ABSTRACT
We report the identification of a new gene,
RRP3
(
Ribosomes in
Saccharomyces cerevisiae
are composed of a small 40S subunit (containing 18S rRNA) and a large 60S
subunit (containing 5S, 5.8S and 25S rRNAs). The 5.8S, 18S and 25S RNAs of
yeast are transcribed as a single 35S RNA polymerase I transcript (
1
,
2
). This 35S pre-rRNA is interrupted by internal transcribed spacers (ITS1 and ITS2) and is
bounded by external transcribed spacers (5'- and 3'-ETS) that must be precisely and efficiently processed for mature RNA formation.
The endonucleotytic cleavages leading to mature yeast rRNA are outlined in
Figure
1
A. Aberrations in any of these steps usually results in defective maturation of
the transcript. The 3'-ETS is most quickly processed and is not detectable
in vivo
. The primary transcript is separated into two intermediate precursors by
cleavage at A
0
, followed immediately by processing at A
1
and A
2
to release the 20S and 27SA intermediates (
2
). Cleavages at A
1
and A
2
are stepwise and appear to be linked. Failure to process at A
1
results in a loss of the A
2
cleavage. The 20S intermediate is further processed to 18S rRNA in the
cytoplasm (
3
), while the 27SA intermediate can be processed by two alternative pathways
which regulates the ratio of the short and long forms of the 5.8S RNA subunit.
Cleavage can occur at A
3
followed by exonucleolytic cleavage to the B
1
(S) site to give a 27SB(S) product (
4
). Processing of 27SB(S) gives rise to the 5.8(S) and the mature 25S RNA.
Alternatively, processing directly at B
1
(L) generates the 27SB(L) and hence 5.8(L) and 25S products.
Though the pathway of rRNA maturation is well characterized in both yeast and
mammals, the biochemical mechanisms involved in rRNA maturation are still under
active study. rRNA processing, like that of pre-mRNA processing, requires a number of RNPs, the snoRNPs (for reviews see
5
,
6
). Many snoRNAs have been characterized in both the yeast and mammalian systems.
They are associated with nucleolar proteins and, quite strikingly, many of them
contain significant regions of homology to mature and flanking sequences in the
pre-rRNA. Crosslinking studies have shown that these interactions occur and
therefore must be meaningful. In several cases the snoRNAs have been directly
implicated in processing events through
in vitro
studies. U3 snoRNA has been shown to be required for a cleavage event in the 5'-ETS in both mouse and frog
in vitro
systems (reviewed in
5
,
6
). Recently it has been conclusively shown that the purified yeast MRP ribozyme
can carry out precise cleavage at A
3
leading to maturation of 5.8S RNA (
7
). It seems likely that this is a RNA-mediated catalytic event (
8
), because MRP is related in structure to the catalytic RNA in RNase P and in
yeast and mammals MRP and RNase P interact with the same protein (the Th/To
antigen in mammals and the Pop1 protein in yeast), leading to the suggestion (
8
) that they share a common evolutionary ancestor. Depletion studies in yeast
have implicated U3, U14 and snR30 snoRNPs in the early cleavages of pre-rRNA leading to maturation of 18S rRNA(
3
,
9
,
10
).
Taq DNA polymerase was obtained from Perkin Elmer Cetus (Norwalk, CT).
Restriction enzymes were obtained from New England Biolabs (Beverly, MA).
Oligonucleotides were synthesized on an Applied Biosystems (Foster City, CA)
DNA synthesizer and purified by urea-polyacrylamide gel electrophoresis. GeneScreen Hybridization Transfer
Membrane was from NEN Research Products (Boston, MA).
Hin
dIII linkers were obtained from New England Biolabs. Sequenase was purchased
from IBI. Exo-mung deletion kit was obtained from Stratagene. Oligonucleotides used for
hybridization were obtained from J.Warner. Quantitation was carried out on an
LKB ultrascan XL densitometer.
