Portal de Periódicos da CAPES

Nuclear export and cytoplasmic processing of precursors to the 40S ribosomal subunits in mammalian cells

2005; Springer Nature; Volume: 24; Issue: 16 Linguagem: Inglês

10.1038/sj.emboj.7600752

1460-2075

Jacques Rouquette, Valérie Choesmel, Pierre‐Emmanuel Gleizes,

RNA modifications and cancer

Article28 July 2005free access Nuclear export and cytoplasmic processing of precursors to the 40S ribosomal subunits in mammalian cells Jacques Rouquette Jacques Rouquette Laboratoire de Biologie Moléculaire des Eucaryotes and Institut d'Exploration Fonctionnelle des Génomes, CNRS and Université Paul Sabatier, Toulouse cedex, France Search for more papers by this author Valérie Choesmel Valérie Choesmel Laboratoire de Biologie Moléculaire des Eucaryotes and Institut d'Exploration Fonctionnelle des Génomes, CNRS and Université Paul Sabatier, Toulouse cedex, France Search for more papers by this author Pierre-Emmanuel Gleizes Corresponding Author Pierre-Emmanuel Gleizes Laboratoire de Biologie Moléculaire des Eucaryotes and Institut d'Exploration Fonctionnelle des Génomes, CNRS and Université Paul Sabatier, Toulouse cedex, France Search for more papers by this author Jacques Rouquette Jacques Rouquette Laboratoire de Biologie Moléculaire des Eucaryotes and Institut d'Exploration Fonctionnelle des Génomes, CNRS and Université Paul Sabatier, Toulouse cedex, France Search for more papers by this author Valérie Choesmel Valérie Choesmel Laboratoire de Biologie Moléculaire des Eucaryotes and Institut d'Exploration Fonctionnelle des Génomes, CNRS and Université Paul Sabatier, Toulouse cedex, France Search for more papers by this author Pierre-Emmanuel Gleizes Corresponding Author Pierre-Emmanuel Gleizes Laboratoire de Biologie Moléculaire des Eucaryotes and Institut d'Exploration Fonctionnelle des Génomes, CNRS and Université Paul Sabatier, Toulouse cedex, France Search for more papers by this author Author Information Jacques Rouquette1, Valérie Choesmel1 and Pierre-Emmanuel Gleizes 1 1Laboratoire de Biologie Moléculaire des Eucaryotes and Institut d'Exploration Fonctionnelle des Génomes, CNRS and Université Paul Sabatier, Toulouse cedex, France *Corresponding author. Laboratoire de Biologie Moléculaire Eucaryote (CNRS-UMR 5099), 118 route de Narbonne, 31062 Toulouse cedex, France. Tel.: +33 561 33 59 26; Fax: +33 561 33 58 86; E-mail: [email protected] The EMBO Journal (2005)24:2862-2872https://doi.org/10.1038/sj.emboj.7600752 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info It is generally assumed that, in mammalian cells, preribosomal RNAs are entirely processed before nuclear exit. Here, we show that pre-40S particles exported to the cytoplasm in HeLa cells contain 18S rRNA extended at the 3′ end with 20–30 nucleotides of the internal transcribed spacer 1. Maturation of this pre-18S rRNA (which we named 18S-E) involves a cytoplasmic protein, the human homolog of the yeast kinase Rio2p, and appears to be required for the translation competence of the 40S subunit. By tracking the nuclear exit of this precursor, we have identified the ribosomal protein Rps15 as a determinant of preribosomal nuclear export in human cells. Interestingly, inhibition of exportin Crm1/Xpo1 with leptomycin B strongly alters processing of the 5′-external transcribed spacer, upstream of nuclear export, and reveals a new cleavage site in this transcribed spacer. Completion of the maturation of the 18S rRNA in the cytoplasm, a feature thought to be unique to yeast, may prevent pre-40S particles from initiating translation with pre-mRNAs in eukaryotic cells. It also allows new strategies for the study of preribosomal transport in mammalian cells. Introduction As is the case for other ribonucleoproteic particles (RNPs), the assembly and maturation of the pre-rRNPs begins cotranscriptionally, and thus takes place in the vicinity of the ribosomal genes. This activity leads to the buildup of the nucleolus, a subdomain of the nucleus, which emerges around the chromosomal clusters of the ribosomal transcription units. The major part of the production of the ribosomal subunits takes place within this structure: the processing and folding of the pre-rRNAs, in which the so-called external and internal