Search

The ribosomal protein L2 (rpL2) of Saccharomyces cerevisiae regulates the accumulation of its own mRNA by a feedback mechanism. An RNA sequence is responsible for this control, initially characterized as a 360 nucleotide-long region, localized at the 5' end of the transcript. This region, fused to an unrelated coding sequence, is able to down-regulate the accumulation of the chimeric transcript when increased levels of rpL2 are induced in the cell. The target regulatory region also responds to regulation when inserted inside an intron, demonstrating that the control process can take place inside the nucleus. Deletion analysis from the 5' and 3' borders have restricted the responsive region to approximately 200 nt. The insertion of a poly-G cassette downstream of the regulatory region allowed the identification of truncated 3' cut-off poly(A)+ RNA molecules. The parallel identification of cut-off molecules containing the 5' portion of the transcript allowed us to deduce that the truncated products originate by endonucleolytic cleavage. Altogether, these results are consistent with a mechanism by which the presence of excess amounts of rpL2 in the cell triggers its own mRNA to a degradative pathway; this involves an initial endonucleolytic cleavage that is followed by exonucleolytic trimming. Such a regulatory mechanism shows interesting analogies with the translational regulation of r-proteins in Escherichia coli.

The gene encoding ribosomal protein L1 in Xenopus laevis is known to be posttranscriptionally regulated; the third intron can be processed from the pre-mRNA in two alternative ways, resulting either in the production of L1 mRNA or in the release of a small nucleolar RNA (U16). The formation of splicing complexes was studied in vivo by oocyte microinjection. We show that spliceosome assembly is impaired on the L1 third intron and that the low efficiency of the process is due to the presence of suboptimal consensus sequences. An analysis of heterogeneous nuclear ribonucleoprotein (hnRNP) distribution was also performed, revealing a distinct site for hnRNP C binding proximal to the 5' end of the L1 third intron. Cleavage, leading to the production of the small nucleolar RNA U16, occurs in the same position, and we show that conditions under which hnRNP C binding is reduced result in an increase of the processing activity of the intron.

It was recently shown that a new class of small nuclear RNAs is encoded in introns of protein-coding genes and that they originate by processing of the pre-mRNA in which they are contained. Little is known about the mechanism and the factors involved in this new type of processing. The L1 ribosomal protein gene of Xenopus laevis is a well-suited system for studying this phenomenon: several different introns encode for two small nucleolar RNAs (snoRNAs; U16 and U18). In this paper, we analyzed the in vitro processing of these snoRNAs and showed that both are released from the pre-mRNA by a common mechanism: endonucleolytic cleavages convert the pre-mRNA into a precursor snoRNA with 5' and 3' trailer sequences. Subsequently, trimming converts the pre-snoRNAs into mature molecules. Oocyte and HeLa nuclear extracts are able to process X. laevis and human substrates in a similar manner, indicating that the processing of this class of snoRNAs relies on a common and evolutionarily conserved mechanism. In addition, we found that the cleavage activity is strongly enhanced in the presence of Mn2+ ions.

We previously reported that the third intron of the X.laevis L1 ribosomal protein gene encodes for a snoRNA called U16. Here we show that four different introns of the same gene contain another previously uncharacterized snoRNA (U18) which is associated with fibrillarin in the nucleolus and which originates by processing of the pre-mRNA. The pathway of U18 RNA release from the pre-mRNA is the same as the one described for U16: primary endonucleolytic cleavages upstream and downstream of the U18 coding region produce a pre-U18 RNA which is subsequently trimmed to the mature form. Both the gene organization and processing of U18 are conserved in the corresponding genes of X.tropicalis and H.sapiens. The L1 gene thus has a composite structure, highly conserved in evolution, in which sequences coding for a ribosomal protein are intermingled with sequences coding for two different snoRNAs. The nucleolar localization of these different components suggests some common function on ribosome biosynthesis.

We report that the third intron of the L1 ribosomal protein gene of Xenopus laevis encodes a previously uncharacterized small nucleolar RNA that we called U16. This snRNA is not independently transcribed; instead it originates by processing of the pre-mRNA in which it is contained. Its sequence, localization and biosynthesis are phylogenetically conserved: in the corresponding intron of the human L1 ribosomal protein gene a highly homologous region is found which can be released from the pre-mRNA by a mechanism similar to that described for the amphibian U16 RNA. The presence of a snoRNA inside an intron of the L1 ribosomal protein gene and the phylogenetic conservation of this gene arrangement suggest an important regulatory/functional link between these two components.

