|Home | About | Journals | Submit | Contact Us | Français|
Given the high metabolic cost required to generate ribosomes, it has been assumed that proteins involved in ribosome synthesis might establish functional cross talk with other intracellular processes to efficiently couple ribosome production and cell growth. However, such interconnections have remained elusive due to the difficulty in separating the intra- and extraribosomal roles of ribosome biogenesis factors. Using a yeast functional screen, I have discovered that Rrp12, a conserved protein involved in ribosome maturation and export, plays roles in the cell cycle and the DNA damage response. These results indicate that Rrp12 participates in a karyopherin Kap121-dependent import route that is crucial for nuclear sequestration of ribonucleotide reductase subunits and, thereby, ensures the proper kinetics of deoxyribonucleotide production during the cell cycle. Within this route, Rrp12 acts as a cofactor important for the full functionality of Kap121. This activity is mechanistically different from the known roles of Rrp12 in ribosome biogenesis. I propose that the functional duality of Rrp12 may couple the control of ribosome production to the regulation of other cellular processes during cell cycle progression.
Eukaryotic ribosome biogenesis is a highly conserved process that initiates in the nucleolus with the synthesis of a polycistronic pre-rRNA and ends in the cytoplasm after export and final maturation of the 40S and 60S ribosomal subunits. Most of the current knowledge of the ribosome biosynthetic pathway has emerged from genetic and biochemical analysis of Saccharomyces cerevisiae, which allowed the characterization of different preribosomal particles that mediate the processing of pre-rRNAs and the assembly of ribosomal proteins onto those RNAs (15, 19). The pathway starts with the transcription of the 35S pre-rRNA by RNA polymerase I and the subsequent association of maturation factors and ribosomal proteins to form the first ribosome precursor, the 90S preribosome. Cleavage in the spacer region between the 18S and the 5.8S rRNAs leads to the split of the 90S preribosome into a pre-40S particle and a pre-60S particle. Those particles then follow two separate maturation routes through the nucleolus and the nucleoplasm before export to the cytoplasm (15, 19).
The complexity of the ribosome synthesis pathway has been underscored by the identification of ~200 nonribosomal factors that are essential for the formation and maturation of preribosomal particles (13, 15, 19, 47). Recent studies have unveiled the specific roles of some of those factors in several steps of ribosome biosynthesis, including the processing/posttranscriptional modification of pre-rRNAs, the assembly of preribosomes, the positioning of ribosomal proteins, the removal/recycling of other biogenesis factors, the incorporation of the 5S ribonucleoprotein, and the export of the pre-40S and pre-60S particles (15, 19, 51). In addition to their intrinsic functions in ribosome biogenesis, experimental evidence indicates that some of the preribosomal factors can participate in extraribosomal functions. Thus, several studies of both yeast and higher eukaryotes have revealed the existence of cross talk mechanisms between the ribosome biosynthesis and cell cycle machineries that are required for committing to the cell cycle, setting cell size during division, and self-monitoring cell homeostasis (6, 16, 46). For example, it has been described for budding yeast that the functional activity of the ribosome biogenesis machinery influences both the regulation of the cell size setpoint and the promotion of Start (2, 16, 17). The study of these cross talk mechanisms may have clinical interest, because mutations in some ribosomal proteins and ribosome biogenesis factors cause human diseases such as Diamond-Blackfan anemia and the Shwachman-Diamond syndrome (29). There is also a large body of experimental evidence suggesting that some specific preribosomal factors are directly involved in prima facie ribosome-independent processes. Thus, the 60S subunit synthesis factor Sda1 has been implicated in the passage of yeast cells through the Start cell cycle transition (56). Yph1/Nop7 and Noc3, two pre-60S particle elements, have been shown to associate with the origin of replication complex (ORC) and favor S-phase entry (9, 52). One pre-60S (Rrp14) and two 90S (Utp6 and Utp7) preribosomal factors play roles during mitosis in the positioning of the mitotic spindle (Rrp14), centrosomal duplication (Utp6), and chromosome segregation (Utp7) (11, 18, 31). Finally, Nop15, a protein associated with early pre-60S particles, has been implicated in cytokinesis (32).
Despite the interest of the above findings, it is important to note that a common issue in the functional analysis of ribosome synthesis proteins is to distinguish direct effects exerted by the elimination of the studied factor from others that are just collaterally or epistatically linked to deficient ribosome production or protein translation. This is because in most studies the proteins under analysis are depleted in a slow and gradual manner, thus making it difficult to formally establish the cause of the resulting phenotype. To circumvent this problem, I decided to use the yeast degron system to screen for defects taking place immediately after the elimination of a specific preribosomal protein. In this system, a specific gene is modified at its 5′ end by inserting ectopic DNA sequences that encode a temperature-sensitive version of dihydrofolate reductase bearing an amino-terminal arginine residue (DHFRts) (37). Such modification leads to the expression of a DHFRts fusion protein that can be targeted very rapidly (≈1 h) for degradation by shifting the culture temperature from 25°C to 37°C. Since the consequences of ribosome loss usually require one or two rounds of division to develop, this experimental system is optimal for identifying cellular functions that are affected just after the degradation of a given ribosome synthesis factor. Using this strategy for a collection of preribosomal proteins, I have discovered that Rrp12, a stable component of both pre-40S and pre-60S particles that mediates ribosomal subunit maturation and export (30, 38, 49), also plays important roles in cell cycle progression and the response to DNA damage. Here, I report the specific implication of Rrp12 in these biological processes.
All yeast strains used in this study are listed in Table 1. Strains with MYC or green fluorescent protein (GFP) carboxy-terminal-tagged alleles were generated by one-step integration of PCR-amplified cassette sequences. All the conditional degron strains used for cell cycle progression studies are from the collection of K. Labib (Paterson Institute for Cancer Research, Manchester, United Kingdom) and were obtained from the Euroscarf archive. The strains referred to in the text as the utp5-td, utp9-td, utp18-td, utp22-td, rrp12-td, nsa1-td, and rix2-td temperature-inducible degron mutants and the CUP1-UBI4-RRP12 mutant correspond to strains Y40033, Y40060, Y40069, Y40046, Y40114, Y40043, Y40061, and YMD356 in Table 1, respectively; the control strain refers to strain Y44020. To delete SML1 at its genomic locus, a PCR was performed using flanking oligonucleotides for the SML1 open reading frame and using genomic DNA from a sml1Δ strain (Euroscarf) as a template.
