Báo cáo Y học: Importin a binds to an unusual bipartite nuclear localization signal in the heterogeneous ribonucleoprotein type I pptx

Importin a binds to an unusual bipartite nuclear localization signal
in the heterogeneous ribonucleoprotein type I
Maria G. Romanelli and Carlo Morandi
Department of Mother and Child, Biology and Genetics, University of Verona, Italy
The heterogeneous nuclear ribonucleoprotein (hnRNP)
type I, a modulator of alternative splicing, localizes in the
nucleoplasm of mammalian cells and in a discrete peri-
nucleolar structure. HnRNP I contains a novel type of
bipartite nuclear localization signal (NLS) at the N-terminus
of the protein that we have previously named nuclear
determinant localization type I (NLD-I). Recently, a neural
counterpart of hnRNP I has been identified that contains a
putative NLS with two strings of basic amino acids separated
by a spacer of 30 residues. In the present study we show that
the neural hnRNP I NLS is necessary and sufficient for
nuclear localization and represents a variant of the novel
bipartite NLS present in the NLD-I domain. Furthermore,
we demonstrate that the NLD-I is transported into the nuc-
leus by cytoplasmic factor(s) with active transport modality.
Binding assays using recombinant importin a show an inter-
action with NLD-I similar to that of SV40 large T antigen
NLS. Deletion analysis indicates that both stretches of basic
residues are necessary for binding to importin a. The above
experimental results lead to the conclusion that importin a
acts as cytoplasmic receptor for proteins characterized by a
bipartite NLS signal that extends up to 37 residues.
Keywords: heterogeneous ribonucleoprotein-I; polypyrimi-
dine tract-binding protein; PTB; nuclear localization signal;
importin a.
Transport of proteins and RNA into and out of the nucleus
occurs through nuclear pore complexes (NPCs), which are

plugged through the double membrane of the nuclear
envelope [1,2]. Small molecules and ions may pass the NPC
passively, while macromolecules larger than 40–45 kDa are
actively transported through the NPC. The active nuclear
import and export of proteins is mediated by specific amino-
acid sequences that are referred as nuclear localization
signals (NLSs) [3,4] and nuclear export signal (NESs) [5]. At
least two different types of ÔclassicalÕ NLSs have been
defined: a short stretch of basic amino acids, exemplified by
the SV40 large T antigen NLS (T-ag P
KKKRKV) [6] and a
bipartite NLS composed of two stretches of basic amino
acids separated by a spacer of 10–12 amino acids, exempli-
fied by nucleoplasmin (
KRPAATKKAGQAKKKK). The
two sets of basic residues of bipartite-type NLS are required
for sufficient nuclear localization, while the spacer is
mutant-tolerant in sequence [7]. NLSs are usually recog-
nized by the heterodimeric import receptor complex com-
prising importins a and b, also named karyopherins [8,9].
Importins a contain the NLS-binding site and importins b
are responsible for the docking of the importin–substrate
complex to the cytoplasmic side of the NPC and its
subsequent translocation through the pore. Transfer
through the pore of importin–NLS protein complex
requires two additional soluble proteins, RanGTPase and
nuclear transport factor-2 (NTF2) [1]. Once inside the
nucleus, Ran-GTP binding to importin b causes the disso-
ciation of the import complex and release of the cargo
[10,11]. The directionality of the nuclear import is conferred

by an asymmetric distribution of the GTP and GDP-bound
forms of Ran between the cytoplasm and the nucleus, with
the GTP-form predominant in the nucleus [12,13]. Based on
the similarities of their primary structures, the importins a
have been separated into three subfamilies, each of which
shows distinct substrate specificity and differential expres-
sion [14–16]. Importins a consist of two structural and
functional domains, a short basic N-terminal importin b
binding (IBB) domain, and a large NLS-binding domain
comprising armadillo (Arm) repeats [17–19]. Crystal struc-
tures of karyopherins a complexes with NLS peptide have
revealed the determinants of specificity for the binding of
NLS sequences [20–22].
A number of NLS sequences that do not conform to the
classical NLS consensus motif have also been identified,
such as the M9 sequence, present in the hnRNP A1, which is
recognized by transportin (karyopherin-b2), rich in glycine
rather than basic residues [23–25]. A unique signal, called
KNS, which allows nuclear transport via a mechanism
independent of soluble factors has also been described in the
hnRNP type K [26]. These findings indicate that the most
important level of control for nuclear protein transport is
the targeting sequence.
We have previously identified a novel bipartite NLS in
the hnRNP type I protein [27], also known as polypyri-
midine tract binding protein (PTB) [28,29], a member of the
large set of RNA binding proteins, known as hnRNPs, that
Correspondence to M. G. Romanelli, Department of Mother and
Child, Biology and Genetics, University of Verona,
Strada le Grazie 8, 37134 Verona, Italy.

Fax: + 39 045 8027180, Tel.: + 39 045 8027182,
E-mail:
Abbreviations: NLS, nuclear localization signal; NLD-I, nuclear
localization determinant type I; T-ag, SV40 large tumor antigen;
hnRNP, heterogeneous nuclear ribonucleoprotein; FITC,
fluorescein isothiocyanate; NPC, nuclear pore complex; GST,
glutathione S-transferase; PTB, polypyrimidine tract-binding
protein; GFP, green fluorescent protein.
(Received 10 January 2002, revised 10 April 2002,
accepted 18 April 2002)
Eur. J. Biochem. 269, 2727–2734 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02942.x
have been implicated in mRNA maturation and transport
[30,31]. In mammalian cells, hnRNPI/PTB functions as a
splicing repressor [32,33], and mediates exon skipping of
several genes, such as a-andb-tropomyosin premRNAs
[34,35], neuron-specific exon in the c-src, c-aminobutyric
acid A c2 receptor, clathrin light chain B, N-methyl-
D
-
aspartate premRNA [36,37]. Interestingly, a PTB homo-
logue, abundantly present in the brain and in some neural
cell lines, known as neural PTB (nPTB), has been recently
identified [38]. Such a protein interacts with neuron-specific
RNA binding proteins that participate in the control of
alternative splicing in neurons [39]. nPTB is 74% identical in
amino-acid sequence to PTB and contains four unusual
RNA recognition motif (RRM) domains, and a putative
bipartite NLS near the N-terminus.
HnRNP I localizes in HeLa cells both in the nucleoplasm
and in a discrete perinucleolar structure [29,41]. We

