Wide mutational analysis to ascertain the functional roles of eL33 in ribosome biogenesis and translation initiation

Identification of different rpl33a mutations

We previously identified the rpl33a-G76R mutation as a spontaneous suppressor of the inability of the gcn2 gcn3 double mutant strain H117 to derepress GCN4 mRNA translation under amino acid starvation, producing a Gcd− phenotype (Martin-Marcos et al. 2007). The G76R substitution also impairs the efficient processing of pre-rRNAs and reduces the levels of mature rRNAs, indicating a role for r-protein eL33 in ribosome biogenesis. Gly-76 is localized between two β-sheets in a motif of 22 amino acids well conserved in all the members of the r-protein L35Ae family. To investigate which eL33A residues are important for proper ribosome biogenesis or/and efficient translation, and to explore in yeast the functional consequences of substitutions associated with DBA and cancer in human rpL35A/eL33, we conducted site-directed mutagenesis of the paralogous gene RPL33A to alter amino acid residues that (i) interact with r-proteins of the dII/dVI cluster (rpL6/eL6 and rpL16/uL13), with L32/eL32 and with different nucleotides of the 60S domains I, II and VI and ES7 and ES39 (Ben-Shem et al. 2011; Ohmayer et al. 2015) (Fig. 1A and D and Supplementary Data); (ii) map to the motif of 22 amino acids conserved in all the members the r-protein L35Ae family, located between Gly-69 and Pro-90 (Fig. 1A); (iii) found mutated in human Diamond–Blackfan anemia (DBA) (Farrar et al. 2008) and in different tumors (Catalogue of Somatic Mutations in Cancer, COSMIC) (Forbes et al. 2008) (Fig. 1B); and (iv) are targets of post-translational modifications (Henriksen et al. 2012; Swaney et al. 2013; Weinert et al. 2013) (Fig. 1A).

Fig. 1figure 1

A Amino acid sequence of yeast ribosomal protein eL33A. Secondary structures are represented as a bar (α-helix), and arrows (β-sheets). Residues substituted by site-directed mutagenesis are highlighted in colors: green, red and pink for those amino acids that interact with ribosomal proteins L6 (eL6), L16 (uL13), and L32 (eL32), respectively; blue, for those that map to a conserved motif of 22 amino acids located between Gly-69 and Pro-90 in the carboxy-terminal region of eL33, or those that were found mutated (i) in patients with Diamond-blackfan anemia (DBA) or (ii) in different tumor entities (Catalogue o Somatic Mutations in Cancer, COSMIC); orange, for residues that are targets of post-translational modifications; the black triangle indicates the position of a C-terminal deletion of the L33A protein described in patients with DBA. Multiple-Ala substitutions of consecutive residues are represented with bars. Red bars and red asterisks above indicate amino acid substitutions that produce Gcd− and/or slow growth phenotypes at different temperatures. B Amino acid sequence comparison between yeast and human eL33. Residues highlighted were found mutated in patients with DBA (red) or in several tumor entities (blue) and the correspondent mutation is marked above. The ∆ symbol indicates a leucine deletion C Structure of eL33 where mutated residues that confer different phenotypes are signaled in red. D Position of eL33 in the 60S ribosomal subunit. The yeast 60S ribosomal subunit is shown viewed from the solvent exposed side with the 25S rRNA in yellow and ribosomal proteins in grey. Ribosomal protein L33 (eL33) is shown in blue, L6 (eL6) in green, L16 (uL13) in red, and L32 (eL32) in pink

Single amino acids that interact with eL6 (Ser-4, Arg-6, Leu-7, Tyr-8, Gly-34, Val-35, Asp-40 Phe-43, Tyr-44 and from Leu-102 to Ile-107), with uL13 (Lys-12, Glu-33 and Gly-95) and with eL32 (Val-22), and residues localized close to Gly-76 (Gly-69, Arg-73, Gly-79 and Arg-82) were substituted either by Ala, which replaces bulky or charged side chains with a methyl group, or by Arg or Glu that introduce a positive or a negative charge, respectively. We also generated in the yeast RPL33A paralogous gene the corresponding human substitutions and truncations associated with DBA (∆L29, V35I and ∆99–107) and with different tumor entities (S4P, L7Q, S28T, R48I, A50T, S56N, G79C, R82C and F94T), and the above indicated amino acid residues were also substituted by Ala, Arg and Glu. The Lys residues at positions 31, 47 and 92, which are susceptible to post-translational modifications, were substituted by Ala. In addition, multiple-Ala substitutions of consecutive residues as indicated in Fig. 1A were generated; although in mutants A40-44 and A92-99, not every amino acid residue was substituted (DAQFY40-44AAQAA and KTFGASVR92-99AAAGAAVA). As shown in Fig. 1A, C, the substitutions are localized in different structural elements of eL33 as that would help to elucidate which specific elements or residues are required for proper function of this essential r-protein.

