
fragment of the expression vector pEGFP-C3 (Clontech). To
generate plasmids producing fusion proteins of IRSp53 with
GST (pGST–IRSp53f, pGST–IRSp53n and pGST–IRSp53c),
aBspEI–EcoRV fragment (177–2887), an NcoIfragment
(192–1194) and an NcoI–EcoRV fragment (1190–2887) of the
longer form of IRSp53 cDNA were blunt-ended and then
inserted in-frame into an SmaI site of the pGEX-5X vector
(Pharmacia, Uppsala, Sweden). For control plasmids used in
the yeast two-hybrid system and expression experiments, BD/
murine p53 (pVA3-1) and AD/SV40 large T antigen were
purchased from Clontech and GST–Bcl2 was a generous gift
from Dr John C. Reed (Burnham Institute, La Jolla, CA).
Yeast two-hybrid system
Yeast of the Y190 strain (MATa,ura3-52,his3-200,lys2-801,
ade2-101,trp1-901,leu2-3,112,gal4∆,gal80∆,cyh
r
2,
LYS2::GAL1
UAS
-HIS3
TATA
-HIS3,URA3::GAL1
UAS
-GAL1
TATA
-
LacZ) was used as host in the yeast two-hybrid system. Yeast
cells were transformed with DNA following a modified lith-
ium acetate transformation protocol and grown in YPD or a
selection medium (46,47). Yeast host cells were first trans-
formed with pBDDR335–1185 and subsequently with plasmid
DNA prepared from a human fetal brain Matchmaker cDNA
library fused with the AD (Clontech). The transformants were
grown on selection plates without Trp, Leu and His supple-
mented with 25 mM 3-amino-1,2,4-triazole (Sigma, St Louis,
MO) (selection plates). Apparent His
+
clones were picked and
their phenotype confirmed by restreaking onto the selection
plates and by β-galactosidase filter assay as described below.
Candidates were then plated on medium without Leu supple-
mented with 10 µg/ml cycloheximide (Cyh) to isolate Leu
+
/
Trp–/Cyh
r
segregants, which had lost the bait plasmid. Finally,
plasmid DNA was propagated in E.coli cells (strain HB101)
and used in subsequent experiments.
Measurement of β
ββ
β-galactosidase activity
Filter assays for β-galactosidase were performed in z-buffer
(100 mM sodium phosphate, pH 7.0, 10 mM KCl, 1 mM
MgSO
4
) supplemented with 50 mM 2-mercaptoethanol and
0.07 mg/ml 5-bromo-4-chloro-3-indolyl-β-D-galactoside, after
transferring yeast colonies onto Hybond-N nylon membranes
(Amersham, Little Chalfont, UK). Filters were then incubated
at 37°C for 2–4 h until the color developed. β-Galactosidase
activities were also measured after disruption of cells as fol-
lows. Log phase yeast cells were resuspended in z-buffer and
disrupted by freeze–thawing with liquid nitrogen. An aliquot
was incubated at room temperature with 0.67 mg/ml o-nitroph-
enol-β-D-galactopyranoside in z-buffer supplemented with 2-
mercaptoethanol. Subsequently, OD
420
was measured to calcu-
late the enzyme activity by the equation of Miller (48).
Cell culture
A rat pheochromocytoma cell line (PC12) was maintained in
Dulbecco’s modified Eagle’s medium (DMEM) with heat-
inactivated 10% horse serum and 5% fetal bovine serum, 50 U/
ml penicillin and 0.1 mg/ml streptomycin at 37°C under a
humidified atmosphere of 5% CO
2
. For stimulation with
growth factor, cells were preincubated in serum-free medium
for 3 h and then treated in culture medium supplemented with
insulin or IGF-1 at a final concentration of 10
–6
–10
–9
Malong
with 500 µM sodium vanadate. Culture was continued at 37°C
for 2 h unless otherwise indicated. The treated cells were lysed
in lysis buffer (150 mM NaCl, 1% Triton X-100, 10 mM Tris–
HCl, pH 7.4, 5 mM EGTA, 500 µM sodium vanadate, 1.0 mM
phenylmethylsulfonyl fluoride, 18 µg/ml aprotinin, 50 µg/ml
leupeptin, 1 mM benzamidine and 0.7 µg/ml pepstatin) and the
supernatant fraction was obtained by centrifugation. HeLa
cells were maintained in the same conditions as for PC12 cells
except for using DMEM with 10% heat-inactivated fetal
956 Human Molecular Genetics, 1999, Vol. 8, No. 6
bovine serum and transfected with plasmid DNA by the lipo-
fection method (49).
