Repression of MRP51 in cis does not contribute to the synthetic growth defect conferred by an hphMX4-marked deletion of DBP1 in a ded1-ts mutant

Abstract

Powers et al. recently demonstrated that the hphMX6 cassette used to delete DPB1 in dbp1Δ::hphMX6 yeast mutants leads to reduced expression in cis of the adjacent gene MRP51, encoding the mitochondrial small subunit (SSU) ribosomal protein Mrp51. Here we provide evidence that elimination of Dbp1, not reduced MRP51 expression, underlies the synthetic growth defect of a dbp1Δ::hphMX6 ded1-ts mutant on glucose-containing medium, where respiration is dispensable, consistent with our previous conclusion that Dbp1 and Ded1 perform overlapping functions in stimulating translation initiation on mRNAs burdened with long or structured 5’UTRs in cells cultured with glucose.

INTRODUCTION

Structures in mRNA that impede ribosome binding or subsequent scanning of the 5’untranslated region (5’UTR) to locate the AUG initiation codon reduce translation efficiency. An abundance of evidence indicates that yeast DEAD-box RNA helicase Ded1 promotes translation initiation by enhancing ribosomal scanning of long or structure-laden 5’UTRs (1–6). Previously, we published ribosome profiling data suggesting that Dbp1, a Ded1 paralog, performs similar functions in yeast cells, finding that many more mRNAs exhibit reduced relative translational efficiencies (TEs) in a ded1-ts dbp1Δ double mutant than in either single mutant, with numerous transcripts exhibiting strong dependence on Dbp1 or Ded1 for efficient translation only when the other helicase is impaired in the double mutant. Such “conditionally hyperdependent” mRNAs contain unusually long 5′UTRs with heightened propensities for secondary structure formation—the same characteristics found previously for mRNAs translationally impaired in the ded1-ts single mutant (2). Consistently, overexpressing Dbp1 in a ded1-cs mutant suppressed the growth defect and improved the TEs of many Ded1-hyperdependent mRNAs in this single mutant. Moreover, we found that Dbp1 associates with translating mRNAs in cell extracts and that purified Dbp1 resembled Ded1 in preferentially accelerating 48S PIC assembly on selected mRNAs harboring structured 5’UTRs in a fully reconstituted yeast translation initiation system. Finally, profiling of 40S initiation complexes in the ded1 and dbp1 mutants provided evidence that Ded1 and Dbp1 stimulate PIC attachment or scanning of 5’UTRs on many Ded1/Dbp1 hyperdependent mRNAs in vivo (6).

Subsequently, Powers et al. reported (7) that the hphMX6 cassette we had utilized in deleting DBP1 in dbp1Δ::hphMX6 strains in the aforementioned study (6) leads to reduced expression in cis of the adjacent gene MRP51, encoding the mitochondrial small subunit (SSU) ribosomal protein Mrp51. This defect was shown to be associated with impaired mitochondrial SSU biogenesis, reduced bulk cytoplasmic translation, and diminished cell growth on media containing non-fermentable carbon sources in which respiration is required for energy generation. These phenotypes could be suppressed by introduction of MRP51, but not DBP1, on a plasmid. The inhibition of MRP51 expression appears to involve a transcript emanating from the strong TEF1 promoter of the MX6 cassette that overlaps the native MRP51 promoter and, although encompassing the entire MRP51 ORF, is likely impaired for Mrp51 production owing to alternative translation initiation events at upstream open-reading-frames in the extended transcript. Presumably, production of this extended transcript interferes with synthesis of the native MRP51 mRNA in the manner described previously for long-untranslated-transcript-isoforms (LUTIs) in WT yeast cells (8,9). A similar reduction in MRP51 expression was observed by Powers et al. (7) on interrogating our published ribosome profiling and RNA-Seq data for the dbp1Δ::hphMX4 mutant grown on glucose-containing medium, resulting in a ~4-fold decrease in translation of MRP51 mRNA, from which we had concluded erroneously that Dbp1 enhances the translational efficiency (TE) of MRP51 mRNA (6).