To generate plasmids pMD31, pMD38, and pMD39 (p413GAL1-RRP12-MYC, p415GAL-RRP12-MYC, and p416GAL1-RRP12-MYC), the RRP12-MYC sequence was PCR amplified as an AvrII-XhoI fragment, using as a template genomic DNA from strain YMD238, and inserted into XbaI-XhoI p413GAL1, p415GAL1, and p416GAL1 (28), respectively. To generate plasmids pMD33 and pMD34 (p415GAL1-RRP12-GFP and p415GAL1-rrp12-td-GFP), the RRP12-GFP and rrp12-td-GFP sequences were PCR amplified as AvrII-XhoI fragments from strains YMD375 and YMD380 and cloned into XbaI-XhoI p415GAL1. Plasmid YMD37 (p415GAL-TSR1-GFP) was generated by cloning an AvrII-XhoI TSR1-GFP fragment, obtained by PCR amplification from pJB1 (a gift from S. Lemmon ), into XbaI-XhoI p415GAL1. Plasmids pDL132 (CEN, URA3, and GAL1-RNR4), pGFP2-C-FUS-H31-28 (CEN, URA3, and GFP-tagged histone H3 NLS), and pRS315-NOP1-GFP-ULP1 (CEN, LEU2, and GFP-ULP1) were provided by S. Elledge, L. Pemberton, and V. Panse, respectively (22, 27, 33). Plasmids pPS1069 (CEN, TRP1, and KAP121-GFP) and PS1070 (CEN, TRP1, and KAP123-GFP) were obtained from P. Silver (39), and plasmids pGAL-NOP1-GFP (CEN, URA3, and GAL1-NOP1-GFP) and pRS316-KAP121 (CEN, URA3, and KAP121) from J. Aitchison (23, 24).
Where indicated, α-factor was added to 7.5 mg/ml, hydroxiurea (HU) to 200 mM, and nocodazole to 15 mg/ml. Degron strains were maintained generally at 25°C in rich medium containing 2% glucose and 0.1 mM CuSO4 (yeast extract-peptone-dextrose [YPD]-Cu). The experiments involving the rrp12-td strains (see Fig. 1C [bottom], [bottom],2D2D [middle] and I, I,3B,3B, D, and F, F,4,4, ,5D5D and E, E,6A6A to F and H, H,7,7, and and8)8) were performed under permissive conditions (YPD-Cu, 25°C). The cell cycle progression studies, polysome analysis, and Western blot experiments (see Fig. 1B, C [top], D, and E, E,2C,2C, E, and F, and and3C3C and E) involving the rrp12-td strain and other degron strains required the depletion of DHFRts-fused proteins by shifting cultures to restrictive conditions (YP-Gal, 37°C). To deplete the proteins in asynchronous cultures, cells were grown at 25°C in 1.5% raffinose-0.5% glucose and 0.1 mM CuSO4 (YPRaf-Cu) and transferred to 2% galactose and 0.1 mM CuSO4 (YP-Gal-Cu) for 1 h at 25°C, before changing them to prewarmed YP-Gal and incubating them at 37°C for 1 to 3 h. For analysis of cell cycle progression upon protein depletion in G1 phase, cells were synchronized with α-factor in YP-Gal-Cu at 25°C for 3 h and shifted to YP-Gal containing α-factor for 1 h at 37°C before washing and releasing them into YP-Gal at 37°C. To monitor the effects of protein depletion in S phase or early mitosis, cultures were treated with α-factor in YP-Gal-Cu at 25°C for 3 h, changed to YP-Gal-Cu containing either hydroxyurea or nocodazole for 1 h at 25°C, transferred to YP-Gal in the presence of hydroxyurea or nocodazole for 1 h at 37°C, washed, and released into YP-Gal at 37°C. For fluorescence-activated cell sorter (FACS) analysis, cells were fixed for 1 h in 70% ethanol, centrifuged, resuspended in 50 mM sodium citrate, and incubated with 200 mg/ml RNase A overnight at 37°C. Before analysis in the FACS scan, cells were spun down, resuspended in 50 mM sodium citrate with 1 mg/ml propidium iodide, and sonicated.
Total RNAs from yeast cells were prepared by the hot-phenol method (1). Analysis of RNA integrity and quantitation of 25S/18S ratios and of rRNA relative contents were performed on an Agilent bioanalyzer. Northern blot analysis was performed as previously described (8), and pre-rRNA contents were quantitated using ImageJ 1.140 (National Institutes of Health).
Yeast cell samples were normalized to an optical density at 600 nm (OD600), and protein extracts were prepared by glass bead disruption using a FastPrep homogenizer (Qbiogene). Cells were lysed either in 20% trichloroacetic acid (12) or in IP2 buffer (20 mM Tris-HCl [pH 8.0], 5 mM MgCl2, 150 mM potassium acetate, 1 mM dithiothreitol, 0.2% Triton X-100, supplemented with Complete) (34). Antibodies used in Western blot analysis were anti-Mcm2, anti-Mcm7, and anti-Rad53 goat polyclonal antibodies (Santa Cruz Biotechnology), anti-histone H3 rabbit polyclonal antibody (Abcam), anti-Sml1 rabbit polyclonal antibody (a gift from Rodney Rothstein, Columbia University), anti-GFP mouse monoclonal antibody (Clontech Laboratories), anti-Nop1 monoclonal antibody (EnCor Biotechnology), anti-Cdc11 rabbit polyclonal antibody (Santa Cruz Biotechnology), and anti-Cdc45 rabbit polyclonal antibody (a gift from Bruce Stillman, Cold Spring Harbor Laboratory).
Polysome analysis and fractionation of lysates through 7 to 50% sucrose gradients were performed as described previously (8). Extract equivalents to 15 absorption units (A260) were layered onto each gradient. After centrifugation, gradients were processed in a gradient collector apparatus (Brandel) that allowed the simultaneous reading of absorbance (A254) and the collection of fractions that were subsequently analyzed by Western blotting.
Fluorescence microscopy was performed using either a Zeiss Axioplan 2 microscope fitted with a 63× objective, a Hamamatsu ORCA-ER digital camera and Openlab (Improvision) cell imaging analysis software, or a Leica DMI-6000 confocal microscope equipped with a 63× objective, a Leica TCS-SP5 scanning system, and Leica LAS-AF software. For in vivo localization of GFP-tagged proteins, yeast strains were grown in liquid medium to an OD600 of 0.5 to 0.7 before being mounted on a microscope slide. The distribution of the GFP-tagged protein was considered to be nuclear only when cells presented a very bright epifluorescent signal concentrated within a round-lobate structure of ~1.5 mm in diameter. In the nuclear import assays of GFP-Rnr4 in rrp12-td cells, confocal images were used to measure fluorescence intensities across 4-mm sections of individual cells. All images were background subtracted before analysis. For visualization of MYC-tagged proteins by indirect immunofluorescence, 1 × 108 cells were fixed in 4% formaldehyde at room temperature for 1 h, washed twice with 1 ml PEM buffer (100 mM PIPES [pH 6.9], 1 mM EGTA, 1 mM MgSO4), resuspended in 1 ml of PEM-1.2 M sorbitol, and incubated with 25 mg/ml zymolyase-20T (Seikagaku) and 0.02% β-glucuronidase (Sigma) for 20 min at 37°C. After washing cells twice in PEM-1.2 M sorbitol, cells were permeabilized in 1 ml PEMS-1% Triton X-100 for 1 min at room temperature. The incubations with antibodies and in-between washes were done in PEM-1% bovine serum albumin (BSA). The primary 9E10 anti-MYC antibody (Roche) was diluted to 1:2,000 and incubated with cells at 4°C for 12 h. The secondary Cy3-conjugated goat anti-mouse antibody (Jackson ImmunoResearch) was diluted 1:500 and incubated with cells for 1 h at room temperature. DNA was stained with 0.1 mg/ml 4′,6-diamidino-2-phenylindole (DAPI), and samples were mounted for imaging in 90% glycerol and 1 mg/ml p-phenylenediamine (Sigma).
