Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
1055 Journal of Cell Science 112, 1055-1064 (1999) Printed in Great Britain © The Company of Biologists Limited 1999 JCS0230 Pre-mRNA and mRNA binding of human nuclear DNA helicase II (RNA helicase A) Suisheng Zhang1, Christine Herrmann2 and Frank Grosse1,* 1Department of Biochemistry and 2Department of Electron Microscopy and Molecular Cytology, Institute for Molecular Biotechnology, Beutenbergstrasse 11, D-07745 Jena, Germany *Author for correspondence (e-mail: [email protected]) Accepted 26 January; published on WWW 10 March 1999 SUMMARY Nuclear DNA helicase II (NDH II), alternatively named RNA helicase A, seems to function as a pre-mRNA and mRNA binding protein in human cells. Immunofluorescence studies of NDH II gave a highly diffused nucleoplasmic staining that was similar to that of hnRNP A1 but differed from the localization of the RNA splicing factor Sc-35. Upon transcriptional inhibition, NDH II migrated from the nucleus into the cytoplasm. During mitosis, NDH II was released into the cytoplasm during pro- to metaphase, and was gradually recruited back into telophase nuclei. The timing of nuclear import of NDH II at telophase was found to be later than that of hnRNP A1 but paralleled that of Sc35. At the ultrastructural level, both NDH II and hnRNP A1 were identified within perichromatin ribonucleoparticle fibrils. However, the subnuclear distributions of NDH II and hnRNP A1 were not overlapping. NDH II could be extracted together with poly(A)-containing mRNA from HeLa cell INTRODUCTION Nuclear DNA helicase II (NDH II) was originally isolated from calf thymus as an unwinding enzyme thought to be involved in DNA metabolism (Zhang and Grosse, 1991). Subsequently it was shown that NDH II is able to unwind both DNA and RNA, which brought about the possibility that this enzyme may also be involved in RNA synthesis (Zhang and Grosse, 1994). A function in RNA rather than DNA metabolism was further strengthened by identifying the cDNA sequence of bovine NDH II (Zhang et al., 1995). The sequence revealed that NDH II is a close homologue of human RNA helicase A (RHA) (Lee and Hurwitz, 1992) and the Drosophila ‘maleless’ protein (MLE) (Kuroda et al., 1991), displaying 92% and 80% amino acid similarities, respectively. Only recently it was demonstrated that RHA and MLE also unwind both RNA and DNA, providing evidence that this group of enzymes shares identical biochemical properties (Lee et al., 1997). The three NDH II homologues catalyze nucleic acid unwinding that is fueled by the hydrolysis of any of the four NTPs or dNTPs; the KM value for ATP is only ≈10 µM. All these enzymes need a 3′-single stranded tail as entry site for nucleic acid displacement and migrate from 3′ to 5′ with nuclei and, to a much lesser extent, from the cytoplasm. Following transcriptional inhibition, NDH II was preferentially associated with mRNA from the cytosol, which biochemically confirmed the microscopic observations. Although NDH II is mainly a nuclear enzyme, it is apparently not associated with the nuclear matrix, since it could be extracted with 2 M NaCl from DNase I-treated nuclei. Our cellular and biochemical observations strongly suggest that NDH II is a pre-mRNA and mRNA binding protein. Its significant affinity for ssDNA, but not for dsDNA, points to a transient role in DNA binding during the process of transcript formation. According to our model, singlestranded DNA might be necessary to retain NDH II in the nuclear compartment. Key words: DEXH family, Immunoelectron microscopy, Immunofluorescence, Mitosis, MLE, hnRNP A1, poly(A) RNA, Sc-35 respect to the single-strand they have bound to (Lee et al., 1997; Lee and Hurwitz, 1992; Zhang and Grosse, 1991). The Drosophila MLE protein is involved in sex-specific gene dosage compensation by increasing the transcriptional level of the single X-chromosome of males to reach that of the two female Xchromosomes (Kuroda et al., 1991; Lee et al., 1997). This is obviously achieved by binding of MLE to the only X chromosome of males, whereas it does not decorate the two Xchromosomes of female flies. Specific labeling of the single X chromosome could be destroyed by RNase treatment, pointing to a preferred binding of RNA structures (Richter et al., 1996). Moreover, recently it has been shown that RHA/NDH II binds to the constitutive cytoplasmic transport element of some retroviral genomic RNAs, indicating both recognition of RNA structures and a putative function in nucleocytoplasmic transport (Tang et al., 1997). Also, RHA/NDH II was shown to be required as ‘bridging element’ for the attraction of the transcriptional coactivator CBP/p300 to the cellular RNA polymerase II complex (Nakajima et al., 1997). All these findings strongly suggest a role of NDH II and its homologues in RNA rather than DNA metabolism, despite the fact that NDH II was originally identified as DNA unwinding enzyme. 