The appropriate mutations were generated in RPL33A under its native promoter on low copy (lc) LEU2 or URA3 plasmids; the resulting URA3 plasmids were introduced into the gcn2 gcn3 rpl33aΔ RPL33B Hm538 strain to screen for mutations that confer 3AT resistance (3ATR), indicating the Gcd− phenotype. The Hm538 strain is sensitive to 3AT, an inhibitor of histidine biosynthesis, because its gcn2 and gcn3 mutations impede derepression of GCN4 under histidine starvation conditions imposed by 3AT. We found that in a first set of mutants (from now on Set 1), the single amino acid substitutions rpl33a-Y44R, G69R, L102R and the previously described G76R, as well as the double mutant G76R-G79R, the nine amino acid carboxy-terminal deletion (∆99–107), and carboxy-terminal blocks of Ala substitutions A102-105 and A102-107, all conferred 3ATR/Gcd− phenotypes of different degrees (Fig. 2A). This Set 1 group of rpl33a of substitution mutations affect residues that (i) interact with eL6 (Tyr-44 and from Leu-102 to Ile107)) (Figs. S2-B and S1-B) and the 25S rRNA ES7 (Pro-104 and Asn-106) (Fig. S1-C) (ii) belong to the 22-residue motif conserved in the L35Ae family (Gly-69, Gly-76 and Gly-76, Gly-79), or (iii) were found mutated in patients with DBA (∆99–107) or in tumor entities (Gly-79).

Fig. 2figure 2

Set 1 rpl33a mutations confer both Gcd− and/ or Slg− phenotypes. A 105 cells of gcn2 gcn3 rpl33a∆ RPL33B strain Hm538 containing the indicated RPL33A alleles on low copy (lc) plasmids or an empty vector were spotted on synthetic complete (SC) medium lacking uracil (Ura) and SC medium containing 5 mM or 10 mM 3-Amino-1,2,4-triazole (3AT) and incubated for 5 days at 28 °C. B Derivatives of strain Hm700 his4-301 rpl33a∆ RPL33B containing the indicated RPL33A alleles on lc plasmids or an empty vector also harboring HIS4-lacZ reporter with an AUG start codon (plasmid p367) were cultured in synthetic dextrose minimal medium (SD) supplemented with histidine (His) and tryptophan (Trp) at 28 °C to A600 of ~ 1.0, and β-galactosidase activities were measured in whole-cell extracts (WCEs) in units of nanomoles of o-nitrophenyl-β-D-galactopyranoside cleaved per min per mg. The mean and standard error (SE) from at least four independent transformants are reported. The values in the right column are the results expressed relative to the corresponding WT value (1.0). C Ten-fold serial dilutions of his4-301 rpl33a∆ RPL33B strain Hm700 containing the indicated RPL33A alleles on lc or high copy (hc) plasmids or empty vectors were grown at 28 °C, 37 °C and 18ºC for 3.5, 4.5 and 7 days, respectively, on SC medium lacking leucine (Leu). D Ten-fold serial dilutions of his4-301 rpl33a∆ rpl33b∆ strain Hm701 containing the indicated RPL33A alleles on lc plasmids were grown at 28 °C, 37 °C and 18ºC for 2.5, 3 and 6 days, respectively, on SC-Leu

The rpl33a∆ null allele in strain Hm538 was constructed to preserve the integrity of YPL142C, a hypothetical ORF encoded on the DNA strand opposite to that encoding RPL33A. Since we frequently observed the appearance of Slg+ revertants or spontaneous suppressors in the Hm538 strain, we deleted the complete RPL33A ORF to produce strain Hm700, of genotype GCN2 GCN3 rpl33a∆ RPL33B his4-301, which did not produce Slg+ colonies. This new strain also allowed us to investigate both, whether any of the mutant rpl33a alleles produce 3AT-sensitivity indicative of a Gcn− phenotype, and/or reduce the stringency of AUG start codon selection by suppressing the His− phenotype conferred by an initiation codon mutation of his4-301.

