ABSTRACT
RNase P is a ribonucleoprotein endoribonuclease responsible for the 5
'
maturation of precursor tRNAs in all organisms. While analyzing mutations in
conserved positions of the yeast nuclear RNase P RNA subunit, significant
accumulation of an aberrant RNA of
~
193 nucleotides was observed. This abundant RNA was identified as a 3
'
extended form of the 5.8S rRNA. This strain also displays a slightly elevated
level of other rRNA processing intermediates with 5
'
-ends at processing site A2 in the internal transcribed spacer 1 (ITS1)
region of the rRNA primary transcript. To test whether pre-rRNA in the region of ITS1/5.8S/ITS2 is a substrate for RNase P
in vitro
, nuclear RNase P was partially purified to remove contaminating nucleases.
Cleavage assays were performed using an rRNA substrate transcribed
in vitro
which includes the 5.8S region and its surrounding processing sites in ITS1 and
ITS2. Discrete cleavages of this rRNA substrate were coincident with the peak
fractions of nuclear RNase P, but not with fractions corresponding to
mitochondrial RNase P or ribonuclease MRP RNA. The cleavage activity is
sensitive to treatment with micrococcal nuclease, also consistent with an
activity attributable to RNase P. The strong RNase P cleavage sites were mapped
and their possible relationships to steps in the rRNA processing pathway are
considered. These observations suggest an intimate relationship between the
processes of tRNA and rRNA maturation in the eukaryotic nucleus.
RNase P is one of a growing list of ribonucleoproteins (RNPs) found to be
involved in a wide variety of biosynthetic pathways in eukaryotes. Essential
RNPs and their functions include the small nuclear RNPs (snRNPs) in mRNA
splicing (
1
,
2
), ribosomes in translation (
3
-
5
), telomerase in chromosome end formation (
6
,
7
), small nucleolar RNPs (snoRNPs) in rRNA processing (
8
,
9
), and RNase P in precursor tRNA (pre-tRNA) 5'-end maturation (
10
). Of these, only RNase P is considered a ribozyme, following demonstration that
the
E.coli
RNA subunit is catalytic
in vitro
in the absence of protein (
11
). RNase P from Bacteria, Archaea and Eucarya contains a similar sized RNA (~300-400 nt), but, unlike the bacterial enzymes, the RNA subunits from
Archaea and Eucarya are not active without protein (
12
,
13
). The RNA subunits from different phyla show only small patches of sequence
conservation, but retain similar structural features (
14
,
15
). Only two protein subunits have been purified to homogeneity from the
eukaryotes, the
S.cerevisiae
mitochondrial (105 kDa ) (
16
) and the
S.pombe
nuclear (100 kDa) (
17
) proteins. The eukaryotic proteins are of a significantly greater molecular
weight compared to the 14 kDa
E.coli
C5 protein (
18
).
Although the 369 nt RNase P RNA from
S.cerevisiae
nuclei has been extensively characterized, the protein content of the holoenzyme
remains to be determined (
15
,
19
-
21
).
E.coli
RNase P is able to recognize a number of different RNA substrates in addition to
pre-tRNAs. Examination of temperature sensitive mutants revealed accumulation
of precursor RNAs including 4.5S RNA (
26
,
27
), 10Sa (
28
,
29
) and the polycistronic mRNA from the histidine operon (
30
). In all instances the substrate resembles a pre-tRNA and is consistent with the proposed minimal structure requirement for
an RNase P substrate (
31
). A recent study (
32
) suggests that bacterial RNase P holoenzyme has a greater substrate range than
the RNA subunit alone, implicating the subunit in recognizing substrates other
than pre-tRNAs.
The eukaryotic RNase P that processes cytoplasmic pre-tRNAs is thought to function in the nucleus and 5' and 3' processing precede splicing (
22
,
22
). Since yeast mutants that fail to splice tRNAs accumulate end-matured tRNAs in the nucleus (
24
,
25
), RNase P cleavage (an earlier event) is probably occurring in the nucleus.
