Beyond the ORF: Paralog-specific regulation of RPS7/eS7 mRNAs via 3’-UTRs and promoter sequences
Sachiko Hayashi, Tohru Yoshihisa, Koppolu Raja Rajesh Kumar, Koppolu Raja Rajesh Kumar, Koppolu Raja Rajesh Kumar, Koppolu Raja Rajesh Kumar
TL;DR
This study shows that two similar yeast genes, RPS7A and RPS7B, are regulated differently, affecting how much of their protein is made.
Contribution
The paper reveals that RPS7 paralog mRNAs are regulated differently via their 3’-UTRs and promoter sequences.
Findings
Deleting RPS7A increases RPS7B mRNA expression but not vice versa.
The 3’-UTR sequences are critical for the stability of RPS7A and RPS7B mRNAs.
Fhl1p is essential for RPS7B transcription but not RPS7A.
Abstract
In a classical view, each paralogous ribosomal protein (RP) is equally synthesized and integrated into the ribosome. Therefore, RP-paralog mRNAs are generally believed to be similarly regulated on their transcription and/or stability. In this paper, we report that two Rps7p/eS7 paralogs of Saccharomyces cerevisiae are differently regulated; deletion of RPS7A upregulates RPS7B paralogous mRNA expression but not vice versa. Their 3’-UTR sequences critically regulated the stabilities of both RPS7A and RPS7B mRNAs. Alterations in these sequences led to a diminished expression of RPS7A and RPS7B mRNAs in a transcript-dependent manner, suggesting that RPS7-paralog mRNAs have different properties for their expression and/or stability. The C-terminal tagging of the ORF and mutation analyses in the 3’-UTR indicate that both RPS7-paralog mRNAs critically rely on their 3’-UTRs for mRNA expressions…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Fig 1
Fig 2
Fig 3
Fig 4
Fig 5- —JSPS KAKENHI
- —JSPS KAKENHI
- —JSPS KAKENHI
- —http://dx.doi.org/10.13039/100007449Takeda Science Foundation
- —http://dx.doi.org/10.13039/100021128Hyogo Science and Technology Association
- —http://dx.doi.org/10.13039/100007802Institute for Fermentation, Osaka
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsRNA and protein synthesis mechanisms · RNA Research and Splicing · RNA modifications and cancer
Introduction
Ribosomes are central molecular machines dedicated to translation in all of life. In Saccharomyces cerevisiae, 79 ribosomal proteins (RPs) and four ribosomal RNA (rRNA) are assembled into the ribosome with the assistance of 76 different small nucleolar RNAs and more than 200 assembly factors [1,2]. Fifty nine of the 79 RPs are encoded by a pair of paralogous genes named A and B, and these RP paralogs are considered to be generated mainly by whole genome duplication [3]. Indeed, 21 of the 59 RP-paralog pairs encode identical proteins while the others show high sequence similarities [4]. These paralog pairs are thought to serve as a repository for buffering deleterious changes in either copy [5], e.g., to permit adjustment of the RP dose to match that of rRNA synthesis, thereby ensuring ribosome assembly. Emerging evidence showed that RP-paralog pairs possess distinct or even opposite expression patterns and paralog-specific functions [6]. Thus, heterogeneous RP-paralog pairs not only maintain a proper ribosomal level by complementing each other’s expression as dose amplification but also offer the potential to generate heterogeneous ribosomes by swapping distinct RP isoforms to fine-tune the translational program known as paralog-specific regulation [6,7]. Further, deletion of RP genes can lead to aneuploidy in yeast as a compensatory response to the loss of the RP copies [4,8]. Cells prefer to duplicate chromosomes containing the remaining RP copies to restore balance, enhancing their fitness despite the aneuploidy.
To date, yeast RP paralog-specific differences were observed on a wide range of cellular and metabolic functions: sensitivity/resistance against various types of drugs, localization and translation of ASH1 mRNA [9] as well as translational accuracy of the luciferase-based reporter mRNAs [10]. Quantitative mass spectrometry analysis revealed a compositional change of a Rpl8p paralog pair in the 80S ribosome in response to a shift from glucose to glycerol media [11]. RPL7 paralogous genes generate two types of acetylated RPs, Rpl7Ap and Rpl7Bp, leading to modulated mRNA translation depending on the ORF length and achieving different drug resistance [12]. Although reducing the availability of duplicated RP genes does not impair overall ribosome function or viability, it still can lead to distinct phenotypes [4,13,14]. In addition, these duplicated RPs are differently expressed, and such differential expression under normal conditions is mainly achieved not only by transcriptional regulation but also by post-transcriptional events, such as splicing, mRNA stabilization and translation control [3,7]. To comprehensively understand the intricate regulatory mechanisms governing these paralog-specific disparities in yeast RP genes and their impacts on cellular processes and phenotypes, further investigations and detailed analyses are imperative.
Here, we revealed paralog-specific regulations of RPS7 mRNAs, encoding a eukaryotic ribosomal protein Rps7p/eS7. Originally, the tRNA-intronless tl(caa)Δint* mutant (see Results in detail) that found to have a long insertion in the RPS7B 3’-UTR that destabilized RPS7B mRNA by nonsense-mediated decay (NMD). This mutation, akin to the rps7bΔ, did not induce RPS7A expression, while the rps7aΔ mutation upregulated RPS7B mRNA to raise Rps7Bp production. Modified 3’-UTRs in both cases of *RPS7-*paralog mRNAs impaired the corresponding mRNA level in a transcript-dependent manner. Additionally, reporter gene assay using the promoter regions of RPS7A and RPS7B revealed different contributions of Fhl1p, a prominent transcription factors almost exclusively associated with ribosomal protein genes in yeast, to their promoter activities. This suggests that transcription also forms an overlooked layer for paralog-specific regulation. Overall, yeast has a different buffering system for Rps7p production between the two RPS7-paralog mRNAs that are sensitive to a variation on their 3’-UTRs.
