INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

About 5000 genes are expressed both during growth and development of Dictyostelium discoideum (the so-called constitutive genes). A few hundred genes are expressed only in the pre-aggregation stage of development (early genes), and about 3000 genes are expressed only after the formation of tight cell aggregates (late genes) (1, 2). The last group can be subdivided into genes expressed preferentially or exclusively in pre-spore or in pre-stalk cells and genes expressed in both cell types (pre-spore-specific, pre-stalk-specific, and common genes, respectively) (3).

In vivo ³²P-labeling experiments have indicated that the expression of a large fraction of pre-spore genes is controlled mainly at the level of mRNA stability (4). Transcription of these mRNAs occurs in vegetative cells or is induced by cell starvation at the very beginning of development and continues at a relatively constant rate throughout development. However, these mRNAs fail to accumulate in the pre-aggregation stage because they are highly unstable. They become stable and start to accumulate when formation of tipped aggregates begins, their stabilization being probably dependent on this process (5). If aggregates are dispersed, the same mRNAs are destabilized and disappear quickly from the dissociated cells (6, 7). The addition of cAMP to the disaggregated cells prevents mRNA destabilization (8, 9).

Since the mRNA molecules that are in the cytoplasm in aggregated cells as stable molecules are the same molecules that are destabilized upon cell disaggregation and restabilized if cells are allowed to reaggregate (9), it is likely that they contain an element (a specific sequence or a sequence with a defined secondary structure) that targets them as mRNA species whose stability has to be regulated. On the other hand the control must be mediated by one or more cytoplasmic factors, which in turn must be responsive to cell-cell contacts and/or to the extracellular concentration of cAMP. In order to consolidate our previous findings and to try to unravel the mechanism of mRNA stabilization/destabilization, we have studied in detail some of the parameters of this process for the AC914 mRNA, which has been chosen as representative of the class of pre-spore-specific mRNAs. The findings reported here indicate that a critical step involved in the control of mRNA stability is the interaction between the 5'UTR of the mRNA and the 40 S ribosomal subunit. The outcome of this interaction is dependent on the state of phosphorylation of ribosomal protein S6.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Cell Growth and Development-- D. discoideum strain AX2 was grown, allowed to develop, and disaggregated as previously described (9). Where specified, cells disaggregated at the stage of first finger were shaken in suspension in the presence of 200 µg/ml pactamycin (The Upjohn Co.), 1 mg/ml puromycin (Sigma), or 200 µg/ml cycloheximide (Sigma).

RNA Isolation and Northern Analysis-- Total RNA was extracted by using Ultraspec II-RNA, following the instructions given by the manufacturer (Biotex Laboratories). 6 µg of total RNA per lane were run onto a 1.2% agarose gel in denaturing conditions as described previously (9), until the migrating dye reached the front of the gel. At the end of the run the amount and integrity of rRNA in each lane were evaluated by ethidium bromide staining. After blotting for 18 h in 10× SSC onto Hybond-N (Amersham Pharmacia Biotech), RNA was covalently linked to the membrane by 3 min exposure to UV light as described in Ref. 10.

DNA Labeling and Hybridization-- Plasmid DNA probes were labeled by the random priming method (10), purified by spin-column chromatography on Sephadex G-50M in TE buffer, and hybridized to Northern blots for 18 h at 37 °C in 50% formamide, 5× SSC, 2% SDS, 2% BSA,¹ 2% Ficoll, 2% PVP. Radioactive membranes were washed four times for 15 min in 250 ml of 2× SSC, 1% SDS at 65 °C and exposed for autoradiography or analyzed with the Bio-Rad PhosphorImager.

DNA Clones-- AC914 is a genomic clone isolated by Dr. A. Ceccarelli from a partial Sau3A1 D. discoideum genomic library inserted into BamHI-cut pAT153. The clone contains 1.1 kbp of sequence upstream of the Cap site of the AC914 gene² and the coding sequence flanked by 1.4 kbp of sequences downstream of the stop codon.

