Portal de Periódicos da CAPES

The asymmetric distribution of the constituents of the Ran system is essential for transport into and out of the nucleus

1997; Springer Nature; Volume: 16; Issue: 21 Linguagem: Inglês

10.1093/emboj/16.21.6535

1460-2075

Elisa Izaurralde, Ulrike Kutay, Cayetano von Kobbe, Iain W. Mattaj, Dirk Görlich,

Trace Elements in Health

Article1 November 1997free access The asymmetric distribution of the constituents of the Ran system is essential for transport into and out of the nucleus Elisa Izaurralde Elisa Izaurralde University of Geneva, Department of Molecular Biology, 30 quai Ernest-Ansermet, CH-1211 Geneva 4, Switzerland European Molecular Biology Laboratory, Meyerhofstraße 1, 69117 Heidelberg, Germany Search for more papers by this author Ulrike Kutay Ulrike Kutay Zentrum für Molekulare Biologie der Universität Heidelberg, Im Neuenheimer Feld 282, 69120 Heidelberg, Germany Search for more papers by this author Cayetano von Kobbe Cayetano von Kobbe University of Geneva, Department of Molecular Biology, 30 quai Ernest-Ansermet, CH-1211 Geneva 4, Switzerland Search for more papers by this author Iain W. Mattaj Iain W. Mattaj European Molecular Biology Laboratory, Meyerhofstraße 1, 69117 Heidelberg, Germany Search for more papers by this author Dirk Görlich Corresponding Author Dirk Görlich Zentrum für Molekulare Biologie der Universität Heidelberg, Im Neuenheimer Feld 282, 69120 Heidelberg, Germany Search for more papers by this author Elisa Izaurralde Elisa Izaurralde University of Geneva, Department of Molecular Biology, 30 quai Ernest-Ansermet, CH-1211 Geneva 4, Switzerland European Molecular Biology Laboratory, Meyerhofstraße 1, 69117 Heidelberg, Germany Search for more papers by this author Ulrike Kutay Ulrike Kutay Zentrum für Molekulare Biologie der Universität Heidelberg, Im Neuenheimer Feld 282, 69120 Heidelberg, Germany Search for more papers by this author Cayetano von Kobbe Cayetano von Kobbe University of Geneva, Department of Molecular Biology, 30 quai Ernest-Ansermet, CH-1211 Geneva 4, Switzerland Search for more papers by this author Iain W. Mattaj Iain W. Mattaj European Molecular Biology Laboratory, Meyerhofstraße 1, 69117 Heidelberg, Germany Search for more papers by this author Dirk Görlich Corresponding Author Dirk Görlich Zentrum für Molekulare Biologie der Universität Heidelberg, Im Neuenheimer Feld 282, 69120 Heidelberg, Germany Search for more papers by this author Author Information Elisa Izaurralde1,2, Ulrike Kutay3, Cayetano von Kobbe1, Iain W. Mattaj2 and Dirk Görlich 3 1University of Geneva, Department of Molecular Biology, 30 quai Ernest-Ansermet, CH-1211 Geneva 4, Switzerland 2European Molecular Biology Laboratory, Meyerhofstraße 1, 69117 Heidelberg, Germany 3Zentrum für Molekulare Biologie der Universität Heidelberg, Im Neuenheimer Feld 282, 69120 Heidelberg, Germany *Corresponding author. E-mail: [email protected] The EMBO Journal (1997)16:6535-6547https://doi.org/10.1093/emboj/16.21.6535 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The GTPase Ran is essential for nuclear import of proteins with a classical nuclear localization signal (NLS). Ran's nucleotide-bound state is determined by the chromatin-bound exchange factor RCC1 generating RanGTP in the nucleus and the cytoplasmic GTPase activating protein RanGAP1 depleting RanGTP from the cytoplasm. This predicts a steep RanGTP concentration gradient across the nuclear envelope. RanGTP binding to importin-β has previously been shown to release importin-α from -β during NLS import. We show that RanGTP also induces release of the M9 signal from the second identified import receptor, transportin. The role of RanGTP distribution is further studied using three methods to collapse the RanGTP gradient. Nuclear injection of either RanGAP1, the RanGTP binding protein RanBP1 or a Ran mutant that cannot stably bind GTP. These treatments block major export and import pathways across the nuclear envelope. Different export pathways exhibit distinct sensitivities to RanGTP depletion, but all are more readily inhibited than is import of either NLS or M9 proteins, indicating that the block of export is direct rather than a secondary consequence of import inhibition. Surprisingly, nuclear export of several