Under amino acid starvation conditions, the GCN4 protein activates transcription of the HIS4 gene among many other amino acid biosynthetic genes (Hinnebusch 1988). Gcd− mutations increase expression of a HIS4-lacZ reporter under non-starvation conditions, because they constitutively derepress GCN4 expression (Harashima and Hinnebusch 1986). To provide a quantitative estimation of the strength of the mutant Gcd− phenotypes, an empty vector, WT RPL33A, or mutant rpl33a alleles were introduced on lc LEU2 plasmids into the rpl33a∆ RPL33B his4-301 strain Hm700. The transformants were cultured in non-starvation conditions (SD medium) and β-galactosidase activities synthesized from a HIS4-lacZ reporter with a WT AUG start codon were measured in whole-cell extracts (WCEs). Consistent with their Gcd− phenotypes, all the 3ATR mutants showed a two- to threefold increase in β-galactosidase expression under amino acid sufficiency conditions, which were similar to those of the rpl33a∆ transformants containing only empty vector, in comparison to the WT RPL33A strain (Fig. 2B).

In addition to the 3ATR/Gcd− phenotypes, the Set 1 of rpl33a mutations confer a slow growth (Slg−) phenotype on synthetic complete (SC) medium at 28 °C, a severe temperature sensitivity at 37 °C (Ts−) and cold sensitivity (Cs−) at 18 °C (Fig. 2C, lc panels), suggesting that these mutations impair functions of eL33 that affect essential processes of the cell. The Slg− phenotypes at 28 °C are relatively more severe in the rpl33a-G69R, -G76R, -G76R-G79R mutants and the strong Slg− phenotype of rpl33a-∆99–107 is comparable to that of the rpl33a∆ mutant. At 37 °C, all the rpl33a mutants showed similar Ts− phenotypes; and at 18 °C, the Slg− phenotypes are similar or stronger than those observed at the permissive temperature in all the mutants, except for rpl33a-Y44R.

The parental rpl33a∆ RPL33B his4-301 Hm700 strain cannot grow on medium lacking histidine (His), because the mutant his4-301 mRNA lacks an AUG start codon (His− phenotype). Sui− mutations increase initiation at the third (UUG) codon at his4-301 and restore the ability to grow in a medium without His (His+ phenotype). However, we found that none of the rpl33a mutations allowed detectable growth on medium without His (His− phenotype) suggesting the absence of marked Sui− phenotypes (data not shown).

The rpl33a mutant alleles were then introduced on high copy (hc) plasmids into the same rpl33a∆ RPL33B his4-301 Hm700 strain, to examine whether overexpression of the mutant proteins would compensate for possibly suboptimal levels of eL33 and attendant reductions of 60S subunits when the alleles are expressed from lc number plasmids instead. As shown in Fig. 2C, at 28 °C, hc rpl33a-Y44R suppressed the Slg− phenotype observed with lc rpl33a-Y44R. Surprisingly, although expression of all rpl33a mutant alleles from hc plasmids slightly suppressed the Ts− phenotypes of all Set 1 mutants at 37 °C, the same phenotypes were observed in the corresponding mutants bearing a high copy number empty vector, (Fig. 2C, central panels, bottom raw) indicating that the suppression, at least at 37ºC cannot be attributed to increased expression of the eL33 variants. At 18 °C, the Cs− phenotypes of rpl33a-Y44R, G69R, L102R, A102-105 and A102-107 mutants were slightly suppressed when mutant rpl33a alleles were expressed from hc plasmids. This suppression is most pronounced in the rpl33a-G69R mutant (Fig. 2C). In contrast, we did not observe any suppression of the Cs− phenotype by rpl33a-G76R, G76R-G79R or 99–107 mutant alleles expressed from hc plasmids (Fig. 2C).

