Synthetic Lethality with Conditional dbp6 Alleles Identifies Rsa1p, a Nucleoplasmic Protein Involved in the Assembly of 60S Ribosomal Subunits

Abstract

Dbp6p is an essential putative ATP-dependent RNA helicase that is required for 60S-ribosomal-subunit assembly in the yeast Saccharomyces cerevisiae (D. Kressler, J. de la Cruz, M. Rojo, and P. Linder, Mol. Cell. Biol. 18:1855–1865, 1998). To identify factors that are functionally interacting with Dbp6p, we have performed a synthetic lethal screen with conditional dbp6 mutants. Here, we describe the cloning and the phenotypic analysis of the previously uncharacterized open reading frame YPL193W, which we renamed RSA1 (ribosome assembly 1). Rsa1p is not essential for cell viability; however, rsa1 null mutant strains display a slow-growth phenotype, which is exacerbated at elevated temperatures. The rsa1 null allele synthetically enhances the mild growth defect of weak dbp6 alleles and confers synthetic lethality when combined with stronger dbp6 alleles. Polysome profile analysis shows that the absence of Rsa1p results in the accumulation of half-mer polysomes. However, the pool of free 60S ribosomal subunits is only moderately decreased; this is reminiscent of polysome profiles from mutants defective in 60S-to-40S subunit joining. Pulse-chase labeling of pre-rRNA in the rsa1 null mutant strain indicates that formation of the mature 25S rRNA is decreased at the nonpermissive temperature. Interestingly, free 60S ribosomal subunits of a rsa1 null mutant strain that was grown for two generations at 37°C are practically devoid of the 60S-ribosomal-subunit protein Qsr1p/Rpl10p, which is required for joining of 60S and 40S subunits (D. P. Eisinger, F. A. Dick, and B. L. Trumpower, Mol. Cell. Biol. 17:5136–5145, 1997). Moreover, the combination of the Δrsa1 and qsr1-1 mutations leads to a strong synthetic growth inhibition. Finally, a hemagglutinin epitope-tagged Rsa1p localizes predominantly to the nucleoplasm. Together, these results point towards a function for Rsa1p in a late nucleoplasmic step of 60S-ribosomal-subunit assembly.

The synthesis of ribosomes is one of the major cellular activities, which, in eukaryotes, takes place primarily, although not exclusively, in a specialized subnuclear compartment termed the nucleolus (33, 39). There, the ribosomal DNA is transcribed as precursors (pre-rRNAs), which undergo processing and covalent modification. Maturation of pre-rRNAs and their concomitant assembly with the ribosomal proteins (r-proteins) are dependent on various cis-acting elements and require a large number of nonribosomal trans-acting factors. Experimental evidence suggests that the basic outline of ribosome synthesis is conserved throughout eukaryotes. However, most of our knowledge comes from the combination of molecular genetics and biochemical approaches applied to the yeast Saccharomyces cerevisiae (reviewed in references 14, 46, 55, and 62).

In S. cerevisiae, the large 60S ribosomal subunits are composed of 46 r-proteins and three rRNA species (5S, 5.8S, and 25S), while the small 40S ribosomal subunits contain 32 r-proteins and the 18S rRNA (34, 62). Three of the four rRNAs (18S, 5.8S, and 25S) are transcribed as a single 35S pre-rRNA by RNA polymerase I, whereas the fourth rRNA (5S) is transcribed independently by RNA polymerase III (62). In the 35S pre-rRNA, the mature rRNA sequences are separated by two internal transcribed spacer sequences, ITS1 and ITS2, and are flanked by two external transcribed spacer sequences, 5′ ETS and 3′ ETS (see Fig. 1). Maturation of the 35S pre-rRNA requires a multitude of different trans-acting factors (≥50), including, among others, small nucleolar RNAs (snoRNAs), components of small nucleolar ribonucleoprotein particles (snoRNPs), endonucleases, exonucleases, putative RNA helicases, and rRNA-modifying enzymes (8, 31, 46, 47, 55). Although the pre-rRNA-processing pathway and its intermediates have been fairly well characterized (Fig. 1), the assembly process of the rRNAs and the approximately 80 r-proteins into mature ribosomal subunits is still poorly understood (62). Furthermore, the precise function of the trans-acting factors is only known in a few cases, and many are assumed to play an as yet ill-defined role in ribosome assembly.

FIG. 1.

Pre-rRNA processing and ribosome assembly in S. cerevisiae. The RNA polymerase I-transcribed pre-rRNA contains the sequences for the mature 18S, 5.8S, and 25S rRNAs that are separated by two internal transcribed spacer sequences, ITS1 and ITS2, and flanked by two external transcribed spacer sequences, 5′ ETS and 3′ ETS. The mature rRNA species are shown as bars, and the transcribed spacer sequences are shown as lines. The processing sites and their locations are indicated. The primary RNA pol I transcript undergoes covalent modifications (2′-O-ribose methylation and pseudouridylation), and it is processed at its 3′ end to yield the 35S pre-rRNA, which is the longest detectable precursor. Early-associating 40S and 60S r-proteins as well as trans-acting factors (proteins and snoRNAs) assemble on this precursor to form a 90S preribosomal particle (90S RNP). The 35S pre-rRNA is first cleaved at the U3 snoRNP-dependent site A₀ to generate the 33S pre-rRNA. This molecule is subsequently processed at sites A₁ and A₂; the latter cleavage results in the separation of the pre-rRNAs destined for the small and large ribosomal subunits and allows the 90S RNP to separate into a 43S RNP and a 66S RNP. The early pre-rRNA cleavages at A₀ to A₂ require snoRNP components, Rrp5p, and the putative ATP-dependent RNA helicases Dbp4p, Fal1p, Rok1p, and Rrp3p. Additional nucleolar assembly reactions probably occur concomitantly to the early cleavages and include incorporation of the Rpl5p-5S RNP and of later-associating r-proteins. The structural rearrangements within early or intermediate preribosomal particles are likely to require the putative ATP-dependent RNA helicases Dbp6p, Dbp7p, and Drs1p. The 43S RNP is exported to the cytoplasm, where endonucleolytic cleavage of the 20S precursor at site D yields the mature 18S rRNA. Then, the newly formed 40S subunits associate with translation initiation factors and are recruited to capped mRNAs, which they search for the first start codon. The 27SA₂ precursor within the 66S RNP is processed by two alternative pathways that both lead to the formation of nuclear pre-60S particles containing the mature 5.8S and 25S rRNAs. In the major pathway, the 27SA₂ precursor is cleaved at site A₃ by the RNase MRP complex. Rrp5p and the putative ATP-dependent RNA helicase Dpb3p assist in this processing step. The 27SA₃ precursor is exonucleolytically digested 5′→3′ up to site B_1S to yield the 27SB_S precursor, a reaction requiring the exonucleases Xrn1p and Rat1p. A minor pathway processes the 27SA₂ molecule at site B_1L, producing the 27SB_L pre-rRNA. While processing at site B₁ is being completed, the 3′ end of mature 25S rRNA is generated by processing at site B₂. The subsequent ITS2 processing of both 27SB species appears to be identical. Cleavage at sites C₁ and C₂ releases the mature 25S rRNA and the 7S pre-rRNA. The putative ATP-dependent RNA helicase Spb4p is a good candidate for assisting cleavage at site C₁ or C₂. The 7S pre-rRNA undergoes exosome-dependent 3′→5′ exonuclease digestion to the 3′ end of the mature 5.8S rRNA. It has been proposed that Dob1p/Mtr4p, a putative ATP-dependent RNA helicase, assists the exosome activity. The data presented in this study suggest that Rsa1p is involved in a nucleoplasmic assembly step of pre-60S ribosomal subunits, which is required for the efficient recruitment of the exchangeable 60S r-protein Qsr1p/Rpl10p. Cytoplasmic assembly of Qsr1p/Rpl10p and formation of 60S ribosomal subunits that are competent for 60S-to-40S subunit joining are likely to require the trans-acting factors Sqt1p and Nmd3p. Mature 60S subunits containing Qsr1p/Rpl10p can bind to 40S subunits to form 80S monosomes that can then engage in translation elongation. The 90S, 66S, and 43S RNPs, as well as the pre-60S and the mature 40S and 60S subunits, are shown as ovals. The nuclear envelope is represented by the stippled bars.

In the nucleolus, the 35S pre-rRNA associates with many of the r-proteins to form a 90S preribosomal particle (Fig. 1). From this particle, 66S and 43S preribosomes containing the 27S and 20S pre-rRNAs, respectively, are formed (50). The 66S particle remains in the nucleus until exonucleolytic trimming of the 7S pre-rRNA to 5.8S rRNA is completed (48, 49), while the 43S preribosome is rapidly exported to the cytoplasm, where the final maturation step in the synthesis of the 18S rRNA takes place (48, 52). A large number of r-proteins associate with the nucleolar preribosomes at early steps during ribosome maturation, whereas others assemble at later steps or even are added only in the cytoplasm (30). It has been shown that at least three large-subunit r-proteins can exchange on mature 60S subunits in vivo, and the exchangeability of Qsr1p/Rpl10p in particular, which is required for 60S and 40S subunit joining, suggests a possible translational regulatory mechanism (12, 16, 65). The available data are, however, far from being sufficient to establish a definitive assembly pathway for the different r-proteins. In addition to r-proteins, the nucleolar preribosomes have long been known to contain non-r-proteins (50); the identity of these proteins has not been clearly established, but they presumably correspond to trans-acting factors required for pre-rRNA processing and modification or are involved in the assembly of the pre-rRNAs with the r-proteins.

