Structural bases for recognition of Anp32⁄ LANP proteins
Cesira de Chiara, Rajesh P. Menon and Annalisa Pastore
National Institute for Medical Research, The Ridgeway, London, UK
The leucine-rich repeat acidic nuclear protein (Anp32a ⁄
LANP) is a member of the Anp32 family of acidic
nuclear evolutionarily-conserved phosphoproteins,
which present a broad range of activities [1]. They are
characterized by the presence of a highly conserved
N-terminal domain containing leucine-rich repeats
(LRRs), motifs known to mediate protein–protein inter-
actions, and of a C-terminal low-complexity region,
mainly composed of polyglutamates.
Since their first description in the neoplastic B-lym-
phoblastoid cell line and their reported association
with proliferation [2], several Anp32a homologs, all
derived from a common ancestor gene by subsequent
duplication events, have been isolated in different tis-
sues and differently named [1]. Members of the Anp32
family are widely recognized as nucleo-cytoplasmic
shuttling phosphoproteins that are implicated in differ-
ent signaling pathways and in a number of important
cellular processes, which include cell proliferation, dif-
ferentiation, caspase-dependent and caspase-indepen-
dent apoptosis, tumor suppression, regulation of
mRNA trafficking and stability, histone acetyltransfer-
ase inhibition, and regulation of microtubule-based
functions [1,3].
The diverse activities of Anp32 proteins are achieved
through an articulated network of interactions with
several cellular partners. Among them, two proteins
are of particular relevance from the clinical point of

unmodified domains bind with very weak (millimolar) affinity, thus sug-
gesting the necessity either for an additional partner (e.g. other regions of
either or both proteins or a third molecule) or for a post-translational
modification. Finally, we identified by two-hybrid screening a new partner
of the LRR domain, i.e. the microtubule plus-end tracking protein
Clip 170 ⁄ Restin, known to regulate the dynamic properties of microtubules
and to be associated with severe human pathologies.
Abbreviations
Anp32a ⁄ LANP, leucine-rich repeat acidic nuclear protein; Atx1, ataxin-1; AXH, ataxin-1 homology; Gal-X, 5-bromo-4-chloroindol-3-yl
b-
D-galactoside; GST, glutathione S-transferase; HSQC, heteronuclear single quantum coherence; LRR, leucine-rich repeat; MAP,
microtubule-associated protein; PP2A, phosphatase 2A; RDC, residual dipolar coupling; SCA1, spinocerebellar ataxia 1; TCEP,
Tris(2-carboxyethyl)phosphine; +TIP, plus-end tracking proteins.
2548 FEBS Journal 275 (2008) 2548–2560 ª 2008 The Authors Journal compilation ª 2008 FEBS
domain of Anp32 binds and strongly inhibits the
enzyme catalytic subunit PP2A-C [3–7], whose struc-
ture in a heterotrimeric complex with the scaffolding
A subunit and the regulatory B¢⁄B56 ⁄ PR61 subunit
was solved recently [13,14]. Although the role of phos-
phorylation in recognition remains debatable, interac-
tion between Anp32a and PP2A-C has been
independently confirmed by high-throughput yeast
two-hybrid screening [15].
Involvement of Anp32 in spinocerebellar ataxia
type 1 (SCA1) pathogenesis was also suggested, on the
basis of the observation of an interaction with the
SCA1 gene product ataxin-1 (Atx1) [16]. This protein
belongs to a family involved in neurodegenerative dis-
eases caused by anomalous expansion of polyQ tracts
[17]. In SCA1, expanded Atx1 forms nuclear inclusions