previously reported that the N-terminal sequence of hnRNP
I/PTB contains a 60-amino-acid sequence that is both
necessary and sufficient to target the protein to the nucleus
[27]. The sequence, named nuclear localization determinant
type I (NLD-I), is characterized by a NLS containing two
clusters of basic amino acids (KR and KKFK) that
resemble the nucleoplasmin bipartite signal, but are separ-
ated by an unusually long stretch of 30 amino acids.
Compared to that of hnRNP I, the nPTB bipartite NLS
conserves the basic stretches whereas 14 residues differ in the
30-amino-acid spacer sequence.
In the present work, we have characterized the NLD
domain present in the neural counterpart of hnRNPI/PTB
and investigated whether the hnRNPI/PTB import involves
the karyopherin a/b pathway. We show that the hnRNP I
NLD-I domain displays significant binding to importin a,
which is diminished or eliminated by mutations in both
basic stretches. The present data support the model for the
recognition of bipartite NLS derived by crystallographic
analysis.
EXPERIMENTAL PROCEDURES
Plasmid construction
A plasmid containing the full length cDNA of human nPTB
[39] was kindly provided by D. L. Black (Howard Hugnes
Medical Institute, Los Angeles, CA, USA). To generate
fusion constructs with the green fluorescent protein (GFP),
the nPTB entire coding region, or its fragments (Fig. 2),
were obtained by PCR from the original cDNA. The
forward and reverse primers used for PCR included SalI
and BamHI restriction enzyme site at the 5¢ and 3¢ ends,

respectively. The PCR products were digested with SalIand
BamHIandclonedinframewitha5¢-GFP coding sequence
into mammalian expression vector pEGFP-C1 (Clontech).
To introduce deletions into the NLD-I sequence of nPTB
we used a two-step PCR method. In the first step we used
the following oligonucleotides: forward primer D11–13
(5¢-ATGGACGGAATCGTCACTGAAGTTGCAGTTA
GAGGATCTGACGAACTACTCTCAGGC-3¢)and
reverse primer R1 (5¢-ATTGGATCCTTATACACGAGA
AGGAGCACC-3¢) to generate the pnPTB-NLD-I D11-13
mutant; forward primer F1 (5¢-GGCAGGCATTCAGTC
GACATGGACGGAATCGTCACT-3¢) and reverse pri-
mer D45-47 (5¢-TACACGAGAAGGAGCACCATCCA
TTTTATCTTCTCCTTTACTATCATTACCATTGGCT
GT-3¢) to generate the pnPTB-NLD-I D45–47; forward
primer D11–13 and reverse primer D45–47 to generate
the pnPTB-NLD-I D11–13; 45–47 mutant (the numbers
indicate the residues deleted within nPTB amino-acid
sequence). In the second PCR step, all the fragments were
amplified with primers F1 and R2 that introduced the SalI
and BamHI sites, respectively, and cloned in the pEGFP-C1
vector.
The glutathione S-transferase (GST) fusion system was
used to generate chimeric proteins [42]. HnRNP I-NLD I,
and mutants D11–13, D45–47, and D11–13; 45–47 frag-
ments, previously cloned in a pA1-CAT vector [27,43], were
amplified by PCR with oligonucleotides that introduced
EcoRIandHindIII restriction sites, and ligated into a
modified pGEX-5X-1 vector (Amersham Pharmacia Bio-
tech) where the XhoI had been mutated in a HindIII site.

Cell culture and transfection
Adherent HeLa cells were maintained in exponential growth
in Dulbecco’s modified Eagle’s medium supplemented with
10% fetal bovine serum. Cells were grown on glass
coverslips and transfected using Lipofectamine (Life Tech-
nologies) according to manufacturer’s instructions.
Forty-eight hours after transfection, cells were washed
with NaCl/P
i
and fixed in 2% paraformaldehyde in NaCl/P
i
at room temperature for 20 min. Cellular localization of the
nPTB–GFP fusion proteins was examined under a fluores-
cent microscope.
Expression of fusion proteins in
Escherichia coli
The pGEX-NLD-I, pGEX-NLD-I D11–13, pGEX-NLD-I
D45–47 and pGEX-NLD-I D11–13; 45–47 constructs were
transformed into the Escherichiacoli strain BL 21(DE3). Cell
culture and batch purification of the GST fusion proteins
were performed essentially according to the manufacturer’s
instructions (Amersham Pharmacia Biotech). All the proce-
dures were carried out at 4 °C; 1 m
M
EGTA and 2 m
M
dithiothreitol were included in the buffers throughout the
purification procedures. Recombinant GST proteins, after
elution, were dialyzed at 4 °C against binding buffer (20 m
M

