Cofactor requirements for nuclear export of Rev response element (RRE)- and constitutive transport element (CTE)-containing retroviral RNAs. An unexpected role for actin.

Nuclear export of proteins containing leucine-rich nuclear export signals (NESs) is mediated by the export receptor CRM1/exportin1. However, additional protein factors interacting with leucine-rich NESs have been described. Here, we investigate human immunodeficiency virus type 1 (HIV-1) Rev-mediated nuclear export and Mason-Pfizer monkey virus (MPMV) constitutive transport element (CTE)-mediated nuclear export in microinjected Xenopus laevis oocytes. We show that eukaryotic initiation factor 5A (eIF-5A) is essential for Rev and Rev-mediated viral RNA export, but not for nuclear export of CTE RNA. In vitro binding studies demonstrate that eIF-5A is required for efficient interaction of Rev-NES with CRM1/exportin1 and that eIF-5A interacts with the nucleoporins CAN/nup214, nup153, nup98, and nup62. Quite unexpectedly, nuclear actin was also identified as an eIF-5A binding protein. We show that actin is associated with the nucleoplasmic filaments of nuclear pore complexes and is critically involved in export processes. Finally, actin- and energy-dependent nuclear export of HIV-1 Rev is reconstituted by using a novel in vitro egg extract system. In summary, our data provide evidence that actin plays an important functional role in nuclear export not only of retroviral RNAs but also of host proteins such as protein kinase inhibitor (PKI).

Microtubule-dependent assembly of the nuclear envelope in Xenopus laevis egg extract.

Microtubules take part in several mechanisms of intracellular motility, including organelle transport and mitosis. We have studied the ability of Xenopus egg extract to support nuclear membrane and pore complex formation when microtubule dynamics are manipulated. In this report we show that the formation of a nuclear envelope surrounding sperm chromatin requires polymerized microtubules. We have observed that microtubule-depolymerizing reagents, and AS-2, a known inhibitor of the microtubule motor protein kinesin, do not inhibit the formation of a double nuclear membrane. However these double membranes contain no morphologically identifiable nuclear pore complexes and do not support the accumulation of karyophilic proteins. In contrast, the assembly of annulate lamellae, cytoplasmic structures containing a subset of pore complex proteins, was not affected. Our data show that not only polymerized microtubules, but also the microtubule motor protein kinesin, are involved in the formation of the nuclear envelope. These results support the conclusion that multiple nuclear envelope-forming mitotic vesicle populations exist, that microtubules play an essential and selective role in the transport of nuclear envelope-forming vesicle population(s), and that separate mechanisms are involved in nuclear envelope and annulate lamellae formation.

Distribution of emerin during the cell cycle.

Human emerin is a nuclear membrane protein that is lost or altered in patients with Emery-Dreifuss muscular dystrophy (EMD). While the protein is expressed in the majority of human tissues analyzed, the pathology predominates in cardiac and skeletal muscles of patients with EMD. Our results show that emerin can be detected by immunocytochemistry and immunoblotting in the nuclear envelope of all vertebrates studied from man to Xenopus. Immunolocalizations and nuclear envelope extraction experiments confirm that emerin possesses properties characteristic for integral membrane proteins of the inner nuclear membrane. Some nuclear envelope proteins are localized also in annulate lamellae (AL), i.e. cytoplasmic flattened membrane cisternae penetrated by pore complexes. To verify whether emerin is contained in these membrane stacks, we have induced the formation of AL by exposure of rat cells (line RV-SMC) to sublethal doses of the antimitotic drug vinblastine sulfate and found that emerin is present in the nuclear envelope, but is absent from AL. In contrast to the homogeneous distribution of emerin in the nuclear envelope of interphase cells, this protein shows a focal accumulation in the nuclear membranes of late telophase cells. During early reassembly of the nuclear envelope at this mitotic stage emerin colocalizes with lamin A/C but not with lamin B and LAP2 proteins. Confocal laser scanning microscopy after double-labeling experiments with emerin and tubulin shows that emerin is concentrated in areas of the mitotic spindle and in the midbody of mitotic cells suggesting a close interaction of these proteins. Our data suggest that emerin participates in the reorganisation of the nuclear envelope at the end of mitosis.

Cellular uptake and nuclear delivery of recombinant adenovirus penton base.