transcribed spacers (ETS and ITS) are eliminated (see Figure 1), and the concomitant assembly with the ribosomal proteins imported from the cytoplasm. At some point during this process, the newly synthesized ribosomal subunits must exit the nucleus and reach the cytoplasm to ensure translation of mRNAs. Nuclear export must be tightly coordinated with pre-rRNA maturation. The mechanisms underlying the nuclear export of preribosomes in vertebrates and their coordination with preribosome maturation remain largely unknown. Figure 1.Pre-rRNA processing in HeLa cells. Schematic and nomenclature according to Hadjiolova et al (1993). Two alternative pathways are presented. Download figure Download PowerPoint Pre-rRNA maturation starts with the large 80–90S RNP particle which splits into the pre-40S and pre-60S particles after cleavage within the internal transcribed spacer 1 (ITS1). Proteomic analysis of these precursor rRNPs in yeast has revealed that their composition is highly dynamic and includes over 150 nonribosomal proteins putatively involved in ribosome biogenesis (Fatica and Tollervey, 2002; Milkereit et al, 2002; Fromont-Racine et al, 2003; Tschochner and Hurt, 2003). Many of these factors have homologs in vertebrates and some have been found in mammalian preribosomes (Takahashi et al, 2003). Through mutation analysis in yeast, it has been found that nuclear export of the pre-40S and pre-60S particles requires the GTPase Ran, a central component in nuclear–cytoplasmic exchanges, as well as the exportin Crm1p (Hurt et al, 1999; Moy and Silver, 1999; Gleizes et al, 2001). In the case of the pre-60S particles, the current model states that Crm1p is recruited by Nmd3p, a preribosomal factor associated with late pre-60S particles, that contains a leucine-rich nuclear export signal (NES) (Ho et al, 2000; Gadal et al, 2002). How Crm1p binds to the precursors of the small subunits has not been established yet. One ribosomal protein of the small subunit, Rps15p, has been shown to be required for the exit of the pre-40S particles from the nucleus, and may directly or indirectly interact with the nuclear transport machinery (Leger-Silvestre et al, 2004). Some preribosomal maturation factors accompany the precursors of the small subunit from the nucleus to the cytosol and may play a role in this transport process (Schafer et al, 2003; Oeffinger et al, 2004). Translocation through the nuclear pore complex (NPC) involves a subset of nucleoporins located on the cytoplasmic side of the pore, the so-called Nup82 complex, which includes Nup82p, Nup159p, Nsp1p, Nup116p and Gle2p (Hurt et al, 1999; Moy and Silver, 1999, 2002; Gleizes et al, 2001). These nucleoporins have overlapping but dissimilar functions in preribosome and mRNP nuclear export (Gleizes et al, 2001). In higher eukaryotes, the only factors required for preribosomal transport that have been characterized to date are Crm1 and Nmd3, whose roles in the production of the large subunit have been found to be conserved in HeLa cells and in Xenopus laevis oocytes (Thomas and Kutay, 2003; Trotta et al, 2003). The main steps that lead to the production of the mature 18S, 5.8S and 28S rRNAs in vertebrates have been defined with varying degrees of precision; some have been thoroughly studied using both in vitro and in vivo assays, others are still poorly characterized (for reviews, see Hadjiolov, 1985; Eichler and Craig, 1994; Gerbi and Borovjagin, 2004). The overall scheme of this process is comparable to that which has been established in more detail for the yeast Saccharomyces cerevisiae (Venema and Tollervey, 1999). However, the sequences of the ETS and ITS diverge greatly between yeast and vertebrates, and more generally from one eukaryote to the other, which makes it difficult to predict the cleavage sites by direct sequence comparison. Noticeably, in vertebrates, the order of the endo- and exonucleolytic cleavages that eliminate the transcribed spacers seems to vary according to species, cell type or physiological state (Hadjiolova et al, 1993; Eichler and Craig, 1994; Gerbi and Borovjagin, 2004). In vertebrates, although the assembly of the ribosomal subunits is only completed in the cytoplasm by addition of the last ribosomal proteins, it is widely assumed