Yeast genomic DNA was prepared as described by Sherman (
24
). Amplification of DNA was done according to Chang
et al
., except for the difference in primers (
14
). We used the primers IIA used by Chang and the newly constructed oligonucleotide 5'-ACGGGATCC(T/G)NCCNGT(A/G)CG(A/G)TGNAT(A/G)TA-3'. DNA was denatured for 1 min at 92oC, the primer was annealed for 45 s at 47oC and primer extension was carried out for
75 s at 72oC. PCR reactions were performed on a Perkin Elmer Thermocycler. After 30
rounds of amplification, the reaction was allowed to extend for 10 min at 72oC. The cloning and sequencing of PCR products was also done as described
except that the PCR product was first blunt end ligated with a 100-fold excess of phosphorylated
Hin
dIII linkers (
14
). The mixture was then digested with
Bam
HI and
Spe
I at 37oC for 2 h. The digests were phenol/CHCl
3
extracted, ethanol precipitated and ligated into the Bluescript KS vector. The
resulting transformants were sequenced using T3 and T7 primers. Sequence
analysis of a number of clones revealed three unique genes. The PCR clone was
then used as a probe to further clone the genomic copy of the gene. One of
these genes, termed
RRP3
was found in a YCP50B library (
25
). Northern and Southern analysis was performed according to Chang
et al
. (
14
).
The genes were further mapped and 4 kb fragments were inserted into the
Bluescript KS vector (Stratagene). The fragments were then sequenced by
performing exo-mung deletions according to the Stratagene kit and sequenced with
Sequenase according to the manufacturer's instructions (IBN).
Yeast cells were grown in YPD or synthetic media supplemented with 2% dextrose
or galactose as a carbon source. Yeast were transformed by the lithium acetate
method (
26
). A 4 kb
Bgl
II fragment from the YCP50 genomic clones was cloned into pUC18 and termed
pCO93.
The haploid strains SS330 (MAT
a
his 200 tyr1 ade2-101 ura3-52 GAL
+
suc2
) and SS328 (MAT
a
his 200 lys2-801 ade2-101 ura3-52 GAL
+
suc2
) were crossed to derive the diploid strain (SS329X) for disruption experiments.
Construction of the disrupted gene was by ligating a blunt end
HIS3
gene fragment into the
Xho
I site of pCA093 to create the disrupted plasmid pCO09HS. The
Eco
RI-
Xba
I fragment of pCO09HS was then used for integration into the genome as described
by Struhl (
27
). The disruption was verified by Southern analysis. This strain was designated
CM101.
The disrupted diploids were sporulated by growing the cells in YPA to 1 OD
600
(2% Bacto-peptone, 1% yeast extract, 1% potassium acetate). They were then
inoculated into 1% potassium acetate supplemented with 0.004% adenine, uracil
and histidine. They were allowed to grow at 30oC for 2-3 days. Tetrad analysis was performed by incubating the sporulated
diploids for 15 min with glusalase and dissecting the ascus into the four
spores on YPD plates.
A plasmid containing the
RRP3
gene controlled by a galactose-inducible promoter was constructed by placing the
Eco
RV-
Ssp
I fragment of pCO093 into the
Bam
HI site of vector pSEY68 by blunt end ligation. The
Eco
RV restriction site is 97 bp upstream of the ATG start codon of
RRP3
and the
Ssp
1 site is 58 bp downstream of the stop codon. This plasmid was transformed into
the diploid disruption strain, sporulated and dissected on YP galactose for
identification of a haploid cell containing both the genomic disruption and the
galactose-inducible plasmid. This haploid strain is designated CM201.
Depletion analysis was performed as described by Hughes and Ares (
9
) except that the cells were grown in minimal medium (2% galactose) to an OD
600
of 1.0. The cells were then washed and inoculated into minimal medium (2%
glucose). The cells were kept in logarithmic phase by dilution. CM201 and SS330
were grown for 5 h in glucose and JH44 for 12 h before pulse-chase analysis. Pulse-chase analysis was performed following the method of Sachs and
Davis (
20
). An aliquot of 3 ml logarithmically growing cells was pulsed with 60 [mu]Ci/ml [
3
H-methyl]methionine (75 Ci/mmol; Amersham) at room temperature for 2.5 min.
The cells were then chased with 50 [mu]g/ml cold methionine. Samples were taken at 0, 3 and 12 min after the chase,
centrifuged, washed and recentrifuged. The samples were then frozen at -70oC.
