Plant Importin α Binds Nuclear Localization Sequences with High Affinity and Can Mediate Nuclear Import Independent of Importin β*

Nuclear import of conventional nuclear localization sequence (NLS)-containing proteins initially involves recognition by the importin (IMP) α/β heterodimer, where IMPα binds the NLS and IMPβ targets the IMPα/NLS-containing protein complex to the nuclear pore. Here we examine IMPα from the plantArabidopsis thaliana (At-IMPα), which exhibits nuclear envelope localization typical of IMPβ rather than IMPα in other eukaryotic cell systems. We show that At-IMPα recognizes conventional NLSs of two different types with high affinity (K d of 5–10 nm), in contrast to mouse IMPα (m-IMPα), which exhibits much lower affinity (K d of 50–70 nm) and only achieves high affinity in the presence of m-IMPβ. Unlike m-IMPα, At-IMPα is thus a high affinity NLS receptor in the absence of IMPβ. Interestingly, At-IMPα was also able to bind with high affinity to NLSs recognized specifically by m-IMPβ and not m-IMPα, including that of the maize transcription factor Opaque-2. Reconstitution of nuclear import in vitroindicated that in the absence of exogenous IMPβ subunit but dependent on RanGDP and NTF2, At-IMPα was able to mediate nuclear accumulation to levels comparable with those mediated by m-IMPα/β. Neither m-IMPα nor -β was able to mediate nuclear import in the absence of the other subunit. At-IMPα’s novel NLS recognition and nuclear transport properties imply that plants may possess an IMPα-mediated nuclear import pathway independent of IMPβ in addition to that mediated by IMPα/β.

Although NLS receptors from different species share structural and functional homology, experimental evidence suggests that nuclear import in plant cells has unique features compared with that in other eukaryotes. In contrast to the latter, in vitro transport in plant cells appears not to be inhibited by the nucleoporin-binding lectin wheat germ agglutinin and to occur at low temperature and in the absence of exogenously added cytosol (see Refs. 29 and 30). In addition, the NLS-binding IMP␣ subunit from Arabidopsis thaliana (At-IMP␣) shows nuclear envelope association (30,31) in similar fashion to IMP␤ in mammalian and other cell systems (8,10,11); IMP␣ in mammalian and yeast systems shows predominantly nucleoplasmic location as well as cell cycle-dependent localization in either cytoplasm or nucleus in Drosophila (32). A linkage of At-IMP␣ with the cytoskeleton has also recently been demonstrated, with a mechanistic role in nuclear import surmised (33). Because of these novel properties, we set out to quantitate the NLS binding properties of At-IMP␣ for the first time using an ELISA-based assay (34,35). We find that At-IMP␣ binds NLSs of different types with high affinity independent of an IMP␤ subunit, in contrast to the IMP␣ subunits from mouse and yeast, which require their respective IMP␤ subunits to achieve high affinity binding (21, 34 -36). At-IMP␣, together with Ran/ TC4 and NTF2 and in the absence of IMP␤, was able to mediate nuclear import in vitro to levels comparable with those mediated by mouse IMP␣/␤ (m-IMP␣/␤). m-IMP␣ was unable to mediate nuclear import in the absence of m-IMP␤. At-IMP␣ thus shows unique properties, being able to fulfill both NLS recognition and nuclear import in the absence of IMP␤.

MATERIALS AND METHODS
Fusion Proteins and Peptides-The amino acid sequences of the NLSs contained within the fusion proteins and peptides used are shown in Table I. The plasmid expressing Xenopus laevis N1N2 amino acids 465-581 fused amino-terminal to ␤-galactosidase was derived by inserting a 351-base pair EcoRV/NcoI-filled in fragment from the N1N2 cDNA (39) into the unique SmaI site of the plasmid vector pPR2 (37). The T-ag peptides pep101Lys and pep101Thr (40) and maize transcription factor (TF) Opaque-2 (O2) bipartite NLS-containing pepO2 (31,41) have all been described previously (see Table I).
Protein concentrations were determined using the dye binding assay of Bradford (45), with bovine serum albumin as a standard.