One class of trans-acting factors involved in ribosome biogenesis comprises the putative ATP-dependent RNA helicases of the DEAD-box and related families. These protein families are defined by several evolutionarily conserved motifs, and their members are involved in various RNA metabolic processes, including pre-mRNA splicing, translation initiation, RNA degradation, and ribosome biogenesis (8). To date, 14 putative RNA helicases have been implicated in ribosome biogenesis in S. cerevisiae (8), and the following functions can be envisaged for these proteins. (i) An RNA-unwinding activity could be required to establish and/or dissociate snoRNA:pre-rRNA base pairings; such base pairings and the final folding of the rRNA in the mature ribosome are in most cases mutually exclusive (32, 53). (ii) Putative RNA helicases may functionally assist endo- and exonucleases (10, 61). (iii) Finally, they may recruit, rearrange or dissociate trans-acting factors and r-proteins within preribosomal particles during the processing and assembly reactions by modulating specific intramolecular rRNA, rRNA-protein, or even protein-protein interactions. In the absence of such a putative RNA helicase, the lack or retardation of the required structural changes may lead to an abortive assembly, which can either entail the disassembly of preribosomal particles and destabilization of pre-rRNA intermediates or the accumulation of preribosomal particles and stabilization of pre-rRNA intermediates (6, 9, 27).

Dbp6p is an essential putative RNA helicase of the DEAD-box protein family that is exclusively localized in the nucleolus. The phenotypes observed upon Dbp6p depletion suggest that Dbp6p is required for the assembly or stability of an early preribosomal particle leading to the formation of 60S ribosomal subunits (27). In order to gain more insight into the functional environment of Dbp6p and to understand in more detail the assembly of 60S ribosomal subunits, we have performed a synthetic lethal (SL) screen with conditional dbp6 alleles. Synthetic enhancement (7, 21) has proven to be one of the most successful genetic approaches for dissecting macromolecular structures and their assembly, as exemplified by the nuclear pore complex and ribosome biogenesis (4, 13, 18, 56). Here, we describe the cloning and the phenotypic analysis of the previously uncharacterized open reading frame (ORF) YPL193W, which is hereafter referred to as RSA1 (ribosome assembly 1). Disruption of RSA1 confers slow growth and temperature sensitivity, and it results in a moderate decrease in the pool of free 60S ribosomal subunits and in the accumulation of half-mer polysomes. This phenotype is similar to that of mutants defective in 60S-to-40S subunit joining. Accordingly, the free 60S subunits of rsa1 null mutant strains, grown for two generations at 37°C, are practically devoid of the 60S-ribosomal-subunit protein Qsr1p. Moreover, the combination of the Δrsa1 and qsr1-1 mutations leads to a strong synthetic growth inhibition. Finally, a hemagglutinin (HA) epitope-tagged Rsa1p localizes predominantly to the nucleoplasm. Together, these results point to a function for Rsa1p in a late nucleoplasmic step of 60S-ribosomal-subunit assembly.

MATERIALS AND METHODS

Strains, media, and genetic methods.

The S. cerevisiae strains in this study are derivatives of the diploid strain W303 (MATa/MATα ura3-1/ura3-1 ade2-1/ade2-1 his3-11,15/his3-11,15 leu2-3,112/leu2-3,112 trp1-1/trp1-1) (45). YDK8-1A (MATα dbp6::kanMX4) and YDK8-2A (MATa dbp6::kanMX4) are meiotic segregants of the previously described diploid strain YDK8 (27); they require a plasmid-borne copy of DBP6 for cell viability. YDK9 (MATa/MATα dbp6::HIS3MX6/DBP6) was obtained by disrupting one DBP6 ORF copy with the HIS3MX6 marker module; YDK9-4A (MATα dbp6::HIS3MX6) is a meiotic segregant of YDK9 that requires a plasmid-borne copy of DBP6 for cell viability. YDK11 (MATa/MATα ade3::kanMX4/ADE3) was obtained by disrupting one ADE3 ORF copy with the kanMX4 marker module; YDK11-5C (MATa ade3::kanMX4) is a meiotic segregant of YDK11. YDK44 (MATa/MATα rsa1::kanMX4/RSA1) and YDK45 (MATa/MATα rsa1::HIS3MX6/RSA1) were obtained by disrupting one RSA1 ORF copy with the kanMX4 and the HIS3MX6 marker modules, respectively. Sporulation and subsequent tetrad dissection resulted in the haploid disrupted strains YDK44-1B (MATa rsa1::kanMX4), YDK44-1C (MATα rsa1::kanMX4), YDK45-11A (MATa rsa1::HIS3MX6), and YDK45-11B (MATα rsa1::HIS3MX6). MMY3-3B (MATα qsr1-1; this strain was originally described to be MATa, although in our hands it is MATα) has been described previously (16). Preparation of standard media and genetic manipulations were according to established procedures (1, 25). Yeast cells were transformed by a lithium acetate method (19). One-step gene replacements were done as described by Rothstein (36). For tetrad dissection, a Singer MSM micromanipulator was used.

Deletion disruptions.

The deletion disruption of the DBP6 ORF by the HIS3MX6 marker module was accomplished as described for the deletion disruption of the DBP6 ORF by the kanMX4 marker module (27), except that the EcoRV-linearized plasmid pFA6a-HIS3MX6 was the template for PCR amplification of the HIS3MX6 marker module (58). Deletion disruptions of the ADE3 and RSA1 ORFs were accomplished by transformation of PCR-synthesized HIS3MX6 and/or kanMX4 marker cassettes with short flanking homology regions (SFH-PCR) into W303 (3, 58, 59). Briefly, heterologous kanMX4 or HIS3MX6 marker modules flanked on each side by short regions of 45 bp with homology to the ADE3 or RSA1 loci, respectively, were generated by PCR with Vent polymerase (New England Biolabs), plasmid pFA6a-kanMX4 or pFA6a-HIS3MX6 as the DNA template, and the following oligonucleotides: ADE3-SFH5′, 5′GGT AAC GAG ACG AAC ACA ACT TTA CAA GTC AAA TAA GAA ATC ATG CGT ACG CTG CAG GTC GAC3′ (the ADE3 5′ upstream sequence, with the start codon underlined, is in boldface, and the 5′ sequence homologous to the kanMX4 marker module is in lightface); ADE3-SFH3′, 5′TTT TGC ATT TGT CTT TAT TAA ATT CTA TAT AAT TAA GTT GTC TTA ATC GAT GAA TTC GAG CTC G3′ (the reverse complement of the ADE3 3′ downstream sequence, with the stop codon underlined, is in boldface, and the 3′ sequence homologous to the kanMX4 marker module is in lightface); RSA1-SFH5′, AAT TAT AAT AAC TTT GAA AAT TCG AAG GGT GAT GGA CAT TCT AGG CGT ACG CTG CAG GTC GAC3′ (the RSA1 coding sequence, with the deletion starting at position +49, is in boldface, and the 5′ sequence homologous to the kanMX4 and HIS3MX6 marker modules is in lightface); and RSA1-SFH3′, 5′TCC CAA TTC TCT TAT AAA GTC CAA AAG TTG GGA GTT TTC GTT TGC ATC GAT GAA TTC GAG CTC G3′ (the reverse complement of the RSA1 3′ coding sequence, with the deletion ending 192 nucleotides upstream of the stop codon, is in boldface, and the 3′ sequence homologous to the kanMX4 and HIS3MX6 marker modules is in lightface). The SFH-PCR products were extracted with phenol, concentrated by ethanol precipitation, and then used to transform W303. Transformants were selected either on yeast extract-peptone-dextrose (YPD) plates containing 200 mg of G418 (Gibco BRL) or on synthetic dextrose minimal medium (SD) plates lacking His (SD-His plates).

Random PCR mutagenesis of DBP6.

Cold- and temperature-sensitive (CS and TS, respectively) dbp6 alleles were generated by random-PCR mutagenesis. Briefly, PCRs with Taq DNA polymerase (Gibco BRL), primers P5′ (5′ATT TCA GTC CCA CGA ACT GA3′ [starting 418 bp upstream of the DBP6 start codon]) and P3′ (5′CAA ACG AGC ATT CCA ACG T3′ [starting 180 bp downstream of the DBP6 stop codon]), and, as the template, the XhoI-restricted plasmid pRS416-DBP6 were performed under conditions where one of the four deoxynucleoside triphosphates (dNTPs) is present in gradually diminishing amounts (from the normal concentration of 200 μM to 25 μM). This was done separately for all four dNTPs, resulting in a total of 32 independent PCRs. The PCR products were divided into eight pools, digested with NheI and BglII, gel purified (Gene Clean; Bio 101, Inc.), and cloned as 1.71-kb fragments into the NheI/BglII-restricted plasmid pRS415-DBP6 (27). The eight ligations and subsequent transformations yielded 80,000 Escherichia coli transformants. Plasmid DNA was prepared from the independent pools, and DNA from six of them was transformed into YDK9-4A pRS416-DBP6. A total of 5,000 yeast transformants were obtained on SD-Leu plates, and they were subsequently replica plated on SD-Leu- and 5-fluoroorotic acid (5-FOA)-containing plates at the permissive (30°C) and the nonpermissive (18 and 37°C) temperatures. Putative conditional mutants were restreaked at all three temperatures, and their plasmid DNA was extracted (60), amplified in E. coli, and transformed into YDK9-4A pRS416-DBP6. Upon plasmid shuffling a total of 13 dbp6 alleles were obtained. The following plasmids bearing dbp6 alleles were used in this study: pRS415-dbp6-2 (pDK263; TS at 37°C), pRS415-dbp6-3 (pDK264; CS at 14.5°C, weakly TS at 37°C), pRS415-dbp6-4 (pDK265; slow growth, tightly CS at 18°C), pRS415-dbp6-10 (pDK285; TS at 37°C), and pRS415-dbp6-13 (pDK288; slow growth, tightly TS).