Materials and methods
Strains, primers, plasmids, and media
S. cerevisiae BY418-based strains are summarized in S1 Table in S1 File. Primers and plasmids used in this study are listed in S2 and S3 Tables in S1 File, respectively. The plasmids utilized in the reporter gene assay were created using the pSA144 plasmid (gift from Prof. Toshifumi Inada, The University of Tokyo, Japan) as the backbone. Details of plasmid construction are also described in S3 Table in S1 File. Standard yeast genetic techniques and other molecular biological techniques used here are essentially described in [15,16]. Gene disruptions and marker insertions were carried out using a PCR-based strategy; a marker cassette with appropriate flanking sequences of a target gene was amplified by PCR, and the target gene on the chromosome was replaced with the PCR fragment by homologous recombination. A C-terminally HA-tagged RPS7A fragment was similarly made using PCR with a 3 × HA tagging cassette containing an ADH1 terminator and a CgHIS3 marker, and were integrated into the indicated parental strains, leading to RPS7A-HA::CgHIS3 strains. The RPS7B-FLAG strains were generated by the two-step gene replacement strategy; the chromosomal RPS7B gene was first disrupted by a URA3 marker, and subsequently the disrupted RPS7B gene region was replaced with an amplified RPS7B-3 × FLAG fragment. Final transformants expecting ura3 RPS7B-3 × FLAG strains were selected by 5-fluoroorotic acid plates, and correct integration was confirmed by sequencing. For introducing the mutated RPS7B 3’-UTR appearing in TYSC2148 (rps7b-102 mutation) into a tL(CAA)N wild-type strain, the corresponding DNA fragment with the RPS7B second exon was amplified from TYSC2148 chromosomal DNA using RPS7B_int_242_fw and eS7B_3UTR_158–227 primers, and then was integrated into SHSC0322 (tL(CAA)N rps7bΔexon::URA3) to replace the corresponding region including the URA3 marker. Correct integration was confirmed by sequencing. For constructing SHSC0633 with fhl1∆::CgHIS3, gene disruptions and marker insertions were conducted using the PCR-based strategy described above.
All strains were essentially grown at 30˚C in YPD [1.0%(w/v) yeast extract, 2.0%(w/v) polypeptone, and 2.0%(w/v) D-glucose]. In case of solid media, final 2.0%(w/v) agar was added.
Crude RNA preparation, northern blotting, and RT-PCR
RNA preparation and northern blotting were performed as essentially described in [16]. Total RNAs from logarithmically growing yeasts were extracted by the hot phenol method with the Na-acetate/SDS buffer [50 mM Na-acetate, pH 5.2, 10 mM EDTA, 1.0%(w/v) SDS] and Phenol, Saturated with Citrate Buffer [pH 4.5] (Nacalai Tesque, Kyoto, Japan). RNA (5.0 µg) per lane was analyzed on 1.2% agarose gel with formaldehyde, and transferred to Hybond N^+^ charged nylon membrane (GE Healthcare, Chicago, Illinois, USA) by capillary transfer. Antisense RNA probes of RPS7A and RPS7B were labeled with digoxigenin using DIG Northern Starter Kit (Roche Diagnostics, Basel, Switzerland). After hybridization in DIG Easy Hyb (Roche Diagnostics) at 68˚C followed by immunodecoration with ALP-labeled anti-digoxigenin IgG, signals were visualized by CDP-star (Roche Diagnostics). Chemiluminescence was captured and analyzed by a cooled CCD camera system EZ-Capture and imaging analysis software CS Analyzer 3 (ATTO, Tokyo, Japan). For RT-PCR, poly-A RNAs in total yeast RNA samples prepared from appropriate strains were converted into cDNAs by reverse-transcription using SuperScript III reverse transcriptase (Thermo Fisher Scientific, Waltham, Massachusetts, USA) with oligo dT_20_ as a primer. Using the cDNA, 3’-UTR regions of RPS7B alleles etc. were amplified by PCR with appropriate primer sets and Ex-Taq DNA polymerase (Takara Bio, Otsu, Japan). Primers used in RT-PCR are listed in Table S4 in S1 File (Supporting Information).
Total protein extraction and western blotting
Crude proteins from mid log-phase yeast cells (0.5 OD_660_ unit) were prepared by alkaline extraction [17], and separated with 10% SDS-polyacrylamide gel followed by transfer to a PVDF membrane (Immobilon P; Millipore, Burlington, Massachusetts, USA). The primary antibodies used here were anti-Srp1p rabbit polyclonal antibodies, anti-HA tag mouse monoclonal antibody (TANA2: MBL, Tokyo, Japan), and anti-DYKDDDDK tag mouse monoclonal antibody (1E6: FUJIFILM Wako Pure Chemical, Osaka, Japan). After incubation with horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse IgG, immunoblot signals were developed via home-made ECL system, and chemiluminescence signals were captured and analyzed as northern blotting.
Results
The tl(caa)Δint* strain operates NMD for RPS7B mRNA due to its aberrant longer 3’-UTR, whereas deletion of RPS7A results in the upregulation of RPS7B
We had previously generated the complete set of S. cerevisiae tRNA-intron deletion strains [18] using a pop-in-pop-out strategy with a GALp-GIN11M86 marker [19]. The integration-marker cassette, after replacing an intron-containing tRNA gene on a chromosome with an intron-less one, was removed by homologous recombination between the hisG sequences, or between flanking sequences that were used for chromosomal integration of the tRNA gene using negative selection of the galactose-inducible toxic marker (Fig 1A). One of these tRNA intron-deletion strains, tl(caa)Δint (TYSC2148) that converted all the tL(CAA) genes in the parental genome to the intron-less ones, was found to have a 2.3-kb insertion of loxP::CmR::hisG::loxP located between the tL(CAA)N/YNCN0007W coding region and the RPS7B/eS7B ORF, while the reported genome-wide transcriptome data indicated that the marker was inserted into the 3’-UTR of RPS7B [20,21]. Hereafter, this tl(caa)Δint strain with the marker insertion is referred to as tl(caa)Δint*.
The tl(caa)Δint has a 2.3-kb insertion in 3’-UTR of RPS7B mRNA and its insertion activates NMD for RPS7B mRNA.(A) A 10-kb chromosome region harboring tL(CAA)N and RPS7B in the wild-type S. cerevisiae genome is shown schematically (upper panel). A magnified view of the chromosome region of tL(CAA)N and RPS7B in the tl(caa)Δint* strain (TYSC2148) is also shown (lower panel). The tl(caa)Δint* has the 2.3-kb sequence insertion containing loxP::hisG::CmR::loxP between tL(CAA)N and RPS7B genes: rps7B-102 allele. Black box, tL(CAA)N; blue arrow with a thick striped-line insertion, RPS7B (the striped line represents intron); gray box, yeast ORFs other than RPS7B; beige box, marker genes. A black or white triangle in a box indicates the direction of transcription. In the lower panel, the 3’-UTR of RPS7B is highlighted in light blue. Gene names of tL(CAA)N and RPS7B genes are in italicized bold face in red and black, respectively. (B) Northern analyses of RPS7A and RPS7B mRNAs. Each lane received an RNA sample (5 µg) prepared from the wild-type strain (WT) or the indicated-mutant strains grown at 30°C in YPD. For RPS7B mRNA, light-contrast (upper) and dark-contrast (lower) images of the same blot are displayed. Black triangles indicate WT RPS7A or RPA7B mRNAs. A white triangle indicates a longer transcript of RPS7B in the tl(caa)Δint* strain. (C) Total RNAs in an agarose gel prior to northern analysis were visualized by GelRed staining.*
To examine effects of unintended marker insertion in the 3’-UTR of RPS7B, we performed northern analyses of RPS7B mRNA. The RPS7B band was clearly detected with RPS7B-specific probe in the wild-type lane (black triangle), whereas it was barely seen in the tl(caa)Δint* lane (Fig 1B, lanes 1 and 2). Total RNAs visualized by GelRed staining prior to northern analysis were similar between the two strains (Fig 1C). Thus, the insertion compromised expression of RPS7B mRNA selectively.