Purification of Rrp12-GFP, Rrp12-td-GFP, Kap121-GFP, and Kap123-GFP was performed by immunoprecipitation with GFP-Trap (Chromotek). Cells with an OD600 equivalent to 40 were lysed with glass beads in 0.5 ml IP2 buffer supplemented with Complete, using the FastPrep homogenizer. Precleared lysates were incubated with 25 ml of GFP-TRAP beads at 4°C for 2 h and washed three times with IP2 buffer at 4°C. For mass spectrometry analysis, the immunoprecipitated material was resuspended in 30 ml of SDS loading buffer, boiled, and separated on 8% or 6.5% acrylamide gels. The gels were stained with silver nitrate, and the protein bands were processed and analyzed for their identification by mass spectrometry in the Genomics and Proteomics Facility of the Centro de Investigación del Cáncer de Salamanca, as previously described (34). For the coimmunoprecipitation experiment (see Fig. 6G), GFP-tagged proteins were purified following the same procedure as that for mass spectrometry analysis, but only one-third of the purified material was used for analysis by Western blotting.
Isolation and quantification of nucleotides from yeast was done essentially as described previously (4, 42). In brief, 6 × 108 cells were harvested by filtration on 25-mm Metricell membrane filters (0.8 mm; Life Sciences). The filters were immersed in 700 ml of cold EB buffer (12% trichloroacetic acid, 15 mM MgCl2) and allowed to stand 15 min on ice with occasional vortexing. After centrifugation, supernatants were transferred to a new tube and extracted twice with 800 ml of a mixture of Freon (78%)-trioctylamine (22%). One-tenth of the sample was used directly for quantification of nucleoside triphosphates (NTPs). The rest of the sample was adjusted to 15 mM MgCl2 and to pH 8.9 with ammonium bicarbonate, and the deoxyribonucleotides were separated from the ribonucleotides by boronate affinity chromatography using an Affi-Gel-601 column (Bio-Rad). Flowthrough fractions of 1 ml were collected and adjusted to pH 5.0 with 85% phosphoric acid before high-pressure liquid chromatography (HPLC) analysis. Nucleotide separation by HPLC was performed on a 4.6- by 200-mm PolyWAX LP column (5-mm particle size; PolyLC, Inc.) using a gradient (20 to 98%) of mobile phase B (7.2 M ammonium phosphate, 2.5% acetonitrile [pH 5.0]) over 35 min and was followed by an elution with 98% mobile phase B and mobile phase A (2.5% acetonitrile) over 20 min. Quantifications were performed by determining all peak areas and comparing them to a standard curve.
I initially chose 24 degron strains to analyze the possible implication of preribosomal factors in cell cycle events. Preliminary experiments led me to disregard 17 of the strains because they either showed severe growth defects at the permissive temperature or did not get efficiently arrested in cell synchronization experiments. This showed that a large proportion of degron strains for ribosome synthesis factors exhibit inadequate expression levels or defective functionality of the DHFRts fusion proteins. The seven remaining strains, which corresponded to factors Utp5, Utp9, Utp18, Utp22, Rrp12, Nsa1, and Rix1, presented normal growth rates and generated normal ratios and amounts of 25S and 18S rRNAs at 25°C (data not shown). All seven strains displayed the expected behavior and induced an unbalanced production of 25S/18S rRNAs at 37°C (data not shown). Such changes were consistent with the specific role of the targeted protein factor in the biogenesis of preribosomal 40S and 60S particles (7, 13, 15, 19). Thus, while degron strains for Utp5, Utp9, Utp18, Utp22, and Rrp12 exhibited defects in the synthesis of the 18S rRNA, those for Nsa1 and Rix1 showed a specific impairment in the generation of the 25S rRNA when shifted to 37°C (data not shown).
To investigate the implication of those seven preribosomal factors in cell cycle progression, I first studied how their respective degron strains progressed through S phase. To this end, I arrested the strains in early S phase with hydroxyurea (HU) and, upon removal of the drug, monitored the DNA content of each cell population when cultured at either the permissive or nonpermissive temperature (the scheme in Fig. 1A). I restricted the analysis to the first 120 min after the temperature shift because during that time frame the cells did not exhibit significant defects in ribosome content (data not shown). Six out of the seven selected degron strains (the utp5-td, utp9-td, utp18-td, utp22-td, nsa1-td, and rix1-td mutants) progressed through S phase with kinetics comparable to that of the control strain, showing peaks of maximal DNA synthesis at 15 to 30 min after HU removal at the restrictive temperature (Fig. 1B and C; also data not shown). In contrast, the degron strain for Rrp12 (the rrp12-td mutant) exhibited a clear delay in S-phase progression when cultured at 37°C and, to a minor extent, when incubated at the permissive temperature (Fig. 1B and C). This was not due to collateral problems derived from the rapid development of a ribosome deficiency, because the relative contents of the two rRNA species in rrp12-td cells were comparable to those shown by degron strains that did not exhibit any S-phase transition defect (Fig. 1D). In fact, the relative levels of 35S and 20S pre-rRNA species were similar in rrp12-td and control cells (Fig. 1E and data not shown), indicating that the rrp12-td strain exhibited normal ribosome biogenesis activity. These data indicate that Rrp12 is the only preribosomal factor analyzed that has an influence in S-phase progression.