by NMR spectroscopy. This technique, which does not
need crystallization, also provides a powerful and flexi-
ble method for mapping binding interfaces. Our struc-
ture, as described in the following sections, reveals the
presence of two extra N-terminal LRR motifs not
observed in the crystal, and allows accurate definition
of the C-terminal domain boundary. Experimental
determination of the dynamic features of the domain
in solution, together with a comparison with the struc-
ture of the spliceosomal U2A¢ in complex with U2B¢¢
[20], suggests new insights into the mechanism of
Anp32 LRR–protein recognition. By a combination of
chemical shift perturbation techniques, molecular
docking and two-hybrid screening, we also probed the
interaction with Atx1 and PP2A, and identified a new
partner of the Anp32a LRR domain.
Results
Description of the Anp32a LRR domain structure
in solution
The construct used for structure determination covers
residues 1–164 of the mouse Anp32a sequence [21].
These boundaries were chosen to include the sequence
up to the beginning of the acidic repeats, where
sequence conservation breaks down (data not shown).
The resulting sample was stable and well behaved, pro-
viding NMR spectra typical of a folded monodisperse
globular domain. The final representative family of the
10 lowest-energy structures after water refinement
could be superimposed on the average structure
with overall rmsd values of 0.71 ± 0.15 A

b
6
b
7
b
8
) shows the secondary structure elements spatially
arranged in the typical right-handed solenoid, which
forms a curved horse-shoe fold. A canonical parallel
b-sheet is present on the concave side, whereas the
convex surface contains both well-defined but irregular
secondary structure elements (in the first and second
repeats) and helical regions. Among these, h
1
and h
6
are regular a-helices whereas h
2
,h
3
,h
4
and h
5
share
features of 3
10
-helices. A search for tertiary structure
similarity performed by dali [22] indicates that the
Anp32a LRR domain structure belongs to the SDS22-

whereas the first repeat (residues 19–43 in our struc-
ture) was considered by these authors to be an
N-CAP. The presence of the first N-terminal LRR had
also not been predicted [1], probably because of the
low sequence conservation in this region. In our opin-
ion, this region constitutes instead a bona fide full
repeat.
Residues 147–149, which are truncated in the crystal
structure, form a b-hairpin (b
7
) with the strand 143–
145 (b
6
). This region shares a remarkable similarity
with U2A¢, the two protein with the highest structural
homology: the two proteins can be superimposed with
2A
˚
rmsd as calculated over 140 residues and have a
dali z score of 17 (Fig. 2A,C). Although mainly
unstructured, a short additional strand C-terminal to
the hairpin (b
8
) is present in some of the NMR struc-
tures in region 149–164 (residues 151–153). This region,
which constitutes the linker between the LRR domain
and the acidic repeats, is thought to be involved in
interactions with the INHAT complex and the phos-
phorylation-dependent tumor suppressive ⁄ proapoptotic
activity, which have been mapped to residues 150–180

1
H–
15
N heteronuclear single quantum coherence
(HSQC) spectrum at pH 7 (Met3, Asp4, Ile30, Glu31,
Ile34, Glu35 and Val52) the last five belong to regions
without a regular secondary structure. All seven resi-
dues, including Met3 and Asp4 at the N-terminus of
h
1
, cluster together in the structure, suggesting that
they experience chemical or conformational exchange.
The C-terminus is unstructured and highly mobile
approximately from residue 154 onwards.
It is also interesting to note a clear correlation
between T
1
and RDC values and the secondary struc-
ture: the concave b-sheet is characterized by shorter
T
1
and positive RDC values, the latter indicating that
A
B
Fig. 1. Solution structure of the LRR
domain of murine Anp32a. (A) NMR bundle
of the 10 best structures in terms of
energy. (B) Average structure as obtained
by the
WHEATSHEAF algorithm [62]. Two

proteins on the basis of the U2A¢–U2B¢¢ structure
The structural similarity with U2A¢, whose structure is
known in a complex with its target U2B¢¢ [20], may
provide valuable hints on how Anp32 interacts with its
partners. We therefore analyzed this complex and
compared its features with those of our structure. Rec-
ognition of the two molecules occurs by fitting a helix
of U2B¢¢ (residues 25–35) into the concave surface of
the U2A¢ LRR (Fig. 4). The size complementarity is
almost perfect. The nearby N-terminal b-hairpin of
Table 1. Structural statistics for the calculations of the Anp32a
LRR domain.
Final NMR restraints
Total distance restraints
a
5151
Unambiguous ⁄ ambiguous 3774 ⁄ 1376
Intraresidue 2021
Sequential 1075
Medium (residue i to i + j, j = 1–4) 663
Long-range (residue i to i + j, j > 4) 1392
Dihedral angle restraints
b
F 89
w 89
1
D
NH
RDC 92
Hydrogen bonds 20