Hepes, 150 m
M
KOAc, 2 m
M
Mg (OAc)
2
,2m
M
dithiothre-
itol). The concentration of proteins was determined by the
method of Bradford [44] using the Bio-Rad dye reagent (Bio-
Rad) and BSA as standard. Protein samples were aliquoted,
quick frozen in liquid N
2
andstoredat)80 °C.
The E. coli strain BLR containing GST fusions of a
functional SV40 large T antigen nuclear localization signal
(Tag NLS) or an inverse version of Tag NLS (Tag NLSinv)
were cultured, and the proteins were purified as described
previously [45].
Nuclear import assay
Digitonin permeabilized HeLa cells were prepared essen-
tially as described by Adam et al. [46]. Cells grown on
coverslips were permeabilized with 55 lgÆmL
)1
digitonin
(Sigma) in transport buffer (TB: 20 m
M
Hepes pH 7.3,
110 m

M
potassium acetate, 5 m
M
sodium acetate, 2 m
M
magnesium acetate, 1 m
M
EGTA, 2 m
M
dithiothreitol,
1 lgÆmL
)1
each of aprotinin, leupeptin, and pepstatin A).
2728 M. G. Romanelli and C. Morandi (Eur. J. Biochem. 269) Ó FEBS 2002
A standard 50-lg nuclear import assay was performed
in transport buffer containing an energy-regenerating
system (1 m
M
ATP, 0,5 m
M
GTP, 10 m
M
creatine phos-
phate, and 0.4 UÆmL
)1
creatine phosphokinase), 5 lgof
GST fused proteins, and 30 lg of rabbit reticulocyte
lysate (Promega) as a cytosol source. The reaction was
allowed to proceed for 45 min at 30 °C. Where indicated,
wheat germ agglutinin (WGA); Sigma) at 50 lgÆmL

)1
,or
hexokinase at 100 UÆmL
)1
and glucose at 10 m
M
,were
included. Import assays were terminated by washing the
cellsincoldNaCl/P
i
followed by fixation in 2%
paraformaldehyde for 30 min.
Immunofluorescence staining
Fixed cells were permeabilized for 3 min in )20 °C acetone,
washed with NaCl/P
i
and incubated for 40 min with
primary antibody diluted in 3% BSA/NaCl/P
i
(1 : 100
monoclonal anti-GST Ig, Santa Cruz Biotechnology),
washed in NaCl/P
i
and incubated for 40 min in 1 : 50
FITC-conjugated goat anti-(mouse IgG) Ig. Cells were
finally washed with NaCl/P
i
, coated with 90% glycerol in
NaCl/P
i

, and observed under a Leitz Orthoplan microscope
with an epifluorescence attachment.
In vitro binding assay
The plasmid pRSET-hSRP1 containing a cDNA of human
importin a [47] was used to produce a [
35
S]methionine-
labeled protein using the Promega TNT T7 Quick Coupled
Translation System. Fifteen micrograms of GST–NLD-I or
NLD-I mutants and 7 lg of GST–Tag NLS or GST–
TagNLSinv were incubated with 40 lL of glutathione-
agarose beads (Amersham Pharmacia Biotech) in 0.5 mL of
binding buffer (20 m
M
Hepes, pH 6.8, 150 m
M
KOAc,
2m
M
Mg (OAc)
2
,2m
M
dithiothreitol, 0.1% Tween 20 for
2hat4°C. The beads were collected and washed three times
with binding buffer. After washing, one-twentieth of the
beads were removed and the amount of immobilized GST
fusion proteins were analyzed by SDS/PAGE. The rest of
the beads were incubated with 90 lLofin vitro translated
importin reaction mixture for 4 h at 4 °C. The beads were

then washed six times in binding buffer, boiled in 30 lLof
sample buffer, and the immobilized proteins were resolved
on a SDS/10% polyacrylamide gel. The
35
S-labeled importin
bound to the GST fusion proteins was detected by fluoro-
graphy using Amplify Reagent (Amersham Pharmacia
Biotech).
RESULTS
The NLD-I of neuronal PTB is capable of targeting
a heterologous protein to the nucleus
In a previous study, we identified the sequence implicated
in the nuclear transport of the human hnRNPI/PTB at
the N-terminal of the protein [27]. This region, which we
called NLD-I, extends in the first 60 amino acids,
upstream to the RRM1 (Figs 1A). The sequence contains
two short basic sequences (KR and KKFK) that resemble
the SV40 T-ag NLS, separated by an unusual long spacer
sequence of 30 amino acids. Both short basic sequence
were necessary for transport and, taken alone, were not
able to target the protein to the nucleus. Sequence
comparisons among NLD-I and the N-terminal region
of neural PTB and PTB homologues isolated from pig,
rat, mouse and Xenopus show that the basic stretches are
highly conserved in all the hnRNPI/PTB homologues,
whereas the sequence of the spacer, that may vary from
29 to 33 amino acids, is less conserved (Fig. 1B). The
nPTB NLD-I sequence is identical in human and mouse,
whereas the variations in NLD-I sequence between human
nPTB and hnRNPI/PTB far exceed that among PTBs