Detection with antibody
A rabbit polyclonal antibody specific to human IRSp53 (anti-
IRSp53) was raised with the purified product of pGST–
IRSp53c and absorbed on GST–Sepharose to diminish the
reactivity to GST. Western blotting was performed following a
standard method (8,46). An aliquot containing 30 µgprotein
was fractionated by SDS–PAGE and transferred onto a nitro-
cellulose membrane (Schleicher & Schuell, Dassel, Germany)
by electroblotting. To detect IRSp53, membranes were treated
with a 1:2000 dilution of the anti-IRSp53 antibody, followed
by a 1:4000 dilution of horseradish peroxide-conjugated goat
anti-rabbit IgG (Sigma). To detect GFP fusion proteins, mem-
branes were treated with a 1:500 dilution of the monoclonal
anti-GFP antibody (Clontech) followed by a 1:1000 dilution of
horseradish peroxide-conjugated rabbit anti-mouse IgG (Dako,
Carpinteria, CA). The DRPLA protein was visualized with the
anti-DRPLA antibody as previously described (8). For immu-
noprecipitation, an aliquot containing 500 µg protein was first
incubated at 4°C with 10 µl of an indicated antibody for 3 h,
then overnight with protein G/protein A–agarose (Calbiochem,
San Diego, CA). Precipitates were washed three times and dis-
sociated by boiling with the SDS–PAGE buffer and subjected
to SDS–PAGE. Co-immunoprecipitated proteins were visual-
ized using the anti-GFP antibody, anti-DRPLA antibody or a
1:2000 dilution of an anti-phosphotyrosine antibody, RC20H
(Transduction Laboratories, Lexington, KY).
Confocal microscopy
Cells were treated as described previously (31). For IRSp53
detection, cells were treated with a 1:1000 dilution of anti-
IRSp53 antibody, followed by a 1:20 dilution of TRITC-conju-
gated swine anti-rabbit immunoglobulin (Dako). A fluorescent
image was obtained using a confocal microscope (Fluoview;
Olympus, Tokyo, Japan) equipped with an Ar laser with exci-
tation at 488 nm and detection at 510–530 nm bandpass for
GFP or with an He/Ne laser with excitation at 543 nm and
detection at 565 to 590 nm bandpass for TRITC.
In vitro translation and binding assay
In vitro translation was performed using the TNT coupled
reticulocyte lysate system (Promega, Madison, WI) with plas-
mid DNA. Translated products (2 µl) were fractionated by
SDS–PAGE and visualized with an antibody. In several exper-
iments, products were radiolabeled with [
35
S]methionine and
detected by autoradiography or quantitatively measured with a
phosphorimager (BAS2000; Fuji Film, Tokyo, Japan). For
binding assays, radiolabeled in vitro translation products were
incubated with 10 µg of GST fusion proteins prebound to
glutathione–Sepharose 4B beads (Pharmacia) in HKM
solution(10mMHEPES,pH7.2,142.5mMKCl,5mM
MgCl
2
, 1 mM EGTA, 0.2% NP-40) at 4°C for 2 h and then
washed with the HKM solution four times. Proteins bound to
the beads were dissociated and analyzed by SDS–PAGE fol-
lowed by autoradiography.
ABBREVIATIONS
AD, GAL4 activation domain; BD, GAL4 DNA-binding
domain; Cyh, cycloheximide; DMEM, Dulbecco’s modified
Eagle’s medium; DRPLA, dentatorubral-pallidoluysian atro-
phy; GFP, green fluorescent protein; GST, glutathione S-
transferase; HD, Huntington’s disease; IGF-1, insulin-like
growth factor-1; IRS, insulin receptor substrate;
IRSp53,insulin receptor substrate protein of 53 kDa; NLS,
nuclear localization signal; SBMA, spinal and bulbar muscular
atrophy; SCA, spinocerebellar ataxia.
