| Nucleic Acids Research | Pages |
Two 5[prime]-ETS regions implicated in interactions with U3 snoRNA are required for small subunit rRNA maturation in Trypanosoma brucei
Introduction
Materials And Methods
Cell culture and RNA isolations
RNA methodologies
Plasmid constructs and trypanosome transfections
Production of T.brucei transformants
Results
Two distinct 5[prime]-ETS cleavages occur in T.brucei pre-rRNAs
5[prime]-ETS cleavage is not the requisite first step of processing in T.brucei
Placement of sequence tags within SSU and LSU1 rRNAs
Two 5[prime]-ETS regions that crosslink with U3 are required for SSU rRNA accumulation
The 5[prime]-ETS site1a element potentially pairs to U3 snoRNA
Discussion
Acknowledgements
References
Two 5[prime]-ETS regions implicated in interactions with U3 snoRNA are required for small subunit rRNA maturation in Trypanosoma brucei
Received May 5, 1999; Revised and Accepted June 24, 1999
ABSTRACT Early pre-rRNA processing events were examined in the ancient protozoan parasite Trypanosoma brucei and found to have both distinctive and conserved features. Two 5[prime]-ETS cleavages occur: A[prime] and the newly discovered A0. A[prime] and A0 appear related to vertebrate and yeast primary pre-RNA cleavage sites, respectively. However, trypanosomatid primary rRNA transcripts can first be processed at the ITS1/5.8S boundary and 5[prime]-ETS sequences then removed by consecutive cleavages at A[prime], A0 and A1 at the 5[prime]-ETS/SSU rRNA junction. 5[prime]-ETS sequences previously crosslinked to U3 snoRNA were tested for their roles in rRNA processing using our new tagged rRNA system. Two distinct A[prime]-adjacent sequence elements, which may pair with U3 hinge bases, were specifically required for SSU rRNA production, as was a downstream element. The latter element appears conserved with the yeast 5[prime]-ETS U3 binding sequence, required for A0, A1 and A2 cleavages, in that they both share 10 bases complementary with U3 hinge sequences and lie upstream from A0 and A1 sites located in a potential stem-loop structure. The distinctive positioning of putative trypanosomatid U3 binding sites with respect to A[prime] and A0 cleavages suggests that different U3-dependent mechanisms may direct each processing event.
INTRODUCTION
The complex process of eukaryotic ribosomal biogenesis, primarily understood from studies in vertebrates and yeasts (reviewed in 1-3), occurs in the nucleolus which assembles on tandemly repeated copies of rRNA genes. The primary RNA polymerase I transcript contains 18S, 5.8S and 25-28S ribosomal RNAs flanked and separated by external (ETS) and internal (ITS) transcribed spacer regions, respectively. Post-transcriptional processing involves endo- and exonucleolytic activities, nucleotide modification and assembly of mature rRNAs into ribosomal subunits. Transcribed spacer sequences vary widely in size and structure between organisms, as do the positions of cleavage sites within pre-rRNAs. A generally conserved order of processing events commences by cleavage within 5[prime]-ETS sequences, followed by cuts within ITS1 to separate pre-small subunit (SSU) and pre-large subunit (LSU) rRNAs which are subsequently resolved to mature rRNA species. Several small nucleolar RNAs (snoRNAs) and their associated proteins are required for pre-rRNA cleavages (reviewed in 2,4-5; 6). U3, U14, yeast snR30 and snR10, and vertebrate U22, E1 and E2 snoRNAs are needed for 18S rRNA maturation; whereas LSU rRNA production requires U8 in vertebrates. U3 snoRNA is multifunctional; it affects not only pre-rRNA cleavages flanking 18S rRNA sequences, but also early cleavage of the 5[prime]-ETS (7-9) and perhaps cleavage near the ITS1/5.8S boundary (10,11).
The roles of U3 in 5[prime]-ETS processing are relatively well studied, though not fully resolved. In yeast, three U3 snoRNA-dependent cleavages generate the immediate precursor to 18S rRNA (7): A0 within 5[prime]-ETS sequences, followed by A1 at the 5[prime]-ETS/18S rRNA boundary, then A2 within ITS1. A0 occurs near the 3[prime]-end of the 5[prime]-ETS, only 90 nt upstream of the A1 cleavage that forms the mature end of the SSU rRNA. In contrast, the vertebrate primary pre-rRNA cleavage occurs relatively near the 5[prime]-end of the 5[prime]-ETS. This cleavage has been duplicated in mouse and Xenopus cell extracts using synthetic 5[prime]-ETS substrates and shown to be a U3-dependent event that occurs in large ribonucleoprotein complexes (8,9,12). These complexes may be equivalent to the 5[prime]-terminal balls visualized on nascent pre-rRNAs in chromatin spreads (13). U3 hypothetically nucleates formation of a multi-snoRNP processing complex on the 5[prime]-ETS in both vertebrates and in yeasts (4,14). The 5[prime]-ETS cleavage events noted between organisms are distinctive, not only in their relative locations within the 5[prime]-ETS, but perhaps in the manner in which they are processed. The primary 5[prime]-ETS cleavage site of vertebrates may be digested by a single-strand-specific riboendonuclease (15,16), whereas the yeast A0 site occurs in a stem structure that may be cleaved by a RNase III-like enzyme (17).