from different vertebrates.
In order to characterize the novel type of bipartite NLS
we first examined the ability of the NLD-I motif of the
neuronal PTB to target a heterologous protein to the
nucleus. GFP–nPTB fusion proteins were constructed in
which NLD-I motif was deleted from nPTB or was the only
sequence fused to GFP. Following transfection with GFP
constructs, HeLa cells were fixed and visualized by direct
fluorescence (Fig. 2). GFP is a small protein (30 kDa) that
can passively diffuse through the nuclear pores and does not
produce a subcellular localization bias (Fig. 2A), whereas
fusion of the entire nPTB ORF to GFP led to a peptide that
accumulates exclusively in the nucleus (Fig. 2B). The first 60
amino acids corresponding to the NLD-I domain are
necessary and sufficient to localize the fusion protein
completely in the nucleus, as demonstrated by the pGFP-
nPTBD60 construct that was confined to the cytoplasm,
Fig. 1. Functional domains of the human hnRNPI/PTB protein.
4
(A) Diagrammatic representation of the human hnRNP I/PTB protein, with the
four RNA recognition motifs (RRM1–4) and the nuclear localization determinant type I (NLD-I). (B) Amino acid alignment of the hnRNP-I
NLD-I domain with homologous domains from human and mouse nPTB, and from mouse, rat, pig and Xenopus PTB. The basic clusters are in
bold letters. Alignment was performed by
CLUSTAL W
program. Asterisks and dots show identical and similar amino acids, respectively.
Ó FEBS 2002 Binding of importin a to hnRNP I bipartite NLS (Eur. J. Biochem. 269) 2729
whereas when the first 60 amino acids are fused to GFP
(pGFP–nPTB60), the protein localized completely into the
nucleus (Fig. 2D,C, respectively).
Deletions of three amino acids, including a lysine in the

first basic stretch (deletion Gly-Val-Lys in pGFP–
nPTBD11–13) abolished the nuclear localization and left
the fusion protein to diffuse passively through the nuclear
pore (Fig. 2E). A similar cellular distribution was observed
when a serine and two lysines were deleted in the second
basic stretch (pGFP–nPTBD45–47) or when both type of
deletions were present in the basic stretches (construct
pGFP–nPTBD11–13; 45–47) (Fig. 2F,G). The above obser-
vations indicated that the two basic motifs in NLD-I of
nPTB are interdependently required for full nuclear local-
ization of the chimeric protein. Similar results were previ-
ously obtained with the NLD-I present in hnRNP I/PTB
that differ from that of nPTB essentially in the spacer
sequence (Fig. 1). These data show that the intervening
sequence does not contribute to the nuclear localization and
may be modified without effect on localization, as it has
been shown for the bipartite NLS of nucleoplasmin [7].
Taken together, these results indicate that the identified
NLS motif is a bona fide nuclear import signal for PTB like
proteins. The NLS contained in the NLD-I is a new variant
of bipartite NLS sequences.
Nuclear import of GST–NLD-I is an energy-dependent
process that requires soluble cytoplasmic factors
and is inhibited by WGA
Due to the fact that NLD-containing NLSs motifs
3
can be
imported by different mechanisms [48], we have undertaken
a study on the identification of the receptor pathway that
mediates nuclear import of the PTB NLS. An in vitro

nuclear transport assay was used in which the plasma
membrane of HeLa cells was permeabilized with the weak
nonionic detergent digitonin that leaves the nuclear envel-
ope intact [46]. NLD-I motif was fused to a GST protein,
that itself does not accumulate into the nucleus [49]
(Fig. 3A). Nuclear transport of GST–NLD-I was examined
in the presence of a transport buffer containing rabbit
reticulocyte lysate as a source of cytosolic proteins and an
ATP-regenerating system, to provide energy for transloca-
tion. The subcellular distribution of the GST–NLD-I was
determined by indirect immunofluorescence microscopy
using an antibody against GST. In such experiments, we
Fig. 2. The NLD-I domain of nPTB directs the
nuclear import of a heterologous protein.
(A) Schematic representation of the nPTB
regions used to produce GFP fusion proteins.
Structural domains are diagrammed with
shaded boxes representing the RRMs. Black
boxes or black boxes interrupted by a white
strip represent the basic stretches or the dele-
ted sequences in the basic stretches, respect-
ively, at the N-terminus of the protein.
(B) Plasmids expressing the native GFP (a),
GFP fused to the entire nPTB (b), the 60-
amino-acid region at N-terminus of nPTB (c),
or nPTB deleted of the first 60 amino acids (d),
were transiently transfected, expressed in
HeLa cells and visualized by fluorescent
microscopy. Likewise, GFP fused to the
60-amino acid region containing deletions at

amino acids 11–13 (e), or 44–47 (f), or both
type deletions (g) were also expressed in HeLa
cells.
2730 M. G. Romanelli and C. Morandi (Eur. J. Biochem. 269) Ó FEBS 2002
found that GST–NLD-I was clearly visible within the nuclei
in cells were the plasma membrane was permeabilized by
digitonin, followed by incubation at 30 °Cfor45minwith
rabbit reticulocyte lysate and an ATP energy-regenerating
system (Fig. 3C). A similar distribution was observed for
GST–TagNLS, which is transported to the nucleus by the
conventional NLS-mediated nuclear protein import path-
way utilizing Ran and the importin ab/heterodimer [45]
(Fig. 3B). This accumulation is ATP dependent as no
nuclear accumulation was observed when the permeabilized
cells were incubated with hexokinase and glucose to deplete
residual ATP, and when import assay was performed in the
absence of an ATP-regenerating system (Fig. 3D). As
expected, the transport reaction was also temperature-
dependent; if the assay was carried out at 4 °C, no transport
was observed (Fig. 3E), confirming the necessity of ATP
hydrolysis for nuclear protein import. When permeabilized
cells were preincubated in transport buffer containing
50 lgÆmL
)1
WGA, a lectin that associates with glycosylated
nucleoproteins and inhibits nuclear pore complex function,
the nuclear accumulation of NLD-I was inhibited (Fig. 3F).
To characterize the amino-acid sequence requirement for
nuclear import, we examined whether amino-acid deletions
into NLD-I would perturb import. When we used GST–

