Turnover rate of yeast PGK mRNA can be changed by specific alterations in its trailer structure.

The effect of insertions in the 5'- and 3'-untranslated regions (UTR) of the Saccharomyces cerevisiae mRNA encoding phosphoglycerate kinase (PGK) on the stability of the transcript in vivo was determined. None of the structural alterations in the 5'-UTR affected mRNA turnover significantly, despite the strong negative effect on translational efficiency of some of these alterations previously observed. We conclude that the structure of the 5'-UTR is not important for the relatively high affinity of PGK mRNA in yeast cells. Moreover, translation cannot be a major factor in determining the rate of turnover of this mRNA. Insertion of either a polyG or polyU, but not a polyA or polyC, tract into the 3'-UTR of PGK mRNA increased its half-life by a factor of about two. Introduction of a hairpin structure containing 18 G.C base pairs had only a slight stabilizing effect. We argue that the stabilization by the structural changes in the 3'-UTR is due to altered folding of the mutant mRNA which retards a rate-limiting endonucleolytic cleavage step in the normal turnover pathway of PGK mRNA. The stabilizing effect of local structural alterations in the 3'-UTR opens the possibility for further increasing the product yield of a (heterologous) gene cloned in yeast cells.

Interaction of ribosomal proteins L25 from yeast and EL23 from E. coli with yeast 26S and mouse 28S rRNA.

The interaction of ribosomal protein EL23 from E. coli and L25 from yeast with yeast 26S rRNA was analysed by nitrocellulose filter binding and RNase protection experiments using both intact rRNA and various fragments prepared by in vitro transcription of cloned yeast rDNA regions in the SP6 system. The results show that EL23 efficiently and specifically interacts with the region of 26S rRNA previously identified as the binding site for the yeast ribosomal protein L25. A comparison of the oligonucleotides resulting from limited RNase T1 digestion of the heterologous EL23/26S rRNA complex with those obtained by the same treatment of the homologous L25/26S rRNA complex showed that the molecular details of the two r-protein/rRNA interactions are highly similar if not identical. Using the synthetic 26S rRNA fragments we could demonstrate that all information for the formation of a biologically active binding site is located within the region of the rRNA delimited by the sequences protected by L25 against RNase T1 digestion. Part of the sequence at the 3' end of the 5'-distal protected region, however, was found not to be essential for r-protein binding although it does enhance the efficiency of this binding. Binding experiments using synthetic mouse 28S rRNA fragments showed that neither EL23 nor L25 interact with the structural equivalent of their respective cognate binding sites present in this mammalian rRNA. We argue that the structure of the expansion sequence present in this region of mouse 28S rRNA is a major cause of this failure.

Ribosomal proteins EL11 from Escherichia coli and L15 from Saccharomyces cerevisiae bind to the same site in both yeast 26 S and mouse 28 S rRNA.

The heterologous interaction of Escherichia coli ribosomal protein EL11 with yeast 26 S and mouse 28 S rRNA was studied by analysing the ability of this protein to form a specific complex with various synthetic rRNA fragments that span the structural equivalent of the EL11 binding site present in these eukaryotic rRNAs. The fragments were obtained by SP6 polymerase-directed in-vitro run-off transcription of parts of the yeast or mouse large rRNA gene cloned behind the SP6 promoter. EL11 was found to protect an oligonucleotide fragment of 63 nucleotides from both the yeast and mouse transcripts against digestion by RNase T1. In both cases, the position of this fragment in the L-rRNA sequence coincides almost exactly with that of the fragment previously found to be protected by EL11 in E. coli 23 S rRNA. Moreover, the protected yeast fragment was shown to be able to re-bind to EL11 by a nitrocellulose filter binding assay. A ribosomal protein preparation from Saccharomyces cerevisiae containing L15 (YL23) as well as the acidic proteins L44', L44 and L45 protects exactly the same oligonucleotide fragment as does EL11 in both the yeast and mouse transcripts. Evidence is provided that L15, which is known to be structurally and functionally equivalent to EL11, is the rRNA-binding protein in this preparation. Thus the structural equivalent of the EL11 binding site present in yeast 26 S rRNA constitutes the second example of functional conservation of a ribosomal protein-binding site on rRNA between prokaryotes and eukaryotes.

Yeast ribosomal protein L25 binds to an evolutionary conserved site on yeast 26S and E. coli 23S rRNA.

The binding site of the yeast 60S ribosomal subunit protein L25 on 26S rRNA was determined by RNase protection experiments. The fragments protected by L25 originate from a distinct substructure within domain IV of the rRNA, encompassing nucleotides 1465-1632 and 1811-1861. The protected fragments are able to rebind to L25 showing that they constitute the complete protein binding site. This binding site is remarkably conserved in all 23/26/28S rRNAs sequenced to date including Escherichia coli 23S rRNA. In fact heterologous complexes between L25 and E. coli 23S rRNA could be formed and RNase protection studies on these complexes demonstrated that L25 indeed recognizes the conserved structure. Strikingly the L25 binding site on 23S rRNA is virtually identical to the previously identified binding site of E. coli ribosomal protein EL23. Therefore EL23 is likely to be the prokaryotic counterpart of L25 in spite of the limited homology displayed by the amino acid sequences of the two proteins.