ELISA-based Binding Assay-An ELISA-based binding assay (34,35,42,46) was used to examine the binding affinity between importin subunits (with and without GST moieties) and NLS-containing proteins or peptides. The latter were coated onto 96-well microtiter plates and incubated with increasing concentrations of IMP subunits, and detection of bound IMP-GST was performed using a goat anti-GST primary antibody, an alkaline phosphatase-coupled rabbit anti-goat secondary antibody, and the substrate p-nitrophenyl phosphate (34). In some experiments, including all those with At-IMP␣, affinity-purified antibody specific to At-IMP␣ (30,31), which cross-reacts with mouse and yeast IMP␣ but not ␤ (Ref. 30 and data not shown), was used instead of anti-GST antibody. Absorbance measurements were performed over 90 min using a plate reader (Molecular Devices), with values corrected by subtracting absorbance both at 0 min and in wells incubated without IMP (34). To examine importin binding to the NLSs in the case of the T-ag-and N1N2-␤-galactosidase fusion proteins, quantitation was performed in parallel for ␤-galactosidase, and these values were subtracted from those for the respective fusion proteins (34,35,42). Fusion proteins were also subjected to a parallel ␤-galactosidase ELISA (see Ref. 34) to correct for any differences in coating efficiencies and enable a true estimate of bound importin (34 -36, 42, 46). Experiments were also performed in which IMP␣s (GST-free) were coated onto microtiter plates, and m-IMP␤-GST and y-IMP␤-GST binding was assessed in similar fashion to the NLS-binding ELISAs above.
In Vitro Nuclear Transport-Nuclear import kinetics were analyzed at the single cell level using mechanically perforated HTC rat hepatoma cells, cultured as previously (37,38), in conjunction with confocal laser scanning microscopy (35,42,46). Experiments were performed as described (36) for 40 min at room temperature in a 5-l volume containing 30 mg/ml bovine serum albumin, 2 M GTP, and an ATP regenerating system (0.125 mg/ml creatine kinase, 30 mM creatine-phosphate, 2 mM ATP), transport substrate (0.2 mg/ml IAF-labeled fusion protein), and where indicated, 4 M RanGDP, 0.15 M NTF2, 1 M m-IMP␤, 1 M m-IMP␣, or 0.6 M At-IMP␣. In inhibition experiments, specific antibody to either IMP␣ (30, 31) (final concentration of 20 g/ml) or IMP␤ (10) (mAb3E9; final concentration of ϳ200 g/ml from purified ascites fluid) was included in the assay. Image analysis and curve-fitting was performed as described (35,42,46); the level of accumulation at the nuclear envelope, relative to medium fluorescence, was determined using NIH Image 1.60 in line plot mode as previously (47).

At-IMP␣ Displays High Binding Affinity to Different Types of
NLSs-Our previous results using an ELISA-based assay to determine the binding affinities of IMP subunits for different NLSs indicated relatively low affinity on the part of m-IMP␣ compared with that for the m-IMP␣/␤ heterodimer (34 -36, 42). We set out to quantitate the NLS binding affinity of At-IMP␣ using this assay, where proteins and peptides were used that contained the NLS of T-ag, the bipartite NLS of the X. laevis nuclear factor N1N2 (36,39), or the less well defined NLSs of the yeast TF GAL4 (48) and the plant TF O2 (49) (see Table I). We found that At-IMP␣ bound all of these NLSs, present in fusion proteins and/or as peptides, with 2-5-fold higher affinity than m-IMP␣ (Figs. 1 and 2, Table II, and data not shown), the latter only attaining binding affinity comparable with that of At-IMP␣ in the presence of m-IMP␤ (Table II). The specificity of binding in all cases was indicated by the fact that binding to NLS-mutant derivatives was severely reduced (less than 8% maximal binding compared with the respective wild type derivative; Table II). The results indicate that At-IMP␣ is able to recognize different types of NLS with high affinity independent of an IMP␤ subunit (see below).
That At-IMP␣ bound GAL4 so well was somewhat surprising, since its NLS is recognized exclusively by IMP␤ subunits and not by IMP␣ (see Ref. 23). The data in Fig. 2 and Table II (see legend) illustrate this, showing clearly that m-IMP␣, in contrast to At-IMP␣ and m-IMP␤, does not recognize the GAL4 NLS. Similarly, the NLS contained within pepO2 is recognized with high affinity by m-IMP␤ (K d of 10.1 nM) rather than m-IMP␣ (K d of 80 nM), and pepO2 is bound by At-IMP␣ with high affinity (K d of 11.7 nM; see Table II). Since At-IMP␣ recognizes the GAL4 and O2 NLSs with high affinity, it clearly possesses some of the NLS binding properties of IMP␤.
