| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 279, Issue 20, 20613-20621, May 14, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Department of Biochemistry, Emory University School of Medicine, Atlanta, Georgia 30322
Received for publication, February 17, 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
. In this study, we utilize in vitro solution binding assays and in vivo analyses to directly test this model. We demonstrate that mimicking phosphorylation at a site adjacent to an NLS decreases the binding affinity of the NLS for importin . This decrease in cargo affinity for importin correlates with a decrease in nuclear accumulation in vivo. Through these analyses, we show that the cell cycle-dependent nuclear import of the Saccharomyces cerevisiae transcription factor Swi6p correlates with a phosphorylation-dependent change in affinity for importin . Furthermore, we present data using the SV40 NLS to suggest that this form of regulation can be utilized to artificially modulate the nuclear import of a cargo, which is usually constitutively targeted to the nucleus. This work defines one molecular mechanism for regulating nuclear import by the classic NLS-mediated transport pathway. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
All macromolecules that move into and out of the nucleus are transported through nuclear pore complexes, large proteinaceous channels that are embedded in the nuclear envelope (1, 2). Soluble factors are required to recognize, target, and transport most macromolecules through the nuclear pores (3–5). The best characterized nuclear transport process occurs via receptor recognition of classic nuclear localization signals (NLSs)1 on protein cargoes targeted for nuclear import (3, 6). These classic NLS cargo proteins are recognized in the cytoplasm by a heterodimeric receptor composed of importin/karyopherin and 
(3, 5, 7–9). Importin recognizes and binds the NLS, and importin translocates the trimeric import complex through the nuclear pore (3, 5). Delivery into the nucleus is dependent on the small GTPase Ran, which governs the interactions between the nuclear transport receptors and macromolecular cargoes and thus confers directionality to nucleocytoplasmic transport (3, 5, 10). Once inside the nucleus, the cargo is delivered and the transport receptors are recycled to the cytoplasm (3, 5).
Classic NLSs are typified by a single cluster of basic amino acids (monopartite) or two clusters of basic amino acids separated by a 10- to 12-amino acid linker (bipartite) (11, 12). The prototypical monopartite NLS is that of the SV40 large-T antigen (PKKKRKV) and the prototypical bipartite NLS is that of nucleoplasmin (KRPAATKKAGQAKKKK). Recent studies that have examined the structural and energetic contributions made by individual amino acid residues within the sequence have refined our definition of a functional NLS and have thereby allowed easier identification of potential NLS signals within a protein sequence (13–15).
The dynamic compartmentalization of NLS-containing proteins requires regulated changes in the relative import and/or export rates of a protein. In numerous cases, transport of an NLS cargo into the nucleus is regulated by phosphorylation (16, 17). There are at least three ways in which phosphorylation could regulate protein import into the nucleus: 1) phosphorylation could cause a conformational change in the protein, which reveals or masks an NLS sequence; 2) phosphorylation could cause the release or binding of an NLS masking protein; or 3) phosphorylation could directly modulate the affinity of an NLS for the import receptor importin . Examples of each of these mechanisms include the growth regulatory protein STAT (signal transducers and activators of transcription) where phosphorylation causes the protein to dimerize creating an NLS (18), the p65 subunit of NF-
B where phosphorylation of NF-B-bound I-B leads to degradation of I-B and unmasking of the NF-B NLS (19), and the v-Jun oncoprotein where phosphorylation may directly modulate interactions with the NLS receptor (20). Direct regulation of the interaction between an NLS cargo and importin (regulation method 3 above) requires that the phosphorylation site is within or adjacent to the NLS. In contrast, the other two forms of regulation, methods 1 and 2 above, could be due to phosphorylation at any site within the protein. Each of these modes of regulation is non-exclusive, and thus a combination could regulate the nuclear transport of a particular cargo. Here we will focus on the direct modulation of importin binding by phosphorylation.