1056 S. Zhang, C. Herrmann and F. Grosse To further characterize the cellular function of NDH II we analyzed its intracellular distribution by immunofluorescence and immunoelectron microscopy. Our immunofluorescence studies revealed a similar cellular distribution pattern as the mRNA packaging factor hnRNP A1 and a different localization than the RNA splicing factor Sc-35. Observations by immunoelectron microscopy revealed a largely nonoverlapping association of NDH II and hnRNP A1 with perichromatin RNP fibrils. Biochemically, we could show that NDH II copurifies with poly(A)-containing RNAs. Nevertheless, purified NDH II binds equally well to singlestranded RNA (ssRNA), double-stranded RNA (dsRNA), and with only a slightly decreased affinity to single-stranded DNA (ssDNA), while double-stranded DNA (dsDNA) is obviously not a substrate. From the combined cell biological and biochemical data we suggest a model, in which NDH II initially binds to ssDNA provided by the transcriptional complex, then moves to the nascent transcript, unwinds part of it to allow splicing and transport, and then returns to singlestranded parts of chromatin DNA in order to start a new cycle of pre-mRNA association. MATERIALS AND METHODS Immunofluorescence HeLa cells, grown on coverslips, were rinsed with PBS (10 mM sodium phosphate, pH 7.4, 140 mM NaCl, 3 mM KCl), fixed in 4% paraformaldehyde in PBS for 15 minutes and then permeabilized in 0.5% Triton X-100 in PBS for another 15 minutes. To reduce the nonspecific background, cells were blocked by incubation with 5% bovine serum albumin (BSA) in PBS for 15 minutes. After washing with PBS, primary antibodies in PBS containing 0.5% BSA were added with different titers, as indicated in the figure legends. Incubation with the antigen-specific antibody was for 1 hour, followed by a PBS wash and further incubation for 1 hour with Cy3-labelled anti-rabbit IgG (Amersham, Braunschweig, Germany) at a dilution of 1 to 400 or antimouse IgG conjugated with fluorescein isothiocyanate (FITC) (Boehringer-Mannheim, Mannheim, Germany) at a dilution of 1 to 60. Immunofluorescence was viewed with a Zeiss fluorescence microscope and photographed under 100-fold magnification. Doubleimmunofluorescent labeling was achieved by co-incubation of cells with rabbit and mouse primary antibodies, which were recognized by Cy3- or FITC-coupled secondary antibodies against rabbit or mouse IgG. Electron microscopy HeLa cells were fixed in 4% paraformaldehyde in 0.1 M Sörensen phosphate buffer, pH 7.3, at 4°C for 6 hours. After washing in 0.1 M Sörensen buffer for three times, the cell pellet was dehydrated in ethanol with gradually increasing concentrations (from 30% to 75%) and then embedded in LR White resin (Plano, Cardiff, UK). Ultrathin sections were treated with 5% BSA in PBS for 1 hour and incubated for 6 hours with rabbit and mouse primary antibodies at the same dilutions described for immunofluorescence, followed by an overnight incubation with gold-conjugated secondary antibodies against rabbit IgG and mouse IgG at dilutions of 1 to 100 (Plano). After immunolabeling, samples were proceeded for the EDTA regressive method developed for electron microscopy observations of RNA-containing structures (Bernhard, 1969). Purification of recombinant human NDH II Human NDH II was overexpressed in baculovirus-infected Sf9 cells and purified by Ni2+-NTA-agarose chromatography and subsequent chromatography on poly(rI•rC) agarose, exactly as described (Zhang and Grosse, 1997). Co-purification of NDH II with poly(A)-containing RNA from HeLa cells HeLa cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM, C.C.Pro GmbH, Neustadt/W., Germany) supplemented with 10% fetal bovine serum (Gibco-BRL, Karlsruhe, Germany). To examine binding of NDH II to cellular RNA substrates, HeLa cells (≈4×107) were collected, washed with PBS and suspended in 500 µl of RSB-100 buffer (10 mM NaCl, 1.5 mM MgCl2, 10 mM Tris-HCl, pH 7.2, 0.5% Triton X-100, 0.1% Trasylol (Bayer, Leverkusen, Germany) and 1 mM phenylmethanesulfonyl fluoride (PMSF)). The cells were disrupted by passing the suspension through a 25-gauge needle for 4 times, followed by centrifugation at 3000 g at 4°C for 5 minutes. The cytosolic supernatant fraction was mixed with an equal volume of 2× oligo(dT) column loading buffer (1×, 20 mM Tris-HCl, pH 7.8, 0.5 M NaCl, 1 mM EDTA, 0.1% SDS and 0.1 M 2-mercaptoethanol). Separately, the nuclear pellet was suspended in 500 µl of cold RSB-100 buffer, sonicated and centrifuged at 12,000 g at 4°C for 15 minutes. The supernatant of the nuclear extract was mixed with an equal volume of 2× oligo(dT) column loading buffer. Both cytosolic and nuclear fractions were passed through oligo(dT)-cellulose columns (200 µl), which have been pre-treated by 0.1 N NaOH followed by extensive washing in water until the pH was 7.0 and then equilibrated with 10 volumes of 1× oligo-dT column loading buffer. After loading, the columns were washed with 20 volumes of 1× oligo(dT) column-loading buffer. The columns were eluted with 500 µl of 10 mM Tris-HCl, pH 7.6, 1 mM EDTA and 0.05% SDS. The eluted fractions were incubated with 0.1 mg/ml RNase A at room temperature for 10 minutes, followed by protein precipitation in 10% trichloroacetic acid. After centrifugation the protein precipitates were dissolved in 50 µl of SDS-PAGE loading buffer. One-twentieth volume of each sample was loaded for western-blot analysis. HeLa cell fractions and preparation of the nuclear matrix HeLa cells (≈4×107) were collected and washed with cold PBS. Cells were disrupted with 5 strokes of a Teflon homogenizer in 1 ml buffer A (50 mM Tris-HCl, pH 7.8, 5 mM KCl, 5 mM MgCl2, 250 mM sucrose, 10 mM Na2S2O5, 7 mM β-mercaptoethanol, 1 mM PMSF) followed by centrifugation at 600 g at 4°C for 10 minutes. The supernatants were designated as cytosolic fraction. The pellets were suspended in 1 ml of buffer A containing 150 mM KCl and purified by centrifugation at 2000 g at 4°C for 10 minutes through a cushion of buffer A containing 150 mM KCl and 0.7 M sucrose. The purified nuclei were suspended in 500 µl of buffer B (20 mM Tris-HCl, pH 7.4, 5 mM MgCl2, 250 mM sucrose, 1 mM PMSF) and DNase I was added