substrates including importin-α and -β, transportin, HIV Rev and tRNA appears to require nuclear RanGTP but may not require GTP hydrolysis by Ran, suggesting that the energy for their nuclear export is supplied by another source. Introduction Transport between nucleus and cytoplasm proceeds through the nuclear pore complexes (NPCs) that allow diffusion of small molecules of up to 40 kDa and can accommodate active, energy-dependent transport of particles as large as several million daltons. Import and export of different classes of transport substrates through the NPC appears to be mediated by distinct, saturable transport receptors that are thought to shuttle between nucleus and cytoplasm (for review see Gerace, 1995; Görlich and Mattaj, 1996; Koepp and Silver, 1996; Nigg, 1997). According to this model, an import receptor would bind the substrate in the cytoplasm then carry it through the NPC into the nucleus. After the release of the import substrate into the nucleoplasm, the receptor would return to the cytoplasm without the cargo and accomplish the next round of import. Conversely, an export factor has to bind the export substrate in the nucleus and release it in the cytoplasm. This predicts asymmetric import/export cycles and implies that the binding of the substrate to its receptors is somehow regulated by the different environments of nucleus and cytoplasm. The import of proteins with a classical, basic-type nuclear localization signal (NLS) serves as a paradigm for transport with a shuttling receptor (for reviews see Görlich and Mattaj, 1996; Panté and Aebi, 1996; Schlenstedt, 1996). The NLS of the import substrate is bound in the cytoplasm by the heterodimeric importin-α/β complex. Importin-α provides the NLS-binding site, whereas importin-β accounts for the subsequent interactions with the NPC that drive translocation. The trimeric NLS/importin-α/β complex is transferred through the NPC probably as a single entity, and becomes disassembled upon termination of translocation. The two subunits are returned to the cytoplasm, most likely separately. The small GTPase Ran appears to fulfil at least two distinct functions in the import process. First, Ran‘s GTP cycle appears to supply at least a substantial proportion of the energy required for translocation into the nucleus (Melchior et al., 1993a; Moore and Blobel, 1993; Weis et al., 1996). This is presumed to involve nucleotide exchange and GTP hydrolysis on NPC-bound Ran. One nucleoporin, RanBP2, has been shown to bind RanGTP and RanGAP1 specifically (Wu et al., 1995; Yokoyama et al., 1995; Mahajan et al., 1997). However, the identity of the NPC component(s) that would translate Ran's GTP cycle into a directed movement through the NPC is not yet known. The second function of Ran, which has been more clearly established, is to regulate the interaction between importin-α and -β. The direct binding of RanGTP to importin-β dissociates the importin heterodimer (Rexach and Blobel, 1995; Chi et al., 1996; Görlich et al., 1996a). That this dissociation is a specific nuclear event that follows translocation into the nucleus can be explained by free RanGTP being available only inside the nucleus. Ran itself is predominantly, but not exclusively nuclear (Bischoff and Ponstingl, 1991a). Ran's major nucleotide exchange factor is the chromatin-bound RCC1, that generates RanGTP inside the nucleus (Ohtsubo et al., 1989; Bischoff and Ponstingl, 1991b). In contrast, the principal GTPase activating protein RanGAP1 is excluded from the nucleoplasm (Hopper et al., 1990; Melchior et al., 1993b; Bischoff et al., 1994, 1995a; Matunis et al., 1996; Mahajan et al., 1997). RanGAP1 stimulates conversion of RanGTP to the GDP-bound form and thereby depletes RanGTP from the cytoplasm. The Ran-binding protein RanBP1 (Coutavas et al., 1993) is also cytoplasmic (Richards et al., 1996). It binds specifically to the GTP-bound form of Ran and stimulates the GTPase activation by RanGAP1 10-fold (Beddow et al., 1995; Bischoff et al., 1995b). Thus the asymmetric distribution of Ran, RCC1, RanGAP1 and RanBP1 should result in a steep RanGTP gradient across the nuclear envelope with a high nuclear concentration and a very low level in the cytoplasm. Ran and its effector proteins have also been studied in