Sequences corresponding to the third intron of the X.laevis L1 ribosomal protein gene were isolated from the second copy of the X.laevis gene and from the single copy of X.tropicalis. Sequence comparison revealed that the three introns share an unusual sequence conservation which spans a region of 110 nucleotides. In addition, they have the same suboptimal 5' splice sites. The three introns show similar features upon oocyte microinjection: they have very low splicing efficiency and undergo the same site specific cleavages which lead to the accumulation of truncated molecules. Computer analysis and RNAse digestions have allowed to assign to the conserved region a specific secondary structure. Mutational analysis has shown that this structure is important for conferring the cleavage phenotype to these three introns. Competition experiments show that the cleavage phenotype can be prevented by coinjection of excess amounts of homologous sequences.

The splicing of the third intron of the L1 r-protein gene of X.laevis was studied in the heterologous in vitro HeLa nuclear system. Despite the evolutionary distance, the cis-elements responsible for the default process play a similar role in the two organisms. Analysis of the splicing of various mutant substrates showed that the 5' splice site is primarily responsible for the low efficiency of splicing of the third intron. The suboptimal 5' splice site sequence leads to the utilization of an upstream alternative site which corresponds to the one utilized in vivo. The accumulation of splicing intermediates in the in vitro system allowed the identification of the branch site and of the branch consensus sequence. In contrast, the in vivo regulatory mechanism involving cleavage of the pre-mRNA is not mimicked in the HeLa extract.

Splicing of the regulated third intron of the L1 ribosomal protein gene of Xenopus laevis has been studied in vivo by oocyte microinjection of wild-type and mutant SP6 precursor RNAs and in vitro in the heterologous HeLa nuclear extract. We show that two different phenomena combine to produce the peculiar splicing phenotype of this intron. One, which can be defined constitutive, shows the same features in the two systems and leads to the accumulation of spliced mRNA, but in very small amounts. The low efficiency of splicing is due to the presence of a noncanonical 5' splice site which acts in conjunction with sequences present in the 3' portion of the intron. The second leads to the massive conversion of the pre-mRNA into site specific truncated molecules. This has the effect of decreasing the concentration of the pre-mRNA available for splicing. We show that this aberrant cleavage activity occurs only in the in vivo oocyte system and depends on the presence of an intact U1 RNA.

The expression of the yeast L2 r-protein gene is controlled at the level of mRNA accumulation. The product of the gene appears to participate in this regulation by an autogenous feedback mechanism. This control does not operate at the level of transcription but instead affects L2 mRNA accumulation. This autogenous regulation of mRNA accumulation provides an interesting analogy to the autogenous translational regulation of r-proteins in Escherichia coli.

The same factor, ABF1, binds to the promoters of the two gene copies (L2A and L2B) coding for the ribosomal protein L2 in Saccharomyces cerevisiae. In vitro binding experiments and in vivo functional analysis showed that the different affinities of the L2A and L2B promoters for the ABF1 factor are responsible for the differential transcriptional activities of the two gene copies. The presence of ABF1-binding sites in front of many housekeeping genes suggests a general role for ABF1 in the regulation of gene activity.

A specific control regulates, at the level of RNA splicing, the expression of the L1 ribosomal protein gene in Xenopus laevis. Under particular conditions, which can be summarized as an excess of free L1 protein, a precursor RNA which still contains two of the nine introns of the L1 gene accumulates. In addition to the splicing block the two intron regions undergo specific endonucleolytic cleavages which produce abortive truncated molecules. The accumulation of mature L1 RNA therefore results from the regulation of the nuclear stability of its precursor RNA. We propose that a block to splicing can permit the attack of specific intron regions by nucleases which destabilize the pre-mRNA in the nucleus. Therefore the efficiency of splicing could indirectly control the stability of the pre-mRNA.

We have previously constructed and selected six recombinant plasmids containing cDNA sequences specific for different ribosomal proteins of Xenopus laevis (Bozzoni et al., 1981). DNA cloned in these plasmids have been isolated and sequenced. Amino acid sequences of the corresponding portions of the proteins have been derived from DNA sequences; they are arginine- and lysine-rich as expected for ribosomal proteins. One of the cDNA sequences has an open reading frame also on the strand complementary to the one coding for the ribosomal protein; this fragment has inverted repeats twenty nucleotides lone at the two ends. The codon usage for the six sequences appears to be non-random with some differences among the ribosomal proteins analysed.

The activity of antisera against ribonucleoproteins containing U1 small nuclear RNA (Sm and RNP) has been analysed on pol II transcripts in an in vivo system. Xenopus laevis ribosomal protein gene transcripts are accumulated in the form of precursor RNA when either of the two kinds of antisera are injected into the germinal vesicles of X. laevis oocytes before the injection of purified L1 and L14 ribosomal protein genes. No effect on the accumulation of mature histone mRNA is detected when X. laevis histone genes are injected together with the RNP antiserum. These results strongly suggest that U1-RNP complexes play an essential role in intron removal in vivo.