Eukaryotic nuclear RNase P is closely related to ribonuclease MRP
ribonucleoprotein, an endonuclease known to be involved in rRNA processing (
33
-
37
). In humans, RNase P appears to share Th antigen association with RNase MRP (
38
-
40
). A 100.5 kDa polypeptide has been shown to associate with both enzymes in
yeast and may be analogous to the Th antigen. A temperature sensitive mutation
in this yeast protein, Pop1, results in defects in rRNA processing and tRNA
processing, as might be anticipated for a protein playing a functional role in
both enzymes (
41
). In addition to the common protein association, the RNA components of RNase P
and RNase MRP possess some sequence and structural similarity (
14
,
42
,
43
), leading to suggestion that the two eukaryotic enzymes are derived from a
common progenitor. These observations, in addition to
in situ
localization of at least some RNase P RNA to the perinucleolar compartment with
RNase MRP (
44
,
45
), has led to further speculation that there might be undiscovered similarities
in the substrates recognized by the two enzymes (
43
,
46
,
47
).
In the process of characterizing mutants with defects in the
S.cerevisiae
nuclear RNase P RNA subunit, the appearance of an anomalous RNA prompted us to
investigate whether RNase P was involved in the maturation of rRNA, as well as
tRNA.
For construction of the temperature sensitive mutant used in these studies,
S.cerevisiae
strain JLY-1 (
19
) containing a
rpr1
::HIS3 disruption in strain W3031A (
MAT
a
ade2-1 his3-11,15 leu2-3,112 trp1-1 ura3-1 can1-100
) and plasmid YCp50-
RPR1
with wild-type copy of the
RPR1
gene were used. JLY-1 was transformed with plasmid pRS315 (
LEU
marker) containing either the
RPR1
mature coding sequence or the mutated
RPR1
coding sequence inserted between wild-type
RPR1
5' and 3' flanking gene sequences; the wild-type
RPR1
containing plasmid (
URA3
marker) was eliminated by growth on 5-fluorooritic acid (
48
). Protease-deficient strain PP1002 (
MAT
[alpha]
ade2 leu2-3,112 pep4-3 rna3 rna82
) (gift of P. Piper & Kerstin Stråby) was used as the source of RNase P holoenzyme (
20
). Media were prepared as described (
19
). Growth of the temperature sensitive (
ts
) mutant used in these studies and its isogenic wild-type strain was maintained in selective medium lacking histidine and
leucine at 25oC for isolation of RNA.
RPR1
mutation T
315
[Delta]T
307
was derived by sequence randomization at highly conserved positions in the
RPR1
RNA (Fig.
1
). Detailed characterization is described elsewhere (
49
). Nucleotide T
315
represents an A to T transversion and a single nucleotide deletion is found at
position 306 or 307. DNA sequencing revealed no other mutations in the plasmid-born
RPR1
gene (
50
).
RNA was labeled by addition of 500 [mu]Ci [
32
P]orthophosphate to 1 ml of cells at OD
600
of 1.0 for 5 min. RNA was prepared as previously described (
51
). Samples were subjected to electrophoresis on an 8% denaturing acrylamide gel
and exposed on a Molecular Dynamics PhosphorImager 445 SI.
Oligonucleotides used for PCR, primer extension and northern hybridization are
designated P1-P5. P1 and P5 were used as PCR primers in construction of the rDNA
template (see RNase P reactions). P1 included a T7 promoter. Oligonucleotides P2, P3 and P4 were end-labeled with T4 polynucleotide kinase (Gibco-BRL) and [[gamma]-
32
P]ATP (New England Nuclear) and purified by 15% denaturing polyacrylamide gel
electrophoresis. Oligonucleotide sequences are as follows:
P1, 5'-TAATACGACTCACTATAGGCCAAACGGTGAGAGA-
TTTCTGTGC-3';
P2, 5'-TCCAGTTACGAAAATTCTTGTTTT-3';
P3, 5'-CGCATTTCGCTGCGTTCTTCATCG-3';
P4, 5'-ACAGAATGTTTGAGAAGGAAATG-3'; and
P5, 5'-ACTTTAAGAACATTGTTCGCCT-3'.