A number of important regulatory proteins contain phosphorylation sites located within or adjacent to classic NLS sequences. Recent studies have shown that transport into and out of the nucleus correlates with regulated phosphorylation at these sites. For example, p53 (21), the adenomatous polyposis protein (APC) (22), and the Saccharomyces cerevisiae Swi6 protein (23) all have documented phosphorylation sites within or proximal to their NLS sequences. Phosphorylation of these sites is associated with cytoplasmic protein localization and, conversely, hypophosphorylation correlates with nuclear protein localization. Thus, phosphorylation appears to represent an important mechanism to regulate the nuclear transport and consequently the function of an NLS cargo protein. Furthermore, the proximal position of the phosphorylation site to the NLS sequence suggests that the nuclear import of these cargoes may be modulated by directly regulating the binding affinity of the NLS for the NLS receptor.
Importin structural studies revealed the determinants for specific recognition of classic NLS sequences by the NLS-binding pocket of importin (13, 14, 24). Structures of the NLS binding domain of S. cerevisiae importin (amino acid residues 89–530) bound to various NLS peptides showed that this domain of importin consists of ten helical repeats known as armadillo motifs (13, 14). These armadillo motifs form a concave NLS binding groove, which is lined by conserved tryptophan and asparagine residues and surrounded by acidic amino acids. This structure creates specific binding pockets for NLS cargoes that combine both hydrophobic interactions and electrostatic interactions with the positively charged residues of the NLS (13, 14). These observations suggest that the addition of a negatively charged phosphate group proximal to an NLS could decrease binding of the NLS to importin by disrupting the electrostatic interactions. Furthermore, a similarly positioned negative group within any classic NLS sequence may change its binding affinity for importin and thus modulate the intracellular localization of a cargo protein.
Recent studies that correlate differences in the rate of import with changes in the phosphorylation state of NLS-containing proteins (17, 25) have not examined the change in binding affinity between the NLS and its receptor using a quantitative assay. Furthermore, although one of these studies solved the co-crystal structure of a phosphorylated peptide containing an NLS sequence and non-autoinhibited importin , the phosphorylation site was 14 amino acids upstream of the NLS sequence. In contrast, in our analyses the phosphorylation sites examined are located one amino acid upstream of the NLS (26). Thus, it is not known to what extent phosphorylation proximal to or within an NLS changes the binding affinity for importin . Furthermore, it is important to determine if this change in affinity is sufficient to account for the observed changes in protein localization. A complete understanding of phosphorylation-mediated regulation of nuclear import by modulation of the interaction between an NLS and importin requires a quantitative model for the import of a cargo that correlates the in vitro interaction energies with the in vivo localization of a protein.
Here, we examine the effect of mimicking phosphorylation on the affinity of various NLS sequences for importin . We have utilized site-directed mutagenesis, in vitro binding assays, and in vivo analyses to investigate this mechanism of regulating nuclear transport. Through these analyses we show that mimicking phosphorylation of residues adjacent to an NLS decreases the affinity of that NLS for importin . This decrease in affinity correlates with a decrease in nuclear accumulation of an NLS cargo reporter protein. We use a model cargo, the S. cerevisiae transcription factor Swi6p, to demonstrate that this mechanism of regulation occurs in vivo. We propose that this mode of regulation could be exploited to artificially manipulate the steady state localization of proteins.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
) and Swi6p. His6-
IBB-importin (residues 89–530), His6-Swi6p, His6-Swi6p K163A, His6-Swi6p S160E, and various His6-NLS-GFP proteins were expressed in the Escherichia coli strain BL21(DE3) and purified by nickel affinity chromatography essentially as described previously (15, 29).
Fluorescence Anisotropy Assay—Fluorescence anisotropy measurements were carried out using an ISS PC1 fluorometer fitted with polarization filters. The dissociation constants for the binding of the NLS-GFP proteins to importin were measured essentially as described previously (15, 29). Briefly, the NLS-GFP was diluted in PBS to the desired concentration (
20 nM) in a total volume of 2 ml in a 1-cm quartz cuvette. Changes in the anisotropy of the GFP fluorophore were monitored as increasing amounts of IBB-importin (IBB-) protein were added to the assay volume. Changes in anisotropy were used to calculate the fraction of the GFP fluorophore bound, yielding a binding isotherm for the reaction. The binding isotherm was then fit through non-linear regression to a simple binding equation to obtain dissociation constants. All Kd values are calculated as detailed at www.biochem.emory.edu/Hodel/Research/BindingCurves/fitting_curves.htm. The dissociation constant for full-length Swi6p binding to IBB- was measured using a competition assay (30) with Swi6 NLS-GFP. The assay was carried out as described above with changes in the anisotropy of Swi6 NLS-GFP monitored in the presence of increasing amounts of IBB-. This yielded a dissociation constant for Swi6 NLS-GFP binding to IBB-. To measure the binding of different full-length Swi6p variants to IBB-, a competition assay was carried out with each full-length protein. The binding of Swi6 NLS-GFP was examined in the presence of four different concentrations of recombinant full-length wild-type or mutant Swi6p. The Kd values for wild-type and mutant Swi6 proteins binding to IBB- were determined by fitting the resulting binding curves to an equation for the fraction of Swi6 NLS-GFP bound as a function of the Kd for Swi6 NLS-GFP, the Ki for the full-length Swi6 protein, the total Swi6 NLS-GFP concentration, the total IBB- concentration, and the total concentration of the full-length Swi6 protein. Binding energies were calculated using G = RT ln Kd, where RT = –0.59 kcal/mol.
