(Received for publication, December 24, 1996, and in revised form, May 21, 1997)
From the Department of Clinical and Biological Sciences, University of Turin, Ospedale San Luigi Gonzaga, Regione Gonzole 10, 10043 Orbassano-Torino, Italy
40 and 60 S ribosomal subunits have been reconstituted in vitro from purified ribosomal RNA and ribosomal proteins of Dictyostelium discoideum. The functionality of the reconstituted ribosomes was demonstrated in in vitro mRNA-directed protein synthesis. The reassembly proceeded well with immature precursors of ribosomal RNA but poorly if at all with mature cytoplasmic RNA species. Reassembly also required a preparation of small nuclear RNA(s), acting as morphopoietic factor(s).
In vitro reconstitution of functional ribosomal subunits from free Escherichia coli rRNA and proteins was reported almost 30 years ago (1-3). More recently it has been accomplished also with archaebacteria (4, 5). However, all attempts to reconstitute eukaryotic ribosomes have failed, although their biogenesis in vivo has been well characterized (6-13). Here we describe the in vitro reassembly of functional ribosomal subunits from free ribosomal RNA and proteins of Dictyostelium discoideum.
The pioneer work of Nomura and co-workers (1, 2) has established for the first time that the information sufficient to assemble a cell organelle are contained in its components. It has been extremely useful in studying the function of single ribosomal proteins both in ribosome assembly and in their interactions with the other components of the translation machinery. Nevertheless, the experimental conditions used to reconstitute bacterial ribosomal subunits in vitro were probably much different from those in which the assembly process occurs in vivo. We have shown that a step requiring high energy and that could be accomplished in vitro only by the exposure of the reconstitution mixture to high, nonphysiological temperatures (1, 2) was probably not the limiting step in ribosome assembly in vivo (14). Although ribosome assembly also occurs in E. coli cells in discrete steps (15), the high energy requirement is probably overcome in vivo by some factor(s) not present among the components derived from mature 70 S ribosomes.
We and others have found that in E. coli newly formed 30 S ribosomal subunits still contain immature pre-16 S rRNA (16, 17). The extra sequences present in immature rRNA must facilitate the assembly process, since in vitro pre-16 S rRNA, but not mature 16 S rRNA, was incorporated instantaneously into 30 S particles in the presence of ribosomal proteins even at 0 °C (18).
Our interest in the in vitro reconstitution of eukaryotic ribosomes has been prompted by the finding that during development of D. discoideum the stability of a class of mRNAs is controlled by a mechanism that involves the modification of one or several components of 40 S ribosomal subunits (19-21).1,2 In trying to accomplish the reconstitution of eukaryotic ribosomes, we took advantage of our previous studies on ribosome assembly in E. coli. Here we report that D. discoideum immature precursor rRNAs are also better substrates than mature cytoplasmic 17 and 26 S rRNAs for the in vitro reconstitution process. Furthermore, a nuclear component, probably one or several small nucleolar RNAs, which are not present in 40 or 60 S ribosomal particles, plays an essential role.
Dictyostelium discoideum AX2 cells were grown as described previously (22).
Differential Radioactive Labeling and Isolation of Nuclear and Cytoplasmic rRNAGrowing cells at a density of 2 × 106/ml were washed twice by centrifugation in Sorensen
buffer and resuspended in the same buffer at a concentration of 2 × 108 cells/ml. In a typical experiment, 200 µCi of
[H3]uracil were added to 10 ml of cell suspension for 30 min. Cells were then pelleted and lysed in ice-cold 25 mM
HEPES/KOH, pH 7.5, 25 mM potassium acetate, 10 mM magnesium acetate, 5% sucrose, and 2% Tergitol (buffer
A). The lysate was spun in an Eppendorf microcentrifuge at top speed
for 20 s to remove cell debris and then for 5 min to pellet the
nuclei, which contained most of the incorporated label. Nuclei were
lysed by resuspension in buffer A not containing sucrose and Tergitol,
and the lysate was centrifuged in a Beckman SW40 rotor at 170,000 × g for 18 h at 4 °C through a 15-35% sucrose
gradient in 20 mM Tris-HCl, pH.7.8, 0.5 M
NH4Cl, 50 mM magnesium acetate, 6 mM -mercaptoethanol (buffer B). Fractions sedimenting at
40 and 60 S were collected separately, ethanol-precipitated, resuspended in buffer B, and centrifuged again through a sucrose gradient to isolate pure 40 and 60 S particles. From each of these preparations, RNA was extracted with Ultraspec-II RNA (Biotecx Laboratories), following the procedure suggested by the manufacturer. RNA (about 3-4 µg from 40 S particles and 6-8 µg from 60 S
particles) was ethanol-precipitated and redissolved in 10 mM Tris-HCl, pH 7.5, 1 mM EDTA (TE buffer).