Plasmids.

All recombinant DNA techniques were done according to established procedures using E. coli DH10B for cloning and plasmid propagation (38). Plasmids pRS416-DBP6 (pDK186), pRS415-DBP6 (pDK211), and pRS415-HA-DBP6 (pDK250) have been previously described (27). A 2.89-kb XhoI/SacI fragment from pRS415-HA-DBP6 was subcloned into pRS414 (42) to generate pRS414-HA-DBP6 (pDK327). Subcloning the dbp6 alleles as 2.82-kb XhoI/SacI fragments into pRS414 resulted in the plasmids pRS414-dbp6-2 (pDK345), pRS414-dbp6-3 (pDK346), pRS414-dbp6-4 (pDK324), pRS414-dbp6-10 (pDK347), and pRS414-dbp6-13 (pDK325). Plasmid pHT4467-HA-DBP6 (pDK321) was constructed by cloning a 2.88-kb SalI/SacI fragment, with the SalI site being blunt ended by T4 DNA polymerase treatment, into the SalI/SmaI-restricted plasmid pHT4467 (CEN URA3 ADE3) (56). YCplac33-RSA1 (pDK468) was obtained by cloning a 2.71-kb SalI/HindIII fragment from pDK433, which was previously generated by cloning a 3.1-kb PstI fragment from the original library clone pDK427 (see below) into the PstI-restricted vector YCplac111, into the SalI/HindIII-restricted vector YCplac33. YCplac111 and YCplac33 have been previously described (20).

Isolation of SL mutants.

Yeast strains bearing mutations in the DBP6 gene were screened for SL mutations based on a combination of the ade2/ade3 red/white colony-sectoring assay and counter-selection on 5-FOA-containing plates (5, 26). To construct the starting strain for the SL screen, YDK8-1A pRS414-dbp6-4 was crossed to YDK11-5C pHT4467-HA-DBP6, the resulting diploid, YDK15 (MATa/MATα ade3::kanMX4/ADE3 dbp6::kanMX4/DBP6 pHT4467-HA-DBP6 pRS414-dbp6-4), was sporulated, and tetrads were dissected. Two spore clones of opposite mating type, YDK15-5B (MATa ade3::kanMX4 dbp6::kanMX4 pHT4467-HA-DBP6) and YDK15-5D (MATα ade3::kanMX4 dbp6::kanMX4 pHT4467-HA-DBP6), originating from a nonparental ditype tetrad that had already lost the plasmid pRS414-dbp6-4 were transformed with pRS414-dbp6-2, pRS414-dbp6-3, and pRS414-dbp6-10. These six “SL screen starting strains” were grown in liquid SD medium lacking Trp and Ura (SD-Trp-Ura medium) to an optical density at 600 nm (OD₆₀₀) of around 0.5 and plated on SD-Trp plates at a density of approximately 500 cells/plate. The plates were then UV irradiated, resulting in 25 to 55% survival, and incubated for 5 days at 30°C in the dark. Red colonies were restreaked once on SD-Trp plates and then twice on SD-Trp and 5-FOA-containing plates. To confirm that the nonsectoring, 5-FOA-sensitive phenotype was neither due to genomic integration of the plasmid pHT4467-HA-DBP6 nor linked to the plasmid-borne dbp6 alleles, candidate SL strains were transformed with pRS415-HA-DBP6, pRS415-dbp6-2, pRS415-dbp6-3, pRS415-dbp6-10, and pRS415; true SL strains should only show restored sectoring and growth on 5-FOA-containing plates upon transformation with pRS415-HA-DBP6. From a total of approximately 40,000 screened colonies, six strains (derived from YDK15-5B: sl263210 [dbp6-2], sl263309 [dbp6-2], sl264205 [dbp6-3], sl264234 [dbp6-3], and sl285204 [dbp6-10]; derived from YDK15-5D: sl264409 [dbp6-3]) showed a strong synthetic enhancement or an SL phenotype and were thus retained for further analyses.

Cloning of RSA1, rsa1-1, and rsa1-2.

Strain sl264234 was transformed with a YCplac111-based yeast genomic library (29) yielding approximately 20,000 transformants on five SD-Leu plates. Transformants were replica plated onto 5-FOA-containing plates and incubated for 3 days at 30°C to identify colonies that could lose the plasmid pHT4467-HA-DBP6. These colonies were then restreaked on SD-Leu plates that were incubated at 30°C. Since the slow-growth phenotype of the SL strain is more pronounced than that of the dbp6-3 single-mutant strain, candidate colonies with more or less wild-type growth behavior were selected for plasmid isolation (60). Yeast plasmid preparations from two candidate colonies were transformed into E. coli KC8, and selection of transformants was done on minimal M9 medium lacking leucine and supplemented with ampicillin. E. coli plasmid preparations containing the library plasmid were retransformed into sl264234 to confirm the complementation of the synthetic-enhancement phenotype. Sequencing the inserts of the two complementing plasmids revealed that both contained overlapping regions from chromosome XVI. One clone (pDK427) had a 5.8-kb insert that included the three ORFs YPL192C, RSA1, and YPL194W. Subcloning indicated that the presence of the RSA1 ORF was sufficient to complement the synthetic-enhancement phenotype of sl264234. Transformation of pDK427 into the remaining SL strains revealed that the synthetic-enhancement phenotype could only be complemented for sl285204.

To determine whether sl264234 and sl285204 indeed had mutations in the RSA1 gene, the RSA1 ORF, including 394 bp of the promoter and 310 bp of the terminator region, was sequenced. To this end, RSA1 was amplified (Vent polymerase) by PCR from genomic DNA prepared from a wild-type control strain (YDK15-5B) and the two SL strains with oligonucleotides spanning or introducing the restriction sites EcoRI (RSA1-5′EcoRI, 5′CAA GGT CTG TTG AAT TCC3′; the EcoRI site is in boldface) and HindIII (RSA1-3′HindIII, 5′GCC ATG AAG CTT CTC GAG CTG GTA GAG TCA GGA AGC3′; the HindIII site is in boldface), respectively. The 1.85-kb PCR products were digested separately with EcoRI/BamHI and BamHI/HindIII and cloned in a three-piece ligation into the EcoRI/HindIII-restricted vector YCplac111, yielding the plasmids YCplac111-RSA1 (pDK477), YCplac111-rsa1-1 (pDK478), and YCplac111-rsa1-2 (pDK482). YCplac111-RSA1 complemented the slow-growth and TS phenotype of the rsa1 null allele, and its sequence is identical to the one deposited at the Saccharomyces Genome Database.

RSA1 HA epitope tagging and cloning under the control of its cognate promoter.

To express an N-terminally double-HA-tagged Rsa1 protein from its cognate promoter at approximately wild-type levels, the RSA1 promoter region (up to position −394) was amplified (Vent polymerase) by PCR from pDK433 with oligonucleotides spanning or introducing the restriction sites EcoRI (RSA1-5′EcoRI; see above) and NcoI (RSA1-NcoI, 5′CAT GCC ATG GTC CAT GTC GCA ATA AGC T3′; the NcoI site is in boldface, and the reverse complement of the start codon is underlined), respectively. This PCR product was digested with EcoRI/NcoI and cloned into the EcoRI/NcoI-restricted plasmid pAS24 (41) to replace the GAL1-10 promoter by the cognate RSA1 promoter. The promoter sequence of the obtained plasmid, YCplac111-pRSA1-2xHA (pDK480), was confirmed by sequencing. Meanwhile, the RSA1 ORF and 310 bp of its terminator region were amplified (Vent polymerase) by PCR from pDK433 with oligonucleotides introducing the restriction sites SalI (RSA1-SalI, 5′GAA TTC GTC GAC AAT TAT AAT AAC TTT GAA AAT TCG AAG3′; the SalI site is in boldface, and the second codon of the RSA1 ORF is underlined) and HindIII (RSA1-3′HindIII; see above), respectively. The PCR product was digested with SalI/HindIII and cloned into the SalI/HindIII-restricted plasmid pAS24-DBP6 (pDK240) (27) to yield plasmid pAS24-RSA1 (pDK481). The RSA1 ORF and the terminator sequence were confirmed by sequencing. YCplac111-HA-RSA1 (pDK483) was obtained by subcloning a 1.54-kb NcoI/HindIII fragment from pDK481 into the NcoI/HindIII-restricted pDK480. YCplac111-HA-RSA1 complemented the rsa1 null allele to the wild-type extent at 30 and 37°C, and the HA-tagged Rsa1p was detected by Western blotting as a faint band that migrated at the molecular mass of approximately 50 kDa.

Synthetic-interaction crosses.

To determine if different mutants affecting assembly of 60S ribosomal subunits were showing synthetic-enhancement phenotypes, the following crosses were performed.

(i) dbp6 qsr1-1.

