| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 279, Issue 46, 47661-47671, November 12, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
¶
From the
Department of Biochemistry and Molecular Biology, Southern Illinois University, Carbondale, Illinois 62901-4413 and Department of Molecular and Structural Biochemistry, North Carolina State University, Raleigh, North Carolina 27695-7622
Received for publication, August 3, 2004 , and in revised form, September 3, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Box C/D RNAs are also found in Archaea, in which they are designated snoRNA-like RNAs or sRNAs (8–11). Archaeal sRNAs are generally smaller than the eukaryal snoRNAs and typically possess C' and D' boxes that vary little from the terminal box C and D sequences. The primary function of the archaeal box C/D sRNAs also is to guide the 2'-O-methylation of targeted nucleotides, and their mechanism of nucleotide modification is analogous to the eukaryal snoRNAs.
Both the eukaryal and archaeal box C/D RNAs are bound to core proteins to establish snoRNP and sRNP complexes, respectively. Four snoRNP core proteins are bound to the box C/D snoRNAs: fibrillarin (Nop1p), Nop56p, Nop58p (Nop5p), and the 15.5-kDa (Snu13p) protein (4–7). The methylation activity resides in the proteins of the snoRNP, with fibrillarin functioning as the methyltransferase. The differential distribution of the 15.5-kDa Nop56 and Nop58 proteins on the box C/D and C'/D' motifs establishes an asymmetric snoRNP complex (12, 13). The archaeal box C/D sRNP complex possesses three core proteins. Ribosomal protein L7Ae (the archaeal homolog of the eukaryal 15.5-kDa protein), afibrillarin, and aNop5p (a single homolog of eukaryal Nop56p and Nop58p) bind both the terminal C/D and internal C'/D' RNA motifs to establish a symmetrical RNP complex (4, 5, 14–19). In vitro assembly systems using purified archaeal sRNAs and recombinant core proteins have reconstituted enzymatically active box C/D sRNPs (16, 18, 20, 21). Core protein binding follows an order of assembly in which L7Ae binds first followed by aNop5p and then afibrillarin. Efficient catalysis requires that the box C/D and C'/D' RNPs be juxtaposed within the full-length sRNA (18).
The tRNATrp of Haloferax volcanii is derived from an intron-containing pre-tRNA (22) and possesses 2'-O-methylated nucleotides at positions 34 (Cm) and 39 (Um) (where "m" is 2'-O-methylation of the residue) (23). Analysis of the pre-tRNATrp structure (see Fig. 1, upper left) revealed that the intron contains box C/D and C'/D' motifs with guide sequences complementary to the pre-tRNA regions encompassing modified nucleotides Cm34 and Um39, respectively (8, 10, 24). Subsequent investigations that involved deleting regions of the pre-tRNA intron and then examining nucleotide methylation in vitro led to the conclusion that these intron-encoded box C/D motifs are indeed responsible for guiding the methylation of pre-tRNATrp nucleotides C34 and U39 (21, 24). These investigations also led to the proposal that the box C/D RNPs of the unspliced intron are responsible for guiding in cis the methylation of the pre-tRNATrp nucleotides (21, 24). The presence of a cis mechanism for intramolecular methylation would be unique and provide a stark contrast from the box C/D RNP-guided trans-methylation in eukaryal and other archaeal systems. Although box C/D snoRNAs are frequently found in the introns of eukaryal pre-messenger RNAs, they are excised from the host pre-mRNA as snoRNPs before they guide the 2'-O-methylation of target nucleotides.
|
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
In Vitro RNA Synthesis—Generally, in vitro transcription was carried out in 20-µl reactions at 37 °C for 2–3 h in buffer containing 40 mM Tris-Cl, pH 7.9, 6 mM MgCl2, 10mM dithiothreitol, 2 mM spermidine, 10 µCi of [
-32P]ATP (specific activity, 3000 Ci/mmol) (MP Biomedicals), 0.6 mM unlabeled ATP, and unlabeled GTP, CTP, and UTP each at 1.0 mM, PCR-amplified DNA, and 50 units of T7 RNA polymerase (New England Biolabs). Radiolabeled RNA transcripts were purified by denaturing PAGE, and amounts were approximated by Cerenkov counting. High specific activity transcripts were prepared similarly except that unlabeled ATP was omitted.