ACKNOWLEDGEMENTS
We thank J. Reed for providing clones, A. Asaka and Y. Oht-
suka for technical assistance and K. Saito for manuscript prep-
aration. This study was supported in part by Grants for
Genome Research, Brain Research and Pediatric Research
from the Ministry of Health and Welfare, a Grant-in-Aid for
Specific Research on Priority Areas (Neuron Death) from the
Ministry of Education, Science and Culture and a Grant for
Liberal Harmonious Research Promotion System from the Sci-
ence and Technology Agency.
REFERENCES
1. Naito, H. and Oyanagi, S. (1982) Familial myoclonus epilepsy and chore-
oathetosis: hereditary dentatorubral-pallidoluysian atrophy. Neurology,
32, 798–807.
2. Nagafuchi, S., Yanagisawa, H., Sato, K., Shirayama, T., Ohsaki, E.,
Bundo, M., Takeda, T., Tadokoro, K., Kondo, I., Murayama, N., Tanaka,
Y., Kikushima, H., Umino, K., Kurosawa, H., Furukawa, T., Nihei, K.,
Inoue, T., Sano, A., Komure, O., Takahashi, M., Yoshizawa, T.,
Kanazawa, I. and Yamada, M. (1994) Dentatorubral and pallidoluysian
atrophy expansion of an unstable CAG trinucleotide on chromosome 12p.
Nature Genet.,6, 14–18.
3. Koide, R., Ikeuchi, T., Onodera, O., Tanaka, H., Igarashi, S., Endo, K.,
Takahashi, H., Kondo, R., Ishikawa, A. and Hayashi, T. (1994) Unstable
expansion of CAG repeat in hereditary dentatorubral-pallidoluysian atro-
phy (DRPLA). Nature Genet.,6, 9–13.
4. Komura, O., Sano, A., Nishino, N., Yamauchi, N., Ueno, S., Kondoh, K.,
Sano, N., Takahashi, M., Murayama, N., Kondo, I., Nagafuchi, S.,
Yamada, M. and Kanazawa, I. (1995) DNA analysis in hereditary denta-
torubral-pallidoluysian atrophy: correlation between CAG repeat length
and phenotypic variation and the molecular bases of anticipation. Neurol-
ogy,45, 143–149.
5. Nagafuchi, S., Yanagisawa, H., Ohsaki, E., Shirayama, T., Tadokoro, K.,
Inoue, T. and Yamada, M. (1994) Structure and expression of the gene
responsible for the triplet repeat disorder, dentatorubral and pallidoluy-
sian atrophy (DRPLA). Nature Genet.,8, 177–182.
6. Yazawa, I., Nukina, N., Hashida, H., Goto, J., Yamada, M. and
Kanazawa, I. (1995) Abnormal gene product identified in hereditary
dentatorubral-pallidoluysian atrophy (DRPLA) brain. Nature Genet.,10,
99–103.
7. Takano, T., Yamanouchi, Y., Nagafuchi, S. and Yamada, M. (1996)
Assignment of the dentatorubral and pallidoluysian atrophy (DRPLA)
gene to 12p 13.31 by fluorescence in situ hybridization. Genomics,15,
171–172.
8. Miyashita, T., Okamura-Oho, Y., Mito, Y., Nagafuchi, S. and Yamada,
M. (1997) Dentatorubral pallidoluysian atrophy (DRPLA) protein is
cleaved by caspase-3 during apoptosis. J. Biol. Chem.,272, 29238–
29242.
9. Nishiyama, K., Nakamura, K., Murayama, S., Yamada, M. and
Kanazawa, I. (1997) Regional and cellular expression of the dentatoru-
bral-pallidoluysian atrophy gene in brains of normal and affected individ-
uals. Ann. Neurol.,41, 599–605.