YDK8-2A pRS416-DBP6 was crossed to MMY3-3B pRS413, the resulting diploid YDK53 (MATa/MATα dbp6::kanMX4/DBP6 qsr1-1/QSR1 pRS416-DBP6 pRS413) was sporulated, and tetrads were dissected. The spore clones YDK53-18A (MATa dbp6::kanMX4 qsr1-1 pRS416-DBP6) and YDK53-18C (MATa dbp6::kanMX4 pRS416-DBP6), originating from a tetratype tetrad, were transformed with the plasmids pRS414, pRS414-HA-DBP6, pRS414-dbp6-2, pRS414-dbp6-3, pRS414-dbp6-4, and pRS414-dbp6-13. Transformants were restreaked on SD-Trp plates and subjected to plasmid shuffling on 5-FOA-containing plates. Viable dbp6/qsr1-1 double mutants could be recovered, and subsequent restreaking on YPD plates at 18, 30, and 37°C showed that the different dbp6 mutations were only weakly, if at all, enhancing the slow-growth phenotype of the qsr1-1 mutant. This result was confirmed by measuring the doubling times of the wild type and some single and double mutants in liquid YPD medium at 30°C. The times were as follows: wild type, 1.55 h; qsr1-1 mutant, 3.2 h; dbp6-4 mutant, 3.5 h; dbp6-13 mutant, 3.8 h; dbp6-4/qsr1-1 mutant, 3.95 h; and dbp6-13/qsr1-1 mutant, 5 h.

(ii) dbp6 rsa1::HIS3MX6.

YDK8-1A pRS416-DBP6 was crossed to YDK45-11A, the resulting diploid YDK46 (MATa/MATα dbp6::kanMX4/DBP6 rsa1::HIS3MX6/RSA1 pRS416-DBP6) was sporulated, and tetrads were dissected. The spore clones YDK46-7A (MATa dbp6::kanMX4 rsa1::HIS3MX6 pRS416-DBP6), and YDK46-7B (MATα dbp6::kanMX4 pRS416-DBP6), originating from a tetratype tetrad, were transformed with the plasmids pRS414, pRS414-HA-DBP6, pRS414-dbp6-2, pRS414-dbp6-3, pRS414-dbp6-4, and pRS414-dbp6-13. Transformants were restreaked on SD-Trp plates and subjected to plasmid shuffling on 5-FOA-containing plates. No viable dbp6-4/rsa1::HIS3MX6 and dbp6-13/rsa1::HIS3MX6 double mutants could be recovered, indicating that the rsa1-null allele and the dbp6-4 and dbp6-13 mutants were synthetically lethal. However, viable, but slow-growing, dbp6-2/rsa1::HIS3MX6 and dbp6-3/rsa1::HIS3MX6 double mutants could be recovered, which showed that the rsa1 null allele was strongly synthetically enhancing the very mild growth defect of the dbp6-2 and dbp6-3 mutants (see Results).

(iii) qsr1-1 rsa1::HIS3MX6.

MMY3-3B YCplac33-RSA1 was crossed to YDK45-11A, the resulting diploid YDK49 (MATa/MATα qsr1-1/QSR1 rsa1::HIS3MX6/RSA1 YCplac33-RSA1) was sporulated, and tetrads were dissected. Upon restreaking complete tetrads on 5-FOA-containing plates to select for loss of YCplac33-RSA1, several tetratype tetrads, as judged from the phenotypes of the spore clones (one with wild-type growth, His⁻; one with weak slow growth, His⁺; one with slow growth, His⁻; one with strong slow growth, His⁺), could be recovered. The strong slow-growth phenotype of rsa1::HIS3MX6/qsr1-1 double mutant spore clones indicated that the rsa1-null mutation was synthetically enhancing the slow-growth phenotype of the qsr1-1 mutant. Mean values for doubling times were obtained by monitoring the growth rates of spore clones from two tetratype tetrads in liquid YPD medium at 30°C. The times were as follows: wild type, 1.35 h; rsa1::HIS3MX6 clone, 2.35 h; qsr1-1 clone, 3.1 h; qsr1-1/rsa1::HIS3MX6 clone, 8.3 h (see Results).

Information on the crosses dbp6 nip7-1, dbp6 nop4, dbp6 rpl16b::LEU2, dbp6 spb4-1, qsr1-1 spb4-1, and rsa1::HIS3MX6 spb4-1, as well as the description of the deletion disruption of SPB4, the cloning of spb4-1 by gap repair, and the description of the strains used for the crosses will be provided by the authors upon request or can be found on the Linder laboratory Web site (http://www.expasy.ch/linder/Kressler_suppl.html).

Sucrose gradient analyses.

Polyribosome preparations, polysome analyses, and ribosomal subunit preparations were done exactly as described previously (28). Gradient analysis was performed with an ISCO UA-6 gradient UV detection and fraction collection system with continuous monitoring at A₂₅₄.

For fractionation analyses, extract preparations and gradient centrifugation conditions were identical to those used for the polysome analysis, except that 10 A₂₆₀ units of cell extract was layered onto the gradients and the gradients were centrifuged for 3 h. Fractions of approximately 500 μl were collected, and proteins were precipitated by addition of trichloroacetic acid to a final concentration of 10% followed by incubation on ice for at least 10 min. Proteins were pelleted by centrifugation in a microcentrifuge for 10 min at 4°C. Pellets were washed twice with 1 ml of ice-cold acetone and finally resuspended in 30 μl of protein gel loading buffer (1). Aliquots of 10 μl were loaded on sodium dodecyl sulfate–12% polyacrylamide gels and analyzed by Western blot analysis according to standard procedures (1, 38). As a control, 0.125 A₂₆₀ unit of cell extract was run alongside the fractions. Monoclonal mouse anti-Rpl3p antibodies at a dilution of 1:10,000 (57) and polyclonal rabbit anti-Qsr1p antibodies at a dilution of 1:2,000 (51) were used as primary antibodies. Blots were decorated with goat anti-mouse immunoglobulin G (IgG) and goat anti-rabbit IgG horseradish peroxidase-conjugated secondary antibodies (Bio-Rad) at a dilution of 1:15,000 and developed with the ECL detection kit (Amersham). Signal intensities were quantified with the WinCam, version 2.1, program (Cybertech). The relative levels of Qsr1p within each fraction were determined by dividing the Rpl3p-to-Qsr1p ratio of the respective fraction by the Rpl3p-to-Qsr1p ratio of the total extract lane. The obtained values for the wild-type and Δrsa1 strains were then compared, and the results were expressed as the fold reduction in the levels of Qsr1p in the respective fractions. Note that the reduction of Qsr1p on free 60S ribosomal subunits is probably underestimated due to saturation of the Rpl3p signal on the autoradiographs.

Pulse-chase labeling of pre-rRNA.

Cells of strains YDK44-1A and YDK44-1B were grown at 30 or 37°C (6 h at 37°C) to an OD₆₀₀ of around 0.8 in 40 ml of SD-Met medium. Cells were concentrated by centrifugation for 3 min at 3,000 rpm in an SS34 rotor at room temperature and resuspended in 1 ml of prewarmed SD-Met medium. Then, the pre-rRNA was pulse labeled for 1 min with 250 μCi of [methyl-³H]methionine (Amersham; specific activity, 70 to 85 Ci/mmol). The chase was initiated by diluting 250-μl aliquots of the pulse-labeled cells in 1.75 ml (2.2-ml Eppendorf tubes) of prewarmed SD medium containing 1 mg of methionine per ml. The cells were harvested after 0, 2, 5, and 15 min of chase by centrifugation (15 s at full speed), washed in ice-cold water, centrifuged again, and frozen in liquid nitrogen. Total RNA was extracted by the acid-phenol method (1). The methyl group incorporation was measured by scintillation counting, and 20,000 cpm per RNA extract was loaded and resolved on 1.2% agarose-formaldehyde gels (54). RNA was transferred to Hybond-N⁺ nylon membranes (Amersham) by capillary blotting. After being baked for 2 h at 80°C, the membranes were sprayed with EN³HANCE (Du Pont), dried, and exposed to X-ray films for 3 days at −80°C with an intensifying screen.

Indirect immunofluorescence.

YDK44-1B YCplac111-HA-RSA1 and YDK44-1B YCplac111-RSA1 strains were grown to an OD₆₀₀ of around 0.5 in SD-Leu medium, and 5 ml of cells was harvested by centrifugation. Preparation of yeast cells for immunofluorescence, immunofluorescence microscopy, and image acquisition and processing were done as previously described (9).

Miscellaneous.

Total yeast protein extracts were prepared and analyzed by Western blotting according to standard procedures (1, 38). The monoclonal mouse 16B12 antibody (BAbCO) was used as a primary antibody, and blots were decorated with goat anti-mouse IgG horseradish peroxidase-conjugated secondary antibodies (Bio-Rad) and developed with the ECL detection kit (Amersham). Sequencing was kindly performed by C. Rossier by cycle sequencing with dideoxy dye terminator by using an ABI377 instrument (Perkin-Elmer). For classical dideoxy-sequencing a T7 Sequencing kit (Pharmacia) was used. Sequence comparisons were performed at the Saccharomyces Genome Database (Stanford) and at the National Center for Biotechnology Information.

RESULTS

Synthetic lethality with dbp6 alleles identifies the novel RSA1 gene.

In order to gain more insight into the functional environment of Dbp6p in particular and into the assembly of 60S ribosomal subunits in general, we have undertaken an SL screen with conditional dbp6 alleles, which was based on a combination of the ade2/ade3 red/white colony sectoring assay and counter-selection on 5-FOA-containing plates (5, 26). The principle of an SL screen is that, although a mutation in a single gene is tolerated by the cell, combination with a mutation in another functionally related component leads to a synergistic growth inhibition (synthetic enhancement) or to cell death (synthetic lethality). Synthetic phenotypes thus should identify both physically and functionally interacting and nonphysically but functionally overlapping components (13). Moreover, linking different interconnected biological systems by this method should be possible (17).