Immunoblot Analysis—Immunoblot analysis was performed by standard methods (31). Briefly, cultures were grown to log phase in appropriate media at 30 °C. Cells were harvested by centrifugation and washed twice in water and once in PBSMT (100 mM KH2PO4, pH 7.0, 15 mM (NH4)2SO4, 75 mM KOH, 5 mM MgCl2, 0.5% Triton X-100). Cells were subsequently lysed in PBSMT with protease inhibitors (0.5 mM phenylmethylsulfonyl fluoride, 3 µg/ml each of aprotinin, leupeptin, chymostatin, and pepstatin) by glass bead lysis. Equal amounts of total protein (generally 10 µg) were resolved by SDS-PAGE and immunoblotted with polyclonal anti-GFP antibody (1:10,000 dilution) (32).
Microscopy—Direct fluorescence microscopy was used to localize GFP fusion proteins in live cells. For all experiments, cells were stained with 1 µg/ml 4',6-diamidino-2-phenylindole to visualize the DNA and confirm the location of the nucleus. The localization of the fusion proteins was monitored by directly viewing the GFP signal in living cells through a GFP-optimized filter (Chroma Technology) using an Olympus BX60 epifluorescence microscope equipped with a Photometrics Quantix digital camera.
Two-photon microscopy was employed to image the yeast cells for quantitative analysis of the distribution of the GFP fluorescence for each NLS fusion protein. The images were generated using a custom built two-photon microscope (33, 34). The fluorescence quantification and nuclear:cytoplasmic ratio (Rn/c) calculation was carried out by sectoring the images into nuclear and cytoplasmic compartments through a semi-automated procedure using the programs Mathematica (Wolfram Research) and Corel Photopaint (Corel). In each image, the average fluorescence per pixel was calculated for each sectored compartment. The compartments from different images in the Z-scan were grouped into continuous three-dimensional spaces representing multiple slices through a single yeast cell. The cytoplasmic fluorescence for each cell was then determined as the mean of the fluorescence per pixel averages across Z-scans containing that compartment. Because the dimensions of the nuclei of the cells were of the same order of magnitude as the distance between Z-scans (1 µm), the nuclear fluorescence was calculated using just the value determined from the single Z-scan image that sliced through the center of the nucleus of each cell. This analysis yields a ratio, Rn/c, of the amount of nuclear:cytoplasmic fluorescence within a population of cells, which serves as a measure of how much cargo protein is present in the nucleus.
Generation of SWI6 Mutants—Amino acid substitutions were introduced in the S. cerevisiae SWI6 coding region using an overlap PCR strategy (35). The desired DNA mutations were designed into the appropriate oligonucleotides. Yeast genomic DNA was used as a template to amplify SWI6 with the mutations. The PCR products were further amplified by PCR to generate the SWI6 open reading frame expressed from the endogenous SWI6 promoter. The resulting PCR product was cloned into the yeast centromeric (CEN) plasmid pRS315 (36). A similar strategy was used to clone the SWI6 mutations into the C-terminal GFP yeast expression vector (pAC242) and the bacterial expression vector pET28a (Novagen) (pAC762). For all constructs generated, the presence of each desired mutation and the absence of any other mutations was confirmed by DNA sequencing.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
. Furthermore, a similarly positioned negatively charged group within any classic NLS sequence might change its binding affinity for importin . To test these predictions, we designed sequences where the residue proximal to two SV40 NLS derivatives (15) was either a serine (SV40 and SV40A7) or glutamic acid residue (SV40E and SV40A7E) (Fig. 1B). These studies utilized both the SV40 NLS and the weaker binding SV40A7 NLS (15), because it was also hypothesized that to regulate the localization of an NLS-containing protein, the NLS would need to bind importin with a relatively weak affinity close to that of the threshold for a functional NLS. Within the recombinant proteins, the serine residue of SV40 and SV40A7 represents a hypophosphorylated NLS sequence and the negative glutamic acid residue of SV40E and SV40A7E mimics the negative charge created by phosphorylation and thus represents a constitutively phosphorylated NLS.
