| Edited by Fujio Kawamura. Bun-ichiro Ono: Corresponding author. E-mail: ono@se.ritsumei.ac.jp |
Eukaryotic nonsense (termination) suppressors have been best studied in the budding yeast Saccharomyces cerevisiae. Codon-specific suppressors were obtained in the early stage of the study because of their strong suppressor activity. And, they provided useful information about the tRNA genes (Guthrie and Abelson, 1982). Later, less efficient and codon-nonspecific (omnipotent) suppressors became the main targets of investigation and yielded valuable information about the translational machinery (Hinnebusch and Liebman, 1992). It is also found in this line of studies that the cytoplasmic omnipotent suppressor, [PSI+], is the conformational isomer of eRF3 encoded by the SUP35 gene; [PSI+] which is also referred to as ψ, psi, PSI or [PSI] is now widely known as a yeast prion (Wickner, 1994; Lindquist, 1997; Chernoff, 2001).
In the study of S. cerevisiae termination suppressors, it has been a common practice to use strains carrying previously defined suppressors to search new suppressors. Increasingly weak suppressors are unveiled in turn with this strategy and have enriched the list of termination suppressors (Sherman, 1982). Although most of the termination suppressors obtained by now have been elucidated for their molecular nature, there still remain a few whose molecular nature is obscure. Among them are so-called [PSI+]-dependent recessive omnipotent suppressors. They were first identified in a strain with the genetic background of [PSI+] SUP29 (Ono et al., 1982); SUP29 is a leucine-inserting UAA-specific termination suppressor (Ono et al., 1979). In this genetic background, the leu2-1UAA mutation was suppressed, but the lys1-1UAA and his5-2UAA mutations were not. That is, a strain of the genotype [PSI+] SUP29 leu2-1UAA lys1-1UAA his5-2UAA was Leu+ Lys– His–. After UV mutagenesis of such a strain, Lys+ His+ revertants were selected and were found to have new mutations that act as weak and recessive suppressors even in the absence of SUP29. These suppressors were divided into three complementation groups, sup111, sup112 and sup113, which were mapped on the right arms of chromosomes VIII, VII and XIII, respectively (Ono et al., 1986). Although we attempted to screen the wild-type gene of SUP111, from an S. cerevisiae genome library, by mean of complementation of sup111, the attempt was unsuccessful. In this attempt, we however obtained HSP104 and found that over-expression as well as disruption of HSP104 caused elimination of [PSI+] (Chernoff et al., 1995). It was also became apparent that sup111, sup112 and sup113 do not manifest suppressor activity, as judged by the action on leu2-1 UAA, in the [PSI–] cytoplasm; here, [PSI–] indicates absence of [PSI+].
Independent from the above line of studies, it has become evident that S. cerevisiae rapidly degrade mRNAs that bear premature translational termination codons. This function, which is widely referred to as NMD (nonsense-mediated mRNA decay), is first described by Leeds et al. (1991). It is well established by now that three genes are essential for NMD, though they are named differently by different researchers; YMR080C (UPF1, NAM7, IFS2, MOF4) (Leeds et al., 1991; Leeds et al., 1992; Altamura et al., 1992; Cui et al., 1996), YHR077C (UPF2, NMD2, IFS1, SUA1) (Leeds et al., 1992; Cui et al., 1995; He et al., 1996; He et al., 1997) and YGR072W (UPF3, SUA6) (Leeds et al., 1992; He et al., 1997). We think NMD is the most appropriate name representing their function, but we use UPF in this report because it is not only chronologically oldest but also convenient to collectively refer to the genes of our interest. The UPF genes drew our attention because of the facts that disruption of UPF1 confers termination suppressor activity (Leeds et al., 1991) and that some of the frameshift suppressors, which also act as omnipotent suppressors, are allelic to the UPF genes (Culbertson et al., 1980). More significantly, UPF1, UPF2 and UPF3 were located on the right arms of chromosomes XIII, VIII and VII, respectively; note that the map positions of sup111, sup112 and sup113 deduced by linkage analyses very well coincide with those of UPF2, UPF3 and UPF1 defined by the Saccharomyces genome project (Fig. 1). It is therefore intuitively thought that sup111, sup112 and sup113 are allelic to UPF2, UPF3 and UPF1, respectively. However, no definitive evidence for this correlation has ever presented. We thus decided to test complementation of sup111, sup112 and sup113 with UPF2, UPF3 and UPF1, respectively. Here, we present our results.