yeast and all major aspects required for the RanGTPase cycle were found to be conserved (for a review see Koepp and Silver, 1996). The second protein import pathway thus far characterized is mediated by transportin, which is related to importin-β (Nakielny et al., 1996; Pollard et al., 1996; Fridell et al., 1997). hnRNPA1 and several closely related hnRNP proteins are targeted into the nucleus by virtue of the M9-domain, the import signal of the transportin-dependent pathway (Michael et al., 1995; Siomi and Dreyfuss, 1995; Weighardt et al., 1995). The M9 signal binds to transportin directly, with no equivalent of importin-α being involved (Nakielny et al., 1996; Pollard et al., 1996; Fridell et al., 1997). A yeast homologue of transportin that also appears to be involved in the import of hnRNP-like proteins has been characterized (Aitchison et al., 1996). It had not been established whether Ran played a direct role in transportin-mediated import. The export of RNAs out of the nucleus is not yet well understood, but it is thought to require the association with specific RNA-binding proteins that carry the actual export signals (reviewed by Izaurralde and Mattaj, 1995). The paradigm for this model is the export of HIV-1 viral RNAs containing a Rev response element. The HIV Rev protein binds to this element and the complex is then exported by virtue of Rev's leucine-rich nuclear export signal (NES) (Bogerd et al., 1995; Fischer et al., 1994, 1995; Fritz et al., 1995; Stutz et al., 1995; Wen et al., 1995). Pre-mRNAs (hnRNAs), the primary RNA polymerase II transcripts, associate in the nucleus with hnRNP proteins that assist RNA processing and probably mediate export of the matured mRNA out of the nucleus (reviewed in Dreyfuss et al., 1993). Indeed, some hnRNP proteins carry nuclear export signals and shuttle rapidly between nucleus and cytoplasm. The shuttling of hnRNP A1 is the best understood. Its M9 domain confers not only transportin-dependent import, it is also essential for nuclear export of hnRNP A1 (Michael et al., 1995). Saturation experiments in Xenopus oocytes provide evidence for a factor (or factors) that recognizes hnRNPA1 through an interaction that includes, but is not confined to, the M9 domain and is required for the export of at least some mRNAs (Izaurralde et al., 1997; Saavedra et al., 1997). So far it had been unclear if transportin would only be involved in the import of M9 proteins or if it would also play a role in M9-dependent export of proteins like hnRNP A1 and of RNA. In addition to the M9 signal other signals like that recently found in hnRNP K protein (Michael et al., 1997) confer shuttling and may well contribute to mRNA export, which in turn would imply that mRNA export may not occur by a single homogenous mechanism. The nuclear export of U snRNAs is distinct from mRNA export. The 7-methyl cap structure and its interaction with the heterodimeric nuclear cap binding protein complex (CBC) is critical for U snRNA export (Hamm and Mattaj, 1990; Izaurralde et al., 1995). In contrast to mRNA export, export of U snRNAs can be inhibited by saturation of the HIV-1 Rev NES export pathway, suggesting that a leucine-rich export signal plays a role in U snRNA export (Fischer et al., 1995). The asymmetric distribution of RanGTP has been proposed to be required for the directionality of CBC-dependent export of U snRNAs. CBC exists in complex with importin-α. This heterotrimer binds capped U snRNAs. However, when importin-β interacts with the CBC-bound importin-α, capped RNA is released (Görlich et al., 1996b). As discussed above, the asymmetric distribution of RanGTP only allows the importin-α/β interaction to occur in the cytoplasm. Thus, CBC releases RNA in the cytoplasm, and nuclear RanGTP prevents this release in the nucleus. A considerable body of genetic evidence suggests that Ran is involved in nuclear export of RNA. For example, expression of a GTPase-deficient Ran has been shown to cause an mRNA export defect in yeast (Schlenstedt et al., 1995). Furthermore, in vivo studies in yeast and in mammalian cells have suggested that Ran's nucleotide exchange factor RCC1 (Prp20p in yeast) is required for nuclear export of U snRNAs and of the majority of mRNAs (Forrester et al., 1992; Amberg et al., 