Nuclei Isolation, RNA Labeling, and Hybridization-- Vegetative and developing Dictyostelium cells (at 3 h and at first finger stages) were lysed by vortexing at 0 °C in 50 mM HEPES, pH 7.5, 5 mM MgOAc, 10% sucrose, 2% Nonidet P-40 at 10⁸ cells/ml. Nuclei were isolated by centrifugation for 10 min at 6000 rpm in a JA 14.1 Kontron rotor and washed twice in the same buffer. Nuclei resuspended at 10⁸/ml in storage buffer (40 mM Tris-HCl, pH 7.9, 10 mM MgCl₂, 0.1 mM EDTA, 1 mM dithiothreitol, 50% glycerol) were snap-frozen in liquid nitrogen and stored at 80 °C in 200-µl aliquots.

Preparation of Polyribosomes-- Cells at the first finger stage of development were lysed by vortexing them for 15 s at 2 °C in cold buffer containing 2% Tergitol, 10% sucrose, 20 mM MgCl₂, 30 mM KCl, and 20 mM Tris-HCl, pH 7.5. The lysate was centrifuged in an Eppendorf centrifuge at maximum speed for 6 min to obtain the post-lysosomal supernatant. This was centrifuged in a SW40 rotor at 35,000 rpm for 90 min through a 15-35% sucrose gradient containing the same salts as the lysis buffer. The top half of the gradient was discarded, and the bottom half was used directly as a source of polyribosomes.

In Vitro Incubation of Polyribosomes-- To measure the stability of the endogenous mRNAs, 5 A₂₆₀/ml polyribosomes, containing 0.1 A₂₆₀ of total poly(A)⁺ mRNA, but unknown amounts of each specific mRNA, were incubated at 22 °C in the presence of 1 µg of soluble proteins (16), 2 mg/ml ATP, 2 mg/ml GTP, and 1 mg/ml Dictyostelium tRNA.

Isolation of 40 S Ribosomal Subunits-- Cells were lysed in the same buffer described above but containing a higher concentration of KCl (0.5 M). The same buffer (without Tergitol) was used for the centrifugation on sucrose gradients.

In Vitro Assembly of 40 S Ribosomal Subunits-- The procedure described in Ref. 16 was followed.

Two-dimensional Gel Electrophoresis and Autoradiography of 40 S Ribosomal Proteins-- The procedure described in Ref. 17 was followed.

In Vitro Assembly of 40 S Ribosomal Subunits Containing S6 as the only Phosphorylated Protein-- Phosphorylated protein S6 was eluted from a gel similar to the one shown in Fig. 4B, following the procedure described (17). By rerunning a sample of the eluted protein by two-dimensional electrophoresis, we could infer that we had succeeded in solubilizing about 50% of the S6 protein contained in the original gel. The solubilized S6 protein was dialyzed against 6 M urea, 50 mM Tris-HCl, pH 7.2, and added to a 40 S subunit reconstitution mixture (16) containing an amount of ribosomal RNA and unlabeled proteins (derived from vegetative cells) equal to 1/20 of the amount used to display phosphorylated protein S6 on the original gel. At the end of the incubation, the reconstituted 40 S subunits were purified by centrifugation on a sucrose gradient and their protein analyzed by two-dimensional gel electrophoresis. The intensity of the autoradiographic spots corresponding to protein S6 was compatible with the incorporation of a molecule of phosphorylated S6 per 40 S subunit.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

AC914 Is a Late Gene with a Constitutive Promoter-- AC914 mRNA accumulates only late in development starting at the stage of tipped aggregates or first finger (Table I, line 1). Nuclear run-off experiments indicate that AC914 transcription is already in process in vegetative cells and declines during development (Fig. 1, solid circles). This is in agreement with the finding that luciferase mRNA transcribed from the AC914 promoter is present both in vegetative and developing cells (Table I, line 2).

