Isolation of rrp20-1 mutant and identification of RRP20
A screen for mutants that die on loss of mtDNA (ρ0-lethality) was performed after EMS mutagenesis of M2915-6A. Strains that could grow on complete medium but failed to survive after EB elimination of mtDNA were chosen and crossed with a wild-type strain (see Materials and Methods). Upon tetrad analysis, one mutant exhibiting a 2:2 segregation of EBr:EBs phenotypes, indicative of a single nuclear mutation, was isolated and subsequently referred to as rrp20-1. In addition to the EBs phenotype, the rrp20-1 strain exhibits a severe growth rate reduction on rich medium at 30°C (Fig. B). Both phenotypes were recessive, thus cloning of the corresponding wild-type gene was performed by functional complementation after transformation with a yeast DNA library in YEp13-m4. A complementing plasmid, yBCD26, carrying a fragment from chromosome XV containing the entire YOR145c (renamed RRP20) and YOR146w ORFs lying on the opposite strand as well as part of EFD1 and YOR147w, was recovered. In addition, the ORF finder program (NCBI) revealed another possible ORF lying opposite to the 3′-region of the RRP20 gene (Fig. A).
To investigate the location of complementation activity, multicopy subclones, pEFD1, pRRP20 and pRRP85-274 from yBCD26, were generated and transformed into the rrp20-1 mutant. As shown in Figure B, plasmids pRRP20 (row 3) and pRRP85-274 (row 4) were able to complement ρ0-lethality, whereas a strain carrying plasmid pEFD1 (row 2) could not grow on EB. In addition, the transformants containing the complementing plasmids exhibited a better growth on GYP to almost resemble the wild type. As the pRRP85-274 plasmid contains not only the truncated RRP20 but also an uncharacterized ORF (marked by a question mark in Fig. A), the question arises as to which region is required for survival of ρ0 cells and recovery of slow growth. To answer this question, in vitro mutagenesis was carried out to introduce stop codons at nucleotides 6, 27 and 132 of the unidentified ORF that are silent for RRP20 (see Materials and Methods). In parallel, simultaneous disruption of RRP20 and the unidentified ORF in the diploid strain CS5 was done by insertion of a kan marker into the BstEII site (Fig. A), generating the heterozygous strain CS5Δrrp20. This disruption was lethal as tetrad dissection yielded a 2:2 segregation of viable to non-viable spores where all viable spores were G418s (Fig. C). The mutated constructs, pRRP-W2, pRRP-W9 and pRRP-W44 (rows 5, 6, 7, respectively), were transformed into the CS5Δrrp20 diploid strain. After sporulation and ascus dissection, rescue of lethality by the three mutated plasmids is apparent (Fig. C). The G418r clones, indicating that they contain a disrupted allele, displayed Ura+; thus, they harbor the complementing pCXJ15 recombinant plasmids (data not shown). These results not only reveal that RRP20 is responsible for complementation but they also indicate that the lethal phenotype is caused by disruption of this gene. Finally, genetic linkage analysis using the integration of a RRP20 allele associated to the marker URA3 at the RRP20 genomic locus also confirmed that an ability to restore the EBr phenotype of rrp20-1 was in fact due to RRP20 (data not shown).
Rrp20p is a putative RNA-binding protein conserved within eukaryotes
gene encodes a protein of 274 amino acids with a predicted molecular weight of 30.3 kDa and a predicted pI
of 9.87. A BLAST search (21
) demonstrated that Rrp20p shows strong similarity with potential RNA-binding proteins of at least six other organisms, including the fission yeast Schizosaccharomyces pombe
, Arabidopsis thaliana
, Caenorhabditis elegans
, Drosophila melanogaster
, Mus musculus
and human (Fig. ). The functions of these homologs have not previously been investigated. There is 53–72% sequence identity and 73–84% sequence similarity between Rrp20p and these six proteins. The sequence alignment revealed that the N-terminal regions of these proteins lack sequence conservation, whereas the C-terminal regions are more similar and contain a well conserved K homology (KH) domain as reported in the InterPro database (22
) (Fig. ). In Rrp20p, this domain extends from residue 179 to residue 252. The KH domain was first identified in the human hnRNP K and is an evolutionarily conserved sequence of ~70 amino acids found in a wide range of RNA-binding proteins from various organisms (28
). The presence of this domain suggests that Rrp20p is a potential RNA-binding protein.