In certain eukaryotes, such as S. cerevisiae, mRNAs with aberrantly long 3’-UTRs, including artificially elongated variants, are main targets of NMD via the faux 3’-UTR mode [22,23]. To analyze whether this expression defect of RPS7B mRNA in the tl(caa)Δint* was caused by NMD, we deleted the UPF1 gene, which encodes one of the core NMD factors [24,25]. As shown in Fig 1B longer transcript that was detected with the RPS7B-specific probe (white triangle) appeared in the tl(caa)Δint upf1Δ strain, but not in the upf1∆, the wild type nor the tl(caa)Δint (Fig 1B, middle and bottom panels, lanes 1–4). The marker insertion of the tl(caa)Δint* would elongate the 3’-UTR of RPS7B and render the RPS7B mRNA variant susceptible to NMD. We denoted to the rps7B mutant allele in the tl(caa)∆int* strain, characterized by the loxP::CmR::hisG::loxP insertion in the RPS7B 3’-UTR as rps7b-102. Based on the results obtained from RT-PCR experiments using various reverse primers (as depicted in S1 Fig), the 3’-UTR length of the mRNA transcribed from the rps7b-102 variant in the tl(caa)∆int* is estimated to be approximately 1820–2150 nt. This length is 7–13 times longer than that of the wild-type RPS7B, which typically ranges from 170–251 nt [20,21], as consistent with the results of northern blotting (Fig 1B, white triangle). Given that yeast mRNAs with 3’-UTRs of 350 nt or longer are preferentially degraded by NMD [22,26], it was reasonable that NMD targets the longer transcript of RPS7B in the tl(caa)Δint* mutant.
Some paralogous gene pairs encoding a ribosomal protein show interdependent expression [3]. We speculated that reducing RPS7B mRNA in the tl(caa)Δint* may alter RPS7A expression; therefore, we conducted northern analyses of paralogs of RPS7A and RPS7B mRNAs (Fig 1B). RPS7A mRNA was consistently expressed in the wild-type, tl(caa)Δint*, and any upf1Δ background strains (Fig 1B, top panel, lanes 1–4): the signal ratio of RPS7A between the tl(caa)Δint* and the wild-type yeasts was 1.07 ± 0.36. We also tested effect of rps7bΔ and found that the deletion affected RPS7A expression marginally (rps7b∆/WT, 0.69 ± 0.13; by Student’s t-test, *p *= 0.0066; n = 3). However, in the case of RPS7B mRNA, rps7aΔ upregulated RPS7B mRNA by nearly 2-fold (rps7a∆/WT, 1.87 ± 0.53; Student t-test, *p *= 0.024; n = 3; Fig 1B, middle and bottom panels, lanes 5 and 6). The expression of RPS7A appears to dominantly control the Rps7-paralogous expression at the mRNA level for buffering deleterious reduction of itself, whereas that of RPS7B has only a marginal role in the paralogous expression.
The addition of a C-terminal HA-tag to RPS7A gene, along with the ADH1 3’-UTR and terminator, decreased the levels of RPS7A mRNA and compromised cell viability, especially when RPS7B was defective
Paralogous RP gene expression is regulated by multiple layers of transcriptional and post-transcriptional levels [3]. In trypanosomatids, recent findings show that the translation of RP genes can be regulated by RNA-binding proteins interacting with their 5’- or 3’-UTRs [27], contributing to paralog-specific expression [28]. Similarly, in yeast, UTRs may serve as regulatory elements for RP paralog expression, potentially playing a more significant role than previously anticipated. To further explore expression difference between RPS7 paralogs, especially those with alteration in their 3′-UTRs, we modified RPS7A mRNA by introducing a C-terminal HA-tag and replacing its 3’-UTR with the ADH1 3’-UTR and terminator followed by a CgHIS3 marker sequence (RPS7A-HA::CgHIS3). Although this alternation of RPS7A did not inhibit cell growth in the wild-type cell, the introduction of RPS7A-HA::CgHIS3 in the tl(caa)Δint* background diminished growth (Fig 2B, WT RPS7A-HA::CgHIS3 and tl(caa)Δint RPS7A-HA::CgHIS3*). This growth phenotype that appeared in the tl(caa)Δint RPS7A-HA::CgHIS3* was alleviated by an additional UPF1 deletion (Fig 2B, tl(caa)Δint upf1*Δ RPS7A-HA::CgHIS3). The relative levels of RPS7A-HA::CgHIS3 mRNA, which is longer than wild-type RPS7A, were approximately 70–80% lower than those of wild-type RPS7A across all backgrounds (Fig 2C). Regarding protein levels, all RPS7A-HA::CgHIS3 strains produced almost equal amounts of Rps7Ap-HA (Fig 2D; one-way ANOVA, *p = 0.66). Considering the diminished mRNA expression of both RPS7 paralogs in the RPS7A-HA::CgHIS3 strains, it appears that tl(caa)Δint RPS7A-HA::CgHIS3 yields insufficient levels of total Rps7p (Fig 2B–2D). Of note, because upf1Δ did not rescue the phenotype of RPS7A-HA::CgHIS3 (Fig 2C and 2D), NMD was not the reason of decreased mRNA expression from the modified RPS7A gene. Collectively, the 3’-UTR of RPS7A seems to have some cis element(s) to enhance the basal transcription and/or stability of RPS7A mRNA.