Ribosomal protein L1 is encoded by two genes in Xenopus laevis. The comparison of two cDNA sequences shows that the two L1 gene copies (L1a and L1b) have diverged in many silent sites and very few substitution sites; moreover a small duplication occurred at the very end of the coding region of the L1b gene which thus codes for a product five amino acids longer than that coded by L1a. Quantitatively the divergence between the two L1 genes confirms that a whole genome duplication took place in Xenopus laevis approximately 30 million years ago. A genomic fragment containing one of the two L1 gene copies (L1a), with its nine introns and flanking regions, has been completely sequenced. The 5' end of this gene has been mapped within a 20-pyridimine stretch as already found for other vertebrate ribosomal protein genes. Four of the nine introns have a 60-nucleotide sequence with 80% homology; within this region some boxes, one of which is 16 nucleotides long, are 100% homologous among the four introns. This feature of L1a gene introns is interesting since we have previously shown that the activity of this gene is regulated at a post-transcriptional level and it involves the block of the normal splicing of some intron sequences.

By cross-hybridization with a cDNA probe for the Xenopus laevis ribosomal protein L1 we have been able to isolate the homologous genes from a Saccharomyces cerevisiae genomic library. We have shown that these genes code for a ribosomal protein which was previously named L2. In yeast, like in X. laevis, these genes are present in two copies per haploid genome and, unlike the vertebrate counterpart, they do not contain introns. Amino acid comparison of the X. laevis L1 and S. cerevisiae L2 proteins has shown the presence of a highly conserved protein domain embedded in very divergent sequences. Although these sequences are very poorly homologous, they confer an overall secondary structure and folding highly conserved in the two species.

In Saccharomyces cerevisiae, the genes coding for the ribosomal protein L2 are present in two copies per haploid genome. The two copies, which encode proteins differing in only a few amino acids, contribute unequally to the L2 mRNA pool: the L2A copy makes 72% of the mRNA, while the L2B copy makes only 28%. Disruption of the L2B gene (delta B strain) did not lead to any phenotypic alteration, whereas the inactivation of the L2A copy (delta A strain) produced a slow-growth phenotype associated with decreased accumulation of 60S subunits and ribosomes. No intergenic compensation occurred at the transcriptional level in the disrupted strains; in fact, delta A strains contained reduced levels of L2 mRNA, whereas delta B strains had almost normal levels. The wild-type phenotype was restored in the delta A strains by transformation with extra copies of the intact L2A or L2B gene. As already shown for other duplicated genes (Kim and Warner, J. Mol. Biol. 165:79-89, 1983; Leeret al., Curr. Genet. 9:273-277, 1985), the difference in expression of the two gene copies could be accounted for via differential transcription activity. Sequence comparison of the rpL2 promoter regions has shown the presence of canonical HOMOL1 boxes which are slightly different in the two genes.

Using probes to Xenopus laevis ribosomal-protein (r-protein) mRNAs, we have found that in the oocyte the accumulation of r-protein mRNAs proceeds to a maximum level, which is attained at the onset of vitellogenesis and remains stable thereafter. In the embryo, r-protein mRNA sequences are present at low levels in the cytoplasm during early cleavage (stages 2-5), become undetectable until gastrulation (stage 10) and accumulate progressively afterwards. Normalization of the amount of mRNA to cell number suggests an activation of r-protein genes around stage 10; however, a variation in mRNA turnover cannot be excluded. Newly synthesized ribosomal proteins cannot be found from early cleavage up to stage 26, with the exception of S3, L17 and L31, which are constantly made, and protein L5, which starts to be synthesized around stage 7. A complete set of ribosomal proteins is actively produced only in tailbud embryos (stages 28-32), several hours after the appearance of their mRNAs. Before stage 26 these mRNA sequences are found on subpolysomal fractions, whereas more than 50% of them are associated with polysomes at stage 31. Anucleolate mutants do not synthesize ribosomal proteins at the time when normal embryos do it very actively; nevertheless, they accumulate r-protein mRNAs.

Cloned gene for Xenopus ribosomal protein L1 was injected into fertilized eggs, and its expression was analyzed during the period of embryo development when the mRNAs produced by the endogenous ribosomal protein genes are still silent due to a translational control. The injected genes replicated extensively, and a 10-fold excess of L1 mature transcript accumulated in the embryo. This was accompanied by a small amount of incompletely processed L1 RNA that still contained one out of nine introns, a molecule never observed in normal conditions. The excess mature L1 mRNA was distributed between polysomes and messenger ribonucleoproteins (mRNPs) in the same relative proportion observed in control embryos of the same stage. Therefore, more L1 mRNA was loaded onto polysomes and caused the appearance of L1 protein when this was not yet detectable in control embryos. The results suggest a relationship between the excess amount of L1 protein and the alteration in processing of its transcripts.