10. La Spada, A.R., Wilson, E.M., Lubahn, D.B., Harding, A.E. and Fisch-
beck, K.H. (1991) Androgen receptor gene mutations in X-linked spinal
and bulbar muscular atrophy. Nature,352, 77–79.
Human Molecular Genetics, 1999, Vol. 8, No. 6 957
11. The Huntington’s Disease Collaborative Research Group (1993) A novel
gene containing a trinucleotide repeat that is expanded and unstable on
Huntington’s disease chromosome. Cell,72, 971–983.
12. Orr, H.T., Chung, M.Y., Banfi, S., Kwiatkowski, T.J.Jr, Servadio, A.,
Beaudet, A.L., McCall, A.E., Duvick, L.A., Ranum, L.P. and Zoghbi,
H.Y. (1993) Expansion of an unstable trinucleotide CAG repeat in
spinocerebellar ataxia type 1. Nature Genet.,4, 221–226.
13. Pulst, S.M., Nechiporuk, A., Nechiporuk, T., Gispert, S., Chen, X.N.,
Lopes-Cendes, I., Pearlman, S., Starkman, S., Orozco-Diaz, G., Lunkes,
A., DeJong, P., Rouleau, G.A., Auburger, G., Korenberg, J.R., Figueroa,
C. and Sahba, S. (1996) Modulate expansion of a normally biallelic trinu-
cleotide repeat in spinocerebellar ataxia type 2. Nature Genet.,14,269–
276.
14. Kawaguchi, Y., Okamoto, T., Taniwaki, M., Aizawa, M., Inoue, M.,
Katayama, S., Kawakami, H., Nakamura, S., Nishimura, M. and
Akiguchi, I. (1994) CAG expansions in a novel gene for Machado–Joseph
disease at chromosome 14q32.1. Nature Genet.,8, 221–228.
15. Zhuchenko, O., Bailey, J., Bonnen, P., Ashizawa, T., Stockton, D.W.,
Amos, C., Dobyns, W.B., Subramony, S.H., Zoghbi, H.Y. and Lee, C.C.
(1997) Autosomal dominant cerebellar ataxia (SCA6) associated with
small polyglutamine expansions in the alpha 1A voltage-dependent cal-
cium channel. Nature Genet.,15, 62–69.
16. David, G., Abbas, N., Stevanin, G., Durr, A., Yvert, G., Cancel, G.,
Weber, C., Imbert, G., Saudou, F., Antoniou, E., Drabkin, H., Gemmill,
R., Giunti, P., Benomar, A., Wood, N., Ruberg, M., Agid, Y., Mandel,
J.L. and Brice, A. (1997) Cloning of the SCA7 gene reveals a highly
unstable CAG repeat expansion. Nature Genet.,17, 65–70.
17. Ikeda, H., Yamaguchi, M., Sugai, S., Aze, Y., Narumiya, S. and Kaki-
zuka, A. (1996) Expanded polyglutamine in the Machado–Joseph disease
protein induces cell death in vitro and in vivo.Nature Genet.,13,196–
202.
18. Perutz, M.F., Johnson, T., Suzuki, M. and Finch, J.T. (1994) Glutamine
repeats as polar zippers: their possible role in inherited neurodegenerative
diseases. Proc. Natl Acad. Sci. USA,91, 5355–5358.
19. Mangiarini, L., Sathasivam, K., Seller, M., Cozens, B., Harper, A., Heth-
erington, C., Lowton, M., Trottier, Y., Lehrach, H., Davis, S.W. and
Bates, G.P. (1996) Exon 1 of the HD gene with an expanded CAG repeat
is sufficient to cause a progressive neurological phenotype in transgenic
mice. Cell,87, 493–506.
20. Igarashi, S., Koide, R., Shimohata, T., Yamada, M., Hayashi, Y., Takano,
H., Date, H., Oyake, M., Sato, T., Sato, A., Egawa, S., Ikeuchi, T., Tan-
aka, H., Nakano, R., Tanaka, K., Hozumi, I., Inuzuka, T., Takahashi, H.
and Tsuji, S. (1998) Suppression of aggregate formation and apoptosis by
transglutaminase inhibitors in cells expressing truncated DRPLA protein
with an expanded polyglutamine stretch. Nature Genet.,18, 111–117.
