A number of specific mRNAs are destabilized upon disaggregation of developing Dictyostelium discoideum cells. Analysis of a family of cloned genes indicates that only prespore-enriched mRNAs are affected; constitutive mRNAs that are expressed throughout development and mRNAs that accumulate preferentially in prestalk cells are stable under these conditions. The decay of sensitive prespore mRNAs can be halted by allowing the cells to reaggregate, indicating that destabilization occurs by the progressive selection of individual molecules rather than on all members of an mRNA subpopulation at the time of disaggregation. Individual molecules of the sensitive mRNA species remain engaged in protein synthesis in the disaggregated cells until selected. Destabilization of sensitive mRNAs is induced by cell dissociation even in the presence of concentrations of nogalamycin that inhibit RNA synthesis. The reported prevention of disaggregation-induced mRNA decay by actinomycin D and daunomycin is therefore probably a secondary effect unrelated to the inhibition of transcription.

Dictyostelium discoideum is an increasingly popular organism for analysing the regulation of gene expression during cell differentiation. When deprived of nutrients and deposited on a solid substratum, the amoebae aggregate over a period of 8–10 h to form tight multicellular complexes. After forming a tip, each aggregate transforms into a migrating pseudoplasmodium referred to as a grex or slug. At this stage, the presence of two cell types is already apparent; prespore and prestalk cells. Later in development, during culmination, final differentiation of these cells takes place to give rise to a mass of mature spore cells on top of a slender pillar of vacuolated stalk cells.

Many hundred (Jacquet et al. 1981) and possibly several thousand (Blumberg & Lodish, 1980) new mRNA species are expressed during development, some of which accumulate preferentially in prespore or prestalk cells (Alton & Brenner, 1979; Morrissey et al. 1981; Barklis & Lodish, 1983; Borth & Ratner, 1983; Mehdy et al. 1983; Morrissey et al. 1984). Dissociation of cells in developing aggregates leads to the cessation of expression of many late genes and degradation of the mRNA already transcribed (Newell et al. 1971, 1972; Alton & Lodish, 1977; Landfear & Lodish, 1980; Chung et al. 1981; Mangiarotti et al. 1982, 1983; Landfear et al. 1982; Mehdy et al. 1983; Bozzaro et al. 1984). Barklis & Lodish (1983) and Chisholm et al. (1984) have suggested that the genes sensitive to cell dissociation are those expressed for the first time in the aggregation or postaggregation stages and can be either prespore or prestalk enriched, while prestalk-enriched genes, initially expressed prior to aggregation, and genes expressed constitutively in growth and development (Barklis & Lodish, 1983; Mehdy et al. 1983) are insensitive to dissociation.

We are interested in elucidating the mechanisms that control developmental gene expression in Dictyostelium. Here we analyse the phenomenon of specific mRNA destabilization and its relationship to the multicellular state. The data presented suggest that the stability of mRNAs expressed constitutively throughout development or expressed preferentially in prestalk cells is unaffected by cell dissociation. In contrast, all prespore mRNAs examined are rapidly destabilized. We also present evidence concerning the mechanism of destabilization. Our studies indicate that disaggregation induces the instability of sensitive mRNAs in a progressive manner, one by one, rather than affecting the entire mRNA population at the time of cell dissociation; that sensitive mRNAs continue to engage in protein synthesis in disaggregated cells until selected for rapid decay; and that, contrary to a recent report (Amara & Lodish, 1987), destabilization is not dependent upon RNA synthesis.

Cell culture conditions

D. discoideum AX2 was grown in HL5 medium as described elsewhere (Watts & Ashworth, 1970). Cells were harvested during exponential growth at a density of 4×106cells ml-1, washed twice with distilled water, once with PDF (Mangiarotti et al. 1982) and resuspended in PDF at l·5× 10scells ml1. Cells were plated for development at a density of 1×107 cells cm-2 on Whatman 50 filter papers each resting on two Whatman 3 filter pads just saturated with PDF. In experiments involving radiolabelling with [32P]orthophosphate, PDF was substituted by MES-PDF (Mangiarotti et al. 1982).

Monitoring mRNA decay by dot blot analysis

Cells plated on filters in MES-PDF were labelled with [32P]orthophosphate (10mCi per 5×108 cells) between 13–17 h of development. After labelling, the cells were harvested from the filters, washed in PDF by centrifugation (2500g, 5min), resuspended in 5 ml PDF containing 10 mm-EDTA and dispersed by vortex mixing. The dispersed cells were then diluted with PDF-EDTA to 5× 106cells ml-1 and shaken at 230revsmin-1 at 22°C. At intervals (see Results), cells were harvested by centrifugation. Poly (A)+ RNA was isolated from each sample (Mangiarotti et al. 1980) and hybridized to DNA dot blots (Mangiarotti et al. 1982). Each dot contained 2 μg DNA of selected clones (as indicated in the Results sections) immobilized on nitrocellulose.

Monitoring mRNA decay using nogalamycin

Filters bearing cells at the 13 h stage of development were transferred to pads containing 200 μg ml-1 nogalamycin in PDF. Another batch of cells at the same stage of development was disaggregated as described above but in PDF-EDTA containing 200 μg ml-1 nogalamycin and then maintained in shaking suspension (230 revs min-1). At intervals, cells were harvested from both the filters and the disaggregated cell suspension and total RNA was isolated as described by Williams et al. (1987). After electrophoresis on 1·5 % agarose gels in 20 mm-phosphate buffer, (pH 7·0) containing 6% formaldehyde and capillary transfer to Hybond N (Amer-sham), the Northern blots were hybridized with DNA prepared from clones SC79, D19, EB4 or Per97 previously labelled by nick translation (Rigby et al. 1977).