Figure 2 Protein sequence analysis of deduced S.cerevisiae Rrp20p. A BLASTP search of the databases, using the putative Rrp20p protein sequence as query, yields candidate homologs from various organisms. The actual roles of all homologs have not previously been (more ...)
The C-terminal part of Rrp20p containing the KH domain is essential for its function
To further characterize Rrp20p, truncations of RRP20 were subcloned into low copy or multicopy plasmids to identify the essential region. All Rrp20p truncations (Fig. , rows 4 and 8–10) were introduced to the CS5Δrrp20 heterozygous diploid strain followed by sporulation and tetrad dissection. As shown in Figure C, pRRP85-274 (row 4), carrying a N-terminally truncated Rrp20p missing the first 84 amino acids, can complement the lethal phenotype to a similar extent as plasmids carrying the full-length protein (yBCD26 and pRRP20). This indicates that the N-terminal part of Rrp20p is dispensable for its function and is in a good correlation with the low similarity of the N-terminal extensions shown in the sequence alignment (Fig. ). In contrast, pRRP1-110 (row 8), expressing a truncated Rrp20p lacking the last 174 residues and thus the KH domain, failed to abolish lethality as only a 2:2 segregation of viable to non-viable spores was obtained. Interestingly, the smallest truncated version of Rrp20p, which was just able to complement the lethal phenotype of the RRP20 disruption when overexpressed, contains amino acids 85–266 (pRRP85-266; row 9). However, when this truncated protein was expressed at lower level (pMonoRRP85-266; row 10) there was no complementation. All viable spores were G418s and microscopic examination of the non-viable spores showed no evidence of any germination. Interestingly, sequence analysis revealed that rrp20-1 has a single point mutation that converts codon 235 from GGC to GAC resulting in the substitution of Gly235 by an Asp (data not shown). This glycine residue is conserved among the six orthologs and belongs to the KH domain (Fig. ).
rrp20-1 has a severe deficit in 18S ribosomal RNA and accumulates aberrant 22S and 23S intermediates
) has been shown to be an essential gene whose GFP-fusion protein is localized in the nucleolus (30
), an organelle dedicated to ribosomal biosynthesis. To determine if Rrp20p plays a role in this process, we first compared the mature rRNA content of rrp20-1
and wild-type cells. Total RNAs from both strains were separated on an agarose gel, stained by ethidium bromide and visualized under UV light. As shown in Figure A, the rrp20-1
strain (lane 2) exhibits a clear deficit in 18S rRNA indicating that ribosomal biogenesis is impaired. In order to define more precisely which pre-rRNA processing steps are affected in the mutant strain, the steady-state levels of the various pre-rRNAs were assessed by northern blotting. Total RNAs from both wild type and rrp20-1
strains were separated in an agarose-formaldehyde gel, transferred to a nylon membrane and detected by serial hybridizations with specific probes. Prior to hybridization, the membrane was stained with methylene blue to visualize and mark the location of the major 18S and 25S rRNA species (Fig. A′). The location of the probes used to detect the various processing intermediates is shown on the structure of 35S pre-rRNA. As seen in Figure B, probing with oligonucleotide B, which hybridizes upstream of the A0
cleavage site, shows that the rrp20-1
mutant (lane 2) accumulates the 35S pre-rRNA and a 23S aberrant processing product. While 35S precursor can be detected by all probes used, the aberrant product is only detected by probes B, C, D and G, confirming that this atypical intermediate corresponds to the previously described 23S, which extends from the 5′ end of the 5′-ETS to the A3
site. Traces of the 23S RNA species are also detected in the wild type. As shown in Figure C, hybridization with probe C, which binds on the 5′ side of the A2
cleavage site, reveals a reduced amount of the 20S pre-rRNA, which is the direct precursor of the 18S rRNA. In addition, there is an accumulation of a pre-rRNA species below 35S, which presumably corresponds to the 33S precursor since it was also detected by the riboprobe G, delimited the A0
spacer fragment (Fig. G). Furthermore, an unusual precursor migrating between the 23S and the 20S was detected in the mutant (lane 2) but not in the wild type (lane 1). As it was also detected with probe D and the riboprobe G but not with probes B, E and F (Fig. B, D–G), it may correspond to a 22S RNA extending from the A0
to the A3
sites. Consistent with the production of 22S and 23S RNAs, we observed that the 27SA2
pre-rRNA is strongly depleted whereas the level of 27SB intermediates as well as the amount of mature 25S rRNA remains similar to that of the wild type strain (Fig. A, A′ and D–F).