A primer which anneals to positions +24 through +40 of the yeast alcohol
dehydrogenase 1 (
ADH1
) RNA (
52
) was added to control for primer extension reactions. An oligonucleotide which
anneals to positions +62 through +88 of the U6 snRNA was used for total RNA
blot quantitation (
53
,
54
).
The
RPR1
probe was the complement of the coding region cloned into plasmid pGEM-3Z (
20
,
55
) and transcribed using [[alpha]-
32
P]UTP. The ribonuclease MRP RNA (RNase MRP) coding region antisense probe was
synthesized from the RNase MRP gene (
NME1
or
PPR2
) in pBluescript II KS
+
(gift of D. Tollervey).
Total RNA was prepared from yeast carrying wild-type or T
315
[Delta]T
307
RPR1
genes. Two [mu]g of each RNA was subjected to electrophoresis on 8% denaturing
polyacrylamide gels and transferred to Maximum Strength Nytran (Schleicher & Schuell) using a Genie Electrophoretic Blotter (Idea Scientific). For analysis
of both large and small RNAs, ~10 [mu]g of WT and mutant RNAs was added to denaturing RNA loading buffer (95%
formamide, 0.1% SDS, 5 mM EDTA, 0.1% bromophenol blue and xylene cyanol),
incubated for 10 min at 65oC and separated on a 1.2% native agarose gel. A denaturing agarose gel was
also run with these samples and confirmed the results obtained from northern
analysis of the non-denaturing gel for the large rRNAs. For these analyses, RNA was capillary
blotted to Nytran (
56
). Hybridizations with P3 and P4 were done at 55oC and 50oC, respectively, for blots of denaturing polyacrylamide gels. For
quantitation of total RNA loaded, the radioactive U6 probe was used in a
separate hybridization at 40oC. The northern blot of the 1.2% agarose gel was probed with P4 and the
ADH1 probe at 40oC. Following hybridization, membranes were exposed for PhosphorImager
analysis.
For detecting specific RNAs in RNase P purification chromatography, 10 [mu]l from Mono Q fractionation was treated with stop mix (0.25 mg/ml proteinase
K, 0.5% SDS, 25 mM EDTA) extracted, and precipitated. Samples were separated on
6% denaturing acrylamide gels, electroblotted, and hybridized at 55oC to
32
P-labeled antisense
RPR1
or
NME1
RNA.
Primer annealing and extension used SuperScript II (Gibco) at 45oC as recommended by the manufacturer. Products were separated on 8%
denaturing polyacrylamide gels. Gels were dried and exposed to film or phosphor
screen. Positions of primer extension stops were determined by comparison with
adjacent chain termination sequencing reactions performed using identical
primers on the rDNA template used for synthesis of the rRNA substrate. Primer
extension of the aberrant rRNA (Fig.
4
C) was done following elution of this RNA from 8% denaturing poly- acrylamide gel.
Extracts were prepared from
Saccharomyces cerevisiae
strain PP1002 as described (
57
). RNase P activity was purified from the extracts using SP Sepharose
(Pharmacia), DEAE cellulose (Whatman DE52), and FPLC Mono Q H/R 5/5
(Pharmacia). The yeast extract was ammonium sulfate precipitated and
resuspended in buffer HGMDN (20 mM HEPES pH 7.9, 10% glycerol, 10 mM MgCl
2
, 1 mM DTT, 0.5% Nonidet P-40), adjusted to 0.15 M KCl, and applied in a final volume of 300 ml (with
addition of HGMDN + 0.15 M KCl) to 120 ml SP Sepharose. Extensive washing was followed
with single-step elution with HGMDN + 0.4 M KC1. Approximate protein peak
concentrations were assessed by Coomassie brilliant blue staining. The peak protein fractions were pooled, adjusted to 0.2 M KCl with HGMDN, and applied to
a 40 ml column of DEAE cellulose. RNase P was eluted in a single step with
HGMDN buffer containing 0.4 M KCl.