I then investigated the behavior of the degron strains in both S-phase entry and mitosis. To analyze the former process, cells were arrested in G1 with α-factor, maintained at either 25°C or 37°C for 1 h with α-factor, and then released into α-factor-free medium at both the permissive and nonpermissive temperature conditions (for a scheme of this experiment, see Fig. 2A). To analyze the latter process, I arrested cells in G2/M at 25°C using nocodazole, maintained them at 25°C or 37°C for 1 h with nocodazole, and then followed entry in mitosis in cultures kept at either 25°C or 37°C (the scheme in Fig. 2B). These analyses indicated that the rrp12-td strain had a defect in S-phase entry (Fig. 2C) and a milder but significant delay in passage through mitosis (Fig. 2E). The S-phase entry defect of rrp12-td cells was also observed, although less pronounced, at the permissive temperature (Fig. 2D). I also found an rrp12-td mutant-like G1-S transition delay in the case of the utp5-td degron strain (Fig. 2C). The observed cell cycle progression delays were specific for those two proteins, because I did not find any alteration in the degron strains for the other preribosomal factors studied in these experiments (Fig. 2C to E). Because it has been reported previously that cells can sense abnormal accumulations of pre-rRNAs and delay their passage through Start (2), I explored the possibility that the defects in S-phase entry of rrp12-td and utp5-td cells were due to an altered content of pre-rRNA species. Northern blot analysis showed no major changes in the relative levels of 20S or 35S pre-rRNAs in rrp12-td cells relative to control cells 45 min after being released from the G1 block (Fig. 2F and data not shown). Thus, the G1-S delay in rrp12-td cells cannot be attributed to an accumulation of pre-rRNAs at the Start checkpoint. In contrast, the utp5-td strain presented clear reductions in the 20S/35S ratio and pre-rRNA content (Fig. 2F and data not shown), which were indicative of defects in pre-rRNA synthesis and processing. However, those alterations cannot be the sole cause of the G1-S delay, because the utp18-td strain also displayed an impairment in pre-rRNA processing without showing any detectable defect in cell cycle progression (Fig. 2C and F). Taken together, these results indicate that the depletion of both Rrp12 and Utp5 induces defects in S-phase entry and that, in the case of Rrp12-deficient cells, such dysfunctions cannot be attributed to alterations in the content of pre-rRNAs.
Some yeast mutants with problems in S-phase progression also display reduced tolerance to DNA damage insults. This led me to investigate whether the rrp12-td yeast strain had an altered sensitivity to genotoxic agents. As the growth assays required to test sensitivity to DNA damage entail longer culture periods than those needed to visualize cell cycle progression, I resorted to the use of “transcriptional shutoff” to bypass the rapid loss of viability associated with the total elimination of Rrp12. For this purpose, I exploited the fact that in the degron strains the expression of the DHFRts fusion proteins is under the regulation of the Cu2+-dependent CUP1 promoter. Due to this, I could induce a progressive loss of DHFRts fusion proteins in yeast cells when cultured in the absence of Cu2+ at the permissive temperature. Under such experimental conditions, I observed that rrp12-td cells were more sensitive than wild-type and utp5-td cells to treatments with HU or methyl methanesulfonate (MMS) (Fig. 2G), suggesting that Rrp12 does influence the tolerance of yeast cells to DNA damage. As in the cell cycle experiments, I noticed that rrp12-td cells were hypersensitive to HU and MMS even when Cu2+ was present in the cultures and, therefore, when the synthesis of the DHFRts-Rrp12 fusion protein was not blocked (Fig. 2G). This defect was caused by the modifications present at the RRP12 locus in rrp12-td cells because a single-copy vector expressing RRP12 under the control of its own promoter restored the tolerance to DNA damage (Fig. 2G). Furthermore, the DNA damage sensitivity phenotype could be reproduced when the RRP12 gene was replaced by the CUP1-UBI4-rrp12 allele in a wild-type strain (Fig. 2G). These results suggest that the DHFRts-Rrp12 fusion protein has intrinsic defects that affect some cell cycle/DNA damage response-related functions but not the previously known roles in ribosomal biogenesis.
To examine whether the hypersensitivity to DNA-damaging agents was caused by a defective activation of the DNA damage checkpoint, I released cells from an α-factor-mediated G1 block in the presence of HU at the permissive temperature (Fig. 2H) and monitored the phosphorylation kinetics of Rad53, a serine/threonine kinase that mediates cellular responses to DNA damage (3). As a second experimental parameter, I tracked down the disappearance of the ribonucleotide reductase inhibitor Sml1, a known target of the Mec1/Rad53/Dun1 kinase cascade that is degraded in response to DNA damage (54). Whereas wild-type and utp5-td cells responded to this treatment with the expected kinetics of Rad53 phosphorylation and Sml1 degradation (Fig. 2I), rrp12-td cells were defective in those responses even after being exposed to HU for up to 2 h at the permissive temperature (Fig. 2I). Compared to the other strains used in these experiments, I also observed that rrp12-td cells had higher than normal levels of Rad53 (Fig. 2I, compare lanes for time zero in the control and rrp12-td strains). Rad53 levels also increased ≈2-fold in asynchronously growing rrp12-td cells (Fig. 2J). A similar defect in Rad53 accumulation and the lack of activation were observed when the rrp12-td strain was analyzed under nonpermissive culturing conditions (data not shown). These results, combined with those shown in Fig. 1 and and2,2, indicate that Rrp12 has ribosome biogenesis-independent functions that are important for S-phase entry, S-phase progression, mitosis, and the DNA damage response. Given that those properties were not displayed by the rest of preribosomal factors interrogated in this work, I decided to focus subsequent studies exclusively on Rrp12.
The defects in cell cycle progression and response to DNA damage exhibited by the rrp12-td strain under permissive conditions indicated to me that (i) the DHFRts-Rrp12 protein was impaired in some of its biological functions and (ii) those functions were not related to its ascribed role in ribosomal biogenesis, because the rrp12-td strain displayed no detectable defects in the production of pre-rRNAs and rRNAs (see above; Fig. 1 and and2;2; also data not shown). These observations prompted me to further compare the structural and biological properties of the wild-type and DHFRts fusion versions of this protein. First, I observed by sequencing the degron cassette that the DHFRts-Rrp12 fusion protein had lost the leucine residue located at position 9 of the Rrp12 primary structure, probably due to a mutation fortuitously generated during the construction of the degron cassette (Fig. 3A). This was not the cause of the cell cycle-related defects of rrp12-td cells, because the expression of an Rrp12 protein lacking that residue could not recapitulate the phenotype of the rrp12-td strain (data not shown). In addition, I found that the relative amount of the DHFRts-Rrp12 protein in the rrp12-td strain was smaller than the amount of the endogenous Rrp12 in the control cells (Fig. 3B). Similar differences in protein content were observed using GAL1-driven GFP-tagged versions of both proteins, indicating that DHFRts-Rrp12 was expressed at lower levels than wild-type Rrp12 (Fig. 3B). Despite such reduced expression levels, rrp12-td cells could induce normal levels of pre-rRNA intermediaries (Fig. 1E and and2F2F and data not shown), rRNAs (see levels at 25°C in Fig. 3D and data not shown), and polysomes (Fig. 3C). Consistent with this, I also observed that they could synthesize normal amounts of a number of endogenous proteins under permissive conditions (Fig. 3E). I also tracked down the incorporation of Rrp12-GFP and DHFRts-Rrp12-GFP onto preribosomal complexes using sucrose gradient fractionation experiments. As previously described (38), I could detect Rrp12-GFP sedimenting with 40S/pre-40S particles (Fig. 3F, top, fraction 10), preribosome-enriched 90S particles (Fig. 3F, top, fractions 12 and 13), and preribosome-free 15- to 20S fractions (Fig. 3F, top, fractions 4 to 6). The sedimentation profile of DHFRts-Rrp12 was similar to that shown by its wild-type counterpart (Fig. 3F, bottom), although the relative amount detected in each fraction was much lower in the former case due to the diminished expression levels of DHFRts-Rrp12 (see above; Fig. 3B). Finally, I found that the wild-type and Rrp12 fusion proteins displayed similar subcellular distributions (Fig. 3G). Taken together, the results indicate that Rrp12 is required for two different cellular processes, one of them related to ribosome biosynthesis and the other one to cell cycle progression and DNA damage responses. They also suggest that the functions of Rrp12 in these different processes have been uncoupled or are differentially affected in the DHFRts-Rrp12 protein present in the rrp12-td strain.