An Ad2 capsid component, the penton base, expressed as recombinant protein, was found to be capable of affecting the entire entry pathway of adenovirion in HeLa cells, i.e., cell attachment, endocytosis, vesicular escape, intracytoplasmic movement, and translocation through the nuclear pore complex. Data with pentamerization-defective mutants suggested that none of these successive steps depended upon penton base pentamer status, indicating that the peptide domains responsible for these functions were carried by the monomer. Observations performed with wild-type (WT) and an integrin-binding-site double-mutant (K288E340) suggested that the penton base could enter the cell via an alternative, RGD- and LDV-independent, pathway. Of three mutants that were found to be defective in nuclear addressing in insect cells, only one, W165H, was also altered in nuclear transport in HeLa cells. The other two, W119H and RRR547EQQ, showed a WT pattern of nuclear localization in HeLa cells, suggesting that the region including tryptophan-119 and the basic signal at position 547 did not act as a nuclear localization signal in the human cell context. The integrity of cellular structures and the cytoskeleton seemed to be required for the vectorial movement and nuclear import of WT penton base, as suggested by experiments using permeabilized HeLa cells, isolated nuclear membranes, and cytoskeleton-targeted drugs.

Nuclear pore localization and nucleocytoplasmic transport of eIF-5A: evidence for direct interaction with the export receptor CRM1.

Eukaryotic initiation factor 5A (eIF-5A) is the only cellular protein known to contain the unusual amino acid hypusine. The exact in vivo function of eIF-5A, however, is to date unknown. The finding that eIF-5A is an essential cofactor of the human immunodeficiency virus type 1 (HIV-1) Rev RNA transport factor suggested that eIF-5A is part of a specific nuclear export pathway. In this study we used indirect immunofluorescence and immunogold electron microscopy to demonstrate that eIF-5A accumulates at nuclear pore-associated intranuclear filaments in mammalian cells and Xenopus oocytes. We are able to show that eIF-5A interacts with the general nuclear export receptor, CRM1. Furthermore, microinjection studies in somatic cells revealed that eIF-5A is transported from the nucleus to the cytoplasm, and that this nuclear export is blocked by leptomycin B. Our data demonstrate that eIF-5A is a nucleocytoplasmic shuttle protein.

Preassembly of annulate lamellae in egg extracts inhibits nuclear pore complex formation, but not nuclear membrane assembly.

Annulate lamellae (AL) are cytoplasmic structures containing pore complexes similar in composition and morphology to nuclear envelope pore complexes. We have tested the ability of Xenopus egg extract to support nuclear membrane and pore complex formation when chromatin is added to extract only after annulate lamellae had been allowed to assemble (preincubated extract). We have observed that preassembly of AL does not inhibit the formation of a double membrane surrounding sperm chromatin. However, these double membranes are often distended, do not support accumulation of karyophilic proteins, and do not possess immunologically or morphologically identifiable pore complexes. We have demonstrated that nuclear pore complex assembly and function can be rescued by adding isolated egg vesicles to the preincubated extract. Our data support the conclusion that multiple vesicle populations are utilized in the formation of a nuclear envelope, including: vesicle population(s) which are common to pore formation in nuclear envelopes and annulate lamellae, and vesicle population(s) which seem to be preferentially utilized for nuclear membrane assembly.

A biochemical and immunological comparison of nuclear and cytoplasmic pore complexes.

Pore complexes are not confined to the nuclear envelope but can also be found in the cytoplasm of numerous cell types in the form of annulate lamellae (AL). We have induced formation of AL by exposure of rat cells (line RV) to sublethal doses of the antimitotic drug vinblastine sulfate, and compared the distribution of several nuclear pore complex proteins (nucleoporins) in the nuclear envelope and AL by immunocytochemistry, cytochemical lectin binding studies and immunoblot analyses of nuclear and AL-enriched fractions. All the antibodies used yielded punctate nuclear surface staining in immunofluorescence microscopy which is characteristic for nuclear pore complex components. When we applied antibodies against the nucleoporin p62, AL were visualized as numerous cytoplasmic dot-like structures. Immunogold electron microscopy confirmed the correspondence of the cytoplasmic bodies with stacks of AL. Antibodies to constituents of the cytoplasmic (nup180) and nucleoplasmic (nup153) filaments extending from both sides of nuclear pore complexes also stained the AL, indicating that pore complexes are intrinsically asymmetric assemblies independent of their specific intracellular topology. By contrast, AL were negative with five different antibodies against the transmembrane nuclear pore glycoprotein gp210 and the lectin concanavalin A (ConA) known to bind to the oligosaccharide side chains of gp210. Similarly, there was no staining of the AL with antibodies to the other nuclear pore membrane protein so far known in higher eukaryotes, POM121. Immunoblot analyses confirmed the presence of p62, nup180 and nup153 in both the nuclear and AL fractions and the absence of gp210 and POM121 from AL. Our results do not support the generally held view that gp210 and POM121 function in anchoring the pore complex scaffold to the pore membrane. Rather, they point to a role for these proteins in transport processes through the nuclear pore complexes. Since AL are not involved in nucleocytoplasmic transport processes they may lack components of the transport machinery.