The conditional dbp6 alleles needed for the SL screen were generated by random-PCR mutagenesis of DBP6 (see Materials and Methods). We obtained a total of 13 CS or TS dbp6 alleles, five of which were used in this study. The alleles dbp6-2, dbp6-3, and dbp6-10 conferred almost wild-type growth at 30°C (doubling times in YPD at 30°C were around 1.9 h for these mutants versus 1.5 h for the wild-type strain), and dbp6-4 and dbp6-13 resulted in a slow-growth phenotype (doubling times in YPD at 30°C were around 3.5 h) and in tight cold or temperature sensitivity, respectively. Polysome analysis of these five mutants revealed that all of them, although to different extents, were deficient in 60S ribosomal subunits and that they accumulated half-mer polysomes (see Fig. 6; data not shown).

FIG. 6.

The rsa1 null mutation enhances the polysome profile phenotype of the dbp6-3 mutant. Polysome profiles are shown for the following strains: wild type (YDK46-7B pRS414-HA-DBP6) (A), dbp6-3 strain (YDK46-7B pRS414-dbp6-3) (B), Δrsa1 strain (YDK46-7A pRS414-HA-DBP6) (C), and Δrsa1/dbp6-3 strain (YDK46-7A pRS414-dbp6-3). Strains were grown in YPD at 30°C, and cells were harvested at an OD₆₀₀ of around 0.8. Cell extracts were resolved in 7-to-50% sucrose gradients. Gradients were analyzed by continuous monitoring at A₂₅₄. Sedimentation is from left to right. The peaks of free 40S and 60S subunits, 80S ribosomes (free couples and monosomes), and polysomes are indicated. Half-mers are indicated by arrows.

We first tested the specificity or reliability of the SL approach with respect to the conditional dbp6 mutants. To this end, we determined whether different mutations affecting the biogenesis of 60S ribosomal subunits—nip7-1 (64), nop4-1 and nop4-2 (44), rpl16b::LEU2 (35), and spb4-1 (37)—showed synthetic-enhancement phenotypes when combined with dbp6 alleles (see Materials and Methods). To avoid genetic background heterogeneity that could make the comparison of growth phenotype differences between single and double mutants difficult, we introduced the nip7-1 and the nop4::TRP1 alleles by three subsequent crosses into the W303 background. The spb4-1 mutation was cloned by gap repair from the original spb4-1 strain (YAS168) and then introduced into a strain with the SPB4 gene disrupted in the W303 background (see Materials and Methods). None of the above mutations synthetically enhanced the growth phenotype of the dbp6 mutants, and the double mutants always adopted the growth phenotype of the less healthy single-mutant strain (data not shown). So far, the only synthetic-enhancement interaction involving dbp6 alleles is the previously reported one between the dbp6-2 and dbp6-3 mutations and the dbp7 null allele (6). Since the pre-rRNA processing defects observed upon Dbp6p depletion and in the dbp7 null mutant are similar, it was speculated that Dbp6p and Dbp7p might act at related steps in 60S-ribosomal-subunit assembly (6). We conclude that combining mutations that affect 60S-ribosomal-subunit biogenesis does not necessarily lead to synthetic lethality, implying that an SL screen with dbp6 alleles, and even more so with the mildly affected dbp6-2, dbp6-3, and dbp6-10 alleles, could potentially reveal specific interactions.

Yeast cells were mutagenized by UV irradiation, and 40,000 surviving colonies were screened for strains that carried mutations that were synthetically enhancing or synthetically lethal with conditional dbp6-2, dbp6-3, or dbp6-10 alleles. Six SL mutant strains were obtained (see Materials and Methods). In agreement with a role for Dbp6p in the assembly of 60S ribosomal subunits, the SL phenotype of two semidominant mutants (sl263309 and sl264205) was complemented by the RPL3 gene (29), which encodes the L3 r-protein of the large ribosomal subunit. Two of the SL mutant strains are still being characterized, and the remaining two are the subject of this article.

The RSA1 gene was cloned by complementation of the slow-growth and synthetic-enhancement phenotype of strain sl264234 (see Materials and Methods). Upon transformation into the remaining SL strains, RSA1 only complemented the synthetic-enhancement phenotype of sl285204. To determine whether these two SL strains actually had mutations in RSA1, the RSA1 ORF and its 5′ and 3′ flanking regions were amplified by PCR from genomic DNA that was prepared from a wild-type control strain and the two SL strains (see Materials and Methods). Sequencing revealed that the rsa1-1 mutation of sl264234 was an A-to-T conversion at nucleotide position +673, which changes the lysine codon (AAG) at amino acid position 225 to a premature stop codon (TAG). The rsa1-2 mutation of sl285204 was found to be an A-to-T conversion at nucleotide position +1, which changes the start codon to TTG (Fig. 2). The recessive rsa1-1 and rsa1-2 mutations most likely correspond to complete loss-of-function mutations (see below).

FIG. 2.

Amino acid sequence of Rsa1p. A putative seven-amino-acid pattern nuclear localization signal is present near the C terminus of the protein (amino acids 363 to 369; boldface and underlined). The amino acids that are changed in the rsa1-1 (K) and rsa1-2 (M) mutants are in boldface. The rsa1-1 mutation changes the lysine codon (AAG) at amino acid position 225 to a premature stop codon (TAG), and the rsa1-2 mutation changes the ATG start codon to TTG. The asterisk defines the stop codon of the RSA1 ORF.

RSA1 (YPL193W) is an ORF of 1,143 bp which encodes a protein of 381 amino acids with a predicted molecular mass of 44 kDa. Rsa1p is predicted to be slightly acidic (pI 5.77) and to localize to the nucleus (SwissProt PsortII program) due to the presence of a classical “seven-amino-acid pattern” nuclear localization signal that encompasses amino acids 363 to 369 (Fig. 2). Rsa1p probably exists in low intracellular abundance based on a codon adaptation index of 0.12. Sequence comparison searches showed that Rsa1p has neither an orthologue in the Caenorhabditis elegans genome nor homologues in higher eukaryotic genome or expressed sequence tag databases. Moreover, the amino acid sequence had no similarities to any characterized functional motifs.

Rsa1p is required for optimal cell growth.

As a first step in the functional analysis of Rsa1p, we constructed a rsa1 null allele. We replaced most of one RSA1 ORF copy in the diploid strain W303 with the kanMX4 or HIS3MX6 marker modules (see Materials and Methods). Tetrad analysis revealed that two of the four spore clones had a slow-growth phenotype, with the slow-growing spore clones always being G418 resistant or His⁺ (data not shown). Restreaking of complete tetrads on YPD plates at 18, 30, and 37°C showed that the rsa1 null mutant spore clones were clearly growing slower than the wild-type control spore clones at 18 and 30°C (Fig. 3A; data not shown); however, the most pronounced effect was at 37°C (Fig. 3A). Doubling times of 1.3 h for the wild-type and of 2.3 h for the rsa1 null mutant spore clones were obtained in liquid YPD medium at 30°C. The doubling time of the rsa1 null mutant was very similar to those of the SL strains sl264234 (rsa1-1) and sl285204 (rsa1-2), indicating that the two point mutations and the null mutation had identical effects on cell growth. Together these results indicate that Rsa1p is required for optimal cell growth and that the absence of Rsa1p leads to a TS growth phenotype.

FIG. 3.

The rsa1 null mutation confers slow growth and temperature sensitivity, and it synthetically enhances the weak slow-growth phenotype of the dbp6-3 mutant. (A) YDK44 (MATa/MATα rsa1::kanMX4/RSA1) was sporulated, and tetrads were dissected. A wild-type and a Δrsa1 spore clone from a representative tetrad are shown on YPD plates that were incubated for 48 h at 30 and 37°C, respectively. (B) Wild-type (YDK46-7B pRS414-HA-DBP6), Δrsa1 (YDK46-7A pRS414-HA-DBP6), dbp6-3 (YDK46-7B pRS414-dbp6-3), and Δrsa1/dbp6-3 (YDK46-7A pRS414-dbp6-3) strains are shown on a YPD plate that was incubated for 60 h at 30°C.

rsa1 and dbp6 synthetically interact.

To confirm that the synthetic-enhancement phenotype of the SL strains sl264234 and sl285204 was solely due to mutations in the RSA1 gene, we tested if the rsa1 null allele had synthetic effects on different dbp6 mutants (see Materials and Methods). As expected, the growth rates of strains carrying both the rsa1 null and the mild dbp6-2 or dbp6-3 mutation were substantially reduced compared to the growth rate of the single-mutant strains at 30°C (Fig. 3B; data not shown). Doubling times of 1.35 (wild type), 1.75 (dbp6-2 and dbp6-3), 2.35 (Δrsa1), and around 6 h (dbp6-2/Δrsa1 and dbp6-3/Δrsa1) were obtained at 30°C in YPD medium. On the other hand, no viable dbp6-4/Δrsa1 and dbp6-13/Δrsa1 double mutants could be recovered (see Materials and Methods), indicating that the rsa1 null allele and the strong dbp6-4 and dbp6-13 mutations were synthetically lethal (data not shown). However, 2μm plasmids harboring either RSA1 or DBP6 were unable to suppress the slow-growth or TS phenotype of different dbp6 alleles or the rsa1 null allele, respectively (data not shown).

These results confirm that rsa1 and dbp6 synthetically interact, and they show that the rsa1 null and point mutations similarly enhance the mild growth defect of weakly affected dbp6 mutants. Moreover, the basis of this synthetic interaction is likely to be specific, because the rsa1 null mutation only mildly exacerbates the strong slow-growth phenotype of the spb4-1 mutant (data not shown; see Materials and Methods).

Absence of Rsa1p leads to the accumulation of half-mer polysomes.