Generously allowed regions 0.4
Disallowed regions 0.8
a
Calculated for the 10 lowest-energy structures after water refine-
ment.
b
Derived from
3
J(HN, Ha) coupling constants and TALOS [48].
c
Calculated for residues 3–154. The more positive the score, the
better it is. Problematic structures typically have scores around )3.
Wrong structures have scores lower than )3.
d
Calculated for resi-
dues 1–154.
A
B
C
Fig. 2. Comparison between the NMR (A) and the X-ray (B) struc-
tures of the Anp32a LRR domain, and U2A¢ (C) [20]. The coordi-
nates were first superimposed using the
DALI server, and then
displaced.
C. de Chiara et al. Study of the interactions of the Anp32a LRR module
FEBS Journal 275 (2008) 2548–2560 ª 2008 The Authors Journal compilation ª 2008 FEBS 2551
U2A¢ (residues 13–26) provides further interactions by
wrapping around the other molecule on one side.
There is also a good charge complementarity, as the
concave surface of the U2A¢ LRR is negatively

subunit B¢⁄B56 ⁄ PR61, and the catalytic domain C.
Interaction with Anp32 has been shown to involve the
catalytic subunit [15,32] and to inhibit its catalytic
activity, both in the absence and in the presence of the
scaffold subunit A and the regulatory subunit B, with
apparent K
i
in the low nanomolar range [4]. This implies
that the interaction involves an exposed region of
PP2A-C, without appreciable contributions from the
other two subunits. Anp32 is also known to inhibit
PP2A in a noncompetitive manner, i.e. without binding
to the active site of the enzyme [4]. Finally, antibodies
recognizing the fourth LRR of Anp32e ⁄ Cpd1 (resi-
dues 87–101) are known to block the inhibitory PP2A
activity of Anp32e in protein extracts [7]. Taken
together, these findings limit the region of interaction to
the only exposed surface of PP2A-C that contains a
semiexposed helix (residues 222–232).
The model of an Anp32 LRR–PP2A complex, built
using complex U2A¢–U2B¢¢ as a template, shows that,
by analogy with this structure, helix 222-232 of PP2A-
C protrudes out enough to fit well into the groove
formed by the concave surface of the LRR domain.
Stabilizing interactions could form between His230 of
PP2A-C and Asp119 and Asn94 of Anp32a. A salt
bridge could form between Glu226 of PP2A-C and
Lys67, Lys68 and Lys91.
Testing the interaction with Atx1 experimentally
Interaction between the Anp32a LRR domain and the

suggesting a decrease in the Trp solvent exposure con-
sequent to interaction (data not shown). However, the
decrease in fluorescence intensity was far from reach-
ing a plateau even at the highest Anp32a ⁄ Atx1 ratio
tested (60 : 1).
Fig. 5. Probing the interaction between the Anp32a LRR domain
and the AXH domain of Atx1 by chemical shift perturbation. Super-
imposition of the HSQC spectra of a 0.2 m
M solution of
15
N-labeled
Anp32a LRR domain in 20 m
M Tris (pH 7.0) and 2 mM TCEP,
recorded at 600 MHz and 27 °C in the absence (blue) and in the
presence (red) of a three-fold excess of unlabeled Atx1 AXH
domain.
C. de Chiara et al. Study of the interactions of the Anp32a LRR module
FEBS Journal 275 (2008) 2548–2560 ª 2008 The Authors Journal compilation ª 2008 FEBS 2553
This evidence indicates that interaction between the
two domains is very weak, i.e. with binding constants
in the millimolar range. Although such binding is defi-
nitely too weak to be significant, it is certainly possible
that, in vivo, the interaction is enhanced either by other
regions of the two molecules or by post-translational
modifications that are absent in our assays.
Identification of new potential partners of the
Anp32 LRR domain
To identify new partners specific for the Anp32a LRR
domain, we used a construct spanning the same region
studied by structural techniques (residues 1–164) as a

