Structure and expression of ribosomal protein genes

Structure and expression of ribosomal protein genes in Xenopus laevis

Biochem. Cell Biol. Downloaded from www.nrcresearchpress.com by Guangzhou Jinan University on 06/04/13 For personal use only.

Francesco Amaldi, Olga Carnacho-Vanegas, Beatrice Cardinali, Francesco Cecconi, Claudia Crosio, Fabrizio Loreni, Paolo Mariottini, Livio Pelliuoni, and Paola Pierandrei-Amaldi

Abstract:In Xenopus laevis, as well as in other vertebrates, ribosomal proteins (r-proteins) are coded by a class of genes that share some organizational and structural features. One of these, also common to genes coding for other proteins involved in the translation apparatus synthesis and function, is the presence within their introns of sequences coding for small nucleolar RNAs. Another feature is the presence of common structures, mainly in the regions surrounding the 5' ends, involved in their coregulated expression. This is attained at various regulatory levels: transcriptional, posttranscriptional, and translational. Particular attention is given here to regulation at the translational level, which has been studied during Xenopus oogenesis and embryogenesis and also during nutritional changes of Xenopus cultured cells. This regulation, which responds to the cellular need for new ribosomes, operates by changing the fraction of rp-mRNA (ribosomal protein rnRNA) engaged on polysomes. A typical 5' untranslated region characterizing all vertebrate rp-mRNAs analyzed to date is responsible for this translational behaviour: it is always short and starts with an 8-12 nucleotide polypyrimidine tract. This region binds in vitro some proteins that can represent putative trans-acting factors for this translational regulation. Key words: ribosomal proteins, snoRNA, translational regulation, Xenopus laevis.

Resum6 : Chez Xenopus laevis et d'autres vertkbrks, les proteines ribosomiques sont codees par une famille de gbnes ayant des caract6ristiques structurales et organisationnelles communes. Une de ces caractkristiques, Cgalement partagte par les gbnes codant d'autres prottines intervenant dans la synthbse et la fonction du mtcanisme de traduction, est la presence de sequences codant les petits ARN nuclColaires dans les introns de ces gbnes. Une autre caractkristique est la presence de structures communes, principalement dans les regions prbs des extrtmitts 5', intervenant dans la rkgulation coordonnCe de leur expression. La regulation de cette expression se fait aux niveaux transcriptionnel, post-transcriptionnel et traductionnel. Dans cet article, une attention particulibe est portke B la regulation de la traduction durant l'ovogenbse et l'embryogenbse chez Xenopus et Cgalement aprbs un changement du milieu nutritif des cellules de Xenopus en culture. Cette rtgulation, consCcutiveau besoin de nouveaux ribosomes dans les cellules, s'effectue par une modification de la fraction d' ARNm des protkines ribosomiques sur les polyribosomes. Ce comportement au cours de la traduction est attribuable ii une region non traduite B I'extrtmitC 5', caractkristique de tous les ARNm des protkines ribososrniques des verttbrks ktudiks jusqu'k maintenant : cette rCgion est toujours courte et dCbute par une sCquence de 8 21 12 nuclCotides pyrimidiques. Des protkines, possiblement des facteurs trans intervenant dans la regulation de la traduction, se lient B cette rkgion in vitro. Mots clis : protkines ribosomiques, petit ARN nuclkolaire, regulation, traduction, Xenopus laevis. [Traduit par la rkdaction]

Received May 20, 1995. Accepted August 18, 1995.

Abbreviations:r-proteins, ribosomal proteins; rp-mRNA, ribosomal protein mRNA; S'UTR, 5' untranslated region; bp, base pair(s); nt, nucleotide(s), 3'UTR, 3' untranslated region; rp-genes, ribosomal protein genes; snoRNA, small nucleolar RNA; ETS, external transcribed spacer; RNP, ribonuclear protein. F. ~maldi,'0.Camacho-Vanegas,F. Cecconi, and F. Loreni. Dipartimento di Biologia, UniversitB di Roma "Tor Vergata", 00133 Roma, Italy. B. Cardinali, C. Crosio, L. Pellizzoni, and P. Pierandrei-Amaldi.Istituto di Biologia Cellulare, C.N.R., 00137 Roma, Italy. P. Mariottini. Dipartimento di Biologia, Terza Universiti di Roma, 00154 Roma, Italy. Author to whom all correspondence should be addressed. Biochem. Cell Biol. 73: 969-977 (1995). Printed in Canada I Imprim6 au Canada

Biochem. Cell Biol. Vol. 73, 1995


Introduction The frog Xenopus laevis has proved to be very useful for the study of ribosome synthesis regulation in higher eukaryotes. This is partly because of the possibility of studying the regulation of ribosome synthesis during oogenesis, when it is extremely high, and during embryogenesis, when it resumes after a period of inactivity, and partly because of the possibility of interfering with the normal situation by microinjection of DNA, RNA, proteins, and antibodies, etc., into oocytes and fertilized eggs. Furthermore, the Xenopus anucleolate mutant provides the unique opportunity to study the synthesis of rproteins in the absence of rRNA production. The availability of the X. laevis cell line obviates the practical limits presented by the Xenopus embryo system for some kinds of experiments such as, for instance, those involving administration of drugs or in vivo labeling with radioactive precursors or attempts to modify cell growth conditions.

Nomenclature of ribosomal proteins In our previous papers, some r-proteins have been named according to the numbering system introduced in our first study of Xenopus r-proteins (Pierandrei-Amaldi and Beccari 1981). Considering that many more r-proteins have been sequenced by now in several systems and that, in general, homologous r-proteins are easily identified, we think it is appropriate to use a unified nomenclature. Accordingly, we now adopt the rat system as a standard (Wool et al. 1990) and change the previous names as indicated in Table 1. For instance, the X. laevis r-protein previously indicated as L1 is now identified as XL4, meaning that it is the Xenopus r-proteins homologous to the rat r-protein LA.