View this table: [in this window] [in a new window]   Table I Kinetics of mRNA accumulation during development Quantitative assays of mRNAs in developing cells were carried out by Northern analysis as described under "Experimental Procedures." Half-lives of mRNAs in cells disaggregated and kept in shaken suspension were measured as described (9). In the experiment reported in line 8, 200 µg/ml pactamycin were added to disaggregated cells. The mRNAs analyzed in lines 1, 5, and 8 were those transcribed from the endogenous genes (AC914 and A15), and the same cloned genes were used as 32P-labeled probes. The mRNAs analyzed in lines 2-4, 6, and 7 were those transcribed by the construct whose organization is described using the following symbols: , promoter; , 5'-UTR; , coding region and 3'UTR. In this case the 32P-labeled probe was a c-Myc oligonucleotide.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Nuclear run-off assay on Dictyostelium cells at different times of development. RNA labeled in isolated nuclei was hybridized to different DNAs spotted onto Hybond-N, and the hybridized radioactivity was measured with a Bio-Rad PhosphorImager. Solid squares, DNA from AC914 gene fused to A15 promoter; solid circles, AC914 DNA.

AC914 mRNA Accumulates Only in Tight Aggregates Even When Transcribed from a Strong Constitutive Promoter-- To confirm that the kinetics of accumulation of AC914 mRNA depends on the modulation of its stability, we have fused the AC914 gene to the Actin 15 (A15) promoter, generating line 3 listed in Table I. To be able to distinguish the mRNA transcribed from the heterologous promoter and the mRNA transcribed from the endogenous gene, we have inserted in the coding region of construct 3 a 33-mer oligonucleotide derived from c-myc (13). By using this oligonucleotide as a ³²P-labeled probe, we were able to detect the AC914 mRNA transcribed from the A15 promoter by Northern analysis. The kinetics of accumulation of this mRNA (Table I, line 3) is similar to that of the mRNA transcribed from the endogenous gene (Table I, line 1). Since the run-off transcription activity of the A15 promoter (Fig. 1, solid circles) parallels the one of the AC914 promoter, this result reinforces the notion that the accumulation of AC914 mRNA is controlled during development at the level of mRNA stability.

Substituting the 5'UTR of AC914 mRNA with One of a Stable mRNA Prevents Destabilization-- The coding region and the 3'UTR of AC914 gene were fused to the A15 promoter joined to the A15 5'UTR and to the nine nucleotides coding for the first three A15 amino acids (construct 4, Table I, line 4). Also in these cases a 33-mer oligonucleotide from c-myc was inserted in frame in the coding region of the gene, to allow detection only of the mRNA transcribed from the A15 promoter. The mRNAs transcribed from A15 promoter was already present in vegetative cells and did not accumulate further during development (Table I, line 4). The endogenous AC914 mRNA in the same cells (distinguishable by Northern analysis because of its smaller size) maintained its regulation and was detectable only after 9 h of development (data not shown). The decay of the endogenous AC914 mRNA and of that transcribed from construct in Table I, line 4 in disaggregated cells was monitored by Northern blot analysis. As expected the first one decayed rapidly (Table I, line 1), and the latter one remained stable (Table I, line 4). These results indicate that the 5'UTR of AC914 mRNA is necessary for the mRNA to be destabilized during growth and early development and in disaggregated cells.

Adding the 5'UTR of AC914 mRNA to a Stable mRNA Destabilizes It-- To determine whether the 5'UTR of AC914 mRNA was capable of destabilizing an mRNA intrinsically stable, we substituted the 5'UTR of A15 mRNA, which is stable when transcribed both from its own promoter and from the AC914 promoter (Table I, lines 5 and 6) with the 5'UTR of the AC914 mRNA. The hybrid mRNA began to accumulate at the stage of tipped aggregates and was destabilized in disaggregated cells (Table I, line 7). We conclude that the 5'UTR of AC914 is not only necessary but also sufficient to destabilize a stable mRNA in non-aggregated cells.