RNP Assembly and Electrophoretic Mobility Shift Assay—Recombinant M. jannaschii L7Ae, aNop5p, and afibrillarin proteins were prepared, and RNP complexes were assembled as described previously (18). Briefly, 
0.2 pmol of radiolabeled RNA was incubated at 70 °C with 10 pmol of L7Ae in 20-µl reactions (20 mM HEPES, pH 7.0, 150 mM NaCl, 0.75 mM dithiothreitol, 1.5 mM MgCl2, 0.1 mM EDTA, 10% glycerol) for 10 min. Assembly of higher order RNP complexes was accomplished by incubating 10 pmol of L7Ae, 32 pmol of aNop5p, and 33 pmol of afibrillarin with radiolabeled RNA transcripts in the presence of 10 µg of Escherichia coli tRNA. Assembled RNP complexes were resolved by electrophoretic mobility shift assay using 4% polyacrylamide gels containing 25 mM potassium phosphate, pH 7.0, and 2% glycerol as described previously (18). Resolved RNPs were visualized by phosphorimaging using a Packard Cyclone system.
In Vitro RNP-directed Nucleotide 2'-O-Methylation and Thin Layer Chromatography Analysis of Modified Nucleotides—Generally, RNP complexes were assembled in the presence of 0.05 mM S-adenosylmethionine (AdoMet) using recombinant core proteins and 0.2 pmol of [-32P]ATP-labeled RNAs as described above. After incubation at 70 °C for 2 h, radiolabeled RNA was purified by phenol/chloroform extraction and ethanol precipitation. In cases in which two different transcripts were included in the reaction, the amount of each transcript was 0.1 pmol. RNA samples were digested with RNase T2, and the digestion products were resolved on cellulose plates (EM Science) using two-dimensional TLC. The solvents for TLC were isobutyric acid, 0.5 N NH4OH (5:3, v/v) for the first dimension and isopropanol/HCl/H2O (70:15:15, v/v/v) for the second dimension (23). Radiolabeled nucleotides resolved by TLC analysis were visualized and quantified by phosphorimaging. The identity of dinucleotides was established based on our previous study (23). A small aliquot of RNA was obtained before the RNase T2 digestion and saved. These RNAs were checked by denaturing PAGE for RNA stability and integrity, especially for those cases in which methylation was not observed. Target nucleotide 2'-O-methylation was calculated by dividing the amount of radioactivity in the corresponding dinucleotide spot on the TLC plate by 33 (number of A residues in the pre-tRNA) of the sum of the total radioactivity in all spots and expressed as the percentage of modification.
The amount of radiolabeled wild type pre-tRNA transcript used in the reactions and their incubation times varied (up to 2 h) for the kinetic study experiments (see Fig. 2D). Incubation at 70 °C beyond this time leads to RNA degradation. In some reactions, small amounts (<0.08 fmol) of high specific activity transcript were included as tracers to quantify the minimal levels of detectable modification. Two specific and more sensitive experiments using the G117C mutant pre-tRNA were carried out. In one experiment, instead of the typical 0.2 pmol, 2 pmol of pre-tRNA was used for the standard 2-h reaction. For wild type pre-tRNA, more than 80 and 40% of C34 and U39, respectively, were methylated under these conditions (see Fig. 2D). In another experiment, 0.2 pmol of normally radiolabeled and 2.25 fmol of high specific activity tracer transcript (in place of the typical <0.08 fmol transcript) were used in the reaction. We are able to detect methylation of a given residue as low as 2.5% under these conditions.