 View Details | Fig. 1. Comparison of map positions of sup111, sup112 and sup113 with those of UPF2, UPF3 and UPF1, respectively. Map positions of the genes of our interest were deduced from information available in the Saccharomyces genome database (http://www.yeastgenome.org/). Thick lines represent regions of chromosomes VIII (A), VII (B) and XIII (C) relevant to the present study. Linkage distances (cM) and physical distances (approximate kbp) between the indicated loci are shown above and below the thick lines, respectively. |
S. cerevisiae strains used in this study are listed in Table 1. YC13-6C, IA35-2B-ura3 and YC9-1A had sup111, sup112 and sup113, respectively in the [PSI+] cytoplasm. These strains contained the leu2-1UAA mutant gene, and we used this gene to monitor suppressor activity throughout this study. Plasmid pRS316 (Sikorski and Hieter, 1989) was used as the vector for the genes of our interest. Plasmid pBluescript II SK+ (Toyobo, Tokyo, Japan) was used for sequence analysis. Escherichia coli strain DH10B (Sambrook et al., 1989) was used as the host to amplify the plasmids.
 View Details | Table 1. S. cerevisiae strains used in this study |
The fragment containing UPF1, UPF2 or UPF3 was PCR-amplified using primers which were made in a way that the obtained fragment to have the SacI and SmaI sites near the two ends, respectively. The amplified fragment was digested with SacI and SmaI and ligated with pRS316 digested with the same set of restriction enzymes. Strain DH10B was treated with the reaction mixture, and Ampr transformants were obtained. Plasmid was recovered from each transformant and subjected to sequence analysis. The plasmids which were confirmed to have the sequences corresponding to UPF1, UPF2 and UPF3 were designated pRS-UPF1, -UPF2 and -UPF3, respectively, and were used in this study.
S. cerevisiae standard media (Sherman et al., 1974) were used. YPD contained 1% Bacto yeast extract, 2% Bacto peptone and 2% glucose. SD contained 0.67% Difco yeast nitrogen base without amino acids and 2% glucose. SD medium was supplemented with appropriate nutrients to test growth characteristics of the strains. Growth temperature was 30°C. For growth of E. coli, LB medium (Sambrook et al., 1989) was used; growth temperature was 37°C.
Standard DNA manipulation procedures (Sambrook et al., 1989) were adopted. PCR was achieved using a thermal cycler (PTC-150; M. J. Research, Inc. Waltham, MA, USA) and Pyrobest DNA Polymerase (Takara Shuzo Co., Ltd., Kyoto, Japan). DNA amplification was accomplished by the following program: 1 min at 94°C (1 cycle); 10 sec at 98°C, 30 sec at 55°C, 2 min at 72°C (25 cycles), and 10 min at 72°C (1 cycle). Restriction enzymes and ligase were purchased from Takara Shuzo Co. Ltd. (Kyoto, Japan) and/or Nippon Gene (Toyama, Japan) and used as suggested by the manufacturers.