Destabilization of AC914 mRNA Does Not Require Protein Synthesis but Is Inhibited by Pactamycin-- The fact that the AC914 mRNA containing A15 5'UTR cannot be destabilized whereas A15 mRNA containing AC914 5'UTR has a regulated stability may be interpreted in different ways. One possibility is that the interaction between the 5'UTR of the mRNA and the small ribosomal subunit plays a crucial role in the destabilization mechanism. To test whether the 40 S subunit-5'UTR interaction is involved in the control of mRNA stability, cells were disaggregated in the presence of pactamycin, a drug that blocks the movement of 40 S ribosomal subunits from the mRNA Cap site to the first AUG codon. The addition of 200 µg/ml pactamycin (a concentration sufficient to inhibit completely the incorporation of [³⁵S]methionine into polypeptides (data not shown)) prevented completely the decay of the endogenous AC914 mRNA (Table I, line 8).

mRNAs in Polyribosomes Isolated from Aggregated Cells Are Relatively Stable When Incubated in Vitro-- Fig. 2, A and B, shows the in vivo stability of the two test mRNAs, AC914 and SC253. Both mRNAs are stable at the first finger stage. Upon disaggregation, the pre-stalk-specific SC253 mRNA remains stable, whereas the pre-spore-specific AC914 mRNA is destabilized and disappears rapidly.

View larger version (45K): [in this window] [in a new window]   Fig. 2.   Stability of AC914 and SC253 mRNAs in vivo and in vitro. The decay of the two test mRNAs at the stage of first finger (A) and in disaggregated cells (B) was followed by in the presence of the transcription inhibitor nogalamycin as described (9). CE, five A260 units of polyribosomes isolated as described in the text from aggregated, disaggregated, or reaggregated cells were incubated in vitro for the indicated times. RNA was isolated from the polyribosomes and analyzed by Northern blot hybridization, using AC914 and SC253 DNAs as 32P-labeled probes.

Only the mRNA Destabilized in Vivo by Cell Disaggregation Is Unstable in Vitro in Polyribosomes Isolated from Disaggregated Cells-- In polyribosomes isolated from disaggregated cells and incubated in vitro, AC914 mRNA disappeared with a half-life of about 15 min (Fig. 2D). SC253 mRNA remained stable as in polyribosomes derived from aggregated cells. When disaggregated cells were replated and allowed to reaggregate (9) before polyribosome isolation, both test mRNAs were stable upon subsequent incubation in vitro (Fig. 2E). Thus the stability of these two mRNAs in vitro reproduces exactly that observed in vivo.

AC914 mRNA Is Destabilized by the Presence of 40 S Ribosomal Subunits Derived from Cells Not Yet Aggregated or Disaggregated-- We have exchanged components between in vitro systems consisting of polyribosomes isolated from aggregated and from disaggregated cells to determine which was involved in controlling mRNA stability. The addition of 40 S ribosomal subunits from polyribosomes isolated from disaggregated cells destabilized AC914 mRNA contained in polyribosomes isolated from aggregated cells, while having no effect on SC253 mRNA stability (Table II, line 4). The exchange of any other component (soluble proteins, 60 S subunits, tRNA, and mRNAs) had no effect on AC914 mRNA stability (data not shown). The addition of 40 S ribosomal subunits derived from vegetative AX2 cells or from cells in the pre-aggregation stage to polyribosomes derived from aggregated cells also destabilized selectively AC914 mRNA (Table II, lines 1 and 2).

Proteins Are the 40 S Subunit Components Modified by Cell Aggregation-- We have recently described the in vitro reconstitution of ribosomal subunits from free Dictyostelium ribosomal RNA and proteins (16). Hybrid 40 S subunits containing pre-17 S rRNA from vegetative cells and ribosomal proteins derived from vegetative, pre-aggregated, aggregated, and disaggregated cells were constructed and tested for their ability to destabilize AC914 mRNA contained in polyribosomes derived from aggregated cells. All types of particles were active in this respect, except those containing proteins from aggregated cells (Table III).