to 1/10 of the amount of DNA. After incubation at room temperature for 30 minutes, the DNase I-digested nuclei suspension was centrifuged at 1000 g for 10 minutes. Then the pellet was extracted with 2 M NaCl in buffer C (100 mM Tris-HCl, pH 7.4, 0.1 mM MgCl2, 290 mM sucrose) for 10 minutes on ice, followed by centrifugation at 15,000 g at 4°C for 10 minutes. High-salt extraction of nuclear pellets was repeated for three times and samples were prepared from the supernatant and pellet of each extraction for western-blot analysis. Filter binding assays Filter binding assays were performed as described before (Grosse et al., 1986). Purified NDH II was incubated at different concentrations with 32P-labeled RNA and DNA substrates in 10 µl binding buffer containing 20 mM Hepes, pH 8.0, 50 mM NaCl, 1 mM DTT, and 1 mM EDTA. After incubation at room temperature for 20 minutes, the samples were loaded to a nitrocellulose membrane (Protran B85, 0.45 µM, Schleicher and Schuell, Dassel, Germany) in a 96-well vacuum blotter. After washing for three times by filling up the slots with ≈500 µl binding buffer (each), the filter was removed, dried, and cut out at the positions of sample loading. The retained radioactivity was quantified by scintillation counting. In order to decrease unspecific binding, the nitrocellulose membrane was treated with 0.3 M NaOH for 10 minutes, washed two times for 5 minutes each with distilled water, and then equilibrated in binding buffer for at least 16 hours. Intracellular localization of NDH II 1057 Gel mobility-shift assays RNA probes were transcribed by T3 and T7 RNA polymerases (Boehringer-Mannheim) starting at the corresponding promoters of the Bluescript KS+ plasmid (Stratagene, Amsterdam, The Netherlands). An 85 bp long ssRNA was synthesized from the T3 promotor of this plasmid after linearization with BamHI. The transcript was annealed to the complementary strand that was synthesized from the T7 promotor of the Pvu II-linearized plasmid. Single-stranded protrusions were cut-off by treating the hybrid with RNases A and T1. The RNA probe was labeled by including [α-32P]GTP into the transcription mixture. Purified NDH II was incubated with ≈60 µM (nt) labeled RNA in 10 µl binding buffer (see above). After incubation at room temperature for 25 minutes, each sample was mixed with 3 µl loading buffer containing 40% sucrose, 0.25% Bromophenol Blue, and 0.25% Xylene cyanol FF, and then electrophoresed through a 6% native polyacrylamide gel in 45 mM boric acid, 45 mM Tris base, 1 mM EDTA at 100-150 V for 50 minutes. After electrophoresis the gel was exposed to an X-ray film at –70°C for 16-24 hours. Similarly, binding of NDH II to M13 ssDNA was measured by incubating various concentrations of the purified enzyme with 6 µM (nt) M13mp18 DNA, primed with the 32P-labeled 45-mer oligodeoxynucleotide containing a 3′-tail (Zhang and Grosse, 1997). The samples were electrophoresed through 1% agarose in 90 mM boric acid, 90 mM Tris base, 2 mM EDTA at 2 V/cm for 24 hours. RESULTS Intracellular localization of NDH II The intracellular localization of NDH II was investigated by double-immunofluorescence of HeLa cells with rabbit antiserum against NDH II and mouse monoclonal antibodies against hnRNP A1 or Sc-35. The latter ones were used as reference markers for the intracellular distribution of NDH II under various experimental conditions. In HeLa cells NDH II was mainly localized in the nuclei, although very weak cytoplasmic signals could also be seen (Fig. 1A,C). The nuclear distribution of NDH II could be observed as a widespread homogeneous staining within the nucleoplasmic region that excluded the nucleoli. Similar staining patterns were also found for hnRNP A1 (Fig. 1A′). hnRNP A1 undergoes relocalization into the cytoplasm when transcription is inhibited (Piñol-Roma and Dreyfuss, 1992). A similar effect was also observed for NDH II after treatment of cells with actinomycin D (Fig. 1B). However, following transcriptional inhibition the increased cytoplasmic signal of NDH II was less apparent than that of hnRNP A1 (Fig. 1B′). This might either reflect different levels of abundance of these two proteins in HeLa cells, or, different mechanisms regulating their cellular localization. In contrast to hnRNPs, snRNPs and non-snRNP RNA splicing factors are concentrated in nuclear speckles, which occupy distinct portions of the nucleoplasmic region (Spector et al., 1991). Some DEXD/H helicases involved in RNA splicing colocalize with Sc-35 in nuclear speckles (Gee et al., 1997; Molnar et al., 1997; Ono et al., 1994; Sukegawa and Blobel, 1995). Sc-35 speckles underwent dynamic changes that correlated with the transcriptional activity of the cell. In transcriptionally active cells Sc-35 was dispersed throughout the nucleoplasmic regions. But following transcriptional inhibition Sc-35 assembled into speckles (Fig. 1D′). The nucleoplasmic immunostaining of NDH II did not show the same pattern as the nucleoplasmic distribution of Sc-35 (Fig. 1C,C′). Moreover, in HeLa cells treated with the transcription inhibitor 5,6-dichloro-1-β-D-ribofuranosylbenzimidazole (DRB), NDH II concentrated at a few faintly visible nuclear foci, whose positions did not overlap with Sc-35 (Fig. 1D). Redistribution of NDH II in mitotic HeLa cells During mitosis most of the proteins with functions in transcription and RNA processing are released from the chromosomes and transcription activities become suppressed or ceased for the period of chromosomal condensation. NDH II was excluded from the mitotic nucleus as early as