U3 snoRNA evidently functions through base pairing interactions with pre-rRNAs. Direct contacts between the 5[prime]-domain of U3, including conserved box A and variable hinge region bases and various 5[prime]-ETS sequences, have been implicated by psoralen RNA crosslinking in yeast, mammalian and trypanosome cells (18-22). Only one of these, a yeast U3-5[prime]-ETS crosslink, has yet been shown to indicate a functionally relevant interaction. This crosslink occurred 139 nt upstream of A0, within a sequence having 10 perfect bases of complementarity with the U3 hinge region. Deletion or mutation of the 5[prime]-ETS binding mimicked the effects of U3 depletion in yeast (23). Further genetic complementation analysis demonstrated that base pairing between U3 and this 5[prime]-ETS sequence was required for A0, A1 and A2 cleavages and subsequent SSU rRNA production (24). Phylogenetically supported base paired interactions have been proposed between U3 box A bases and sequences that form the 5[prime] pseudoknot structure of SSU rRNAs (25,26). In fact, recent genetic complementation experiments conducted in yeast have shown that a U3 interaction with the 5[prime]-end of 18S rRNA is required for A1 and A2 cleavages (27).
Studies of rRNA processing in the ancient eukaryote Trypanosoma brucei have revealed both common and unusual features of this pathway. Trypanosomatid LSU rRNAs are fragmented as a consequence of having novel transcribed spacers dispersed throughout their coding sequences (28-30) and early processing events may be distinctive as well. The major 5[prime]-ETS cleavage of T.brucei, called A[prime] herein, superficially resembles that of vertebrates in its 5[prime] location (28). We previously crosslinked the diminutive T.brucei U3 snoRNA (31) to three positions within the 5[prime]-ETS, and U3 snoRNA bases involved in two of these were clearly located (19). In the U3-site1 crosslink, non-conserved bases in the U3 single-stranded hinge domain contacted processed 5[prime]-ETS sequences 3[prime]-adjacent to the A[prime] site at nt 115-116. This crosslink represents the only known close contact between U3 and a major 5[prime]-ETS cleavage site and implies that U3 may affect its cleavage. The U3-site2 5[prime]-ETS interaction, interesting in that it involved highly conserved U3 box A bases, occurred much further downstream in the 5[prime]-ETS, as did the less characterized U3-site3 crosslink.
In this study, 5[prime]-ETS sequences implicated in contacts with U3 were examined by mutational analysis to test their roles in rRNA processing. We developed a tagged rRNA system for this work. Two of three sites implicated in interactions with U3 by crosslinking were determined to be required specifically for SSU rRNA production. Furthermore, examination of pre-SSU rRNA sequences indicated that trypanosomal pre-rRNA processing is unusual in that cleavage between SSU and LSU sequences can precede 5[prime]-ETS cleavage and that two cleavage sites occur in the 5[prime]-ETS: one of these is the previously determined major cleavage site, called A[prime] herein, and one newly identified site shares structural similarity to the yeast A0 site.
MATERIALS AND METHODS
Cell culture and RNA isolations
Trypanosoma brucei strain 427 procyclic forms were grown in a modified version of BSM medium (32) (custom made by Sigma Chemical Co.), supplemented with 5% fetal bovine serum (FBS) (Gemini-Bio Products, Inc.). Cultures were harvested at a density of 8-10 × 106 cells/ml and RNA was isolated by a modified version of the one-step acid phenol-guanidinium isothiocyante-chloroform extraction method (33). Following the initial extraction of RNA from cells, subsequent acid phenol/chloroform/isoamyl alcohol (APCIA) extractions were done till the interface cleared, followed by chloroform extraction. RNAs were then precipitated from the aqueous layer by addition of 2.5 vol of 100% ethanol (EtOH). The RNA precipitate was collected by centrifugation, washed in 70% EtOH and suspended in water plus 1/10 vol of 3 M sodium acetate, pH 5.2. The suspended RNA was subjected to rounds of APCIA and chloroform extraction, then RNAs were precipitated following addition of 2.5 vol of EtOH and collected by centrifugation.