|
Pre-tRNA Splicing Reactions—Splicing endonuclease and ligase reactions to produce linear and circular forms of H. volcanii pre-tRNATrp introns were carried out as described previously (25, 26). RNA products were separated by denaturing PAGE, and specific introns were eluted from the gels.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
H. volcanii pre-tRNATrp as well as the linear and circular forms of its spliced intron can bind recombinant M. jannaschii box C/D sRNP core proteins in vitro to assemble sRNP complexes (Fig. 1, lower panels). Consistent with previous investigations (16, 18, 20, 21), electrophoretic mobility shift assay also demonstrated that core protein binding is ordered, with L7Ae binding first followed by aNop5p and then afibrillarin (data not shown). The mouse U14 box C/D core motif (19) is an effective competitor for protein binding (data not shown), indicating that core protein binding is specific for the box C/D motif. When the RNA products of a pre-tRNATrp splicing reaction were incubated with L7Ae, only the introns (both linear and circular forms) bound this core protein (data not shown). Thus, box C/D RNP assembly requires the box C/D and C'/D' RNA motifs contained within the intron of this pre-tRNA.
In Vitro Assembled Box C/D RNPs Guide the 2'-O-methylation of Target Nucleotides in Pre-tRNATrp—Box C/D RNPs assembled on the H. volcanii pre-tRNATrp with the three recombinant M. jannaschii sRNP core proteins guided the AdoMet-dependent, 2'-O-methylation of C34 and U39 target nucleotides in the tRNATrp precursor. Following RNP assembly and incubation in the presence of AdoMet, TLC of the nucleotides generated by RNase T2 digestion of the [-32P]ATP-labeled pre-tRNATrp transcript revealed radiolabeled residues corresponding to the dinucleotides CmCp and UmCp (where "p" is 3'-phosphate of the residue) (Figs. 2, B and C, and 3A). (RNase T2 cleaves after every ribonucleotide to produce ribonucleoside 3'-monophosphate (Np), except when nucleotides are methylated at the 2' position of the sugar.) Dinucleotides CmCp and UmCp are derived from 2'-O-methylation of residues C34 and U39, respectively. Specificity of methylation for these two nucleotides has been confirmed by dNTP concentration-dependent, primer extension (27) analysis (data not shown) as well as by mutation studies described later. Nucleotide methylation required box C/D RNP assembly because neither dinucleotide was observed when any one of the three sRNP core proteins was omitted from the reaction (data not shown). These studies therefore demonstrate that the in vitro assembled box C/D RNPs function to guide the 2'-O-methylation of pre-tRNATrp nucleotides C34 and U39. The amount of the CmCp dinucleotide was always greater than the UmCp dinucleotide (Fig. 2, B and C, and Fig. 3A), suggesting that modification of U39 follows that of C34 and/or occurs at a significantly slower rate.
|
At the highest pre-tRNA concentration (2 pmol), the sum of C34 and U39 modifications exceeded 100% at time points of 30 min or more (Fig. 2D). This indicated that at least some transcripts are modified at both positions and precluded the possibility that, in this assay, only one of the nucleotides could be modified in any given pre-tRNA molecule. In these studies it was also noted that for all pre-tRNA concentrations tested, C34 methylation was more rapid than that of U39 (Fig. 2D). Additionally, there was a lag in U39 methylation, and its rate of modification as well as its amount was always lower than that of C34. These kinetics suggest that the methylation of these two nucleotides in a pre-tRNA molecule is sequential, with U39 methylation occurring after C34 is modified (see below).