1993; Kadowaki et al., 1993, 1994; Cheng et al., 1995), and that the yeast RanGAP1, Rna1p, is also required for mRNA export (Traglia et al., 1989). However, these genetic experiments do not distinguish whether Ran is directly involved in export or is required solely for the re-import of shuttling export mediators from the cytoplasm. Here we show that Ran is involved directly not only in the import of NLS proteins, but also in M9-dependent nuclear import, in the recycling of transportin, importin–α and -β back to the cytoplasm, and in the export of HIV Rev, tRNA, U snRNA, and several mRNAs. Our data suggest that the requirement of Ran for nuclear export is direct rather than for re-import of potential export mediators. We show that the asymmetric distribution of the Ran system across the nuclear envelope is crucial for nucleocytoplasmic transport and that major transport pathways are blocked when RanGAP1 or RanBP1 are mislocalized into the nuclear compartment. The sensitivity towards these treatments and the requirement for GTP hydrolysis by Ran is different for different import or export pathways. Furthermore, we show that RanGTP releases the M9 signal from transportin and provide evidence that this happens upon entry into the nucleus. This indicates that transportin can import an M9-containing substrate but has to return to the cytoplasm without the cargo. This suggests that the export function of the M9 signal is probably mediated by a receptor distinct from transportin. Results RanGTP dissociates the M9/transportin complex RanGTP binding to importin-β precludes the binding of importin-α. This interaction, together with the asymmetric distribution of the components of the Ran system, is critical in making importin-β a uni-directional import factor for NLS-proteins. To date, it has not been clear whether transport mediators other than importin-β might be regulated in a similar way. Transportin is responsible for nuclear import of proteins with an M9 domain (Pollard et al., 1996; Fridell et al., 1997). The M9 domain, however, is not only an import signal, it is also essential for nuclear export of hnRNP A1 (Michael et al., 1995) which might mean that transportin is responsible for both import and export. However, it is also possible that transportin mediates import, but that the export receptor for hnRNP A1 is distinct from transportin. If transportin was both import and export receptor for hnRNP A1 then it should also bind this substrate in the nucleus, i.e. in the presence of RanGTP. To test this, we performed the experiments shown in Figure 1A. In vitro translated hnRNP A1 efficiently binds to immobilized transportin. However, binding was completely prevented if assayed in the presence of saturating amounts of RanQ69L (GTP form), whereas wild-type RanGDP had no effect on the interaction (Figure 1A). The RanQ69L mutant is GTPase deficient and remains in the GTP-bound form even in the presence of cytoplasmic RanGAP (Bischoff et al., 1994; Klebe et al., 1995). Figure 1B shows that RanGTP can also dissociate a pre-formed M9/transportin complex. Transportin was expressed in Escherichia coli and the total lysate was passed over an immobilized M9 domain to allow transportin to bind. Transportin could be eluted from this column not only with 1 M magnesium chloride, but also specifically with RanGTP demonstrating that RanGTP releases the M9 domain from transportin. As transportin and RanGTP form a stable complex under the conditions of elution as judged by gel filtration (Figure 1C), the release is probably the consequence of the direct binding of RanGTP to transportin. This would also be consistent with the observation that transportin closely resembles importin-β in its effects on the RanGTPase (F.R.Bischoff, S.Nakielny and G.Dreyfuss, personal communication). Figure 1.(A) Regulation of the M9/transportin interaction by Ran. hnRNP A1 was translated in vitro (Input) and subjected to binding to IgG Sepharose alone (Background) or to IgG Sepharose to which z-tagged transportin had been pre-bound. Where indicated, 0.15 μM or 1.5 μM RanQ69L (GTP-form), or 1.5 μM Ran wt (GDP-form) were added before binding. The input and the bound fractions were analysed by SDS–PAGE followed by fluorography. (B) RanGTP dissociates transportin from M9. Transportin was expressed at 15°C in E.coli. A lysate was prepared in binding buffer: 50 mM Tris–HCl pH 7.5, 400 mM NaCl, 10 mM magnesium acetate. 