(Received for publication, November 6, 1996, and in revised form, December 5, 1996)
From the Bristol-Myers Squibb Pharmaceutical Research Institute, Seattle, Washington 98121
The process of nuclear protein transport requires
the interaction of several different proteins, either directly or
indirectly with nuclear localization or targeting sequences (NLS).
Recently, a number of karyopherins , or NLS-binding proteins, have
been identified. We have found that the karyopherins hSRP1 and hSRP1
are differentially expressed in various leukocyte cell lines and could
be induced in normal human peripheral blood lymphocytes. We show that
the two karyopherins bind with varied specificities in a sequence
specific manner to different NLSs and that the sequence specificity is
modulated by other cytosolic proteins. There was a correlation between
binding of karyopherins
to different NLSs and their ability to be
imported into the nucleus. Taken together, these data provide evidence
for multiple levels of control of the nuclear import process.
Active nuclear transport of proteins with molecular weights
greater than 40-60 kDa requires at least four different proteins, which act in a sequential manner with karyophilic proteins containing nuclear localization targeting sequences
(NLS)1 (1-4). There appear to be several
discrete steps in the import process which involves: 1) binding of the
NLS-binding protein, karyopherin , to an NLS; 2) interaction of this
complex with karyopherin
; 3) targeting to nuclear pore proteins;
and 4) the ATP/GTP-dependent translocation through the
nuclear pore mediated by ran (1, 5, 6).
Recently, the proteins involved in NLS binding and transport have been
identified. Those proteins that interact directly with the NLS have
been termed karyopherins (7-11). The Xenopus protein importin 60 was the first karyopherin
to be cloned, sequenced, and
shown to be involved in nuclear protein import (7). Subsequently, a
number of other karyopherins
have been identified, which suggests that there is a family of these NLS-binding proteins. The two major
groups of karyopherins
include 1) the yeast protein SRP1 (12) and
the human proteins hSRP1 and NPI-1 (8, 9), and 2) importin 60 (7) and
the human proteins hSRP1
(11) and Rch1 (10). In this report we have
termed hSRP1 and hSRP1
, K1 and K2, respectively. Each of these
karyopherins
are capable of binding to NLSs and facilitating
nuclear import. Recently it was shown that there was tissue-specific
expression of the mouse K1 (mSRP1) and K2 (mPendulin). The levels of K1
RNA appear higher in the brain and cerebellum, whereas K2 RNA was found
mostly in the thymus and spleen (13).
Similar to karyopherin , there are several homologs of karyopherin
, (also called importin 90 or p97) (11, 14). The function of
karyopherin
appears to be the targeting of the
karyophile-karyopherin
complex to the nuclear pore (11, 16). The
interaction of karyopherin
with karyopherin
has been shown to
enhance the latter protein's affinity for the NLS containing protein
(1). Although the protein factors described above are sufficient to support nuclear protein transport, there are accessory factors which
are also important for regulating nuclear transport. These factors
include p10 (17-20), Rna1p (20), and the heat shock protein Hsc73
(21).
Perhaps most important for nuclear protein transport is the
targeting sequence or NLS. Although there is a consensus for other organellar targeting sequences, there is little amino acid sequence homology among the large number of NLSs that have been identified (22).
The most conserved feature of either the "simple" (5-7 amino acid
sequences) or "bipartite" (two sets of positively charged amino
acids separated by 10-11 amino acids) is the presence of two basic
amino acids which constitutes the core of the NLS. Two mechanisms for
regulating the activity of the NLS include protein phosphorylation and
masking of the NLS to prevent its recognition by karyopherin (23,
24).
In this report we show that there are multiple levels of control of
nuclear import. These control points include the sequence specific
binding of karyopherins to various NLSs and modulation of this
interaction by other cytoplasmic proteins. In addition, the
differential and inducible expression of karyopherins
may play a
role in regulating nuclear protein transport.
hSRP1 and hSRP1 anti-peptide antibodies were
prepared using the sequences: CMSTPGKENFRLKS and CMSTNENANTAARLHR
corresponding to the amino termini of hSRP1 and hSRP1
, respectively
(11). The peptides were coupled to keyhole limpet hemocyanin and
injected into rabbits to raise polyclonal antibodies. Both antibodies
were affinity purified over the corresponding peptide affinity column and shown not to be cross-reactive with one another (data not shown).