prophase when the chromosomes started condensing and the nuclear envelope disrupted (Fig. 2A,B). The exclusion reached a maximum at metaphase when the two sets of sister chromatids were highly condensed and aligned at the metaphase plate (Fig. 2C,D). At telophase, NDH II re-entered the nucleus at a stage when the chromosomes in the dividing cells started to decondense and the nuclear envelope was gradually reformed (Fig. 2E,G, and H,J). For comparison, we observed the redistribution of hnRNP A1 and Sc-35 by double-immunolabeling in mitotic HeLa cells. hnRNP A1 and Sc-35 were similarly excluded from prophase and metaphase chromosomes as NDH II (data not shown). However, hnRNP A1 was recruited earlier than NDH II into the telophase nucleus (Fig. 2E,F), while the re-entering of Sc-35 into telophase nuclei paralleled that of NDH II (Fig. 2H,I). Further differences between the nuclear import of hnRNP A1, NDH II, and Sc-35 were found by subjecting mitotic cells to transcriptional inhibition. The nuclear import of hnRNP A1 at telophase was sensitively blocked by actinomycin D (Fig. 3B) or DRB (data not shown). A similar effect was observed earlier on the transcription-dependent nuclear transport of other hnRNP proteins (Piñol-Roma and Dreyfuss, 1991). However, we found that actinomycin D did not block the re-import of NDH II into telophase nuclei which otherwise excluded hnRNP A1 (Fig. 3A). Neither was Sc-35 subjected to a similar inhibition of nuclear import as hnRNP A1. Sc-35 re-entered the telophase nucleus together with NDH II irrespectively of a treatment with actinomycin D (Fig. 3D,E). DAPI stainings (see Fig. 2B,D,G,J and Fig. 3C,F) are shown to demonstrate the state of chromosomal condensation and cell division. Association of NDH II with perichromatin RNP fibrils The subnuclear localization of NDH II in HeLa cells was studied by immunoelectron microscopy. An EDTA-regressive staining method was used to contrast RNP-structures against condensed chromosomal areas in the nucleus (Bernhard, 1969). NDH II, identified by 10 nm gold-particles, was found to be associated with RNP-structures at the chromosomal periphery in the nucleoplasm (Fig. 4A,B) that have been designated as perichromatin fibrils. Some NDH II molecules showed up in cytoplasmic RNP structures, implicating possible functions of NDH II in RNA export or translation. We also studied the distribution of hnRNP A1 in HeLa cells, which was detected with 5 nm gold-particles. We found that hnRNP A1 was also associated with perichromatin fibrils (Fig. 4C). Some spherical RNP particles were associated with a high density of hnRNP A1 (indicated by arrows in Fig. 4C). An association of hnRNP A1 with RNP particle or fibrils was also observed in the cytoplasm (Fig. 4C). Although NDH II and hnRNP A1 displayed a similar distribution pattern on perichromatin RNP fibrils, which might indicate an association with nascent hnRNA transcripts, the localization of differently sized gold 1058 S. Zhang, C. Herrmann and F. Grosse particles hardly showed a coincidence between these two proteins (Fig. 4D). Therefore, the two pre-mRNA binding proteins hnRNP A1 and NDH may bind to the same substrate at different stages of processing. Alternatively, these two proteins may have a non-overlapping substrate specificity. NDH II is associated with poly(A)-containing RNA from HeLa cells Our cell biological studies indicated that NDH II is an RNAbinding protein with a preference for pre-mRNA and possibly mRNA. To support these observations by biochemical means, we studied whether NDH II bound to poly(A)-containing RNA. Some RNA binding proteins of the hnRNP group (Dreyfuss et al., 1993) and a DEXD/H RNA helicase (Ladomery et al., 1997) Fig. 1. Immunofluorescence of NDH II in HeLa cells. The cellular localization of NDH II was examined by immunofluorescence with rabbit antiserum against bovine NDH II (1:1000), and detected by a Cy3-secondary antibody against rabbit IgG (A-D). HeLa cells were either grown under normal culturing conditions (A and C), or treated by 5 µg/ml actinomycin D (B), or 100 µM DRB (D) for 3.5 hours. Double-immunofluorescence studies were performed by coincubation of NDH II antibodies with monoclonal antibodies (1:1000) against hnRNP A1 (A′ and B′) or Sc-35 (Sigma) (C′ and D′). The latter were detected by an FITC-labeled antibody against mouse IgG. have been found to co-elute from oligo(dT) columns together with the retained poly(A)-containing RNA. Analogously, cytoplasmic and nuclear fractions of HeLa cells were loaded onto oligo(dT) cellulose columns in the presence of 500 mM NaCl. The columns were washed and eluted with salt-free buffer. Highsalt buffers support binding of poly(A)-containing RNA to the oligo(dT) matrix, and disfavor protein-nucleic acid interactions, while low-salt buffers destabilize poly(A)•oligo(dT) interactions and stabilize protein-RNA interactions. NDH II turned out to be a poly(A)-containing RNA-binding protein that could be identified in both cytosolic and nuclear fractions (Fig. 5, lanes 1 and 2). Only full-length NDH II of 142 kDa was retained, whereas degradation products, which always showed up in the cytosol or nuclei (see Fig. 6), were not visible. Inhibition of A A′ B B′ C C′ D D′ Intracellular localization of NDH II 1059 A B C D E F H G I J nuclear RNA synthesis with actinomycin D altered the intracellular distribution of RNA-bound NDH II. As already observed by immunofluorescence microscopy, the amount of NDH II increased in the cytoplasm and decreased in the nuclei A B C D E F Fig. 2. Immunofluorescence of NDH II