All of the experiments in our previous analysis of dbp1Δ::hphMX6 mutants (6) were conducted on cells growing in medium containing glucose as carbon source, where Mrp51 and respiration in mitochondria are not essential for energy generation. Indeed, we showed that the dbp1Δ::hphMX6 single mutant is indistinguishable from WT both in cell growth rate and bulk polysome assembly (6), indicating that any reduction in mitochondrial translation that results from reduced MRP51 expression has no discernible impact on bulk cytoplasmic translation in glucose-grown cells. Similar to our findings, De la Cruz et al. analyzed a dbp1Δ:TRP1 allele and found a synthetic slow growth phenotype (Slg⁻) in combination with a ded1/spp81–3 mutation on glucose-containing medium, despite the absence of any Slg⁻ for the dbp1Δ::TRP1 mutation; and no synthetic interaction was observed on combining dbp1Δ::TRP1 with mutations impairing other initiation factors, including tif1–1 (eIF4A), cdc33–1 and cdc33–42 (eIF4E), Δstm1 (eIF4B), or Δtif4631 (eIF4G1) (10). We consider it unlikely that a reduction in MRP51 expression conferred by dbp1Δ::hphMX4 or dbp1Δ:TRP1 would produce such a highly specific exacerbation of the phenotypes of ded1 mutations in glucose-grown cells, especially considering that the TRP1 promoter is >10 times weaker than the S. cerevisiae TEF1 promoter on glucose media (11).

Nevertheless, we could not eliminate the possibility that a reduction in MRP51 expression was contributing to the synthetic reductions in growth conferred by the dbp1Δ::hphMX4 mutation in the dbp1Δ::hphMX ded1-ts double mutant—a key observation indicating functional cooperation between Dbp1 and Ded1 in stimulating translation (6). We had shown that the synthetic Slg⁻ phenotype of the ded1-ts dbp1Δ::hphMX4 double mutant was fully complemented by DBP1 on a high-copy plasmid (lacking MRP51); however, this is an imperfect experiment because DBP1 overexpression can suppress the growth defect of ded1 mutations (as noted above) in addition to complementing the dbp1Δ deletion. As such, it was still possible that reduced MRP51 expression conferred by dbp1Δ::hphMX4, rather than loss of DBP1 function, was responsible for exacerbating the growth defect conferred by the ded1-ts mutation in the ded1-ts dbp1Δ::hphMX4 double mutant on glucose medium. In an effort to rectify this shortcoming, we tested whether a low-copy plasmid containing DBP1 but lacking MRP51 could likewise complement the synthetic Slg⁻ phenotype of the dbp1Δ::hphMX4 ded1-ts double mutant, whereas a plasmid harboring MRP51 alone could not.

RESULTS AND DISCUSSION

As expected, the low-copy DED1 plasmid fully complemented the strong Slg⁻ of the dbp1Δ::hphMX4 ded1-ts double mutant examined in our previous study (6) and conferred WT growth at 34°C on medium with glucose as carbon source (Figure 1A, row 5 vs. rows 4, 1, & 3). These results are consistent with our previous finding that deletion of DBP1 alone by dbp1Δ::hphMX4 does not appreciably impair cell growth or bulk translation on glucose-containing medium (6). Importantly, introducing the low-copy DBP1 plasmid into the double mutant increased growth at 34°C compared to the empty vector control, but left a residual growth defect intact in comparison to the WT strain, comparable to that exhibited by the ded1-ts single mutant (row 6 vs. rows 1–2). In contrast, introducing low-copy MRP51 afforded no growth improvement at 34°C and was indistinguishable from empty vector (row 7 vs. row 4). These results indicate that the stronger growth defect of the dbp1Δ::hphMX4 ded1-ts double mutant compared to the ded1-ts single mutant results from loss of DBP1, not MRP51, function.

Figure 1. Low-copy DBP1 but not MRP51 complements the slow-growth phenotype conferred by dbp1Δ::hphMX4 in the ded1-ts dbp1Δ::hphMX4 double mutant on glucose-containing medium.