2'-O-Methylation of Pre-tRNATrp Nucleotides C34 and U39 Is Sequential and Prefers but Does Not Require Watson-Crick Pairing between the Target and Guide Nucleotides—Analysis of 2'-O-methylated pre-tRNAs having single mutations at the target and/or corresponding guide nucleotide revealed that Watson-Crick pairing between guide and target residues was not absolutely required for methylation. The importance of target and guide nucleotide pairing was first assessed for the box C/D RNA motif by mutating target nucleotide C34 and/or corresponding guide nucleotide G117. These experiments revealed that the non-canonical base pairs G:G, U:G, A:G, C:A, and C:U yielded methylated target residues, albeit at lower efficiencies (pre-tRNA mutants C34G, C34U, C34A, G117A, and G117U in Table I and Fig. 3B). For all of these non-canonical pairs, the amount of modification at position 34 exceeded that of U39, again consistent with sequential modification of these two nucleotides as observed in our kinetic studies of the wild type pre-tRNA. Surprisingly, mutation of G117 to C117 to establish C:C pairing of target and guide nucleotides resulted in the loss of methylation of both C34 and U39 (pre-tRNA mutant G117C in Table I and Fig. 3C). This observation was verified by more sensitive analyses of nucleotide methylation. Specifically, a high concentration (2 pmol) of G117C pre-tRNA was used in one experiment, whereas a very high amount of high specific activity G117C tracer was included in another. Both experiments yielded negligible and non-quantifiable amounts of modified nucleotides, confirming that C:C pairing between the C34 target and C117 guide nucleotides of G117C pre-tRNA did not result in C34 methylation during our standard analysis. However, methylation of both mutant G34 and wild type U39 was restored in the double mutant pre-tRNAs (C34G/G117C) in which Watson-Crick pairing between the guide and target nucleotides is re-established by creating a second, compensatory mutation, i.e. C34G, in the target nucleotide of the G117C pre-tRNA (mutant C34G/G117C in Table I and Fig. 3D). The absence of U39 methylation in the single mutant G117C and its restoration in double mutant C34G/G117C support the conclusions of our kinetic studies indicating that modification of U39 requires prior modification of nucleotide 34. The G117C mutation that disrupts box C/D-guided methylation at position 34 also disrupts U39 methylation guided by the non-mutated, wild type box C'/D' motif; the pre-tRNA here contains normal guide A70 nucleotide, which is complementary to the U39 target (Fig. 3C). On the other hand, alternative Watson-Crick as well as non-canonical base pairs between the target and guide nucleotides of the box C/D motif that promote methylation at position 34 do allow subsequent C'/D' RNA-guided U39 methylation (Fig. 3, B and D, and mutants C34G, C34U, C34A, G117A, G117U, and C34G/G117C in Table I). The fact that U39 methylation levels were below those at position 34 in these cases (Table I) is again consistent with sequential nucleotide methylation.
|
Unspliced Pre-tRNATrp Can Guide the Intermolecular or trans-2'-O-Methylation of Pre-tRNATrp Nucleotides at Positions 34 and 39—Our kinetic studies suggested an intermolecular or trans mechanism for the methylation of pre-tRNA nucleotides C34 and U39. To assess this possibility, an assay was developed in which two different tRNATrp precursors were co-incubated with recombinant core proteins in the same methylation reaction. The first pre-tRNA of this pair was the box C/D mutant G117C. This pre-tRNA presents a C34:C117 target guide nucleotide pair and is incapable of box C/D-guided methylation (Fig. 3C and Table I). The second pre-tRNA mutant of the pair possessed various mutations at target site 34 (mutants C34G, C34U, or C34A; three different mutant target nucleotides but not the wild type C). The box C/D guide nucleotide of this second pre-tRNA was always G117 (wild type) and would therefore be capable of guiding C34 (wild type) methylation of the first pre-tRNA using a trans mechanism. Thus, the appearance of the dinucleotide CmCp in TLC analysis for any of these three co-incubated pre-tRNA pairs could only result from intermolecular or trans-methylation of the non-mutated target nucleotide C34 of the first pre-tRNA by the non-mutated guide G117 nucleotide of the second pre-tRNA and not by cis-methylation by either of the two pre-tRNAs in the assay. For all three pre-tRNA pairs tested (Table I and Fig. 4A), a CmCp dinucleotide was detected, indicating that trans-methylation does occur. The appearance of the dinucleotides GmCp, UmCp, and AmCp for the respective pairs (Table I) indicated 2'-O-methylation of target nucleotide 34 in the second pre-tRNA molecule. These second pre-tRNA methylations could result from trans-methylation but are not inconsistent with a cis mechanism. The appearance of UmCp for all three pairs was the product of sequential box C'/D'-guided U39 modification for both pre-tRNAs, except for mutant C34U in which methylation of residue U34 in the second pre-tRNA molecule would also contribute to UmCp levels.