The defects in S-phase progression found in rrp12-td cells are reminiscent of those previously reported for mutants of Nop7 and Noc3, two ribosome biogenesis factors that interact with the DNA replication machinery (9, 52). However, unlike Nop7 and Noc3, Rrp12 does not interact with ORC proteins and does not play any role in the assembly or stability of prereplicative complexes (data not shown). To shed light on the cell cycle-related functions of Rrp12, I decided to identify the spectrum of wild-type Rrp12- and DHFRts-Rrp12-interacting proteins using mass spectrometry. To this end, I used yeast strains encoding the above protein versions with a C-terminal GFP tag, thus facilitating the purification of protein complexes using the GFP-Trap technique (36). Since I expected that the purification of the bulk of those proteins would yield a high proportion of preribosomal factors that would obscure the detection of proteins involved in other functions, I decided to characterize the Rrp12 binding proteins present in unfractionated Rrp12 preparations and, at the same time, in Rrp12 complexes obtained from yeast lysates depleted of high-molecular-weight preribosomal complexes. Consistent with my a priori criteria, I observed that the Rrp12-GFP and DHFRts-Rrp12-GFP complexes contained proteins present in both 90S (Utp20, Rrp5, Kre33, and Utp4) and pre-40S (Tsr1, Rio1, Enp1, Nob1, and Rio2) preribosomal particles when purified from unfractionated lysates (Fig. 4A). Those complexes also contained karyopherin 121 (Kap121, also known as Pse1), the splicing factor Prp22, and a functionally uncharacterized protein encoded by the YJR141W gene. Those three proteins had not been identified previously as stable components of preribosomes. When Rrp12 proteins were purified from preribosome-free supernatants, I observed that only Kap121 remained stably associated with Rrp12 (Fig. 4B). Confirming the reliability of the Rrp12-Kap121 interaction, I could also identify the wild-type Rrp12 and DHFRts-Rrp12 in purified Kap121 complexes obtained from unfractionated extracts of Kap121-GFP-expressing cells (Fig. 4C). Another protein, the preribosome export receptor Arx1, was also reproducibly detected in Kap121 complexes (Fig. 4C). These experiments also showed that the complex formed by DHFRts-Rrp12 and Kap121 is either more efficiently established or more stable than those formed by its wild-type counterpart. Thus, despite its lower expression levels relative to those shown by the wild-type protein (see above; Fig. 3B), I observed that DHFRts-Rrp12 was present in larger amounts than wild-type Rrp12 in Kap121-GFP complexes (Fig. 4C). The interaction of Rrp12 with Kap121 was not due to a general tendency of the former protein to bind karyopherins, because parallel experiments indicated that the wild-type and DHFRts-tagged versions of Rrp12 did not interact with another karyopherin, Kap123 (Fig. 4D). To further verify that Kap121 was not a component of preribosomal complexes, I analyzed its sedimentation profile in sucrose gradient ultracentrifugation experiments. Kap121 distributed exclusively in polysome- and preribosome-free fractions (Fig. 3F, bottom), thus suggesting that the Rrp12-Kap121 interaction is unrelated to the ascribed roles of Rrp12 in ribosomal biogenesis.
Kap121 is known to work as a nuclear import factor and a nuclear envelope anchor for a selected number of cargos, such as histone H3 and the Ulp1 protease (27, 33). To investigate whether Rrp12 was another cargo of this karyopherin, I examined the subcellular localization of GFP-tagged versions of Rrp12, histone H3, and Ulp1 in pse1-1 cells, a yeast strain that carries a temperature-sensitive Kap121ts mutant protein that is functionally impaired at both permissive and nonpermissive temperatures (39, 40). I observed that, compared to control cells, pse1-1 cells growing at both the permissive and nonpermissive temperatures did not localize GFP-histone H3 and GFP-Ulp1 at their normal locations in the nucleoplasm and nuclear envelope, respectively (Fig. 5A). In contrast, the import of the de novo-induced Rrp12-GFP protein to the nucleolus and nucleoplasm did not change between control and pse1-1 cells under both permissive or nonpermissive conditions (Fig. 5A). These findings argue against Rrp12 being a passive cargo of Kap121.
Given the above results, I next investigated whether Rrp12 could play a regulatory and/or structural role within the Rrp12/Kap121 complex. If so, I hypothesized that the overexpression of Rrp12 might influence the phenotype of pse1-1 cells. Consistent with this idea, I observed that overexpression of wild-type Rrp12 could rescue the growth defects of pse1-1 cells at the nonpermissive temperature (Fig. 5B). The effects of Rrp12 on pse1-1 cell viability were similar to those obtained by the expression of wild-type Kap121 (Fig. 5B). Instead, I did not find any rescue effect when proteins involved in either ribosomal biogenesis (Tsr1 and Nop1) or nuclear trafficking (Kap123) events were used (Fig. 5B), further indicating that Rrp12 and Kap121 exhibit a specific functional relationship. The overexpression of Rrp12 also restored the normal subcellular localization of GFP-histone H3 in the nucleoplasm of pse1-1 cells (Fig. 5C). However, it could not correct the mislocalization of GFP-Ulp1 observed with those cells (Fig. 5C), suggesting that Rrp12 may influence only the carrier function of Kap121 toward specific cargo subsets.
Given the pernicious effects of DHFRts-Rrp12 at the permissive temperature (Fig. 1 and and2),2), I next investigated whether the expression of this protein could be blocking the normal function of Kap121. Consistent with this possibility, I observed that the Kap121-GFP protein showed lower expression levels in rrp12-td cells than in control cells at the permissive temperature (Fig. 5D). Furthermore, its subcellular localization changed significantly, depending on the type of Rrp12 protein expressed. Thus, in the presence of wild-type Rrp12, Kap121-GFP showed the expected localization at the nuclear rim (Fig. 5E). In contrast, in DHFRts-Rrp12-expressing cells, it was found distributed throughout the whole cell, including the vacuoles (Fig. 5E). These changes were not due to massive alterations in the structure of the nuclear envelope or the nuclear pore complexes, because the GFP-Ulp1 (data not shown) and MAb414-reactive FXFG nucleoporins (data not shown) kept their wild-type nuclear envelope and nuclear pore localizations in rrp12-td cells, respectively. These results suggest that the normal recycling and/or functionality of the Kap121/Rrp12 complexes is altered when the Rrp12 protein is replaced by its DHFRts-tagged version.
