|Home | About | Journals | Submit | Contact Us | Français|
Recent studies reveal that maturation of the 40S ribosomal subunit precursors in mammals includes an additional step during processing of the internal transcribed spacer 1 (ITS1), when compared with yeast Saccharomyces cerevisiae, even though the protein content of the pre-40S particle appears to be the same. Here, we examine by depletion with siRNA treatment the function of human orthologs of two essential yeast pre-ribosomal factors, hEnp1/bystin and hTsr1. Like their yeast orthologs, bystin is required for efficient cleavage of the ITS1 and further processing of this domain within the pre-40S particles, whereas hTsr1 is necessary for the final maturation steps. However, bystin depletion leads to accumulation of an unusual 18S rRNA precursor, revealing a new step in ITS1 processing that potentially involves an exonuclease. In addition, pre-40S particles lacking hTsr1 are partially retained in the nucleus, whereas depletion of Tsr1p in yeast results in strong cytoplasmic accumulation of pre-40S particles. These data indicate that ITS1 processing in human cells may be more complex than currently envisioned and that coordination between maturation and nuclear export of pre-40S particles has evolved differently in yeast and mammalian cells.
During the past 10 years, proteomic and genetic studies have allowed the identification of approximately 190 proteins that play a role in eukaryotic ribosome biogenesis, from the initial nucleolar steps to final processing in the cytoplasm (1–6). These proteins, called pre-ribosomal factors, display different functions including endo- or exoribonucleases, RNA helicases, chaperones, NTPases or methyl transferases (2,5). They are implicated in ribosome biogenesis but do not form part of mature ribosomes.
Early ribosomal proteins and pre-ribosomal factors assemble co-transcriptionally in the nucleolus with the neo-synthesized pre-ribosomal RNA (pre-rRNA), forming the so-called 90S pre-ribosomal particle (3,7). The large polycistronic rRNA precursor contains the future 18S, 5.8S and 28S mature ribosomal RNAs flanked by external and internal transcribed spacers (ETSs and ITSs), which are progressively removed through endonucleolytic cleavages and exonucleolytic processing (Figure 1A) (8,9). Early processing steps include enzymatic modifications of nucleotides, by 2′-O-ribose methylation or pseudo-uridylation (10,11), together with removal of the 5′- and 3′-ETS and endonucleolytic cleavage of the ITS1 (Figure 1B and C). This latter step splits the 90S particle into pre-40S and pre-60S particles, which then follow two separate processing pathways that involve distinct pre-ribosomal factors (4,12). Both particles are exported from the nucleus and their maturation is completed in the cytoplasm (13–16). Association and dissociation of these numerous pre-ribosomal factors along the maturation pathway renders the composition of pre-ribosomal particles highly dynamic. The exact activity or function of most of the pre-ribosomal factors remains unclear, even when the step at which they are implicated is known.
The list of pre-ribosomal factors composing the pre-40S particles is much shorter than that for the large subunit (3,17–19). Following cleavage of the ITS1, the majority of the factors present in the 90S particle dissociate, although a few remain part of the pre-40S particle, including Enp1p, Dim1p, Rrp12p and Bms1p (3,17). Dim1p, a methyltransferase associated with both the 90S and the pre-40S particles, methylates two adjacent nucleotides in the 3′-end of the 18S rRNA (20). Enp1p and Bms1p are known to be implicated in the processing of ETS1 and ITS1 (21–23), but their molecular functions remain unclear. Rrp12p, a HEAT repeat containing protein, is able to interact with nuclear pore complex components and appears to be required for the nuclear export of the pre-40S particle (24). These proteins are joined by other pre-ribosomal factors that assemble specifically with the pre-40S particle. Hence, Ltv1p in yeast, as well as the kinase hRio2 in human, were proposed as adapters for exportin Crm1p which is required for transport to the cytoplasm (25,26). Hrr25p, a protein kinase, regulates association of the ribosomal protein Rps3p to the pre-40S particle, allowing a structural modification of the pre-40S particle beak, which could be required for passage through the nuclear pore complex (27). The Rio kinases and Tsr1p are also required for the final cytoplasmic steps, but their molecular function remains to be elucidated (22,28,29). Finally, Nob1p is the endoribonuclease responsible for the cleavage at site D, which matures the 20S pre-rRNA into the 18S rRNA in the cytoplasm (30–33).