The pooled RNase P activity was applied to a Mono Q H/R 5/5 (1 ml) column in
HGMDN + 0.275 M KCl. A 20 ml gradient of 0.0275-0.625 M KCl was used for
elution, and 0.5 ml fractions were collected and tested for RNase P activity,
protein content, and the presence of specific RNAs by northern analysis. In
addition to the activity peaks observed with fractions 28 and 32/33 (0.47 and
0.49 M KCl) (Fig.
5
), another peak of RNase P activity was found in fractions 37 and 38 (0.52-0.53 M KCl). This coincided with a peak of mitochondrial RNase P, as
analyzed by northern blot hybridization to an oligonucleotide probe
complementary to mitochondrial RNase P RNA, 5'-GGACTCCTGCGGGGTCCGCCCGCG-3' (
16
).
RNase P activity from chromatographic separation was assayed using a
mitochondrial pre-tRNA
Asp
substrate (
58
), [[alpha]-
32
P]UMP-labeled as described for the
RPR1
as probe. The substrate was purified and assayed described (
59
).
RNase P activity was assayed as previously described for RNase MRP (
34
) using a yeast pre-rRNA substrate and 0.5-1 [mu]l of the Mono Q fractions. This pre-rRNA substrate is defined by the PCR primers P1 and P5
(see Figs
3
B and
8
) which were used to synthesize this region by PCR amplification from the yeast
rDNA repeat from plasmid pRibH7 (contains the 2.7 kb
Hin
dIII fragment in pBR322, gift of H. Federoff & J. Warner). This template was transcribed from the T7 promoter included in
primer P1 and purified. The
in vitro
transcribed pre-rRNA was 3'-end labeled by addition of [5'-
32
P]cytidine 3',5'-bis(phosphate) with RNA ligase (
60
). Electrophoretic separations of all assay products were visualized using a
PhosphorImager.
Sensitivity of RNase P to micrococcal nuclease digestion was tested by combining
0.5 [mu]l of the peak fraction from FPLC Mono Q separation in a 5 [mu]l reaction with 4 mM CaCl
2
and 2 U micrococcal nuclease with incubation at 37oC for 15 min. In some cases 1 [mu]l 200 mM EGTA (ethylene glycol-bis([beta]-aminotheyl ether)-
N,N,N
'
,N
'-tetraacetic acid) was first added specifically to inhibit
micrococcal nuclease before RNase P addition. Micrococcal nuclease digestion
was stopped by adding 1 [mu]l 200 mM EGTA to those reactions to which it had not been previously added.
RNase P reactions were then performed using the pre-rRNA substrate.
Our laboratory has examined the effects of simultaneous mutations at highly
conserved positions in
S.cerevisiae
nuclear RNase P RNA. Filled circles on the secondary structure model of the RNA
in Figure
1
highlight one region of conserved nucleotides that was targeted for
randomization of the sole copy of the
RPR1
gene, encoding the RNA component of yeast nuclear RNase P. One of these
variants, T
315
[Delta]T
307
(Fig.
1
), was found to accumulate a previously unidentified RNA. Growth of the mutant
is reduced at 25oC as well as 37oC, allowing analysis of non-lethal RNA processing defects at 25oC. To confirm the presence of a nuclear RNase P functional
defect, a northern blot prepared with total RNA from both the mutant and wild-type strains was hybridized with a probe to tRNA
Leu3
(Fig.
2
A). Examination of this single tRNA population revealed defects in the tRNA
Leu3
steady-state population consistent with results from a previously characterized
RNase P mutant (
19
,
61
,
62
). RNA isolated from both the wild-type and mutant strains showed products corresponding in size to the pre-tRNA
Leu3
containing only its intron and the fully-processed, mature tRNA
Leu3
. In the RNA from the mutant, the probe identified additional RNAs consistent
with an accumulation of the primary transcript (intron with 5'- and 3'-ends) and a spliced tRNA
Leu3
carrying its 5' leader and 3' trailing sequences. This lack of 3'-end processing in an RNase P mutant has been observed
previously (
19
) and may be due to extensive pairing between the 5' leader and 3' trailing sequences. The appearance of pre-tRNA
Leu3
which is spliced, but not end-matured, suggests that maturation of termini before splicing is not
obligatory (
19
).