|
H. volcanii Pre-tRNATrp C34 and U39 Methylation in an H. volcanii Cell Extract Is Also Sequential and Guided Exclusively via a trans Mechanism Utilizing Pre-existing Box C/D RNPs Contained in the Extract—Previous investigators (21, 24) have examined archaeal pre-tRNATrp 2'-O-methylation in vitro using both recombinant proteins and cell extracts and have concluded that nucleotide modification occurs via a cis mechanism. This conclusion contrasts with our observations of trans-methylation when the pre-tRNATrp box C/D RNPs are reconstituted with recombinant proteins. We therefore decided to also carry out pre-tRNATrp methylation in the cell extracts (Fig. 3, AA–GG) using the same pre-tRNA mutants used to study methylation of the box C/D RNP complexes assembled with recombinant proteins (Fig. 3, A–G). TLC analyses of the reaction products produced for the various pre-tRNA mutants incubated in cell extracts is presented in parallel in Fig. 3 to allow direct comparison of the two in vitro methylation systems. Incubation of wild type H. volcanii pre-tRNATrp transcripts in H. volcanii cell extract clearly resulted in the 2'-O-methylation of C34 and U39 in the pre-tRNA as anticipated (Fig. 3AA). A time course analysis of nucleotide methylations in cell extract (data not shown) for the wild type pre-tRNATrp revealed similar reaction kinetics as that with the complexes assembled with recombinant proteins, consistent with the sequential methylation of C34 and U39.
Methylation of neither G34 (mutant) nor U39 (wild type) nucleotide of the C34G pre-tRNA was observed when this pre-tRNA was incubated in the extract (Fig. 3BB), even when the pre-tRNA substrate concentration was doubled and the cell extract was increased 10-fold (data not shown). The lack of G34 methylation was unexpected because the same mutant pre-tRNA assembled with recombinant proteins was able to utilize this purine:purine/target:guide pair (G34:G117) to guide methylation (Fig. 3, compare B with BB). This suggested that the cell extract system does not permit non-Watson-Crick pairing for box C/D-guided methylation. The coincident lack of U39 methylation, however, would be anticipated because the methylation of this target nucleotide sequentially follows and requires nucleotide 34 modification. Unexpectedly, when pre-tRNA G117C, a mutant pre-tRNA that does not show any modification when incubated with recombinant proteins (Fig. 3C), was incubated in the cell extract, both target nucleotides C34 and U39 were methylated (Fig. 3CC). This suggested that mutant G117C is functioning as the methylation substrate in the cell extract. The results of these C34G and G117C pre-tRNA mutants prompted us to consider the possibility that pre-tRNA modifications in the cell extract were occurring exclusively via a trans mechanism and were utilizing endogenous box C/D RNPs.
To examine these possibilities, additional pre-tRNATrp mutants were incubated in the cell extract, and methylation of the two target nucleotides was assessed. Surprisingly, incubation of the pre-tRNA double mutant C34G/G117C, which contains the compensatory G34:C117 pair, resulted in no methylation of either G34 or U39 target nucleotides (Fig. 3DD). This pre-tRNA double mutant did show both G34 and U39 methylation when it was incubated with recombinant core proteins (Fig. 3D). The lack of G34 methylation in the cell extract suggests the inability of this double mutant pre-tRNA to guide its own methylation in cis as well as to guide methylation of another identical pre-tRNA in trans. Again, a lack of U39 modification in the absence of nucleotide 34 modification in the double mutant is consistent with the normal sequential methylation of the two target nucleotides. In sum, the results of C34G and G117C single mutants and their combinatorial double mutant are consistent with the proposition that the pre-tRNA fails to assemble a competent box C/D RNP in the cell extract but serves as a pre-tRNA substrate for trans-methylation guided by endogenous RNP complexes.
Incubation of pre-tRNA mutant U39A in the extract (Fig. 3EE) resulted in C34 methylation but not mutant target A39 methylation, again suggesting that modification of target nucleotide 34 occurs independently of nucleotide 39. Target nucleotide C34 would be methylated by endogenous RNP using wild type guide nucleotide G117, but target nucleotide A39 cannot be methylated by A70 guide nucleotide of either the U39A mutant pre-tRNA (cis or trans) or the wild type endogenous RNP. Pre-tRNA mutant A70U, which is not methylated at U39 with recombinant proteins (Fig. 3F), does reveal U39 methylation in the cell extract (Fig. 3FF). This is again consistent with the utilization of endogenous RNP containing wild type A70 guide nucleotides. Finally, the pre-tRNA double mutant U39A/A70U, in which Watson-Crick pairing (A39:U70) between target and guide is restored, thus producing A39 methylation with the recombinant proteins (Fig. 3G), failed to be methylated at A39 in the cell extracts (Fig. 3GG). Although this double mutant could, in principle, be guided by its own pre-tRNA either in cis or trans, the absence of wild type target nucleotide U39 prevented its methylation via endogenous box C/D RNPs. This is similar to the absence of nucleotide 34 methylation of the double mutant C34G/G117C (Fig. 3DD) when incubated in the cell extract.