Organization and structure of rmprotein genes The genes coding for r-proteins are generally present in one or two copies in eukaryotic genomes. In the specific case of X. laevis, there are two gene copies for most of the r-proteins studied, namely X U , XL18, XS7, and XS3 (Loreni et al. 1985; Beccari et al. 1986; Mariottini et al. 1988; Di Cristina et al. 1991), due to a duplication of the whole genome, which occurred in this species about 30 million years ago (Bisbee et al. 1977). Up to now, four Xenopus r-protein genes have been completely isolated and sequenced: X U , XL18, XS3, and XS7 (Loreni et al. 1985, Beccari and Mazzetti 1987; Pellizzoni et al. 1995; Mariottini et al. 1993). For another gene, encoding r-protein XS24, the 5' portion with the promoter region is available (Mariottini and Amaldi 1990). Comparison of the structure of these genes reveals a common architecture that reflects the fact that they belong to a class of genes sharing common regulations at the transcriptional and translational level. Figure 1 compares the 5' ends of these genes, showing that the transcription start site is always located within a stretch of 12-20 pyrimidines flanked by short C+G-rich sequences. Similar transcription start sites have also been observed in mouse r-protein genes and, more generally, in several housekeeping genes in vertebrates (Wagner and Perry 1985). In all cases, the first intron is localized exactly after the ini-

Table 1. Nomenclature of Xenopus r-proteins.

Original namesa

New adopted namesb

'Numbering according to Pierandrei-Amaldi and Beccari (1981). %umbering according to the rat system (Wool et al. 1990).

tiation ATG codon, as in the genes for XL4, XL18, and XS24, or very close to it, as in the genes for r-proteins XS7 and XS3, thus separating the 5'UTR from the coding portion of the gene. The first exons, which code for the 5UTR of the mRNAs, are always quite short, i.e., between 35 and 50 bp (Fig. 1). Besides starting with a run of 8-12 pyrimidines, as a result of the above mentioned initiation in the middle of a run of pyrimidines, they display other sequence similarities (Mariottini et al. 1988). Similarities are also observed in the 100 nucleotides both upstream and downstream in the first intron. It is possible that this last region is involved in the transcription of this class of genes, as it has been shown for two mouse r-protein genes (Moura-Neto et al. 1989). The 3'UTRs of the Xenopus rp-genes analyzed also present some common features: they are particularly short, 40-60 nucleotides, and display sequence similarities among themselves (Mariottini et al., 1988). This region is probably involved in the control of transcript stability. For instance, it has been shown that at the end of oogenesis, during oocyte maturation, rp-mRNAs are rapidly degraded, which is due to the absence of specific stabilizing sequences in the 3UTR (Varnum and Wormington 1990).

The introns of r-protein genes often encode snoRNAs The Xenopus genes for r-proteins XL18, XS3, and XS7 contain six introns each, and the gene for r-protein X U contains nine. The most remarkable property of the introns of some rprotein genes is the presence of evolutionarily conserved sequences, often repeated in other introns of the same gene (Loreni et al. 1985; Mariottini et al. 1990), that code for snoRNAs as indicated in Table 2. Thus, Bozzoni and collaborators have shown that the Xenopus gene for r-protein X U contains, in its third intron, a 106-nt sequence that codes for U16 RNA (Fragapane et al. 1993) and a 60-nt sequence repeated in introns 2, 4, 7, and 8 (Loreni et al. 1985; Cutruzzolh et al. 1986) that codes for U18 RNA (Prislei et al. 1993). Similarly, the six introns of the Xenopus gene for r-protein XS7 contain a 220-nt repeated sequence coding for U17 RNA (Mariottini et al. 1993; Cecconi et al. 1994), and introns 3, 5, and 6 of the Xenopus gene for r-protein XS3 code for U15 RNA (Pellizzoni et al. 1994). At present, no snoRNA coding sequence has been identified in the six introns of the gene for r-protein XL18. Thus, four snoRNAs have been found to be

Amaldi et al.


Fig. 1. Nucleotide sequences surrounding the 5' end of the genes for different X. laevis r-proteins. Sequences are shown from position -30 relative to the cap site (arrow). The polypyrimidine tracts are underlined with a solid line, and the flanking GC-rich tracts are underlined with a dotted line. The initiation AUG codons are boxed and the positions of the first introns are indicated. References are as follows: XL4 (Loreni et al. 1985); XL18 (Beccari and Mazzetti 1987); XS3 (Pellizzoni et al. 1995); XS7 (Mariottini et al. 1993); XS24 (Mariottini and Amaldi 1990).

Table 2. Sequenced Xenopus r-protein genes and their intron encoded snoRNAs. snoRNA Host gene Encoded r-protein

XS3 XS7 XL4

XL18

Intron no.a

1-2-3-4-5-6 1-2-3-45-6 1-2-3-4-5-6-7-8-9 1-2-3-4-5-6-7-8-9 1-2-3-4-5-6

Name

Length (nt)

U15 U17 U16 U 18

150 220 116 60

No. boxes complemenatry to rRNA

2 2

1

Ref. Pellizzoni et al. 1994 Cecconi et al. 1994 Fragapane et al. 1994 Prislei et al. 1994

-

aIntrons that contain snoRNA coding sequences are bolded.

encoded in the introns of three of the four r-protein genes sequenced to date. This suggests that the number of the intron encoded snoRNAs might be as high as 5Cb100. Intron-encoded snoRNAs have also been described in other vertebrates. They have been found in r-protein genes or in genes for other proteins somehow related to ribosome synthesis or function (see reviews Sollner-Webb 1993 and Filipowicz and Kiss 1993). In general, a good conservation is observed among the various snoRNA sequence copies in a given species and also among different vertebrate species. When the primary sequence is not well conserved, as in the case of the very poor conservation of U15 snoRNA between Xenopus and man and among the various Xenopus copies, homology is still present at the level of secondary structure (Pellizzoni et al. 1994). In some cases, the snoRNA coding sequences are found in different introns in the same gene of different species, for example, U15 RNA is encoded in introns 3, 5, and 6 of the Xenopus XS3 gene (Pellizzoni et al. 1994) but is present in the first and possibly other introns of the human S3 gene (Tycowski et al. 1993). In other cases, the snoRNA coding sequences are found in the introns of different genes in different species, for example, U17 RNA, which is encoded in the gene for r-protein XS7 in Xenopus (Cecconi et al. 1994) but is present in the introns of the RCCl gene in mammals (Kiss and Filipowicz 1993). Three of the four Xenopus snoRNAs studied, U15 (Pellizzoni et al. 1994), U16 (Fragapane et al. 1993), and U18 (Prislei et al. 1993), contain the C and D motifs typical of snoRNAs interacting with the nucleolar protein fibrillarin (Tyc and Steitz 1989), whereas U17 does not (Cecconi et al. 1994).