Assembly of the pre-ribosomal factors into pre-ribosomes is presumably highly coordinated with pre-rRNA processing and intracellular transport of the particles. Recently, the pre-40S particle composition was characterized in human cells (26). It appears that the pre-ribosomal factors assembled in the pre-40S particle are highly conserved between Saccharomyces cerevisiae and human cells. However, ITS1 processing in yeast and mammals is clearly different. After ITS1 cleavage at site A2 in yeast, elimination of the ITS1 within the pre-40S particles requires a single cytoplasmic cleavage step at the 3′-end of the 18S rRNA (Figure 1B) (13). In contrast, one detects two abundant intermediates in pre-40S particles in mammalian cells, namely the 21S and 18S-E pre-rRNAs. These RNAs correspond to two processing steps, the 21S pre-rRNA being cleaved in the nucleus and the 18S-E species in the cytoplasm (15) (Figure 1B). Therefore, even if the protein content of pre-40S particles is conserved, the pre-ribosomal factors must have evolved to accommodate an increase in complexity of pre-rRNA maturation from yeast to mammals.
To further investigate similarities and differences between yeast and human 40S ribosomal subunit biogenesis, we have studied the role of the human orthologs of two pre-ribosomal factors implicated at two distinct steps of pre-40S particle processing in yeast: Enp1p and Tsr1p. These factors are both essential in yeast. In addition, they are involved at different steps of pre-40S maturation: Enp1p is required for efficient cleavage of the ITS1 in the nucleus (7), whereas Tsr1p is necessary for cleavage at the 3′-end of the 18S rRNA in the cytoplasm (22,28,29). hTsr1, the human ortholog of Tsr1p, was recently found in human pre-40S particles, but its specific role in 40S subunit production is still unclear (26). Bystin/Enp1, the human ortholog of Enp1p, was first identified as a component of a cytoplasmic complex-mediating adhesion of human trophoblast and endometrial epithelial cells in vitro (34,35), as well as an essential protein for early embryonic development after implantation of the embryo in mouse and rat (36–38). Recently, it was reported that human bystin is overexpressed in hepatocellular carcinoma (HCC) and is essential for cell growth and tumor development (39). The involvement of bystin in ribosome biogenesis has also been described (38,40). Further analysis of the functions of bystin and hTsr1 in human ribosome biogenesis indicates that these two proteins are indeed functional orthologs of their yeast counterparts, but highlights differences in pre-40S particle biogenesis between yeast and human.
To construct pEGFP-bystin and pEGFP-hTsr1, bystin and hTsr1 cDNAs were generated by RT-PCR and cloned into the HindIII and XbaI sites of pEGFP-C3 (Clontech Laboratories, Mountain View, CA, USA), in-frame with the sequence encoding EGFP. EGFP alone was expressed from pEGFP-C3. Transfection of HeLa cells was performed by electro-transformation with 10µg of plasmidic DNA, as described below for siRNAs.
Human cervical carcinoma HeLa cells were grown in Dulbecco’s Modified Eagle Medium (DMEM) containing Glutamax® as a supply of l-glutamine, and supplemented with 10% fetal calf serum (FCS), 1mM sodium pyruvate, 100U/ml penicillin and 100µg/ml streptomycin (Invitrogen, Paisley, UK). Cells were incubated at 37°C in 5% CO2. RNA polymerase I transcription was selectively inhibited by 50ng/ml actinomycin D for 2h.
Two or three 21-mer siRNA duplexes were designed and purchased from Eurogentec (Seraing, Belgium) to knockdown expression of human genes encoding bystin (GenBank accession number: NM_004053) and hTsr1 (GenBank accession number: NM_018128), which are two ribosomal factors associated with the small 40S subunit. After washing the cells once in DMEM (without serum and antibiotics), 10µl of 100µM siRNA solution were added on ice to 107 HeLa cells in suspension in 200µl of the same medium. Electro-transformation was performed at 250 V and at 950µF with a Gene Pulser (Bio-Rad, Hercules, CA, USA), in a cuvette with a 4-mm inter-electrode distance. The cell suspension was then plated on a 200-cm2 Petri dish containing 20ml DMEM supplemented with calf serum and antibiotics. Control samples were electro-transformed with a scramble siRNA (siRNA-negative control duplex; Eurogentec). The efficiency of the downregulation induced by each siRNA was assessed by quantitative reverse transcriptase-PCR, taking GAPDH as an internal control to normalize expression of the target genes. As the most efficient knockdowns were obtained with a mixture of siRNAs bystin-1 (5′-cugcccaaggcauuuaagadtdt-3′) and bystin-2 (5′-gagguugagacagucaugudtdt-3′), and with siRNA tsr1-1 (5′-cuggaacaguacacuugaadtdt-3′), these conditions were used for the experiments herein described. The siRNA targeting hRPS15 (siRNA rps15-2: 5′-UCACCUACAAGCCCGUAAA-3′) has been described previously (15).