in mitotic HeLa cells. To enrich mitotic cells, 2 mM thymidine was added to the culture for 14-16 hours to block cells at the beginning of S phase. The thymidine-block was removed by returning to fresh medium. Typically, after 9-10 hours, cells were progressing into mitosis. Immunofluorescence of NDH II is shown for prophase (A), metaphase (C), and telophase (E and H). A double-immunofluorescence micrograph of NDH II and hnRNP A1 (F) or Sc-35 (I) at telophase is also shown. DAPI stainings (B, D, G and J) are presented for visualizing the condensation state of mitotic chromosomes. after inhibition of RNA synthesis (Fig. 5, lanes 3 and 4). Only very little NDH II was eluted from the oligo(dT) cellulose column when, before loading of the column, the cellular extracts were treated with RNase A (Fig. 5, lanes 5 and 6). Fig. 3. Effect of actinomycin D on the recruitment of NDH II, hnRNP A1 and Sc-35 into HeLa telophase nuclei. HeLa cells, 9 hours after the release from a thymidine block (see legend of Fig. 2), were treated with 5 µg/ml actinomycin D for 1 hour. Doubleimmunofluorescence of NDH II with hnRNP A1 (A and B) and NDH II with Sc-35 (D and E) are shown. DAPI stainings (C and F) are presented for visualizing telophase nuclei. 1060 S. Zhang, C. Herrmann and F. Grosse Fig. 4. Immunoelectron microscopy of NDH II in HeLa cells. Electron microscopic detection of NDH II required post-embedding immunolabeling, followed by an EDTAregressive method for contrasting RNP structures. NDH II and hnRNP A1 can be distinguished by two different sizes of goldparticles conjugated with the secondary antibody against rabbit IgG (10 nm) and mouse IgG (5 nm). Results are shown for the individual immunolabeling of NDH II (A and B), hnRNP A1 (C), and for double-immunolabeling of NDH II and hnRNP A1 (D). Larger arrowheads indicate gold-particles corresponding to NDH II (10 nm) and smaller arrowheads indicate goldparticles corresponding to hnRNP A1 (5 nm) in nuclei or cytoplasm. The subcellular compartments are indicated as (C) for cytoplasm and (N) for nucleoplasm. Magnification: ×50, 000 (A,C,D); ×100,000 (B). Bars: 200 nm (A,C,D); 100 nm (B). Intracellular localization of NDH II 1061 Fig. 5. Co-elution of NDH II and poly(A)-containing RNA from oligo(dT) cellulose. HeLa cytosolic and nuclear extracts were passed through oligo(dT) cellulose (200 µl) in the presence of 500 mM NaCl. After washings, the columns were eluted with buffer without salt. The eluted fractions were treated with RNase A and the proteins that co-eluted with the oligo(dT)-retained RNAs were analyzed by western blotting with NDH II antiserum (1 to 1000 diluted). Cytosolic and nuclear extracts were prepared from HeLa cells grown under normal culturing conditions (lanes 1 and 2) or treated with 5 µg/ml actinomycin D for 3.5 hours (lanes 3 and 4). Only minor amounts of NDH II co-eluted from the oligo(dT) cellulose column after nuclear and cytosolic extracts were treated with 0.1 mg/ml RNase A for 15 minutes at room temperature (lanes 5 and 6). Salt-solubility of NDH II in HeLa cell nuclei Salt-extraction provides a physical means to distinguish between salt-soluble and salt-insoluble nuclear proteins. The latter ones are thought to be associated with the nuclear matrix, a proteinaceous network that might constitute the solid support for the co-ordination of transcription and RNA processing (for a recent review see Pederson, 1998). To examine whether NDH II was associated with the nuclear matrix, we fractionated HeLa cells into cytosol and nuclei. The latter were further purified via sedimentation through a sucrose cushion of highdensity (0.7 M), followed by chromatin depletion with DNase I and high salt extraction of the non-chromatin remnant. western blots were performed to analyze the presence of NDH II in different fractions of the matrix preparation. It turned out that NDH II did not belong to the nuclear matrix proteins since repeated extractions with 2 M NaCl depleted the HeLa nuclei of NDH II (Fig. 6A). We studied also the high salt-solubility of hnRNP A1 (Fig. 6B) and Sc-35 (Fig. 6C). hnRNP A1 displayed a salt-solubility similar to that of NDH II, while Sc35 was not extractable with high salt concentrations, indicating that Sc-35 is a nuclear matrix protein. Nucleic acids binding properties of NDH II A salient feature of NDH II and its homologues is binding to and unwinding of both DNA and RNA. The cell biological studies given above as well as the previously determined domain structure of NDH II point to a role in RNA rather than DNA metabolism. In an approach to better understand the substrate specificity of this enzyme, we measured the relative affinities of NDH II for dsRNA, ssRNA, and ssDNA using filter binding assays (Fig. 7). Binding of NDH II to poly(rI•rC) and to poly(rI) exhibited a steep linear increase up to a plateau level when the protein concentration was increased. In contrast, NDH II binding to ssDNA occurred with a sigmoidal curve that leveled off at a lower plateau than that observed with RNA. From the binding curves, apparent association constants (KA) were calculated as 1.6•107 M−1, 1.5•107 M−1, and 7.4•106 M−1 for binding to dsRNA, ssRNA, and ssDNA, respectively. Hence, NDH II bound equally well to both forms of RNA, and only by a factor of two less well to DNA. To further discriminate between the various substrate affinities, gel mobility shift assays were performed. Radioactively labeled 85-mer dsRNA or M13 ssDNA were incubated with different amounts of NDH II and subsequently subjected to gel electrophoresis under