BSA-FITC nuclear localization sequence conjugates were prepared by
mixing 25 mg of BSA-FITC (Molecular Probes) with 10 mg of sulfo-SMCC
(Pierce) for 2 h at room temperature. The complex was dialyzed
into PBS and the NLS peptides added and coupled overnight at 4 °C.
The FITC-BSA-NLS conjugate was finally dialyzed against transport
buffer prior to use. Anti-Hsc73 antibodies were from StressGen
Biotechnologies.
Human peripheral blood lymphocytes were purified by Ficoll gradient centrifugation. Cytosolic extracts from either peripheral blood lymphocyte or cell lines were prepared by resuspending PBS washed cells in buffer "A" (20 mM HEPES, pH 7.2, 20 mM NaCl, 2.5 mM MgCl2, 0.1% Nonidet P-40), 5 min, 4 °C. Nuclei were pelleted at 2000 × g. for 4 min at 4 °C and vortexing. The cytosolic extract was gently removed by pipette and diluted 1:1 with cold PBS. Nuclear proteins were extracted using buffer A plus 0.42 M NaCl. A BCA protein assay (Pierce) was used to determine protein concentration.
Preparation of NLS-BSA-Sepharose ResinsBSA-Sepharose was
prepared using activated CH-Sepharose as described by the manufacturer
(Pharmacia Biotech Inc.). In order to couple the NLS peptides to the
BSA- Sepharose, sulfo-succinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate was added to
a 3-fold weight excess to the BSA-Sepharose and mixed at room
temperature for 2 h. The resin was washed five times in PBS
followed by incubation with the NLS peptides overnight at 4 °C.
Peptides were synthesized with a cysteine at either the C or N terminus
for coupling to the affinity resin. Equal molar amounts of each of the
different peptides was coupled to the BSA resin. There was
approximately 1.0 peptide bound per BSA molecule. The amount of peptide
bound to BSA-Sepharose was determined by measuring the amount of free cysteine (derived from the peptide) using Ellmans reagent before and
after coupling. This procedure agrees well with amino acid analysis to
measure coupling efficiency. The sequences of the NLSs are as follows:
SV40 large T antigen, CGGGPKKKRKV; SV40 mutant, CGGGPKKAAAV; ICP8,
CHRIEEKRKRTYETFKSI; HIV1422, CGGGKKKYKLK; HIV 1423, CGGGKSKKKAQ (15);
NF-B P50, CYPEIKDKEEVQRKRQKL; NF-
B P50 mutant,
CYPEIKDKEEVQRAAQKL; Myc, CGGGPAAKRVKLD. Peptides were synthesized on a
Gilson multiple peptide synthesizer using Fmoc (N-(9-fluorenyl)methoxycarbonyl) amino acids. All of the
peptides were analyzed by mass spectrometry and yielded the correct
molecular weight. Similar results were obtained with three different
preparations of affinity resins.
Raji or Jurkat cell extracts were prepared as described above. The cytosolic extracts were precleared for 2 h at 4 °C with the BSA-Sepharose resin with no NLS attached. After preclearing, 25 µl of Sepharose-BSA-NLS resin was incubated with cytosolic extract from 5 × 106 cells with mixing at 4 °C for 1 h. Samples were then washed four times with ice-cold PBS (1 ml each wash). 100 µl of SDS-gel sample buffer were then added to each sample.
For the experiment using purified K2-GST fusion protein, a cytosolic extract from 1 × 107 Jurkat cells was mixed with 2 µg of K2-GST fusion protein for 2 h at 4 °C prior to incubating with the NLS affinity resins. The purified protein alone (2 µg) was incubated in the presence of 1 mg/ml BSA in a 1:1 mixture of buffer A and PBS with the affinity resins. After incubation, the samples were washed four times with cold PBS and analyzed by Western blot.