After treatment with siRNAs for 48 h, the cells were successively washed at 4°C with DMEM, PBS and Buffer A (10mM HEPES, pH 7.9, 1.5mM MgCl2, 10mM KCl). One-tenth of the suspension was kept for total RNA extraction. The cells were then mechanically disrupted with a Dounce homogenizer in Buffer A containing 0.5mM DTT. After centrifugation at 4°C and at 1000g for 10min, the supernatant (i.e. cytoplasmic fraction) was frozen. The pellet containing the nuclei was washed with 10mM Tris–HCl, pH 7.5, 3.3mM MgCl2 and 250mM sucrose. After centrifugation, the nuclei were suspended in a solution containing 10mM MgCl2 and 250mM sucrose, and further purified by centrifugation for 10min at 500g on a sucrose cushion (0.5mM MgCl2, 350mM sucrose). The pellet was then lyzed in 10 volumes of 50mM Na acetate, pH 5.1, 140mM NaCl, 0.3% SDS and the nuclear RNAs were extracted twice with phenol. Total and cytoplasmic RNAs were purified with Trizol reagent. After alcohol precipitation, RNA pellets were dissolved in H2O, quantified using a Nanodrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and diluted to 1mg/ml.
Forty-eight hours after transfection, HeLa cells were treated with 100µg/ml cycloheximide (Sigma) for 10min. The cytoplasmic fractions were prepared on ice as described above, except that cycloheximide was added to all buffers. On each 10–50% (w/w) sucrose gradient, 1mg proteins was loaded, prepared with a Gradient Master former (BioComp Instruments, Fredericton, NB, Canada). The tubes were centrifuged at 4°C and at 36000 r.p.m. for 105min in a SW41 rotor (Optima L100XP ultracentrifuge; Beckman Coulter, Villepinte, France). The gradient fractions were collected at OD254nm with a Foxy Jr. gradient collector (Teledyne Isco, Lincoln, NE, USA).
Pre-rRNA species were analyzed by northern blot. RNAs (3µg/well for total and nuclear extracts, 6µg/well for cytoplasmic extracts), mixed with 3 volumes of loading dye (25mM MOPS, 6% formaldehyde, 1.5% formamide, 0.16µg/ml BET) were separated on a 1% agarose gel in the presence of MOPS buffer (20mM MOPS, 6mM sodium acetate, 7×10−2 N NaOH, 0.5 M EDTA pH 8.0) containing 6% formaldehyde, and run in MOPS/formaldehyde buffer (1×MOPS, 6% formaldehyde) at 140 V. RNAs were then transferred to a Hybond N+ nylon membrane (GE Healthcare, Orsay, France). After fixation by UV-crosslinking, the membranes were pre-hybridized for 1 h at 45°C in 6×SSC, 5×Denhardt’s solution, 0.5% SDS, 0.9µg/ml tRNA. The 32P-labeled oligodeoxynucleotide probe was then added and incubated overnight at 45°C. The probes used in the present study were 18S (5′-tttacttcctctagatagtcaagttcgacc-3′), 5′ITS1 (5′-cctcgccctccgggctccgttaatgatc-3′), ITS1-160 (5′-tctccctcccgagttctcgg-3′), ITS1-605 (5′-ttaaacctccgcgccggaac-3′), ITS1-639 (5′-gacccacgggcggacgggcgatc-3′), ITS1-721 (5′-ggagcggagtccgcggtg-3′), ITS1-779 (5′-gtaaagcccccacccgacggc-3′), ITS2b (5′-ctgcgagggaacccccagccgcgca-3′), ITS2d/e (5′-gcgcgacggcggacgacaccgcggcgtc-3′) and 28S (5′-cccgttcccttggctgtggtttcgctagata-3′). For detection of ITS2, probes ITS2b and ITS2d/e were mixed in equal amounts. After hybridization, the membranes were washed twice for 10min at room temperature in 2×SSC, SDS 0.1% and once in 1×SSC, SDS 0.1%. Labeled RNA signals were acquired with a FLA2000 PhosporImager (Fuji, Stamford, CT, USA) and quantified with ImageGauge software. The 3′RACE analysis was adapted from Kiss and Filipowicz (41). The forward primer used for PCR amplification spanned nucleotides 582–599 in the ITS1: the ITS1-583 primer (5′-ccgccgtccaggtaccta-3′). The amplified fragments were subcloned and automatically sequenced.