Because strain Hm700 contains intact the RPL33B allele, which could be alleviating the phenotypes of rpl33a mutations, we constructed the double mutant rpl33a∆ rpl33b∆ strain Hm701, harboring a WT RPL33A allele on an URA3 lc plasmid. The rpl33a mutant alleles generated by site-directed mutagenesis that did not show 3ATR or Slg− phenotypes at 28 °C in strains Hm538 and Hm700, respectively, were introduced into the new strain Hm701 on a LEU2 plasmid, (Set 2 group mutants), and the URA3 RPL33A plasmid was evicted by counter-selection on medium containing 5-FOA (Boeke et al. 1987). At 28 °C, only rpl33a-Y103R and A40-44 conferred a slight/modest Slg− phenotype. At 37 °C, rpl33a-F43R, G79C, L102A, Y103A, P104R and S105R produced a slight Ts− phenotype (Fig. 2D and data not shown), which is more pronounced in the rpl33a-L7R, ∆L29 and Y103R mutants (Fig. 2D). After 3 days of incubation at 37 °C on SC medium, no detectable growth was observed in the rpl33a-V35R, G79R, A40-44 and A104-107 mutants, indicating severe Ts− phenotypes. None of the analyzed mutants showed Cs− at 18 °C or His+ phenotypes (Fig. 2D and data not shown). These substitutions affect residues that (i) interact with eL6 (Leu-7, Val-35, Phe-43, Tyr-103, Asp-40-Tyr-44 and Pro-104-Ile-107) (Figs. S3 and S1-B), (ii) were found mutated in human DBA and in different tumor entities (Leu-29, Val-35 and Leu-7), or (iii) interact with different nucleotides of the 60S rRNA domain II (Leu-29) (Fig. S4-A), domain I and ES39 (from Lys-92 to Arg-99) (Fig. S4-B) and ES7 (Pro-104 and Asp106) (Fig. S1-C).

Defects in pre-rRNA processing caused by mutations in rpl33a

To investigate defects in the pre-rRNA maturation pathway caused by mutations in different regions of RPL33A, we conducted Northern analysis of pre-rRNA and mature mRNA species in WT and the two sets of rpl33a mutant cells grown to mid-logarithmic phase in SC medium at 28ºC (Figs. 3B, C and 4A), and after shifting the Set 2 group of mutants to 37 ºC for 6 h (Fig. 4B). The pre-rRNA processing pathway in S. cerevisiae and the probes used for Northern analysis are shown in Fig. 3A. The RNAs were quantified relative to the level of SCR1, the RNA component of the Signal Recognition Particle (SRP) transcribed by RNA polymerase III, used as an internal control.

Fig. 3figure 3

Defects in ribosomal RNA processing caused by Set 1 rpl33a mutations. A Scheme of the yeast rRNA processing pathway. The 35S pre-rRNA contains the sequences for mature 18S, 5.8S, and 25S rRNAs separated by two internal transcribed spacer sequences (ITS1 and ITS2) and flanked by two external transcribed spacer sequences (5’ ETS and 3’ETS). The rRNAs are represented as bars and the transcribed spaces as lines. The processing sites and the annealing positions of oligonucleotides used as probes are indicated by the letters A to E above the diagram and by the numbers 1 to 6 beneath all rRNAs, respectively. B Derivatives of strain Hm700 (his4-301 rpl33a∆ RPL33B) containing the indicated RPL33A alleles on lc or hc plasmids or empty vectors were cultured in liquid SC-Leu at 28 °C to A600 of ~ 1.0. Total RNA was extracted and samples containing 10 µg were resolved on 1.2% agarose 4% formaldehyde gels and subjected to Northern blot analysis. The RNA species detected by consecutively hybridizations of the blot are labeled on the right. The steady-state levels (%) of 25S and 18S rRNAs normalized with SCR1 are indicated

Fig. 4figure 4

Defects in ribosomal RNA processing caused by Set 2 rpl33a mutations. Derivatives of strain Hm701 (his4-301 rpl33a∆ rpl33b∆ containing the indicated RPL33A alleles on lc plasmids were cultured in liquid SC-Leu at 28 °C to A600 of ~ 1.0 (A) or grown to mid-logarithmic phase in SC-Leu at 28 °C and then shifted to 37 °C for 6 h (B). Total RNA was extracted and samples containing 10 µg were subjected to Northern analysis as indicated in Fig. 3B