Based on the above results, I decided to investigate whether DHFRts-Rrp12 could impair the nuclear import of all or some Kap121 cargos. Against the idea of a general interference with Kap121 function, I observed that rrp12-td cells did not have major alterations in the subcellular localization of GFP-histone H3 or GFP-Ulp1 at permissive or nonpermissive temperatures (data not shown). However, it is known that histone H3 can be transported to the nucleus by other karyopherins that do not associate with either Rrp12 or DHFRts-Rrp12 proteins (i.e., Kap123). This led me to consider that DHFRts-Rrp12 could induce its cell cycle and/or DNA damage phenotypes by blocking the transport of factors exclusively imported into the nucleus by Kap121. Since the Kap121-specific cargos remain largely unknown, I decided to test if I could identify some of them by examining the subcellular localization in wild-type and rrp12-td cells of a selected number of proteins known to be transported into the nucleus in a cell cycle-regulated manner. To facilitate the direct evaluation of nuclear import defects, I used GAL1-GFP reporter plasmids that allowed me to induce the expression of the protein of interest a few hours before fluorescence microscope analyses. Using this approach, I identified Rnr4 as a potential cargo for the Kap121/Rrp12 complex. Rnr4 is one of the subunits of ribonucleotide reductase (RNR), the rate-limiting enzyme for dNTP synthesis in S. cerevisiae (45). Rnr4 and the other small RNR subunit, Rnr2, remain within the nucleus during G1 and G2/M. However, when cells enter S phase or when they are exposed to DNA damage or replicative stress, the Rnr2/Rnr4 heterodimer exits the nucleus presumably to associate with the cytosolic Rnr1 homodimer and form a catalytically competent RNR holoenzyme, leading to active dNTP production (21, 22, 48, 50, 53). Consistent with this known regulatory model, I observed that GAL1-driven GFP-Rnr4 accumulated in the nuclei of most control cells after a 5-h galactose induction at 25°C, although some diffuse epifluorescence signals were also detected the cytoplasm (Fig. 6A and B; data not shown). In contrast, when expressed in rrp12-td cells under the same culture conditions, GFP-Rnr4 showed little nuclear enrichment and was present in the nucleoplasm, punctate areas at the nuclear periphery, and the cytoplasm (Fig. 6A and B and data not shown). Quantification of the GFP signal by confocal microscopy further revealed the differences between the two strains, showing that the levels of enrichment of GFP-Rnr4 in the nucleus versus the cytoplasm were ≈3-fold and ≈1.4-fold in control and rrp12-td cells, respectively. Immunoblot analyses confirmed that GFP-Rnr4 was expressed at similar levels in the control and rrp12-td strains (Fig. 6C). This defect was specific for Rnr4, because the nuclear import of the preribosomal protein Tsr1 did not change regardless of the type of Rrp12 protein present in cells (Fig. 6A). Further proof of this specificity was evidenced by other experiments showing that the nuclear import of other ribosome synthesis-related (Nop1, Sof1, Rpl25, and Rps3) and cell cycle-related (Mcm4, Cdc14, and histone H4) proteins was normal in rrp12-td cells (data not shown). Using strains encoding Myc-tagged versions of the two members of the Rnr2/Rnr4 heterodimer, I could demonstrate that the expression of DHFRts-Rrp12 was accompanied by an increase in the total number of cells with cytoplasmic Rnr4 and Rnr2 (Fig. 6E). Such mislocalization was due neither to reduced expression levels of Rnr2 or Rnr4 (Fig. 6D) nor to an increase in the percentage of cells in S phase in rrp12-td cells (data not shown). In fact, 25 to 35% of G0/G1 unbudded cells in the rrp12-td strain had Rnr2-MYC or Rnr4-MYC delocalized to the cytoplasm (Fig. 6F). In contrast, those percentages ranged between 2 and 8% in control cells (Fig. 6F).
I could not detect a physical interaction of endogenous Rrp12 with either Rnr2 or Rnr4. However, under conditions of Rrp12 overexpression, I observed by the use of coimmunoprecipitation experiments that Rrp12 did associate with Rnr2 (Fig. 6G), but I could not find any detectable interaction of Rrp12 with Rnr4 in the same experiments (Fig. 6G), suggesting that Rrp12 might establish an interaction with the Rrn2/Rrn4 dimer through a direct association with Rnr2.
Since the export of the Rnr2/Rnr4 heterodimer from the nucleus to the cytosol is thought to increase the pool of catalytically competent RNR holoenzyme, I hypothesized that this enzyme could display higher activities in rrp12-td cells. Consistent with this, I observed that the steady-state levels of all dNTP species were 1.2- to 1.5-fold higher in exponentially growing cultures of rrp12-td cells than in the control strain (Fig. 6H). A modest, but reproducible, increase in the dNTP concentration was also detected in rrp12-td cells arrested in G1 (Fig. 6H), the phase of the cell cycle associated with the lowest levels of dNTPs. These results further indicate that Rrp12 plays a role in the nuclear localization of the small RNR subunits and the regulation of dNTP content in the cell.
Consistent with an implication of Kap121 in this process, I observed that GFP-Rnr4 did not efficiently accumulate in the nucleus in pse1-1 cells growing at the permissive temperature even with long induction times of the reporter protein (Fig. 6I). Instead, the nuclear and nucleolar localizations of the preribosomal Tsr1-GFP protein were similar in control and pse1-1 cells (Fig. 6I). As previously shown for histone H3 (Fig. 5C), the mislocalization defect of GFP-Rnr4 was rescued by overexpressing wild-type Rrp12 in pse1-1 cells (Fig. 6J). Taken together, my findings suggest that the Rrp12/Kap121 complex is important for the proper nuclear localization of Rnr2 and Rnr4 and the regulation of RNR activity levels.
Improper production of dNTPs at the beginning of S phase can cause inefficient initiation of DNA replication, and this might compromise Rad53 activation when cells are exposed to genotoxic stress (43, 44). To explore whether this could be the cause of the deficient DNA repair in rrp12-td cells, I evaluated the generation of dNTP pools during the cell cycle of control and DHFRts-Rrp12-expressing cells. With the control strain, I observed that the concentration of all dNTP species was maximal in S phase, in which they increased ≈3- to 6-fold relative to the levels present in the G1 phase (Fig. 7B to E). rrp12-td cells instead displayed a lower increment of dNTP levels (≈2- to 4-fold relative to the concentration in G1) and a delay in the kinetics of dNTP production of more than 20 min (Fig. 7B to F). Interestingly, the expression profile of the S-phase marker Clb5 showed that S-phase entry in rrp12-td cells was retarded by less than 10 min relative to that in control cells (Fig. 7H and I). Therefore, once it reaches S-phase entry, the rrp12-td strain requires a longer period of time to produce maximal dNTP levels than control cells. These results indicate that the synthesis of dNTPs during S phase is slow and inefficient in DHFRts-Rrp12-expressing cells.