Effect of nogalamycin on RNA synthesis in developing and disaggregated cells

D. discoideum cells were allowed to develop on filters to the 15 h stage of development. One batch of filters was then transferred to pads saturated with 200 μg ml-1 nogalamycin (Ennis, 1981) in PDF whilst a second (control) batch was transferred to filters saturated with PDF alone. After 5 min, 100 μCi [3H]uracil was added by spotting uniformly onto each filter using a syringe. Other cells at 15 h of development were harvested and disaggregated in PDF-EDTA as described above. The cell suspension was divided into two halves, to only one of which was added nogalamycin at 200 μg ml-1. After 5 min, 100 μCi [3H]uracil was added to each suspension. At intervals, cell samples were harvested from the two sets of filters and the two sets of suspension cultures and total RNA was isolated (Mangiarotti et al. 1980). For each RNA sample, one sample was treated with 1 μg pancreatic RNAse for 30min at room temperature, then precipitated with ice-cold 10% TCA and TCA-insoluble radioactivity determined by liquid scintillation counting. Another sample was TCA precipitated and counted without RNase digestion. Control RNA samples treated with RNase contained negligible counts. Thus the radioactivity of samples not treated with RNase was used as a direct measure of RNA radioactivity.

Monitoring mRNA decay after disaggregation and reaggregation

Cells at the 17 h stage of development were dissociated as described above and then maintained in shaking suspension. After 30min incubation, the cell suspension was divided into three fractions. One fraction was kept shaking without any further addition. To a second fraction of cells was added nogalamycin to 200 μg ml-1 and then shaking continued. The third fraction of cells was collected by centrifugation and then replated on filters at 107 cells cm-2 in the presence of 200Jμg ml-1 nogalamycin. At intervals during this protocol as described in Results, cell samples were harvested from all three fractions and total RNA was extracted (Williams et al. 1987). After electrophoresis of 10gg of each RNA on 1·5% agarose gels and blotting into Hybond N, the Northern blots were hybridized with selected cloned DNAs previously labelled by nick translation (see Results).

The effect of actinomycin D and daunomycin on mRNA decay

Cells allowed to develop to the 15 h stage of development were dispersed into PDF-EDTA shaking suspension culture as described above. The suspension culture was then divided into two halves. Shaking of one fraction was continued without further addition whilst to the other actinomycin D and daunomycin were added to 125μg ml-1 and 250 μg ml-1 final concentrations respectively and then shaking continued. Total RNA was isolated from cell samples taken at intervals during the incubation of each culture and analysed by Northern blot analysis as described above.

Intracellular localization of mRNA in developing and disaggregated cells

Cells at the 15 h stage of development on filters or disaggregated at this time and kept in shaking suspension in PDF-EDTA for 30 min (see above) were lysed and centrifuged on 15–30 % sucrose gradients as described by Mangiarotti et al. (1981). After centrifugation, gradients were analysed by continuous monitoring at 260 nm. Fractions corresponding to 80S ribosomes and lighter material, polyribosomes up to 6–7 ribosomes per mRNA and polyribosomes more than 6–7 ribosomes per mRNA were separately pooled. Total RNA was extracted (Williams et al. 1987) from each pooled sample and analysed by Northern blotting with labelled cloned DNA as described above.

Only prespore-enriched mRNAs are destabilized upon disaggregation

Earlier studies have concluded that mRNAs that accumulate preferentially in either prespore or prestalk cells during the aggregation and postaggregation stages become destabilized and are rapidly degraded when the cells are disaggregated (Mehdy et al. 1983; Barklis & Lodish, 1983). In contrast, constitutive mRNAs present in both cell types throughout growth and development are unaffected by cell disaggregation. Since earlier studies used relatively few clones to reach these conclusions, we have repeated the analysis using a large bank of clones which are now available. D. discoideum AX2 cells were allowed to develop for 13 h and then labelled with 32P for 4h. At 17h of development, the cells were dissociated and shaken fast in PDF-EDTA buffer to maintain them as single cells. At intervals, samples were taken and used to prepare poly (A)+ RNA which was then hybridized to DNA dot blots corresponding to a large panel of clones. Some of these results are given in Fig. 1. The data show that several constitutive mRNAs, SC79, SC29 and CZ22 (Chung et al. 1981; Barklis & Lodish, 1983; Mangiarotti et al. 1983), and a developmentally regulated non-cell-type-specific mRNA, D15 (Barklis & Lodish, 1983), are indeed stable for the time course of the experiment. Surprisingly, so are five mRNAs that accumulate preferentially in prestalk cells; Dll, D14 (Barklis & Lodish, 1983), Per79 (E. Barklis; personal communication), 5D and 5G (Corney et al. 1989). In contrast, the large number of prespore-enriched mRNAs examined [D18, D19, EB4 (Barklis & Lodish, 1983); Per97 (E. Barklis, personal communication); GM55b, GM27 (Mangiarotti et al. 1980); A3 (Mangiarotti et al. 1982, 1983); IA, 3B, 2C, 2D, 4D, 2H (Corney et al. 1989)] are degraded upon disaggregation. The rate of degradation is specific for each prespore mRNA, some mRNAs decaying with a half-life of 30 min or less. In addition to the clones indicated in Fig. 1, constitutive mRNAs CZ5 and CZ12 (Mangiarotti et al. 1982, 1983) and prestalk mRNAs Bl (Mangiarotti et al. 1982) and Per74 (R. Giorda; unpublished observations) were also examined and found to be stable to disaggregation whilst prespore mRNAs EB5 (E. Barklis, personal communication) 6C, IF, 3F, 7F, 1H and II (Corney et al. 1989) were all unstable (data not shown). Thus, of the mRNAs we have examined, the only class that are unstable upon disaggregation are those which accumulate preferentially in prespore cells during development.