First, we analyzed the effects of Set 1 rpl33a mutations expressed from lc plasmids in rpl33a∆ RPL33B strain Hm700 on the pre-rRNA processing pathway and on the level of RPL33A mRNA (Fig. 3B). Appreciable reductions (~ 40–90%) in the steady-state levels of 25S rRNA were observed in all the Set 1 lc rpl33a mutants; and in lc rpl33a-Y44R, G76-G79R and ∆99–107 those reductions were comparable to that of the empty vector transformant of Hm700 (~ 90%) (Fig. 3B, Table 4). The abundance of 18S rRNA was also reduced in all the rpl33a mutants, but in general to a lesser extent than the reductions of 25S rRNA, (~ 25–60% and ~ 70% in rpl33a∆) (Fig. 3B, Table 4), indicating a stronger effect of these rpl33a mutations in the maturation pathway of the 60S versus 40S subunit. Early cleavages at the A0-A1-A2 sites of the 35S pre-rRNA were reduced in all the rpl33a mutant strains, with attendant increases in 35S/32S, 35S/27SA2, 35S/27S, and 35S/20S ratios, with relatively greater increases observed in rpl33a-Y44R, G76R, G76-G79R, ∆99–107, A102-107 and rpl33a∆ mutants (Fig. 3B, Table 5). Almost no detectable amounts of 33-32S precursors, and substantial reductions in the levels of 27SA2, 27S, 7S and 20S pre-rRNAs were observed in all the rpl33a mutants, which is consistent with the lower amounts of 25S, 5.8S and 18S rRNAs detected in the mutants in comparison with the WT. However, the 27SA2/25S, 27S/25S and 27S/7S ratios were higher in the mutants than in WT (Table 5), likely reflecting defects in processing of the 27S pre-rRNA species at A3, B1S and C2 sites, as previously reported for the rpl33-G76R mutant and strains lacking RPL33A (Farrar et al. 2008; Martin-Marcos et al. 2007; Poll et al. 2009). Although mature 18S rRNA is less abundant in these mutants than in the WT (~ 25–70%), there was also a comparable reduction in the level of 20S pre-rRNA, indicating that these mutations do not affect processing of 20S pre-rRNA to 18S rRNA. Although the levels of 5.8S were diminished (~ 40–65%) in the rpl33a mutants and ~ 70% in rpl33a∆ (Fig. 3B, Table 4), the ratio 7S/5.8S was similar or slightly lower in the mutants than in the WT (Table 5), suggesting that the processing of 7S to 5.8S is normal in rpl33a mutant cells.

Table 4 Steady-state levels of mature rRNAs and mRNA RPL33A, normalized with SCR1Table 5 Ratios between different species of pre- and mature rRNAs

As expected, the RPL33A mRNA is absent in the strain transformed with an empty vector (rpl33a∆) (Fig. 3B). From lc plasmids, rpl33a-102R, A102-105 and A102-107 mutants showed RPL33A mRNA levels similar than those of WT, whereas rpl33a-Y44R, G69R, G76R, G76R-G79R, and -∆99-107 showed significant increases in RPL33A mRNA abundance (Fig. 3B, Table 4). When WT RPL33A is expressed from a hc plasmid, the levels of RPL33A mRNA increase ~ fivefold with respect to lc RPL33A (Fig. 3B and C); and similar increases from ~ 3 to ~ ninefold (A102-107 mutant) in the amounts of RPL33A mRNA were detected when rpl33a mutant alleles were expressed from hc plasmids (Fig. 3C). Although stabilization of the specific RPL33A mRNAs by an unknown mechanism cannot be excluded, the amounts of 25S and 60S subunits in these mutants did not increase (Table 4), suggesting that there would not be an increase in the amount of the corresponding overexpressed eL33A mutant proteins assembled into ribosomes in these mutants.

We found similar defects in pre-rRNA processing produced by Set 1 rpl33a mutations on hc plasmids (Fig. 3C). Amounts of 25S, 18S and 5.8S rRNAS were reduced to a similar extent as observed in cells harboring the corresponding lc rpl33a mutants, with a slight increase in the 25S rRNA level in the hc rpl33a-Y44R versus the lc version (Fig. 3B and C; and Table 4), which is consistent with the less accentuated Slg− phenotype of the hc rpl33a-Y44R strain (Fig. 2C). In contrast, hc rpl33a-G69R leads to a greater reduction of 25S rRNA than does lc rpl33a-G69R. This might be explained by the observation that in the lc rpl33a-G69R mutant we frequently observe the appearance of revertants or suppressors that could be masking the defects produced by the G69R mutation, or by a possible dominant-negative effect of the overexpressed G69R mutant in the hc rpl33a-G69R strain. Ratios of pre-rRNAs to mature rRNAs for Set 1 rpl33a mutations on hc plasmids are shown in Table 5. Thus, with the notable exception of rpl33a-Y44R, the pre-rRNA processing defects are very similar in cells with rpl33a mutant alleles expressed from hc or lc plasmids, suggesting that those defects do not arise merely from reduced expression of the eL33A mutant proteins when rpl33a mutant alleles are present in lc plasmids. Rather, the mutations appear to affect the function of eL33A in contributing to 60S biogenesis.