For fluorescence in situ hybridization (FISH) experiments, cells grown on glass cover slips were washed twice in PBS and then fixed for 30min with 4% paraformaldehyde in PBS (EMS, Hartfield, PA, USA). After being rinsed in PBS, cells were permeabilized for 18 h in 70% ethanol at 4°C. After two washes in 2×SSC containing 10% formamide, hybridization was performed in the dark at 37°C for 5 h in a buffer containing 10% formamide, 2×SSC, 0.5mg/ml tRNA, 10% dextran sulfate, 250µg/ml BSA, 10mM ribonucleoside vanadyl complexes and 0.5ng/µl of each probe. Precursors to the 18S rRNA were localized with the 5′-ITS1 probe conjugated to Cy3 or Cy5 (GE Healthcare) on amino-modified deoxythymidine (a probe complementary to the 3′-end of the 18S rRNA and the 5′-end of the ITS1), with the ETS1-b probe (5′-agacgacaacgcctgacacgcacggcac-3′) conjugated to Cy5 (localized between the 5′-end of the ETS1 and the 01/A′ cleavage site), with the ETS1-1399 probe (5′-cgctagagaaggcttttctc-3′) conjugated to Cy5 (which hybridizes to the 47S pre-rRNA between 01/A′ and A0 cleavage sites) or with the 5.8S-ITS2 probe (5′-gcgattgatcggcaagcgacgctc-3′) conjugated to Cy3 (a probe complementary to the 3′-end of the 5.8S rRNA and the 5′-end of the ITS2). After two washes at 37°C with 2×SSC containing 10% formamide, the cover slips were rinsed in PBS and DNA was counterstained with 1µg/ml Hoechst 33342 (Molecular Probes). After 10min incubation at room temperature, the cover slips were rinsed twice in PBS and mounted in Mowiol 4.88 (Polyscience Inc., Eppelheim, Germany).
For detection of fluorescent proteins after transient transfection, cells grown on glass cover slips were washed twice in PBS and then fixed for 30min with 4% paraformaldehyde in PBS (EMS). After being washed twice in PBS, DNA was counterstained with Hoechst 33342 for 10min at room temperature. The cover slips were then rinsed twice in PBS and mounted in Mowiol 4.88.
Observations were made with an inverted IX-81 microscope (Olympus, Rungis, France), equipped with a motorized stage (Märzhäuser, Wetzlar-Steindorf, Germany) and a Coolsnap HQ camera (Photometrics, Tucson, AZ, USA), driven by MetaMorph (Molecular Devices, Downington, PA, USA). Cells were observed using a planapochromat×100, 1.4 numerical aperture oil immersion objective, mounted on a PIFOC® piezo stepper (PI, Karlsruhe, Germany). The light source was a Polychrome V monochromator (Till Photonics GmbH, Gräfelfing, Germany) equipped with a 150 W Xenon source and used with a 15-nm bandwidth. For multi-labeling experiments, multiband dichroic mirrors (Chroma, Rockingham, VT, USA) were used to avoid pixel shift in the acquisition plane. Specific single-band emission filters were mounted on a motorized wheel (PI). Images were captured with the same exposure times for different samples. The gray levels were scaled within the same limits in order to allow sample comparison. Fluorescence quantification of regions of interest was performed on 12-bit images using MetaMorph software. After background subtraction, the results were expressed as an amount of gray levels/pixel. Fluorescence was quantified in four identical regions of interest in the nucleoplasm and the cytoplasm in at least 24 cells per sample.