|
—To compare the binding of the SV40, SV40A7, SV40E, and SV40A7E NLS sequences to importin , we used a quantitative solution binding assay based on fluorescence anisotropy (15, 29). The NLS sequences were cloned as in-frame N-terminal fusions to the green fluorescent protein (GFP). The strategy was designed such that each NLS is in the same context within the GFP fusion proteins (15). Thus, when comparing the different NLSs, there should be no contribution of conformational change to any of the binding affinities measured (15).
The binding of an NLS to N-terminally truncated importin (IBB-) is used as a measure of binding to cytoplasmic importin . This simplifies the interpretation and collection of data, because the N-terminal autoinhibitory domain, which is normally bound by importin in the cytoplasm (37), is absent. We have previously shown that the binding affinities of NLS sequences for either IBB- or full-length importin in the presence of importin are similar (15, 29). Furthermore, this IBB- is identical to the protein that was crystallized in complex with different NLSs (13, 14) and used in the quantitative dissection of an NLS (15), facilitating direct comparison to these analyses.
To examine SV40-NLS-GFP, SV40E-NLS-GFP, SV40A7-NLS-GFP, and SV40A7E-NLS-GFP binding to IBB-, we performed the fluorescence anisotropy assay as described under "Experimental Procedures." Typical curves for SV40A7-NLS-GFP (
) and SV40A7E-NLS-GFP (
) binding are shown in Fig. 2A. As described under "Experimental Procedures," these types of curves are used to calculate Kd values for the interaction between the NLS cargo and importin (Table II).
|
|
IBB- binds to SV40-NLS-GFP and SV40A7-NLS-GFP with Kd values of 9 and 80 nM, respectively (15). The SV40E and SV40A7E mutants bind less tightly (Kd 39 nM and Kd 283 nM) than SV40 and SV40A7, an 3.5- to 4-fold decrease in affinity. This shows that the glutamic acid residue, which mimics phosphorylation, causes a decrease in the affinity of each NLS for importin .
Steady-state Localization of NLS Fusion Proteins—To determine if the difference in affinity of the NLS proteins for importin measured in vitro correlates with a change in the in vivo localization of the protein, we constructed a reporter cargo for use in live yeast cells. The spatial context of the reporter is identical to the in vitro NLS-GFP fusion protein except that a second GFP protein was fused in-frame to the C terminus to increase the size of the reporter (molecular mass of 55 kDa) and presumably decrease transport by passive diffusion. Direct fluorescence microscopy was used to analyze the intracellular localization of each NLS-GFP-GFP fusion protein (Fig. 2B). Controls demonstrate that a GFP-GFP protein localizes throughout the cell while wild-type SV40-NLS-GFP-GFP accumulates within the nucleus (Fig. 2B, compare panels A and B). The SV40A7-NLS-GFP-GFP protein also accumulates in the nucleus (Fig. 2B, panel C). In contrast, the SV40A7E-NLS-GFP-GFP protein appears to have less nuclear accumulation than SV40A7-NLS-GFP-GFP (Fig. 2B, compare panels C and D). The localization of SV40A7E-NLS-GFP-GFP is similar to the GFP-GFP protein (Fig. 2B, compare panels A and D). Immunoblotting of the GFP fusion proteins demonstrates that each protein is expressed at approximately the same level suggesting that changes in signal are not due to changes in the amounts of the various proteins (Fig. 2C, compare lanes 2–5).