We next investigated the effects in pre-rRNA processing of Set 2 rpl33a mutations expressed from lc plasmids in strain Hm701 with a deletion of both paralogous genes RPL33A and RPL33B, either at 28 °C (Fig. 4A), or after 6 h at 37 °C (Fig. 4B). At 28 °C, only rpl33a-Y103R and A40-44 showed a modest reduction of ~ 15% and ~ 25%, respectively, in 25S rRNA levels (Fig. 4A, Table 4), which is consistent with the mild Slg− phenotype at 28 °C observed in these Set 2 mutants (Fig. 2D). The 35S/32S, 35S/27SA2, 35S/27S, and 27SA2/25S ratios were slightly higher in rpl33a-Y103R and A40-44 mutants than in WT (Table 6), with attendant reductions (10–30%) in the amounts of pre-rRNA species 32S, 27SA2, 27S and 25S mature rRNA (Fig. 4A). Thus, the Y103R and A40-44 mutants exhibit modest defects in the processing reactions that produce mature 25S rRNA at 28 °C. As shown in Fig. 4A (lowest panel) and Table 4, at 28 °C all the mutants showed RPL33A mRNA amounts similar than that of the WT, except rpl33a-A92-99 that exhibits a ~ 20% reduction, and rpl33a-G79R that shows a ~ 1.7-fold increase in the amount of RPL33A mRNA.

Table 6 Ratios between different species of pre- and mature rRNAs

At 37 °C, the RPL33A mRNA levels were similar, or higher than those of the WT in rpl33a-L7R, V35R, F43R, and G79R mutants; however, they are reduced in rpl33a-∆L29 (~ 50%), Y103R (~ 30%), A40-44 (~ 15%), A92-99 (~ 45%) and A104-107 (~ 40%) mutants (Fig. 4B and Table 4). Strong reductions in the amounts of 25S rRNA were found in rpl33a-V35R, G79R, A40-44 and A104-107 (~ 60–80%), which is consistent with the Ts− phenotypes shown by these mutants at 37 °C, whereas more moderate reductions occur in rpl33a-L7R, ∆L29, F43R, Y103R and A92-99 mutants (~ 10–55%, Fig. 4B, Table 4). Similar WT levels or small reductions in the abundance of 18S rRNA (~ 15–30%) were also observed in Set 2 rpl33a mutants at 37 °C, except for the rpl33a-G79R (~ 40%), A40-44 (~ 60%), and A104-107 (~ 45%) (Fig. 4B, Table 4). The observable amounts of the 35S pre-rRNA were similar in the mutants and in the WT, with the A40-44 exception that showed a ~ 40% decrease in 35S pre-rRNA levels. In addition, the 35S/32S, 35S/27SA2, 35S/27S, and 35S/20S ratios were elevated in these mutants at 37ºC (Table 6), with attendant reductions in the amounts of 27SA2, 27S, 20S and 7S and, consequently, of the 25S, 18S and 5.8S mature rRNAs (Fig. 4B and Table 4). Moreover, the 27SA2/25S and 27S/25S ratios are slightly increased in the Set 2 mutants (except in rpl33a-L7R and F43R) indicating specific pre-rRNA processing defects at those two steps of the maturation pathway (Table 6).

In summary, the Northern analyses in Figs. 3, 4 revealed defects indicative of impaired processing of 35S pre-rRNA at sites A0-A1-A2 in all of the analyzed Set 1 rpl33a mutants, in Set 2 rpl33a mutants Y103R and A40-44 at both 28 °C and 37 °C, and the rest of the Set 2 rpl33a mutants at 37 °C, leading to strong reductions in the levels of 33-32S, 27SA2, 27S, 7S and 20S pre-rRNAs in all of the mutants. Defects in processing of 27S pre-rRNAs and probably a pronounced destabilization of early and intermediate 60S pre-RNAs, could contribute to the reduced production of 25S, and 5.8S mature rRNAs. Together, these data support the idea that a wide range of mutations affecting key eL33A residues substantially impair the efficiency of different pre-rRNA processing steps in the ribosomal maturation pathway.