RNA was extracted from cells essentially as described by Elion and Warner (
28
). Cells were harvested by centrifugation at 4oC, washed in TE buffer and resuspended in 20 [mu]l SB3 buffer (50 mM Tris-HCl, pH 8.0, 1 M sorbital, 10 mM MgCl
2
, 3 mM DTT). Zymolase was added to 1 [mu]g/ml and incubated for 30 min at 30oC. The cells were then diluted in 350 [mu]l 50 mM NaOAc, pH 5.3, 10 mM EDTA, 0.5% SDS and immediately mixed
with an equal volume of hot phenol (equilibrated in the above buffer). The
samples were vortexed at 65oC for 15 min and centrifuged to separate the phases. The aqueous phase was
further extracted with CHCl
3
and ethanol precipitated.
Yeast RNA (5 [mu]g) was denatured with glyoxal and separated on a 1.2% agarose gel containing
10 mM sodium phosphate, pH 7.0 (
29
). RNA was transferred onto GeneScreen membrane by pressure blotting and
crosslinked.
Hybridization with oligonucleotides complementary to pre-rRNA was carried out at the calculated
T
m
of the oligonucleotide in 6* SSPE, 1% SDS, 10* Denhardt's, 50 [mu]g/ml tRNA and 50 [mu]g/ml denatured salmon sperm DNA. The filters were washed
at room temperature twice for 10 min in 6* SSPE, 0.1% SDS and once at the calculated
T
m
.
An
Nde
I site was introduced, by site-directed mutagenesis, into the
RRP3
gene at the initiating methionine codon. The corresponding
Nde
I-
Xba
I fragment was then ligated into vector pet19 (Novagen). The vector contains a
series of 10 histidine residues followed by the thrombin protease cleavage
site. The plasmid was placed in
E.coli
strain Bl21(DE3) and grown to an OD
600
of 0.4. At this point the cells were induced with 0.5 mM IPTG and grown for 4 h
more. Cells were harvested and lysed in 50 mM Tris, pH 8.0, 500 mM NaCl, 5 mM
imidizole, 10% glycerol, 5 mM [beta]-mercapatoethanol. The extract was loaded onto a 1.5 ml NTA column,
washed with a similar buffer containing 50 mM imidazole and eluted with 200 mM
imidazole. The eluate was dialyzed against 100 mM NaCl, 50 mM Tris, pH 7.5, 1
mM EDTA, 10% glycerol, 1 mM DTT. The protein was loaded onto a Sepharose S FPLC
column and eluted with a gradient of 0.1-1 M NaCl.
Proteins involved in the rRNA processing pathway have been difficult to identify
by conventional mutation screens (
20
,
30
,
31
). Recently, one processing protein, Drs1, was identified as a cold-sensitive mutation (
21
). Study has been further complicated by the fact that very few mutants have
been found to accumulate intermediates in the processing pathway (possibly due
to the instability of incorrectly processed intermediates). Only rrp1 was found
to accumulate an intermediate to 25S rRNA (
32
,
33
). For the mutant clp8 the efficiency of processing of 18S rRNAfrom the 20S
intermediate is reduced, but this is believed to be caused by a defect in
export, since final processing takes place in the cytoplasm (
34
). Therefore few
trans
-acting factors have been identified and most by means other than their
biochemical function. Several, processing proteins were identified by
localization in the nucleolus (
35
,
36
).
We have cloned a new gene by PCR analysis in a search for additional putative
RNA helicases in
S.cerevisiae
. Degenerate oligonucleotides
complementary to the conserved regions IV and VIII of DEAD box genes were
synthesized and used as primers to identify new genes. A PCR product of the
appropriate size was isolated, cloned and 45 clones were sequenced, revealing
three new genes. A YCP50 library was then screened to detect genomic clones.
The genes were then mapped and sequenced. These new genes have all the highly
conserved motifs characteristic of the `DEAD' family of RNA helicases (Fig.
2
). One of these genes,
RRP3
, codes for a 60.9 kDa protein and contains a large (130 amino acid) N-terminus prior to the helicase domain. Furthermore, Rrp3 has several GR
repeats that are common to nucleolar proteins in the C-terminus between amino acids 524 and 532.