1 ml post-ribosomal supernatant (corresponding to 20 ml culture) was subjected to binding to 30 μl Streptavidin agarose to which the biotinylated nucleoplasmin core-M9 fusion had been pre-bound. After washing with 5 times 1 ml equilibration buffer and 1 h wash in Ran buffer (20 mM potassium phosphate pH 7.0, 100 mM NaCl, 1 mM magnesium acetate), the bound protein was either eluted with 100 μl of 1 M magnesium chloride or 100 μl of 10 μM RanQ69L (GTP form). Analysis was by SDS–PAGE followed by Coomassie staining. (C) RanGTP and transportin form a stable complex. Two nmoles RanQ69L GTP either alone or together with 2 nmoles z-tagged transportin were incubated in 200 μl for 30 min on ice and applied at 0.4 ml/min to a 25 ml Superdex 200 column (Pharmacia) which was equilibrated in Ran buffer. One ml fractions were collected of which 10 μl were analysed by Western blotting with an rabbit anti Ran antibody. Transportin was detected by virtue of its z-tag (the IgG binding domain from Protein A). Note that free RanGTP runs at an elution volume of ∼19 ml whereas in the presence of transportin, most RanGTP elutes as a complex with transportin ∼12 ml. Download figure Download PowerPoint M9 import is Ran-dependent If the interaction between RanGTP and transportin was significant for M9 import, one would predict that M9 import is Ran dependent. Figure 2A confirms this assumption and shows that the efficient import of a fluorescently labelled M9 fusion protein into the nuclei of permeabilized HeLa cells requires the addition of both transportin and Ran. Wild type Ran cannot be substituted by two mutant forms, RanT24N or RanQ69L that are unable to stably bind GTP or unable to hydrolyse the bound GTP respectively (see below). This is consistent with a previous report that RanQ69L can inhibit M9-dependent protein import (Nakielny et al., 1996). Figure 2.(A) Ran-dependence of M9 import. The substrate for import into nuclei of permeabilized cells was a fluorescently labelled fusion between the M9 domain from hnRNP A1 and the core domain from nucleoplasmin. Where indicated, 0.8 μM transportin and 1.5 μM Ran wild type, RanQ69L, or RanT24N were added. Import in the presence of an energy-regenerating system was allowed for 10 min at 20°C. Confocal sections through the equators of the permeabilized cell nuclei are shown. For more details see Materials and methods. (B) The M9 substrate and transportin separate after entry into the nucleus. Import in the presence of wild type Ran was essentially as in (A), with the modifications that the core M9 fusion (1 μM) was labelled with Texas Red maleimide and transportin (0.1 μM) was used in a fluorescein-labelled form. After fixation, the distribution of the M9 substrate and transportin was determined by confocal laser scanning microscopy. The panels shows the two channels separately and the merged image. Note that the M9 fusion strongly accumulated inside the nucleus, whereas transportin gave a weak intranuclear signal and was bright at the nuclear envelope. Download figure Download PowerPoint One would predict from these data that the M9 protein would dissociate from transportin upon nuclear entry, in a manner analogous to the dissociation of importin-β from importin-α. To show this directly, we used a limiting concentration of fluorescein-labelled transportin (0.1 μM) together with the Texas Red labelled M9 reporter to perform an in vitro import reaction in the presence of Ran and an energy-regenerating system. After fixation, the distribution of transportin and the import substrate were determined by confocal fluorescence microscopy (Figure 2B). Whereas the M9 fusion (shown in red) had strongly accumulated inside the nuclei, transportin (green) was bright at the nuclear envelope but gave only a weak signal inside the nuclei. These results suggest that, although capable of entry into the nucleoplasm, transportin does not remain tightly associated with the M9 import substrate inside the nucleus. Rather, it is likely to be efficiently returned to the cytoplasm, leaving the M9 protein behind. It should be noted that the distribution of transportin