The C-terminal HA-tagging of the RPS7A gene with the ADH1 terminator reduces RPS7A mRNA expression in an NMD-independent manner.(A) The panel illustrates the WT RPS7A and the RPS7A-HA::CgHIS3 genes. Stop codons are marked in magenta. A number in parentheses represents nucleotide length of the corresponding part (bp). (B) Growth comparison on a YPD plate at 30°C. The upper set of wild type (WT) and tl(caa)Δint strains has the wild-type RPS7A gene. The lower set of the WT and tl(caa)Δint* strains has the C-terminally HA-tagged RPS7A gene with the ADH1 terminator and CgHIS3 marker (RPS7A-HA::CgHIS3) in the presence or absence of the UPF1 gene. Saturated cultures of the indicated strains were serially diluted by 10-fold as shown in the bottom and dropped onto the plate. (C) Northern analyses of RPS7A and RPS7B mRNAs in the strains used in (B). Cells were grown at 30°C in YPD. Black triangles indicate WT RPS7A or RPS7B mRNAs. Grey and white triangles indicate RPS7A-HA and the longer transcript of rps7b in tl(caa)Δint*, respectively. For RPS7B mRNA, light-contrast (upper) and dark-contrast (lower) images of the same blot are displayed. The bottom panel represents total RNAs visualized by GelRed staining prior to northern analysis. (D) Western blot analysis of the RPS7A-HA::CgHIS3 strains used in (B) and (C). Equal amounts of total protein were extracted from the cells, and these extracts were subjected to immunoblotting analysis. An antibody specifically designed to detect the HA tag was used, with anti-Srp1p antibodies serving as a control for loading. Numbers represent relative abundance (RA) of HA-tagged Rps7Ap that normalized by Srp1p in n ≥ 3. The mean value of the WT was set to 1.00. ns: not significant by one-way ANOVA, p > 0.05.*
The 3’-UTR of RPS7B mRNA, derived from the tl(caa)Δint*, but not the CgHIS3 sequence, triggers NMD for RPS7B in the wild-type strain
The presence or absence of an intron in a tRNA gene influences its ability to function as an insulator, preventing the spread of silenced chromatin and thereby altering the silencing status of nearby pol II genes [29–31]. Thus, it is possible that the tRNA intron in tL(CAA)N additionally affect RPS7B expression in the tl(caa)∆int* strain. To verify whether only the loxP::hisG::CmR::loxP insertion into the RPS7B mRNA 3’-UTR, but not the intron removal from the tL(CAA) genes, resulted in the observed phenotypes, we generated two types of rps7b derivatives, rps7b-101 and rps7b-102, which possess different insertions within the 3’-UTR of RPS7B at the same position as the tl(caa)Δint* mutant, and introduced into strains with the wild-type tL(CAA) genes. In the rps7b-101 strain, a CgHIS3 marker was introduced as an inserted sequence. In contrast, the rps7b-102 strain carries the rps7b-102 allele, which contains the rps7b::loxP::CmR::hisG::loxP insertion retrieved from the tl(caa)Δint* strain (Fig 3A). Unlike tl(caa)∆int*, neither rps7b-101 nor rps7b-102 alone conferred any growth defect at 30°C on YPD (Fig 3B). Northern analyses revealed that the authentic RPS7B mRNA was barely detected in the rps7b-102 mutant while the longer transcript (white triangle) was stabilized by introducing upf1Δ like the tl(caa)∆int* mutant (Fig 3C, lanes 3 and 7). Consequently, the rps7b-102 recapitulated the phenotypic characteristics of tl(caa)Δint* in terms of RPS7B mRNA expression, which is subject to NMD, and these phenotypes remained unaffected by the presence or absence of intron in the tRNA^Leu^CAA genes. We further examined the effects of very low-level expression of RPS7B in the rps7b-102, possessing the RPS7A-HA::CgHIS3 background. Again, the synthetic growth defect observed in the tl(caa)∆int* was reproduced in the rps7b-102 derivative, and the defect was suppressed by upf1∆ (Fig 3D). Stabilization of rps7b-102 mRNA by upf1∆ in the RPS7A-HA::CgHIS3 background was confirmed by northern blotting (Fig 3C, lanes 9 and 10). RT-PCR analyses were conducted with cDNA obtained from the rps7b-102 mutant that harbors the wild-type tL(CAA) genes and upf1∆. These analyses revealed that the 3’-UTR of the rps7b-102 mRNA terminates approximately 1820–2150 nt downstream of the stop codon under the wild-type tL(CAA) background, mirroring the findings observed in cDNA from the tl(caa)∆int* mutant. And, the tRNA intron had no effect on RPS7B expression with the long 3’-UTR.
NMD for RPS7B mRNA with altered 3’-UTR is affected by the inserted sequences.(A) The panel illustrates the 3’-part of RPS7B of the wild-type, rps7b-101 (CgHIS3 insertion), and rps7b-102 alleles. A number in parentheses represents nucleotide length of the corresponding part (bp). (B) Growth comparison of the wild-type, tl(caa)∆int, rps7b-102, and rps7b-101 strains on a YPD plate at 30°C. (C) Northern analyses of RPS7A and RPS7B mRNAs in the indicated rps7b-101- and rps7b-102-based strains. Cells were grown at 30°C in YPD. Black triangles indicate WT RPS7A or RPS7B mRNAs. Grey and white triangles indicate RPS7A-HA and the longer transcript of rps7b-102 mutants, respectively. For RPS7B mRNA, light-contrast (upper) and dark-contrast (lower) images of the same blot are displayed. The bottom image is GelRed-stain of the RNA gel as in Fig. 2C. (D) Growth comparison of the indicated RPS7A-HA::CgHIS3 variants on a YPD plate at 30˚C.*
Intriguingly, the rps7b-101 mutant that has the CgHIS3 insertion, instead of the rps7b-102 allele, produced RPS7B mRNA with similar length and signal intensity to the wild type. The mutant did not express any longer transcripts stabilized by upf1Δ (Fig 3C, lanes 4 and 8). RT-PCR analyses showed that RPS7B mRNA in the rps7b-101 mutant has its major 3’-terminus between the 206–265 nt downstream the termination codon and its minor terminus between the 265–598 nt (S2 Fig). Indeed, the former major terminus is similar to that of the wild-type (170–251 nt) [20,21]. Therefore, it seems that the inserted CgHIS3 sequence coincidentally introduces cryptic poly-A signals within a permissive range to avoid NMD.