that processing of the preribosomal transcripts (pre-rRNAs) into mature 18S, 5.8S and 28S rRNAs takes place entirely in the nucleus. In contrast, in S. cerevisiae, the final maturation process of the small subunit occurs in the cytoplasm and includes the last cleavage at the 3′ end of the 18S rRNA (Udem and Warner, 1973; Vanrobays et al, 2001). This step, whose mechanism remains to be defined, requires the participation of various factors associated with the pre-40S particles, including the Rio proteins (Vanrobays et al, 2001, 2003). Here, we have addressed the nuclear transport of the pre-40S particles and the coordination of this process with the final steps of pre-18S rRNA maturation in mammalian cells. We show that cytoplasmic maturation of the 18S rRNA is not an exception restricted to yeast, but also takes place in mammalian cells. In HeLa cells, the pre-40S particles exiting the nucleus contain a precursor of the 18S rRNA whose final processing at the 3′ end occurs in the cytoplasm. Nuclear export of these pre-40S particles depends on the human ribosomal protein Rps15 and is also perturbed when exportin Crm1 is inhibited. These data suggest that, despite a strong divergence in the sequences of the transcribed spacers and differences in the pre-rRNA processing modes, the localization of the final cleavage of the pre-18S rRNA to within the cytoplasm is a common feature in eukaryotes. Nuclear export of the small ribosomal subunit as an incomplete form may prevent premature translation initiation on pre-mRNAs within the nucleus. Results Northern blot analysis of pre-rRNA processing in HeLa cells reveals a new precursor to the 18S rRNA Pre-rRNA processing in HeLa cells has been proposed to occur through alternative pathways (Figure 1) (Hadjiolova et al, 1993). We first re-examined pre-rRNA processing in HeLa cells by Northern blot analysis of total RNA with probes complementary to the ETS1, the 5′-most nucleotides of the ITS1 (5′-ITS1 probe), or to the ITS2 (Figure 2A). The highest molecular weight precursor detected corresponds to a mixture of 45S and 45S′ pre-rRNAs, that is, before and after processing at the early cleavage site. The 30S pre-rRNA, produced by cleavage of the 45S/45S′ RNAs in the 3′ part of the ITS1 (site 2), was detected with the ETS1 and 5′-ITS1 probes. The complementary fragment generated by cleavage at site 2, namely the 32S pre-rRNA, was revealed by the ITS2 probe, along with the 12S pre-rRNA, which contains the 5.8S rRNA and the ITS2. A species migrating between the 45S/45S′ and the 30S pre-rRNAs did not contain the ETS1, but showed up with both probes complementary to the ITS1 and to the ITS2, as expected for the 41S pre-rRNA. Figure 2.A new late precursor to the 18S rRNA in HeLa cells, the 18S-E pre-rRNA. (A) Northern blot analysis of total RNA extracts from HeLa cells. Northern blots probed with oligonucleotides complementary to the ETS1, the 5′ of the ITS1 (5′-ITS1) and the ITS2. The 5′-ITS1 probe hybridizes with a rapidly migrating precursor of the 18S rRNA, which we called 18S-E. (B) Synthesis of the 18S-E pre-rRNA is sensitive to treatment with actinomycin D. Analysis by Northern blot (probes: 5′-ITS1 and actin) and ethidium bromide staining (18S and 28S RNAs) of total RNA extracts from HeLa cells treated for 1 h with 40 ng/ml actinomycin D. (C) Northern blot analysis of total RNA extracts from HeLa cells probed with the 5′-ITS1 and 5687-ITS1 oligonucleotides. Download figure Download PowerPoint Two other major 18S rRNA precursors were detected with the 5′-ITS1 probe, but not with the ETS1- or ITS2-specific probes, indicating that they corresponded to precursors downstream of cleavage at sites 1 and 2. The slowest migrating precursor ran above the 18S rRNA, and corresponded to the 21S pre-rRNA, which is thought to result from cleavage at sites 1 and 2. In addition, we detected a new pre-18S species, which was as abundant as the 21S pre-rRNA and migrated with an electrophoretic mobility close to that of the mature 18S RNA. Treatment of the cells for 1 h with 40 ng/ml actinomycin D, which inhibits RNA polymerase I, resulted in the disappearance of this band as well as the other bands detected with the ITS1 probe, confirming that they all corresponded to pre-rRNAs (Figure 