Bystin presents 55% similarity and 36% identity with yeast Enp1p, whereas hTsr1 and Tsr1p share 51% similarity and 30% identity. Similarity between bystin and Enp1p reaches 75% (50% identity) in the C-terminal core domain which defines the specific ‘bystin domain’, whereas their N-termini are more divergent. The bystin domain of Enp1p has a truncated 5′ extremity (Figure 2A). Two highly conserved domains are found in hTsr1 and Tsr1p: the DUF663 domain, whose function is unknown but which is also found in Bms1; and the AARP2CN domain, which is weakly similar to the GTP-binding domain of elongation factor TU.
When expressed as GFP fusions, both proteins were mainly located in the nucleus and concentrated in the nucleolus, while the cytoplasm was faintly stained (Figure 2B). For bystin, this staining pattern was similar to immunodetection of the endogenous protein (data not shown) and confirmed data from previous reports (26,38,40). After arrest of RNA polymerase I transcription and segregation of the nucleolar components by actinomycin D treatment (42), both GFP-bystin and GFP-hTsr1 were detected in the crescent-shape granular component, consistent with their putative association with pre-ribosomal particles in the nucleolus (Figure 2B).
Requirement of bystin or hTsr1 for ribosome synthesis was assessed by knocking-down expression of these proteins with siRNAs. In both cases, mRNA levels were reduced by 84–90% (data not shown). Ribosome production after bystin- or hTsr1 depletion was analyzed from cytoplasmic fractions prepared in the presence of cycloheximide and separated on sucrose density gradients (Figure 3A). Control cells electro-transformed with a scrambled siRNA displayed a typical profile with peaks corresponding to the 40S and 60S subunits, 80S ribosomal particles and polysomes of increasing lengths. After knocking-down bystin, the levels of free 40S subunits and 80S particles were severely decreased, while 60S subunits accumulated. A similar phenotype was also observed upon depletion of hTsr1, although the decrease in 40S subunits was less pronounced. This imbalance in the levels of 40S and 60S subunits indicates a deficit of 40S production. Consistently, the 18S/28S rRNA ratio was ~30% lower in bystin- or hTsr1-deficient cells than in control cells (see histogram in Figure 3C).
In order to determine the pre-rRNA processing steps of the 40S particle maturation pathway involving bystin and hTsr1, pre-rRNAs were analyzed by northern blot with probes hybridizing to the ITS1 or the ITS2 (Figures 3 and and4).4). In parallel, the intracellular fate of the pre-ribosomes was monitored by FISH with probes complementary to the 5′-ETS/ETS1, the ITS1 or the ITS2 (Figure 5).
Cells treated with bystin-specific siRNAs over-accumulated 41S and 21S pre-rRNAs, while the amount of 18S-E pre-RNA decreased (Figure 3B and 3C). The build-up of 41S pre-rRNA was also evidenced with the ITS2 probes, while the 32S and 12S large subunit precursors were unaffected compared with control cells. These pre-rRNA patterns indicate delayed cleavage in ITS1 at site 2 and deficient maturation of the 21S pre-rRNA within the pre-40S particles.
In addition, we observed the conspicuous build-up of a shorter form of the 21S pre-rRNA, that we named 21S-C (court is the French for short). Such a form was previously described upon depletion of RPS19, and named 20S pre-rRNA (43); however this nomenclature is confusing considering the 20S pre-rRNA in S. cerevisiae. Careful examination of total RNAs’ northern blot profiles showed the reproducible presence of a shoulder in the 21S peak at this position in control cells, suggesting that the 21S-C is a regular processing intermediate that accumulates upon bystin depletion (data not shown). Indeed, this intermediate could be clearly detected in the nuclear fraction of control cells (Figure 3D).