Although the path to production of the major aberrant RNA is unclear, we tested
whether the RNase P defect might cause more subtle problems in maturation of
the 5.8S 5'-end after the A2 cleavage (Fig.
3
B). Primer extension was performed using primer P2 (Fig.
3
A) for quantitation of the relative levels of the precursors with 5'-ends at A2 and A3. ADH1 mRNA was used for normalization of signals.
Levels of transient pre-rRNAs with 5'-ends at A3 appear unchanged while the pre-rRNAs with their 5'-ends at A2 are increased 5-fold in the mutant (Fig.
5
). The product seen below site A2 is attributable to a primer extension pause
site as seen on longer exposure of the wild-type reaction. Increase in the levels of pre-rRNA with 5'-ends at A2 suggests that the mutant may be partially
deficient in its ability to convert the precursors cleaved at A2 to those with
mature 5.8S termini. It is unclear why this difference in rRNAs with 5'-ends at A2 is not manifest when the pre-rRNA is quantitated in a northern analysis (see above; Fig.
4
D). It is possible that only the level of the relatively minor 27Sa intermediate
(5'-end at A2) is increasing and the level of the more abundant 27Sb (5'-ends at B1
S
and B1
L
) is not changing. Alternatively, the RNAs that give rise to the increase in 5'-ends at site A2 might have quite heterogeneous 3'-ends and be spread over multiple bands in Figure
4
D that are not visible individually.
To test whether nuclear RNase P was capable of cleaving rRNA
in vitro
, we partially purified the RNase P holoenzyme from wild-type and mutant yeast. The results from RNase P assays using a pre-tRNA
Asp
substrate and the most highly purified FPLC Mono Q fractions of wild-type enzyme are shown in Figure
6
A. The RNase P activity produced two peaks centered in fractions 28 and 32/33.
To determine the location of the RNA subunit of RNase P, samples were
deproteinized and separated on a denaturing polyacrylamide gel (Fig.
6
B). Two species of RNA were detected with an
RPR1
probe: the 369 nt fully processed
RPR1
RNA and the previously characterized ~453-483 nt precursor (
19
). The strongest hybridization signal, corresponding to the size of the 369 nt
mature
RPR1
RNA, was seen in fraction 28. Fraction 32 RNA showed another peak of
hybridization with both the mature
RPR1
RNA and its precursor. The first peak of RNase P activity corresponds to the
first peak of mature
RPR1
RNA and the second activity peak corresponds to the
RPR1
precursor-containing fractions. A third peak of pre-tRNA cleavage activity was seen in fractions 37 and 38, which was
found to contain intact mitochondrial RNase P RNA by hybridization with a
mitochondrial RNase P RNA probe (not shown).
To determine if RNase MRP was able to cleave a pre-rRNA substrate, a 603 nt
pre-rRNA substrate was transcribed from a PCR template (primers P1 and P5
shown in Figs
3
A and 8). The substrate spanned much of ITS1, 5.8S, and all of ITS2. In assays
using this 3'-end labeled pre-rRNA substrate we were unable to find any evidence of
endonucleolytic cleavage with fractions corresponding to the RNase MRP RNA
(Fig.
6
B). Instead, a set of discrete cleavage products were detected in those
fractions coinciding with nuclear RNase P (Figs
6
and
7
A). However, this does not discount the possibility of an undescribed activity
as responsible for these rRNA cleavages. No similar activity corresponding to
the mitochondrial RNase P was detected using the pre-rRNA substrate. In a further effort to identify RNase MRP activity in
these fractions, a previously characterized
in vitro
substrate for yeast RNase MRP was used in a cleavage assay (not shown) (
64
). Little activity was detected in any MRP fractions, suggesting activity did
not survive well under these purification methods, although full-length RNase MRP RNA could be detected as separate from the RNase P RNA
peaks (Fig.