In our analysis the SV40E-NLS-GFP-GFP protein showed similar localization to SV40-NLS-GFP-GFP (data not shown). To determine if there was a subtle difference in localization between the SV40-NLS-GFP-GFP and SV40E-NLS-GFP-GFP proteins, we utilized srp1-31 cells. srp1-31 cells express a conditional allele of importin . Consequently, these cells are defective in the import of NLS cargoes (38). This defect in NLS cargo import is minor at 25 °C but can be utilized to shift the dynamic range for assaying import. Thus, by examining the localization of NLS cargo proteins in srp1-31 cells, it is possible to more easily observe relative differences in nuclear accumulation. At 25 °C the SV40-NLS-GFP-GFP protein accumulates in the nucleus of srp1-31 cells, but the SV40E-NLS-GFP-GFP protein localizes throughout the cell (data not shown). Our analysis of the engineered SV40 NLSs suggests that the glutamic acid substitution in both SV40 and SV40A7 decreases the binding to the import receptor. Furthermore, although SV40A7E can bind to importin , the affinity appears to be below the affinity required for efficient nuclear import in vivo.
Swi6p as a Model Protein Cargo—Our data with the SV40 NLS variants suggest that phosphorylation at a site adjacent to an NLS can directly modulate its binding to importin . To determine if a similar effect is observed for a protein regulated by phosphorylation in vivo, we utilized the S. cerevisiae protein Swi6p. Swi6p is a transcription factor that regulates the cell cycle-specific expression of several genes (39). It enters the nucleus only during the G1 phase of the cell cycle (23, 40). During this phase of the cell cycle, a single serine at position 160 of Swi6p is hypophosphorylated (23). In contrast, during the other phases of the cell cycle, serine 160 is phosphorylated (23). This cell cycle-dependent phosphorylation of Swi6p correlates with its in vivo localization (23). When Swi6p is phosphorylated, it localizes primarily to the cytoplasm, and when it is hypophosphorylated, it localizes to the nucleus (23). This phosphorylation site (serine 160) is adjacent to the NLS of Swi6p (Fig. 3) (23). Thus, Swi6p is a model protein cargo to study the effects of phosphorylation on the affinity of an NLS for importin .
|
has not been examined. To directly measure the binding of the Swi6 NLS to IBB-, we fused the NLS sequence of Swi6 (residues 157–169) in-frame with GFP (Swi6 NLS) in the same context as the SV40 NLSs. Additionally, to test our hypothesis, we changed the serine residue within the Swi6p NLS to either alanine (Swi6A), to create an NLS that could not be phosphorylated, or glutamic acid (Swi6E), to mimic a constitutively phosphorylated Swi6 NLS (Fig. 3).
Swi6p NLS Binding to Importin —To examine Swi6 NLS binding to IBB-, we performed the fluorescence anisotropy assay as described for the SV40 NLSs. Typical binding curves for Swi6-NLS-GFP () and Swi6E-NLS-GFP () binding to IBB- are shown in Fig. 4A. As would be expected for the recombinant proteins, substitution of the serine residue with an alanine residue in the Swi6 NLS did not have a significant effect on binding to IBB- (data not shown). The binding affinity of Swi6-NLS-GFP for IBB- is 26 nM (Table II). In contrast, substitution of the serine with a glutamic acid residue in the Swi6 NLS (Swi6E) decreased the affinity for IBB- by 4.8-fold (Kd 124 nM) (Table II). In addition, as a control for the change in size upon substitution of the glutamic acid for serine, we also substituted the glutamic acid with a glutamine (Swi6Q). The Swi6Q-NLS-GFP protein showed similar binding to IBB- as was observed for Swi6-NLS-GFP, suggesting that it is the charge of the glutamic acid rather than the size that alters the affinity of Swi6E-NLS-GFP for importin (data not shown).
|
To quantitatively compare the localization of each of these NLS-GFP-GFP proteins, we utilized a two photon microscopy approach as described under "Experimental Procedures." The rate of import of each of the NLS-GFP-GFP fusion proteins was measured using a steady-state method. This analysis yields a ratio of the nuclear to cytoplasmic (Rn/c) fluorescence, which is a measure of the relative import rate of each NLS fusion protein at steady state. The Rn/c for each NLS is shown below the respective direct microscopy panel in Fig. 4B. This analysis yields an Rn/c for the Swi6 NLS fusion protein (2.2) that is higher than when the serine is mutated to a glutamic acid residue (1.5). Our results suggest that the import rate of Swi6 NLS is faster than the import rate for the Swi6E NLS.