21. Sisodia, S.S. (1998) Nuclear inclusions in glutamine repeat disorders: are
they pernicious, coincidental, or beneficial? Cell,95, 1–4.
22. Goldberg, Y.P., Nicholson, D.W., Rasper, D.M., Kalchman, M.A., Koide,
H.B., Graham, R.K., Bromm, M., Kazemi-Esfarjani, P., Thornberry,
N.A., Vaillancourt, J.P. and Hayden, M.R. (1996) Cleavage of huntingtin
by apopain, a proapoptotic cysteine protease, is modulated by the poly-
glutamine tract. Nature Genet.,13, 442–449.
23. Wellington, C.L., Ellerby, L.M., Hackam, A.S., Margolis, R.L., Trifiro,
M.A., Singaraja, R., McCutcheon, K., Salvesen, G.S., Propp, S.S.,
Bromm, M., Rowland, K.J., Zhang, T., Rasper, D., Roy, S., Thornberry,
N., Pinsky, L., Kakizuka, A., Ross, C.A., Nicholson, D.W., Bredesen,
D.E. and Hayden, M.R. (1998) Caspase cleavage of gene products associ-
ated with triplet expansion disorders generates truncated fragments con-
taining the polyglutamine tract. J. Biol. Chem.,273, 9158–9167.
24. Yeh, T.C., Ogawa, W., Danielsen, A.G. and Roth, R.A. (1996) Character-
ization and cloning of a 58/53-kDa substrate of the insulin receptor tyro-
sine kinase. J. Biol. Chem.,271, 2921–2928.
25. Johnston, B.M., Mallard, E.C., Williams, C.E. and Gluckman, P.D.
(1996) Insulin-like growth factor-1 is a potent neuronal rescue agent after
hypoxic-ischemic injury in fetal lambs. J. Clin. Invest.,97, 300–308.
26. Parrizas, M., Saltiel, A.R. and LeRoith, D. (1997) Insulin-like growth fac-
tor 1 inhibits apoptosis using the phosphatidylinositol 3'-kinase and
mitogen-activated protein kinase pathways. J. Biol. Chem.,272, 154–161.
27. Semenov, M.V. and Synder, M. (1997) Human dishevelled genes consti-
tute a DHR-containing multigene family. Genomics,42, 302–310.
28. Zhou, J., Liyanage, U., Mediana, M., Ho, C., Simons, A.D., Lovett, M.
and Kosik, S. (1997) Presenilin 1 interaction in the brain with a novel
member of the armadillo family. NeuroReport,8, 2085–2090.
29. Watanabe, M., Isobe, T., Ichimura, T., Kuwano, R., Takahashi, Y. and
Kondo, H. (1993) Molecular cloning of rat cDNA for beta and gamma
subtypes of 14-3-3 protein and developmental changes in expressing of
their mRNAs in the nervous system. Brain Res. Mol. Brain Res.,17, 135–
146.
30. Cadigan, K.M. and Nusse, R. (1997) Wnt signaling: a common theme in
animal development. Genes Dev.,11, 3286–3305.
31. Miyashita, T., Nagao, K., Ohmi, K., Yanagisawa, H., Okamura-Oho, Y.
and Yamada, M. (1998) Intracellular aggregate formation of dentatoru-
bral-pallidoluysian atrophy (DRPLA) protein with the extended poly-
glutamine. Biochem. Biophys. Res. Commun.,249, 96–102.
32. Yu, H., Chen, J.K., Feng, S., Dalgarno, D.C., Brauer, A.W. and Schreiber,
S.L. (1994) Structural basis for the binding of proline rich peptides to SH3
domains. Cell,76, 933–945.
33. Feng, S., Chen, J.K., Yu, H., Simon, J.A. and Schreiber, S.L. (1994) Two
binding orientations for peptides to the Src SH3 domain: development of
a general model for SH3–ligand interactions. Science,266, 1241–1247.
34. Gusella, J. and MacDonald, M.E. (1998) Huntingtin: a single bait hooks
many species. Curr. Opin. Neurobiol.,8, 425–430.