NLD-ID11–13, GST–NLD-ID45–47, or GST–NLD-ID11–
13;45–47 as the cargo in different import assays, no
transport in the nucleus was seen (Fig. 3G,H,J).
NLD-I domain binds to importin a
The previous data suggest that the import of the PTB NLS
might be mediated by cytosol receptor like the importin a/b
complex. The crystal structure of importin a reveals two
NLS peptide binding pockets and the distance between the
two binding sites allows a 10-residue spacer to link the two
peptide segments [20,50]. Importin a binding to a bipartite
signal with a spacer sequence longer than 30 amino acids
have not been tested thus far. The unusual sequence of the
PTB NLS raised the possibility that it might not be
recognized directly by importin a.
TotestifNLD-Iwouldbindtoimportina, we first
performed a binding assay using in vitro translated importin
a and recombinant GST–NLD-I fusion protein. The GST
fusions with SV40 tag NLS (GST–NLS) and an inverse
version of Tag NLS (GST–NLSinv) served as positive and
negative controls, respectively, for the importin a binding.
As shown in Fig. 4A, importin a is able to bind NLD-I
(lane GST–NLD-I). To ascertain whether the amino acid
that impaired nuclear transport in NLD-I deletion mutants,
mediate the importin a binding, we tested importin a
binding to NLD-I D11–13, NLD-ID45–47, or NLD-ID11–
13;45–47, expressed as GFP-fusions. All mutants showed no
detectable or very weak binding to importin a (Fig. 4B).
These results indicate that importin a binds to NLD–I by
interaction to the residues of the two short basic stretches,
when both basic stretches are present in the NLS sequence.

DISCUSSION
In this report, we have demonstrated that nuclear translo-
cation of nPTB and hnRNPI/PTB occurs via a polybasic
NLS sequence present in the N-terminus NLD-I motif. This
sequence functions in the nuclear import of PTB-like
proteins via an active energy-dependent process and binds
to the importin a. PTB NLS shares common features with
the known bipartite type NLSs; in fact, it contains two
clusters of basic amino acids, a smaller one of two basic
amino acids (KR) and a larger one with three basic amino
acids in a group of five (KKFKG); both are essential for full
nuclear localization and importin a binding. However, it
Fig. 3. Nuclear import of NLD-I is an energy-
dependent process inhibited by WGA. Digito-
nin-permeabilized HeLa cells were incubated
with GST, GST–TagNLS (a control for the
conventional importin a/b mediated nuclear
import pathway directed by SV40 NLS),
GST–NLD-I, or mutated NLD-I and visual-
ized by indirect immunofluorescence, using a
GST monoclonal antibody. In vitro nuclear
transport (see ÔMaterial and methodsÕ)was
carried out in the presence of cytosol and an
ATP-regenerating system (A–J). The effects
on nuclear transport of the ATP-regenerating
system omission (D), or 4 °C incubation (E),
were examined. Transport studies were also
carried out after preincubation with WGA
(F). Images are representative of at least three
independent experiments.

Ó FEBS 2002 Binding of importin a to hnRNP I bipartite NLS (Eur. J. Biochem. 269) 2731
represents a novel member of bipartite signal because there
are differences between the critical basic residues of NLS
and the consensus sequences of other bipartite signals, and
because the spacer between the basic motifs is unusually
longer (30 amino acids) than that of other well characterized
bipartite NLSs (Table 1). Searching SWISSPROT and
standard databases by PROSITE, and analyzing data
deposited at the PredictNLS server (c.
columbia.edu/predictNLS/) [56], we find that functional
bipartite NLSs are characterized by an intervening region of
different lengths, but not longer than 18 amino acids, like
the human androgen receptor NLS motif [57]. In transfected
cells, it has been shown that the segment spacer of
nucleoplasmin bipartite NLS can be replaced by a sequence
up to 20 alanine residues, without disrupting efficient
nuclear import [8]. Recently, a bipartite NLS, containing a
spacer between the basic motifs of 32 amino acids, has been
functionally characterized in hypoxia inducible factors 1a
[58]. We indicate a consensus sequence for bipartite long
type NLS as KRx(30–32)K[K/R]xK, according to Cokol
et al. [56]. Searching the PredictNLS database using this
motif, we found 61 proteins, with a true nuclear protein
percent of 83.6.
Nuclear import of most proteins requires both impor-
tins a and b, with importin a as the adapter between
importin b and the cargo protein. Some proteins undergo
nuclear import via direct binding to importin b without
involvement of importin a. Our results suggest that nuclear
import of PTB-like proteins, and probably of all the

proteins that contain the NLD-I type NLS, is an energy-
dependent process mediated by importin a. Deletions in
the basic regions within the N-terminal domain of
hnRNP I and nPTB inhibit nuclear import and/or accu-
mulation and reduces importin a binding. This conclusion
fits with the in vitro binding studies, which showed that the
NLD-I domain binds strongly to importin a and the bind-
ing is diminished or eliminated even if only one of the two
basic stretches are mutated. The affinity of the importin-
targeting sequence interaction is a critical parameter in
determining transport efficiency [5]. This is the first study
to demonstrate that importin a recognizes and binds an
NLS sequence that extends up to 37 residues. The crystal
structures of importin a [20–22,50] clearly reveal two
distinct binding sites that can accommodate both essential
elements of the bipartite NLS. The larger binding site is
structured optimally for the recognition of five lysine or
arginine residues, while the smaller binding site allows
specific recognition of two basic residues, and the interac-
tion is simultaneous at both sites. The distance between the
two binding sites allow a 10-residue spacer to link the two
peptide segments, while a shorter linker would impair the
simultaneous binding of the two clusters. The smaller basic
cluster is required to be upstream of the larger cluster. The
Table 1. Bipartite type nuclear localization signal. The single-letter amino acid code is used. The bold letters indicate the two arms of basic residues
of the bipartite NLS. aa, amino acids.
Protein Bipartite NLS Reference
Bipartite short type NLSs
Nucleoplasmin
KRPAATKKAGQAKKKKLDK [7]