Fig. 1.

The decay of specific mRNAs following disaggregation. Cells were labelled between 13h and 17 h of development, then disaggregated and shaken fast in suspension culture to maintain them as single cells. Poly (A)+ RNA was isolated at various times and then hybridized to dot blots of DNA (2 μg per dot) from constitutive, non-cell-type-specific, prestalk and prespore clones as indicated. The dots for each clone, left to right, represent hybridization to the poly(A)+ RNA isolated at 0, 30, 60, 90 and 120 min of incubation in shaking suspension culture.

Fig. 1.

The decay of specific mRNAs following disaggregation. Cells were labelled between 13h and 17 h of development, then disaggregated and shaken fast in suspension culture to maintain them as single cells. Poly (A)+ RNA was isolated at various times and then hybridized to dot blots of DNA (2 μg per dot) from constitutive, non-cell-type-specific, prestalk and prespore clones as indicated. The dots for each clone, left to right, represent hybridization to the poly(A)+ RNA isolated at 0, 30, 60, 90 and 120 min of incubation in shaking suspension culture.

Prespore destabilization is reversed by reaggregation

If cells dispersed from tight aggregates and kept in suspension for 30 min (a time period sufficient to halve the amount of prespore mRNAs) are replated on filters at a concentration of 107 cells cm-2, they reaggregate to form visible mounds in less than 15 min. In principle it should be possible to monitor the effects of reaggregation upon mRNA stability by the 32P labelling protocol used in Fig. 1. However, in practice, we were unable to obtain sufficient incorporation of 32P into poly(A)+ RNA to allow this experimental approach to succeed. We therefore chose the alternative approach of following the decay of unlabelled mRNAs by Northern blot analysis. For this protocol, it is necessary to inhibit further mRNA synthesis. The antibiotic nogalamycin is ideal for this purpose; at 200 μg ml – 1 nogalamycin, RNA synthesis both by cells developing normally on filters and by disaggregated cells incubated in shaking suspension is inhibited greater than 98% (Fig. 2). Furthermore, nogalamycin does not alter the rate of decay of mRNA in disaggregated cells (see Fig. 3 below) and so has no effect on the intrinsic stability of these mRNAs.

Fig. 2.

Inhibition of RNA synthesis by nogalamycin. Cells at 15 h of development were either allowed to continue development unperturbed (•) or in the presence of 200 μg ml – ?1 nogalamycin (○). Other cells were disaggregated at 15 h and placed in suspension culture. One half of the suspension was shaken in PDF (▴) and the other in PDF containing 200 μg ml–1 nogalamycin (▵). [3FI]uracil was added to each cell fraction and samples taken at the time points shown. The data represent [3H] radioactivity in TCA-insoluble material.

Fig. 2.

Inhibition of RNA synthesis by nogalamycin. Cells at 15 h of development were either allowed to continue development unperturbed (•) or in the presence of 200 μg ml – ?1 nogalamycin (○). Other cells were disaggregated at 15 h and placed in suspension culture. One half of the suspension was shaken in PDF (▴) and the other in PDF containing 200 μg ml–1 nogalamycin (▵). [3FI]uracil was added to each cell fraction and samples taken at the time points shown. The data represent [3H] radioactivity in TCA-insoluble material.

Fig. 3.

mRNA stabilization on reaggregation. Cells at 17 h of development were disaggregated, kept in suspension for 30min, and then divided into three fractions. Fraction A was kept shaking without further addition and fraction B was shaken after the addition of nogalamycin to 200 μg ml –1. Cells from fraction C were collected by centrifugation and replated on filters in the presence of 200 μg ml –1 nogalamycin. Sample 0 was taken at the time of cell disaggregation. For each of fractions A, B and C, samples were also taken at the time of replating (sample 1), and then at 30min (sample 2), 60min (sample 3) and 90 min (sample 4) later. Northern blots of RNA isolated from each sample and probed with SC79, D19, A3 and GM55b [32P] DNA are shown.

Fig. 3.

mRNA stabilization on reaggregation. Cells at 17 h of development were disaggregated, kept in suspension for 30min, and then divided into three fractions. Fraction A was kept shaking without further addition and fraction B was shaken after the addition of nogalamycin to 200 μg ml –1. Cells from fraction C were collected by centrifugation and replated on filters in the presence of 200 μg ml –1 nogalamycin. Sample 0 was taken at the time of cell disaggregation. For each of fractions A, B and C, samples were also taken at the time of replating (sample 1), and then at 30min (sample 2), 60min (sample 3) and 90 min (sample 4) later. Northern blots of RNA isolated from each sample and probed with SC79, D19, A3 and GM55b [32P] DNA are shown.