Binding of Full-length Swi6p to Importin —The NLS of Swi6p binds 4.8-fold less tightly to importin when we mimic constitutive phosphorylation (Table II). Furthermore, this change in binding affinity correlates with a change in protein localization (Fig. 4B). To investigate if the effects observed for these isolated NLS sequences can account for the phosphorylation-mediated regulation of Swi6p nuclear localization in vivo, we examined the modulation of the full-length Swi6 protein using a similar approach.
To measure the binding affinity of full-length Swi6p for importin , we used a competition assay where we examined the ability of full-length Swi6p to compete with the Swi6-NLS-GFP for binding to IBB-. We first used fluorescence anisotropy to measure the binding of Swi6-NLS-GFP to IBB- (Fig. 5A). This yielded a Kd of 26 nM. The competition experiment was then carried out in the presence of four concentrations of the full-length Swi6 proteins in competition with Swi6-NLS-GFP. This analysis yields an equilibrium binding constant for the interaction of full-length Swi6 protein with IBB- of 45 nM (Fig. 5A and Table II).
|
, we carried out the competition assay with a variant of full-length Swi6p where the serine was changed to glutamic acid (S160E) (binding curves not shown). Swi6 S160E binds to IBB- 3.5-fold less tightly (Kd 163 nM) than full-length wild-type Swi6p (Kd 45 nM) (Table II). The difference in binding affinity between the full-length Swi6 protein (
) and the Swi6E () protein for IBB- is revealed when comparing the binding curves for competition with Swi6-NLS-GFP (Fig. 5B). At equal concentrations, the Swi6 protein competes better than the Swi6E protein for binding to IBB- in the fluorescence anisotropy competition assay with Swi6-NLS-GFP (Fig. 5B, compare full-length Swi6 protein () and the Swi6E () protein binding curves at 1.5 x 10–7 M IBB-).
Monopartite NLS sequences have an essential lysine residue that is necessary for a high affinity interaction with importin (15). Therefore, as a control for binding specificity, we also carried out the competition assay with full-length Swi6p where the predicted essential lysine residue within the NLS was changed to alanine (K163A). The Swi6 K163A protein did not compete with Swi6 NLS-GFP for binding to IBB- even at high concentrations (data not shown).
Comparison of the Change in Free Energy—The decrease in binding of the mutant SV40, SV40A7 NLS, Swi6 NLS, and full-length Swi6p to IBB- can be expressed as a change in free energy (G). The G is the difference in binding between each wild-type protein and the corresponding glutamic acid mutant protein (Fig. 5C). As shown in Fig. 5C, the G between the wild-type and glutamic acid substituted full-length Swi6 proteins is similar to that obtained for the isolated NLS sequences. This analysis suggests that a negatively charged amino acid proximal to an NLS directly impacts binding to importin .
Localization of Full-length Swi6-GFP—To determine whether the measured binding affinities between importin and the Swi6p variants correlate with the nuclear import of each protein, full-length Swi6-GFP proteins were generated. These tagged proteins were expressed from the SWI6 promoter on a centromeric plasmid. At steady state, the localization of these proteins is the result of the ratio between import and export. In each of the different Swi6 variants, we are presumably directly targeting import, whereas export should be similar for each protein. The localization of these proteins was examined in both wild-type and srp1-31 cells (38).