The intron-encoded snoRNAs are not produced by independent transcription but by processing the intron sequences of the host gene transcripts (see reviews Sollner-Webb 1993 and Filipowicz and Kiss 1993). In the case of Xenopus U16 RNA, it has been proposed that the processing pathways of the r-protein gene transcript are alternative: either the primary transcript is spliced to produce r-protein XL4 mRNA or it is processed to produce the snoRNAs (Fragapane et al. 1993). For other snoRNAs, no such evidence has been obtained. An intriguing structural feature common to many of the intron-encoded snoRNAs is the presence of sequences complementq to regions of rRNA transcripts (Table 2). Of the four Xenopus intron-encoded snoRNAs, U17 contains a 12-nt sequence that is perfectly complementary to an 18s rRNA region, and also a 9+7-nt complementarity, with a 15-nt loop, to a 16-nt sequence in the ETS of the Xenopus rRNA precursor (Cecconi et al. 1994). U 18 snoRNA contains a 13-nt sequence complementary to a region of 28s RNA (Cutruzzoli et al. 1986) and U15 RNA contains two sequences, 9 and 10 nt in length, that are complementary to two regions of 28s rRNA (Pellizzoni et al. 1994). Although the precise functional role has not been determined for any one of these snoRNAs, it seems reasonable to think that they are involved in some step of rRNA processing and (or) ribosome assembly. Depletion experiments carried out with antisense oligonucleotides against the U22 snoRNA in Xenopus have demonstratedthat this RNA is involved in the maturation of 18s ribosomal RNA (Tycowski et al. 1994). The almost general localization of snoRNA coding sequences within the introns of genes coding for ribosomal or nucleolar proteins also supports the assumption that this pecu-

Biochem. Cell Biol. Vol. 73, 1995

liar gene organization is not coincidental but must have an important functional relevance.


Ribosomal protein production is regulated at various levels The synthesis of r-proteins has been studied during Xenopus oogenesis and embryogenesis by following the activity of rprotein genes at the various levels of transcription, transcript stability, translation, and protein stability (Pierandrei-Amaldi et al. 1982, 1985a; Baum and Wormington 1985). The results obtained indicated that r-protein production is controlled at various levels proceeding from transcription to protein stability. These levels are as follows. First, studies on promoters and on transcription of Xenopus r-protein genes (Carnevali et al. 1989; Marchioni et al. 1993) have revealed that they share a common architecture with mammalian r-protein genes and that they are recognized by transcription factors active on these and other housekeeping genes (Hariharan et al. 1989 and Hariharan and Perry 1990 and references therein). In general, r-protein genes are always transcriptionally active even if ribosome synthesis is not required, as in early embryogenesis when transcription is resumed, or in conditions in which ribosomes cannot be synthesized, as in anucleolate embryos. However, a modulation of transcription activity is necessary to adjust the production of ribosomes to the different growth and differentiation states of the cell. Recently, a model has been proposed for the transcriptional regulation of the human r-protein S14 gene that involves two stable RNAs transcribed from the antisense strand of the first intron of the r-protein S14 gene (Tasheva and Roufa 1995). Second, posttranscriptional controls that regulate the level of r-protein mRNA in the cell have been observed during Xenopus embryogenesis (Pierandrei-Amaldiet al. 1985a) and during oocyte maturation (Varnum and Wormington 1990). Moreover, a unique posttranscriptional regulation has been described for r-protein XL4 (Bozzoni et al. 1984; Caffarelli et al. 1987; Fragapane et al. 1992). This involves the modulation of a specific step of the splicing of the third intron of X U transcript (the same intron that, as mentioned above, codes for the U16 snoRNA). Third, a translational regulation also operates in controlling the partition of rp-mRNA between mRNPs and polysomes (references in Amaldi and Pierandrei-Amaldi 1990). This regulation, also observed and studied in mammalian systems (references in Perry and Meyuhas 1990), represents a very fast response to changing amounts of available unutilized ribosomes and will be considered in more detail in the rest of this paper. Fourth, under certain conditions, r-proteins can be synthesized in excess of the amount needed for ribosome assembly; in such cases, they are degraded (Pierandrei-Amaldi et al. 1985a; Bowman 1987; Baum et al. 1988).

The translational control of r-protein synthesis In Xenopus, the first evidence of translational regulation of rprotein synthesis was obtained by studying the events that lead to the onset of the synthesis of new ribosomes in the embryo (Pierandrei-Amaldi et al. 1982, 1985a; Baum and Worming-