non-denaturing conditions. Also, nucleic acid binding of NDH II was challenged with various amounts of poly(rI•rC), poly(rI), poly(rC), M13 ssDNA, and M13 dsDNA in competition binding assays. NDH II bound to both dsRNA and ssDNA in a concentration-dependent manner (Fig. 8A,B). Binding of NDH II to the dsRNA probe was most efficiently competed by dsRNA and ssDNA, but apparently not by ssRNA and dsDNA (Fig. 8B). Hence, dsRNA-bound NDH II might dissociate from this substrate when a ssDNA entry site is provided. When ssDNA was used as binding probe, the most efficient competitor turned out to be ssRNA rather than dsDNA (Fig. 8B). DISCUSSION The intracellular localization of NDH II was revealed by immunofluorescence and immunoelectron microscopy. Both methods confirmed the previously observed nucleoplasmic localization (Zhang et al., 1995), but showed also the presence of minor amounts of the protein in the cytoplasm. Moreover, shuttling of NDH II between the nucleus and the cytoplasm has been demonstrated by immunofluorescence after transcriptional inhibition and during mitosis. Furthermore, NDH II was shown to be a poly(A)-RNA binding protein that could be extracted from nuclei under normal circumstances and from the cytosol after transcriptional inhibition. The distribution of NDH II in HeLa interphase cells was very similar to that of hnRNP proteins but differed from the speckled distribution of snRNPs and splicing factors. Until now the mechanisms that regulate the subnuclear localization of these proteins is unclear. In interphase, the chromosomes are decondensed and dispersed as distinct chromosomal territories throughout the whole nucleus. Transcription occurs on the surface of the chromosomal domains, to which transcription factors and RNA polymerases have easy access. Nascent RNA transcripts are packaged by RNA-binding proteins to form ribonucleoprotein complexes, which undergo further post-transcriptional processing and nucleocytoplasmic transport. It has been realized that these molecular events may be represented by fibrillar structures 1062 S. Zhang, C. Herrmann and F. Grosse Fig. 6. Comparison of the salt-solubility of NDH II, hnRNP A1 and Sc-35 in HeLa cell nuclei. A nuclear-matrix preparation protocol was applied to the fractionation of cell extracts, followed by western blot analysis of NDH II (A), hnRNP A1 (B) and Sc-35 (C). Rabbit anti-serum against NDH II and monoclonal antibodies against hnRNP A1 and Sc-35 (Sigma) were diluted at 1 to 1000. Fractions were from the cytosol (C) (lane 1), nuclei (N) (lane 2), the nuclear supernatant (NS1) and pellet (NP1) after the first 2 M NaCl extraction of DNase-I treated nuclei (lanes 3 and 4), the supernatant (NS2) and pellet (NP2) after the second 2 M salt extraction (lanes 5 and 6), and the supernatant (NS3) and pellet (NP3) after the third salt extraction (lanes 7 and 8). observed at the periphery of the chromosomal domains (Lamond and Earnshaw, 1998; Misteli and Spector, 1998). hnRNPs are associated with perichromatin fibrils and immunofluorescence studies revealed that they are homogeneously distributed throughout the nuclear regions excluding the nucleoli (Dreyfuss et al., 1993). The occupation of these fibrils by both NDH II and hnRNPs provided evidence for common functions in posttranscriptional RNA processing and transport. Indeed, NDH II displays some similarities with hnRNP proteins by having a modular protein structure consisting of two dsRNA binding domains (dsRBDs) at the N terminus and an arginine/glycinerich RGG box at the C terminus (Dreyfuss et al., 1993; Zhang and Grosse, 1997). The additional RNA binding elements may contribute to the observed nuclear localization pattern (Dreyfuss et al., 1993). dsRBDs are part of many nucleic acid binding proteins, such as the components of hnRNPs, mRNPs, and ribosomes (Bass et al., 1994; Eckmann and Jantsch, 1997). The dsRBDs are also essential domains for some proteins involved in RNA transport and translation (Bycroft et al., 1995; McMillan et al., 1995). An RGG-box is not only found in NDH II but also in many hnRNP proteins; this domain may mediate both nucleic acid binding and protein-protein interactions (Dreyfuss et al., 1993). Hence, it is tempting to speculate that both the dsRBDs and the RGG-box of NDH II support binding to hnRNAs for subsequent processing and transport. On the other hand, a family of pre-mRNA binding proteins, which participate in spliceosome-mediated splicing events, contain arginine/serine (RS)-rich nucleic acid binding domains (Kramer, 1996). RS-rich pre-mRNA binding proteins, such as Sc-35, have a nuclear localization different from hnRNP proteins. Although they are also found in perichromatin fibrils, where transcription, splicing and other co-transcriptional events occur, most of the RNA splicing factors are accumulated in interchromatin granules, which probably represent sites for their recycling and/or storage. Corresponding to their distribution on perichromatin fibrils and their enrichment at interchromatin granules, splicing factors give rise to a diffused nucleoplasmic staining pattern and the characteristic speckled localization. RS-domains and other nucleic acid binding motifs, such as RRM (RNA recognition motif), seem to direct splicing factors to these subnuclear structures, possibly via specific protein-protein and/or protein-nucleic acid interactions Intracellular localization of NDH II 1063 0.8 0.7 Fraction Bound 0.6 0.5 0.4 0.3 0.2 0.1 0 0 40 80 120 160 200 240 280 NDH II (nM) poly(rI•rC) [KA = 1.6 x 107 M-1 ] poly(rI) [KA = 1.5 x 107 