Abundance of FLAG-tagged Rps7Bp demonstrates a positive correlation with mRNA abundance: it is upregulated in the rps7a∆ strain and downregulated in the tl(caa)Δint* strain
According to Ghulam et al. [7], Rps7p expression is regulated by both transcriptional and translational levels. To assess the transcriptional effect of RPS7B, we first performed semi-quantitative RT-PCR. As shown in S3 Fig, the primary transcript of RPS7B having its intronic sequence was not reduced in the tl(caa)Δint* nor the rps7b-102 mutants, implying that the transcriptional process of RPS7B gene was not compromised by the marker insertion in its 3’-UTR. We then examined whether the tl(caa)Δint* strain impaired RPS7B mRNA expression post-transcriptionally, hypothesizing that mRNA abundance directly reflects protein levels due to aberrant 3’-UTR. To analyze whether Rps7Bp expression correlates with its mRNA abundance, we performed western blotting using strains expressing FLAG-tagged Rps7Bp in the wild-type, the tl(caa)Δint* and the rps7aΔ backgrounds. The FLAG tags were introduced at the end of the ORF, and conjugated with the stop codon and 3’-UTR of RPS7B in each strain (Fig 4A). As shown in Fig 4B and 4C, the yield of FLAG-tagged Rps7Bp reduced to one-fourth in the tl(caa)Δint* mutant (tl(cca)∆int*/WT, 0.26 ± 0.13; pairwise comparison with Benjamini-Hochberg False Discovery Rate (FDR) used as a multiple test correction method, hereafter BH; *p *< 0.0001), but rose more than 1.5 times in the rps7aΔ (rps7a∆/WT, 1.71 ± 0.09; BH, *p *< 0.0001). mRNAs encoding RPS7B-FLAG variants were less stable compared with the untagged RPS7B mRNA (Fig 4D, lanes 1 and 4), The FLAG-tagged mRNAs retained expression difference of their untagged counterparts (Fig 4D, lanes 1–3 versus lanes 4–5). These findings, in conjunction with others (Figs 1 and 2), demonstrate that disruption of proper RPS7A expression leads to an enhancement in RPS7B mRNA expression, subsequently results in an upregulating of Rps7Bp production in accordance with the mRNA level.
Translational status of Rps7Bp in the tl(caa)Δint and rps7aΔ strains.(A) The panel illustrates the RPS7B region of the wild-type, RPS7B-FLAG, and RPS7B-FLAG with the tl(caa)Δint*-derived insertion alleles. (B) Western blot analysis of FLAG-tagged Rps7Bp in the wild type (WT), tl(caa)Δint* and rps7aΔ background. Total cell extracts were analyzed as in Fig. 2D. (C) A bar chart represents relative abundance of FLAG-tagged Rps7Bp quantified and normalized by abundance of Srp1p in (B). Multiple comparisons were performed using the Benjamini-Hochberg (BH) method (***, p < 0.0001, n = 3 or 4). The wild-type level (1.0) is shown as a dashed line. Error bars indicate standard deviation. (D) Northern analyses of RPS7A and RPS7B mRNAs in the indicated strains. Cells were grown at 30°C in YPD, and log-phase cells were harvested. Black triangles indicate WT RPS7A and RPS7B mRNAs. A grey triangle indicates RPS7B-FLAG mRNA. The bottom panel represents total RNAs visualized by GelRed staining prior to northern analysis.
General RP gene transcriptional regulators, Fhl1p and Rap1p, play major roles of RPS7-paralog responsive expression
In the traditional view, regulation of RPs often occurs at protein level or in response to the amount of assembled ribosomes, with feedback process frequently leading to transcriptional repression of RP genes [32]. Thus, RPS7 paralog-specific feedback response would initiate at transcriptional level, and in this process, the upstream and/or downstream regions of the ORF may provide clues to understanding the RPS7-specific gene regulation mechanism in vivo. To explore difference of transcriptional features between RPS7A and RPS7B genes, we first focused on the upstream region and searched for known cis elements located in the promoter region of RPS7A and RPS7B on Saccharomyces Genome Database. Indeed, the RPS7 promoters have consensus sequences for prominent transcription factors, Fhl1p and Rap1p [33] (Fig 5A). To assess impacts of these cis elements on RPS7 expression, we conducted reporter gene assay utilizing both wild-type and mutated versions of their respective promoters and the common CYC1 terminator (S2 Fig). RPS7A contains one binding site each for Rap1p and Fhl1p with a 5-base pair overlap between the two sites. We deleted this entire sequence covering the Rap1p and Fhl1p binding sites to make a mutated version of the RPS7A promoter (Fig 5A, rps7a-mut). Conversely, RPS7B contains two Fhl1p binding sites, one of which overlaps with the Rap1p consensus sequence; we simultaneously deleted these two sites from the RPS7B promoter region (rps7b-mut).
*The transcriptional response of RPS7 paralogs relies on Rap1p and Fhl1p binding.(A) The left panel illustrates the upstream regions of RPS7A and RPS7B genes. Areas where Rap1p and Fhl1p bind simultaneously are highlighted in red, while a region where Fhl1p bind alone is marked in orange. A number in parentheses represents nucleotide length of the 5’-UTRs (bp). The right panel shows enlarged sequences of the dotted line areas from the left panel. The numbers represent the upstream region relative to the transcription start site. The simultaneous binding sites for Rap1p (RPS7A: 5’-ATGTTTGGGT-3’; RPS7B: 5’-AGGTACGGAT-3’) and Fhl1p (RPS7A: 5’-ACCCAAACAT-3’; RPS7B: 5’-ATCCGTACCT-3’) are highlighted in red, while the Fhl1p binding site alone in RPS7B (5’-ATCCGTACAT-3’) is shown in orange. (B) Western blot analyses to assess expression of the GFP-FLAG-HIS3 reporter gene under the control of wild-type or mutated RPS7 paralogous promoters in the indicated strains. The mutated versions of the promoters (rps7a-mut and rps7b-mut) lack the Fhl1p and/or Rap1p binding sites. Total cell extracts were analyzed as in Fig 2D. A bar chart represents the relative abundance of the reporter-gene product normalized by the abundance of Srp1p across at least three biological replicates. Multiple comparisons were performed using with the BH method (ns, not significant; *, p < 0.05; **, p < 0.01, n = 3 or 4). Error bars indicate standard deviation. (C) Western blot analyses of the GFP-FLAG-HIS3 reporter gene driven by the ADH1 promoter and harboring 3’-UTR of RPS7A or RPS7B gene in the indicated strains. Total cell extracts were analyzed as in Fig 2D. A bar chart represents the relative abundance of the reporter-gene product normalized by the abundance of Srp1p in more than three biological replicates. Multiple comparisons were performed using with the BH method (ns, not significant; , p < 0.05, n = 4 or 5). (D) Western blot analyses of the GFP-FLAG-HIS3 reporter gene described in (B) in the wild-type and fhl1∆ strains. A bar chart represents the relative abundance of the reporter-gene product normalized and presented as in (B).