ton 1985; Wormington 1989). In fact, new mRNA specific for r-proteins (rp-mRNA) starts to be accumulated after the socalled mid-blastula transition (stage 8) as occurs for many other mRNAs. At the beginning, however, the newly transcribed rp-mRNA is poorly translated, being largely retained in light cytoplasmic RNPs, whereas only 20-30% is loaded on polysomes. After stage 26, about 20 h later, the fraction of rpmRNAs associated with polysomes increases to about 6580%, and appreciable amounts of new r-proteins begin to be synthesized and accumulated. An uncoupling between the synthesis of rp-mRNA and its use has also been observed during Xenopus oogenesis (Baum and Wormington 1985; Cardinali et al. 1987). To study some aspects of this translational regulation, we used the X. laevis kidney cell line B 3.2 in which DNA, RNA, and protein synthesis decrease rapidly after a downshift to serum-free medium (Loreni and Amaldi 1992). Addition of serum to the culture after a few hours of deprivation causes a rapid recovery. During these growth rate changes, we observed a shift of rp-mRNA distribution between polysomes and mRNPs (Fig. 2). The percentage of mRNA on polysomes for the four r-proteins analyzed changes from 70-80% during rapid growth to 25-35% during the downshift and back to 7080% after the upshift. Changes of the distribution of rp-mRNA between polysomes and mRNPs have also been observed in various mammalian systems following growth rate changes: mouse fibroblasts after serum starvation (Geyer et al. 1982), mouse lymphosarcoma cells after dexamethasone treatment (Meyuhas et al. 1987), and mouse myoblasts during differentiation (Agrawal and Bowman 1987). Thus, it is now well established that control at the translational level is an important mechanism of r-protein synthesis regulation in all higher eukaryotes. A general feature of this control is that the fraction of rp-mRNA associated with polysomes, and therefore translationally active, is regulated according to the protein synthesis activity of the cell. Since protein synthesis activity is generally dependent on cell growth conditions, rp-mRNAs are loaded mostly on polysomes in rapidly growing cells and stored mostly as inactive rnRNPs in resting cells. In Xenopus B 3.2 cells, we have demonstrated a reversible transfer of rp-mRNAs between polysomes and mRNPs without alteration of messenger stability (Loreni and Amaldi 1992). This rp-mRNA relocation between the two compartments is very fast. In fact, after only 5 min of serum deprivation, the change in distribution is detectable and is complete within 40 min (Loreni and Amaldi 1992). During the upshift, the change is slightly slower. The rapidity and reversibility of the regulation suggest two considerations. The first is that the mechanism is probably based, at least at the beginning, on posttranslational modifications, since the synthesis of new transcripts or even new proteins would require a longer period of time. The second is that this kind of regulation is very advantageous in the case of rapid adaptations to extracellular stimuli, even though it can be used during terminal differentiation (Aloni et al. 1992; Agrawal and Bowman 1987). Another question concerning rp-mRNA translational control is its specificity. As stated above, there is a correlation between general protein synthesis activity and the fraction of polysome-associated rp-mRNAs. However, the changes in mRNA localization appear to specifically involve only rpmRNAs. In fact, during nutritional shifts of B 3.2 cells (Loreni


Amaldi et al. Fig. 2. Polysome-mRNP distribution of rp-mRNA in Xenopus cultured cells during growth rate changes induced by serum deprivation and readdition (Loreni and Amaldi 1992). (A) Examples of optical density profiles of sucrose gradients of cytoplasmic extracts prepared from exponentially growing and down-shifted cells. (B) Dot blot analysis of the RNA extracted from sucrose gradient fractions of cytoplasmic extracts prepared from cells in the following growth conditions: exponential growth, 1.5 h and 3 h of down-shift, and 0.5 hand 1 h of up-shift. Hybridization was performed with probes for transcripts for the genes for r-protein XS24 and for calmodulin (calmod.) as a control.

+sedlrnentation -

A

polysomes

I mRNPs E

1.5 down I

B

4 n

*

.

.

.

.

.

a

-

-

-

.

1

1 2 3 4 56789101112 Fraction number

growing 3 h down 0.5 h up I~ U D

.+++@+OWm

*W++*rcsn @*@+Q&*'~

3 h down 0.5hup I~ U D

w* 4 $ :

P

Calmod. *a ~ . b mRNA

brt*&q@B a r we*,+

and Amaldi 1992), as well as in other experimental systems, unrelated control mRNAs do not show any alteration of polysoma1 association. A few exceptions to this observation have been recently described. For instance, EF-la mRNA is translationally regulated like an rp-mRNA both in Xenopus embryogenesis and in B 3.2 cells (Loreni et al. 1993), but of course it is not surprising that translation factors would be coordinately regulated with r-proteins. From the analysis of the data obtained in Xenopus, as well as in other vertebrate systems, we could hypothesize the existence of a feedback mechanism in the translational control of rp-mRNAs. However, a direct autogenous control, similar to the one described in prokaryotes (see reference in Nomura 1990), can be ruled out. In fact, purified r-proteins, individually or in small groups, had no inhibitory effect on the synthesis of new r-proteins when microinjected into the cytoplasm of Xenopus oocytes or added in an in vitro translation system programmed with a Xenopus mRNA (Pierandrei-Amaldi et al.

1985b). Moreover, in Xenopus anucleolate mutants, in spite of the absence of rRNA synthesis, rp-mRNAs are normally synthesized and, after stage 26, translated as in controls. The unutilized r-proteins, which do not find rFWA for assembly, are degraded with a half-life of about 1 h, but do not interfere with further translation of their own mRNAs (PierandreiArnaldi et al. 1985~).A similar conclusion has been reached for mammalian cells, in which overexpression of r-proteins in transfected mouse myoblasts was found to cause rapid degradation of excess proteins without affecting rp-mRNA translation (Bowman 1987). An alternative feedback mechanism could be a regulation mediated by nontranslating free ribosomes present in the cell. In fact, in most of the experimental systems analyzed, one can observe an inverse proportionality between the amount of free ribosomes present in the cell and the percentage of rp-mRNA loaded on polysomes. Moreover, the amount of maternal free ribosomes in developing Xenopus embryos has been experimentally modified (Pierandrei-Amaldi et al. 1991). An increase was obtained by microinjection of purified ribosomes into fertilized eggs and a decrease was induced by treatment of embryos with low doses of cycloheximide, which caused a build-up of polysomes and thus a reduction in the pool of free subunits and monosomes. These experimental alterations in the amount of free ribosomes showed that increased amounts of ribosomes inhibit rp-mRNAs translation, whereas a ribosome depletion increases it. Finally, during the kinetic analysis of changes in rp-mRNA localization in B 3.2 cells, we have examined the disribution of ribosome in polysomes and in free monomers (Loreni and Amaldi 1992). The results show that, at least during the upshift, the decrease of free ribosomes precedes rp-mRNA localization change and, therefore, could be the cause of the polysomal recruitment. The contribution of components of the translation apparatus to rp-mRNA translational control is still an open issue. In fact, a general model of translational regulation proposed by Lodish (1974) is based on the differential affinity of mRNAs for a component of the translation apparatus. According to this model, when the amount or the activity of this component is limiting, low affinity mRNAs would be selectively excluded from polysomes. In the last few years, two candidates have been proposed as possible determinants of rp-mRNA translational control. One is initiation factor 4E (eIF-4E), which is considered a limiting component of the initiation event; it is phosphorylated and its overexpression can induce cell transformation (Lazaris-Karatzas et al. 1990). The fact that eIF-4E phosphorylation shows a correlation with rp-mRNA polysomal localization has been used as evidence for its involvement in translational control (Kaspar et al. 1990). The second candidate is ribosomal protein S6, which is also phosphorylated at multiple sites. The phosphorylation status of S6 correlates with general protein synthesis activity of the cell and it has recently been shown to also correlate with changes in rpmRNA distribution (Jefferies et al. 1994). We can make two considerations regarding the possibility of involvement of these components of the translational apparatus in rp-mRNA translational control. The first, as indicated above, concerns the clear correlation between protein synthesis activity of the cell and rp-mRNA distribution (polysomal versus rnRNPs). It is not surprising to find that a modification of the activity of key components of protein synthesis appara-