M-1 ] ssM13 [KA = 7.4 x 106 M-1 ] Fig. 7. Quantification of NDH II binding to ssRNA, dsRNA, and ssDNA. The indicated amounts of purified NDH II were incubated with 32P-labeled poly(rI•rC) (open circles), poly(rI) (filled circles), or M13 ssDNA (filled squares) in binding buffer and then passed through nitrocellulose filters. The retained radioactivities were determined by scintillation counting. (Caceres et al., 1997; Li and Bingham, 1991). According to both the subnuclear localization and primary structure, NDH II does not belong to the group of pre-mRNA binding proteins that functions as component of the spliceosome. Cell biological studies revealed that NDH II most likely is a pre-mRNA and mRNA binding protein that might be involved in structural rearrangements of nascent transcripts for subsequent processing and transport. This explanation, however, did not explain the still enigmatic property of this enzyme (and its homologues) to bind to both DNA and RNA. This led us to measure the relative affinities of NDH II for both nucleic acid substrates. Interestingly, NDH II bound dsRNA, ssRNA, and ssDNA nearly equally well, but displayed a much lower affinity for dsDNA. The low affinity for dsDNA might explain that upon transcriptional inhibition and during mitosis, NDH II is not retained within the nucleus. This also explains the well-known biochemical finding that NDH II requires a single-stranded tail as entry site for DNA unwinding (Zhang and Grosse, 1991). Moreover, a detailed biochemical analysis of the binding properties of the individually expressed auxiliary binding domains of NDH II revealed that the dsRBDbound dsRNA could be competed by ssRNA and by ssDNA, but not by dsDNA. Furthermore, the RGG box bound equally well to ssDNA and ssRNA and again not to dsDNA (S. Zhang and F. Grosse, unpublished results). Thus, from combining cell biological and biochemical data one can envision the following scenario for the action of NDH II on pre-mRNA and mRNAs: NDH II binds equally well to single-stranded and double-stranded regions of RNA, where single-stranded parts may act as entry-sites for the subsequent unwinding of dsRNA. Those unwound regions may be reannealed by hnRNPs to form correctly folded transcripts for subsequent splicing and transport. NDH II, when in the proximity of chromatin fibers may dissociate from re-folded Fig. 8. Competitive gel shift assays for studying NDH II’s relative affinities for various nucleic acids. (A) Recombinant and highly purified human NDH II was incubated with the dsRNA probe and then subjected to gel electrophoresis. The left panel indicates titrations with 0-160 nM of NDH II, and the resulting mobility shifts. The following panels (from left to right) indicate competition of the mobility shift obtained with 40 nM NDH II by the indicated amounts of poly(rI•rC), poly(rI), poly(rC), M13 ssDNA, and M13 dsDNA. (B) The same experiment as described above, but using 6 µM (nt) M13 ssDNA as binding substrate for NDH II. Here, binding was competed with 1.5-15 µM (nt) poly(rI•rC) and poly(rI), respectively. 1064 S. Zhang, C. Herrmann and F. Grosse dsRNAs to ssDNA that is transiently formed during transcription and thereby may facilitate a new round of transcript formation. Alternatively, these ssDNA domains may act as retention sites to keep NDH II within the nucleus. When the lengths of the newly formed transcript exceeds ≈17 nucleotides, i.e. the minimal length for nucleic acid binding (Zhang and Grosse, 1994), NDH II might use this singlestranded part of the nascent transcript as binding site to restart the removal of randomly formed secondary structures from the newly formed pre-mRNA. Therefore, NDH II might act as molecular chaperone that removes unwanted hybrids between two neighboring pre-mRNA molecules in heavily transcribed genes. Upon transcriptional inhibition or during mitosis, NDH II cannot be kept in the nucleus because of the lack of singlestranded nucleic acids. Hence, it migrates together with mRNA into the cytoplasm, where, due to its nuclear localization sequence, at least a fraction of it is immediately transported back into the nucleus. Here, it can only be retained when ssDNA (or ssRNA) is available as association site for the next round of binding to single-stranded parts of newly formed transcripts, and transient unwinding of unfavorable doublestranded regions of newly synthesized pre-mRNAs. This model is in agreement with the finding that NDH II binds to retroviral type D genomic RNA and migrates together with it to the cytoplasm (Tang et al., 1997). However, from our studies it is not yet clear whether NDH II recognizes specific RNA sequences or structures. The model would also explain the close proximity between NDH II and the transcriptionally active RNA polymerase, as recently exemplified by showing that NDH II acts as a ‘bridging factor’ between RNA polymerase II and the transcriptional coactivator CBP/p300 (Nakajima et al., 1997). Indeed, we have biochemical evidence that NDH II copurifies together with RNA polymerase II, which indicates complex formation between these two enzymes. Moreover, our model might not only hold true for transcription of class II genes, but might be extended to the synthesis of ribosomal genes, at least in specific cells and under specific conditions (unpublished observations). We are grateful to S. Monajembashi and G. Gothe for their assistance in preparing microscopic photographs and figures, to A. Willitzer for synthesis of the oligonucleotides, and to H.