Anp32a or with antibodies to histone H3 as a negative
control. The proteins from immunoprecipitation com-
plexes were subjected to western blot analysis using
antibodies to Clip 170. Clip 170 was associated with
the complex pulled down by antibodies to Anp32a
but not with the one pulled down by antibodies to
histone H3 (Fig. 7B,C).
Discussion
Here, we have explored the interaction properties of
the LRR domain of Anp32, a family of LRR proteins
potentially implicated in several important cellular
pathways. Two particularly interesting interactions
have been described, with the PP2A phosphatase and
with Atx1, two proteins of high medical importance.
We first determined the domain boundaries of the
domain by solving the solution structure at high reso-
lution of a fragment spanning the whole conserved
region up to some highly acidic repeats containing EA-
EEE motifs. We show that the domain contains a
compact and rigid fold with five LRRs and a
C-capping motif. The structural information was used
to model the interaction with PP2A, which is known
to be mainly mediated by the PP2A-C subunit. We
suggest that, by analogy with the mode of recognition
of U2B¢¢ by U2A¢, which has the highest structural simi-
larity with the Anp32a LRR domain, the interaction
+
+
-
-

was no growth on QD plates when either
the N-terminus of Clip 170 (residues 1–
1164, BD-Clip.NT) or the Gal4 DNA-binding
domain was used as prey. (B) Amino acid
sequence of the region of Clip 170 interact-
ing with Anp32a.
Study of the interactions of the Anp32a LRR module C. de Chiara et al.
2554 FEBS Journal 275 (2008) 2548–2560 ª 2008 The Authors Journal compilation ª 2008 FEBS
involves the helix-spanning residues 222–230 of PP2A-
C [13,14]. This region is the only element of PP2A-C
protruding out from the PP2A trimer, and its size and
shape mean that it could easily fit into the complemen-
tary concave surface of the Anp32a LRR domain.
We tested binding to the AXH domain of Atx1
experimentally by chemical shift perturbation assays.
We observed only very minor effects, which are com-
patible, at the very best, with millimolar affinities.
The effects could be observed only at low ionic
strength, suggesting that the interaction is mainly of
an electrostatic nature and is nonspecific. Would our
results shed doubts on an interaction originally
observed by two-hybrid screening? On the one hand,
it is interesting to note that none of the high-through-
put studies of the Atx1 interactome has reported any
evidence for this interaction [34,35]. On the other
hand, however, very recent data provide the first evi-
dence of a functional link between Anp32a and Atx1,
showing that Atx1 relieves the transcriptional repres-
sion induced by Anp32a in complex with E4F [36].
As addition of exogenous Anp32a restores repression,