Biochem. Cell Biol. Vol. 73. 1995 Fig. 3. Nucleotide sequence of the S'UTRs of mRNAs (cDNAs) for different X. laevis r-proteins. Sequences are shown from the cap site (+ 1 nt) to the initiation AUG codon (boxed). The typical

Fig. 4. Schematic representation of the localization, within the S'UTR of XL4 mRNA, of the binding sites for proteins A and C and their cleavage products B and D (from Cardinali et al. (1993)

polypyrimidine tracts are underlined. For references, see Amaldi and Pierandrei-Amaldi (1990).

and unpublished results).

sequence at the very 5' end of the mRNA for r-protein S16 is necessary, although probably not sufficient, for translational regulation in the mouse system (Levy et al. 1991).

tus is associated with a change in rp-mRNA distribution. A much tougher task will be to show a causal relationship. The second consideration is that, in the systems analyzed in detail, rp-mRNA distribution is always bimodal, either fully loaded with ribosomes or on mRNPs (Meyuhas et al. 1987; Amaldi and Pierandrei-Amaldi 1990). The fact that we never observe rp-mRNA on small polysomes is more consistent with a specific modification of the messenger (e.g., a binding factor) that makes it translatable or not. In fact, in the case of alteration of a component of the translational apparatus, we would expect a statistical decrease in the initiation events and, therefore, in the number of ribosomes interacting with rp-mRNA.

The 5' UTR as cis-acting elements of the translational control To understand the molecular mechanism involved in translational regulation, it is important to identify the cis-acting elements and the trans-acting factors that, analogously to the better understood transcription control mechanisms, are responsible for this specific control. A number of observations have suggested that the 5'UTR of rp-mRNAs might contain the cis-acting elements, sequence(s) or structure(s), responsible for the typical translational behaviour. In fact, as shown in Fig. 3, the 5'UTRs of the mRNAs for the different r-proteins in Xenopus and in other vertebrates, share conserved features: they are always short, starting with a run of 8-12 pyrimidines, and they contain other similarities (Mariottini et al. 1988; Amaldi and Pierandrei-Amaldi 1990; Hariharan et al. 1989). Also, the observation that all r-protein genes retain a first intron localized exactly at, or very close to, the initiation ATG codon supports the idea that these untranslated sequences might have some important role in the regulation of translation. To verify this hypothesis, Xenopus fertilized eggs were microinjected with a gene construct in which the upstream sequences (promoter) and the 5'UTR (first exon) of the gene coding for Xenopus r-protein XS24 had been joined to the coding portion of a reporter gene deprived of its own 5'UTR (Mariottini and Amaldi 1990).The analysis of the expression of this fused gene during embryogenesis has demonstrated that the XS24 5'UTR has conferred to the fused mRNA the translation behaviour typical of rp-mRNAs. That the 5'UTRs of rpmRNAs have a role in their translational control has also been shown with analogous experiments carried out in the mouse (Levy et al. 1991; Hammond et al. 1991) and in the Dictyostelium (Steel and Jacobson 1991) systems. In addition, sitedirected mutagenesis was used to show that the pyrimidine

Interaction of proteins with the SUTR of rp-mRNA as putative mediators of translational regulation It is possible that common factors, interacting with the similar responsive sequences in the 5'UTR of the rp-mRNAs, may cooperate in the translational control, resulting in the coordinate synthesis of the r-proteins. To verify this hypothesis, we have characterized the particles containing the translationally repressed rp-mRNA in vivo and have analyzed the proteins that specifically bind the 5'UTR of this RNA in vitro (Cardinali et al. 1993). The RNA used in the experiment was mRNA for r-protein L1. Sedimentationanalysis and isopycnic centrifugation studies carried out with embryo extracts have shown that the repressed rp-mRNAs are associated in slow sedimenting particles containing more RNA than protein (2.3: 1). When the particle is reconstituted in vitro by incubating mRNA and cytoplasmic extracts, a similar density was observed in isopycnic gradients. To gain information about the proteins associated with XL4 mRNA, a fine analysis of in vitro RNA-protein complex formation was performed paying particular attention to the 5'UTR (Cardinali et al. 1993). In vitro binding experiments between XL4 mRNA and embryo cytoplasmic extracts followed by gel shift and UV-cross-linking experiments have shown the interaction of some proteins with the 5'UTR. Competition experiments have confirmed that the observed binding is specific. Four proteins were found to interact with this region. The binding site of two of them, A (57 kDa) and B (45 m a ) , is located in the pyrimidine tract common to all rpmRNAs. Proteins of the same size also interact with the analogous regions of r-protein XS3 and XL18. The other two proteins, C (31 kDa) and D (24 m a ) , bind in the downstream region and the interaction seems to be affected by an alternative conformation of the RNA. One can speculate that the mRNA can assume alternative conformations according to different translational situations. Recent results, based on peptide mapping analysis, have shown that proteins B and D are cleavage products of A and C, respectively, but still maintain binding capacity (manuscript in preparation). Thus, as schematically shown in Fig. 4, two proteins bind specifically to the 5'UTR of XL4 mRNA. A protein of 56 kDa, possibly homologous to protein A, has been described to bind the pyrimidine tract of mouse L32 mRNA (Kaspar et al. 1992). Although the interaction of these proteins with rp-mRNA is demonstrated, their role in translational control is not yet clear. It is also possible that they are constantly associated with the regulatory region of the mRNA, thus facilitating the action of other factors. To clone their genes and permit functional experiments,

Amaldi et al.

purification of the proteins by chromatographic and affinity methods is in progress.