CBP80
a
RRRHSDENDGGQPHKRRK [14]
N1N2
b
RKKRKTEEESPLKDKAKKSK [54]
SW15
c
KKYENVVIKRSPRKRGRPRK [55]
Human IL 5
KKYIDGQKKKCGEERRRVNQ [51]
Human RB
KRSAEGSNPPKPLKKLR [52]
Human p53
KRALPNNTSSSPQPKKKP [53]
Consensus
KK-10–12 aa -KKK
RR
RRR
Bipartite long type NLSs
HnRNPI/PTB_HUMAN
KRGSDELFSTCVTNGPFIMSSNSASAANGNDSKKFKGDS [29]
PTB_HUMAN
KRGSDELLSGSVLSSPNSNMSSMVVTANGNDSKKFKGED [39]
HIF1a
d
KRKMEHDGSLFQAUGIGTLLQQPDDHAATTSLSWKRVKG [58]
Consensus
KR-30–32 aa -K[K/R]XK
a

CAP-binding protein 80.
b
Xenopus laevis phosphoprotein.
c
S. cerevisiae transcription factor.
d
Hypoxia inducible factor 1a.
Fig. 4. Binding of importin a to wild-type or mutated NLD-I. (A) Wild-
type GST–NLD-I, or GST–NLS and GST–NLSinv, immobilized on
glutathione sepharose 4B, were incubated with 90 lLofin vitro
translated protein a labeled with [
35
S]methionine for 4 h at 4 °C. The
proteins were separated by SDS/10% PAGE gel, and bound importin
a was analyzed by fluorography (upper panel). An amount repre-
senting 1/20th of the beads incubated with GST fusion proteins,
extensively washed, was resolved by SDS/PAGE and stained with
Comassie-blue (lower panel, CB staining). (B) Binding of importin a
was also analyzed using immobilized GST–NLD-I peptides mutated
by deletions. The results are representative of two independent
experiments.
2732 M. G. Romanelli and C. Morandi (Eur. J. Biochem. 269) Ó FEBS 2002
spacing between a defined number of binding sites acts as
molecular ruler that sets further constraints on target
specificity. Binding properties of NLD-I type NLS to
importin a are consistent with crystallographic structure of
interaction with bipartite signal and further characterize
the bipartite signals, confirming that the linker sequence
does not contain a consensus, and may be as long as 30
amino acids.

Mammalian paralogs of importin a have recently been
discovered and six genes for importin a have been found in
human. The importin a that we have used in our binding
experiments (hSRP1a) represents one of the three sub-
families of importins a [15]. Several experiments clearly
support the hypothesis that importins a might be specialized
in their efficiency to transport different nuclear proteins [59].
It will be interesting to analyze the binding specificity of the
different types of importin a toNLD-INLSandtothe
putative NLD-I type NLSs present in other nuclear
proteins.
ACKNOWLEDGEMENTS
We wish to thank Pamela Lorenzi for excellent technical help. We
thank Michael F. Rexach for his generous gift of GST–NLS and GST–
NLSinv constructs and Karsten Weis for his generous gift of pRSET-
hSRP a plasmid. This work was supported by grants from Ministery of
Scientific Research and Technology (ex 60%).
REFERENCES
1. Gorlich, D. & Kutay, U. (1999) Transport between the cell nucleus
and the cytoplasm. Annu. Rev. Cell. Dev. Biol. 15, 607–660.
2. Gant, T.M., Goldberg, M.W. & Allen, T.D. (1998) Nuclear
envelope and nuclear pore assembly: analysis of assembly inter-
mediates by electron microscopy. Curr. Opin. Cell. Biol. 10, 409–
415.
3. Doye, V. & Hurt, E. (1997) From nucleoporins to nuclear pore
complexes. Curr. Opin. Cell Biol. 9, 401–411.
4. Dingwall, C. & Laskey, R.A. (1991) Nuclear targeting sequences –
a consensus? Trends Biochem. Sci. 16, 478–481.
5. Jans, D.A., Xiao, C.Y. & Lam, M.H. (2000) Nuclear targeting
signal recognition: a key control point in nuclear transport?

Bioessays 22, 532–544.
6. Nigg, E.A. (1997) Nucleocytoplasmic transport: signals, mecha-
nisms and regulation. Nature 386, 779–786.
7. Robbins, J., Dilworth, S.M., Laskey, R.A. & Dingwall, C. (1991)
Two interdependent basic domains in nucleoplasmin nuclear tar-
geting sequence: identification of a class of bipartite nuclear tar-
geting sequence. Cell 64, 615–623.
8. Makkerh, J.P.S., Dingwall, C. & Laskey, R.A. (1996) Compara-
tive mutagenesis of nuclear localization signals reveals the
importance of neutral and acidic amino acids. Curr. Biol. 6, 1025–
1027.
9. Weis, K. (1998) Importins and exportins: how to get in and out of
the nucleus. Trends Biochem. Sci. 23, 185–189.
10. Yoneda, Y. (2000) Nucleocytoplasmic protein traffic and its sig-
nificance to cell function. Genes Cells 5, 777–787.
11. Bischoff, F.R. & Ponstingl, H. (1995) Catalysis of guanine
nucleotide exchange of Ran by RCC1 and stimulation of hydro-
lysis of Ran-bound GTP by Ran-GAP1. Methods Enzymol. 257,
135–144.
12. Izaurralde, E., Kutay, U., von Kobbe, C., Mattaj, I.W. & Gorlich,
D. (1997) The asymmetric distribution of the constituents of the
Ran system is essential for transport into and out of the nucleus.
EMBO J. 16, 6535–6547.
13. Mattaj, I. & Englmeier, L. (1998) Nucleocytoplasmic transport:
the soluble phase. Annu. Rev. Biochem. 67, 265–306.
14. Miyamoto, Y., Imamoto, N., Sekimoto, T., Tachibana, T., Seki,
T., Tada, S., Enomoto, T. & Yoneda, Y. (1997) Differential modes
of nuclear localization signal (NLS) recognition by three distinct
classes of NLS receptors. J. Biol. Chem. 272, 26375–26381.
15. Kohler, M., Speck, C., Christiansen, M., Bischoff, F.R., Prehn, S.,