Localization of the Ran-GTP binding protein RanBP2 at the cytoplasmic side of the nuclear pore complex.

A partial cDNA clone coding for the mouse homologue of the human Ran-GTP binding protein, RanBP2, has been isolated by screening of a murine expression library with antibodies to nup180, a previously identified nuclear pore complex protein (nucleoporin). Whether the antibodies cross-reacted with the polypeptide encoded by the cDNA clone or, alternatively, nup180 is proteolytically related to RanBP2, has not been determined. The 3795-bp open reading frame of the cDNA encodes a polypeptide consisting of 1265 amino acids with three Ran-GTP binding domains (RanBD) that are almost identical with published partial amino acid sequences of human RanBP2 as deduced from several partial cDNA clones of other authors. Sequence analysis further revealed that murine RanBP2 contains tandemly repeated zinc fingers of Cys2-Cys2 type and multiple copies of the FXFG nucleoporin "signature" motif clustered in regions preceding the RanBDs. Antibodies raised against a synthetic peptide of the derived amino acid sequence decorated the cytoplasmic rings of nuclear pore complexes (NPCs) as shown by immunogold electron microscopy. We suggest that the cytoplasmically disposed nucleoporin RanBP2 provides docking sites for import substrate-receptor complexes and, further, that the affinity of these sites to the transport substrate is modulated in a Ran-dependent fashion.

Nup180, a novel nuclear pore complex protein localizing to the cytoplasmic ring and associated fibrils.

Using an autoimmune serum from a patient with overlap connective tissue disease we have identified by biochemical and immunocytochemical approaches an evolutionarily conserved nuclear pore complex (NPC) protein with an estimated molecular mass of 180 kD and an isoelectric point of approximately 6.2 which we have designated as nup180. Extraction of isolated nuclear envelopes with 2 M urea and chromatography of the solubilized proteins on WGA-Sepharose demonstrated that nup180 is a peripheral membrane protein and does not react with WGA. Affinity-purified antibodies yielded a punctate immunofluorescent pattern of the nuclear surface of mammalian cells and stained brightly the nuclear envelope of cryosectioned Xenopus oocytes. Nuclei reconstituted in vitro in Xenopus egg extract were also stained in the characteristic punctate fashion. Immunogold EM localized nup180 exclusively to the cytoplasmic ring of NPCs and short fibers emanating therefrom into the cytoplasm. Antibodies to nup180 did not inhibit nuclear protein transport in vivo nor in vitro. Despite the apparent lack of involvement in NPC assembly or nucleocytoplasmic transport processes, the conservation of nup180 across species and its exclusive association with the NPC cytoplasmic ring suggests an important, though currently undefined function for this novel NPC protein.

In vitro assembly of prenucleolar bodies in Xenopus egg extract.

Nuclei assembled in Xenopus egg extract from purified DNA or chromatin resemble their natural counterparts in a number of structural and functional features. However, the most obvious structural element of normal interphase nuclei, the nucleolus, is absent from the in vitro reconstituted nuclei. By EM, cytological silver staining, and immunofluorescence microscopy employing antibodies directed against various nucleolar components we show that nuclei assembled in vitro contain numerous distinct aggregates that resemble prenucleolar bodies (PNBs) by several criteria. Formation of these PNB-like structures requires pore complex-mediated nuclear transport of proteins but is independent of the genetic content of the in vitro nuclei as well as transcriptional and translational events. Our data indicate that nuclei assembled in vitro are capable of initiating early steps of nucleologenesis but that the resulting PNBs are unable to fuse with each other, probably due to the absence of a functional nucleolus organizer. With appropriate modifications, this experimental system should be useful to define and analyze conditions promoting the site-specific assembly of PNBs into a coherent nucleolar body.

Spontaneous assembly of pore complex-containing membranes ("annulate lamellae") in Xenopus egg extract in the absence of chromatin.