DNA for sequence analysis was prepared using a DYEnamic Terminator Cycle Sequencing Kit (Amersham Pharmacia, Piscataway, NJ, USA). Amplification of DNA was accomplished by 25 cycles of 95°C, 20 sec –50°C, l5 sec –60°C, 60 sec using a Mini CyclerTM (MJ Research, Inc., Tokyo, Japan). After completion of the cycles, the reaction mixtures of four bases were combined and subjected to ethanol precipitation. The precipitate was vacuum-dried, dissolved in the mixture of deionized-formamide and 15 mM EDTA (5:1), and applied to a DNA sequencer (ABI 373A; Applied Biosystems Japan Co., Ltd.). For electrophoresis, 6% polyacrylamide gel (50% urea, 5.7% acrylamide, 0.3% methylene bis-acrilamide, 0.05% APS and 0.04% TEMED) and lx TBE buffer (90 mM Tris base 90 mM borate and 2.2 mM EDTA) were used. Electrophoresis (2,500 V, 40 mA and 30 W) was run at 40°C for 14 hr. Data analysis was made with the 373A software (Applied Biosystems Japan Co., Ltd.).
Total RNA was extracted by the method of Schmitt et al. (1990), and an aliquot containing 50 μg RNA was subjected to agarose gel electrophoresis (Sambrook et al., 1989). RNAs were then transferred to nitrocellulose membrane by the method of Southern (1975). The membrane was challenged with the PCR-amplified LEU2-containing fragment, which was end-labeled with digoxigenin-11-ddUTP by means of terminal transferase (Roche Diagnostics, Tokyo, Japan). RNA-DNA hybridization was monitored using a digoxigenin-detecting kit (Roche Diagnostics, Tokyo, Japan). rRNAs revealed by ethidium bromide (EtBr) staining were used as internal controls.
Strains YC13-6C (sup111), IA35-2B-ura3 (sup112) and YC9-1A (sup113) were treated with plasmids pRS-UPF2, pRS-UPF3 and pRS-UPF1, respectively, and Ura+ transformants were obtained. A transformant representing each plasmid was tested for growth. As shown in Fig. 2A, strain YC13-6C (lane 1) did not grow on -Ura +Leu and -Ura -Leu media because of its possession of ura3. But, it grew, though somewhat weakly, on +Ura -Leu medium due to suppression of leu2-1UAA by [PSI+] sup111. The same strain transformed with pRS316 (lane 2) grew well not only on -Ura +Leu medium but also on +Ura -Leu and -Ura -Leu media because of its possession of the URA3-bearing plasmid. When strain YC13-6C was transformed with pRS-UPF2 (lane 3), it grew well on -Ura +Leu medium but not on -Ura -Leu medium, indicating that suppression of leu2-1UAA by [PSI+] sup111 is diminished by the presence of UPF2. Here, it should be mentioned that strain YC13-6C/pRS-UPF2 showed residual growth on +Ura -Leu medium. This was attributed to the loss of pRS-UPF2 in nonselective conditions because the cells grown up were all Ura– if tested on the separate plates (data not shown). It should be also added that strains IA35-2B-ura3 and YC9-1A showed good growth on -Leu medium even after transformation with pRS-UPF2 (data not shown), indicating that the effect of pRS-UPF2 was specific to sup111.
 View Details | Fig. 2. Complementation of sup111, sup112 and sup113 with the UPF genes. Strains YC13-6C (A), IA35-2B-ura3 (B) and YC9-1A (C) were used in the experiment. Cells with no plasmid (lane 1), control plasmid (lane 2) or test plasmid (3) were suspended in water at a density of 106 per ml. Then, a 0.1 ml aliquots were spotted on the indicated agar plates and incubated at 30°C for four days; control plasmid was pRS316, and test plasmids were pRS-UPF1 (A), pRS-UPF2 (B) and pRS-UPF3 (C). |
We achieved the parallel experiments for strains IA35-2B-ura3 (Fig. 2B) and YC9-1A (Fig. 2C) using plasmids pRS-UPF3 and pRS-UPF1, respectively. The results obtained were exactly identical with those obtained for strain YC13-6C with plasmid pRS-UPF2. From these results, we concluded that sup111, sup112 and sup113 were complemented with UPF2, UPF3 and UPF1, respectively.