cAMP-dependent Kinase Is Involved in the Control of mRNA Stability-- Since cAMP prevents mRNA destabilization induced by cell disaggregation (8, 9), it was reasonable to suppose that cAMP-dependent protein kinase A might be involved in the stability of control mechanisms. The availability of transformed Dictyostelium cells which overexpress the catalytic subunit of protein kinase A (K cells) (18, 19) allowed this hypothesis to be tested. When K cells were disaggregated, AC914 mRNA was destabilized and decayed in vivo by about a factor of 1.5 in the first 15 min but then became stable again and no longer decayed (Fig. 3). Overexpression of the catalytic subunit of protein kinase A therefore is not sufficient to prevent mRNA destabilization initially, but it rapidly overcomes the destabilization and reimposes mRNA stability.

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Effect of the overexpression of the catalytic subunit of protein kinase A on the in vivo stability of AC914 mRNA. K cells were disaggregated at the first finger stage of development and shaken in suspension. RNA extracted at the indicated times was analyzed by Northern blot hybridization using 32P-labeled AC914 DNA as a probe. Before autoradiography of the filter, the amount of 32P present in each band was measured with a Bio-Rad PhosphorImager. The label decayed by 30% between 0 and 15 min and then remained constant.

Phosphorylation of Protein S6 Renders 40 S Ribosomes Incapable of Destabilizing mRNA-- It has been reported (17) that the only modification of ribosomal proteins which consistently occurs during development of Dictyostelium is the phosphorylation of protein S6, which occurs on three residues of serine during cell aggregation. To determine the relevance of S6 phosphorylation on mRNA stability, the proteins of 40 S ribosomal subunits labeled with [³²P]orthophosphate in vivo were extracted and analyzed by two-dimensional gel electrophoresis and autoradiography as described (17). From Fig. 4A, it is apparent that no 40 S protein is phosphorylated in the pre-aggregation stage of development, whereas three phosphorylated derivatives of protein S6 are visible in the post-aggregation stage (Fig. 4B). They disappear if cells are disaggregated in the absence of cAMP (Fig. 4C) but remain if cells are disaggregated in the presence of 1 mM cAMP (Fig. 4D). The three phosphorylated derivatives reappear if cells are allowed to reaggregate (Fig. 4E). Finally in a mutant in which development is blocked before the formation of tight aggregates (Agg), protein S6 is not phosphorylated even at late times of development (Fig. 4F). The trypsin fingerprint analysis of protein S6 was altered in the mutant, indicating that the mutation had occurred in protein S6 gene. Phosphorylation of protein S6 may therefore play a crucial role in the progress of Dictyostelium development.

View larger version (47K): [in this window] [in a new window]   Fig. 4.   Phosphorylation of ribosomal protein S6 at various stages of development. Cells were labeled and ribosomal proteins analyzed by gel electrophoresis and autoradiography as described (17). A, cells labeled immediately after starvation; B, cells labeled at the first finger stage; C, cells labeled at the first finger stage and disaggregated; D, as in C, but 1 mM cAMP was added at the time of disaggregation; E, cells labeled at the time of first finger, disaggregated, and allowed to reaggregate; F, Agg cells labeled 15 h after starvation (corresponding to the first finger stage in wild type strain).

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

We have shown that AC914 mRNA undergoes two major changes in its stability, one during the formation of tight cell aggregates and the other when cells are disaggregated. The nuclear run-off data for the endogenous gene show that it is already active during vegetative growth. The accumulation of a reporter gene coupled to the AC914 promoter shows clearly that the control elements present in this fragment of DNA can direct transcription in a similar way to the endogenous promoter. These results consolidate our previous findings based on ³²P-labeling kinetics (4, 7, 9) showing that the expression of a group of pre-spore genes is regulated at the level of mRNA stability.

Both transitions in stability require that the 5'UTR of AC914 mRNA is present and is not substituted by the 5'UTR of a constitutively stable mRNA. On the other hand the 5'UTR of AC914 fused in front of a stable mRNA is sufficient to destabilize this mRNA in non-aggregated cells.