A rrp3 null allele was created by inserting the
HIS3
gene into the helicase domain. The clone containing the disrupted gene was
excised and integrated into the chromosome to disrupt one allele. Southern
analysis verified the disruptions. Upon sporulation the tetrads segregated 2:0,
indicating that
RRP3
is an essential gene. Northern and Southern analysis also verified that
RRP3
is a single copy gene (data not shown). A haploid disruption containing the
rrp3 disruption was created by sporulating from a heterozygous diploid
containing the plasmid pCO09HS. This plasmid contains the wild-type
RRP3
gene controlled by the Gal1 promoter (pCO09HS). This strain was termed CM201
and grows on galactose but is inviable on glucose.
In order to deplete the cell of Rrp3, the strain CM201 was grown in minimal
medium containing galactose. The cells were transferred to glucose-containing medium and grown until the doubling time deviated from the wild-type. After the shift to galactose the rate of growth of CM201
quickly decreases as the medium is changed and deviates from control after 5-6 h (Fig.
3
). The control strain, SS330 and a wild-type containing the
RRP3
plasmid was used as a control and has a doubling time of ~90 min in glucose. The strain JH44 (a U3 depletion strain) was also
subjected to depletion and pulse-chase analysis as another control (data not shown). After 5 h growth in
glucose, both strains were pulsed with [
3
H-methyl]methionine, chased and samples were taken at the indicated time
points (see Fig.
4
). The ability to observe the 35S precursor varied from experiment to experiment
because of its instability. For the Rrp3-depleted cells, mature 18S rRNA and its 20S intermediate did not form and
a faint 23S band was sometimes observed, as seen at the 3 min time point for
CM201. The control strains showed normal processing of the 35S substrate. The
U3-depleted strain showed a similar defect to that seen in depleted CM201,
but the time course was slower (data not shown).
Many of the intermediates in rRNA processing shown in Figure
1
A can be detected by Northern analysis. Some of these intermediates cannot be
detected in the Rrp3-depleted strain. Processing, as determined by Northern analysis, of the
RNA in Rrp3-depleted cells occurs as indicated in Figure
1
B. A similar scheme was also determined for other genes involved in 18S rRNA
processing (
9
,
10
,
35
). We were able to tell from the detectable intermediates that cleavages at A
0
, A
1
and A
2
were lacking (Fig.
1
A). Interestingly, this is the same phenotype seen when the three essential
snoRNAs are depleted. Either cleavage at B
1
and/or A
3
separates the 18S precursor from the 27S RNA. The 27S intermediate is further
processed to mature 5.8 and 25S rRNA as discussed below.
The oligonucleotides used for hybridization were obtained from J.Warner and are
listed below (
30
; Fig.
6
B). The oligonucleotide JW127 anneals to the 5'-ETS and weakly labels the 35S precursor band, only visible under
prolonged exposure in both wild-type and depleted cells (Fig.
6
A). As also observed by Warner, this oligonucleotide hybridizes poorly. In
depleted CM201 cells, an additional band at 23S and one slightly smaller band
appears, probably a 22S intermediate from the 5'-end of the primary transcript. It is unstable and does not
accumulate. The length and presence of this intermediate indicates that
cleavage at A
0
did not occur.
As described in Materials and Methods, we have adapted Rrp3 for metalloaffinity
chromatography and have purified the protein to near homogeneity. The ATPase
activity of the recombinant enzyme was measured in the presence or absence of a
number of RNA cofactors. There was a barely detectable level of ATPase activity
in the presence of poly(A), but yeast rRNA did not affect the activity. Though
more study of this enzyme is required, the low ATPase activity may indicate
that the enzyme requires a quite specific RNA ligand. The activity of DbpA, for
example, is nil with all RNA effectors other than its specific substrate (
40
).
As previously stated, the pattern of cleavage events seen in Rrp3-depleted cells is identical to that seen when other processing factors are
depleted (
3
,
9
,
10
,
36
-
39
). In each case the A
1
and A
2
cleavage events are blocked. This suggests that Rrp3 could mediate structural
transitions between one of the essential snoRNAs, U3, U14 or snR30, and the pre-rRNA substrate. Since U3 and U14 contain sequences complementary to the
substrate, Rrp3 could be required as a helicase to mediate ATP-dependent changes in these interactions.
We would like to thank Jon Warner for oligonucleotides for hybridization and
John Woolford for advice. We would also like to thank John Hughes and Manuel
Ares Jr for helpful discussion and for control strains and David Tollervey for
a helpful reading of the manuscript. This work has been supported by a grant
from the National Institutes of Health.
REFERENCES
Return