Collectively, pre-tRNA methylation carried out in the cell extract led to the following conclusions: (a) pre-tRNATrp precursors added to the cell extract function only as substrate RNAs and not as guide RNAs; (b) the failure of exogenously added pre-tRNAs to guide methylation suggests that these pre-tRNAs are not assembling functional C/D RNPs; (c) pre-existing wild type C/D box guide sRNPs are acting in trans to methylate the exogenously added pre-tRNAs at both target positions; (d) 2'-O-methylation of residues at positions 34 and 39 occurs sequentially in the cell extract as it does when the C/D RNPs are assembled with recombinant core proteins; and (e) methylation of target nucleotide 39 requires prior and independently occurring modification of target nucleotide 34.
Both Linear and Circular Forms of the Excised H. volcanii Pre-tRNATrp Intron Can Act as Guide RNAs for the trans-Methylation of Pre-tRNATrp—Excised linear and circular forms of the pre-tRNATrp intron possess both box C/D and C'/D' motifs and assemble box C/D RNPs (Fig. 1). We therefore assessed the ability of these excised introns to guide the methylation of their pre-tRNA target nucleotides via a trans mechanism. Both linear and circular introns were isolated and used in separate methylation reactions with recombinant core proteins that also contained G117C pre-tRNA mutant (C34:C117 pair), a pre-tRNA unable to methylate its C34 target nucleotide (see Fig. 3C). In reactions with either a circular or linear intron, pre-tRNA nucleotide C34 was methylated (Fig. 5, A and B). The methylation of C34, in turn, permitted the sequential methylation of U39 and hence the appearance of the dinucleotide UmCp. A time course study using linear introns as guide RNAs also confirmed the sequential modification of these two nucleotides with respect to relative amounts of CmCp versus UmCp produced and a coincident lag in U39 methylation (data not shown). Similarly, both linear and circular forms of the excised intron are able to guide methylation of C34 within the 78-nt 5' halfmer pre-tRNA molecule (Fig. 5, D and E), thus demonstrating that trans-methylation of C34 does not require a folded, full-length pre-tRNA. Electrophoretic analysis of these RNAs after completion of the pre-tRNA methylation confirmed that the circular intron had not been linearized during the reaction (Fig. 5F), thus demonstrating that circular RNA can also assemble a functional guide RNP. Finally, we have also attempted to methylate target nucleotide U39 of a pre-tRNA 3' halfmer using both linear and circular introns. Strikingly, neither form of the intron-encoded RNP was able to modify U39, either when the 3' halfmer was alone or co-incubated with the 5' halfmer (data not shown). Absence of U39 methylation in these cases may suggest that this nucleotide modification requires a correctly folded, full-length pre-tRNA precursor or that misfolding of the 3' halfmer does not allow it to properly pair with the intron-encoded guide RNP.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
H. volcanii pre-tRNATrp can be assembled in vitro into a functional box C/D RNP with recombinant core proteins, but RNP assembly apparently does not occur in the H. volcanii cell extract. Thus, pre-existing box C/D RNPs in the extract are responsible for guiding the methylation of exogenously added pre-tRNA substrate. Clearly, the difference in protein composition (isolated recombinant core proteins versus total cell protein extract) between the two in vitro systems and their ability or inability to assemble functional box C/D RNPs are responsible for the observed differences in methylation results for the same pre-tRNA substrates (Fig. 3, compare A–G with AA–GG). Differences in methylation capabilities among various pre-tRNATrp substrates assembled with recombinant core proteins and total cell extracts have been observed previously (compare results of Ref. 21 with 24). This suggests that caution must be exercised when using these different in vitro box C/D RNP assembly systems because the results obtained in one system may not be directly comparable with those observations and conclusions made in the other. Observed differences may be attributed to additional (accessory?) proteins available in the extract that affect the efficiency of the methylation reaction that are not present in the minimal system using recombinant core proteins.