Haller,H.,Gorlich,D.&Hartmann,E.(1999)Evidencefordis-
tinct substrate specificities of importin alpha family members in
nuclear protein import. Mol. Cell. Biol. 19, 7782–7791.
16. Nadler,S.G.,Tritschler,D.,Haffar,O.K.,Blake,J.,Bruce,A.G.
& Cleaveland, J.S. (1997) Differential expression and sequence–
specific interaction of karyopherin alpha with nuclear localization
sequences. J. Biol. Chem. 272, 4310–4315.
17. Gorlich, D., Henklein, P., Laskey, R.A. & Hartmann, E. (1996) A
41 amino acid motif in importin-alpha confers binding to
importin-beta and hence transit into the nucleus. EMBO J. 15,
1810–1817.
18. Moroianu, J., Blobel, G. & Radu, A. (1996) The binding site
of karyopherin alpha for karyopherin beta overlaps with a
nuclear localization sequence. Proc. Natl Acad. Sci. USA 93,
6572–6576.
19. Peifer, M., Berg, S. & Reynolds, A.B. (1994) A repeating amino
acid motif shared by proteins with diverse cellular roles. Cell 76,
789–791.
20. Conti, E., Uy, M., Leighton, L., Blobel, G. & Kuriyan, J. (1998)
Crystallographic analysis of the recognition of a nuclear locali-
zation signal by the nuclear import factor karyopherin alpha. Cell
94, 193–204.
21. Conti, E. & Kuriyan, J. (2000) Crystallographic analysis of the
specific yet versatile recognition of distinct nuclear localization
signals by karyopherin alpha. Struct. Fold Des. 8, 329–338.
22. Fontes, M.R.M., Teh, T. & Kobe, B. (2000) Structural basis of
recognition of monopartite and bipartite nuclear localization
sequences by mammalian importin-alpha. J. Mol. Biol. 297, 1183–
1194.
23. Bogerd, H.P., Benson, R.E., Truant, R., Herold, A., Phingbodhi-

pakkiya, M. & Cullen, B.R. (1999) Definition of a consensus
transportin-specific nucleocytoplasmic transport signal. J. Biol.
Chem. 274, 9771–9777.
24. Bonifaci, N., Moroianu, J., Radu, A. & Blobel, G. (1997) Kar-
yopherin beta2 mediates nuclear import of a mRNA binding
protein. Proc. Natl Acad. Sci. USA 94, 5055–5060.
25. Siomi, H. & Dreyfuss, G. (1995) A nuclear localization domain in
the hnRNP A1 protein. J. Cell Biol. 129, 551–560.
26. Michael, W.M., Eder, P.S. & Dreyfuss, G. (1997) The K nuclear
shuttling domain: a novel signal for nuclear import and nuclear
export in the hnRNP K protein. EMBO J. 16, 3587–3598.
27. Romanelli, M.G., Weighardt, F., Biamonti, G., Riva, S. &
Morandi, C. (1997) Sequence determinants for hnRNP I protein
nuclear localization. Esp. Cell. Res. 235, 300–304.
28. Gil, A., Sharp, P.A., Jamison, S.F. & Garcia-Blanco, M.A. (1991)
Characterization of cDNAs encoding the polypyrimidine tract-
binding protein. Genes Dev. 5, 1224–1236.
29. Ghetti, A., Pinol-Roma, S., Michael, W.M., Morandi, C. &
Dreyfuss, G. (1992) hnRNP I, the polypyrimidine tract-binding
protein: distinct nuclear localization and association with
hnRNAs. Nucleic Acids Res. 20, 3671–3678.
30. Krecic, A.M. & Swanson, M.S. (1999) hnRNP complexes: com-
position, structure, and function. Curr. Opin. Cell. Biol. 11, 363–
371.
31. Weighardt, F., Biamonti, G. & Riva, S. (1996) The roles of het-
erogeneous nuclear ribonucleoproteins (hnRNP) in RNA meta-
bolism. Bioessays 18, 747–756.
32. Valcarcel, J. & Gebauer, F. (1997) Post-transcriptional regulation:
the dawn of PTB. Curr. Biol. 7, 705–708.
Ó FEBS 2002 Binding of importin a to hnRNP I bipartite NLS (Eur. J. Biochem. 269) 2733