One mCi of [14C]uracil was added to another 10-ml cell suspension for 8 h, followed by a large excess of unlabeled uracil for 2 h. Cells were lysed, nuclei and cytoplasm were separated, and rRNA was extracted from isolated cytoplasmic 40 and 60 S ribosomal subunits as described above.
The effectiveness of nuclei and cytoplasm separation was verified by the fact that nuclei did not contain any 14C-labeled RNA, while cytoplasm did not contain any 3H-labeled 36 S RNA (the large precursor of ribosomal RNAs).
Gel Electrophoretic Analysis of rRNA3H-Labeled (nuclear) and 14C-labeled (cytoplasmic) rRNAs, isolated as described above, were mixed together in an appropriate ratio, denatured by heating at 65 °C in 50% formamide for 5 min, and analyzed by electrophoresis on 5% polyacrylamide gels in 6 M urea. At the end of the run, the gel was cut into 1.5-mm slices, which were counted in Triaton solution (Packard). 3H-Labeled rRNA was also extracted from in vitro reconstituted ribosomal particles isolated by sedimentation through sucrose gradients, and it was analyzed by gel electrophoresis as described above.
Fingerprint Analysis of Precursor and Mature rRNAs108 cells were starved, plated on a 2.8-cm Millipore filter embedded with 10 mCi of 32P in MES-PDF3 buffer (19). After 2 h, cells were collected and lysed, and RNA was extracted from nuclear and cytoplasmic 40 and 60 S particles as described above. Each RNA preparation was sedimented through sucrose gradients to display nuclear precursor 19 and 28 S RNAs and cytoplasmic mature 17 and 26 S RNAs (23-25). Each fraction was concentrated by ethanol precipitation, digested with RNase T1, and analyzed as described by Batts-Young et al. (25).
In Vitro Labeling and Analysis of a 7 S Nuclear RNA FractionTotal RNA was extracted with Ultraspec-II RNA from
nuclei derived from 109 unlabeled growing cells,
ethanol-precipitated, dissolved in TE buffer, and chromatographed on an
oligo(dT)-cellulose column to eliminate poly(A)+ RNA, as
described previously (19). Poly(A) RNA was fractionated
on a 15-35% sucrose gradient containing 20 mM Tris-HCl,
pH 7.3, and 0.5 M NH4Cl in a Beckman SW40 rotor at 170,000 × g for 36 h to display 19 S RNA and
smaller RNAs. Each fraction from the top half of the gradient was
tested for its ability to support in vitro ribosome
reconstitution. A fraction moving at about 7 S, which turned out to be
active, was labeled with [32P]pCp (26) and analyzed by
electrophoresis on a 8% polyacrylamide gel in 6 M urea,
followed by autoradiography.
The nuclear
RNA fraction sedimenting at about 7 S and labeled in vitro
with [32P]pCp was separated by gel electrophoresis into
two major bands. A faint band was visible in front of each of the two
major bands, while another 13 bands appeared after an exposure longer
than the one shown in Fig. 5. However, the two major bands in the
latter case obscured part of the gel. Therefore, the number of
different small nuclear or nucleolar RNAs present in the fraction
sedimenting at about 7 S could not be determined precisely. RNA was
extracted as described in Ref. 24 from the regions of the gel
corresponding to each of the two major bands and the regions containing
the faint bands. The RNA corresponding to each region of the gel was tested both for its ability to substitute for nuclear lysate in supporting in vitro reconstitution of ribosomal subunits
(see below) and to hybridize to cloned DNAs containing rRNA genes.