Importin ␤ Subunits Can Recognize At-IMP␣ without Affecting NLS Binding Affinity-Binding of IMP␤ to IMP␣ has been shown to increase the NLS binding affinity of the latter (21,34,35). We decided to investigate whether At-IMP␣ is able to dimerize with the IMP␤ subunits m-IMP␤ and y-IMP␤ employing our ELISA-based assay. We found higher affinity binding of At-IMP␣ to y-IMP␤ (K d of 10 Ϯ 2.5 nM, n ϭ 5) than to m-IMP␤ (K d of 20 Ϯ 5.1 nM, n ϭ 13). For comparison, m-IMP␣ was also tested, exhibiting highest affinity for m-IMP␤ (K d of 2.2 Ϯ 0.9 nM, n ϭ 4), as expected, and a K d of 21 Ϯ 1.2 nM (n ϭ 3) for y-IMP␤. a The single letter amino acid code is used; residues represented by lowercase letters indicate mutations abrogating nuclear targeting activity (37).
Since At-IMP␣ exhibited high affinity binding to m-IMP␤, we also tested whether binding of m-IMP␤ influenced NLS recognition by At-IMP␣, using our ELISA-based assay and antibodies to either At-IMP␣ or m-IMP␤-GST. The affinity of At-IMP␣ for the NLSs within the various fusion proteins and peptides after precomplexation to m-IMP␤ was not significantly different from that in its absence (Figs. 1 and 2; Table  II). This was in stark contrast to the effect of m-IMP␤ on NLS recognition by m-IMP␣ (Table II), which significantly increased the binding affinity for the various NLSs 4 -8-fold (see also Refs. 34 and 35).
At-IMP␣ Can Mediate Nuclear Protein Import Independent of IMP␤-The NLS-recognition properties of At-IMP␣ implied that it possessed some of the properties of IMP␤ as well as IMP␣, so that we decided to compare the ability of At-IMP␣ to that of the mouse importins to mediate nuclear import in a reconstituted in vitro system in the presence of purified RanGDP and NTF2. Since plant protoplasts retain sufficient amounts of transport factors, including IMP␣, to support nuclear import in the absence of exogenously added cytosol (29,30), we used our well characterized system of mechanically perforated rat hepatoma cells, which requires the addition of cytosol and an ATP-regenerating system for NLS-dependent nuclear accumulation (35-37, 42, 46). We initially examined the ability of At-IMP␣ and the mouse importins to mediate binding to the nuclear envelope in the absence of Ran/NTF2 (Fig. 3A). No significant targeting of T-ag-CcN-␤-Gal to the nuclear envelope was observed in the absence of exogenously added importin subunits (Fig. 3A, bottom right panel, and results not shown); although m-IMP␣ (Fig. 3A, top right panel) or -␤ alone could not mediate binding as expected, the combination of the two was able to do so with high efficiency (Fig. 3A,  Table II. bottom left panel). Significantly, At-IMP␣ was able to perform the same function in the absence of m-IMP␤ (Fig. 3A, top left  panel); the maximal level of accumulation at the nuclear envelope, as determined by image analysis using the line plot mode (47), was 2.6 Ϯ 0.2 (n ϭ 16) times that in the cytoplasm, compared with a level of 2.4 Ϯ 0.1 (n ϭ 15) effected by m-IMP␣/␤. This activity was consistent with At-IMP␣'s ability to localize at the nuclear envelope/NPC (30,31).