Additional probes were used to determine the 3′-end of this species by northern blot (Figure 4A). Both the 21S and 21S-C pre-rRNAs were revealed with probes spanning position 170 (ITS1-160) and 615 (ITS1-605) in the ITS1. In contrast, probes hybridizing around nucleotides 650 (ITS1-639) or 730 (ITS1-721) no longer revealed the 21S-C RNA. Considering that at least 15nt are required for specific hybridization of these oligonucleotidic probes in our experimental conditions, these results indicate that the 3′ extremity of the 21S-C pre-rRNA is roughly located between nucleotides 620 and 655 of the ITS1 (Figure 4B). If the 21S-C pre-rRNA resulted from an alternative endonucleolytic cleavage in the ITS1, one would expect the appearance of a complementary cleavage product. However, no pre-rRNA migrating above the 32S RNA was revealed with the ITS1-721 or the ITS2 probes in bystin-depleted cells (Figure 4C). We observed with these probes the disappearance of the 36S pre-rRNA (Figure 4C), a rare precursor that forms upon direct cleavage of early precursors at site E (Figure 4D), which confirms that bystin depletion blocks cleavage at site E.
To further delineate the 3′-end of the 21S-C pre-RNA, we performed 3′ RACE experiments with an oligonucleotide spanning nucleotides 582–599 in the ITS1. First, an anchor oligonucleotide was ligated into the 3′-termini of RNAs isolated from whole-cell extracts in order to perform reverse transcription. RNAs with their 3′-end between nucleotide 600 and the anchor were then amplified by PCR, subcloned and sequenced. As shown in Figure 4E, we found several RNAs ending between positions 620 and 650, both in control and in bystin-deficient cells. RNAs from bystin-depleted cells were enriched in species ending around positions 635 and 650. These species were also found in wild-type cells and the heterogeneity of the 3′ extremities is consistent with the often-smeary appearance of the band found below the 21S pre-rRNA on northern blots (Figure 3B). This suggests that the 21S pre-rRNA is processed by a 3′–5′ exonuclease after cleavage at site 2. In this hypothesis, the 21S-C could correspond either to the pause or the stalling of this exonuclease, or to the end product of this exonucleolytic processing.
Thus, these data show that bystin participates in efficient cleavage at site 2 and is strictly required for further processing of the 21S pre-rRNA. In addition, they uncover a new processing mechanism in the ITS1.
In contrast to bystin, hTsr1 knockdown did not affect the levels of the 18S rRNA precursors when compared with control cells (Figure 3B). Formation of the large subunit rRNAs was not affected either, as seen on the same northern blot revealed with ITS2 probes. However, in the absence of hTsr1, the 18S/28S rRNA ratio strongly decreased, confirming the impact of hTsr1 depletion on 40S ribosomal subunit formation observed by sucrose-gradient analyses. These data indicate that hTsr1 performs a late function in 18S rRNA maturation and that hTsr1-deficient pre-40S particles are rapidly turned over.
We next analyzed the impact of hTsr1 and bystin knockdown on the intracellular fate of pre-ribosomes by FISH, using probes complementary to transcribed spacers 5′-ETS/ETS1, ITS1 and ITS2 (Figure 5). We included in these experiments cells depleted of RPS15, a ribosomal protein required for nuclear export (15). With the 5′ITS1 probe, cells treated with scrambled siRNAs displayed a strong nucleolar labeling characteristic of the early pre-rRNA species, as well as a cytoplasmic signal indicating the presence in this compartment of 18S-E pre-rRNA, the last precursor to the 18S rRNA (Figure 5A). The labeling level of the nucleoplasm was comparatively weak, attesting rapid trafficking of pre-ribosomal particles through this compartment. In bystin-depleted cells, cytoplasmic labeling decreased whereas nucleoli were conspicuously brighter than in control cells, suggesting accumulation of pre-40S particles in this domain. In addition, nucleoplasmic foci were observed in ~50% of bystin-depleted cells, as previously reported after knockdown of RPS19 or RPS24 (44,45). Using probes complementary to the 5′ ETS, no significant differences were detected in bystin-depleted cells compared with control cells (Figure 5C): this suggests that transcription and early pre-rRNA maturation are not affected, which is in accordance with the northern blot results.
In hTsr1-depleted cells, nucleolar labeling was also more intense. In addition, fluorescence was detected in the whole nucleoplasm, similar to that observed in RPS15-depleted cells (Figure 5A), indicating that, in the absence of hTsr1, pre-ribosomal particles are retained in the nucleus. In parallel, the cytoplasm showed almost no fluorescence. This change in pre-40S particle distribution was attested by quantification of the fluorescence signal (Figure 5B). Nucleoplasmic retention in cells depleted of hTsr1 was specific of the late pre-40S particles, as it was not observed with the probes complementary to the 5′ ETS (Figure 5C).