The localization of each of the Swi6-GFP proteins in srp1-31 cells at 25 °C is shown in Fig. 6. The wild-type Swi6 protein shows cell cycle-dependent localization (Fig. 6, panel A). As a control for import of full-length Swi6p via importin , we examined the localization of Swi6 K163A-GFP where the mutant NLS no longer mediates binding to importin (data not shown). The Swi6 K163A protein localized throughout the cell with no nuclear accumulation (Fig. 6, panel B). In contrast, Swi6 S160A-GFP showed nuclear accumulation during all phases of the cell cycle (Fig. 6, panel C). This is different than the localization of the wild-type Swi6 protein where it is cytoplasmic at all stages of the cell cycle except G1. This further supports the hypothesis that phosphorylation of serine 160 directly inhibits import into the nucleus. Mimicking phosphorylation of Swi6p (Swi6 S160E) rendered the protein cytoplasmic at G1, whereas the wild-type Swi6-GFP protein had obvious nuclear accumulation (Fig. 6, compare panels A and D). The localization of Swi6 K163A-GFP and Swi6 S160E-GFP was similar (Fig. 6, compare panels B and D). In wild-type (data not shown) and srp1-31 cells we observed similar effects on the localization of the various proteins, although the effects were more pronounced in srp1-31 cells. An immunoblot of the Swi6-GFP proteins showed that each variant was expressed at approximately the same level (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
. In this study, we utilize in vitro solution binding assays and in vivo analyses to directly test this model. We demonstrate that mimicking phosphorylation at a site adjacent to an NLS decreases the binding affinity of the NLS for importin . This decrease in cargo affinity for importin correlates with a decrease in nuclear accumulation in vivo. These analyses suggest that the cell cycle-dependent nuclear import of the S. cerevisiae transcription factor Swi6p correlates with a phosphorylation-dependent change in affinity for importin . Furthermore, we present data with the SV40 NLS that suggest that this form of regulation can be utilized to artificially modulate the nuclear import of a cargo, which is usually constitutively targeted to the nucleus. Taken together, these analyses define one molecular mechanism for regulating nuclear import by the classic NLS-mediated transport pathway.
The recognition of an NLS by the NLS binding pocket of importin is dependent on electrostatic interactions. This is demonstrated by structural studies of importin bound to various NLS peptides, which revealed extensive electrostatic interactions between the positively charged residues of the NLS and the negatively charged residues that surround the NLS binding pocket (13, 14, 24). Furthermore, in vitro binding studies showed that mutation of basic residues to uncharged residues within the NLS can impact the binding to importin (15, 29). Conversely, a mutant of importin , ED-importin (D203K/E402R), where two of the negatively charged residues surrounding the NLS binding pocket are mutated to positively charged residues has significantly reduced binding to NLS cargo (42, 43). Thus, our data, where the addition of a negatively charged group proximal to an NLS decreases binding to importin , is consistent with these observations. The negatively charged phosphate presumably disrupts the electrostatic interactions between the NLS and importin .
We demonstrate that the binding of two classic monopartite NLSs can be modulated by mimicking phosphorylation. In this analysis, the phosphorylation sites are both similarly positioned adjacent to the N terminus of the NLS sequences. Previous studies with the APC protein suggest that its nuclear import is regulated by phosphorylation of a serine residue at the C terminus of a monopartite NLS sequence (22). It has not been formally tested if phosphorylation at this site impacts the binding of APC to importin . It will be interesting to use a similar analysis to determine if phosphorylation at the C terminus of an NLS regulates nuclear import by a similar molecular mechanism.
Although we demonstrate this mode of regulation for classic monopartite NLS, it is clear that it may also regulate the binding of classic bipartite NLS sequences to importin . For example, the S. cerevisiae transcription factor Swi5p has a bipartite NLS with a phosphorylation site in a similar position as that within the monopartite NLS sequences studied here (Fig. 1A) (41). Swi5p also has cell cycle-dependent nuclear localization that correlates with its phosphorylation (41). Thus, it appears that the regulation of nuclear import by phosphorylation of classic NLS-cargoes may be a common mechanism for regulating protein localization and function. Interestingly, there are numerous clinically important proteins that contain documented phosphorylation sites within classic NLS sequences. For example, the bipartite NLS of the tumor suppressor p53 has a cdc2/cyclin kinase site located within the linker of the bipartite NLS (Fig. 1A) (21). It will be interesting for future studies to examine the impact of phosphorylation on import of clinically important cargoes.
The nucleocytoplasmic dynamic localization of proteins reflects relative changes in the import and export rates of nuclear transport. Swi6p shows cell cycle-dependent localization in vivo (23). This nucleocytoplasmic dynamic localization of Swi6p presumably reflects relative changes in the import and export rates. The Swi6 protein accumulates in the nucleus during G1, presumably because the rate of import exceeds the rate of export. Our data suggest that this accumulation is mediated by an increased affinity of hypophosphorylated Swi6 protein for importin . Conversely, Swi6p is mainly cytoplasmic during other stages of the cell cycle apparently due to phosphorylation decreasing its affinity for importin . The kinase(s) and phosphatase(s) that regulate the phosphorylation of serine 160 of Swi6p have, to this date, not been identified. It will be interesting to identify these enzymes and characterize their role in the cell cycle-dependent localization of Swi6p. Furthermore, the role of export in the dynamic compartmentalization of Swi6p has recently been investigated by Queralt and Igual (40). Deletion of the Swi6p export receptor, Msn5p, renders Swi6p constitutively nuclear (40). Interestingly, Swi6p maintains its cell cycle-dependent localization when a classic nuclear export signal (44) is fused to the C terminus (40). This supports the hypothesis that import is the major regulator of the cell cycle-dependent nuclear localization of Swi6p.