Anucleolate Xenopus embryos as models for human ribosome deficiencies A recessive lethal mutation in which the nucleolus is absent was described in X. laevis by Elsdale et al. (1958). Frogs homozygous for the mutation (0-nu) are unable to synthesize 18s and 28s rRNAs and die at the end of embryogenesis; in fact, they survive to the swimming tadpole stage by virtue of the large number of ribosomes provided by the egg (Brown and Gurdon 1964). The defect is due to a deletion of the reiterated rRNA genes (Wallace and Birnstiel 1966); less than 1% of the normal amount of rDNA is found in the density gradient region where rDNA normally bands (Brown and Weber 1968). The residual rDNA in the anucleolate embryos appears to include a few normal repeats together with a variety of unusual fragments containing either spacer or gene sequences (Steele et al. 1984). The anucleolate mutation of Xenopus was useful in proving that the nucleolus organizer is the site of the genes encoding the 18s and 28s rRNAs and also in studying the coordinate regulation of the various components of the ribosome, 5 s RNA (Brown 1967; Miller 1973), and r-proteins (Pierandrei-Amaldiet al. 1982, 1985a) during development. It has been found that r-protein genes are normally activated after the mid-blastula transition and, as in normal embryos, the rp-mRNA produced is scarcely translated up to stage 26 and is translated more efficiently at later stages. However, the synthesized proteins, lacking rRNA for assembly, are rapidly degraded. At the end of embryogenesis, when the maternal ribosomal store is completely depleted but before death, the full complement of rp-mRNA becomes completely loaded on polysomes (Pierandrei-Amaldiet al. 1985a). The existence of the anucleolate mutant in Xenopus and the occurrence of a similar situation in Drosophila (Ritossa et al. 1966) suggest that analogous ribosome deficiencies may also occur in humans. TheXenopus 0-nu mutant thus provides a useful animal model that can help to identify a possible human phenoptype and to set up a test for the analysis of this deficiency.

Acknowledgements Our research was supported by grants from the Commission of the European Communities (contract SCl*-0259), from Progetto Finalizzato Biotecnologie and Progetto Finalizzato Ingegneria Genetica of CNR, and from Telethon-Italy (grant E.02). B. Cardinali and 0 . Camacho-Vanegas are recipients of UNIDO-ICGEB fellowships.

References Agrawal, A.G., and Bowman, L.H. 1987. Transcriptional and translational regulation of ribosomal protein formation during mouse myoblast differentiation. J. Biol. Chem. 262: 48684875. Aloni, R., Peleg, D., and Meyuhas, 0. 1992. Selective translational control and nonspecific posttranscriptional regulation of ribosomal protein gene expression during development and regeneration of rat liver. Mol. Cell. Biol. 12: 2203-2212. Amaldi, F., and Pierandrei-Amaldi, P. 1990. Translational regulation of the expression of ribosomal protein genes in Xenopus laevis. Enzyme, 44: 93-105.

Baum, E.Z., and Wormington, W.M. 1985. Coordinate expression of r-protein genes during Xenopus development. Dev. Biol. 111: 488498. Baum, E.Z., Hyman, L.E., and Wormington, W.M. 1988. Post-translational control of ribosomal protein L1 accumulation in Xenopus oocytes. Dev. Biol. 126: 141-146. Beccari, E., and Mazzetti, P. 1987. The nucleotide sequence of the ribosomal protein L14 gene of Xenopus laevis. Nucleic Acids Res. 15: 1870-1872. Beccari, E., Mazzetti, P., Mileo, A.M., Bozzoni, I., PierandreiAmaldi, P., and Amaldi, F. 1986. Sequences coding for the ribosomal protein L14 in Xenopus laevis and Xenopus tropicalis: homologies in the 5' non translated region are shared with other r-protein mRNAs. Nucleic Acids Res. 14: 7633-7646. Bisbee, C.A., Baker, M.A., Wilson, A.C., Hadji-Azimi, I., and Fishberg, M. 1977. Albumin phylogeny for clawed frogs (Xenopus). Science (Washington, D.C.), 195: 785-787. Bowman, L.H. 1987. The synthesis of ribosomal proteins S16 and L32 is not autogenously regulated during mouse myoblast differentiation. Mol. Cell. Biol. 7: 4464-4471. Bozzoni, I., Fragapane, P., Annesi, F., Pierandrei-Amaldi, P., Amaldi, F., and Beccari, E. 1984. Expression of two Xenopus laevis ribosomal protein genes in injected frog oocytes. A specific splicing block interferes with the L1 RNA maturation. J. Mol. Biol. 180: 987-1005. Brown, D.D. 1967. The genes for ribosomal RNA and their transcription during amphibian development. Curr. Top. Dev. Biol. 2: 47-73. Brown, D.D., and Gurdon, J.B. 1964. Absence of rRNA synthesis in the anucleolate mutant of Xenopus laevis. Proc. Natl. Acad. Sci. U.S.A. 51: 139-146. Brown, D.D., and Weber, C.S. 1968. Gene linkage by RNA-DNA hybridization. I. Unique DNA sequences homologous to 4 s RNA, 5 s RNA, and ribosomal RNA. J. Mol. Biol. 34: 661680. Caffarelli, E., Fragapane P., Gehring C., and Bozzoni, I. 1987. The accumulation of mature RNA for the Xenopus laevis ribosomal protein L1 is controlled at the level of splicing and turnover of the precursor RNA. EMBO J. 6: 3493-3498. Cardinali, B., Campioni, N., and Pierandrei-Amaldi, P. 1987. Ribosomal protein, histone and calmodulin mRNAs are differently regulated at the translational level during oogenesis of Xenopus laevis. Exp. Cell Res. 169: 4 3 2 4 1 . Cardinali, B., Di Cristina, M., and Pierandrei-Amaldi. P. 1993. Interaction of proteins with the mRNA for ribosomal protein L1 in Xenopus: structural characterization of in vivo complexes and identification of proteins that bind in vitro to its 52TTR. Nucleic Acids Res. 21: 2301-2308. Carnevali, F., La Porta, C., Ilardi, V., and Beccari E. 1989. Nuclear factors specifically bind to upstream sequences of a Xenopus laevis ribosomal protein gene promoter. Nucleic Acids Res. 17: 8171-8184. Cecconi, F., Mariottini, P., Loreni, F., Pierandrei-Amaldi,P., Campioni, N., and Amaldi, F. 1994. a small nucleolar RNA with a 12 nt complementarity to 18s rRNA and coded by a sequence repeated in the six introns of Xenopus laevis ribosomal protein S8 gene. Nucleic Acids Res. 22: 732-741. CutruzzolB, F., Loreni, F., and Bozzoni, I. 1986. Complementarity of conserved sequence elements present in 28s ribosomal RNA and in ribosomal protein genes of Xenopus laevis and Xenopus tropicalis. Gene, 49: 37 1-376. Di Cristina, M., Menard, R., and Pierandrei-Amaldi, P. 1991. Xenopus laevis ribosomal protein S l a cDNA sequence. Nucleic Acids Res. 19: 1943. Elsdale, T.R., Fischberg, M., and Smith, S. 1958. A mutation that reduces nucleolar number in Xenopus laevis. Exp. Cell Res. 14: 642-643.