Defects in polysome assembly caused by rpl33a mutations

To investigate whether the rpl33a mutations reduced steady-state levels of 60S subunits and also confer reductions in general translation of mRNAs, we analyzed profiles of free-ribosomal subunits, 80S monosomes and polysomes from WT and Set 1 and Set 2 rpl33a mutants by sucrose gradient-velocity sedimentation. The ratio of polysomes to monosomes (P/M) was calculated to determine the effects of the mutations on the rate of bulk translation initiation.

As shown in Fig. 5A, all the Set 1 rpl33a mutants showed an appreciable decrease in the P/M ratio and reduced polysome abundance, in comparison with the WT at 28 °C, suggesting that these rpl33a mutations reduce the rate of bulk translation initiation. The pool of free 40S ribosomal subunits is elevated and the levels of free 60S subunits is severely reduced in all the mutants, concomitant with the presence of “halfmer” shoulders on the monosome (80S) and disome peaks of the polysome profiles. The appearance of halfmers, representing mRNAs with one or more 80S ribosomes plus a single 43S or 48S PIC is characteristic of a reduced level of 60S ribosomal subunits, resulting in a delay in the 60S subunit joining reaction at the AUG codon. The reduction in P/M ratio was strongest in the rpl33a- G69R, G76R-G79R and ∆99-107 mutants and comparable to that of the rpl33a∆ mutant. In contrast, rpl33a-102R and the two blocks of Ala substitutions in residues of the C-terminal region of eL33 (A102-105 and A102-107) produced only a ~ 20% reduction in the P/M ratios whereas G76R and Y44R decreased the P/M ratio by ~ 30–40%. The hc rpl33a-Y44R mutant showed an increase in both the P/M ratio (~ 40%) and the polysome content when compared with the lc rpl33a-Y44R mutant, which is in accordance with the partial suppression of the Slg− phenotype and the modest increase in the levels of 25S rRNA seen above on overexpressing this variant. The decrease in the P/M ratios of polysomes in the Set 1 rpl33a mutants correlated with the intensity of the Slg− phenotypes displayed by these mutants at 28 °C (Fig. 2C, left panels and Table 7), suggesting that this phenotype resulted, at least in part, from a reduction in the amounts of 60S subunits that generally reduces the initiation rate of protein synthesis.

Fig. 5figure 5

Substitutions in the yeast ribosomal protein eL33A result in a deficit of 60S subunits and accumulation of halfmer polysomes. A Derivatives of strain Hm700 (his4-301 rpl33a∆ RPL33B) containing different RPL33A alleles on lc or hc plasmids or an empty vector as indicated (Set 1 mutants), were cultured in liquid SC-Leu at 28 °C to A600 of ~ 1.0. Cycloheximide was added at 100 µg/ ml before harvesting the cells and WCE were prepared in the presence of 30 mM Mg2+. 10 A260 of each extract were resolved on 7–50% sucrose gradients and analyzed by continuous monitoring of A254. In the left top panel, peaks representing free-ribosomal 40S and 60S subunits, 80S monosomes and polysomes are indicated. Mean Polysome/Monosome (P/M) ratios were calculated from two or three independent experiments. B-C WCEs of derivatives of strain Hm701(his4-301 rpl33a∆ rpl33b∆ containing the indicated RPL33A alleles (Set 2 mutants) were obtained and analyzed as in Fig. 5A from cells cultured in SC-Leu at 28 °C (B) or shifted to 37ºC for 6 h (C)

Table 7 Comparison between the Slg− phenotype and P/M ratio in different Set 1 rpl33a mutants

We also analyzed total polysome profiles from WT and selected Set 2 rpl33a mutants. At 28 °C, we observed in the rpl33a-A40-44 mutant an increase in free 40S ribosomal subunits, a decrease in free 60S subunits, 80S monosomes and polysomes, and the presence of halfmers (Fig. 5B), all consistent with the ~ 25% reduction in the levels of 25S rRNA revealed by this mutant at the permissive temperature (Fig. 4A and Table 4). The P/M ratio of A40-44 was very similar to that of the WT; however, the proportion of > 4-mer polysomes was somewhat reduced (~ 50% of WT), suggesting a modest reduction in the translation initiation rate at the permissive temperature. After 6 h of incubation at 37 °C, we observed accumulation of free 40S subunits accompanied by a marked reduction in the amount of free 60S subunits and the appearance of halfmers in all five mutants. Notably, the levels of free 40S subunits and those of 80S monosomes are quite similar in the rpl33a-V35R, Y103R, and A40-44 mutants (Fig. 5C). The P/M ratios are reduced in all the mutants compared with the WT, as is the abundance of 80S ribosomes and polysomes. The rpl33a-A40-44 mutant showed a P/M decrease of ~ 50%, the largest reduction in P/M ratio among the Set 2 mutants analyzed, whereas rpl33a-∆L29 displays only a slight reduction of ~ 15% in the P/M ratio. Interestingly, the P/M ratio is lower in the rpl33a-L7R versus V35R and Y103R mutants, although the defects in ribosome biogenesis were more severe in the latter mutants, thus indicating that L7R elicits a relatively stronger defect in translation initiation beyond its deleterious effect on ribosome assembly.