The above results suggested that the delayed production of dNTPs during the cell cycle could be the cause of the slow S phase and defective DNA damage response exhibited by rrp12-td cells. I investigated therefore whether those defects could be rescued by artificially increasing the concentration of dNTPs in the rrp12-td strain. To this end, I took advantage of previous findings showing that the deletion of the SML1 gene leads to an ~2.5-fold increase in the dNTP concentration in yeast cells. I observed that the deletion of SML1 did not suppress the G1-S delay exhibited by rrp12-td cells (data not shown), indicating that other defects, aside from the low content of dNTPs at S-phase entry, must affect the progression of DNA replication in these cells (see Discussion). In contrast, the resistance of the rrp12-td strain to both HU and MMS was significantly increased when cells lacked the SML1 gene (Fig. 8A). A similar result was observed when the Rnr1 subunit was overexpressed in rrp12-td cells (data not shown), further suggesting that the rescue activity is due to enhanced production of dNTPs. Notably, I also observed that the SML1 deletion restored the activation of the DNA damage checkpoint in rrp12-td cells (Fig. 8B), thus indicating that a defective regulation of dNTP production is the most likely cause of the impaired response to DNA damage present in DHFRts-Rrp12-expressing cells.
Here I describe that Rrp12, a protein involved in ribosome subunit maturation and export, plays critical roles in cell cycle-related processes. Such roles have been unveiled through two independent approaches. First, by promoting the degradation of this protein in a degron strain, I have shown that the loss of Rrp12 leads to severe defects in S-phase entry and progression, inefficient DNA damage responses, and a milder delay in the M/G1 transition. Those dysfunctions were clearly dissociated from any defect in ribosomal production. Second, I found serendipitously that the DHFRts-Rrp12 protein expressed by the rrp12-td degron strain had some intrinsic structural defects that impaired its cell cycle/DNA damage response-related functions while keeping intact its ribosomal biosynthesis-linked roles. This property made it possible to dissociate the functional roles of Rrp12 in vivo by studying the rrp12-td strain under permissive conditions.
Interestingly, I observed that Rrp12 has unique functions during the cell cycle that are not shared by other proteins involved in ribosome biosynthesis. Thus, unlike results for other preribosomal components previously linked to cell cycle events, I could not detect any major defect in the Start cell cycle transition, in the loading of prereplicative complexes onto chromatin during the G1 phase, or in cytokinesis. Consistent with this idiosyncratic role, our protein complex purification experiments and coimmunoprecipitation experiments with epitope-tagged proteins (unpublished results) did not detect an interaction of Rrp12 with prereplicative components as previously reported for the pre-60S factors Noc3 and Noc7 (9, 52). Instead, I have found that the influence of Rrp12 on the activation of the DNA damage response and to some extent on S-phase progression is due to its implication in a very specific process that requires Kap121-dependent transport of proteins from the cytosol to the nucleus. Such a process favors the proper sequestration of the small Rnr2 and Rnr4 subunits of ribonucleotide reductase in the nucleus, thus contributing to the controlled production of dNTPs during specific cell cycle stages. In agreement with such a role, I have observed that yeast strains expressing defective versions of Kap121 and Rrp12 do not properly localize Rnr2 and Rnr4 inside the nucleus. Instead, these RNR subunits accumulate in the cytosol, leading to an abnormal production of dNTPs in both asynchronous and G1-arrested cultures, in the case of DHFRts-Rrp12-expressing cells.
What is the role of Rrp12 in the context of Kap121-mediated transport routes? Prima facie, two possible activities could be envisioned. One of them is that Rrp12 is just another Kap121 cargo that, when malfunctioning, leads to a dominant negative effect on the nuclear import machinery. I believe that such a scenario does not fit the experimental data, because I have observed no changes in the subcellular localization of Rrp12 in cells expressing nonfunctional versions of Kap121. Furthermore, although a dominant negative effect could be justified in the case of cells expressing the DHFRts-Rrp12 protein, it is not compatible with the data obtained when the Rrp12 protein is degraded in the degron strain at the nonpermissive temperature. Another possibility is that Rrp12 is in fact a functional element of the Kap121 transport machinery. Several independent observations support this model. First, I observed that these two proteins form stable associations that survive the rather stringent conditions of the complex purification experiments. Indeed, I have detected Kap121, but not Kap121-dependent cargos, bound to Rrp12 and DHFRts-Rrp12 proteins. Likewise, Rrp12, but not Kap121 cargos, was detected in immunotrapped GFP-Kap121 complexes. Second, and perhaps more importantly, I have found that the overexpression of wild-type Rrp12 can rescue some of the nuclear import defects present in yeast cells expressing a Kap121ts mutant, including the transport of histone H3 and Rrn4. These results suggest that the Rrp12/Kap121 interaction might help the Kap121ts mutant to overcome its intrinsic structural defects. Third, I have observed that wild-type Kap121 is less stable and becomes mislocalized to the cytosol and vacuoles in cells expressing the DHFRts-tagged version of Rrp12, further indicating that Kap121 requires the presence of wild-type Rrp12 to ensure full functionality. Given that Rrp12 is a shuttling HEAT repeat-containing protein that can bind directly to nucleoporins (30) and, in addition, that Kap121 requires interactions with the nuclear pore components for its association to the nuclear envelope (25, 26), Rrp12 may be also important to ensure the tethering of Kap121 to the nuclear pore. This last possibility remains to be determined. However, given that the DHFRts-Rrp12 maintains its functionality in nucleolus-dependent functions, such as its assembly with preribosomes and the subsequent nuclear-to-cytoplasmic export of ribosomal subunits, I surmise that the basis of the Kap121 defect found in rrp12-td cells must be prior to any alteration linked to nuclear pore interaction or the ensuing mobilization into the nucleus.