Biochem. Cell Biol. Vol. 73, 1995 Filipowicz, W., and Kiss,-T. 1993. Structure and function of nucleolar snRNPs. Mol. Biol. Rep. 18: 149156. Fragapane, P., Caffarelli, E., Lener, M., Prislei, S., Santoro, B., and Bozzoni, I. 1992. Identification of the sequences responsible for the splicing phenotype of the regulatory intron of the L1 ribosomal protein gene of Xenopus laevis. Mol. Cell. Biol. 12: 1117-1125. Fragapane, P., Prislei, S., Michienzi, A., Caffarelli, E., and Bozzoni, I. 1993. A novel small nucleolar RNA (U16) is encoded inside a ribosomal protein intron and originates by processing of the pre&A. EMBO J. 12: 2921-2928. Geyer, P.K., Meyuhas, O., Perry, R.P., and Johnson, L.F. 1982. Regulation of ribosomal protein mRNA content and translation in growth-stimulated mouse fibroblasts. Mol. Cell. Biol. 2: 685693. Hamrnond, M.L., Merrick, W., and Bowman, L.H. 1991. Sequences mediating the translation of mouse S16 ribosomal protein mRNA during myoblast differentiation and in vitro and possible control points for the in vitro translation. Genes Dev. 5: 1723-1736. Hariharan, M.D., and Perry, R.P. 1990. Functional dissection of a mouse ribosomal protein promoter: significance of the polypyrimidine initiator and an element in the TATA-box region. Proc. Natl. Acad. Sci. U.S.A. 87: 1526-1530. Hariharan, M., Kelley, D., and Perry, R.P. 1989. Equipotent mouse ribosomal protein promoters have a similar architecture that includes internal sequence elements. Genes Dev. 3: 1789-1800. Jefferies, H.B.J., Reinhard, C., Kozma, S.C., and Thomas, G. 1994. Rapamycin selectively represses translation of the "polypyrimidine tract" mRNA family. Proc. Natl. Acad. Sci. U.S.A. 91: 444145. Kaspar, R.L., Rychlik, W., White, M.W., Rhoads, R.E., and Moms, D.R. 1990. Simultaneous cytoplasmic redistribution of ribosomal protein L32 mRNA and phosphorylation of eukaryotic initiation factor 4E after mitogenic stimulation of swiss 3T3 cells. J. Biol. Chem. 265: 3619-3622. Kaspar, R.L., Tomohito, K., Cranston, H., Morris, D.R., and White, M.W. 1992. A regulatory cis element and a specific binding factor involved in the mitogenic control of murine ribosomal protein L32 translation. J. Biol. Chem. 267: 508-514. Kiss, T., and Filopowicz, W. 1993. Small nucleolar RNAs encoded by introns of the human cell cycle regulator gene RCCI. EMBO J. 12: 2913-2920. Lazaris-Karatzas, A,, Montine, K.S., and Sonenberg, N. 1990. Malignant transformation by a eukaryotic initiation factor subunit that binds to mRNA S'cap. Nature (London), 345: 544-547. Levy, S., Avni, D., Hariharan, M., Perry, R.P., and Meyuhas, 0. 1991. Oligopyrimidine tract at the 5' end of mammalian ribosomal protein mRNAs is required for their translational control. Proc. Natl. Acad. Sci. U.S.A. 88: 3319-3323. Lodish, H.F. 1974. Model for the regulation of mRNA translation applied to haemoglobin synthesis. Nature (London), 251: 385388. Loreni, F., and Amaldi, F. 1992. Translational regulation of ribosomal protein synthesis in Xenopus cultured cells: mRNA relocation between polysomes and RNP during nutritional shifts. Eur. J. Biochem. 105: 1027-1032. Loreni, F., Ruberti, I., Bozzoni, I., Pierandrei-Amaldi, P., and Amaldi, F. 1985. Nucleotide sequence of the L1 ribosomal protein gene of Xenopus laevis: remarkable sequence homology among introns. EMBO J. 4: 3483-3488. Loreni, F., Francesconi, A,, and Amaldi, F. 1993. Coordinate translational regulation in the syntheses of elongation factor la and ribosomal proteins in Xenopus laevis. Nucleic Acids Res. 21: 47214725. Marchioni, M., Morabito, S., Salvati, A.L., Beccari, E., and Carnevali, F. 1993. XrpFI, an amphibian transcription factor composed of multiple polypeptides immunologically related to the GA binding protein a and P subunits, is differentially