Hybridization of 7 S Nuclear RNA Fraction to Ribosomal DNA
Phage DNAs each containing a different gene for rRNAs (17, 5.8, 26, and 5 S rRNA) were cloned with the hybridization competition technique described by Mangiarotti et al. (22). The 7 S nuclear RNA corresponding to different regions of the gel shown in Fig. 5 was hybridized to each of the cloned DNA containing an rRNA gene, in the absence or in the presence of an excess of the corresponding unlabeled rRNA species, following the procedure described in Ref. 19.
Isolation of NucleoliNucleoli were isolated following the procedure described in Ref. 27, as modified by Frankel et al. (24).
Preparation of Ribosomal ProteinsGrowing cells were
collected and lysed in buffer B containing 2% Tergitol. Ribosomal 40 and 60 S subunits were isolated by centrifugation through sucrose
gradients as described above. To extract ribosomal proteins, ribosomal
subunits (50 A260/ml) were adjusted to 0.1 M magnesium acetate, and 2.2 volumes of glacial acetic acid
were added dropwise with constant stirring in the cold, as described in
Ref. 28. Stirring was continued for 1 h, and rRNA was removed by
centrifugation at 27,000 × g for 10 min. The extracted
proteins were dialyzed against 300 volumes of 6 M urea, 6 mM -mercaptoethanol for 18 h in the cold.
To facilitate the analysis of ribosomal proteins by two-dimensional gel electophoresis, growing cells were resuspended in Sorensen buffer and labeled with 250 µCi of [35S]methionine and cysteine for 3 h. The labeled proteins were extracted as described above from native or in vitro reconstituted 40 and 60 S ribosomal subunits and analyzed as described by Ramagopal and Ennis (28).
In Vitro Reconstitution of Ribosomal SubunitsIn a typical
experiment, 10 µg of 40 S ribosomal proteins in 0.2 ml of 6 M urea were mixed with 4 µg of 3H-labeled 19 S RNA or 14C-labeled 17 S RNA in 0.1 ml of TE buffer and 20 µg of 60 S ribosomal proteins in 0.4 ml of 6 M urea were
mixed with 8 µg of 3H-labeled 5, 5.8, and 28 S RNAs or
14C-labeled 5, 5.8, and 26 S RNAs in 0.2 ml of TE buffer.
The RNA and protein mixtures were dialyzed at 23 °C for 12 h
against a buffer containing 20 mM Tris-HCl, pH 7.3, 0.5 M NH4Cl, 20 mM
MgCl2, 10 mM -mercaptoethanol.
Growing cells were washed twice in Sorensen buffer and lysed in buffer A. The lysate was spun in an Eppendorf microcentrifuge at top speed for 20 s to remove cell debris and then for 5 min to pellet nuclei. The nuclei were lysed by resuspension in TE buffer and centrifuged in a 55 Ti rotor in a TL Beckman ultracentrifuge at 250,000 × g for 2 h to discard any particle the size of ribosomal subunits or larger. When nuclear lysate was added to ribosomal subunit reconstitution mixtures, the amount of nuclear lysate and the amount of rRNA utilized were derived from a comparable number of cells.
Preparation of a Cell-free System for Protein SynthesisCells were lysed in buffer A. The lysate was spun in an Eppendorf microcentrifuge at top speed for 10 min to pellet microsomes and lysosomes. The supernatant was centrifuged in a 55 Ti rotor in a TL Beckman ultracentrifuge at 250,000 × g for 5 h to pellet all ribosomal particles. The supernatant (S55) was dialyzed for 3 h against buffer A not containing sucrose and Tergitol. S55 represented the source of all factors required for protein synthesis except ribosomes and mRNA.