Dynamic compartmentalization appears to be one mechanism that can be used to modulate protein activity. Using phosphorylation to regulate protein localization is advantageous due to the ease of reversibility for rapid changes, which may be required at a specific time in the cell cycle or in response to a stimulus. Indeed, the import of Swi6p into the nucleus is regulated to ensure transcription of specific target genes at a specific point in the cell cycle. A previous study shows that substitution of serine 160 with an aspartic acid residue, which is also thought to mimic phosphorylation, results in a more cytoplasmic localization than wild-type Swi6p during the G1 phase of the cell cycle (23). Furthermore, this mimicking of phosphorylation at serine 160 does not have an appreciable effect on the timing or periodicity of Swi6-specific gene transcripts, but it does lead to a decrease in the amount of Swi6 responsive mRNA transcribed (23). This is consistent with our analysis where mimicking of phosphorylation does not completely block binding to importin but rather decreases it. Thus, we observe a decrease in the nuclear accumulation of Swi6p and not an exclusion from the nucleus.
The import of an NLS-containing protein into the nucleus appears to correlate with the affinity with which it binds to importin . In this study, we demonstrate that mimicking phosphorylation of an NLS can decrease the affinity of the NLS for importin by 3.5–5-fold. Thus, the regulation of direct binding to importin by phosphorylation is not an on/off switch, but rather a change in the ratio of nuclear: cytoplamic amounts of the modified NLS-protein. Hence, the functional significance of this form of regulation appears to be dependent on how much of a particular NLS protein is required in the cellular compartment for it to perform its function.
Many proteins localize differentially between the nucleus and cytoplasm in response to cell cycle progression, growth signals, or environmental stimuli. Although dynamic compartmentalization modulates the activities of these proteins, our understanding of the mechanisms that regulate nuclear transport is relatively poor. Furthermore, phosphorylation directly regulates the binding of protein cargoes to other members of the importin family of nuclear transport receptors (16, 17). For example, phosphorylation within the NLS sequence of the S. cerevisiae protein, Pho4p, reduces its affinity for the importin family receptor, Pse1p, and impedes import of Pho4p in vivo (45, 46). Understanding these mechanisms of regulation for different nuclear transport receptors and cargoes may allow us to selectively target a particular receptor and/or cargo for regulation. Studies that correlate structural analyses, in vitro interaction energies, and in vivo functionality are necessary to understand these modes of regulation at a mechanistic level. The recognition of nuclear transport as a mechanism to regulate protein activity should enhance our understanding of many biological processes and ultimately may be utilized to help control disease states.
|
FOOTNOTES |
|---|
Note Added in Proof—While this manuscript was under review, the identification of both the kinase, Clb6/Cdc28, and the phosphatase, Cdc14, that regulate the phosphorylation status and cellular localization of Swi6 was published (Geymonat, M., Spanos, A., Wells, G. P., Smerdon, S. J., and Sedgwick, S. G. (2004) Mol. Cell. Biol. 24, 2277–2285).
* This work was supported by National Institutes of Health Grant GM-58728 (to A. H. C.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 
Both authors contributed equally to this work.
To whom correspondence should be addressed: Dept. of Biochemistry, Emory University School of Medicine, 1510 Clifton Rd., NE, Atlanta, GA 30322. Tel.: 404-727-4546; Fax: 404-727-3452; E-mail: acorbe2{at}emory.edu.
1 The abbreviations used are: NLS, nuclear localization signal; CEN, centromeric; DIC, differential interference contrast; GFP, green fluorescent protein; IBB, importin binding; IBB-, importin residues 89–530; APC, adenomatous polyposis protein; PBS, phosphate-buffered saline.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