As in vivo depletion of Dbp6p leads to a deficiency in 60S ribosomal subunits and to the accumulation of half-mer polysomes (27), we determined if the absence of Rsa1p would give a similar phenotype. Polysome profile analysis of a rsa1 null mutant strain (YDK44-1B), grown at 30°C in YPD medium, revealed that the absence of Rsa1p resulted in the appearance of half-mer polysomes. The pool of free 60S ribosomal subunits, however, was only moderately decreased (Fig. 4B). When YDK44-1B was grown for 5 h at 37°C, a similar, but somewhat stronger, phenotype was observed. At 37°C, the amount of free 60S ribosomal subunits was more strongly decreased, the accumulation of half-mer polysomes was enhanced, and the overall reduction in polysomes became more apparent (Fig. 4C). Similar results were obtained when polysomes from the rsa1-disrupted sister spore clone YDK44-1C were analyzed (data not shown). In contrast, wild-type strains originating from the same tetrad (YDK44-1A and YDK44-1D) had normal polysome profiles at 30 and 37°C (Fig. 4A; data not shown). Furthermore, the similar loss-of-function natures of the rsa1 null mutation and the two point mutations were further confirmed by the fact that polysome profiles from these strains were essentially identical at 30°C (data not shown).

FIG. 4.

Absence of Rsa1p leads to the accumulation of half-mer polysomes. (A) YDK44-1A (wild type) grown at 30°C. YDK44-1B (Δrsa1) was grown at 30°C (B) or shifted to 37°C for 5 h (C). Cells were grown in YPD and harvested at an OD₆₀₀ of around 0.8. Cell extracts were resolved in 7-to-50% sucrose gradients. Gradients were analyzed by continuous monitoring at A₂₅₄. Sedimentation is from left to right. The peaks of free 40S and 60S subunits, 80S ribosomes (free couples and monosomes), and polysomes are indicated. Half-mers are indicated by arrows.

Quantification of total ribosomal subunits in low-Mg²⁺ sucrose gradients revealed that there was indeed only a moderate deficit in total 60S ribosomal subunits relative to 40S subunits; a ratio of A₂₅₄ for 60S ribosomal subunits to that for 40S ribosomal subunits of around 2 was observed for the wild type at 30 and 37°C in YPD medium and for the rsa1 null mutant at 30°C. This ratio decreased to around 1.7 when rsa1 null mutant strains were grown for 5 h at 37°C. The overall ribosomal subunit content was also reduced in the rsa1 null mutant at 30°C and more drastically at 37°C, as evidenced by the increased relative intensity of the “free-material” peak compared to the ribosomal subunit peaks (data not shown). Such a decrease in ribosomal material of both ribosomal subunits is a common feature of mutants affected in 60S-ribosomal-subunit production (see below).

Taken together, polysome profile analysis and ribosomal subunit quantification indicate that the absence of Rsa1p leads to a moderate decrease in the pool of free 60S ribosomal subunits and a relatively more drastic accumulation of half-mer polysomes at 30°C. Such a phenotype is characteristic of mutants defective in 60S-to-40S subunit joining, and similar polysome profiles have been observed upon depletion of or mutation in the large-subunit r-protein Qsr1p (16) or upon depletion of Sqt1p, a high-copy-number suppressor of a dominant-negative qsr1 allele (15). At 37°C, however, the total number of 60S ribosomal subunits is more drastically reduced, which is indicative of a deficiency in the formation of 60S ribosomal subunits in the absence of Rsa1p.

Formation of mature 25S rRNA is reduced in the rsa1 null mutant.

To study in more detail the role of Rsa1p in the metabolism of 60S ribosomal subunits, we analyzed the effects of the rsa1 null mutation on the synthesis and processing of pre-rRNA by [methyl-³H]methionine pulse-chase labeling experiments. For this purpose, strains YDK44-1A (wild type) and YDK44-1B (Δrsa1) were grown at 30 or 37°C (6 h at 37°C) in SD-Met medium to an OD₆₀₀ of around 0.8. The cells were pulse-labeled for 1 min and chased for 2, 5, and 15 min with an excess of cold methionine, and total RNA was extracted and analyzed (see Materials and Methods). In the wild-type RSA1 strain, formation of pre- and mature rRNAs was similar at 30 (Fig. 5A, lanes 1 to 4) and 37°C (Fig. 5B, lanes 1 to 4), with the mature rRNA species appearing slightly quicker at 37°C. In the rsa1 null mutant, there was an overall reduction in the formation of labeled pre- and mature rRNA species, which was more pronounced at 37°C. This is in agreement with the decrease in the steady-state levels of ribosomal material observed by polysome analysis and ribosomal subunit quantification (see above). Similarly, reduced formation of labeled rRNA species has been previously reported for other mutants affecting 60S-ribosomal-subunit assembly (9, 27). At the level of the formation of the mature rRNA species, we observed a slight decrease in the ratio of labeled 25S to 18S rRNA at 30°C (Fig. 5A, lanes 6 to 8) and a more drastic decrease at 37°C (Fig. 5B, lanes 6 to 8). However, the stability of the mature 25S rRNA was not affected throughout the duration of the chase. The processing pathway leading to formation of the 18S rRNA was also mildly impaired, as revealed by the lower levels of mature 18S rRNA (Fig. 5, lanes 6 to 8) and its 20S precursor (Fig. 5, lanes 5). Moreover, the 32S pre-RNA was absent and there was a weak accumulation of an aberrant 23S species (Fig. 5; compare lanes 1 and 5). Such a delay of pre-rRNA processing at sites A₀, A₁, and A₂ on the pathway to 18S rRNA formation is a general feature of mutations affecting the synthesis of 60S ribosomal subunits, and this has been proposed to be a consequence of a feedback mechanism that slows production of 18S rRNA when the formation of 25S and 5.8S rRNA is inhibited (6, 27, 64).

FIG. 5.

Absence of Rsa1p leads to reduced synthesis of the mature 25S rRNA. Strains YDK44-1A (RSA1) and YDK44-1B (Δrsa1) were grown in SD-Met medium at 30°C (A) or shifted for 6 h to 37°C (B). Cells were pulsed-labeled (p) for 1 min with [methyl-³H]methionine and then chased (c) for 2, 5, and 15 min with an excess of unlabeled methionine. Total RNA was extracted, and 20,000 cpm was loaded and separated on a 1.2% agarose–formaldehyde gel, transferred to a nylon membrane, and visualized by fluorography. The positions of the different pre-rRNAs and mature rRNAs are indicated.

These results indicate that the formation of mature 25S rRNA is preferentially, but mildly, reduced at 30°C and more drastically reduced at 37°C in the absence of Rsa1p. These findings are in agreement with the moderate decrease in the pool of free 60S ribosomal subunits, which is also more pronounced at the nonpermissive temperature.

Polysome profile analysis of dbp6/Δrsa1 double mutants.

To elucidate the basis of the synthetic enhancement or the synthetic lethality observed for the combination of the rsa1 null mutation with dbp6 mutations, the viable dbp6/Δrsa1 double mutants were subjected to polysome profile analysis. In addition, this analysis should also indicate whether Dbp6p was acting upstream of Rsa1p in the pathway of 60S-ribosomal-subunit assembly, as suggested by the distinct polysome profile phenotypes observed for the dbp6 mutants and the rsa1 null mutant (60S assembly versus 60S-to-40S subunit joining). To this end, wild-type, dbp6-2 and dbp6-3, Δrsa1, and dbp6-2/Δrsa1, and dbp6-3/Δrsa1 strains were grown at 30°C in YPD medium and their polysome profiles were recorded. The profiles obtained for the wild-type strain (Fig. 6A) and the rsa1 null mutant strain (Fig. 6C) were in agreement with the ones shown in Fig. 4A and B, respectively. As expected, the weak dbp6 mutants were deficient in free 60S versus 40S ribosomal subunits, and they accumulated half-mer polysomes (Fig. 6B; data not shown). The strength of these defects was correlated with the growth phenotypes of the dbp6 mutants, since the more severely affected dbp6-4 and dbp6-13 mutants (data not shown) and a strain genetically depleted of Dbp6p (27) were almost devoid of free 60S subunits and their total polysome content was strongly decreased. In the dbp6-2/Δrsa1 (data not shown) and the dbp6-3/Δrsa1 (Fig. 6D) double mutants, the overall polysome content was markedly decreased compared to that for dbp6-2 and dbp6-3 or Δrsa1 single mutants. Furthermore, the deficit in free 60S versus 40S ribosomal subunits and the accumulation of half-mer polysomes were more pronounced. However, the reduction in the levels of free 60S subunits was not as dramatic as in the stronger dbp6 mutants or upon depletion of Dbp6p, despite the longer or similar generation time of the dbp6/Δrsa1 double mutants. Overall, the polysome profiles of the dbp6/Δrsa1 double mutants were more similar to single-mutant profiles of dbp6-2 and dbp6-3 than to those of Δrsa1, indicating that the dbp6-2 and dbp6-3 mutations were epistatic over the rsa1 null mutation.

Taken together, these results indicate that Dbp6p is most likely acting upstream of Rsa1p in the pathway of 60S-ribosomal-subunit assembly and that the basis of the synthetic interaction between dbp6 and rsa1 is a combination of qualitative 60S assembly defects entailing specifically a synergistic decrease in the formation of 60S ribosomal subunits, which leads to a substantial deficit in overall polysome content. The presence of a considerable amount of free 60S ribosomal subunits in the dbp6/Δrsa1 double mutant indicates that the rsa1 null mutation affects not only the net formation of 60S ribosomal subunits but also an assembly step that is required for rendering 60S ribosomal subunits competent for 60S-to-40S subunit joining.

rsa1 and qsr1 synthetically interact.