Fig. 7. Anp32a and Clip 170 associate with each other in HeLa and transfected COS cells. (A) Colocalization of Clip 170 and Anp32a in COS
cells that were transfected with a plasmid vector carrying V5-tagged Clip 170 and c-Myc-tagged Anp32a. Cells were analyzed by confocal
microscopy. Clip 170 was localized in the microtubule network (green), and Anp32a (red) was predominantly nuclear, with some cells show-
ing localization in the microtubules. The merged image shows colocalization of the proteins in the microtubules. (B) Expression of endoge-
nous proteins in HeLa cells. HeLa cells were lysed in RIPA buffer, and input controls and immunoprecipitated samples were probed with
the antibodies shown. (C). Interaction of endogenous Clip 170 and Anp32a in HeLa cells. HeLa cells were lysed in RIPA buffer and immuno-
precipitated as above with antibodies to histone H3 or antibodies to Anp32a. Proteins were transferred onto a poly(vinylidene difluoride)
membrane and probed with antibodies to Clip 170.
C. de Chiara et al. Study of the interactions of the Anp32a LRR module
FEBS Journal 275 (2008) 2548–2560 ª 2008 The Authors Journal compilation ª 2008 FEBS 2555
with microtubules and with other MAPs, and to regu-
late the dynamic properties of microtubules [33]. Iden-
tification of this new potential partner is particularly
interesting, because Anp32a has already been reported
to be involved in microtubule dynamics via its interac-
tion with several members of the family of MAPs, i.e.
MAP1B, MAP2, and MAP4 [39–41]. The interaction
with MAP1B was suggested to modulate the effects of
MAP1B in neurite extension [41]. Microtubule +TIPs
have also been shown to be involved in modulating
neuronal growth cones, the motile tips of growing
axons [42,43]. Interaction of Clip 170 with micro-
tubules has been suggested to be influenced by
phosphorylation, as phosphorylation by a rampamy-
cin-sensitive kinase (fluorescence recovery after photo-
bleaching; FRAP) increases the interaction of Clip 170
with microtubules [44]. Interestingly, in our coimmu-
noprecipitation experiments, the Clip 170 band
appeared to be more intense when the cell lysates
incorporated a cocktail of phosphatase inhibitors, sug-

N]ammonium sulfate
and [
13
C]glucose as the sole sources of nitrogen and carbon
respectively. The cells were grown at 37 °C until an attenu-
ance (D) at 600 nm of 0.5 was reached, and then cooled
to 18 °C, induced with isopropyl thio-b-d-galactoside
(0.5 mm), and harvested after overnight expression. A stan-
dard purification protocol was performed, using Pharmacia
GST–Sepharose resin (GE Healthcare). Cleavage of the
GST tag was achieved overnight at room temperature using
the PreScission protease (GE Healthcare). The protein was
further purified by HPLC size exclusion chromatography,
using a prepacked HiLoad 16 ⁄ 60 Superdex 75 prep grade
column (Pharmacia). The concentration of the NMR sam-
ple used for structural studies was typically in the range
0.3–0.7 mm, in a buffer containing 10 mm Tris ⁄ HCl and
2mm Tris(2-carboxyethyl)phosphine (TCEP) at pH 7.0 in
90% H
2
O ⁄ 10% D
2
O. All the NMR experiments were per-
formed at 27 °C on Bruker Advance and Varian Inova
spectrometers, both equipped with cryoprobes and operat-
ing at 14.1 and 18.8 T, respectively, and on a Varian Inova
spectrometer operating at 14.1 T. Samples of the Atx1
AXH domain (residues 567–689 and 567–694) were pro-
duced as previously described [18].
Experimental restraints

a hydrogen bond was consistently observed in at least
50% of the structures inspected at an advanced stage of
the refinement.
1
D
NH
RDCs were measured at 27 °C,
aligning the protein in 5% n-dodecyl-penta(ethylene gly-
col) ⁄ n-hexanol (r = 0.92) using a buffer composed of
20 mm Tris ⁄ HCl, 2 mm TCEP and 0.02% NaN
3
at
pH 7.0. The liquid crystalline medium gave a stable quad-
rupolar splitting of the D
2
O signal of 21 Hz. The final
concentration of the protein in this medium was
 0.37 mm.92
1
J
NH
splittings were obtained from a
J-modulated
15
N–
1
H HSQC spectrum [49] for NH vectors
with a heteronuclear NOE value higher than 0.75 and
used for the purpose of structure validation using the
program module [50]. The rmsd in hertz from RDC