Poly-A+ mRNA from Xenopus laevis oocytes, partially enriched for r-protein coding capacity has been used as starting material for preparing a cDNA bank in plasmid pBR322. The clones containing sequences specific for r-proteins have been selected by translation of the complementary mRNAs. Clones for six different r-proteins have been identified and utilized as probes for studying their genomic organization. Two gene copies per haploid genome were found for r-proteins L1, L14, S19, and four-five for protein S1, S8 and L32. Moreover a population polymorphism has been observed for the genomic regions containing sequences for r-protein S1, S8 and L14.

We have studied the regulation of ribosomal protein (r-protein) synthesis in Xenopus anucleolate mutants, which lack the genes for rRNA. The accumulation of mRNA for the two r-proteins analyzed parallels the controls up to stage 30. This mRNA is mobilized onto polysomes and is translated as in normal embryos, but r-proteins are unstable in the absence of rRNA to assemble with. A translational control of rp-mRNA distribution between polysomes and mRNPs is observed, but this is not due to an autogenous regulation by r-proteins. After stage 30 the amount of rp-mRNA declines specifically in the mutants because the transcripts are unstable. Considering the temporal correlation between this event and the onset of r-protein synthesis we suggest that an autogenous control operates at the level of transcript stability.

Studies on ribosome biosynthesis in developing Xenopus oocytes and embryos, and after microinjection of cloned ribosomal-protein genes, have revealed that the synthesis of ribosomal proteins (r-proteins) is controlled by two types of regulation: (1) a post-transcriptional regulation, operated by feedback of the r-proteins themselves, controls processing and stability of r-protein transcripts and thus the amount of the corresponding mRNA present in the cell; and (2) a translational regulation controls the efficiency of utilization of r-protein mRNA (rp-mRNA) in response to the cellular needs for new ribosomes.

The expression of two Xenopus laevis ribosomal protein genes (L1 and L14) has been analysed by microinjection of the cloned genomic sequences into frog oocyte nuclei. While the injection of the L14 gene causes the accumulation of the corresponding protein in large excess with respect to that synthesized endogenously, the L1 gene does not. Analysis of the RNA shows that both genes are actively transcribed. The seven-intron-containing L14 transcript is completely processed to a mature form, while two out of nine intron sequences persist in the L1 transcript. This precursor RNA is confined to the nucleus; its accumulation is due to a specific block of splicing operating at the level of two defined introns and not to saturation of the processing apparatus of the oocyte. The different behaviour of the two genes may reflect different mechanisms of regulation which, in the case of the L1 gene, could operate at the level of splicing.

In the haploid genome of Xenopus laevis there are two genes coding for the r-protein L14. It is not known if they are located on the same chromosome. cDNA clones deriving from the transcripts of the two genes have been isolated from an oocyte messenger cDNA bank showing that they are both expressed. We have studied the structure of one of the L14 genes by Electron Microscopy, restriction mapping and sequencing. An allelic form of the L14 gene was also isolated. It contains a large deletion covering the 5' end region up to the middle of the third intron. The 5' end of the X. laevis L14 gene was compared to that of the corresponding gene in the closely related species X. tropicalis and found to be highly conserved. The L14 gene has multiple initiation sites, but the large majority of the transcripts start in the middle of a pyrimidine tract not preceded by a canonical TATA box as in other eukaryotic housekeeping genes. The X. laevis L1 and L14 genes have a common decanucleotide in the first exon in the same position with regard to the initiator ATG which just precedes the first intron. The decanucleotide shows homology with the X. laevis 18S rRNA.

Whether ribosomal-protein synthesis in Xenopus laevis is autogenously controlled at the translational level as is known to occur in prokaryotes has been studied. For this purpose ribosomal (r) proteins were prepared from X. laevis ribosomal subunits and group fractionated by ion-exchange chromatography. They were then added to an in vitro translation system directed by an oocyte mRNA fraction which contains template activity for r proteins. The synthesized radioactive products were analyzed by 2D gel electrophoresis and compared with controls. Similarly in vivo experiments were performed by microinjection of the fractionated proteins into the cytoplasm of Xenopus oocytes followed by incubation with [35S]methionine for different times. In all the experiments no evident effect of r proteins on the translation of their own mRNA was observed.

The sequence analysis of the L1 ribosomal protein (r-protein) gene of Xenopus laevis has revealed a strong homology in four out of the nine introns of the gene; this homology region spans 60 nucleotides (nt) with 80% homology [Loreni et al., EMBO J. 4 (1985) 3483-3488]. We have extended our analysis to X. tropicalis, a species which is closely related to X. laevis. Partial sequencing of the isolated L1 gene has revealed that these 60-nt homology regions are also present in at least two introns of the X. tropicalis L1 gene. Computer analysis has revealed that perfect nt sequence complementarity exists between 13 nt of this intron region and the 28S ribosomal RNA in a region which is conserved in all eukaryotes, suggesting a possible base-pairing interaction between these two sequences.