In wild-type cells (WT), deletion of the Fhl1p and Rap1p cis elements from the RPS7A promoter (rps7a-mut) did not change the reporter expression (Fig 5B; 0.78 ± 0.22 vs 0.91 ± 0.25, BH within the same RPS7 genetic background, *p *> 0.05), while the mutated version of the RPS7B promoter (rps7b-mut) failed to drive the reporter expression (0.68 ± 0.25 vs 0.04 ± 0.09; BH within the same RPS7 genetic background, *p *= 0.013). The Fhl1p- and/or Rap1p-binding sites are selectively crucial for RPS7B expression under the WT background. Consistent with the expression difference of endogenous RPS7B between WT and rps7a∆ strains (Fig 4), the wild-type RPS7B promoter led to a 2.9-fold increase in the reporter expression in the rps7a∆ compared to that in the WT cells (Fig 5B, BH over different RPS7 genetic backgrounds, *p *= 0.038). Similarly, the wild-type RPS7A promoter also yielded a 7.7-fold increase in reporter expression in the absence of chromosomal RPS7A (BH over different RPS7 genetic backgrounds, *p *< 0.0001). This suggests the presence of a transcriptional feedback mechanism affecting both RPS7A and RPS7B in response to the level of Rps7Ap. Although this cis element mutation on the RPS7A promoter did not affect reporter expression in the WT background, the same mutation resulted in reduction of the reporter expression in the rps7a∆ (6.02 ± 1.55 vs 2.99 ± 0.96; Fig 5B, BH within the same RPS7 genetic background, p = 0.0078). Thus, the cis elements for Fhl1p and/or Rap1p are required for full response of RPS7A expression in the rps7a∆ background. Further, as expected, the rps7b-mut promoter could not drive reporter expression in the rps7a∆ background. In the rps7b∆ cells, both the wild-type RPS7A and RPS7B promoters led to comparable reporter expression to that in the WT cells (Fig 5B; 1.09 ± 0.21 and 1.06 ± 0.48, respectively). While disruption of the cis element for Fhl1p and Rap1p in the RPS7A promoter seemed to reduce reporter expression, this was not statistically significant (Fig 5B; BH within the same RPS7 genetic background, *p *> 0.05). Again, no reporter expression was seen with the rps7b-mut promoter in the rps7b∆. Collectively, the upstream region, particularly the Fhl1p and/or Rap1p binding sites, play a crucial role in RPS7 paralog specific response.
Secondly, to evaluate the impact of the downstream region on their paralog-specific expressions, we additionally generated reporter genes in which constructs share a common ADH1 promoter and GFP-FLAG-HIS3 gene, with the only difference being the 3’-UTRs and terminators (RPS7A or RPS7B), in contrast to the upstream region. As shown in Fig 5C, although the overall expression of these reporter genes was mildly decreased in the rps7a∆ background, either cases that carry the RPS7A or RPS7B 3’-UTR and terminator showed similar reporter expression among wild-type, rps7a∆, and rps7b∆ strains. Thus, the 3’-UTRs of RPS7 paralogs alone had a limited impact on paralog-specific gene regulation, prompting us to further investigate the role of the promoter regions.
Since Fhl1p and/or Rap1p binding sites are crucial for RPS7-paralog specific response (Fig 5B), we generated an fhl1∆ strain and repeated the reporter assay to analyze whether Fhl1p acts on RPS7 paralog expression. Because RAP1 is essential for yeast growth we could not conduct deletion analysis of this transcription factor. As shown in Fig 5D, fhl1∆ did not lose reporter expression driven by the wild-type RPS7A promoter (lane 2), while the expression was completely abolished in the other cases (lanes 3–5). Considering that the wild-type RPS7A promoter was still active in the fhl1∆ strain (Fig 5D, lane 2) and the rps7a-mut promoter functioned in the WT cells (Fig 5B, lane 3), the Fhl1p-site in the RPS7A promoter region may be recognized not only by Fhl1p but also by another factor; probably Rap1p, and Fhl1p may bind an additional site in this region or via interaction with other DNA-binding proteins. Conversely, transcription of RPS7B strictly depends on Fhl1p, suggesting that Fhl1p drives its expression by binding at least one of the two Fhl1p-binding sites of the RPS7B promoter. Thus, the different contributions of Fhl1p (and/or Rap1p) to the RPS7-paralog promoter regions are key determinants of RPS7-paralog expression, which is influenced by the levels of Rps7Ap (or total Rps7p).
Discussion
Our study revealed an interesting finding: the deletion of RPS7A led to an upregulation in the expression of RPS7B, as shown in Fig 1B. However, the opposite effect was not observed; deleting RPS7B did not reciprocally impact the expression of RPS7A. This observation suggests that RPS7A, but not RPS7B, works as an expression modulator of its counterpart (Fig 1B). Our serendipitous finding began with an unexpected genomic change near a tRNA gene during construction of the tl(caa)∆int strain. Although our case is completely artificial, insertion are often introduced at genomic fragile sites by transposons [34]. Indeed, tRNA loci exhibit mutation rates approximately 7–10 times greater than the genome-wide average [35]. Thus, genomic changes or mutations similar to our mutant might arise under natural conditions during evolution.
Asymmetrical regulation has been described in several RP-paralog pairs [3]. It is considered that not only transcriptional initiation but also transcriptional termination and/or splicing acts as the decisional step defining amounts of the RP mRNAs, and rather RNA degradation is not a major determinant of RP mRNA abundance [3]. Since the production of FLAG-tagged Rps7Bp was well correlated with its mRNA level (Fig 4B–4D), the critical regulatory step in Rps7Bp expression driven by RPS7A products seems to rely on mRNA abundance. Indeed, our reporter assay revealed that reporter expression from the RPS7B promoter is upregulated by deletion of the RPS7A gene, but also indicated that the RPS7A promoter is also under feedback regulation. Additionally, well-known transcription factors of RP genes, Fhl1p and/or Rap1p, contributed to the full activation of RPS7A and RPS7B (Fig 5B and 5C). Notably, the RPS7B but not RSP7A promoter-driven expression was completely abolished when FHL1 was deleted (Fig 5D and 5E; see below). These results showed asymmetric features of RPS7 paralogous genes in transcriptional regulation. In addition, we found that transcription of RPS7A is also under the feedback regulation by Rps7Ap, suggesting that maintaining total level of Rps7p/eS7 is important for yeast physiology. Despite of such regulatory mechanism, deletion of one of RP paralog pairs can lead pivotal effects on growth competition among yeast cells, which sometimes result in survival of mutants with drastic genomic alterations, such as aneuploidy [4,8]. Therefore, in the case of Rps7p regulation, the increase in RPS7B mRNA levels that observed in rps7a∆ is possibly due to aneuploidy, specifically a duplication of the chromosome that carries RPS7B.