Nuclear Import AssayIn vitro nuclear import was measured using a modification of previously published procedures (25). Cytosol extracts were prepared from either Jurkat, HSB2, or 70Z/3 cells as follows. Cells were suspended in 1.5 volumes of lysis buffer (5 mM HEPES-KOH, pH 7.35, 110 mM potassium acetate, 5 mM sodium acetate, 2 mM magnesium acetate, 0.1 mM EGTA, 2 mM dithiothreitol, and protease inhibitor mixture containing 0.2 mM phenylmethylsulfonyl fluoride, 1 µg/ml pepstatin, 1 µg/ml leupeptin, and 1 µg/ml tosylphenylalanyl chloromethyl ketone. The cells were allowed to swell for 10 min on ice and then homogenized with a motorized Teflon pestle for 15 strokes. The cytosol was dialyzed against transport buffer (20 mM HEPES-KOH, pH 7.35, 110 mM potassium acetate, 5 mM sodium acetate, 2 mM magnesium acetate, 0.1 mM EGTA, 2 mM dithiothreitol) overnight. Protein concentration of the extract was between 8 and 16 mg/ml. For the import assay, cells were diluted into carbonate buffer and attached to glass slides using Cell-Tak (Collaborative Research). To permeabilize cells, digitonin was diluted from a frozen 20 mg/ml stock in Me2SO to 40 µg/ml in transport buffer. Cells were rinsed twice with cold transport buffer after permeabilization.
The final transport reaction consists of 1 mM ATP, 10 mM creatine phosphate, 0.5 unit/ml creatine phosphokinase, varying concentrations of NLS-FITC-BSA, and 50% cytosolic extract in a 25-µl aliquot. This mixture was placed on the digitonin-permeabilized cells, and coverslips were incubated for 1 h in a 37 °C incubator. Cells were rinsed twice with cold PBS and fixed in 2% paraformaldehyde, PBS for 15 min, protected from light. Cells were again rinsed twice and aspirated to near dryness, and 7 µl of Slo-Fade (Molecular Probes) were added per circle. Samples were analyzed by fluorescent confocal microscopy (Bio-Rad).
Production of K2 Fusion ProteinThe cDNA for the Rch1
(K2) protein was amplified from a phorbol myristic acid-activated human
T cell library by polymerase chain reaction and inserted into pCDM8.
The GST-K2 fusion protein was produced by amplifying from this
construct, inserting the polymerase chain reaction fragment into the
BamHI-EcoRI sites of pGEX-2T (Pharmacia) and
transforming into Escherichia coli (DH5, Life
Technologies, Inc.). Protein was purified on a glutathione-agarose affinity column (Sigma).
A 96-well
Immulon plate was coated overnight with 20 ng of a p50
NF-B-ovalbumin conjugate. Varying concentrations of unconjugated soluble inhibitor peptides were then added and incubated for 15 min,
followed by the addition of 100 ng of biotinylated K2-GST. The plate
was incubated for 1 h at 22 °C followed by four washes with PBS
containing 0.01% Tween. The biotinylated K2-GST was detected using
strepavidin-horseradish peroxidase. All reactions were performed in PBS
plus 0.01% Tween. The IC50 was determined to be the
concentration of soluble peptide which yielded 50% inhibition of
binding of K2-GST to the NLS conjugate bound to the plate.
In order to study the
expression of karyopherin , cytosolic and nuclear extracts from
various lymphocyte cell lines were prepared and analyzed by Western
blot for levels of K1 and K2. Although equal amounts of protein were
loaded for each sample, there were clear differences in the levels and
cytoplasmic:nuclear distribution of K1 and K2 in the various cell lines
as detected by ECL analysis (Fig. 1A). All of
the cell lines except for the murine 70Z/3 cells are of human origin.