In order to test whether the transformed UPF genes truly affect the intracellular level of leu2-1 mRNA, we achieved Northern blot analysis using the LEU2 gene as a probe. As shown in Fig. 3A, strain YC13-6C (lane 1) had about twice more leu2-1 mRNA comparing to the same strain transformed with pRS-UPF2 (lane 2). Similarly, strains IA35-2B-ura3 (Fig. 3B) and YC9-1A (Fig. 3C) had about twice and three times more leu2-1 mRNA than strains YC13-6C/pRS-UPF3 and IA35-2B-ura3/pRS-UPF1, respectively. These results are in agreement with the results of the preceding complementation analyses and support our contention that sup111, sup112 sup113 cause accumulation of leu2-1 mRNA that bears a premature translational termination codon.
 View Details | Fig. 3. Effects of the UPF genes on the cellular level of leu2-1 mRNA. Strains YC13-6C (A), IA35-2B-ura3 (B) and YC9-1A (C) were transformed with control plasmid (lane 1) or UPF-bearing plasmid (lane 2). Total RNA was extracted from the cells grown in -Leu medium (A), or in -Ura -Leu medium (B and C), to late logarithmic phase, fractionated by agarose gel electrophoresis, transferred to nitrocellulose membrane and then probed with the DNA fragment containing LEU2 as described in Materials and Methods. Relative intensity of the 1.2 kb band revealed by Northern blot hybridization was deduced by dividing the densitometric intensity of this band (upper panels) with that of the sum of rRNA bands revealed by EtBr-staining (lower panels). |
To further confirm the above results and also to seek the nature of the sup111, sup112 and sup113 mutations, we tried to analyze their nucleotide sequences. To this purpose, we PCR-amplified the entire region of UPF2, UPF3 and UPF1 using genomic DNA of strains YC13-6C (sup111), IA35-2B-ura3 (sup112) and YC9-1A (sup113), respectively, as templates. The amplified DNA fragments were inserted into the SmaI site of plasmid pBluescript II SK+ and subjected to sequence analyses; for each experiment, we examined three independent fragments and obtained the identical sequence. The results are summarized in Fig. 4. It is seen that sup111 had a stretch of 7 As, in place of a stretch of 8 As, at the position starting from the 1609th nucleotide of UPF2. Similarly, sup112 had a stretch of 5 As, in place of a stretch of 6 As, at the position starting from the 585th nucleotide of UPF3. We therefore conclude that sup111 and sup112 are frameshift mutations arisen in the regions of A repeat in UPF2 and UPF3, respectively. Contrastingly, sup113 had a base substitution of G to T at the 1289th nucleotide of UPF1. That is, sup113 is a missense mutation of UPF1; with this mutation, the 431st amino acid of Upf1 is expected to be valine instead of glycine.