Extract prepared from activated Xenopus eggs is capable of reconstituting nuclei from added DNA or chromatin. We have incubated such extract in the absence of DNA and found that numerous flattened membrane cisternae containing densely spaced pore complexes (annulate lamellae) formed de novo. By electron and immunofluorescence microscopy employing a pore complex-specific antibody we followed their appearance in the extract. Annulate lamellae were first detectable at a 30-min incubation in the form of short cisternae which already contained a high pore density. At 90-120 min they were abundantly present and formed large multilamellar stacks. The kinetics of annulate lamellae assembly were identical to that of nuclear envelope formation after addition of DNA to the extract. However, in the presence of DNA or chromatin, i.e., under conditions promoting the assembly of nuclear envelopes, annulate lamellae formation was considerably reduced and, at sufficiently high chromatin concentrations, completely inhibited. Incubation of the extract with antibodies to lamin LIII did not interfere with annulate lamellae assembly, whereas in the presence of DNA formation of nuclear envelopes around chromatin was inhibited. Our data show that nuclear membrane vesicles are able to fuse spontaneously into membrane cisternae and to assemble pore complexes independently of interactions with chromatin and a lamina. We propose that nuclear envelope precursor material will assemble into a nuclear envelope when chromatin is available for binding the membrane vesicles, and into annulate lamellae when chromatin is absent or its binding sites are saturated.

Expression of the neuronal noradrenaline transporter in Xenopus laevis oocytes.

RNA isolated from both rat phaeochromocytoma (PC12) cells and bovine adrenal medulla was size-selected and injected into oocytes of Xenopus laevis. Expression of the neuronal noradrenaline transporter was assayed 3 days after injection by measuring the nisoxetine-sensitive noradrenaline uptake. A 2 kilobase fraction of either total RNA derived from both tissues or of polyadenylated mRNA obtained from bovine adrenal medulla caused Na(+)-dependent transport of 3H-noradrenaline in the injected oocytes.

Identification of a soluble precursor complex essential for nuclear pore assembly in vitro.

We analysed the soluble form in which the nuclear pore complex protein p68 is stored in Xenopus laevis eggs and its involvement in pore complex assembly processes. We have shown previously that p68, which is the major wheat germ agglutinin (WGA)-binding glycoprotein of nuclear pore complexes from Xenopus oocytes, is located in the pore channel and participates in mediated transport of karyophilic proteins. Using a monoclonal antibody directed against p68 (PI1) we removed this protein from Xenopus egg extract by immunoadsorption. On addition of lambda DNA the immuno-depleted extract supported reconstitution of nuclei which were surrounded by a continuous double-membrane envelope but lacked pore complexes and were unable to import karyophilic proteins such as nucleoplasmin or lamin LIII. Essentially identical results were obtained with extract depleted of WGA-binding proteins. Our finding that both the anti-p68 antibody and WGA efficiently removed components from the extract necessary for pore complex assembly but did not interfere with nuclear membrane formation demonstrates that these processes are independent of each other. Analysis of the immunoprecipitate on silver-stained SDS-polyacrylamide gels indicated that the antibody adsorbed other proteins besides p68, notably two high molecular weight components. By sucrose gradient centrifugation and gel filtration we showed that p68 together with associated protein(s) forms a stable, approximately globular complex with an Mr of 254,000, a Stokes radius of 5.2 nm and a sedimentation coefficient of 11.3 S. Our finding that p68 occurs in the form of larger macromolecular assemblies offers an explanation for the distinctly punctate immunofluorescence pattern observed in the cytoplasm of mitotic cells after staining with antibodies to p68.

Parthenogenesis in Xenopus eggs requires centrosomal integrity.

Xenopus eggs are laid arrested at second metaphase of meiosis lacking a functional centrosome. Upon fertilization, the sperm provides the active centrosome that is required for cleavage to occur. The injection of purified centrosomes mimics fertilization and leads to tadpole formation (parthenogenesis). In this work we show that the parthenogenetic activity of centrosomes is inactivated by urea concentrations higher than 2 M. The loss of activity is correlated with a progressive destruction of the centriolar cylinder and extraction of proteins. This shows that centrosomes are relatively sensitive to urea since complete protein unfolding and solubilization of proteins normally occurs at urea concentrations as high as 8-10 M. When present, the parthenogenetic activity is always associated with a pelletable fraction showing that it cannot be solubilized by urea. The parthenogenetic activity is progressively inactivated by salt concentrations higher than 2 M (NaCl or KCl). However, only a few proteins are extracted by these treatments and the centrosome ultrastructure is not affected. This shows that both parthenogenetic activity and centrosomal structure are resistant to relatively high ionic strength. Indeed, most protein structures held by electrostatic forces are dissociated by 2 M salt. The loss of parthenogenetic activity produced at higher salt concentrations, while the structure of the centrosome is unaffected, is an apparent paradox. We interpret this result as meaning that the native state of centrosomes is held together by forces that favor functional denaturation by high ionic strength. The respective effects of urea and salts on centrosomal structure and activity suggest that the centrosome is mainly held together by hydrogen and hydrophobic bonds. The in vitro microtubule nucleating activity of centrosomes can be inactivated at salt or urea concentrations that do not affect the parthenogenetic activity. Since egg cleavage requires the formation of microtubule asters, we conclude that the extracted or denatured microtubule nucleating activity of centrosomes can be complemented by components present in the egg cytoplasm. Both parthenogenetic and microtubule nucleating activities are abolished by protease treatments but resist nuclease action. Since we find no RNA in centrosomes treated by RNase, they probably do not contain a protected RNA. Taken together, these results are consistent with the idea that the whole or part of the centrosome structure acts as a seed to start the centrosome duplication cycle in Xenopus eggs.