To examine the effects of reaggregation on mRNA stability, cells allowed to develop for 17 h on filters were dissociated and maintained in suspension for 30min. The suspension was then divided into three fractions. One fraction (Fig. 3A) was kept shaking without further manipulations. To a second fraction (Fig. 3B) was added 200 μg ml-1 nogalamycin and shaking was then continued. Finally, cells from the third fraction (Fig. 3C) were harvested by centrifugation and replated on hlters at 107 cells cm’2 in the presence of 200 μg ml1 nogalamycin. Samples were taken at intervals and poly (A)+ RNA extracted from each was examined by Northern blot analysis using a labelled DNA probe from SC79 (a constitutive mRNA), and prespore D19, A3 and GM55b probes. The constitutive mRNA, SC79, was stable under the three conditions tested and therefore not affected by either disaggregation or reaggregation of the cells. In contrast, the three prespore mRNAs decayed rapidly upon disaggregation, a result in keeping with the data of Fig. 1. The presence of nogalamycin did not affect this rapid decay (compare fractions A and B in Fig. 3). However, upon reaggregation, the prespore mRNAs D19, A3 and GM55b which had not yet decayed remained undegraded over the next 90min. Reaggregation therefore stabilized the residual prespore mRNA against degradation, that is, the disaggregation-induced decay of prespore mRNA is reversible.

Destabilized prespore mRNAs remain in polyribosomes

The data reported in Fig. 3 suggest that although the average half-life of a given species of mRNA is greatly reduced in disaggregated cells, each mRNA molecule remains unaltered until a sudden event leads to its rapid degradation. In agreement with this interpretation, we find that prespore mRNAs known to be sensitive to cell disaggregation remain in polyribosomes during decay in dissociated cells. The evidence for this is shown in Fig. 4. Lysates of cells allowed to develop normally on filters for 17 h or disaggregated at this time and shaken in suspension for 30 min were centrifuged on sucrose density gradients (Fig. 4A). Fractions corresponding to large polyribosomes (L; over 6 –7 ribosomes per mRNA), small polyribosomes (S; 6 –7 ribosomes per mRNA), single 80S monosomes and lighter material (M) were pooled for RNA extraction. The RNAs were analysed by Northern blot hybridization with clone SC79 (constitutive) and EB4, D19 and Per97 (prespore enriched). Both in normal developing cells and in cells disaggregated and maintained in suspension for 30min, all four species of mRNA are found entirely in the polyribosome fraction (Fig. 4B). This argues that for a given population of mRNAs destabilized by disaggregation, individual mRNAs remain functional in protein synthesis until they are suddenly degraded.

Fig. 4.

Sensitive mRNA remains in polyribosomes in disaggregated cells. (A) Polysome profiles from cells at the 15 h stage of development or disaggregated at this stage and maintained in shaking suspension for 30 min are shown. (B) RNA extracted from the three regions of the gradients indicated in A were analysed by Northern hybridization with DNA from the indicated clones. In the case of Northerns to be hybridized with D19, EB4 and Per97 probes, the loading for the (Disagg) samples was approx, fourfold greater than for the (Dev.) samples to compensate for mRNA decay. Both in normal developing cells (Dev.) and disaggregated (Disagg.) cells the RNAs are found associated with large (L; greater than 6 –7 ribosomes per mRNA) and small (S; up to 6 –7 ribosomes per mRNA) polysomes. A negligible amount is present in 80S monosomes (M) and lighter material.

Fig. 4.

Sensitive mRNA remains in polyribosomes in disaggregated cells. (A) Polysome profiles from cells at the 15 h stage of development or disaggregated at this stage and maintained in shaking suspension for 30 min are shown. (B) RNA extracted from the three regions of the gradients indicated in A were analysed by Northern hybridization with DNA from the indicated clones. In the case of Northerns to be hybridized with D19, EB4 and Per97 probes, the loading for the (Disagg) samples was approx, fourfold greater than for the (Dev.) samples to compensate for mRNA decay. Both in normal developing cells (Dev.) and disaggregated (Disagg.) cells the RNAs are found associated with large (L; greater than 6 –7 ribosomes per mRNA) and small (S; up to 6 –7 ribosomes per mRNA) polysomes. A negligible amount is present in 80S monosomes (M) and lighter material.

The use of nogalamycin verifies that disaggregation dramatically reduces prespore mRNA half-lives

Previous studies have used different experimental protocols to measure mRNA half lives in normal development and after disaggregation. The conclusion that disaggregation leads to mRNA destabilization has never been verified using the same measurement procedure on both sets of cells. Since nogalamycin strongly inhibits RNA synthesis in both developing and disaggregated cells (Fig. 2) without inhibiting mRNA decay (Fig. 3), we have been able to use this drug to compare mRNA half-lives directly in each experimental situation. Fig. 5 follows the decay of several mRNAs by Northern blot analysis after the addition of nogalamycin to cells allowed to develop normally on filters for 13 h (Fig. 5A) and to cells disaggregated at this time and maintained in shaking suspension (Fig. 5B). The half-lives of the mRNAs examined in normal developing cells range from 4 – 6 h, while in disaggregated cells they fall to about 30 min for the prespore enriched D19, EB4 and Per97 mRNAs. The constitutive SC79 mRNA (Fig. 5B) and prestalk-enriched 5G mRNA (data not shown) remain relatively unaffected by the disaggregation.

Fig. 5.

Comparison of mRNA decay in developing and disaggregated cells. (A) Nogalamycin (final concentration 200 μg ml-1) was added to cells on filters at the 15 h stage of development. RNA from cells harvested at the indicated times was then analysed by Northern blot hybridization to DNA from clone SC79 (constitutive) and clones D19, EB4 and Per97 (prespore-enriched). The slight increase in EB4 mRNA at 2h may reflect incomplete inhibition of transcription of this particular mRNA in this experiment. (B) Cells at 15 h of development were disaggregated and kept in shaking suspension in the presence of 200 μg ml-1 nogalamycin. RNA from cells harvested at the indicated times was analysed by Northern blot hybridization as described in A.