Additional observations highlight the specificity of Rrp12 toward Kap121. Thus, I have observed that Rrp12 does not interact with other karyopherins and that its deficiency/malfunction does not disrupt the transport of other proteins to the nucleus. Furthermore, Rrp12 is required for the efficient action of Kap121 on nuclearly localized cargos (i.e., Rnr2, Rnr4, and histone H3) but not on a cargo that is delivered to the nuclear envelope (Ulp1). The latter result is interesting, because it suggests that Kap121 utilizes different structural determinants or associates with different cotransporters to achieve its specificity toward different cargos. The results indicating that other proteins involved in ribosomal biogenesis (Tsr1 and Nop1) and nuclear trafficking events (Kap123) do not rescue the functionality of the Kap121ts mutant protein further underscore the unique role of Rrp12 in nuclear import-dependent events. Interestingly, it has been reported recently that eIF3, a translation initiation factor that participates in the maturation of pre-40S and pre-60S subunits, cooperates with Kap121 in transporting the proteasome into the nucleus in Schizosaccharomyces pombe (41). Thus, it is possible that the Rrp12/Kap121 functional interaction described here is part of a more general regulatory theme in which Kap121 cooperates with preribosome-associated factors to promote the nuclear import of different protein complexes. This network could potentially contribute to the proper coupling of ribosomal biogenesis and parallel biological routes required for optimal cell proliferation or growth. In this context, it is interesting to point out that our experiments have revealed that Kap121 can also associate with the preribosome export factor Arx1 that, intriguingly, is known to be required for a Kap121-mediated mechanism that recycles some ribosome synthesis factors from the cytosol back into the nucleus (20).
I demonstrated that the artificial enhancement of ribonucleotide reductase activity could restore normal DNA damage responses in rrp12-td cells, providing genetic proof for the idea that the defects in the regulation of that enzyme are the main cause for the DNA damage sensitivity present in the rrp12-td strain. Such a strategy, however, did not suppress the delay in S-phase progression exhibited by rrp12-td cells, suggesting that there must be other events mediated by Rrp12 at this stage of the cell cycle. Whether such events occur through Kap121-dependent nuclear import routes or through Kap121-independent mechanisms remains to be determined.
Although the sml1Δ rescue experiments pinpoint a clear implication of ribonucleotide reductase in the cell cycle/DNA damage response defects present in rrp12-td cells, they also raise some interesting questions pertaining to the role of dNTP levels in such processes. Thus, it is known that increased synthesis of dNTP via the activation of the cytosolic ribonucleotide reductase holoenzyme is required for proper S-phase progression and DNA damage checkpoint activation. In this context, it would be expected that rrp12-td cells would show efficient S-phase progression and reduced sensitivity to DNA damage since they contain higher levels of dNTPs than control cells do. Given that rrp12-td cells display instead a slow S phase and hypersensitivity to DNA damage, the results indicate that the elevation of dNTPs is not a benefit for the cell. This is in contrast with previous findings showing that a moderate elevation of dNTPs does not alter cell cycle progression and confers increased viability to DNA damage (4, 5, 55). Therefore, it seems that it is the specific defect in the regulation of RNR exhibited by rrp12-td cells that is particularly negative. One possibility is that a constant presence of Rnr2/Rnr4 in the cytoplasm interferes with the timing or duration of other known regulatory mechanisms that control RNR activity during the cell cycle (i.e., Sml1 degradation, dATP feedback inhibition, and transcriptional regulation of RNR genes) (4, 10, 35, 54). In favor of this possibility, time course experiments have shown that the dNTP production kinetics along the cell cycle is delayed in rrp12-td cells compared to control cells. Therefore, although they contain higher-than-normal dNTP levels in G1, rrp12-td cells reach S phase without having an optimal concentration of dNTPs. This defect may cause inefficient initiation of DNA replication and defective sensing of DNA damage.
The discovery that Rrp12 participates in nuclear import events raises the question of how this function is related to the previously described roles for this protein in ribosome biogenesis. I have shown that the overexpression of Rrp12 promotes the import of ribosome-unrelated proteins and that the partially defective DHFRts-rrp12 mutant exhibits defects in Kap121-mediated functions not implicated in ribosome maturation or export. Therefore, Rrp12 participates in nuclear import activities unrelated to ribosome synthesis. Whether Rrp12 is also involved in the nuclear import of ribosome biogenesis factors is presently unknown, but this is an interesting possibility to be addressed in the future. Rrp12 was proposed to have a role in ribosome subunit export based on its structural similarity with β-karyopherins, its ability to interact with nucleoporins and with Ran, and the fact that its depletion causes the nuclear accumulation of pre-40S and pre-60S particles (30). However, it is generally difficult to distinguish whether a particular factor is required for the actual export process or whether it is required to achieve export competence. In fact, the finding that Rrp12 depletion causes defects in pre-rRNA processing and preribosome maturation has been taken to suggest that its involvement in ribosome subunit export might be indirect (51). The role of Rrp12 in nuclear import described here opens the possibility that the defects in ribosome maturation and export of Rrp12-depleted cells could be caused by a deficient import of essential ribosome biogenesis factors, some of which are already known to be Kap121 cargos.
Although I have focused the present work on the elucidation of the extraribosomal functions of Rrp12, functional screening also revealed that Utp5, a 90S preribosome component essential for ribosome biogenesis, is important for cell cycle progression. However, unlike the case of Rrp12, the elimination of that protein elicits only defects in S-phase entry, indicating that Rrp12 and Utp5 affect different processes during the cell cycle. The G1-S delay of Utp5-depleted cells is accompanied by abnormal pre-rRNA synthesis and processing, but this is not necessarily the cause of their cell cycle phenotype. My finding of a degron strain (the utp18-td mutant) that produces abnormal pre-rRNA levels but displays normal cell cycle progression indicates that an alteration of the pre-rRNAs contents per se does not cause a slow G1-S transition. This is at odds with previously published data showing that defects in pre-rRNA processing are sensed at the Start checkpoint to delay the entry into S phase (2). My results suggest that such a sensing mechanism might become activated only when major disruptions in preribosome formation or nucleolar structure take place or when the accumulation of unprocessed pre-rRNAs reaches some threshold level.
I am grateful to the following colleagues for their kind gifts of reagents: J. Aitchison (pGAL-NOP1-GFP and pRS315-KAP121 plasmids), S. Elledge (pDL132 plasmid), R. Rothstein (anti-Sml1 antibody), V. Panse and E. Hurt (pGFP-ULP1 plasmid), L. Pemberton (pGFP-H3 plasmid), P. Silver (Y1201 strain and pPS1069 and pPS1070 plasmids), and B. Stillman (anti-Cdc45 antibody). I specially thank Arturo Calzada for his advice with cell cycle progression experiments, and the personnel of the CIC Proteomics Facility for their excellent work in the characterization of protein complexes and the quantification of dNTP levels. I also thank J. Pérez-Fernández for the purification of Rrp12 complexes in the initial part of this project, P. Martín-Marcos for help with polysome analysis, and the staff of the CIC Microscopy Facility for assistance with confocal microscopy. I am indebted to Xosé R. Bustelo for fruitful discussions and critically reading the manuscript.
This work is supported by grants from the Spanish Ministry of Science and Innovation (MICINN) (BFU-2008-02729 and RD06/0020/0001), the Castilla-León Autonomous Government (GR95), and the Samuel Solórzano Barruso Foundation (FS/2-2009). MICINN funding is cosponsored by the European Union FEDER Program.
Published ahead of print on 11 April 2011.