Amino Acid Incorporation SystemProtein synthesis was measured at 23 °C for 15 min in a total reaction volume of 100 µl containing 10 µg of S55 proteins in the presence of the following components: 20 mM HEPES/KOH, pH 7.5, 100 mM potassium acetate, 15 mM magnesium acetate, 2 mM dithiothreitol, 0.5 mM spermidine, 2 mM ATP, 0.5 mM GTP, 20 mM creatine phosphate, 3 µg of creatine phophokinase, 250 mM each of 19 amino acids (excluding methionine), 200 µg of Dictyostelium tRNA, 20 µCi of [35S]methionine. Poly(A)+ RNA, prepared as described by Mangiarotti et al. (19), was added in the amount of 10 µg. The composition of the system was such that the rate of incorporation of [35S]methionine over the 15-min period of incubation increased linearly with the amount of native ribosomes added to the reaction mixture. The linearity was checked up to 40 µg of ribosomes/100 µl. Since we used much lower amounts of ribosomes in the experiment described in Table I, the data reported should reflect exclusively the functional capacity of the tested ribosomes without the interference of any other limiting factor.
|
Nuclear and cytoplasmic ribosomal particles were labeled
differentially with [3H]uracil and
[14C]uracil. Nuclear particles cosedimenting with
cytoplasmic 40 and 60 S ribosomal subunits could be isolated. RNA
contained in these particles was analyzed by gel electrophoresis (Fig.
1A). Nuclear 40 S particles
contained an RNA that was larger than the cytoplasmic 17 S rRNA, and
nuclear 60 S particles contained 5 and 5.8 S RNA plus a molecule that
was larger than the cytoplasmic 26 S rRNA. We call the two large RNAs
19 and 28 S from their sedimentation rate through the sucrose gradient
(Fig. 3). These molecules are similar in size to the precursor rRNAs
described previously in D. discoideum (23-25), although
these were isolated from whole nuclei. To determine whether they were
also similar by sequence, we have analyzed the RNase T1 fingerprint of
32P-labeled rRNAs present in nuclear and cytoplasmic 40 and
60 S particles (Fig. 2). The patterns
obtained were practically superimposable to those reported in Ref.
25.
Ribosome Assembly Requires Pre-rRNAs and Morphopoietic Factors
When the labeled rRNAs were incubated with 40 and 60 S ribosomal subunit proteins under different salt and temperature conditions, no particles sedimenting at 40 or 60 S were formed (Fig. 3A). However, when a small amount of nuclear lysate depleted of ribosomal subunits (see "Experimental Procedures") was added to the incubation mixture, 3H-labeled 19 S RNA was incorporated into particles sedimenting as 40 S and 3H-labeled 5, 5.8, and 28 S RNAs were incorporated into particles sedimenting as 60 S with an efficiency of 80-90%, while 14C-labeled mature RNAs were incorporated into similar particles with an efficiency of at most 10% (Fig. 3B). The amount of nuclear lysate required was that derived from a number of cells comparable with the one from which the rRNA had been derived. The amount of ribosomal proteins needed was twice that of the rRNA, probably because some proteins were partially denatured during the extraction procedure.
Ribosome reconstitution therefore proceeds more efficiently with
nuclear immature rRNA than with cytoplasmic mature rRNA, and a nuclear
component plays an important role in the assembly process. The required
nuclear lysate could be replaced by an RNA fraction that was extracted
from nuclei with guanidine and phenol and sedimented at about 7 S in
zonal sedimentation in a sucrose gradient (Fig.
4). The 7 S fraction contained two major
RNA species plus several minor species, which could be separated by gel
electrophoresis (Fig. 5, right
lane). The minor species are only partially visible in Fig. 5, but
they became visible with a longer exposure of the gel. The number of
small nuclear or nucleolar species present in the 7 S fraction was at
least 17, but we cannot exclude the possibility that several species
comigrated in our gel, masking an even greater complexity of this RNA
fraction.
The 7 S Fraction Contains a Precursor of 5.8 S rRNA and Many Small Nuclear RNAs
The RNAs corresponding to each of the two bands
shown in Fig. 5 (right lane) and to the faint bands were
recovered from the gel and hybridized to cloned DNAs containing the
gene of 17, 5.8, 26, or 5 S rRNA. The RNA corresponding to the upper
major band hybridized only to DNA containing the 5.8 S RNA gene, but
its hybridization was completely prevented by the addition of an excess of unlabeled 5.8 S rRNA (Fig. 6). The RNA
corresponding to the lower major band and to the faint bands did not
hybridize to any ribosomal DNA. Thus, the first RNA is probably a
precursor of 5.8 S rRNA, not previously detected in D. discoideum but already described in other eukaryotic species (29,
30), while the other RNAs must be small nuclear RNAs.