Fig. 5.

Comparison of mRNA decay in developing and disaggregated cells. (A) Nogalamycin (final concentration 200 μg ml-1) was added to cells on filters at the 15 h stage of development. RNA from cells harvested at the indicated times was then analysed by Northern blot hybridization to DNA from clone SC79 (constitutive) and clones D19, EB4 and Per97 (prespore-enriched). The slight increase in EB4 mRNA at 2h may reflect incomplete inhibition of transcription of this particular mRNA in this experiment. (B) Cells at 15 h of development were disaggregated and kept in shaking suspension in the presence of 200 μg ml-1 nogalamycin. RNA from cells harvested at the indicated times was analysed by Northern blot hybridization as described in A.

RNA synthesis is probably not required for mRNA destabilization

Nogalamycin does not affect the decay of prespore A3, D19, EB4, GM55b or Per97 mRNAs upon disaggregation (Figs 3 and 5) despite the fact that it inhibits RNA synthesis by disaggregated cells essentially completely (Fig. 2), strongly suggesting the RNA synthesis is not required for mRNA destabilization.

This result is in direct contrast to the conclusions of Amara & Lodish (1987) that blocking RNA synthesis does block the destabilization of prespore mRNAs. These workers used actinomycin D and daunomycin to inhibit RNA synthesis rather than nogalamycin. Since the phenomenon may be prespore mRNA-specific, it is important to examine the decay response of the same prespore mRNAs to each of these drug regimes. Fig. 6 follows the effect of actinomycin D and daunomycin on the disaggregation-induced decay of four prespore mRNAs (D19, EB4, Per97 and A3) which we have already examined using nogalamycin (Figs 3 and 5). Cells were allowed to develop on filters for 15 h, then dissociated and maintained in suspension in the absence (Fig. 6A) or presence (Fig. 6B) of 125 μg ml-1 actinomycin D and 250 μg ml-1 daunomycin, the concentrations previously employed by Amara & Lodish (1987). Total RNA extracted from samples taken at intervals during the incubation were analysed by Northern blot hybridization. Whereas the prespore-enriched D19, EB4, Per97 and A3 mRNAs decayed upon incubation of disaggregated cells in buffer alone as expected, none of these mRNAs decayed in the presence of the antibiotic mixture. Since nogalamycin appears to inhibit RNA synthesis completely yet not affect the decay of the D19, EB4 and Per97 mRNAs (Fig. 5), it seems probable that RNA synthesis is not required for mRNA destabilization and that the block in decay caused by actinomycin D and daunomycin is a secondary effect. Alternatively, however, it is conceivable that a small subclass of RNAs exist whose synthesis is required for destabilization and that only actinomycin D and daunomycin inhibit their synthesis.

Fig. 6.

Effect of actinomycin and daunomycin on mRNA decay in disaggregated cells. Cells at 15 h of development were disaggregated and maintained in shaking suspension in PDF without drugs or in PDF-EDTA containing 125 μg ml-1 actinomycin D and 250 μg ml-1 daunomycin. Total RNA from samples taken at the indicated times was analysed by Northern blot hybridization to DNA from clones D19, EB4, Per97 and A3.

Fig. 6.

Effect of actinomycin and daunomycin on mRNA decay in disaggregated cells. Cells at 15 h of development were disaggregated and maintained in shaking suspension in PDF without drugs or in PDF-EDTA containing 125 μg ml-1 actinomycin D and 250 μg ml-1 daunomycin. Total RNA from samples taken at the indicated times was analysed by Northern blot hybridization to DNA from clones D19, EB4, Per97 and A3.

Discussion

The first indication that dissociation of Dictyostelium cell aggregates leads to destabilization of a class of mRNAs was obtained by Chung et al. (1981). They compared the average rate of decay of total mRNA in normal developing cells, treated with actinomycin D and daunomycin to block transcription, to the decay of both total mRNA and a single species of mRNA in disaggregated cells in the absence of drugs. Since, as we show here and has been reported elsewhere (Amara & Lodish, 1987), actinomycin and daunomycin stabilize mRNA in disaggregated cells, one cannot exclude the possibility that these drugs have a similar effect on normal developing cells. As an alternative approach, we studied the incorporation of 32P into specific mRNAs during development and the decay of the same labelled mRNAs in disaggregated cells (Mangiarotti et al. 1982). Although it was not possible to obtain an accurate measure of mRNA half-life in the developing cells, from our data it was apparent that cell dissociation destabilized a class of mRNAs by at least a factor of 5. This conclusion was questioned by Casey et al. (1983). From a similar analysis on total polyadenylated mRNA, their study found that most mRNA is relatively stable and a minor fraction is unstable both in aggregated and disaggregated cells. They suggested that the latter fraction might include the specific mRNAs studied by us and that we may have overestimated their stability in normal developing cell aggregates.

In a more recent study, Manrow & Jacobson (1988) used a 32P pulse-chase protocol to analyse the rate of decay of individual mRNA species both in aggregated cells and in cells disaggregated in the presence and absence of cAMP. During the chase in aggregated cells, the amount of 32P label decreased slowly in some of the mRNA studied while it remained constant or even increased in other mRNAs. In cells disaggregated without cAMP, the amount of 32P label decreased very slowly in constitutive and non-cell-type-specific mRNAs but decayed with half-lives of 20 to 30 min in cell-type-specific mRNAs, at least during the first 1-1 -5 h after disaggregation. These results are quantitatively similar to those previously reported by us (Mangiarotti et al. 1982). However, Jacobson (1988) accept that their 32 since Manrow & P chase was not completely effective in aggregated cells, it is difficult to quantify the effect of cell disaggregation on Mrna stability on the basis of their data. Nevertheless, Manrow & Jacobson (1988) conclude that cell-type-specific mRNAs do indeed undergo ‘a short-term labilization’ upon disaggregation in the absence of cAMP, in agreement with the data presented by us here and elsewhere (Mangiarotti et al. 1982, 1983, 1985). It is therefore curious and apparently contradictory that these authors also state that the decrease in mRNA abundance upon disaggregation reflects their ‘inherently fast normal rates of decay’.