expressed during Xenopus laevis development. Mol. Cell. Biol. 13: 6479-6489. Mariottini, P., and Amaldi, F. 1990. The 5' untranslated region of mRNA for ribosomal protein S19 is involved in its translation during Xenopus development. Mol. Cell. Biol. 10: 816822. Mariottini, P., Bagni, C., Annesi, F., and Amaldi, F. 1988. Isolation and nucleotide sequences of cDNAs for Xenopus laevis ribosomal protein S8: similarities in the 5' and 3' untranslated regions of mRNAs for different r-proteins. Gene, 67: 69-74. Mariottini, P., Annesi, F., Bagni, C., Chen, Q.-M., Francesconi, A., Pesce, C.D., Serra, M.J., and Amaldi, F. 1990. Ribosomal protein genes in Xenopus laevis: organization, structure and identification of the cis-element responsible for their translational control. In Nuclear structure and function. Edited by J.R. Harris and I.B. Zbarsky, Plenum Press, New York. pp. 83-87. Mariottini, P., Bagni, C., Francesconi, A., Cecconi, F., Serra, M.J., Chen, Q.-M., Loreni, F., Annesi, F., and Amaldi, F. 1993. Sequence of the gene coding for ribosomal protein S8 of Xenopus laevis. Gene, 132: 255-260. Meyuhas, O., Thompson, E.A., and Perry, R.P. 1987. Glucocorticoid selectively inhibits translation of ribosomal protein mRNAs in P1798 lymphosarcoma cells. Mol. Cell. Biol. 7: 2691-2699. Miller, L. 1973. Control of 5 s RNA synthesis during early development of anuclmeolate and partial nucleolate mutants of Xenopus laevis. J. Cell Bioi. 59: 624-632. Moure-Neto, R., Dudor, K.P., and Perry, R.P. 1989. An element downstream of the capsite is required for transcription of the gene encoding mouse ribosomal protein L32 gene. Proc. Natl. Acad. Sci. U.S.A. 86: 39974001. Nomura, M. 1990. History of ribosome research: a personal account. In The ribosome: structure, function and evolution. Edited by W. Hill, A. Dahlberg, R.A. Garrett, P.B. Moore, D. Schlessinger and J.R. Warner. Am. Soc. Microbiol., Washington, D.C. pp. 3-55. Pellizzoni, L., Crosio, C., Campioni, N., Loreni, F., and PierandreiAmaldi, P. 1994. Different forms of U15 snoRNA are encoded in the introns of the ribosomal protein S1 gene of Xenopus laevis. Nucleic Acids Res. 22: 4607-4613. Pellizzoni, L., Crosio, C., and Pierandrei-Amaldi,P. 1995. Sequence of the cDNA and gene coding for ribosomal protein S 1 of Xenopus laevis. Gene, 154: 145-15 1. Perry, R.P., and Meyuhas, 0 . 1990. Translational control of ribosomal protein production in mammalian cells. Enzyme, 44: 83-92. Pierandrei-Amaldi, P., and Beccari, E. 1981. Messenger RNA for ribosomal proteins in Xenopus laevis oocytes. Eur. J. Biochem. 106: 603-611. Pierandrei-Amaldi, P., Campioni, N., Beccari, E., Bozzoni, I., and Amaldi, F. 1982. Expression of ribosomal protein genes in Xenopus laevis development. Cell, 30: 163-171. Pierandrei-Amaldi, P., Beccari, E., Bozzoni, I., and Amaldi, F. 1 9 8 5 ~Ribosomal . protein production in normal and anucleolate Xenopus embryos: regulation at the posttranscriptionaland translational levels. Cell, 42: 317-323. Pierandrei-Amaldi, P., Campioni, N., Gallinari, P., Beccari, E., Bozzoni, I., and Amaldi, F. 1985b. Ribosomal protein synthesis is not autogenously regulated at the translational level in Xenopus laevis. Dev. Biol. 167: 281-289. Pierandrei-Amaldi, P., Campioni, N., and Cardinali, B. 1991. Experimental changes in the amount of maternally stored ribosomes affect the translation efficiency of ribosomal protein mRNA in Xenopus embryo. Cell. Mol. Biol. 37: 227-238. Prislei, S., Michienzi, A., Presutti, C., Fragapane, P., and Bozzoni, I. 1993. Two different snoRNAs are encoded in introns of amphibian and human L1 ribosomal protein genes. Nucleic Acids Res. 21: 582&5830. Ritossa, F., Atwood, K.C., and Spiegelman, S. 1966. A molecular explanation of bobbed mutants of Drosophila as partially deficiencies of ribosomal RNA . Genetics, 54: 8 19-834.

Amaldi et al.


Sollner-Webb, B. 1993. Novel intron-encoded small nucleolar RNAs. Cell, 75: 402405. Steel, L.F., and Jacobson, A. 1991. Sequence elements that affect mRNA translation activity in developing Dictyostelium cells. Dev. Genet. 12: 98-103. Steele, R.E., Thomas, PS., and Reeder, R.H. 1984. Anucleolate frog embryos contain ribosomal DNA sequences and a nucleolar antigen. Dev. Biol. 102: 409416. Tasheva, E.S., and Roufa, D.J. 1995. Regulation of human RPS14 transcription by intronic antisense RNAs and ribosomal protein S 14. Genes Dev. 9: 304-3 16. Tyc, K., and Steitz, J.A. 1989. U3, U8 and U13 comprise a new class of mammalian snRNPs localized in the cell nucleolus. EMBO J., 8: 3113-3119. Tycowski, K.T., Shu, M.-D., and Steitz, J.A. 1993. A small nucleolar RNA is processed from an intron of the human gene encoding ribosomal protein S3. Genes Dev. 7: 1176-1190. Tycowski, K.T., Mei-Di, S., and Steitz, J.A. 1994. Requirement for intron-encoded V22 small nucleolar RNA in 18s ribosomal RNA maturation. Science (Washington, D.C.), 266: 1558-1561.

Varnum, S.M., and Wormington, W.M. 1990. Deadenylation of maternal mRNAs during Xenopus oocyte maturation does not require specific cis-sequences: a default mechanism for translational control. Genes Dev. 4: 227S2286. Wagner, M., and Perry, R.P. 1985. Characterization of the multigene family encoding the mouse S16 ribosomal protein: strategy for distinguishing an expressed gene from its processed pseudogene counterparts by an analysis of total genomic DNA. Mol. Cell. Biol. 5: 3560-3576. Wallace, H.R., and Bimstiel, M.L. 1966. Ribosomal cistrons and the nucleolar organizer. Biochim. Biophys. Acta, 114: 296-310. Wool, I.G., Endo, Y., Chan, Y.L., and Gliick, A. 1990. Structure, function and evolution of mammalian ribosomes. In The ribosome: structure, function and evolution. Edited by W. Hill, A. Dahlberg, R.A. Garrett, P.B. Moore, D. Schlessinger and J.R. Warner. Am. Soc. Microbiol., Washington, D.C. pp. 203-214. Wormington, M.W. 1989. Developmental expression and 5 s binding activity of Xenopus ribosomal protein L5. Mol. Cell. Biol. 9: 5281-5288.