The Morphopoietic Factor Is Probably a Small Nucleolar RNA
The RNAs corresponding to the two major bands shown in Fig. 5
(right lane) and to the faint bands were separately tested
for their ability to promote ribosomal particle reconstitution in the
absence of nuclear lysate. In the presence of the RNA corresponding to
the upper major band, no ribosomal particles formed. Thus, the
requirement for the 7 S nuclear fraction is not due to the requirement
for an immature 5.8 S rRNA. In the presence of the RNA corresponding to
the lower major band, nuclear 3H-labeled rRNA was
incorporated into 40 and 60 S particles to an extent comparable to that
with nuclear lysate (Fig. 7). Thus, the
small nuclear RNA corresponding to the lower major band in Fig. 5 can
fully replace the nuclear lysate in the reconstitution system. In the
presence of the RNAs corresponding to the faint bands, no ribosomal
particles formed.
The active small nuclear RNA fraction has not yet been further characterized. Though it gives origin to only one band by gel electrophoresis analysis, it might contain several small nuclear RNAs that comigrate under our electrophoretic conditions. Thus, although the small nuclear RNA fraction is required for the reconstitution of both 40 and 60 S ribosomal subunits, we cannot know whether this function is carried out by the same RNA species.
We cannot determine the stoichiometry of the small nuclear RNA and of
the rRNA in the assembly process, because we cannot detect the first
RNA by the A260 absorbance in the preparative sucrose gradient. However, since we have labeled with
[32P]pCp the 7 S RNA fraction in parallel with known
amounts of 5 and 5.8 S RNAs, we can roughly estimate that the labeled
major 7 S lower band contained 103 µg of RNA, judging
by the amount of the other two RNAs run on the gel to obtain bands of
intensity comparable with the one of the 7 S RNA. Thus, the
stoichiometry of the assembly reaction should be 1 small nuclear
RNA/100-200 rRNA molecule species. It is evident that the 7 S RNA
functions catalytically in stimulating ribosome assembly.
We have repeated the experiments shown in Figs. 4 and 7 with RNA extracted from nucleoli. This RNA contained a fraction that sedimented at 7 S on a sucrose gradient, could be labeled with [32P]pCp, and gave an electrophoretic pattern similar to the one shown in Fig. 5 (right lane). The lower major band could support the in vitro reconstitution of 40 and 60 S ribosomal subunits with the same efficiency as the 7 S nuclear band in the experiment of Fig. 7. This suggests that the active nuclear 7 S fraction is probably a small nucleolar RNA.
In Vitro Reconstituted Ribosomal Subunits Contain a Full Complement of Ribosomal ProteinsRibosomal proteins labeled with
[35S]methionine and cysteine were extracted from
cytoplasmic 40 and 60 S subunits and analyzed by two-dimensional gel
electrophoresis (Fig. 8, A and
C). An aliquot of these proteins was used to reconstitute
in vitro 40 and 60 S particles starting from nuclear
immature rRNA and in the presence of the 7 S small nuclear RNA
fraction. As shown in Fig. 8, B and D, the
reconstituted particles contained all of the ribosomal proteins present
in native subunits at comparable amounts.
In Vitro Reconstituted Ribosomal Subunits Function in Protein Synthesis
To test their functionality, the 3H-labeled particles reconstituted in the presence of nuclear lysate or of the 7 S small nuclear RNA were incubated in a cell-free system for protein synthesis (31) lacking endogenous ribosomes and programmed with poly(A)+ RNA extracted from Dictyostelium cells. We tested native subunits alone and several amounts of each species of reconstituted particles in the presence of an excess of the other species of native subunit (Table I). As we mentioned under "Experimental Procedures," the added ribosomes are the only limiting factor of the rate of [35S]methionine incorporation. The relative activity of the tested ribosomes can therefore be directly compared. With all types of reconstituted particles, the efficiency of incorporation was 85-90% of that obtained with native ribosomal subunits.