The 5'UTR of AC914 therefore contains all the elements in cis required to modulate the stability of an mRNA as a function of cell aggregation. This is the first example in the literature of destabilizing elements located in the 5'UTR of an mRNA, rather than in the 3' or more rarely in the coding region (20). More work will be required to define these elements.

The 5'UTR of an mRNA interacts only with the small ribosomal subunit and with some initiation factors. Our results with AC914 mRNA thus suggest that 40 S ribosomal subunits can interact with the 5'UTR of AC914 to form a specific complex with the destabilizing elements, but this interaction is not possible if the 5'UTR of AC914 is replaced by A15 5'UTR.

In line with this interpretation, AC914 mRNA destabilization is inhibited by pactamycin, which prevents the movement of 40 S ribosomal subunits from the Cap toward the first AUG codon, but not by other inhibitors of protein synthesis.

The role of 40 S ribosomal subunits in AC914 mRNA destabilization is clearly shown by the in vitro experiments reported in this paper. Several in vitro systems to study mRNA stability have been described (21-23) and have given useful information concerning the mechanism of stability control of specific mRNAs. Our finding that a given molecule of mRNA may be stable or unstable in vitro following the exchange of 40 S subunits between polyribosomes isolated from aggregated and disaggregated cells strongly implicates 40 S subunits as at least one of the trans-acting factors in this mechanism. The fact that 40 S subunits derived from cells in growth or in the pre-aggregation stage of development also destabilize AC914 mRNA suggests that the shift in stability occurring at the time of tight aggregation may be due to a mechanism opposite to that involved in mRNA destabilization following cell disaggregation.

The component of 40 S subunits involved in the stability control is protein. This suggests the possibility that the mRNA stabilization/destabilization mechanism is based on phosphorylation/dephosphorylation of one or several ribosomal proteins, as implied by the involvement in the process of protein kinase A indicated by our results with K cells. This suggestion has been confirmed by the finding that the state of phosphorylation of ribosomal protein S6 is responsible for the activity of 40 S subunits in destabilizing pre-spore-specific mRNAs. This is clearly indicated by the inability of pre-spore-specific mRNAs to become stable if the S6 phosphorylation is prevented by a mutation in the same protein and proven by the fact that in vitro reconstituted 40 S subunits which differ only in the absence or the presence of phosphate groups on protein S6 can or cannot destabilize prespore-specific mRNAs. However, we do not know whether S6 phosphorylation is directly carried out by protein kinase A or is mediated by some other protein kinase.

As we have mentioned before, if disaggregated cells are allowed to reaggregate, the same molecules of mRNA which had become unstable recover stability. To explain the reversibility of the destabilization process, one has to admit that the interaction between the destabilizing elements in the 5'UTR and the destabilizing form of 40 S subunits is not always productive. One possibility is that this interaction determines the decapping of the mRNA, but this event would not be obligatory and could occur only on a fraction of mRNAs. Another possibility is that only a fraction of 40 S ribosomes carry an mRNase. We recall that the hypothesis of a fraction of ribosomes devoted to the degradation of mRNA was advanced years ago for Escherichia coli (24).

Since mRNA destabilization apparently involves the interaction between the 5'UTR and 40 S ribosomal subunits, the latter must function as independent particles, not joined to 60 S subunits. When the cyclic dissociation of ribosomal subunits was discovered (25-28), it was interpreted as due to the modality of the initiation and termination stages of the synthesis of polypeptide chains. The data presented in this paper suggest that at least in some cases 40 S subunits have a function in which they act alone and which is not strictly related to protein synthesis (destabilizing a class of mRNA). It has been reported (29) that calmodulin binds to a protein of 60 S ribosomal subunits and presumably in so doing participates in the regulation of the concentration of cytoplasmic Ca²⁺. Several authors (30-32) have reported that the large subunit of both prokaryotic and eukaryotic ribosomes, and specifically its major RNA, has the capacity to mediate refolding of denatured proteins. Perhaps we should consider that ribosomal subunits may have functions different from protein synthesis.