Structure and function of the nontranscribed spacer regions of yeast rDNA.

The sequences of the nontranscribed spacers (NTS) of cloned ribosomal DNA (rDNA) units from both Saccharomyces cerevisiae and Saccharomyces carlsbergensis were determined. The NTS sequences of both species were found to be 93% homologous. The major disparities comprise different frequencies of reiteration of short tracts of six to sixteen basepairs. Most of these reiterations are found within the 1100 basepairs long NTS between the 3'-ends of 26S and 5S rRNA (NTS1). The NTS between the starts of 5S rRNA and 37S pre-rRNA (NTS2) comprises about 1250 basepairs. The first 800 basepairs of NTS NTS2 (adjacent to the 5S rRNA gene) are virtually identical in both strains whereas a variable region is present at about 250 basepairs upstream of the RNA polymerase A transcription start. In contrast to the situation in Drosophila and Xenopus no reiterations of the putative RNA polymerase A promoter are present within the yeast NTS. The strands of the yeast NTS reveal a remarkable bias of G and C-residues. Yeast rDNA was previously shown to contain a sequence capable of autonomous replication (ARS) (Szostak, J.W. and Wu, R (1979), Plasmid 2, 536-554). This ARS, which may correspond to a chromosomal origin of replication, was located on a fragment of 570 basepairs within NTS2.

The in vivo and in vitro initiation site for transcription of the rRNA operon of Saccharomyces carlsbergensis.

We have performed a detailed analysis of the transcription initiation of the rRNA operon in the yeast Saccharomyces carlsbergensis. Electron microscopic analysis of R-looped pre-rRNA molecules together with a very sensitive S1-nuclease mapping showed the use of only a single transcription start at about 700 bp upstream of the 17S rRNA gene and not of the minor start sites proposed for the very closely related species S. cerevisiae by others [Bayev et al. (5), Swanson and Holland (6)]. The sequence of 730 bp of the initiating region is presented. In vitro transcription in concentrated lysates of yeast spheroplasts in the presence of (gamma-SH)ATP or (gamma-SH)GTP, followed by purification of the in vitro initiated RNA via Hg-agarose, revealed that on the endogenous template exactly the same site is used for transcription initiation as in vivo.

The primary and secondary structure of yeast 26S rRNA.

We present the sequence of the 26S rRNA of the yeast Saccharomyces carlsbergensis as inferred from the gene sequence. The molecule is 3393 nucleotides long and consists of 48% G+C; 30 of the 43 methyl groups can be located in the sequence. Starting from the recently proposed structure of E. coli 23S rRNA (see ref. 25) we constructed a secondary structure model for yeast 26S rRNA. This structure is composed of 7 domains closed by long-range base pairings as n the bacterial counterpart. Most domains show considerable conservation of the overall structure; unpaired regions show extended sequence homology and the base-paired regions contain many compensating base pair changes. The extra length of the yeast molecule is due to a number of insertions in most of the domains, particularly in domain II. Domain VI, which is extremely conserved, is probably part of the ribosomal A site. alpha-Sarcin, which apparently inhibits the EF-1 dependent binding of aminoacyl-tRNA, causes a cleavage between position 3025 and 3026 in a conserved loop structure, just outside domain VI. Nearly all of the located methyl groups, like in E. coli, are present in domain II, V and VI and clustered to a certain extent mainly in regions with a strongly conserved primary structure. The only three methyl groups of 26S rRNA which are introduced relatively late during the processing are found in single stranded loops in domain VI very close to positions which have been shown in E. coli 23S rRNA to be at the interface of the ribosome.

Transcriptional units for ribosomal proteins in yeast.

The effects of ultraviolet irradiation on the rates of synthesis of individual ribosomal proteins in yeast were examined and compared with the ultraviolet sensitivities of the synthesis of other yeast proteins. It was found that the synthesis of yeast ribosomal proteins is much more sensitive to ultraviolet irradiation than that of other yeast cellular proteins. Taking into account the half-life of yeast mRNA, the results obtained indicate that the genes coding for ribosomal proteins form part of long transcriptional units, which are much longer than the DNA stretch needed to code for a ribosomal protein of average molecular weight. Saturation hybridization of total poly(A)-containing mRNA with yeast nuclear DNA revealed that as much as 30% of DNA is complementary to yeast mRNA. Thus, the primary transcript of a protein gene on the average is about 1.7 times the length of the actual messenger. On the basis of the various experimental data we suggest a clustering of the yeast ribosomal protein genes in a number of common transcriptional units.