To strengthen these data, we performed subcellular fractionation followed by northern blot analysis (Figure 6A). Consistent with FISH, the nuclear fraction was enriched in 18S-E pre-rRNA upon hTsr1 knockdown (Figure 6B). In bystin-depleted cells, the nuclear fraction was enriched with 41S, 21S and 21S-C pre-rRNAs, whereas the level of 18S-E pre-rRNA decreased, both in the nuclear and in the cytoplasmic fractions. However, the nuclear and cytoplasmic distribution of the 18S-E pre-rRNA still produced in bystin-depleted cells did not change (Figure 6B). These data show that assembly of hTsr1 is required for normal trafficking of pre-40S particles from the nucleus to the cytoplasm.
The data presented here indicate that bystin and hTsr1 are required at distinct steps of the pathway: bystin is necessary for efficient cleavage of the ITS1 at site 2 and for further processing of the pre-rRNA; hTsr1 is involved downstream in nuclear export of the pre-40S particles and potentially in final cytoplasmic processing.
Yeast Enp1p is a nuclear protein, concentrated in the nucleolus. This pre-ribosomal factor is required for the earliest cleavages at sites A0, A1 and A2 of the 35S pre-rRNA. Enp1p associates with pre-ribosomes from the 90S particle to the cytoplasmic pre-40S particles (17). Depletion of Enp1p affects cleavage of the ETS1 at sites A0 and A1 and blocks processing of the ITS1 at site A2. Alternative cleavage in the ITS1 at site A3 generates the 23S and 21S aberrant intermediates, whereas 20S pre-rRNA production is blocked (21).
Here, we find that the bystin/hEnp1 human protein is also implicated in pre-40S particle maturation. It is largely located in the nucleus and accumulates in the nucleolus, but it also accompanies the pre-40S particles to the cytosol. Our results confirm previous studies on the intracellular localization of bystin, its presence in pre-40S particles (26,40), and its involvement in ITS1 maturation (38,40). We extend these observations by showing that depletion of bystin leads to an accumulation of the 41S and 21S pre-rRNAs, which we interpret as a delay in ITS1 cleavage at site 2 and a blockade of ITS1 processing in the pre-40S particles. In addition, a short form of the 21S pre-rRNA accumulates. Given that this species is also detected at low levels in control cells, we think that it could be a short-lived intermediate between the 21S and 18S-E pre-rRNAs. Interestingly, a similar pre-rRNA is observed in human cells upon knockdown of ribosomal protein RPS19 (43), a protein whose depletion in yeast also results in cleavage at site A3 instead of A2 and strongly blocks ITS1 maturation in pre-40S particles. The presence of 21S-C pre-rRNA could indicate an alternative endonucleolytic cleavage in the ITS1 preferentially used upon deficient cleavage at site 2. However, we did not detect any cleavage product complementary to the 21S-C pre-rRNA by northern blot. In addition, the heterogeneity of the RNA 3′ extremities found in this region by 3′-RACE experiments, together with the fuzzy appearance of this band on northern blots, rather suggest that an exonuclease trims the 21S pre-rRNA after cleavage at site 2. We cannot exclude that this exonuclease is part of a degradation process triggered by the accumulation of 21S pre-rRNA in bystin-depleted cells. However, the presence in control cells of intermediates similar to those found upon bystin knockdown rather supports that this exonucleolytic processing is part of the 21S to 18S-E pre-rRNA conversion and constitutes an additional step in the maturation of the human pre-40S particles.
Accumulation of 21S pre-rRNA as well as disappearance of 36S pre-rRNA upon bystin depletion indicate that bystin is required for cleavage at site E. Processing at site E also requires some RPS proteins, especially RPS18 and RPS19 (46). That RPS19 and bystin depletion yields similar phenotypes may indicate that RPS19 is required for recruitment of bystin to pre-40S particles, as observed in yeast (47). Interestingly, Enp1p and Rps19p in yeast are necessary for ITS1 cleavage at site A2. This draws an interesting parallel between yeast site A2 and human site E, although they are, respectively, positioned at 217nt and ~24nt from the 18S rRNA 3′-end. Similar to site A2, cleavage at site E seems to be dependent on processing of the 5′-ETS, as seen upon depletion of some RPS proteins (45,46). Human site 2 in turn is independent of 5′-ETS processing, like site A3 in yeast.