In the presence of RanGDP and NTF2, m-IMP␣ or ␤ alone could not mediate nuclear import of T-ag-CcN-␤-Gal (Fig. 3B), levels of nuclear accumulation being similar to those in the absence of IMP subunits (Fig. 3C, Table III). This was in contrast to the m-IMP␣/␤ combination, where maximal nuclear accumulation relative to that in the cytoplasm (Fn/c max ) was over 5-fold (half-maximal within 2.6 min). Significantly, At-IMP␣, in the absence of exogenously added IMP␤ subunit, was able to mediate nuclear import to comparable levels (Fn/c max of 3.7; half-maximal within 8 min). The specificity of transport in all cases was demonstrated by the fact that the T-ag-Cc-␤-Gal protein, containing a nonfunctional NLS, was not accumulated to any significant extent (Fig. 3, B and C; Table III). The results thus demonstrated that At-IMP␣ could mediate nuclear import of an NLS-containing transport substrate independent of an IMP␤ subunit. While nuclear import mediated by At-IMP␣ was  Table II.  Table I. c Detection of binding was carried out using anti-GST antibodies; all other analyses were performed using anti-At-IMP␣ antibodies (see Materials and Methods). d Significant differences (p Ͻ 0.022) between columns C and D (Student's t-test). e LB, low binding; K d unable to be determined. See Footnotes f, g, k, and l. f Maximal binding by T-ag-Cc-␤-Gal was 6.7 Ϯ 3.8 (4) and 6.8 Ϯ 3.9 (4) % that for T-ag-CcN-␤-Gal on the part of At-IMP␣ in the presence or absence of m-IMP␤, respectively.  3. Ability of At-IMP␣ to mediate nuclear protein import reconstituted in vitro in the absence of exogenously added IMP␤. Nuclear import was reconstituted in mechanically perforated HTC cells in the presence of an ATP-regenerating system containing GTP/GDP and using NTF2 and GDP-loaded Ran as described under "Materials and Methods." A, visualization (after a 10-min incubation at room temperature) of binding at the nuclear envelope of the T-ag-NLS-containing fusion protein T-ag-CcN-␤-Gal mediated by IMP subunits in the absence of Ran and NTF2. B, visualization (after a 25-min incubation at room temperature) of nuclear accumulation mediated by IMP subunits in the presence of Ran and NTF2. Lack of import of the NLS mutant derivative T-ag-Cc-␤-Gal is shown in the bottom panels. C, quantitative results for nuclear import enhanced by NTF2, especially in terms of the rate of import, it did not absolutely appear to require it, since an Fn/c max of over 3 was observed in its absence ( Fig. 4; Table III).

TABLE II NLS binding parameters of At-IMP␣ in the presence and absence of m-IMP␤ as measured using an ELISA-based assay
Nuclear import mediated by At-IMP␣ was also tested in the presence of the m-IMP␤ subunit (Figs. 3B and 4). Interestingly, nuclear import was reduced by more than 50%, compared with in the absence of m-IMP␤, in either the absence or presence of NTF2 ( Fig. 4; Table III), this inhibition of At-IMP␣-mediated transport presumably resulting from m-IMP␤'s ability to bind At-IMP␣ (see above). To test whether the nuclear import mediated by At-IMP␣ alone was attributable to small amounts of residual IMP␤ in the assay, inhibition experiments were performed using antibodies to either IMP␣ (29) or IMP␤ (10) (Fig.  5). Clear results were obtained, whereby the antibody to IMP␣ reduced nuclear import mediated by At-IMP␣ together with Ran over 90%, while the antibody to IMP␤ had no inhibitory effect but rather enhanced transport by almost 70%, consistent with the idea that IMP␤ inhibited At-IMP␣-mediated nuclear import (Fig. 5). In contrast to the results for At-IMP␣, the antibodies to IMP␣ or IMP␤ both inhibited nuclear import mediated by m-IMP␣/␤ in conjunction with Ran, maximal nuclear accumulation being reduced by over 80 and 95%, respectively (Fig. 5). Clearly, the anti-IMP␤ antibody was functional in inhibiting IMP␣/␤-mediated nuclear import, the results for At-IMP␣ thus demonstrating that its ability to mediate nuclear import was not attributable to residual contaminating IMP␤ in the assay system; rather, the residual IMP␤ appeared to inhibit its action. An additional control was performed of m-IMP␣ alone together with Ran in the absence or presence of the antibodies, indicating that the background level of accumula-tion in the absence of exogenously added m-IMP␤ was not increased by the anti-IMP␤ antibody (Fig. 5), further underlining the functional differences between m-IMP␣ and At-IMP␣.
Nuclear import kinetic measurements were also performed in vitro for the first time using the bipartite NLS-containing substrate N1N2-␤-Gal (Table III and data not shown), producing results similar to those for T-ag-CcN-␤-Gal. Maximal levels of At-IMP␣-mediated import (Fn/c max of about 3.4; half-maximal within 15.5 min) were obtained in the presence of RanGDP; the rate of import was almost twice as fast in the presence of NTF2 (t1 ⁄2 of ϳ9 min; Table III). In the presence of m-IMP␤, accumulation was reduced by ϳ40 and 60% in the presence and absence of NTF2, respectively. Clearly, independent of IMP␤, At-IMP␣ is able to mediate nuclear import not only of a T-ag-NLS-containing protein (above) but also of a bipartite NLS-containing protein.