The rps7b-102 mutation found in the tl(caa)Δint* strain resulted in generation of RPS7B mRNA with an approximately 2-knt 3’-UTR, subsequently triggering its degradation through NMD (Fig 1B and S1 Fig). This occurs irrespective of the presence or absence of the intron in the tL(CAA) loci (Fig 3C). It is well-established that the majority of transcripts with 3’-UTRs of 350 nt or longer are subject to degradation via NMD in the yeast [22,26], while there are some exceptions, such as SSY5, which possess 3’-UTRs ranging in size from 420 to 500 nt [22]. In the case of the rps7B-101, cells utilized some multiple cryptic transcription termination sites for RPS7B mRNA (S2 Fig). Due to the primary termination occurring approximately 200–270 nt downstream the stop codon, which is short enough to escape from NMD, the predominant rps7b-101 transcript exhibited a similar expression level to the wild-type (Fig 3C). So far, we have been unable to identify conventional polyadenylation signal motifs, such as positioning elements AAUAAA/AAAAAA or upstream elements UAUAUA [36], in close proximity to the anticipated 3’-ends of the 3’-UTRs of rps7b-102 or rps7b-101. The mechanism by which yeast cells select these sites for poly-A addition remains unclear.
In the case of HA-tagged RPS7A mRNAs, an ADH1 terminator was utilized instead of the authentic one in RPS7A, resulting in lower expression compared to the wild type (Fig 2C). Since the deletion of UPF1 did not rescue this phenotype, it is conceivable that HA-tagged RPS7A mRNAs may be susceptible to transcriptional repression and/or degraded mediated by an mRNA degradation pathway(s) distinct from NMD. Indeed, the RP-paralog pairs generally have heterogeneous 3’-UTRs. Specifically, the minor paralog tends to exhibit alternative termination sites and/or a longer 3’-UTR [3]. Therefore, alteration of the 3’-UTR could have a significant impact on the abundance of RPS7 mRNAs by potentially losing some cis-elements that are unique to RP mRNAs. We did not investigate the effect of a 3’-UTR insertion in the RPS7A 3’-UTR, specifically a 2.3-kb insertion of loxP::CmR::hisG::loxP, as seen in rps7b-102. Given the genomic context, there are no tRNA genes or fragile sites around RPS7A, suggesting that natural insertion would not occur. However, more comprehensive studies are needed to examine the impact of 3’-UTR insertions in RPS7A to better understand the overall imbalance in regulation between paralogous genes.
Since FLAG-tagged RPS7B mRNAs, utilizing the authentic promoter, displayed reduced expression even with the RPS7B-FLAG gene featuring the genuine 3’-UTR (Fig 4D), it appears that the insertion of an approximately 100 bp fragment into the C-terminal part of the ORF exerts a significant influence on RPS7B expression. Codon utilization and other ORF features are expected to coevolve with mRNA properties, such as length, translational efficiency, and 3’-UTR regulation; consequently, all of these factors, individually or in concert, have potential to impact mRNA stability [37–39]. There may be interdependence between a regulatory sequence(s) in the 3’-UTR and its co-acting element(s) in the ORF of RPS7B. Thus, alteration in length between the ORF and the 3’-UTR may affect stability of RPS7B mRNA.
Our reporter gene assay showed the different contributions of Fhl1p and/or Rap1p on RPS7 paralogs expression, and also showed that RPS7A/B regulation according to the Rps7Ap level is mainly carried out via their promoter regions but not via their 3’-UTRs. The RPS7A promoter is under feedback regulation by abundance of Rps7Ap (or sum of Rps7 proteins), and this regulation is partially compromised by the Fhl1p site deletion. Fhl1p is a RP gene-specific transcription factor collaborating with an essential transcription activator Ifh1p, and has a pivotal role in RP regulation upon various stresses, including unbalanced expression of RPs [40,41]. Deletion of the Fhl1p site in the RPS7A promoter had the least effect on reporter expression, and the FHL1 gene deletion alone has a marginal effect on the RSP7A/B promoter activity. However, simultaneous deletion of these two elements completely abolished the reporter expression. This suggests that the Fhl1p site in the RPS7A promoter is dispensable for Fhl1p function, as Fhl1p can still be recruited to the RPS7A promoter via interactions with other transcription factors, such as Rap1p (a pioneer transcription factor [42]) and Sfp1p (a transcription factor requited by Rap1p [40]), and/or Hmo1p (an HMG family chromatin interactor [43]), which associated with the RPS7A promoter region by themselves [40,43,44]. Indeed, it was reported that IFH1 but not FHL1 is essential for yeast growth [41], and that a DNA-recognizing fork-head domain is not crucial for Fhl1p function [45]. Thus, in the absence of Fhl1p, another factor bound to the Fhl1p site may recruit Ifh1p to the RPS7A promoter. If this Fhl1p-independent recruitment of Ifh1p to the RPS7A promoter requires the Fhl1p site or a surrounding sequence, concomitant deletion of the Fhl1p site and the FHL1 gene may hinder Ifh1p from associating with the RPS7A promoter and its transcription. On the other hand, both the Fhl1p site and Fhl1 protein are indispensable to transcription from the RPS7B promoter. Although RPS7A and RPS7B belong to the same Category I RP genes, where Hmo1p contributes to recruitment of Fhl1p•Ifh1p to their promoter [46], the RPS7A promoter does not require Fpr1p, an FKBP12 homolog in the yeast, for its full function, unlike the RPS7B promoter. This difference in promoter architecture between these paralogs may result in the difference in requirement of the Fhl1p site and Fhl1p itself. Since deletion of either of these genetic elements completely disrupts the RPS7B promoter activity, contribution of these elements to Rps7Ap-dependent regulation of the RPS7B promoter is still to be analyzed. Moreover, Rps7Ap may directly interact with a transcriptional regulator to achieve RPS7A-specific feedback regulation, as known by the general autoregulatory mechanism of RP production [3]. At least, Fhl1p is not the target of Rps7Ap (Fig 5), indicating that Rap1p is the prime target identified so far. While Rap1p is essential, the use of a ts-mutant [47,48] or Auxin-degron system, and the rap1∆sil mutant [49] offers valuable tool for further investigation.