in these experiments depended on the amounts added. At a limiting concentration, i.e. when import was dependent on the recycling of transportin back to the cytoplasm, transportin showed the typical nuclear pore staining pattern and only a weak intranuclear signal. At a saturating concentration (2 μM), transportin strongly accumulated inside the nuclei and the nuclear signal then obscured the NPC staining (not shown), and resembled the transportin staining pattern found in intact cells (Fridell et al., 1997). We assume that at high transportin concentration re-export is more readily saturated than entry into the nucleus. mRNA and U snRNA export, but apparently not tRNA export, require GTP hydrolysis by Ran We next wanted to know if Ran also plays a direct role in any of the various processes of export from the nucleus. To address this question it is essential to separate the requirement for export from that for re-import of export mediators. In addition, it is necessary to investigate two possible functions for Ran in a given export pathway. First, the process could rely on GTP hydrolysis by Ran as an energy source. Second, Ran could be required to regulate compartment-specific interactions of the transport factors, as it does for importin-β and transportin. The export of various substrates can be followed by injecting them into nuclei of Xenopus oocytes and analysing their distribution between nucleus and cytoplasm after dissecting the oocytes at various time points. Likewise, nuclear import can be studied following cytoplasmic injection of the substrate of interest. With these assays one can potentially measure the effects of dominant-negative Ran mutants or that of a disturbed asymmetry of the Ran system on a given transport process. The RanQ69L mutant is GTPase deficient (Bischoff et al., 1994; Klebe et al., 1995) and should block processes that require GTP hydrolysis by Ran in a dominant-negative way, as has been shown for NLS-dependent nuclear import (Palacios et al., 1996). We therefore used the sensitivity towards RanQ69L as a criterion to test if DHFR or histone H4 mRNA, U1 or U5 snRNA, or tRNA would require GTP hydrolysis by Ran for their export out of the nucleus. A mixture of these RNAs was injected into nuclei of Xenopus oocytes together with U6 snRNA, which stays in the nucleus and was the internal control for nuclear integrity. Without inhibitor, the mRNAs, U1 and U5 snRNA and tRNA were all efficiently exported to the cytoplasm within the 210 min of incubation (Figure 3, compare lanes 1–3 with 4–6). The co-injection of 160 μM RanQ69L completely blocked export of U snRNA and mRNA, but left tRNA export essentially unaffected (Figure 3, lanes 7–9). The 160 μM in the injected sample should result in a final nuclear concentration of ∼20 μM RanQ69L which is the same order of magnitude one would expect for endogenous wild type Ran in the nucleus. Export of U snRNAs was severely affected by injection of RanQ69L at 40 μM (lanes 13–15) and is thus the most sensitive export process. Nuclear injection of 160 μM RanQ69L does not block NLS-dependent or M9-dependent protein import (not shown). The export inhibition is therefore unlikely to be a secondary consequence of an import defect. These experiments therefore suggest that U snRNA export and mRNA export require GTP hydrolysis by Ran whereas tRNA export may not have such a requirement. The two classes of RNA that are affected are however differentially sensitive to this treatment. Figure 3.Effects of nuclear injection of Ran mutants on RNA export. X.laevis oocyte nuclei were co-injected with a mixture of 32P-labelled RNAs and recombinant RanQ69L or RanT24N proteins as indicated above the lanes. The mixture of RNAs consisted of: DHFR mRNA, histone H4 mRNA, U1ΔSm, U5ΔSm, U6Δss and human initiator methionyl tRNA. The ΔSm U snRNAs lack the Sm binding site required for re-import into the nucleus. U6Δss does not leave the nucleus and is an internal control for nuclear integrity. Synthesis of DHFR, histone H4, U1ΔSm and U5ΔSm RNAs was primed with the m7GpppG cap dinucleotide, whereas synthesis of U6Δss RNA was primed with γ-mGTP. RNA samples from total oocytes (T) or cytoplasmic (C) and nuclear (N) fractions were collected 210 min after injection in lanes 4–24 or