DISCUSSION
It has become obvious over the last 2 years that there are multiple signal-dependent nuclear import pathways additional to that mediated by the IMP␣/␤ heterodimer. These include those mediated exclusively by IMP␤-related homologs such as transportin, where no IMP␣ subunit homolog is involved (24 -27), and those where soluble cytosolic receptors are not required at all such as nuclear import conferred by the heterogenous nuclear ribonucleoprotein K KNS shuttle sequence (50) and the human immunodeficiency virus type 1 Tat NLS (46). It has also recently been shown that, in the absence of IMP␣, IMP␤ can mediate NLS recognition and nuclear import of U small nuclear ribonucleoproteins and proteins such as human kinetics of the fusion proteins T-ag-CcN-␤-Gal and T-ag-Cc-␤-Gal mediated by mouse IMP subunits (left panel) or At-IMP␣ (right panel) in the presence of Ran and NTF2. Results shown are from a single typical experiment, where each data point represents at least five separate measurements each of Fn, Fc, and background fluorescence (see "Materials and Methods"). Data are fitted for the function Fn/c ϭ Fn/c max * (1 Ϫ e Ϫkt ), where Fn/c is the nuclear/cytoplasmic ratio, k is the rate constant, and t is time (min) (34,35,42). Pooled data are presented in Table III.
where Fn/c max is the maximal level of nuclear accumulation, k is the rate constant, and t is time (min) (34,35). Where n ϭ 1, the S.E. is from the curve fit. b LA, low nuclear accumulation; t 1 ⁄2 unable to be determined.
immunodeficiency virus type 1 Rev (51,52), TCPTP (22), and parathyroid hormone-related protein (24), implying that IMP␤ itself can function independently as a nuclear import receptor. The present study's findings that IMP␣ from plant is able both to bind NLSs with high affinity and to mediate nuclear import independent of IMP␤ represent evidence for a further variant of a signal-dependent nuclear import pathway. The results show that At-IMP␣ can fulfill the NLS binding and NPC dock-ing roles of the IMP␣/␤ heterodimer and thereby may function in analogous fashion to IMP␤/IMP␤-related homologs in the pathways mentioned above. This study thus implies that nuclear import pathways mediated exclusively by IMP␣ subunits may exist in plants and possibly other eukaryotes (see Ref. 53). Rice IMP␤ has recently been shown to be able to interact with rice IMP␣ (54) to mediate nuclear import (55), so that it seems likely that the novel IMP␣-mediated pathway described here may exist in addition to conventional pathways in which IMP␣ acts in conjunction with IMP␤. We and others have shown that the NLS binding affinity of IMP␣ from yeast or mouse increases when it is bound to IMP␤ (21, 34 -36). Furthermore, nuclear import of NLS-containing proteins recognized by IMP␣ is dependent on IMP␤ for the NPC docking and translocation steps in these systems (Refs. 8,11,18,and 36;see Ref. 56). In this study, we found that in the absence of IMP␤, At-IMP␣ binds NLSs with high affinity similar to that possessed by yeast and mammalian IMP␣/␤ heterodimers. Further, in the absence of exogenously added IMP␤, At-IMP␣ is able to mediate the nuclear import of proteins containing two different types of NLS; although as elsewhere (54,55,57) the NLSs used in the transport studies here are not plant-derived, the possibility that this is the basis of At-IMP␣'s apparent novel properties seems unlikely, since plant NLSs resemble those from other eukaryotes in all respects, e.g. in terms of NLS-type (T-ag-NLS-like, bipartite, and Mat␣2-like) (29,30). That residual IMP␤ in our nuclear import assay system is not the basis of the observed results is demonstrated by the lack of inhibition of At-IMP␣-mediated nuclear import by anti-IMP␤ antibodies (Fig. 5); rather, IMP␤ appears to inhibit At-IMP␣-mediated nuclear import (Figs. 4 and 5). Based on the results here and elsewhere, At-IMP␣, in contrast to mouse and yeast IMP␣ subunits (21,35,36), thus appears to possess a number of IMP␤ properties/characteristics: 1) At-IMP␣ localizes at the nuclear envelope/NPC in tobacco protoplasts (30) in analogous fashion to IMP␤ subunits in mammalian and yeast systems (e.g. Ref. 8; see, however, Refs. 53 and 58); 2) like mouse and yeast IMP␤ subunits (23), At-IMP␣ recognizes the O2 and GAL4 NLSs; and 3) At-IMP␣ can mediate targeting to FIG. 4. Nuclear import reconstituted