Mutagenesis experiments revealed that Watson-Crick base pairing between target and guide nucleotides was preferred but not absolutely required for the target nucleotide 34 methylation by recombinant proteins. Utilization of non-canonical pairing between target and guide nucleotides has also been noted previously for eukaryotic box C/D snoRNAs in vivo (28). Notably, disruption of Watson-Crick pairing between the C'/D' target nucleotide 39 and its guide nucleotide for the tested pairs resulted in a complete loss of methylation. Interestingly, a comparison of the RNA duplexes formed between the target and guide sequences for the box C'/D' and C/D motifs revealed significant differences in helix stability. The box C/D target: guide duplex has 11 base pairs, six of which are G-C, whereas the C'/D' duplex has eight base pairs, only two of which are G-C. Thus, a reduction in target-guide duplex stability may require more stringent pairing of the target and guide nucleotides as noted previously (28).
The pre-tRNATrp mutants of this study possessed only site-specific, single nucleotide substitution in guide and/or target nucleotides. These minimal changes in primary sequence and RNA folded structure would be expected to present essentially native, folded pre-tRNA substrates for investigation of these RNA-guided nucleotide modification reactions. Previous investigations (21, 24) in other laboratories have also examined the roles of the intron-encoded box C/D and C'/D' motifs in arriving at their conclusion that pre-tRNATrp methylation is accomplished via a cis mechanism. However, these studies utilized pre-tRNA mutant substrates containing small deletions (i.e. removal of boxes D or D') as well as the removal of significant portions (45 and 75 nt) of the intron. We suspect that such changes are likely to alter the folding of the pre-tRNA transcript, thus complicating the interpretation of results regarding the effects these mutations have on box C/D- and C'/D'-guided methylation.
For many of our pre-tRNA mutants incubated as single precursors bound with recombinant sRNP core proteins, methylation could occur via either a cis mechanism involving a single molecule or a trans mechanism involving two identical molecules. Fortuitously, however, one of our pre-tRNA mutants (G117C), presenting as a non-canonical target:guide pair (C34: C117), did not support in vitro assembled box C/D RNP-guided methylation as a single pre-tRNA transcript. This mutant thus provided a pre-tRNA substrate that could be exploited to demonstrate the trans mechanism for guided methylation. These experiments specifically used two distinct pre-tRNAs with various combinations of altered target and guide nucleotides. Results of these experiments conclusively demonstrated box C/D RNP-guided trans-methylation. Further analysis using linear and circular forms of the excised intron confirmed the trans mechanism. A similar methylation system guided by a linear, derivatized form of the Archaeoglobus fulgidus pre-tRNATrp intron has been described previously (20). However, those experiments utilized RNA oligonucleotides as target RNA substrates. Therefore, box C/D- and C'/D'-guided methylation was not carried out with the corresponding target nucleotides positioned in a pre-tRNATrp context as occurs in vivo. Obviously, these RNA substrates would also not provide the opportunity to investigate the temporal relationship of the two-nucleotide modification reactions.