6
B).
Sites of cleavage of the pre-rRNA substrate were mapped more precisely by primer extension of the
unlabeled substrate treated with fraction 28 RNase P (data not shown). The
cleavage sites are indicated relative to the rDNA sequence in Figure
8
and relative to the region in Figure
3
A. One of the prominent cleavages maps 2 nt downstream of the A3 site (Fig.
3
) which is near the postulated site of action of RNase MRP (
65
). Another major cleavage site maps between sites A3 and B1 of ITS1. Sites that
are less favored (Fig.
7
A), are located at B1
L
and B1
S
(the mature 5'-ends of 5.8S
L
and 5.8S
S
), within the 5.8S sequences, and within ITS2. None of these
in vitro
cleavage sites map immediately downstream of 5.8S rRNA, where failure to cleave
in the RNase P mutant might lead directly to accumulation of the 3'-extended 5.8S rRNA seen in Figures
2
and
4
. Possible indirect routes for accumulation of this rRNA are explored in the
Discussion.
In addition to alterations in the population of the tRNA precursors in a yeast nuclear RNase P RNA mutant, there was an accumulation of a novel, abundant RNA slightly larger than 5.8S rRNA. The 5'-end of the aberrant RNA was shown to correspond to the mature 5'-ends of the 5.8S
L
and 5.8S
S
rRNAs, but the 3'-end extended 35 nt beyond the 3'-end of the mature 5.8S
rRNA into ITS2. The accumulation of this novel species of rRNA suggests that the
RNase P holoenzyme is required either directly or indirectly for proper
formation of the mature 5.8S rRNA 3'-end.
Another strain with a mutation in the RNA subunit of RNase MRP also accumulates
an rRNA species containing 5.8S sequences. However, in that strain the aberrant
5.8S rRNA has a normal 3'-end, but the 5'-end is extended by 149 nt (
33
-
36
). The accumulation of this 5'-extended 5.8S species along with a dramatic shift in the ratio of
the mature forms toward the 5.8S
L
form, led to the conclusion that RNase MRP is involved in proper 5.8S
S
5'-end maturation. Also accumulated in the conditional RNase MRP
mutant is the proposed precursor to 5.8S
S
whose 5'-end lies at A2 in ITS1 (Fig.
3
B), supporting a direct precursor to product relationship for RNase MRP
maturation of the 5'-end of 5.8S
S
rRNA (
65
). The data presented here suggest that RNase P may affect maturation of the
5.8S rRNA in a way that is different from the proposed role for RNase MRP in
5.8S
S
5' processing.
Another yeast enzyme involved in 3' processing of 5.8S has been reported recently to be a 3' -> 5' exonuclease activity, Rrp4p (
66
). A strain carrying the mutant allele,
rrp4-1
, accumulates 5.8S rRNAs with 3' extensions. One of the species characterized in
rrp4-1
and wild-type RRP4 is the approximate length of the 3'-extended form of the 5.8S rRNA described here, suggesting
that this pre-rRNA may be a previously undescribed rRNA precursor resulting from the
characterized processing pathway or from an alternative pathway. It remains
possible that the aberrant RNA that accumulates in the mutant is a cleavage
product derived from the larger 27S pre-rRNAs earlier in the processing pathway, rather than from the 7S
precursors.
Examination of large rRNA precursors of wild-type and mutant RNase P RNAs did not reveal large changes in the
normalized levels of the 27S (Fig.
4
D) and 35S (not shown) rRNAs, nor was there an appearance of any aberrant rRNAs
other than the 3'-extended 5.8S species. However, a more careful examination (Fig.
5
) of the levels of the 27Sa major precursor to 5.8S and 25S rRNA (Fig.
3
) by primer extension analysis revealed an increase in the levels of precursor
with 5'-ends at processing site A2 in the RNase P mutant. RNase P appears
to exert at least some of its influence after A2 cleavage, possibly
participating in an uncharacterized pathway that, if interrupted, leads to the
accumulation of the observed rRNA with additional 3' sequence.