 View Details | Fig. 4. Sequence analyses of sup111, sup112 and sup113. The regions containing the entire UPF2, UPF3 and UPF1 genes were PCR-amplified using genomic DNA of strains YC13-6C (A), IA35-2B-ura3 (B) and YC9-1A (C) as templates, respectively. The obtained fragments were sequenced as described in Materials and Methods. The deduced sequence of each gene was aligned with the sequence of the control strain (S288C) (http://www.yeastgenome.org/). Italics indicate different nucleotides between the control and test strains; number represents the position in the coding region of each gene. |
In this report, we have presented evidence that termination suppressor activity of sup111, sup112 and sup113, in the presence of [PSI+], is diminished by the introduction of UPF2, UPF3 and UPF1, respectively. From these results together with the close proximity of the map positions, we conclude that sup111, sup112 and sup113 are allelic to UPF2, UPF3 and UPF1, respectively. Sequence analyses of the UPF2, UPF3 and UPF1 loci of the strains bearing sup111, sup112 and sup113, respectively, have shown that sup111, sup112 and sup113 are point mutations arisen in the UPF2, UPF3 and UPF1 loci, respectively. From these observations, we contend that sup111, sup112 and sup113 manifest suppressor activity via defects in the NMD machinery. In the present case, leu2-1 mRNA is not degraded and thus accumulates in the cell. Protein synthesis (translation) proceeds using leu2-1 mRNA, but translation termination is malfunctioned because of the possession of [PSI+]. Since [PSI+], the polymerized form of the isoform of a termination factor, eRF3, elongates by stimulating conversion of eRF3 to [PSI+] (Cox et al., 1988; Volkov et al., 2002), the cell with [PSI+] is depleted with functional eRF3 and thus is malfunctioned in termination. That is why [PSI+] acts as a termination suppressor. The [PSI+]-caused increase of readthrough of termination codon, together with increased level of leu2-1 mRNA caused by the impaired NMD machinery, the cell produces functional Leu2 (β-isopropylmalate dehydrogenase) to a level where exogenous supply of leucine is not necessary for growth (Fig. 5). This rationale is contradictory to the claim that the impaired NMD machinery causes increased readthrough (Maderazo et al., 2000) but agrees with the contention of Harger and Dinman (2004). Nevertheless, according to our scheme, sup111, sup112 and sup113 are not termination suppressors. They only promote the suppressor phenotype of [PSI+]; that is, they are phenotypic allosuppressors.
 View Details | Fig. 5. A model for the interaction between [PSI+] and so-called [PSI+]-dependent omnipotent suppressors, sup111, sup112 and sup113. leu2-1UAA mRNA is rapidly degraded by means of the NMD machinery in the wild-type strain. However, if this machinery is completely or partly impaired in the presence of sup111, sup112 or sup113, the level of leu2-1UAA mRNA increases. On the other hand, [PSI+] lowers the level of a termination factor, eRF3, due to its property as a yeast prion and thus causes increased level of readthrough of termination codons. By collaboration of [PSI+] and the impaired NMD machinery, the cell produces sufficient amount of functional Leu2 and eventually grows without exogenous leucine. |
As mentioned in Introduction, a large number of termination suppressors are shown to be mutations of tRNAs, particularly of tRNA anticodons. This group of mutations behaves as dominant and codon-specific suppressors (Guthrie and Abelson, 1982). They manifest termination suppressor activity by faithfully following the codon/anticodon pairing rules. The second group of termination suppressors is codon-nonspecific in their action and has been conventionally referred to as omnipotent suppressors. Dominant suppressors of this group are mutations of ribosomal proteins (Ishiguro et al., 1981; Eustice et al., 1986), whereas recessive ones are mutations of elongation factors (Dinman and Kinzy, 1997) or termination factors (Himmelfarb et al., 1985; Kushnirov et al., 1988). All these mutations act as termination suppressors via reduced stringency of codon/anticodon recognition. While the suppressors mentioned here are termination suppressors regardless whether they act faithfully or not to the codon/anticodon recognition rules, the mode of action of sup111, sup112 and sup113 is quite different from them. It should be stressed that they are suppressors of termination mutations but not termination suppressors. That is, sup111, sup112 and sup113 comprise a new class of mutation distinct from the conventional termination suppressors.
From the study of sup111, sup112 and sup113, we become aware of the fact that mRNA level is an important factor for suppression of termination mutations. In this respect, it may be worth mentioning that we have very recently reported that the ESU1 gene which enhances efficiency of [PSI+] sup111 imposes its activity by acting as a transcriptional activator (Ono et al., 2005). It is now evident that both ESU1 (synthesis) and UPFs (degradation) are genes involved in determination of the levels of mRNAs, particularly degradation of mRNAs bearing premature termination codons in case of UPFs. For us, it is unexpected and surprising that study of termination suppressors deals with not only translation but also synthesis and degradation of mRNA.
The work was partly carried out by funds provided by the Bio-Venture project and the 21st century COE program of Ritsumeikan University.
|
Index