The role of bystin in ITS1 processing and its association with pre-40S particles indicate that it is a true functional ortholog of yeast Enp1p. However, bystin does not complement a yeast enp1 null mutant (21). This may be due to the strong divergence of the N-terminal domains of these proteins, which could be required for species-specific functions. In addition, we find no evidence that bystin/hEnp1 is implicated in 5′-ETS maturation, which may be a significant difference in the functions of these proteins.
Yeast Tsr1p localizes to the cytoplasm and the nucleus, including the nucleolus. It associates specifically with pre-40S particles, as indicated by its sole interaction with 20S pre-rRNA and pre-40S-specific factors (27,29). Tsr1p (Twenty-S rRNA) was named due to the large accumulation of 20S pre-rRNA resulting from depletion of this factor (22,29); when detected by FISH, pre-40S particles are found to pile-up in the cytoplasm (29).
The 40S/60S subunits imbalance detected here on sucrose gradients, confirmed by the reduction of the 18S/28S ratio, argues for a major role of hTsr1 function in 40S subunit production. However, unlike in yeast, we observed no major accumulation of pre-rRNAs in the cytoplasm of HeLa cells treated with tsr1 siRNAs, but instead nucleoplasmic retention of pre-40S particles. This discrepancy may correspond to different coordination between pre-rRNA maturation and nuclear export of pre-40S particles in yeast and mammalian cells. Indeed, production of 18S rRNA requires two pre-rRNA processing steps in human pre-40S particles instead of one in yeast. Our results indicate that hTsr1 is not necessary for the first step, i.e. conversion of 21S to 18S-E pre-rRNA. By analogy with yeast Tsr1p, hTsr1 may thus be involved in the final cytoplasmic processing step. But prior to this second processing step, assembly of hTsr1 is required for efficient nuclear export. Whether hTsr1 has a direct role in pre-ribosomal transport to the cytoplasm, for example, as an exportin adaptor, remains to be demonstrated. We found no leucine-rich nuclear export sequence in hTsr1 by computer search, unlike recently shown for hRio2 (26). This does not formally exclude the presence of such a sequence, given the weak consensus by which it is defined. The absence in the mammalian orthologs of Mex67p and Mtr2p of a loop critical for the function of these factors in yeast pre-60S particle nuclear export illustrates that pre-ribosomal transport has evolved between yeast and mammals (48). Alternatively, hTsr1 may be more indirectly required for correct assembly or folding of the pre-40S particles prior to nuclear export. The fact that 18S-E pre-rRNA does not accumulate in hTsr1-depleted cells suggests that hTsr1-deficient pre-40S particles are very labile and more rapidly degraded in mammalian cells than in yeast.
In conclusion, these data illustrate that, although the pre-40S particle-specific factors have clear functional homologies, maturation and nuclear export of these particles are not governed by strictly similar mechanisms in yeast and human cells. We propose that this is linked to differences in ITS1 processing, which change the timing between the maturation of the particle and its competence for nuclear export. In addition, the potential additional pre-rRNA processing step revealed upon knockdown of bystin or RPS19 indicate that nuclear processing of the ITS1 in mammalian cells may be even more complex than currently envisioned.
CNRS, the University of Toulouse and the French National Research Agency (ANR – RIBEUC project); P.E.G. is a fellow of the Institut Universitaire de France; doctoral fellowship from the Ministère de l’Enseignement Supérieur et de la Recherche and from the Association pour la Recherche sur le Cancer (to C.C.). Funding for open access charge: CNRS.
Conflict of interest statement. None declared.
The authors wish to thank Guillaume Stahl, Célia Plisson-Chastang and Milena Preti for scientific discussions, as well as Ulrike Kutay and Sébastien Fribourg for sharing reagents. They are also indebted to Caroline Monod for proofreading the manuscript. C.C. designed and performed most of the experiments, M.-F.O’D. and V.C. and M.F. performed some of the experiments. P.-E.G. supervised the project. C.C., M.-F.O’D. and P.-E.G. were involved in writing the manuscript.