33. Wagner, E.J. & Garcia-Blanco, M.A. (2001) Polypyrimidine tract
binding protein antagonizes exon definition. Mol. Cell Biol. 21,
3281–3288.
34. Mulligan, G.J., Guo, W., Wormsley, S. & Helfman, D.M. (1992)
Polypyrimidine tract binding protein interacts with sequences
involved in alternative splicing of beta-tropomyosin pre-mRNA.
J. Biol. Chem. 267, 25480–25487.
35. Gooding, C., Roberts, G.C. & Smith, C.W. (1998) Role of an
inhibitory pyrimidine element and polypyrimidine tract binding
protein in repression of a regulated alpha-tropomyosin exon. RNA
4, 85–100.
36. Chan, R.C. & Black, D.L. (1997) The polypyrimidine tract bind-
ing protein binds upstream of neural cell-specific c-src exon N1 to
repress the splicing of the intron downstream. Mol. Cell. Biol. 17,
4667–4676.
37. Zhang, L., Liu, W. & Grabowski, P.J. (1999) Coordinate repres-
sion of a trio of neuron-specific splicing events by the splicing
regulator PTB. RNA 5, 117–130.
38. Ashiya, M. & Grabowski, P.J. (1997) A neuron-specific splicing
switch mediated by an array of pre-mRNA repressor sites:
evidence of a regulatory role for the polypyrimidine tract
binding protein and a brain-specific PTB counterpart. RNA 3,
996–1015.
39. Markovtsov, V., Nikolic, J.M., Goldman, J.A., Turck, C.W.,
Chou, M.Y. & Black, D.L. (2000) Cooperative assembly of an
hnRNP complex induced by a tissue-specific homolog of poly-
pyrimidine tract binding protein. Mol. Cell. Biol. 20, 7463–7479.
40.Polydorides,A.D.,Okano,H.J.,Yang,Y.Y.L.,Stefani,G.&
Darnell, R.B. (2000) A brain-enriched polypyrimidine tract-
binding protein antagonizes the ability of Nova to regulate neu-

ron-specific alternative splicing. Proc.NatlAcad.Sci.USA97,
6350–6355.
41. Matera, A.G., Frey, M.R., Margelot, K. & Wolin, S.L. (1995) A
perinucleolar compartment contains several RNA polymerase III
transcripts as well as the polypyrimidine tract-binding protein,
hnRNP I. J. Cell Biol. 129, 1181–1193.
42. Smith, D.B. & Johnson, K.S. (1988) Single-step purification of
polypeptides expressed in Escherichia coli as fusions with glu-
tathione S-transferase. Gene 67, 31–40.
43. Weighardt, F., Biamonti, G. & Riva, S. (1995) Nucleo-cyto-
plasmic distribution of human hnRNP proteins: a search for the
targeting domains in hnRNP A1. J. Cell Sci. 108, 545–555.
44. Bradford, M.M. (1976) A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal. Biochem. 72, 248–254.
45. Rexach, M. & Blobel, G. (1995) Protein import into nuclei:
association and dissociation reactions involving transport sub-
strate, transport factors, and nucleoporins. Cell 83, 683–692.
46. Adam, S.A., Marr, R.S. & Gerace, L. (1990) Nuclear protein
import in permeabilized mammalian cells requires soluble cyto-
plasmic factors. J. Cell Biol. 111, 807–816.
47. Weis, K., Mattaj, I.W. & Lamond, A.I. (1995) Identification of
hSRP1 alpha as a functional receptor for nuclear localization
sequences. Science 268, 1049–1053.
48. Deane, R., Shafer, W., Zimmermann, H.P., Mueller, L., Gorlich,
D., Prehn, S., Ponstingl, H. & Bishoff, F.R. (1997) Ran-binding
protein 5 (RanBP5) is related to the nuclear transport factor
importin-beta but interacts differently with RanBP1. Mol. Cell.
Biol. 17, 5087–5096.
49. Pollard, V.W.,Michael.W.M., Nakienly, S.,Siomi, M.C.,Wang,F.

& Dreyfuss, G. (1996) A novel receptor-mediated nuclear protein
import pathway. Cell 86, 985–994.
50. Dingwall, C. & Laskey, R.A. (1998) Nuclear import: a tale of two
sites. Curr. Biol. 8, R922–R924.
51. Jans, D.A., Briggs, L.J., Gustin, S.E., Jans, P., Ford, S. & Young,
I.G. (1997) A functional bipartite nuclear localisation signal in the
cytokine interleukin-5. FEBS Lett. 406, 315–320.
52. Zacksenhaus, E., Bremner, R., Phillips, R.A. & Gallie, B.L. (1993)
A bipartite nuclear localization signal in the retinoblastoma gene
product and its importance for biological activity. Mol. Cell. Biol.
13, 4588–4599.
53. Liang, S. & Clarke, M.F. (1999) A bipartite nuclear localization
signal is required for p53 nuclear import regulated by a carboxyl-
terminal domain. J. Biol. Chem. 274, 32699–32703.
54. Hu, W. & Jans, D.A. (1999) Efficiency of importin alpha/beta-
mediated nuclear localization sequence recognition and nuclear
import. Differential role of NTF2. J. Biol. Chem. 274, 15820–
15827.
55. Jans, D.A., Moll, T., Nasmyth, K. & Jans, P. (1995) Cyclin-
dependent kinase site-regulated signal-dependent nuclear locali-
zation of the SW15 yeast transcription factor in mammalian cells.
J. Biol. Chem. 270, 17064–17067.
56. Cokol, M., Nair, R. & Rost, B. (2000) Finding nuclear localization
signals. EMBO Reports 1, 411–415.
57. Jenster, G., Trapman, J. & Brinkmann, A.O. (1993) Nuclear
import of the human androgen receptor. Biochem. J. 293,
761–768.
58. Luo, J.C. & Shibuya, M. (2001) A variant of nuclear localization
signal of bipartite-type is required for the nuclear translocation of
hypoxia inducible factors (1 alpha, 2 alpha and 3 alpha). Oncogene

20, 1435–1444.
59. Welch, K., Franke, J., Kohler, M. & Macara, I.A. (1999)
RanBP3 contains an unusual nuclear localization signal that is
imported preferentially by importin-alpha3. Mol. Cell Biol. 19,
8400–8411.
2734 M. G. Romanelli and C. Morandi (Eur. J. Biochem. 269) Ó FEBS 2002