gram (version 1.2) [51]. A typical run consisted of nine iter-
ations. At each iteration, 20 structures were calculated by
simulated annealing using the standard cns protocol [52]
with numbers of steps equal to 15 000 and 12 000 in the
first and second cooling stages of the annealing, respec-
tively. Floating assignment for prochiral groups and correc-
tion for spin diffusion during iterative NOE assignment
were applied as previously described [53,54]. At the end of
each iteration, the best seven structures in terms of lowest
global energy were selected and used for assignment of
additional NOEs during the following iteration. In the final
aria run, the number of structures generated in iteration 8
was increased to 100, and after refinement by molecular
dynamics simulation in water of the 50 lowest-energy struc-
tures [55], the 10 lowest-energy structures were selected as
representative of the Anp32a LRR domain structure and
used for statistical analysis. In the final iteration, 3774
unambiguous and 1377 ambiguous NOEs were assigned.
Among the 5151 total NOEs, 2021 were intraresidue, 1075
sequential, 663 medium range, and 1392 long range. Struc-
ture quality was evaluated using the programs procheck
[56] and whatif [57]. The coordinates are deposited with
the Protein Data Bank (accession code 2jqd).
Comparative modeling
The structure of an Anp32a–PP2A complex was modeled
on the U2A¢–U2B¢¢ coordinates (1a9n) [20]. The available
information strongly indicates that the interaction is domi-
nated by the C subunit of PP2A. Of this, the main region
that protrudes out into solution and is not protected by
interactions with the other two subunits comprises

stepwise additions of a 0.87 mm stock solution of Anp32a
LRR domain up to a 60 : 1 ratio. The data were evaluated
using the origin program package (Micro-Cal Software,
Bletchley, UK).
Yeast two-hybrid analysis
The DNA fragment encoding the murine Anp32a N-termi-
nus (1–164 amino acids) was cloned into the pGBKT7 vec-
tor (Clontech, Mountain View, CA, USA) for expression as
a Gal4 DNA-binding domain fusion protein. This bait was
transformed into an AH 109 yeast strain and used to screen
a human brain two-hybrid cDNA library from Clonetech
as previously described [60]. DNAs recovered from clones
selected by growth in quadruple-dropout media and Gal-X
overlay assays were sequenced and compared with known
sequences.
Confocal microscopy
cDNAs encoding full-length Anp32a and Clip 170 (Gene-
Service, IMAGE 3592614) were cloned into the
pBudCE4.1 vector (Invitrogen, Paisley, UK). The immu-
nofluorescence assay was carried out essentially as
described previously [61]. Briefly, COS cells were grown
overnight in chamber slides and transfected with
pBudCE4.1 vector expressing V5-tagged Clip 170 and c-
Myc-tagged Anp32a. Forty-eight hours after transfection,
cells were fixed using 4.0% paraformaldehyde, permeabi-
lized with 0.2% Triton X-100 ⁄ NaCl ⁄ P
i
and probed with
fluorescein isothiocyanate-conjugated antibodies to V5
(Invitrogen) and Cy3-conjugated antibodies to c-Myc

times with RIPA buffer before being resuspended in
SDS ⁄ PAGE sample buffer. Samples and input controls
were analyzed by PAGE and western blotting using mono-
clonal antibodies to Clip 170, histone H3 or Anp32a.
Acknowledgements
We thank Drs N. Q. McDonald and B. O’Hara (Bir-
beck College, London) for providing the Lanp clone,
which was produced from a cDNA originally provided
by Dr A. Matilla (ICH, London), Filippo Prischi for
help with the gromacs software, and Dr L Masino for
helpful discussions. The project is under the Eurosca
consortium.
References
1 Matilla A & Radrizzani M (2005) The Anp32 family of
proteins containing leucine-rich repeats. Cerebellum 4,
7–18.
2 Malek SN, Katumuluwa AI & Pasternack GR (1990)
Identification and preliminary characterization of two
related proliferation-associated nuclear phosphopro-
teins. J Biol Chem 265, 13400–13409.
3 Santa-Coloma TA (2003) Anp32e (Cpd1) and related
protein phosphatase 2 inhibitors. Cerebellum 2, 310–320.
4 Li M, Guo H & Damuni Z (1995) Purification and
characterization of two potent heat-stable protein inhib-
itors of protein phosphatase 2A from bovine kidney.
Biochemistry 34, 1988–1996.
5 Li M, Makkinje A & Damuni Z (1996a) The myeloid
leukemia-associated protein is a potent inhibitor of pro-
tein phosphatase 2A. J Biol Chem 271, 11059–11062.
6 Li M, Makkinje A & Damuni Z (1996b) Molecular