Set 1 rpl33a mutants exhibit Gcd− phenotypes

As shown in Fig. 2A, B, the Set 1 rpl33a mutations conferred 3ATR phenotypes in a gcn2 gcn3 rpl33aΔ strain, and increased HIS4-lacZ expression under amino acid sufficiency conditions (Gcd− phenotypes). To obtain direct evidence that the Set 1 rpl33a mutations cause derepression of GCN4 at the translational level, we measured β-galactosidase activities from the GCN4-lacZ reporter on plasmid p180, which contains the WT GCN4 mRNA leader with the four regulatory uORFs (Fig. 6A), in cells grown at 28ºC. To evaluate the effects of the mutations on translational control by the uORFs, the increases in expression from p180 were normalized with the values obtained from measuring the expression of the control GCN4-lacZ reporter on plasmid p227 lacking the four uORFs (Fig. 6B). Modest but significant increases of 1.5- 2.4-fold in normalized GCN4-lacZ expression were observed for all the rpl33a mutants compared with the WT, indicating that all the mutants diminished translational repression by the uORFs. The rpl33a-L102R, A102-105 and A102-107 mutants exhibited greater normalized derepression of WT GCN4-lacZ expression than did the other Set 1 rpl33a mutants.

Fig. 6figure 6

Comparison of the Gcd− and Gcn− phenotypes conferred by mutations in rpl33a. A Schematic depiction of GCN4-lacZ reporters containing the four uORFs in the leader of GCN4 (p180), only uORF4 (p226), an elongated version of uORF1 overlapping with the beginning of GCN4 (pM226) and GCN4 without uORFs (p227). B Strain Hm700 (his4-301 rpl33a∆ RPL33B), marked as rpl33a∆*, derivatives of strain Hm700 containing the indicated RPL33A alleles on lc plasmids (upper panel) or on hc plasmids, or an empty vector (middle panel), and derivatives of strain Hm701(his4-301 rpl33a∆ rpl33b∆ containing the indicated RPL33A alleles (lower panel), were transformed with GCN4-lacZ fusions on plasmids p180, p226, pM226 and p227 and cultured in liquid SD + His + Trp, or for 6 h in SD + His + Trp containing 0.5 μg/ml sulfometuron (SM) at 28 °C to A600 of ~ 1.0 and assayed for β-galactosidase activities as in Fig. 2A. The means and SE from at least four independent transformants are reported. The values highlighted in bold in the right columns are the results normalized to correct for the different expression of p227 and relative to the corresponding WT value (1.0)

It has been previously described that “leaky scanning of uORF4” by fully assembled PICs elicits the Gcd− phenotypes of the rpl16b∆ (Foiani et al. 1991), rpl33a-G76R and rpl33a∆ mutations (Martin-Marcos et al. 2007). Thus, we investigated whether other rpl33a mutations elevate GCN4-lacZ expression from the reporter in p226 containing uORF4 alone (Fig. 6A) (Mueller and Hinnebusch 1986). As shown in Fig. 6B, rpl33a mutants produced increases of two- to threefold in GCN4-lacZ expression from plasmid p226 compared to that seen in WT cells, after normalizing with the values of the expression of GCN4-lacZ from plasmid p227. As this effect could also arise from increased reinitiation after termination at uORF4, we measured GCN4-lacZ expression from the pM226 construct containing a single elongated version of uORF1, which overlaps the beginning of GCN4 and is incompatible with reinitiation downstream (Fig. 6A) (Grant et al. 1994). Thus, only mutations that cause leaky scanning of uORF1 would increase GCN4-lacZ expression from pM226. All Set 1 rpl33a mutants showed a 2–4-fold increase in normalized GCN4-lacZ expression compared to WT (Fig. 

留言 (0)

沒有登入
gif