-P. Nasheuer for many suggestions and critically reading the manuscript. The monoclonal antibody against hnRNP A1 (4B10) was generously provided by G. Dreyfuss (Howard Hughes Medical Institute, University of Pennsylvania School of Medicine, Philadelphia). This work was supported by grant Gr895/12-1 of the Deutsche Forschungsgemeinschaft. REFERENCES Bass, B. L., Hurst, S. R. and Singer, J. D. (1994). Binding properties of newly identified Xenopus proteins containing dsRNA-binding motifs. Curr. Biol. 4, 301-314. Bernhard, W. (1969). A new staining procedure for electron microscopical cytology. J. Ultrastruct. Res. 27, 250-265. Bycroft, M., Grünert, S., Murzin, A. G., Proctor, M. and Johnston, D. S. (1995). NMR solution structure of a dsRNA binding domain from Drosophila staufen protein reveals homology to the N-terminal domain of ribosomal protein S5. EMBO J. 14, 3563-3571. Caceres, J. F., Misteli, T., Screaton, G. R., Spector, D. L. and Krainer, A. R. (1997). Role of the modular domains of SR proteins in subnuclear localization and alternative splicing specificity. J. Cell Biol. 138, 225-238. Dreyfuss, G., Matunis, M. J., Piñol-Roma, S. and Burd, C. G. (1993). hnRNP proteins and the biogenesis of mRNA. Annu. Rev. Biochem. 62, 289321. Eckmann, C. R. and Jantsch, M. F. (1997). Xlrbpa, a double-stranded RNAbinding protein associated with ribosomes and heterogeneous nuclear RNPs. J. Cell Biol. 138, 239-253. Gee, S., Krauss, S. W., Miller, E., Aoyagi, K., Arenas, J. and Conboy, J. G. (1997). Cloning of mDEAH9, a putative RNA helicase and mammalian homologue of Saccharomyces cerevisiae splicing factor Prp43. Proc. Nat. Acad. Sci. USA 94, 11803-11807. Grosse, F., Nasheuer, H. P., Scholtissek, S. and Schomburg, U. (1986). Lactate dehydrogenase and glyceraldehyde-phosphate dehydrogenase are single-stranded DNA-binding proteins that affect the DNA polymerase αprimase complex from calf thymus. Eur. J. Biochem. 160, 459-467. Kramer, A. (1996). The structure and function of proteins involved in mammalian pre-mRNA splicing. Annu. Rev. Biochem. 65, 367-409. Kuroda, M. I., Kernan, M. J., Kreber, R., Ganetzky, B. and Baker, B. S. (1991). The maleless protein associates with the X chromosome to regulate dosage compensation in Drosophila. Cell 66, 935-947. Ladomery, M., Wade, E. and Sommerville, J. (1997). Xp54, the Xenopus homologue of human RNA helicase p54, is an integral component of stored mRNP particles in oocytes. Nucl. Acids Res. 25, 965-973. Lamond, A. I. and Earnshaw, W. C. (1998). Structure and function in the nucleus. Science 280, 547-553. Lee, C.-G. and Hurwitz, J. (1992). A new RNA helicase isolated from HeLa cells that catalytically translocates in the 3′ to 5′ direction. J. Biol. Chem. 267, 4398-4407. Lee, C.-G., Chang, K. A., Kuroda, M. I. and Hurwitz, J. (1997). The NTPase/helicase activities of Drosophila maleless, an essential factor in dosage compensation. EMBO J. 16, 2671-2681. Li, H. and Bingham, P. M. (1991). Arginine/serine-rich domains of the su(wa) and tra RNA processing regulators target proteins to a subnuclear compartment implicated in splicing. Cell 67, 335-342. McMillan, N. A. J., Carpick, B. W., Hollis, B., Toone, W. M., ZamanianDaryoush, M. and Williams, B. R. G. (1995). Mutational analysis of the double-stranded RNA (dsRNA) binding domain of the dsRNA-activated protein kinase, PKR. J. Biol. Chem. 270, 2601-2606. Misteli, T. and Spector, D. L. (1998). The cellular organization of gene expression. Curr. Opin. Cell Biol. 10, 323-331. Molnar, G. M., Crozat, A., Kraeft, S. K., Dou, Q. P., Chen, L. B. and Pardee, A. B. (1997). Association of the mammalian helicase MAH with the pre-mRNA splicing complex. Proc. Nat. Acad. Sci. USA 94, 7831-7836. Nakajima, T., Uchida, C., Anderson, S. F., Lee, C. G., Hurwitz, J., Parvin, J. D. and Montminy, M. (1997). RNA helicase A mediates association of CBP with RNA polymerase II. Cell 90, 1107-1112. Ono, Y., Ohno, M. and Shimura, Y. (1994). Identification of a putative RNA helicase (HRH1), a human homolog of yeast Prp22. Mol. Cell. Biol. 14, 7611-7620. Pederson, T. (1998). Thinking about a nuclear matrix. J. Mol. Biol. 277, 147159. Piñol-Roma, S. and Dreyfuss, G. (1991). Transcription-dependent and transcription-independent nuclear transport of hnRNP proteins. Science 253, 312-314. Piñol-Roma, S. and Dreyfuss, G. (1992). Shuttling of pre-mRNA binding proteins between nucleus and cytoplasm. Nature 355, 730-732. Richter, L., Bone, J. R. and Kuroda, M. I. (1996). RNA-dependent association of the Drosophila maleless protein with the male X chromosome. Genes Cells 1, 325-336. Spector, D. L., Fu, X. D. and Maniatis, T. (1991). Associations between distinct pre-mRNA splicing components and the cell nucleus. EMBO J. 10, 3467-3481. Sukegawa, J. and Blobel, G. (1995). A putative mammalian RNA helicase with an arginine-serine-rich domain colocalizes with a splicing factor. J. Biol. Chem. 270, 15702-15706. Tang, H., Gaietta, G. M., Fischer, W. H., Ellisman, M. H. and Wong-Staal, F. (1997). A cellular cofactor for the constitutive transport element of type D retrovirus. Science 276, 1412-1415. Zhang, S. and Grosse, F. (1991). Purification and characterization of two DNA helicases from calf thymus. J. Biol. Chem. 266, 20483-20490. Zhang, S. and Grosse, F. (1994). Nuclear DNA helicase II unwinds both DNA and RNA. Biochemistry 33, 3906-3912. Zhang, S., Maacke, H. and Grosse, F. (1995). Molecular cloning of the gene encoding nuclear DNA helicase II. J. Biol. Chem. 270, 16422-16427. Zhang, S. and Grosse, F. (1997). Domain structure of human nuclear DNA helicase II (RNA helicase A). J. Biol. Chem. 272, 11487-11494.