2B). As seen in Figure 2C, this small precursor was detected with a probe complementary to the first 20 nucleotides of the ITS1 and spanning seven nucleotides of the 18S, but not with a probe hybridizing with nucleotides 159–183 of the ITS1 (probe ITS1-5687). We refined this data by performing 3′-RACE experiments on total RNAs and found a major cleavage site around position 24 in the ITS1 (Supplementary data). The intensity of this band did not parallel that of the 18S rRNA on Northern blots, as seen in the presence of actinomycin (Figure 2B) or after separation on polysome gradients (see Figure 3C); we can therefore exclude cross-hybridization of the 5′-ITS1 probe with the 18S rRNA on Northern blots. We named this uncharacterized RNA ‘18S-E’, for extended and exported (see below). Figure 3.The 18S-E pre-rRNA is mostly located in the cytoplasm and does not participate in translation. (A) Northern blot analysis of total, cytoplasmic and nuclear RNA extracts from human HeLa cells and murine L929 cells shows localization of the 18S-E pre-rRNA to the cytoplasm. Hybridization was performed with probes complementary to the 5′-most nucleotides of the ITS1 and specific to each species. Only 1/10th of the total and cytoplasmic extracts was loaded on the gel. (B) Intracellular distribution of the 18S-E, 18S and 28S RNAs under various conditions was quantified on Northern blots by phosphorimaging. n: number of experiments. (C) Ribosome/polysome separation on sucrose gradient and Northern blot analysis of the fractions. HeLa cell cytoplasmic extracts were centrifuged on a 10–50% sucrose gradient. The 18S and 28S rRNAs were detected by ethidium bromide staining and the 18S-E species was revealed by Northern blotting with the 5′-ITS1 probe. The 18S-E RNA is found with the free 40S subunits. Download figure Download PowerPoint 18S rRNA precursors are exported from the nucleus in human and murine cells To obtain further insight into the location of the 18S precursors in HeLa cells, we prepared nuclear and cytoplasmic fractions of these cells and analyzed their rRNA content by Northern blotting. As seen in Figure 3A, all the previously characterized 18S precursors were found exclusively in the nucleus, including the 21S pre-rRNA. Strikingly, the 18S-E species was unambiguously detected in the cytoplasm. Using quantitative phosphorimager analysis (Figure 3B), we estimated that the cytoplasm contained 70–80% of the 18S-E pre-rRNA, which indicates that it is efficiently exported from the nucleus. Under the same conditions, 5–8% of the signals detected with probes to the mature 18S or 28S rRNA was found in the nuclear fraction. Upon inhibition of rRNA synthesis in the nucleus with actinomycin D, this number dropped to 3%, which may be considered as an estimate of the contamination of the nuclear fraction with cytoplasmic ribosomes. We extended this observation to another mammalian species by repeating this experiment with murine L929 cells. As shown in Figure 3A, a probe directed to the 5′-most nucleotides of the murine ITS1 detected rRNA precursors in L929 total RNA migrating as in HeLa cells, including the 18S-E species. In addition, upon cell fractionation, this species was also found to be cytoplasmic in L929 cells; up to 90% of 18S-E rRNA was found in the cytoplasmic fraction (Figure 3B). To more precisely establish the fate of the 18S-E species in the cytoplasm of HeLa cells, we next separated the cytoplasmic ribosomal particles on a sucrose gradient. Detection of the mature 18S and 28S rRNAs in the fractions clearly indicated the positions of the free 40S and 60S subunits, the 80S ribosome and the polysomes (Figure 3C). Hybridization with the 5′-ITS1 probe showed that the 18S-E species was restricted to particles migrating in the gradient with the 40S subunits, and was not incorporated into the polysomes, as expected for precursors to the 40S particles. These results confirm that the 18S-E species is exported in ∼40S particles, and strongly suggest that it does not participate in translation. Maturation of the 18S rRNA involves a cytoplasmic protein The results above indicate that the 18S-E RNA is transported to the cytoplasm in pre-40S particles. In S. cerevisiae, cleavage at the 3′ end of the 18S rRNA takes