35. Engelender, S., Sharp, A.H., Colomer, V., Tokito, M.K., Lanahan, A.,
Worley, P., Holzbaur, E.L.F. and Ross, C.A. (1997) Huntingtin-associ-
ated protein 1 (HAP1) interacts with the p150
Glued
subunit of dynactin.
Hum. Mol. Genet.,6, 2205–2212.
36. Liu, Y.F., Deth, R.C. and Devys, D. (1997) The domain-dependent asso-
ciation of huntingtin with epidermal growth factor receptor signaling
complexes. J. Biol. Chem.,272, 8121–8124.
37. Sittler, A., Walter, S., Wedeemeyer, N., Hasenbank, R., Scherzinger, E.,
Eickhoff, H., Bates, G.P., Lehrach, H. and Wanker, E.E. (1998) SH3GL3
associates with the huntingtin exon1 protein and promotes the formation
of polygln-containing protein aggregates. Mol. Cell,2, 427–436.
38. Wood, J.D., Yuan, J., Margolis, R.L., Colomer, V., Duan, K., Kushi, J.,
Kamishky, Z., Kleiderlein, J.J., Sharp A.H. and Ross, C.A. (1998)
Atrophin-1, the DRPLA gene product, interacts with two families of WW
domain-containing proteins. Mol. Cell Neurosci.,11, 149–160.
39. White, M.F. and Yunush, L. (1998) The IRS-signaling system: a network
of docking proteins that mediate insulin and cytokine action. Curr. Top.
Microbiol. Immunol.,228, 179–208.
40. Clark, S.F., Martin, S., Carozzi, A.J., Hill, M.M. and James, D.E. (1998)
Intracellular localization of phosphatidylinositide 3-kinase and insulin
receptor substrate-1 in adipocytes: potential involvement of a membrane
skeleton. J. Cell. Biol.,140, 1211–1225.
41. Yeh, T.C., Li, W., Keller, G.A. and Roth, R.A. (1998) Disruption of a
putative SH3 domain and the proline-rich motifs in the 53-kDa substrate
of the insulin receptor kinase does not alter its subcellular localization or
ability to serve as a substrate. J. Cell Biochem.,68, 139–150.
42. Kozak, M. (1991) Structural features in eukaryocytic mRNAs that modu-
late the initiation of translation. J. Biol. Chem.,266, 19867–19870.
43. Bondy, C.A. (1991) Transient IGF-I gene expression during the matura-
tion of functionally related central projection neurons. J. Neurosci.,11,
3442–3455.
44. Beck, K.D., Powell-Braxton, L., Widmer, H.R., Valverde, J. and Hefti, F.
(1995) Igf1 gene disruption results in reduced brain size, CNS hypomyeli-
nation and loss of hippocampal granule and striatal parvalbumin-contain-
ing neurons. Neuron,14, 717–730.
45. Dudek, H., Datta, S.R., Franke, T.F., Birnbaum, M.J., Yao, R., Cooper,
G.M., Segal, R.A., Kaplan, D.R. and Greenberg, M.E. (1997) Regulation
of neuronal survival by the serine-threonine protein kinase Akt. Science,
275, 661–665.
46. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A
Laboratory Manual, 2nd Edn. Cold Spring Harbor Laboratory Press,
Cold Spring Harbor, NY.
47. Gietz, R.D., Wood, R.A., Manivasakam, P. and Schiestl, R.H. (1996)
Yeast Growth and Yeast Transformation in Cell Biology: A Laboratory
Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
48. Miller, J.H. (1972) Experiments in Molecular Genetics. Cold Spring Har-
bor Laboratory Press, Cold Spring Harbor, NY.
49. Felgner, P.L., Gadek, T.R., Holm, M., Roman, R., Chan, H.W., Wenz,
M., Northrop, J.P., Ringold, G.M. and Danielsen, M. (1987) Lipofection:
a highly effective, lipid-mediated DNA-transfection procedure. Proc.
Natl Acad. Sci. USA,84, 7413–7417.
958 Human Molecular Genetics, 1999, Vol. 8, No. 6