In additional experiments, the 3H-labeled ribosomal
subunits reconstituted in the presence of nuclear lysate or of 7 S
small nuclear RNA entered polyribosomes in a cell-free system to the same extent as native cytoplasmic ribosomal subunits (Fig.
9). The three kinds of subunits entered
polyribosomes at an efficiency of 50-60%, a level close to the one
found in growing cells. Thus, the original activity of the subunits
used for the reconstitution experiment, the efficiency of the
reassembly, and the functionality of reconstituted particles were
comparable with those obtained with E. coli ribosomes
(1).
Data to be published elsewhere2 show that the in vitro reconstituted 40 S ribosomal subunits are also functional in an in vitro test of their ability to destabilize specific mRNAs, when their protein components are obtained from cells at a developmental stage in which the same mRNAs are highly unstable in vivo.
Pre-rRNA Does Not Mature during the Reconstitution ProcessSince in vitro reconstitution of ribosomal subunits requires small nuclear RNA and small nuclear (or nucleolar) RNA is known to be involved in rRNA processing (32-38), one possibility to test was that ribosome reconstitution is dependent upon or concomitant with rRNA maturation. On the contrary, Fig. 1B shows that the rRNA molecules contained in reconstituted particles retained the same size they had in vivo.
Nuclear Ribosomal Subunits Containing Immature rRNA Can Function in Protein SynthesisThe results reported above suggested the possibility that in vivo assembled ribosomal subunits still present in nuclei and containing immature RNA can already function in protein synthesis. To examine this possibility, we tested 3H-labeled ribosomal particles isolated from nuclei in the in vitro protein synthesis system described under "Experimental Procedures." They were as active as unlabeled ribosomal subunits isolated from polyribosomes (Table I).
The data reported here show that, in the presence of a nuclear component, eukaryotic 40 and 60 S ribosomal subunits can be reconstituted in vitro from free D. discoideum rRNA and ribosomal proteins. This should provide a new opportunity to study the structure and function of eukaryotic ribosomes.
The in vitro reconstituted ribosomal subunits have a full complement of ribosomal proteins and are fully functional in protein synthesis. Data to be published elsewhere2 show that they are active also in a newly discovered function, the control of the stability of a class of developmentally regulated mRNAs.
The in vitro system described here for ribosome reconstitution indicates some principles that should be applicable to other eukaryotic organisms and help to solve the problem of the in vitro reconstitution of their ribosomal particles. One is that mature rRNA, extracted from cytoplasmic ribosomes, is an inadequate substrate for the reconstitution reaction. A much better substrate is immature precursor rRNA extracted from nuclei. It is likely that the extra sequences present in precursor RNA play some role in the assembly process. This is in agreement with the finding that nuclear particles sedimenting at 40 and 60 S and therefore presumably fully assembled still contain immature RNA. The same is true for ribosomal particles reconstituted in vitro and containing a full complement of ribosomal proteins. Both nuclear and in vitro reconstituted particles containing immature rRNA are active in protein synthesis. This suggests that in Dictyostelium the last step in ribosome assembly is RNA maturation, as in E. coli (17, 18, 39) and probably in yeast (40). Data to be published elsewhere2 show that in Dictyostelium, as in E. coli (14, 18, 39), ribosomal subunits still containing immature RNA enter polyribosomes directly as they join the cytoplasm. Maturation of rRNA would therefore occur in polyribosomes.
In vitro reconstitution of ribosomal particles does not occur autonomously but instead requires morphopoietic factors that are contained in the nucleus. Among these factors there is/are one or several small nuclear RNAs, since the nuclear lysate that drives the reconstitution reaction in our experimental conditions can be substituted by a small nuclear RNA fraction. Since rRNA incorporated in reconstituted particles is still immature, the small nuclear RNA that functions in our system is probably not involved in rRNA trimming but rather in the assembly process. In any case, the principle that the assembly of many cell organelles is facilitated and/or controlled by morphopoietic factors appears to be valid also for eukaryotic ribosomes, as already suggested by in vivo studies (33, 34).
Some of the data reported here were communicated at an International Meeting on D. discoideum held in Amsterdam in 1993.
We thank Dr. M. Magro for help in some experiments and Pietro Nasillo for the photographic reproductions.