Although RNA folding algorithms did not predict a tRNA-like RNase P substrate structure at the 3'-end of 5.8S rRNA, an attempt was made to use RNase P to
cleave a pre-5.8S rRNA substrate
in vitro.
The nuclear RNase P cleavage positions are indicated in Figure
8
relative to the nucleotide sequence of ITS1, 5.8S and ITS2. The simplest result
from these experiments would have been the existence of a strong cleavage site
at the end of 5.8S rRNA
in vitro
, thus easily explaining why 3'-extended 5.8S rRNA accumulates in the RNase P mutant. No such
cleavage was observed, consistent with predicted folds not forming a structure
appropriate for an RNase P substrate (
67
-
69
).
The two most prominent RNase P cleavage sites, [brvbar] and [brvbar][brvbar], map to ITS1 (between 290 and 291 and between 326 and
328) within 75 nt of the 5'-end of 5.8S
S
sequences. Site [brvbar] is 2 nt downstream of the proposed processing site A3. Cleavage at A3
is required for production of the mature 5'-end of 5.8S
S
rRNA and is the site of RNase MRP cleavage
in vivo
(
37
,
65
). Common substrate recognition for RNase P and RNase MRP has been shown
previously
in vitro
(
70
) and may be a reflection of the evolutionary relatedness of RNase P and RNase
MRP (
14
,
43
,
46
,
47
,
70
). Site [brvbar][brvbar] does not correspond to any proposed site
in vivo
, but may be a site that is rapidly processed further
in vivo
and has not been described. Clearly, other sites that map at positions within the 5.8S sequence are
unlikely to be physiologically relevant and emphasize that the
in vitro
cleavage assay results must be viewed with caution. The extent of the pre-rRNA, the lack of proteins, the non-physiological folding conditions, and other missing components of
rRNA biogenesis are all potential problems.
It is quite possible that RNase P does not directly cleave pre-rRNAs at all
in vivo
, but rather that RNase P mutations affect rRNA processing indirectly. RNase P
might be part of an rRNA processing structure known as the `processome' (
9
), a large complex in the nucleolus composed of snoRNPs and other associated
proteins. Association of RNase P RNA with at least one protein that also
associates with RNase MRP RNA was shown by co-immunoprecipitation of the RNA subunits of both enzymes by antibodies to a
Protein A-Pop1 protein. RNase P RNA may influence the structure of a complex
containing RNase MRP and the shared Pop1 protein, thereby interfering with
proper 5.8S 3'-end processing through misassembly or disruption of the complex. Alternatively, some
other cleavage product of RNase P, such as a snoRNA, might be required for proper folding or cleavage of the pre-rRNA.
An evolutionary and functional relationship between RNase P and RNase MRP has
been suggested to explain a cleavage at site A3 for RNase MRP (
47
). In this scenario, RNase P originally functioned to cleave a tRNA with a
conserved positioned in the equivalent 16S/23S spacer region of the rRNA
operons from both Bacteria and Archaea to ensure separation of the rRNAs. It
was postulated that RNase MRP evolved from RNase P in eukaryotes to cleave this
region whose tRNA was subsequently lost during the course of evolution. While
RNase P continues to process pre-tRNAs, it is possible that RNase P also remained a member of this complex
retaining a functional role in rRNA maturation, but a role that is now distinct
from RNase MRP.
We are grateful to N. Martin and J. Warner for their gifts of materials. We
thank A. Tranguch, W. Ziehler, Y. Lee, P. Good and D. Thiele for advice,
support and critical evaluation of the manuscript. This work was supported by
National Institutes of Health Grant RO1 GM34869 (to D.R.E.). Oligo synthesis
was partially subsidized by National Institutes of Health Grant P30CA46592 to
the University of Michigan Cancer Center. Computer assistance was facilitated
by National Institutes of Health Grant M01 RR00042 to the University of
Michigan Clinical Research Center. J.R.C. was supported by National Institutes
of Health Pre-doctoral Training Grant T32 GM07315.
REFERENCES
Return