Raji (human B) and Jurkat (human T) cells express the highest levels of
the karyopherins and have slightly higher cytoplasmic:nuclear ratios
for K1 than for K2. We found that karyopherin
is present in both
the cytoplasm as well as the nucleus in different nuclear:cytoplasmic
ratios, depending on the cell type. THP-1 (human monocytic), 70Z/3
(mouse pre-B), and HSB-2 (human pre-T) cells express significantly
lower levels of K1 and K2 and appear to express one predominant form per cell type. That is, THP-1 and HSB-2 cells express mostly K2, whereas the 70Z/3 cells express mostly K1. Although the 70Z/3 cells are
of mouse origin, there are only two amino acid differences between the
mouse pendulin (K2) and human hSRP1
(K2) in the sequence we used for
raising antibodies (11). Hence, we believe that K2 is not expressed in
this mouse cell line. As a control, levels of the heat shock protein
Hsc73 were similar in the different cell lines and appeared equally
distributed between the cytoplasm and nucleus (Fig. 1A).
In comparison to the cell lines, the levels of K1 and K2 were much lower, on a per protein basis, in normal human peripheral blood lymphocytes. Although there was variation in the basal levels of K1 and K2 depending on the donor, in each case the levels were lower than in the cell lines. Since the levels were quite low we explored whether the levels could be increased upon cell activation. As seen in Fig. 1B, activation of peripheral blood lymphocyte with lipopolysaccharide (LPS), concanavalin A (ConA), or phorbol myristic acid (PMA) + ionomycin, causes significant increases in the level of expression of K1 and K2 in the cytoplasm and nucleus. In cell lines already expressing K1 and K2, there was no significant increase after cell activation (data not shown).
K1 and K2 Bind Differently to Various NLSsSince there are
many different NLSs (22), we investigated whether K1 and K2 bind with
similar or different specificities to a variety of NLS peptides. To
study this interaction, Sepharose-BSA-NLS affinity resins were
incubated with cytosolic extracts from either Raji or Jurkat cells. As
seen in Fig. 2, using either Raji or Jurkat lysates,
there were clear differences in the ability of either K1 or K2 to
interact with the different NLS resins. K1 and K2 appeared to interact
with greatest affinity for the SV40 T antigen NLS, but there was very
little binding to the SV40 mutant resin which had 3 amino acid
substitutions. The HIV1422 NLS appeared to interact similarly to the
SV40 NLS, while K1 and K2 had a lower affinity for the other NLSs. To
support the affinity chromatography studies, we also quantitated the
binding of soluble free NLS peptides to purified K2-GST using a
competition binding assay as described under "Materials and
Methods." Based on the chromatography data we analyzed low (p50
NF-B) and high (SV40) affinity NLS:K2 interactions in the solution
binding assay. We found that the SV40 and p50 NF-
B NLSs had
IC50 values of 0.1 and 12.5 µM, respectively.
This suggests that there is an approximately two orders of magnitude difference in the affinity of these NLSs for K2. These data support the
results shown in Figs. 2 and 3 and confirm our findings
that the NLSs interact with differential affinities to karyopherin
.
The differential interaction of K1 and K2 with the various NLSs appeared to be dependent on the cell type used for affinity chromatography (Fig. 2). For example, whereas K1 binds to the ICP8 and Myc NLS in the Raji cells, there was no binding of these NLSs in the Jurkat cells. In addition, K2 bound to the Myc, ICP8, and HIV1423 NLS to a greater extent in the Raji cell lysate, as opposed to the Jurkat lysate.
In order to address whether other cytoplasmic proteins could affect the
ability of K2 to interact with NLSs we cloned and expressed K2 as a GST
fusion protein. We then analyzed the interaction of the fusion protein
alone versus the fusion protein incubated with a cytoplasmic
extract. As seen in Fig. 3B, the K2-GST fusion protein had
the highest relative affinity for the SV40 and HIV1422 NLS. Binding
among the NF-B p50, Myc, and HIV1423 NLSs to K2-GST was equivalent,
whereas the amount of binding to ICP8 and the mutant NLSs was much
less. When the fusion protein was mixed with cytosolic extracts the
relative binding specificities were quite different compared to using
the purified protein alone (Fig. 3A). There were some slight
differences in binding between the native K2 and the fusion protein
which migrates at a higher molecular weight; however, the most
significant differences are between Fig. 3, A and
B. The purified K2-GST binds well to the NF-
B p50 NLS,
whereas it does not bind when present in the cytosolic mix of protein.