Functional role of newly formed pore complexes in postmitotic nuclear reorganization.

Many nuclear proteins are released into the cytoplasm at prometaphase and are transported back into the daughter nuclei at the end of mitosis. To determine the role of this reentry in nuclear remodelling during early interphase, we experimentally manipulated nuclear protein uptake in dividing cells. Recently we and others have shown that signal-dependent, pore complex-mediated uptake of nuclear protein is blocked in living cells on microinjection of the lectin wheat germ agglutinin (WGA), or of antibodies such as PI1 that are directed against WGA-binding pore complex glycoproteins. In the present study, we microinjected mitotic PtK2 cells with WGA or antibody PI1 and followed nuclear reorganization of the daughter cells by immunofluorescence and electron microscopy. The inhibitory effect on nuclear protein uptake was monitored by co-injection of the karyophilic protein nucleoplasmin. When injected by itself early in mitosis, nucleoplasmin became sequestered into the daughter nuclei as they entered telophase. In contrast, nucleoplasmin was excluded from the daughter nuclei in the presence of WGA or antibody PI1. Although PtK2 cells with blocked nuclear protein uptake completed cytokinesis, their nuclei showed a telophase-like completed cytokinesis, their nuclei showed a telophase-like organization characterized by highly condensed chromatin surrounded by a nuclear envelope containing a few pore complexes. These findings suggest that pore complexes become functional as early as telophase, in close coincidence with nuclear envelope reformation. They further indicate that the extensive structural rearrangement of the nucleus during the telophase-G1 transition is dependent on the influx of karyophilic proteins from the cytoplasm through the pore complexes, and is not due solely to chromosome-associated components.

Monoclonal antibodies to a Mr 68,000 pore complex glycoprotein interfere with nuclear protein uptake in Xenopus oocytes.

Using a monoclonal antibody (PI1) raised against mouse lymphocyte nuclear matrix fractions we have identified a N-acetylglucosamine (GlcNAc)-containing glycoprotein of Mr 68,000 as a component of the nuclear pore complexes of Xenopus laevis oocytes. The antigenic determinant recognized by antibody PI1 comprises both the sugar moiety and protein sequences since, on the one hand, added GlcNAc competed effectively for antibody binding and, on the other hand, the antibody reacted in immunoblots with only one member of the GlcNAc-containing pore complex glycoprotein family. By using immunogold-electron microscopy we could demonstrate that the Mr 68,000 glycoprotein was located preferentially to the cytoplasmic side of the pore complex channel. When radiolabeled soluble nuclear proteins were injected into the cytoplasm of Xenopus oocytes, their reentry into the nucleus was almost completely inhibited in the presence of antibody PI1 as shown by two-dimensional gel electrophoresis. The results indicate that the evolutionarily conserved Mr 68,000 glycoprotein is involved in transport processes of karyophilic proteins from the cytoplasm into the nucleus.

Role of nuclear material in the early cell cycle of Xenopus embryos.

Activated Xenopus eggs show periodic surface contraction waves and oscillations in endogenous protein phosphorylation, MPF, and kinase activities timed with the cleavage cycle of control fertilized eggs. In this paper, we show that in activated eggs lacking the material that originates from the oocyte nucleus, MPF and kinase oscillations occur in the absence of surface contraction waves. Two mitotic phosphoproteins (M116 and M46), previously described by 32P labeling in nucleated eggs, are no longer detected in the enucleated eggs. We conclude that a cytoplasmic temporal control of MPF and kinase activities is likely to be the essential cell cycle oscillator. The oocyte nuclear components normally stored in the cytoplasm of the embryos are not involved in the clock although they appear to be required for the generation of surface contraction waves.