place in the cytoplasm and requires several proteins associated with the pre-40S particle. We hypothesized that the processing of the 18S-E species in HeLa cells was an orthologous process involving the mammalian counterparts of these proteins. If this hypothesis is true, loss of function of one of these proteins should induce the accumulation of 18S-E RNA in the cytoplasm, without affecting the relative levels of the earlier 18S rRNA precursors, whereas the amount of 18S rRNA should decrease. One of the best characterized proteins involved in the cytoplasmic maturation of the 18S rRNA in S. cerevisiae is Rio2p (ScRio2p), a protein kinase found in late pre-40S particles (Vanrobays et al, 2003). The human predicted counterpart (hRio2) displays 58% sequence similarity with ScRio2p over 419 amino acids and contains an additional C-terminal sequence. Rio2p in yeast is mostly detected in the cytoplasm and is exported from the nucleus via a Crm1p-dependent mechanism (Vanrobays et al, 2003; Leger-Silvestre et al, 2004). As shown in Figure 4A, cells expressing a cDNA encoding this protein fused to the green fluorescent protein (EGFP) displayed a fluorescent signal restricted to the cytoplasm when observed by confocal microscopy. There was no evidence of the presence of hRio2 in the nucleoli. Consistently, hRio2-EGFP was only detected in the cytoplasmic fraction by Western blot (Figure 4B). Upon treatment with leptomycin B (LMB), a drug inhibiting the exportin Crm1, hRio2-GFP shifted to the nucleus (Figure 4C), which indicates that it is actively exported from the nucleus by a Crm1-dependent mechanism, as previously observed for ScRio2p (Vanrobays et al, 2003). Nuclear export of hRio2 does not depend on association with pre-40S particles in the nucleus, since it was not perturbed upon inhibition of ribosomal biogenesis by actinomycin D (Figure 4C). As expected from a factor associated to pre-40S particles, a fraction of hRio2-EGFP cosedimented with free (pre-) 40S subunits in a sucrose gradient (Figure 4D). These results demonstrate that hRio2 is a cytoplasmic protein which is actively exported out of the nucleus through a Crm1-dependent mechanism, and suggest that it is associated with pre-40S particles. Figure 4.Localization of hRio2 in HeLa cells. (A) Visualization of HeLa cells expressing Rio2-EGFP and EGFP by laser-scanning confocal microscopy. (B) Western blot of the nucleus and cytoplasm fractions probed with anti-GFP antibodies shows that hRio2-GFP is a cytoplasmic protein. As a control, the histone deacetylase HDAC2 is detected in the nuclear fraction. (C) hRio2-GFP is actively exported from the nucleus through a mechanism depending on Crm1, but not on preribosome synthesis. At 48 h after transfection with the hRio2-EGFP expression vector, cells were treated with 10 nM LMB for 2 h, or with 0.04 μg/ml actinomycin D for 1 h. (D) A fraction of hRio2-GFP cosediments with (pre-)40S particles on a sucrose gradient. Lysates from cells expressing hRio2-EGFP or EGFP were fractionated by ultracentrifugation on sucrose gradient as in Figure 3B. Fractions were analyzed by SDS–PAGE and Western blot with antibodies against GFP and against S19, a protein of the 40S subunit which is used as a marker of the free (pre-)40S subunits, the 80S monosomes and the polysomes. Note that expression of EGFP was much higher than that of hRio2-EGFP. Download figure Download PowerPoint We designed siRNAs targeting the hRIO2 mRNA. These siRNAs efficiently and specifically affected the stability of the endogenous hRio2 mRNA and inhibited the synthesis of hRio2-EGFP (Supplementary data). Northern blot analysis of total RNA with the 5′-ITS1 probe 72 h post-transfection with Rio2 siRNAs showed the expected phenotype, an increase in the 18S-E signal relative to that of the earlier precursors or of the 28S rRNA (Figure 5A). In parallel, the 18S/28S molar ratio, as measured from agarose gel stained with ethidium bromide or Northern blots, was found to drop by ∼50%, reflecting a significant inhibition of the production of the 18S rRNA. Analysis of the intensity profiles on the autoradiography (normalized to the amount of 28S rRNA) indicated that accumulation of the 18S-E species was the major perturbation of pre-rRNA processing (Figure 