The binding of K2-GST to the Myc and HIV1423 NLSs was also diminished
when mixed with the cytoplasmic proteins, whereas binding to SV40,
HIV1423, and ICP8 was unchanged. These data clearly show that
cytoplasmic proteins can alter the binding specificity of K2 to
different NLSs.
As described above,
we showed that there was differential expression and interaction of
karyopherins with NLSs depending on the cell type. Therefore, using an
in vitro nuclear transport assay (25) we examined whether
there was a correlation between nuclear import and either the
expression or interaction of karyopherins with NLSs. Fig.
4A shows that the SV40 NLS-BSA conjugate is
transported into the nucleus much more efficiently than the ICP8 NLS
conjugate at equimolar concentrations (panel a versus b).
There was also a good correlation between the nuclear transport of the
SV40, p50 NF-B, ICP8, and Myc NLS conjugates and their ability to
bind to the NLS affinity resins (Fig. 4A, panels
c-f, compared to Fig. 2). We predicted that there might be
differences in the ability of different cell extracts to reconstitute
nuclear transport based on the differential expression of the
karyopherins
. As seen in Fig. 4B, there were subtle
differences in the ability of different cell extracts to reconstitute
import of the p50 NF-
B NLS BSA conjugate. It appears that the 70Z/3
extract was most efficient at supporting nuclear import followed by the
Jurkat and the HSB2 extracts. Although there are no dramatic
differences in the ability of the cell extracts to reconstitute import,
there is a correlation with expression of K1 and K2 and their ability
to bind to the p50 NF-
B NLS conjugate. That is, 70Z/3 cells express
K1 which binds best to the p50 NF-
B NLS compared to the HSB2 cells
which express K2 which binds weaker to this particular NLS. We have reproducibly seen that in the HSB2 cells, the import substrate is
mostly concentrated at the nuclear rim. This may be due to limited
expression of other components of the import machinery such as ran
which is required for the energy dependent translocation through the
nuclear pore, among other possibilities. Based on the data presented
above it appears that nuclear transport correlates best with the
ability of karyopherins
to bind to NLSs, as opposed to the
expression of the karyopherins
in the cell.
Nuclear transport of BSA-FITC-NLS
conjugates. Nuclear transport assays were performed as described
under "Materials and Methods." A, nuclear transport of
different NLS conjugates using Jurkat cell extracts. In panels c-f, different
concentrations of the conjugates were used to attain similar levels of
nuclear transport. Panels g and h show that there
is minimal transport of the p50 NF-B mutant-BSA-FITC conjugate or
BSA-FITC alone. B, nuclear transport using cytosolic
extracts from either HSB2, Jurkat, or 70Z/3 cells. 1.0 µM
of the p50 NF-
B BSA-FITC conjugate was used with equivalent amounts
of cytosolic protein from each of the different cell types.
Panels a and b, Jurkat extracts; panel
c, HSB2 extract; panel d, 70Z/3 extract.
Many of the recent studies on nuclear protein transport have
focused on the mechanics of the import process in nonhematopoietic cells such as HeLa and Xenopus oocytes, as well as in
reconstituted in vitro systems (1, 4). In contrast to these
studies, we have focused on the transport machinery in lymphocytes and
leukocyte cell lines. In the studies presented here, we have analyzed
the expression and interactions of human karyopherins with
different NLSs.
Lymphocyte activation induces the expression of a wide variety of
genes, which are dependent upon the translocation of specific transcription factors to the nucleus (26, 27). A key protein in this
translocation process is karyopherin , the NLS-binding protein (7).
We found that, in general, the less differentiated cells (THP-1, 70Z/3,
HSB-2) have lower levels of karyopherin
and tend to express only
one predominant form (Fig. 1). In contrast, the more mature Raji and
Jurkat cell lines express much higher levels of both K1 and K2.