Eukaryotic specific ribosomal protein Rps7p/eS7 locates at the 40S head region contacting with Rps13p/uS15, Rps22p/uS8 and Rps27p/eS27 [2,50]. In humans, Rps7p is one of the RPs have implicated in hematopoietic disorders, most notably Diamond-Blackfan anemia [51]. A splicing variant of the human RPS7 5’-UTR decreases mRNA level followed by protein reduction, implying that the RPS7 expression level itself is a pathological cause [52]. In budding yeast, RPS7A is a major paralog regarding mRNA expression, reflecting RNA polymerase II association to the ORF, post-cap stability of mRNA, as well as protein abundance [7]. Curiously, the differential incorporation of ribosomal protein paralogs into active ribosomes showed cells’ preference for Rps7Bp-containing ribosomes rather than Rps7Ap-containing ones, indicating a discrepancy between paralog utilization in ribosomes and mRNA/protein abundance of the Rps7p-paralog pair [7]. Also at the post-translational levels, only Rps7Ap is ubiquitinated by Not4p, and that cooperated with RQC machinery to support proteostasis positively [53,54]. Thus indeed, yeast duplicated RP genes are unequally controlled at different levels, including transcription, mRNA stability, translation, RP modification, and response to various stimuli [3,7,12].
Our substantiated evidence suggests that the yeast harbors a surveillance system for RPS7A mRNA level, and utilizes this information to modulate expression of only RPS7B. So far, mechanisms by which the yeast distinctly perceives abundance of RPS7A mRNA in a manner independent of RSP7B and the underlying biological meanings for this asymmetric regulatory phenomenon remain uncovered. It is important to note that, in many paralogous RP genes, the surveillance or feedback typically occurs through protein level or in response to the amounts of assembled ribosomes. It is also possible that feedback mechanism exists to regulate the expression of different genes when specific subunits are not properly assembled, particularly if ribosomes become stalled at certain stages of translation due to improper expression of paralogous RPs. Comprehensive studies that include both mRNA and protein levels should be considered in future research. Future continuous investigations will unveil more details of these regulatory processes.
Supporting information
S1 FigRT-PCR of the 3’ regions of rps7b-102 mRNA.(A) Three prime parts of the rps7b-102 gene and positions of primers used in the PCR (small red arrows) were schematically shown. (B) RT-PCR products of the rps7b-102 cDNA from SHSC0386 (rps7b-102 tl(caa)∆int upf1∆) amplified by three different cycle numbers were analyzed by agarose gel electrophoresis. Reverse primers used in the PCR are shown on the top of the gel image, and correspond to those in (A). The rightmost three lanes are PCR products with genomic DNA from SHSC0386 as a template, and are used to confirm ability of primers.(TIF)
S2 FigRT-PCR of the 3’ regions of rps7b-101 mRNA.(A) Three prime parts of the rps7b-101 gene and positions of primers used in the PCR (small red arrows) were schematically shown. (B) The 3’ parts of the rps7b-101 were amplified from SHSC0388 (rps7b-101) cDNA and were analyzed as in S1 Fig. Primers used in the PCR are shown on the top of the gel image, and correspond to those in (A). The rightmost four lanes are PCR products with genomic DNA from SHSC0388 as a template, and are used to confirm ability of primers.(TIF)
S3 FigRT-PCR of unspliced and spliced forms of RPS7A and RPS7B.(A) Schematic drawings of primers (red arrows) used for RT-PCR of RPS7A and RPS7B mRNAs. The F1 primer of RPS7A and the R3 primer of RPS7B are designed to hybridize with only spliced mRNAs, and the F2 of RPS7A and R2 of RPS7B are against intronic regions. (B) Gel electrophoresis of RT-PCR samples. cDNAs were synthesized from total RNAs of the strains shown in the bottom list as in S1 Fig except that random hexamers instead of oligo dT_20_ were used as RT primers. Left panels are RT-PCR of the unspliced form of RPS7A mRNA (upper) and RPS7B mRNA (lower). Right panels are those of their spliced form. In both cases, two PCR cycles (34 and 32 cycles for the unspliced form, and 27 and 25 cycles for the spliced form) are adopted. A primer set used in each amplification is shown under the gel image in red. Lane 1, wild type; lane 2, tl(caa)∆int*; lane 3, rps7b-102. Details of the primers are described in S4 Table in S1 File.(TIF)
S1 FileS1-S4 Tables. Strains and primers used in this study.S1 Table. Yeast strains used in this study. S2 Table. Primers used to construct stains and plasmids. S3 Table. Plasmid used to construct strains. S4 Table. Primers used for RT-PCR.(XLSX)
S2 FileRaw_images.(PDF)
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Woolford JL Jr, Baserga SJ. Ribosome biogenesis in the yeast Saccharomyces cerevisiae. Genetics. 2013;195(3):643–81. doi: 10.1534/genetics.113.153197 24190922 PMC 3813855 · doi ↗ · pubmed ↗
- 2de la Cruz J, Karbstein K, Woolford JL Jr. Functions of ribosomal proteins in assembly of eukaryotic ribosomes in vivo. Annu Rev Biochem. 2015;84:93–129. doi: 10.1146/annurev-biochem-060614-033917 25706898 PMC 4772166 · doi ↗ · pubmed ↗
- 3Petibon C, Malik Ghulam M, Catala M, Abou Elela S. Regulation of ribosomal protein genes: an ordered anarchy. Wiley Interdiscip Rev RNA. 2021;12(3):e 1632. doi: 10.1002/wrna.1632 33038057 PMC 8047918 · doi ↗ · pubmed ↗
- 4Steffen KK, Mc Cormick MA, Pham KM, Mac Kay VL, Delaney JR, Murakami CJ, et al. Ribosome deficiency protects against ER stress in Saccharomyces cerevisiae. Genetics. 2012;191(1):107–18. doi: 10.1534/genetics.111.136549 22377630 PMC 3338253 · doi ↗ · pubmed ↗
- 5Kellis M, Birren BW, Lander ES. Proof and evolutionary analysis of ancient genome duplication in the yeast Saccharomyces cerevisiae. Nature. 2004;428(6983):617–24. doi: 10.1038/nature 02424 15004568 · doi ↗ · pubmed ↗
- 6Shi Z, Barna M. Translating the genome in time and space: specialized ribosomes, RNA regulons, and RNA-binding proteins. Annu Rev Cell Dev Biol. 2015;31:31–54. doi: 10.1146/annurev-cellbio-100814-125346 26443190 · doi ↗ · pubmed ↗
- 7Ghulam MM, Catala M, Abou Elela S. Differential expression of duplicated ribosomal protein genes modifies ribosome composition in response to stress. Nucleic Acids Res. 2020;48(4):1954–68. doi: 10.1093/nar/gkz 1183 31863578 PMC 7038994 · doi ↗ · pubmed ↗
- 8Cheng Z, Mugler CF, Keskin A, Hodapp S, Chan LY-L, Weis K, et al. Small and large ribosomal subunit deficiencies lead to distinct gene expression signatures that reflect cellular growth rate. Mol Cell. 2019;73(1):36-47.e 10. doi: 10.1016/j.molcel.2018.10.032 30503772 PMC 6382079 · doi ↗ · pubmed ↗