5A). We also noticed a moderate accumulation of the 45S/45S′ precursor and of an undescribed species migrating just below the 28S rRNA, which we called ‘26S’. Increase in the levels of these precursors indicate retardation in processing of the 5′-ETS (see below the characterization of the 26S species, Figure 7). A similar phenotype was also observed in yeast upon depletion of Rio2p, and may be secondary to the strong defect in 40S subunit synthesis induced by Rio2 depletion (Vanrobays et al, 2003). Figure 5.Analysis of pre-rRNA processing in HeLa cells after knockdown of hRio2 expression with siRNAs. (A) Northern blot analysis with the 5′-ITS1 probe of total RNAs shows accumulation of the 18S-E pre-rRNA in HeLa cells treated with hRio2 siRNAs. A scrambled siRNA with no defined target was used as a control. Equal amounts of RNA were loaded in each lane. The intensity profiles were determined by phophorimaging and normalized to the amount of 28S rRNA. The levels of 18S and 28S rRNAs were measured by reprobing the blot with complementary oligonucleotides (not shown). The values of the ratios were arbitrarily set to 1.0 in control cells. (B) Polysome analysis of HeLa cells treated with scrambled or hRio2 siRNAs and detection by Northern blot of the 18S-E pre-rRNA. Depletion of hRio2 leads to a strong imbalance between the free 60S and 40S subunits, indicating a defect in 40S subunit production. The 18S and 28S rRNAs were visualized with ethidium bromide. (C) FISH with the 5′-ITS1 probe conjugated to Cy3 shows accumulation of 18S-E pre-rRNA in the cytoplasm. In parallel, the U2 snRNA was detected with a Cy5-labeled probe. Quantification of the fluorescence was performed as described in Materials and methods. For the 5′-ITS1 probe, the results of two separate experiments are shown (***P<0.001, Student's t-test). (D) Northern blot analysis of subcellular nuclear and cytoplasmic fractions of HeLa cells with the 5′-ITS1 probe reveals accumulation of the 18S-E pre-rRNA in the cytoplasm upon treatment with hRio2 siRNA for 72 h. Only 1/10th of the cytoplasmic fraction was loaded on the gel. The levels of 18S and 28S were determined on the same blot with specific probes. The subcellular distribution of the 18S-E, 18S and 28S RNAs was quantified by phosphorimaging. N: nucleus; C: cytoplasm. Download figure Download PowerPoint Figure 6.Characterization of the pre-rRNAs accumulating in LMB-treated cells. (A) Metabolic labeling of RNAs in HeLa cells with [32P]orthophosphate. Time after addition of [32P]orthophosphate to the medium is indicated (see Materials and methods). Total RNAs were extracted, separated by electrophoresis, and revealed by autoradiography. (B) Mapping of the extremities of the 26S RNA with probes complementary to the 5′-ETS (ETS1) and to the ITS1. Comparison of total RNAs from untreated and LMB-treated cells facilitates visualization of the 26S RNA. (C) Determination of the 5′-end of the 26S RNA by primer extension. Reverse transcription was performed with oligonucleotide ETS1-1722 on nuclear RNAs from LMB-treated cells. (D) Proposed boundaries for the 26S and 43S pre-rRNAs. Download figure Download PowerPoint Consistent with a defect in the production of the 40S subunit, analysis of the translation machinery on sucrose gradients showed a decrease in the level of both free 40S subunits and 80S assembled ribosomes, together with a large buildup in the amount of free 60S particles (Figure 5B). Upon Northern blotting, the accumulating 18S-E pre-rRNA was mostly detected in the fractions corresponding to free (pre-) 40S subunits. A small part of it was also found in the 80S fraction, suggesting that, when accumulating in high amounts in the cytoplasm, these pre-40S particles could assemble with the 60S subunits into monosomes. We next performed in situ hybridization with the 5′-ITS1 probe conjugated to Cy3 (Figure 5C). In untreated HeLa cells, we observed a strong labeling of the nucleolus and a weak signal in the cytoplasm, consistent with our cell fractionation data showing the presence of 18S rRNA precursors both in the nucleus and in the cytoplasm. Treatment with the siRNAs resulted in a marked increase in the cytoplasmic signal, as expected from the accumulation of 18S-E pre-rRNA in this compartment. To