Manrow & Jacobson (1988) also question our earlier finding that cAMP stabilizes cell-type-specific mRNAs in disaggregated cells, a conclusion we based not only on pulse-chase data but also on studies using approach to steady-state labelling (Mangiarotti et al. 1983, 1985). In fact, the quantitative effect of cAMP on mRNA stability is difficult to evaluate from their data since the effectiveness of the chase varied between aggregated and disaggregated cells and some prespore mRNAs exhibited first order decay in the disaggregated cells in the presence of cAMP but biphasic decay in the absence of cAMP. Nevertheless, all the prespore mRNAs tested decayed at a substantially lower rate in the 2 hour period following disaggregation when incubated in the presence of cAMP than in its absence. The heterogeneity of the kinetics of mRNA decay observed by Manrow & Jacobson (1988) may indicate that the effect of cAMP on mRNA stability varies with different mRNAs.

As we show in this paper, the uncertainties inherent in previous radiolabelling studies of mRNA decay can be eliminated by the use of nogalamycin, which has no effect upon the half-lives of mRNA in disaggregated cells. Since the mRNA species that we have studied are in the same physical state, undergoing translation in polysomes, both in aggregated and disaggregated cells, it is very unlikely that mRNA stability is altered by the same drug in aggregated cells. Nogalamycin can therefore be used to measure the effect of cell disaggregation on mRNA half-life directly. From the data reported here, it appears that our previous estimates (as well as those of Chung et al. 1981) were correct; the half-life of the prespore mRNAs examined decreases from 4 – 6 h in normal developing cells to 20 – 30 min upon disaggregation. It is interesting that if disaggregated cells are allowed to reaggregate after the rapid decay of prespore mRNAs has already begun, the decay is promptly halted. This observation strengthens the relationship between the multicellular state and mRNA stability. These data are also in line with our previous report that some mRNAs are transcribed from the beginning of development but do not accumulate in the cell because they are unstable. Their stability increases and they begin to accumulate at the aggregation stage (Mangiarotti et al. 1985).

The fact that reaggregation promptly halts any further rapid decay of sensitive prespore mRNAs provides some insight into the mechanisms involved. The data rule out the possibility that upon disaggregation all molecules of a sensitive mRNA species are modified irreversibly, for example by shortening of the 3’ poly (A) tail or removal of 5’ terminal sequences, and that this then eventually leads to degradation of all of the mRNA molecules at a common rapid rate. Rather, it seems likely that whatever factors cause the preferential degradation of this class of mRNA, they act on the mRNA molecules at random, quickly causing the degradation of each molecule soon after it is selected. By this model and taking into account molecules not yet selected, the degradation event itself must be extremely rapid, yielding an mRNA half-life substantially less than the 20 – 30 min measured for that mRNA population as a whole. This model of disaggregation-induced decay is in line with our other finding that all molecules of sensitive mRNA left at a given time in disaggregated cells are associated with polyribosomes, presumably still functioning in protein synthesis.

The nature of the event that selects individual mRNAs for rapid degradation is still unknown. Our nogalamycin data indicate that RNA synthesis is not required to trigger this event. If this is correct, the blocking action of actinomycin D and daunomycin must be a secondary effect due to the action of these drugs on aspects of cell metabolism other than transcription. We cannot exclude the remote possibility that nogalamycin may fail to inhibit the synthesis of a minor species of RNA, essential for induction of mRNA decay, which is inhibited by actinomycin D and daunomycin. However, it is difficult to envisage how RNA synthesis can be sufficiently rapid to play a major role in inducing the destabilization of sensitive mRNAs, which begins only 10 min after disaggregation (Amara & Lodish, 1987).

One way in which destabilization of sensitive mRNAs might occur is via their exclusion from polyribosomes. In this case, the primary control would be at the level of mRNA translation rather than mRNA stability per se. This possibility is not excluded by our data which show only that, if such an event occurs, it does not involve all molecules of sensitive mRNA at the time of cell dispersion, but must occur progressively. Cycloheximide reduces the rate of mRNA decay upon cell disaggregation (Amara & Lodish, 1987; P. Morandini, G. Mangiarotti and A. Ceccarelli, manuscript in preparation) suggesting either that the translation of particular pre-existing mRNA is necessary to induce the decay of sensitive prespore mRNAs or that the movement of ribosomes along the sensitive mRNAs themselves is required for destabilization. Interestingly, the degradation of histone mRNA is also prevented by inhibiting protein synthesis (Sive et al. 1984; Stimac et al. 1984) and has been shown to depend on the translatability of the histone mRNA itself (Sive et al. 1984). Similarly, destabilization of tubulin mRNA depends upon recognition of the first four amino acids of α-tubulin as they emerge from the ribosome (Yen et al. 1988). However, the role of translation in Dictyo-stelium mRNA decay is stiil unclear. Thus puromycin, which inhibits protein synthesis in Dictyostelium cells almost as well as cycloheximide (Amara & Lodish, 1987), and canavanine, [an arginine analogue which leads to the synthesis of faulty proteins (P. Morandini, G. Mangiarotti and A. Ceccarelli, manuscript in preparation] fail to prevent the decay of sensitive Dictyostelium mRNAs upon disaggregation in contrast to the effect of cycloheximide. The demonstration that during disaggregation the residual mRNA is still associated with polysomes is in contrast to the data of Steel & Jacobson (1987) on ribosomal protein mRNA early in development and so probably represents a different mechanism of post-transcriptional control.