in vitro using purified components mediated by At-IMP␣ and m-IMP␣ in the absence and presence of exogenously added m-IMP␤. Nuclear import kinetics of the fusion protein T-ag-CcN-␤-Gal mediated by IMP subunits in the absence (left) and presence (right) of m-IMP␤ was measured in the presence of an ATP-regenerating system containing GTP/GDP, GDP-loaded Ran, and NTF2 as indicated. Experiments for transport were performed as described in the legend to Fig. 3. Results shown are from a single typical experiment, where each data point represents at least four separate measurements each of Fn, Fc, and background fluorescence. Pooled data are presented in Table III. FIG. 5. Inhibition of At-IMP␣-and m-IMP␣/␤-mediated nuclear import by specific antibodies to IMP subunits. Maximal nuclear accumulation of T-ag-CcN-␤-Gal above background (transport in the absence of IMP subunits) mediated by At-IMP␣, m-IMP␣, or m-IMP␣/␤ in the presence of Ran was measured in the absence or presence of anti-IMP␣ or anti-IMP␤ antibodies (see "Materials and Methods"). The results represent averages from three separate experiments (measurements performed as described in the legend to Fig. 3), with the S.E. indicated. In the case of the low m-IMP␣-mediated transport, results are expressed relative to those for m-IMP␣/␤. the nuclear envelope of NLS-containing transport substrates, as well as their nuclear accumulation in the presence of Ran and NTF2, independent of IMP␤. These properties appear to be sufficient to enable At-IMP␣ to bind to a variety of NLSs with high affinity and mediate nuclear import in the absence of IMP␤. In this sense, At-IMP␣ represents a composite of conventional IMP␣ and ␤ subunit activities, although as stated above, this is not meant to imply that it does not have a role in conventional IMP␣/␤-mediated transport (55). The specific mechanism by which IMP␣ may mediate IMP␤-independent targeting to the NPC and subsequent nuclear import is unclear at this stage but may relate to the recent intriguing demonstration that At-IMP␣ associates with the cytoskeleton (33).
The structure and function of the NPC is conserved in all eukaryotes (see Refs. 56,58,and 59), and the signals determining passage through it are functional in a variety of cell types; e.g. the T-ag NLS functions in yeast and plant cells (60,61;see Ref. 29), the NLSs of the yeast TF SWI5 (62), and Drosophila morphogen Dorsal (63) can mediate nuclear import in mammalian cells, and the plant-cell functioning NLS of Agrobacterium tumefaciens VirD2 protein is active in yeast cells (64). That At-IMP␣ can bind IMP␤ from both mouse and yeast as shown here implies functional conservation across eukaryotes in terms of the proteins mediating nuclear transport. This is reflected in the relatively high sequence conservation (69 and 77% similarity, respectively) within the IMP␤ binding domains of At-IMP␣ and the mouse and yeast IMP␣ subunits (30). Other regions of IMP␣, such as the region C-terminal to the IMP␤ binding domain including the first armadillo repeat (At-IMP␣ amino acids 50 -118), do not exhibit high homology, showing only 32 and 25% similarity with the mouse and yeast IMP␣ subunits, respectively. A priority of future work will be to determine whether the differences in this domain of At-IMP␣, and in particular the absence of a series of negatively charged amino acid residues present in the IMP␣-subunits from other species (see Ref. 30), may constitute the basis of At-IMP␣'s ability to target nuclear import substrates to the NPC in the absence of IMP␤.
The fact that At-IMP␣ can reconstitute nuclear import in rat nuclei in combination with human Ran and NTF2, as shown here, indicates that transport factors from different species are able to function together to mediate signal-dependent nuclear accumulation. Consistent with this, plant cell cytoplasmic extract is able to support NLS-dependent nuclear import in mammalian cells (65), At-IMP␣ (AtKap␣) is able to rescue a yeast Srp1 (IMP␣) mutant defective for transport (66), and importins from rice have recently been shown to be functional in nuclear transport by digitonin-permeabilized HeLa cells (55,57). The study here can thus be seen as further demonstration of the conservation of the eukaryotic cell nuclear transport apparatus, as well as providing further evidence for its multifaceted nature.