Since polysome profile analysis suggested that the absence of Rsa1p might lead to a defect in 60S-to-40S subunit joining, we tested whether the rsa1 null mutation synthetically enhanced the slow-growth phenotype of the qsr1-1 mutant. To this end, we crossed the strain MMY3-3B (qsr1-1) (16), carrying the plasmid YCplac33-RSA1, to strain YDK45-11A (Δrsa1) and subjected the resulting diploid to tetrad analysis (see Materials and Methods). After plasmid shuffling on 5-FOA-containing plates, restreaking of tetratype tetrads on YPD plates indicated that the growth of a strain carrying both the qsr1-1 and the Δrsa1 mutation was substantially poorer than the growth of either single-mutant strain at 30°C (Fig. 7). Doubling times of 1.35, 2.35, 3.1, and 8.3 h were obtained for wild-type, Δrsa1, qsr1-1, and Δrsa1/qsr1-1 strains, respectively, grown at 30°C in YPD medium. We also generated dbp6/qsr1-1 and spb4-1/qsr1-1 double mutants (data not shown; see Materials and Methods), and, in agreement with a specific synthetic interaction between Δrsa1 and qsr1-1, we found that the qsr1-1 mutation only weakly exacerbated the slow-growth phenotype of the dbp6-4, dbp6-13, and spb4-1 mutants.

FIG. 7.

Synthetic enhancement of the slow-growth phenotype of the qsr1-1 mutant by the rsa1 null mutation. The strains MMY3-3B (qsr1-1), carrying the plasmid YCplac33-RSA1, and YDK45-11A (Δrsa1) were crossed, the resulting diploid was sporulated, and tetrads were dissected. Complete tetrads were restreaked on 5-FOA-containing plates to counter-select YCplac33-RSA1. A representative tetratype tetrad after 5-FOA counter-selection is shown on a YPD plate that was incubated for 72 h at 30°C.

Absence of Rsa1p leads to decreased Qsr1p levels on free 60S ribosomal subunits.

Since 60S ribosomal subunits devoid of Qsr1p are unable to join with 40S ribosomal subunits (16), we tested if the absence of Rsa1p would lead to decreased levels of Qsr1p on free 60S ribosomal subunits. For this purpose, wild-type control or rsa1 null mutant strains were grown at 30°C in YPD or shifted for 4 h to 37°C. Total cell extracts were prepared, and 10 A₂₆₀ units of cell extract was separated in 7-to-50% sucrose gradients. Figure 8A to C shows the polysome profiles of wild-type cells at 30°C, Δrsa1 cells at 30°C, and Δrsa1 cells at 37°C, respectively. From each gradient 13 fractions were collected; these were, along with 0.125 A₂₆₀ unit of total cell extract (lane T), subjected to Western blot analysis using polyclonal rabbit anti-Qsr1p and mouse monoclonal anti-Rpl3p antibodies. In the wild-type strain (Fig. 8A), Qsr1p and Rpl3p were, as expected and as previously reported (16), present in similar ratios in the total extract, 60S subunits (Fig. 8A, fraction 8), 80S monosomes (Fig. 8A, fractions 9 and 10), and polysomes (Fig. 8A, fractions 12 and 13). The same result was obtained for wild-type strains grown at 37°C (data not shown). However, in the rsa1 null mutant strain at 30°C (Fig. 8B), there was a specific reduction in the intensity of the signal of Qsr1p in the free 60S peak compared to the intensity of the signal of the 60S subunit r-protein Rpl3p (Fig. 8B, fraction 8; ≥2-fold reduction in Qsr1p levels on free 60S ribosomal subunits compared to the wild-type strain; see Materials and Methods), which was used as an internal control. This effect was more pronounced at 37°C (Fig. 8C, fraction 8; ≥10-fold reduction in Qsr1p levels on free 60S ribosomal subunits compared to the wild-type strain). On the other hand, Qsr1p was present in normal amounts in monosomes (Fig. 8B and C, fraction 10) and polysomes (Fig. 8B, fraction 13, and Fig. 8C, fractions 12 and 13).

FIG. 8.

Absence of Rsa1p leads to decreased levels of Qsr1p on free 60S ribosomal subunits. Shown are a polysome profile analysis and fractionation followed by Western blotting to detect the 60S r-proteins Qsr1p and Rpl3p. (A) YDK44-1A (wild type) grown at 30°C. YDK44-1B (Δrsa1) was grown at 30°C (B) or shifted to 37°C for 4 h (C). Cells were grown in YPD and harvested at an OD₆₀₀ of around 0.8. Cell extracts were resolved in 7-to-50% sucrose gradients. Gradients were analyzed by continuous monitoring at A₂₅₄. Sedimentation is from left to right. The peaks of free 40S and 60S subunits, 80S ribosomes (free couples and monosomes), and polysomes are indicated. Half-mers are indicated by vertical arrows. A total of 13 fractions were collected, proteins were concentrated by trichloroacetic acid precipitation, and equal volumes were resolved on sodium dodecyl sulfate–12% polyacrylamide gels and subjected to Western blotting. T, total extract. Numbers correspond to fraction numbers. The same blot was decorated simultaneously with polyclonal rabbit anti-Qsr1p and monoclonal mouse anti-Rpl3p antibodies to detect the 60S r-proteins Qsr1p and Rpl3p, respectively. The Qsr1p and Rpl3p signals are indicated.

Thus, the sucrose gradient fractionation experiments show that Qsr1p is to a substantial extent absent from free 60S ribosomal subunits in rsa1 null mutant strains, with the effect being more pronounced after shifting the mutant cells for 4 h to 37°C. This result suggests that Rsa1p is likely to be involved in an assembly step of pre-60S ribosomal subunits that is required for the efficient recruitment of Qsr1p.

Rsa1p localizes to the nucleoplasm and is excluded from the nucleolus.

To distinguish between a cytoplasmic, nuclear, or nucleolar role for Rsa1p in 60S-ribosomal-subunit assembly, we analyzed the subcellular localization of Rsa1p by indirect immunofluorescence. For this purpose, RSA1 was HA tagged at its 5′ end by fusion PCR and cloned into YCplac111 to express the N-terminally epitope-tagged Rsa1p (HA-Rsa1p) from its cognate promoter at approximately wild-type levels (see Materials and Methods). The resulting plasmid (YCplac111-HA-RSA1) and a control plasmid harboring the untagged RSA1 gene (YCplac111-RSA1) were transformed into the rsa1 null mutant strain YDK44-1B. HA-Rsa1p complemented the rsa1 null allele to the wild-type extent at 30 and 37°C. In addition, Western blot analysis with a monoclonal anti-HA antibody detected a single protein that migrated at a molecular mass of ca. 50 kDa in a total cell extract from a strain expressing HA-Rsa1p but not from a strain expressing untagged Rsa1p (data not shown). The HA-tagged Rsa1p was detected by anti-HA antibodies, followed by decoration with goat anti-mouse rhodamine-conjugated antibodies (Fig. 9B). For precise subnuclear localization, the nucleoplasm was visualized by staining the DNA with DAPI (4′,6-diamidino-2-phenylindole dihydrochloride) (Fig. 9C), and the nucleolus was stained with anti-Nop1p antibodies (40) in combination with goat anti-rabbit fluorescein-conjugated antibodies (Fig. 9A). Anti-Nop1p antibodies gave the crescent-shaped staining characteristic of nucleolar proteins (Fig. 9A), which was largely excluded from the DAPI-stained area (Fig. 9F, overlap in cyan). The HA-tagged Rsa1p localized to the nucleoplasm, as shown by its colocalization with DAPI (Fig. 9E, overlap in magenta). The absence of an HA-Rsa1p signal in the nucleolus (Fig. 9D, overlap in yellow) indicated that the distribution of HA-Rsa1p was restricted to the nucleoplasm. No signal was obtained with the combination of anti-HA and goat anti-mouse rhodamine-conjugated antibodies when cells of strain YDK44-1B YCplac111-RSA1 were analyzed by indirect immunofluorescence. The predominant localization of HA-Rsa1p in the nucleoplasm indicates that Rsa1p is likely to be implicated in a late nucleoplasmic step of 60S-ribosomal-subunit assembly.

FIG. 9.

HA-Rsa1p localizes to the nucleoplasm. Indirect immunofluorescence was performed with cells expressing HA-Rsa1p from the RSA1 promoter (YDK44-1B YCplac111-HA-RSA1). (A) Nop1p was detected by polyclonal rabbit anti-Nop1p antibodies, followed by decoration with a goat anti-rabbit fluorescein-conjugated antibody. (B) HA-Rsa1p was detected by the monoclonal mouse anti-HA 16B12 antibody, followed by decoration with a goat anti-mouse rhodamine-conjugated antibody. (C) Chromatin DNA was stained with DAPI. Pseudocolors were assigned to the digitized micrographs (A to C), and images were merged. The overlapping distributions are revealed in yellow for Nop1p and HA-Rsa1p colocalization (D), magenta for HA-Rsa1p and chromatin DNA colocalization (E), and cyan for Nop1p and chromatin DNA colocalization (F).

DISCUSSION

Dbp6p is an essential putative ATP-dependent RNA helicase of the DEAD-box protein family that is exclusively localized in the nucleolus. The phenotypes observed upon Dbp6p depletion suggest that Dbp6p is involved in assembly reactions within an early preribosomal particle that are required for the synthesis of 60S ribosomal subunits (27). In order to gain more insight into the functional environment of Dbp6p and to understand in more detail the complex process of 60S-ribosomal-subunit assembly, we performed an SL screen with conditional dbp6 alleles. In agreement with a role for Dbp6p in early assembly events on the pathway to formation of 60S ribosomal subunits, the RPL3 gene was cloned by complementation of the SL phenotype of two mutants from the screen (29). RPL3 encodes the L3 r-protein of the large ribosomal subunit, and Rpl3p has been shown to belong to a group of r-proteins that associate early with the preribosomal particle (30).