Mehdy et al. (1983) have asserted that disaggregation leads to rapid loss of both prespore- and prestalk-enriched mRNAs although one of the two prestalk mRNAs examined, 16-G1 (pst-cath), seemed fairly stable. Manrow & Jacobson (1988) also found that three prestalk mRNAs were destabilized upon disaggregation. Barklis & Lodish (1983) and Chishoim et al. (1984) noted that prestalk-enriched mRNAs fell into two distinct classes. Prestalk I mRNAs (e.g. Cl, Dll and D14) are already expressed in preaggregation cells and are fairly stable upon cell disaggregation whereas prestalk II mRNAs (e.g. Al and PL1) are expressed only later in development and become unstable upon disaggregation. Of the clones examined in the present study, the major characteristic that appears to correlate with mRNA sensitivity to disaggregation is cell-specificity, not the time of expression. Thus some of the prestalk-enriched clones (Per74, Per79, 5D) are late genes (Giorda, R. & Mangiarotti, G., unpublished observations; Corney et al. 1989) but their stability is largely unaffected by disaggregation. In contrast, all the prespore-enriched mRNAs examined were sensitive to cell dissociation. How general the correlation between cell specificity and sensitivity to disaggregation is we cannot say. However, we would suggest that constitutive and many prestalk-enriched genes may be controlled only at the level of transcription whereas prespore cells appear to have developed a second major type of control at the post-transcriptional level for all genes expressed preferentially in this cell type. Attempts to analyse this latter regulation using in vitro systems are now in progress.

This work was supported by funds from Italian CNR (Progetto finalizzato di Ingneria Genetica e Gruppo Nazionale di Biologia Molecolare, Cellulare e Evolutiva) and M.P.I. We thank E. Barklis for the generous gift of clones EB4, Per79 and Per97.