(A) Serial dilutions of the following yeast strains were spotted on synthetic complete (SC) medium without uracil containing 2% glucose as carbon source and incubated at 30°C or 34°C: WT strain BY4741 (MATa his3Δ1 leu2Δ0 met15-Δ0 ura3-Δ0) transformed with low-copy URA3 CEN6 vector pRS416 (12) (row 1), strain Y10029 (aka F2030) (MATa his3Δ1 leu2Δ0 met15-Δ0 ura3-Δ0 ded1–952::kanMX4 (13)) transformed with pRS416 or low-copy URA3 plasmid CP3482 containing DED1 (14) (rows 2–3), and strain H5331 (aka NSY80) (MATa his3Δ1 leu2Δ0 met15-Δ0 ura3-Δ0 ded1–952::kanMX4 dbp1Δ::hphMX4 pRS413 [HIS3]) transformed with pRS416, CP3482, low-copy URA3 plasmid pNDS42 (aka p6251) containing DBP1 (6), or low-copy URA3 plasmid pRS416-MRP51 (aka p6749) containing MRP51 (rows 4–7). Strain H5331 is a transformant of NSY15 (6) containing low-copy HIS3 vector pRS413. (B) Serial dilutions of the following strains were spotted on synthetic complete medium lacking uracil and containing 2% glycerol/2% ethanol in place of glucose as carbon source and incubated at 30°C: WT strain BY4741 transformed with pRS416 (row 1), strain H5314 (aka NSY14) (MATa his3Δ1 leu2Δ0 met15-Δ0 ura3-Δ0 dbp1Δ::hphMX4 (6) transformed with pRS416, pNDS42, or p6749 (rows 2–4), and H5331 transformed with pRS416, pNDS42, or p6749 (rows 5–7). Plasmid p6749 was constructed by synthesis of a Bam HI-Hind III DNA fragment containing the 1035 bp of the MRP51 ORF, 360 bp 5’ of the ORF, and 200 bp 3’ of the ORF (carried out by LifeSct LLC) and inserted between the Bam HI and Hind III sites of pRS416. The cloned synthesized fragment was sequenced in its entirety to verify the WT sequence.

In sharp contrast to these last results, on medium containing the non-fermentable carbon sources glycerol/ethanol in place of glucose, introducing the low-copy MRP51 plasmid fully suppressed the Slg⁻ phenotypes displayed by both the dbp1Δ::hphMX4 ded1-ts double mutant and dbp1Δ::hphMX4 single mutant (Figure 1B, row 4 vs. rows 1–2 & row 7 vs 1 and 5). These last findings confirm the conclusion of Powers et al. that the dbp1Δ::hphMX4 mutation acts in cis to reduce expression of chromosomal MRP51, reducing cell growth on non-fermentable carbon sources where mitochondrial translation is required for respiration. The ded1-ts dbp1Δ::hphMX4 double mutant transformed with low-copy MRP51 shows no Slg⁻ phenotype conferred by ded1-ts (Figure 1B, row 7 vs. 1) most likely because the ded1-ts product is functional at 30°C (2).

These genetic complementation results demonstrate that the reduced cell growth conferred by dbp1Δ::hphMX4 in combination with ded1-ts on glucose-containing medium results primarily from loss of DBP1 function rather than reduced MRP51 expression. Thus, although expression of Dbp1 is low in comparison to Ded1, it makes an appreciable contribution to cell growth when Ded1 function is impaired in cells fermenting glucose. In fact, our ribosome profiling data indicated that DBP1 expression is up-regulated 2- to 3-fold in ded1-ts cells. As such, we consider it likely that the effects of dbp1Δ::hphMX4 in exacerbating the deleterious effects of the ded1-ts mutation on translational efficiencies of many Ded1-hyperdependent mRNAs that we observed on glucose medium reflect loss of Dbp1’s ability to compensate for reduced Ded1 function (6). It remains to be determined whether reduced expression of MRP51 contributed to the reduced translational efficiencies of any particular mRNAs in dbp1Δ::hphMX4 or ded1-ts dbp1Δ::hphMX4 cells.