JW, Strack S, Jeoffrey PD & Shi Y (2006) Structure of
the protein phosphatase 2A holoenzyme. Cell 127,
1239–1251.
14 Cho US & Xu W (2007) Crystal structure of a protein
phosphatase 2A heterotrimeric holoenzyme. Nature 445,
53–57.
15 Stelzl U, Worm U, Lalowski M, Haenig C, Brembeck
FH, Goehler H, Stroedicke M, Zenkner M, Schoenherr
A, Koeppen S et al. (2005) A human protein–protein
interaction network: a resource for annotating the pro-
teome. Cell 122, 957–968.
16 Matilla A, Koshy BT, Cummings CJ, Isobe T, Orr HT
& Zoghbi HY (1997) The cerebellar leucine-rich acidic
nuclear protein interacts with ATX1. Nature 389, 974–
978.
17 Masino L & Pastore A (2002) Glutamine repeats: struc-
tural hypotheses and diseases. Biochem Soc Trans 29,
41–60.
18 de Chiara C, Giannini C, Adinolfi S, de Boer J, Guida S,
Ramos A, Jodice C, Kioussis D & Pastore A (2003) The
AXH module: an independently folded domain common
to ataxin-1 and HBP1. FEBS Lett 551, 107–112.
19 Huyton T & Wolberger C (2007) The crystal structure
of the tumour suppressor protein pp32 (Anp32a):
Study of the interactions of the Anp32a LRR module C. de Chiara et al.
2558 FEBS Journal 275 (2008) 2548–2560 ª 2008 The Authors Journal compilation ª 2008 FEBS
structural insights into Anp32 family of proteins. Prot
Sci 16, 1308–1315.
20 Price SR, Evans PR & Nagai K (1998) Crystal structure
of the spliceosomal U2B¢¢–U2A¢ protein complex bound

J Am Chem Soc 104, 4546–4559.
29 Maciejewski M, Liu D, Prasad R, Wilson S & Mullen
G (2000) Backbone dynamics and refined solution struc-
ture of the N-terminal domain of DNA polymerase
beta. Correlation with DNA binding and dRP lyase
activity. J Mol Biol 296, 229–253.
30 Sanders CR, Hare BJ, Howard KP & Prestegard JH
(1993) Magnetically-oriented phospholipid micelles as a
tool for the study of membrane-associated molecules.
Prog Nucl Magn Reson Spectrosc 26, 421–444.
31 Scherly D, Dathan NA, Boelens W, van Venrooij WJ &
Mattaj IW (1990) The U2B¢¢ RNP motif as a site of
protein–protein interaction. EMBO J 9, 3675–3681.
32 Chen S, Li B, Grundke-Iqbal I & Iqbal K (2008) I
1
PP2A
affects Tau phosphorylation via association with the
catalytic subunit of protein phosphatase 2A. J Biol
Chem doi: 10.1074/jbc.M709852200.
33 Galjart N (2005) CLIPs and CLASPs and cellular
dynamics. Nat Rev Mol Cell Biol 6, 487–498.
34 Lim J, Hao T, Shaw C, Patel AJ, Szabo G, Rual JF,
Fisk CJ, Li N, Smolyar A, Hill DE et al. (2006) A
protein–protein interaction network for human inher-
ited ataxias and disorders of Purkinje cell degeneration.
Cell 125, 801–814.
35 Gould R, Hubank M, Hunt A, Holton J, Menon RP,
Revesz T, Pandolfo M & Matilla-Duenas A (2007)
Down-regulation of the dopamine receptor D2 in mice
lacking ataxin 1. Hum Mol Genet 16, 2122–2134.