Our experiments have demonstrated that methylation of box C'/D'-guided nucleotide 39 occurs after methylation of box C/D-guided nucleotide 34. To our knowledge, this is the first example of sequential methylation guided by box C/D RNP complexes in either Archaea or Eukaryota. Sequential nucleotide methylation was first suggested in our kinetic studies in which U39 methylation lagged behind C34 methylation in all cases. Lower levels of methylation at position 39 compared with position 34 were observed in all cases as well. Subsequent mutagenesis experiments confirmed this order of nucleotide methylation in which all mutants disrupting position 34 methylation resulted in the loss of methylation at position 39. In direct contrast, methylation of C34 was independent of U39 modification. The sequential methylation of nucleotides guided by the box C/D and C'/D' motifs of tRNATrp intron raises interesting questions about the temporal relationship of the methylation reactions of other box C/D RNAs that guide more than one nucleotide modification reaction. Is sequential modification a common characteristic of guide RNAs that direct two modification reactions? Is the sequence of modification always C/D-guided and then C'/D'-guided? Is the order of methylation in the target RNA always 5'-nucleotide first and then 3'-nucleotide? Does the second nucleotide modification reaction require conformational changes in target and/or guide RNA? Does methylation by the box C/D RNP convert a juxtaposed but inactive C'/D' complex into an active form, possibly involving inter-RNP interactions? Kinetic analysis of the C34 methylation reaction indicating more than one step is consistent with the notion that conformational changes in the sRNA or sRNP could play an important role in catalysis. Our establishment of an in vitro pre-tRNATrp nucleotide methylation system using recombinant core proteins now provides the opportunity to address some of these questions directly.
It has been assumed that H. volcanii pre-tRNATrp methylation of nucleotides C34 and U39 in H. volcanii cell extracts were carried out via a cis mechanism (21, 24). However, our results clearly demonstrate that pre-tRNAs added to the cell extract do not serve as guide RNAs. We suspect that free core proteins, even if present in sufficient concentrations in the extract, fail to assemble methylation-competent box C/D RNPs on the full size pre-tRNA. Methylation of pre-tRNA nucleotides C34 and U39 in these extracts indicated that endogenous tRNATrp precursors or excised introns possessing assembled box C/D RNPs are responsible for the observed nucleotide methylations; these reactions clearly involve a trans mechanism. As mentioned above, the previous conclusion regarding a cis mechanism for pre-tRNATrp methylation was based on the results of experiments incubating deletion mutant pre-tRNAs in the extract (24). Some of these pre-tRNAs could not be methylated in the extract. Again, we suspect that the structural changes in those pre-tRNA substrates resulting from deletion of the box elements or large regions of intron likely resulted in structural rearrangements of the pre-tRNA molecule. These alterations could have significantly affected the capabilities of these mutant pre-tRNAs to function as target RNA substrates. Structural rearrangements within the pre-tRNA interrupting its pairing with the endogenous box C/D RNP of the cell extract would not allow methylation.
Our results suggest a novel model for intron-encoded, box C/D RNP-guided 2'-O-methylation and splicing of the H. volcanii pre-tRNATrp in vivo (Fig. 6). Newly transcribed tRNATrp precursors serve as the substrate for box C/D RNP-guided 2'-O-methylation of target nucleotides C34 and U39. These methylation reactions are guided sequentially and in trans by pre-existing, intron-encoded, box C/D and C'/D' RNPs. The 2'-O-methylated pre-tRNAs would then be spliced to produce tRNATrp and circular intron. This splicing reaction would produce additional introns that could guide, in trans, pre-tRNATrp methylation. The linear introns observed in vivo (25) could be either intermediates of the splicing reaction or linearized forms of the circular intron. The point at which the intron-encoded box C/D RNPs are assembled is not clear and could occur at different steps in this pathway or even in stages. The full complement of core proteins could bind the pre-tRNA molecule before methylation, thus providing the possibility that this pre-tRNA could also guide methylation in trans. It is also possible that RNP assembly of the pre-tRNA is only partial, perhaps binding only L7Ae. Partial core protein binding could influence the folded structure of the intron and/or contribute to the ability of the substrate to be methylated. Alternatively, L7Ae could act as a chaperone to help fold the pre-tRNA into a splicing competent structure. In this scenario, L7Ae would remain bound to the spliced introns and subsequently recruit aNop5p and afibrillarin to assemble the complete box C/D sRNPs. Alternatively, box C/D RNP assembly could occur entirely on the excised intron.
|
|
FOOTNOTES |
|---|
¶ To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology, Southern Illinois University, Carbondale, IL 62901-4413. Tel.: 618-453-6466; Fax: 618-453-6440; E-mail: rgupta{at}siumed.edu.
1 The abbreviations used are: sno, small nucleolar; sRNA, sno-like RNA; sRNP, sno-like RNP; RNP, ribonucleoprotein; nt, nucleotide; AdoMet, S-adenosylmethionine.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati 
