184 Copyright © 1998 Elsevier Science Ltd. All rights reserved. 0962-8924/98/$19.00 trends in CELL BIOLOGY (Vol. 8) May 1998
PII: S0962-8924(98)01248-3
Our current understanding of the dynamics of nuclear
import has evolved from the initial identification of
the ‘classical nuclear localization signal’ (cNLS), which
consists of a short sequence rich in basic amino-acid
residues
1
. Transport of proteins bearing this signal
into the nucleus is mediated by a karyopherin –1
heterodimeric complex plus additional factors such
as the GTPase Ran and its associated protein p10
(NTF2) (reviewed in Refs 2 and 3). The nomenclature
for the –1 complex remains confusing, owing to its
almost coincident identification by several labora-
tories (see Table 1). Here, we use the general term
karyopherin [from the Greek terms karyon meaning
nucleus and pher(ein) meaning to bring to or to carry
from
4
] for it most accurately describes a family of re-
ceptors that are able to recognize substrates, in either
(or both) the nucleoplasm or cytoplasm, and assist
in their transport into and out of the nucleus. In the
prototypical case, the karyopherin subunit of the
–1 dimer binds to the cNLS, and the complex
then docks at the nuclear pore complex (NPC) in an
energy-independent manner through interactions
between karyopherin 1 and (at least) cytoplasmically
exposed NPC proteins (nucleoporins) containing
characteristic peptide repeats. Subsequent translo-
cation through the NPC and termination of import
require energy and the GTPase Ran (see the accom-
panying hypothesis article by Melchior and Gerace).
Ran appears to play a crucial role in maintaining
vectorial cargo transport and in regulating the bind-
ing and release steps that occur during translocation
through the NPC. Ran exists in the cell in two forms,
either as Ran–GTP or Ran–GDP. Because of the sub-
cellular localization of several proteins that alter the
nucleotide-bound state of Ran, it is likely that the
predominant form of Ran in the cytosol is Ran–GDP,
whereas in the nucleus the major form is Ran–GTP.
Such a gradient of Ran could establish directional
transport by creating separate environments in which
transport complexes are either assembled or dis-
assembled (reviewed in Refs 5 and 6). Consistent with
this idea, the karyopherin 1––cNLS complex is
stable in the presence of Ran–GDP (i.e. in the cytosol),
whereas Ran–GTP promotes its disassembly and the
release of substrate (i.e. in the nucleus). In addition to
these more global effects, Ran also acts as a molecular
switch at sites within the NPC. Here, it is thought
that, in concert with p10 (Ref. 7), conversion of the
nucleotide-bound state of Ran modulates interactions
between the NPC and 1, and the cNLS-containing
cargo in what has been proposed to be a series of
association and dissociation reactions that occur as
these components step their way through the NPC
8
.
While the karyopherin –1 heterodimer is likely
to account for the import of cNLS-containing cargo,
it has recently been established that an extended
family of karyopherin 1-related nuclear transport
factors exists. This article describes recent progress in
characterizing these proteins. Some members of this
family have been shown to be involved directly in
transporting specific classes of macromolecules into
and out of the nucleus, whereas others have been
implicated indirectly in nuclear transport.
Karyopherin 2 and mRNP proteins
Two very different, but complementary, approaches
led to the identification of the first alternative karyo-
pherin in both Saccharomyces cerevisiae (Kap104p
9
)
and mammalian cells (transportin
10
). Because of their
similarity to karyopherin 1, here these are generi-
cally termed karyopherin 2
11,12
.
The mammalian karyopherin 2 (transportin) was
identified by its interaction with a 38-amino-acid
residue sequence (termed M9) found within the het-
erogeneous nuclear ribonuclear protein (hnRNP) A1
(Ref. 10). A1 is a shuttling protein that exits the
nucleus bound to mRNA and re-enters the nucleus
after it is shed from the mRNA in the cytoplasm
13
.
The M9 sequence is sufficient to act as both an NLS
and a nuclear export signal (NES) on a reporter pro-
tein, and it is thought to be the signal respon-
sible for the shuttling activity of A1
14,15
. The nuclear
import leg of this journey was shown, in an in vitro
assay, to require karyopherin 2
10–12,16
. The import
is energy dependent and separate from the cNLS
pathway
10–12,16
. 2 also interacts with several other
hnRNPs, suggesting that they too are imported by
this pathway
17
.
Yeast karyopherin 2 was identified by its homol-
ogy to 1 (Ref. 9). It is a mainly cytosolic protein that
also binds directly to repeat-containing nucleoporins
and to a cytosolic pool of at least two major mRNA-
binding proteins – Nab2p and Hrp1p (Nab4p)
9
. Inter-
estingly, Hrp1p is the closest structural homologue
to hnRNP A1 in yeast
18
. Temperature-sensitive mu-
tants in 2 cause redistribution of Nab2p from its
REVIEWS
Karyopherins and
kissing cousins
Richard W. Wozniak, Michael P. Rout
and John D. Aitchison
In eukaryotic cells, a regulated flux of molecules between the
cytoplasm and the nucleus maintains two very different
environments while allowing the controlled exchange of
macromolecules necessary for their individual functions.
Molecules entering or leaving the nucleus use nuclear localization
signals or nuclear export signals to pass through selective channels
in the nuclear envelope formed by nuclear pore complexes. The
recognition of signal-bearing cargo, its interaction with the nuclear
pore complex and its translocation through the pore complex are
mediated by soluble transport factors. Recently, the list of
potential transport factors has grown rapidly, suggesting a
previously unanticipated level of complexity for nuclear transport.
Richard Wozniak
and John Aitchison
are in the Dept of
Cell Biology and
Anatomy,
University of
Alberta, Edmonton,
Alberta, Canada
T6G 2H7; and
Michael Rout is at
The Rockefeller
University,
New York,
NY 10021, USA.
E-mails:
rwozniak@anat.
med.ualberta.ca
rout@rockvax.
rockefeller.edu
john.aitchison@
ualberta.ca