immediately after injection in lanes 1–3. RNAs were resolved on 8% acrylamide–7 M urea denaturing gels. The concentration of the recombinant proteins in the injected samples were as indicated above the lanes. Download figure Download PowerPoint It should be noted that RanQ69L only has a dominant-negative effect on NLS- and M9-dependent protein import when injected into the cytoplasm (not shown). There it probably causes premature disassembly of import complexes formed by transportin or importin. That nuclear RanQ69L is not inhibitory for protein import is consistent with termination of import requiring the presence of nuclear RanGTP but not GTP hydrolysis by Ran. The RanT24N mutant inhibits export of mRNA, U snRNA and tRNA from the nucleus Nuclear injection of 20 μM RanT24N blocked U snRNA export completely, and severely reduced the rate of export of the two mRNAs and that of tRNA (Figure 3, lanes 22–24). At 40 μM the block was also essentially complete for mRNA and tRNA export (lanes 19–21). The dominant-negative effect might be explained by the failure of this mutant to bind GTP (Klebe et al., 1995). However, the T24N mutation in Ran also reduces the affinity for GDP ∼500-fold. As a consequence, the complex of (nucleotide-free) RanT24N with the exchange factor RCC1 is neither dissociated by GTP nor by GDP. This makes RanT24N an effective competitive inhibitor of RCC1-mediated nucleotide exchange (Ki = 39 nM) (Klebe et al., 1995). The dominant-negative effect of T24N on mRNA, U snRNA and tRNA export can thus also be explained by a blockage of the RCC1-dependent production of nuclear RanGTP. These results suggest an involvement of Ran in the export of all three categories of RNA. Mislocalization of RanGAP1 to the nucleus blocks RNA export RanGAP1 (Rna1p in yeast) is normally excluded from the nucleoplasm (see Introduction). When injected into nuclei it should result in depletion of RanGTP from this compartment and thus would be expected to affect RNA export similarly to the RanT24N mutant. To test this we used Rna1p, the RanGAP1 from Saccharomyces pombe (Melchior et al., 1993b; Bischoff et al., 1995a) which is a highly efficient activator of Ran from higher eukaryotes. For these experiments S.pombe Rna1p has two further advantages: it is very stable and it can easily be produced in a recombinant form. Figure 4A shows the dose-dependence of export inhibition by nuclear Rna1p. U snRNA export is essentially blocked by nuclear co-injection of as little as 0.5 μM Rna1p in the injection mixture and is thus the most sensitive of the pathways tested. To inhibit tRNA export to a similar extent, ∼10 times more Rna1p had to be injected. Note however that tRNA export occurs at a ∼5-fold higher rate than U snRNA export, and that its inhibition is thus underestimated in Figure 4A. Even the smallest amount of Rna1p reduced the export of DHFR and histone H4 mRNA significantly. However the residual export occurring on injection of 1 μM or more Rna1p was rather resistant to further inhibition. This may indicate that two or more mechanisms with different requirement for RanGTP contribute to mRNA export. Alternatively, some mRNA export mediators might have pre-bound RanGTP which is resistant to Rna1p. Interestingly, not all mRNAs whose export was tested were similarly affected by nuclear injection of Rna1p (see Discussion below). Figure 4.(A) Nuclear injection of Rna1p (RanGAP1) inhibits RNA export. Xenopus oocyte nuclei were co-injected with the mixture of radiolabelled RNAs described in Figure 3A together with various concentrations of recombinant Rna1p as indicated above the lanes. Rna1p stays in the nucleus and its resulting nuclear concentration is ∼1/8 of the concentration in the injection mix. Transport was analysed after 3 h incubation as described in Figure 3A. (B) Effect of nuclear Rna1p on protein import. Xenopus oocyte nuclei were injected with recombinant Rna1p or RanBP1 as indicated above the lanes. One h later, a mixture of in vitro translated 35S-labelled human CBP80 and hnRNP A1 were injected into the oocyte cytoplasm. Proteins were extracted after 4 h incubation in lanes 4–15 or immediately after injection in lanes 1–3. Note the concentration of nuclear Rna1p that blocks RNA export only reduces but does not prevent nuclear