Alton
,
T. H.
&
Brenner
,
M.
(
1979
).
Comparison of proteins synthesised by anterior and posterior regions of Dictyostelium discoideum pseudoplasmodia
.
Devi Biol
.
71
,
1
8
.
Alton
,
T. H.
&
Lodish
,
H. F.
(
1977
).
Synthesis of developmentally regulated proteins in Dictyostelium discoideum which are dependent on continued cell-cell interaction
.
Devi Biol
.
60
,
207
216
.
Amara
,
J. F.
&
Lodish
,
H. F.
(
1987
).
Specific mRNA destabilisation in Dictyostelium discoideum requires RNA synthesis
.
Molec. cell. Biol
.
7
,
4585
4588
.
Barklis
,
E.
&
Lodish
,
H. F.
(
1983
).
Regulation of Dictyostelium discoideum mRNAs specific for prestalk cells
.
Cell
32
,
1139
1148
.
Blumberg
,
D.
&
Lodish
,
H. F.
(
1980
).
Changes in the messenger RNA population during differentiation of Dictyostelium discoideum
.
Devi Biol
.
78
,
285
300
.
Borth
,
W.
&
Ratner
,
D.
(
1983
).
Different synthetic profiles and developmental fates of prespore versus prestalk proteins of Dictyostelium
.
Differentiation
24
,
213
219
.
Bozzaro
,
S.
,
Perlo
,
C.
,
Ceccarelli
,
A.
&
Mangiarotti
,
G.
(
1984
).
Regulation of gene expression in Dictyostelium discoideum cells exposed to immobilised carbohydrates
.
EMBO J
.
3
,
193
200
.
Casey
,
L.
,
Palatnik
,
C. M.
&
Jacobson
,
A.
(
1983
).
Messenger RNA half life in Dictyostelium discoideum
.
Devi Biol.
95
,
239
256
.
Chisholm
,
R. L.
,
Barklis
,
E.
&
Lodish
,
H. F.
(
1984
).
Mechanism of sequential induction of cell-type specific mRNAs in Dictyostelium differentiation
.
Nature, Lond
.
310
,
67
69
.
Chung
,
S.
,
Landfear
,
S. M.
,
Blumberg
,
D. D.
,
Cohen
,
N. S.
&
Lodish
,
H. F.
(
1981
).
Synthesis and stability of developmentally regulated mRNAs are affected by cell-cell contact and cAMP
.
Cell
24
,
785
797
.
Corney
,
A. J.
,
Richards
,
A. J.
,
Phillpots
,
T.
&
Hames
,
B. D.
(
1989
).
Developmental regulation of cell-type enriched mRNAs in Dictyostelium discoideum
.
Molecular Microbiol, (in press)
.
Ennis
,
H.
(
1981
).
Nogalamycin inhibits ribonucleic acid synthesis in growing and developing cells of the slime mould, Dictostelium discoideum
.
Antimicrobial Agents and Chemotherapy
19
,
657
665
.
Jacquet
,
M.
,
Part
,
D.
&
Felenbok
,
B.
(
1981
).
Changes in the polyadenylated messenger RNA population during development of Dictyostelium discoideum
.
Devi Biol
.
81
,
155
156
.
Landfear
,
S. M.
,
Lefebvre
,
P.
,
Chung
,
S.
&
Lodish
,
H. F.
(
1982
).
Transcriptional control of gene expression during development of Dictyostelium discoideum
.
Molec. cell. Biol
.
2
,
1417
1426
.
Landfear
,
S. M.
&
Lodish
,
H. F.
(
1980
).
Cyclic AMP: a role in expression of developmentally regulated genes in Dictyostelium discoideum
.
Proc. natn. Acad. Sci. U.S.A
.
77
,
1044
1053
.
Mangiarotti
,
G.
,
Altruda
,
F.
&
Lodish
,
H. F.
(
1981
).
Rates of synthesis and degradation of ribosomal ribonucleic acid during differentiation of Dictyostelium discoideum
.
Molec. cell. Biol
.
1
,
35
42
.
Mangiarotti
,
G.
,
Ceccarelli
,
A.
&
Lodish
,
H. F.
(
1983
).
Cyclic AMP stabilises a class of developmentally regulated Dictyostelium discoideum mRNAs
.
Nature, Lond
.
301
,
616
618
.
Mangiarotti
,
G.
,
Chung
,
S.
,
Zuker
,
C.
&
Lodish
,
H. F.
(
1980
).
Selection and analysis of clones for developmentally-regulated Dictyostelium discoideum genes by hybridisation-competition
.
Nucleic Acid Res
.
9
,
947
965
.
Mangiarotti
,
G.
,
Giorda
,
R.
,
Ceccarelli
,
A.
&
Perlo
,
C.
(
1985
).
mRNA stabilisation controls the expression of a class of developmentally regulated genes in Dictyostelium discoideum
.
Proc. natn. Acad. Sci. U.S.A
.
82
,
5786
5790
.
Mangiarotti
,
G.
,
Lefebvre
,
P.
&
Lodish
,
H. F.
(
1982
).
Differences of mRNA stability in aggregated and disaggregated Dictyostelium discoideum cells
.
Devi Biol
.
89
,
82
91
.
Manrow
,
R. E.
&
Jacobson
,
A.
(
1988
).
mRNA decay rates in late-developing Dictyostelium discoideum cells are heterogeneous and cyclic AMP does not act directly to stabilise cell-type specific mRNAs
.
Molec. cell. Biol
.
8
,
4088
4097
.
Mehdy
,
M. C.
,
Ratner
,
D.
&
Firtel
,
R. A.
(
1983
).
Induction and modulation of cell-type specific gene expression in Dictyostelium discoideum
.
Cell
32
,
763
771
.
Morrissey
,
J. H.
,
Devine
,
K. M.
&
Loomis
,
W. F.
(
1984
).
The timing of cell-type specific differentiation in Dictyostelium discoideum
.
Devi Biol
.
103
,
414
424
.
Morrissey
,
J. H.
,
Farnsworth
,
P. A.
&
Loomis
,
W. F.
(
1981
).
Pattern formulation in Dictyostelium discoideum: An analysis of mutants altered in cell proportioning
.
Devi Biol
.
83
,
1
13
.
Newell
,
P. C.
,
Franke
,
J.
&
Sussman
,
M.
(
1972
).
Regulation of four functionally related enzymes during shifts in the developmental program of Dictyostelium discoideum
.
J. molec. Biol
.
63
,
373
383
.
Newell
,
P. C.
,
Longlands
,
M.
&
Sussman
,
M.
(
1971
).
Control of enzyme synthesis by cellular interaction during development of the cellular slime mould Dictyostelium discoideum
.
J. molec. Biol
.
58
,
541
554
.
Rigby
,
P. W. J.
,
Diekmann
,
M.
,
Rhodes
,
C.
&
Berg
,
P.
(
1977
).
Labelling deoxyribonucleic acid to high specific activity in vitro by nick translation with DNA polymerase I
.
J. molec. Biol
.
113
,
237
251
.
Sive
,
H. L.
,
Heintz
,
N.
&
Roeder
,
R. G.
(
1984
).
Regulation of human histone gene expression during the HeLa cell cycle requires protein synthesis
.
Molec. cell. Biol.
4
,
2723
––
2090
.
Steel
,
L. F.
&
Jacobson
,
A.
(
1987
).
Translational control of ribosomal protein synthesis during Dictyostelium discoideum development
.
Molec. cell. Biol
.
7
,
965
972
.
Stimac
,
E.
,
Groppi
,
V. E.
&
Coffino
,
P.
(
1984
).
Inhibition of protein synthesis stabilizes histone mRNA
.
Molec. cell. Biol.
4
,
2082
2090
.
Watts
,
D. J.
&
Ashworth
,
J. M.
(
1970
).
Growth of myxamoebae of the cellular slime mould Dictyostelium discoideum in axenic culture
.
Biochem. J
.
119
,
171
174
.
Williams
,
J. H.
,
Ceccarelli
,
A.
,
Mcrobbie
,
J. J.
,
Mahbubani
,
H.
,
Kay
,
R. R.
,
Early
,
A. E.
,
Berks
,
M.
&
Jermyn
,
K. A.
(
1987
).
Direct induction of Dictyostelium prestalk gene expression by DIF provides evidence that DIF is a morphogen
.
Cell
49
,
185
192
.
Yen
,
T. J.
,
Machlin
,
P. S.
&
Cleveland
,
D. W.
(
1988
).
Autoregulated instability of cr-tubulin mRNAs by recognition of the nascent amino terminus of a-tubulin
.
Nature, Lond
.
334
,
580
585
.