a lower extent, fluorescence in the nucleoli of hRIO2 siRNA-treated cells was more intense than in control cells, consistent with the higher levels of 45S/45S′ and 26S precursors observed by Northern blot. In parallel, the signal observed with a Cy5-conjugated probe complementary to the U2 snRNA decreased in cells treated with hRio2 siRNAs, indicating that the increase of the fluorescence observed with the 5′-ITS1 probe was not a mere artefact of the FISH procedure. A similar phenotype was found with two other siRNAs targeted against hRIO2 (not shown). After cell fractionation, more than 90% of the 18S-E RNA in hRio2-depleted cells was found in the cytoplasmic fraction (Figure 5D). The 3′-end of the 18S-E pre-rRNA accumulating under these conditions was identical to that observed in untreated cells when mapped by 3′ RACE analysis (Supplementary data). This set of data is consistent with the 18S-E pre-RNA being a bona fide precursor to the 18S rRNA, whose conversion to mature rRNA takes place in the cytoplasm. This processing step involves the human structural homolog of ScRio2p, which is located to the cytoplasm and appears to be the true functional homolog of the yeast protein. Nuclear export of the pre-40S particles in mammalian cells requires the ribosomal protein Rps15 We next moved on to the identification of molecular determinants of the nuclear export of pre-40S particles in HeLa cells. A defect in the nuclear export of the pre-40S particles should translate into retention of the 18S-E RNA in the nucleoplasm together with its disappearance from the cytoplasm, which can be assessed by FISH or by Northern blot analysis of subcellular fractions. Our work on yeast has recently led us to identify the ribosomal protein Rps15 as a new factor required for nuclear export of the pre-40S particles (Leger-Silvestre et al, 2004). The high degree of homology of this protein between S. cerevisiae and mammals suggests that its function in preribosomal transport may be conserved. To test this hypothesis, we used siRNAs to block Rps15 synthesis and, consequently, its assembly into preribosomes. The siRNAs effectively downregulated Rps15 mRNA without affecting the mRNAs encoding other ribosomal proteins, including Rps19 (Figure 6A) and Rps17 (not shown). Analysis of ribosomes on sucrose gradients 48 h after transfection showed a profile clearly corresponding to a defect in the production of the small subunits, with the complete disappearance of the peak corresponding to the 40S subunit (Figure 6A). Consistent with these observations, electrophoretic analysis of the gradient fractions (which are prepared from cytoplasmic extracts) showed a dramatic decrease of the levels of 18S and 18S-E RNAs, whereas the amount of 28S rRNA remained unaffected (Figure 6A). In contrast, the 18S-E pre-rRNA was more abundant in siRNA-treated cells as compared to control cells, when detected by Northern blot analysis of total RNAs, which includes nuclear RNAs (Figure 6B). Indeed, Northern blot analysis of subcellular fractions showed a conspicuous redistribution of the 18S-E pre-rRNA into the nuclear fraction (Figure 6C). When these cells were submitted to FISH with the 5′-ITS1 probe, the pre-40S particles appeared to be trapped in the nucleoplasm, which was clearly labeled, with almost no signal being detected in the cytoplasm (Figure 6D). This defect in nuclear export was not paralleled by strong alterations in the early steps of pre-rRNA processing, except for a moderate increase in the level of the 45S/45S′ precursor, as seen on Northern blot (Figure 6B). Taken together, these results indicate that Rps15 is not required for pre-rRNA processing upstream of the 18S-E species, but is essential for efficient nuclear export of the pre-40S particles. These experiments were carried out in L929 cells with similar results (data not shown). We conclude that the role of Rps15 in preribosomal trafficking has been conserved in evolution. Figure 7.Characterization of nuclear export factors for the pre-40S particles. (A) Rps15 siRNAs specifically downregulate the level of Rps15 mRNA and efficiently block the production of the 40S subunit. RT–PCR analysis showed that treatment with Rps15 siRNAs for 48 h strongly affected the level of th