(2004) NGF-induced axon growth is mediated by local-
ized inactivation of GSK-3beta and functions of the
microtubule plus end binding protein APC. Neuron 42,
897–912.
44 Choi JH, Bertram PG, Drenan R, Carvalho J, Zhou
HH & Zheng XF (2002) The FKBP12-rapamycin-asso-
ciated protein (FRAP) is a CLIP-170 kinase. EMBO
Rep 3, 988–994.
45 Bilbe G, Delabie J, Bruggen J, Richener H, Asselbergs
FA, Cerletti N, Sorg C, Odink K, Tarcsay L, Wiesen-
danger W et al. (1992) Restin: a novel intermediate
filament-associated protein highly expressed in the
Reed–Sternberg cells of Hodgkin’s disease. EMBO J
11, 2103–2113.
46 Delabie J, Shipman R, Bruggen J, De Strooper B, van
Leuven F, Tarcsay L, Cerletti N, Odink K, Diehl V,
Bilbe G et al. (1992) Expression of the novel intermedi-
ate filament-associated protein restin in Hodgkin’s dis-
C. de Chiara et al. Study of the interactions of the Anp32a LRR module
FEBS Journal 275 (2008) 2548–2560 ª 2008 The Authors Journal compilation ª 2008 FEBS 2559
ease and anaplastic large-cell lymphoma. Blood 80,
2891–2896.
47 Coquelle FM, Caspi M, Cordelieres FP, Dompierre JP,
Dujardin DL, Koifman C, Martin P, Hoogenraad CC,
Akhmanova A, Galjart N et al. (2002) LIS1, CLIP-
170’s key to the dynein ⁄ dynactin pathway. Mol Cell
Biol 22, 3089–3102.
48 Cornilescu G, Delaglio F & Bax A (1999) Protein back-
bone angle restraints from searching a database for
chemical shift and sequence homology. J Biomol NMR

NMR: programs for checking the quality of protein
structures solved by NMR. J Biomol NMR 8, 477–486.
57 Vriend G (1990) WHAT IF: a molecular modelling and
drug design program. J Mol Graph 8, 52–56.
58 Berendsen HJC, van der Spoel CD & van Drunen R
(1995) GROMACS: a message-passing parallel molecu-
lar dynamics implementation. Comp Phys Commun 91,
43–56.
59 Van Gunsteren WF, Billeter S, Eising A, Hunenberger
PH, Kruger P, Mark AE, Scott WRP & Tironi IG
(1996) Biomolecular Simulations: the GROMOS96 Man-
ual and User Guide. VdF: Hoshshulverlag AG an der
ETH Zurich and BIOMOS b.v, Zurich, Gronigen.
60 Menon RP, Gibson TJ & Pastore A (2004) The C ter-
minus of fragile X mental retardation protein interacts
with the multi-domain Ran-binding protein in the
microtubule-organising centre. J Mol Biol 343, 43–53.
61 Menon RP & Pastore A (2006) Expansion of amino
acid homo-sequences in proteins: insights into the role
of amino acid homo-polymers and of the protein con-
text in aggregation. Cell Mol Life Sci 63, 1677–1685.
62 Thomas DJ & Pastore A (2004) WHEATSHEAF – an
algorithm to average protein structure ensembles. Acta
Crystallogr
61, 112–116.
Study of the interactions of the Anp32a LRR module C. de Chiara et al.
2560 FEBS Journal 275 (2008) 2548–2560 ª 2008 The Authors Journal compilation ª 2008 FEBS