Primer extension analysis was carried out to determine the precise steps at which pre-rRNA processing is blocked in the rrp20-1 mutant. Total RNAs isolated from exponentially grown wild-type and rrp20-1 cells were annealed with oligonucleotides H or F. Oligonucleotide H hybridizes to the 5′ end of 18S rRNA which allows detection of processing sites A0 and A1, while oligonucleotide F binds to a region in ITS2 which allows A2, A3, B1L and B1S sites to be detected. In the mutant strain (lane 2), there is an increase of the signals of the primer extension stops at sites A0 and A3 associated with a decrease of the signals of the primer extension stops at sites A1 and A2 (Fig. ). This is in good agreement with the detection of a 22S RNA extending from A0 to A3 sites. It also confirms the accumulation of the 33S pre-rRNA and the depletion of the 20S and 27SA2 pre-rRNAs, which results in a deficit of mature 18S rRNA. The levels of 27SBL and 27SBS, as shown by the primer extension stops at sites B1L and B1S, appears slightly increased in the rrp20-1 mutant without affecting the intermediate ratio between the two strains. Finally, processing at all sites of intermediate products tested were accurate at the nucleotide level (Fig. ). The results of both northern hybridization and primer extension analyses indicate that in the rrp20-1 strain, the deficiency in 18S rRNA production is due to an inhibition of the pre-ribosomal RNA early cleavages at sites A0, A1 and A2.
Figure 4 Primer extension analysis of intermediates in rrp20-1. Total RNAs were isolated from exponentially grown wild type (lane 1) and rrp20-1 (lane 2) strains. Primer extension analysis was performed using oligonucleotides F and H. The reaction products were (more ...)
rrp20-1 does not affect the steady-state levels of snoRNAs U3, U14, snR10 and snR30
Four snoRNAs have been previously shown to be necessary for the early A0
cleavages. These are snoRNAs U3, U14, snR10 and snR30 (reviewed in 31
). It has been reported that genetic depletion of these snoRNAs inhibits early cleavages at sites A0
. To ascertain whether the processing deficiency in these cleavages is a direct consequence of the mutation in RRP20
, steady-state levels of the snoRNAs U3, U14, snR10 and snR30 were assessed. Total RNAs from both wild type and rrp20-1
strains were separated in a polyacrylamide–urea gel, transferred to a nylon membrane and detected by serial hybridizations with specific oligonucleotide probes. As shown in Figure , levels of U3 or the other snoRNAs species, including U14, snR10 and snR30 are not altered in the rrp20-1
mutant. This supports a direct involvement of Rrp20p in pre-rRNA processing and possibly in 40S ribosomal subunit biogenesis.
Figure 5 rrp20-1 mutation does not affect the levels of snoRNAs. Equal amounts (4 µg) of total RNAs isolated from wild type (lane 1) and rrp20-1 (lane 2) strains were resolved on an 8% acrylamide denaturing gel and transferred to a nylon membrane. (more ...)