Differences were also seen in the cytoplasmic:nuclear ratios of the
karyopherins. The differences in both overall expression and
intracellular distribution of karyopherins may reflect the transcriptional activity of the particular cell line and the
requirement for transport of particular proteins into the nucleus.
Unexpectedly, the levels of both K1 and K2 were very low in resting
human peripheral blood lymphocytes in comparison to the cell lines.
This suggested that other karyopherins
may be expressed in these
cells, or that the levels are low due to the "resting" state of the
cells which require a low level of transcription. We found that
commonly used stimuli of peripheral blood lymphocytes were able to
induce the expression of K1 and K2. Although it has been shown that
cell activation leads to increased numbers of nuclear pores to
facilitate protein nuclear transport (28), the increased expression of the proteins involved in the transport process may be an additional mechanism for enhancing nuclear transport rates.
Another potential mechanism for regulation of nuclear transport could be at the level of the NLS itself. Previous studies have shown that different NLSs have varying abilities to target proteins to the nucleus (29, 30). In order to determine whether NLSs vary in their ability to interact with the NLS-binding proteins K1 and K2, we precipitated intracellular proteins using NLS affinity resins and found that there were differences in the ability of K1 and K2 to interact with NLSs (Fig. 2). These results were confirmed using a solution binding assay. These data suggest that the different classes of karyopherins have both different and overlapping specificities for the various NLSs.
The specificity for the NLSs was also dependent upon the cell type. For
example there was no binding of K1 to the Myc and ICP8 NLS in the
Jurkat cell, whereas there was binding to these NLSs in the Raji cell
extract. These data suggest that other cytosolic proteins may influence
the way in which K1 or K2 interact with NLSs. To address this
possibility we analyzed the ability of a K2 fusion protein to interact
with NLSs in the presence or absence of other cytosolic proteins. As
seen in Fig. 3, the binding specificity of K2 for the NLS is clearly
altered in the presence of other cytosolic proteins. This modulation of
binding could be explained by a number of possibilities. One
possibility is that karyopherin , which has previously been shown to
increase the affinity of karyopherin
for NLSs, may also alter its
binding specificity. The heat shock protein Hsc73, which has been shown
to play a role in nuclear transport (21), may modulate NLS:karyopherin
interactions. In fact, Hsc73 does bind to NLSs and appears to
modulate the association and dissociation reaction of
karyopherins.2 A third possibility is that
endogenous NLS-containing proteins can compete for binding of
karyopherins
to the NLS affinity resins. These possibilities are
now being explored.
Finally, we analyzed the ability of different NLS substrates to be
targeted to the nucleus. The ability of NLS-BSA conjugates to be
imported into the nucleus appeared to be determined by the strength of
the interaction between the NLS and karyopherin . Although the
Jurkat, HSB-2, and 70Z/3 cells expressed different levels and ratios of
the two karyopherins
, there were only subtle differences in the
ability of extracts from these cells to facilitate nuclear import. This
would suggest that there may be other karyopherins expressed that are
not detected by our antibodies or that other cytoplasmic proteins can
modulate the ability of the existing karyopherins to import proteins to
the nucleus. It appears that the ability of proteins to be imported
into the nucleus is predominantly determined by the interaction of the
NLS with karyopherins and to a limited extent by the differential
expression of the karyopherins.
The process of nuclear protein import plays a key role in gene
regulation based on its ability to modulate transcription factor nuclear localization. The results described above suggest that there
are multiple levels of control of nuclear import in lymphocytes and
leukocyte cell lines. Studies are in progress to determine whether the
activation and induction of karyopherins in resting, G0,
peripheral blood lymphocytes has an impact on nuclear transport. Taken
together, our data suggest their are complex mechanisms for the
regulation of nuclear import in different leukocyte cell lines that are
mainly governed by the NLS and its sequence specific interaction with
karyopherin .
We thank Dr. Jeffrey A. Ledbetter for his support of this project and reading of the manuscript. We also thank Sheri Fujihara and Drs. Peter Kiener and Ira Mellman for critically reviewing the manuscript.