In this paper, we describe the functional analysis of the previously uncharacterized Rsa1p. The RSA1 gene was cloned by complementation of the slow-growth and synthetic-enhancement phenotype of one SL mutant, and it was shown to also complement a second independent SL mutant. Gene disruption revealed that Rsa1p is required for optimal cell growth since the rsa1 null mutant displays a slow-growth and TS phenotype. Several experimental findings suggest that the absence of Rsa1p has an effect on both the efficient (quantity) and correct (quality) formation of 60S ribosomal subunits; these findings are discussed below, and they lead to the proposition that Rsa1p acts on pre-60S ribosomal subunits during a late nucleoplasmic assembly event.

Polysome profile analysis and ribosomal subunit quantification indicate that the absence of Rsa1p leads to a moderate decrease in the pool of free 60S ribosomal subunits and to the accumulation of half-mer polysomes. Half-mer polysomes generally correspond to a 43S complex, consisting of a 40S ribosomal subunit with attached initiation factors awaiting, while stalled at the first start codon, the addition of a 60S ribosomal subunit (22). The appearance of half-mer polysomes can be due to a decreased synthesis of 60S ribosomal subunits that results in a net deficit of free 60S versus 40S ribosomal subunits; this has been described for mutants defective in the 60S r-proteins Rpl3p, Rpl5p, and Rpl16bp (11, 29, 35) and for mutants defective in components involved in pre-rRNA processing and 60S-ribosomal-subunit assembly (9, 10, 24, 27, 43, 63). Alternatively, the appearance of half-mer polysomes can be the consequence of defective assembly of 60S ribosomal subunits, leading to the formation of 60S ribosomal subunits that can no longer join the waiting 40S ribosomal subunits. In this case, there is practically no net deficit in free 60S versus 40S ribosomal subunits. Such polysome profiles have been observed in a strain with both Rpl24p-encoding genes disrupted (2), upon depletion of or mutation in the large-subunit r-protein Qsr1p (16) or upon depletion of Sqt1p (15). It has been shown that both Rpl24p and Qsr1p are only added to pre-60S ribosomal subunits in the cytoplasm (30, 65). Moreover, 60S ribosomal subunits devoid of Qsr1p are unable to join with 40S ribosomal subunits, whereas 60S ribosomal subunits that contain either wild-type or mutant (e.g., qsr1-1p) forms of Qsr1p are capable of subunit joining and of engaging in translation elongation. Therefore, it has been proposed that reduced association of wild-type or mutant Qsr1p with 60S ribosomal subunits may be the primary cause for a 60S-to-40S subunit joining defect and that this may also represent an additional translational regulatory mechanism (12, 16). The cytoplasmic Sqt1p has been cloned as a high-copy-number suppressor of a dominant-negative qsr1 allele. It is conceivable that Sqt1p is involved in assembling Qsr1p onto pre-60S ribosomal subunits late in the assembly pathway (15). Thus, the polysome profiles obtained for the rsa1-null mutant strains are clearly a consequence of the combination of moderately decreased 60S-ribosomal-subunit formation and a defect in 60S-to-40S subunit joining (see below); both of these effects are more pronounced at 37°C.

One effect of the absence of Rsa1p on the metabolism of 60S ribosomal subunits is the moderate decrease in the steady-state levels and in the formation of 60S ribosomal subunits, as observed by polysome profile analysis, ribosomal subunit quantification, and [methyl-³H]methionine pulse-chase labeling of pre- and mature rRNA. This deficit in the formation of mature 25S versus 18S rRNAs, and thus by consequence the formation of 60S versus 40S ribosomal subunits, is clearly visible at 37°C. The effect of Rsa1p’s absence on 60S-ribosomal-subunit formation is also seen in polysome profiles of the viable dbp6/Δrsa1 double mutants. In these double mutants, the deficit in free 60S versus 40S ribosomal subunits and the accumulation of half-mer polysomes are more pronounced than in either single mutant alone. Moreover, the overall polysome content is markedly decreased, which is most likely responsible for the increase in doubling time. Since the rsa1 null mutation only mildly exacerbates the strong slow-growth phenotype of the spb4-1 mutant, we conclude that the synthetic interaction between dbp6 and rsa1 is not simply due to the combination of two mutations that reduce the formation of 60S ribosomal subunits. Therefore, we propose that the basis of the synthetic enhancement of the dbp6 mutant phenotype by the rsa1 null mutation is the specific destabilization of pre-60S ribosomal subunits that are qualitatively altered due to a mutation in dbp6. Comparison of the double-mutant with the single-mutant polysome profiles indicates that the dbp6-2 and dbp6-3 mutations are epistatic over the rsa1 null mutation. This genetic argument, together with the finding that HA-Rsa1p is localized exclusively in the nucleoplasm, strongly suggests that the nucleolar Dbp6p functions upstream of Rsa1p in the pathway of 60S-ribosomal-subunit assembly.

The reduction in the levels of free 60S ribosomal subunits in the dbp6/Δrsa1 double mutants is, however, not as drastic as that in the severely affected dbp6 single mutants or as that upon depletion of Dbp6p, despite the longer or similar generation times of the dbp6/Δrsa1 double mutants. Moreover, the polysome profiles from rsa1 null mutant strains are similar to the ones from mutants defective in 60S-to-40S subunit joining. Therefore, the absence of Rsa1p is likely to have a second effect on 60S-ribosomal-subunit metabolism. We propose that the rsa1 null mutation not only leads to a net decrease in the formation of 60S ribosomal subunits but also affects an assembly step that is required for rendering 60S ribosomal subunits competent for 60S-to-40S subunit joining. In support of this hypothesis, we find that the rsa1 null mutation synthetically enhances the slow-growth phenotype of the qsr1-1 mutant in a specific manner. It has been shown that the amounts of qsr1-1p on 60S ribosomal subunits are reduced in the qsr1-1 mutant (16). Therefore, the basis of the synthetic interaction between rsa1 and qsr1 may be an enhancement of the inefficient loading of qsr1-1p onto pre-60S ribosomal subunits that are slightly misassembled due to the rsa1 null mutation. This genetic suggestion is confirmed by the finding that the amounts of Qsr1p on free 60S ribosomal subunits are indeed reduced in the rsa1 null mutant; again, this effect is more pronounced at 37°C.

On the basis of all the experimental evidence, we propose that Rsa1p is involved in a late nucleoplasmic step of 60S-ribosomal-subunit assembly. The molecular assembly defect associated with the absence of Rsa1p has a dual effect on the integrity of 60S ribosomal subunits. One effect is the reduced formation of 60S ribosomal subunits, and the other effect results in a less efficient recruitment of Qsr1p onto 60S ribosomal subunits, thereby affecting 60S-to-40S subunit joining. Consistent with this dual effect, overexpression of Qsr1p does not suppress the slow-growth and TS phenotype of the rsa1 null allele (data not shown); such a suppression would be expected if the absence of Rsa1p solely resulted in a Qsr1p loading defect. Moreover, it could well be that there are two populations of 60S ribosomal subunits in the rsa1 null mutant strain, with one population corresponding to wild-type 60S ribosomal subunits and the other to 60S ribosomal subunits that cannot recruit Qsr1p at all.

The existence of late maturation steps involving nuclear pre-60S particles containing the mature 25S rRNA was proposed in the 1970s by Trapman and Planta based on the observation that there are large amounts of mature 25S rRNA relative to 20S pre-rRNA in the nucleus (48). Moreover, the appearance of mature 25S rRNA in cytoplasmic 60S ribosomal subunits and in polysomes is not simultaneous, indicating that there also exists a pool of pre-60S particles that has to undergo final maturation in the cytoplasm before being able to associate with 40S ribosomal subunits to form active ribosomes (48). The finding that at least three 60S r-proteins are only incorporated into 60S ribosomal subunits in the cytoplasm (30, 65), as well as the discovery of trans-acting factors such as Nmd3p (23) and Sqt1p (15) that are involved in steps for the cytoplasmic assembly of 60S ribosomal subunits, is clearly in agreement with the latter proposition. On the other hand, Rsa1p is, to the best of our knowledge, the first protein that may function in a late nucleoplasmic step on pre-60S particles. This interpretation of the phenotypic analysis of Rsa1p would therefore directly support the postulated existence of maturation steps involving nuclear pre-60S particles containing mature 25S rRNA.

At present, due to the lack of a mechanistic model for Rsa1p function, we cannot rule out the formal possibility that the nucleoplasmic role of Rsa1p consists of the control of the expression, either at the level of transcription or pre-mRNA splicing, of one or a few r-proteins or trans-acting-factor-encoding genes. In such a scenario, the absence of Rsa1p would strongly decrease or abolish the production of an r-protein or a trans-acting factor, which in turn would mimic mutation in or genetic depletion of the putative Rsa1p target protein, and this could potentially account for the phenotypes observed in rsa1 null mutant strains. Since the sequence of Rsa1p does not have any homology to known proteins or characterized functional domains, it does not contribute to the understanding of Rsa1p’s function. Therefore, future genetic and biochemical experiments are required to elucidate the precise mechanism of action of Rsa1p. The identification of proteins that physically interact with Rsa1p in particular should shed light on the functional environment of Rsa1p and reveal whether Rsa1p is indeed directly involved in a late nucleoplasmic step of 60S-ribosomal-subunit assembly.