RNA methodologies
Primer extension analysis utilized 10 µg of total RNA and oligonucleotide cr9, crSSU5[prime] or cr10 (Table 1) that had been 5[prime]-end-labeled employing T4 DNA kinase and [32P]ATP. For the S1 nuclease protection assay, these radiolabeled oligonucleotides were used to prime cDNA synthesis, by asymmetric PCR, from HinfI-digested pG.5[prime]-ETS DNA (34), to produce run-off transcripts that were gel-purified before use in standard protection assays.
Table 1. Oligonucleotides used in indicated projects
RNAs were separated on 1% agarose formaldehyde gels, then transferred to Pall Biodyne Plus membranes by the capillary method. These were crosslinked with 254 nm UV light, then baked for 1 h at 80°C under vacuum. Membranes were incubated with radiolabeled oligonucleotide probe ctag1 or ctag2 (Table 1) in 0.25 Na2HPO4, pH 7.0, and 7% SDS at 50°C overnight, then washed three times at 50°C in 3× SSPE (0.54 M NaCl, 30 mM NaPO4, 3 mM EDTA) and 0.2% SDS. Blots were exposed to Kodak XOMAT XAR-5 film. A Molecular Dynamics Storm 860 PhosphorImager was used to analyze radiolabeled RNAs by volume quantitation.
Plasmid constructs and trypanosome transfections
Placement of sequence tags within pG.rDNA sequences (19) was done by standard PCR mutagenesis methods. PCR-amplified DNA fragments were produced using pG.rDNA as the template and these were pieced together in sequential subcloning steps to create the 15 kb pG.rDNAt1t2. The 5[prime] 1.6 kb fragment, made using oligonucleotides rRNA1 and crSSUt1, contained the rDNA promoter, 5[prime]-ETS and 5[prime]-SSU sequences and had synthetic XbaI and BamHI (tag1) ends; the second 2.0 kb fragment, made with rSSUt1 and crSSU3, had BamHI (tag1) and native PstI ends; the third 1.2 kb fragment, made with rSSU1 and crLSUt2, had PstI and KpnI (tag2) ends; the fourth 2.1 kb fragment, made with rLSUt2 and crLSU2, had KpnI (tag2) and PstI ends; the 3[prime] terminal fragment was derived directly from pG.rDNA by digestion with PstI and XbaI. The tagged rDNA gene sequences, present on a 12.1 kb XbaI fragment, were ligated into the XbaI site of a modified pBS.XS2 DNA (kindly donated by Jay Bangs), to make pX.rDNAt1t2. This construct contained the neomycin (NEO) drug resistance cassette driven by the PARP promoter for expression in T.brucei and a unique MluI site at the 3[prime]-terminus of LSU2 coding sequences used to target integration into chromosomal rDNA genes.
Deletion and substitution mutations were made by two-step nested PCR mutagenesis. Using outside primers rRNA1 and crSSU1 and the inside mutagenic primers noted below, two arms were first generated from pG.rDNAt1t2, then these were used as templates to amplify 3.6 kb DNA fragments. Products were subsequently digested with Eco47III and BamHI and ligated into similarly digested pX.rDNAt1t2 to replace wild-type 5[prime]-ETS sequences. Clones were confirmed by chain termination DNA sequence analysis (using Amersham T7 Sequenase v.2.0). Mutagenic inside primers were (see Table 1): crpps and rpps3 for pX.rRNAt1t2s1a; cr19b and rds1b for pX.rRNAt1t2s1b; cr24 and rd1c for pX.rRNAt1t2s1c; crds2a and r7s2.3 for pX.rRNAt1t2s2a; crds2b and r7s2.3 for pX.rRNA-t1t2s2b; cr5 and rd2c for pX.rRNAt1t2s2c; crd3a and rds3 for pX.rRNAt1t2s3a; crsub1a1 and rpps3 for pX.rRNAt1t2sub1a1; crsub1a2 and rpps3 for pX.rRNAt1t2sub1a2. Plasmid DNAs were purified with Qiagen Maxi Preparation.
Production of T.brucei transformants
Plasmid DNAs were digested with MluI, extracted once each with PCIA and chloroform, then precipitated with 1/10 vol 3 M sodium acetate, pH 5.2, and 2 vol 100% ETOH. DNAs were collected by centrifugation, washed with 70% EtOH, then suspended in 100 µl sterile ddH2O. Trypanosoma brucei procyclic forms grown in BSM plus 20% FBS were collected at a cell density of 1 × 107 cells/ml and cells were prepared for transfection according to the BTX Electro Cell Manipulator ECM 600 operation manual, except that cells were suspended at a final concentration of 8 × 107 cells/ml for 3-4 × 107 total cells/4 mm gap cuvette. 500 µl of cells and 100 µg of DNA were electroporated using the 2.5 kV setting, 24 [Omega] resistance and 1.5 V charge. After electroporation, cells were immediately placed into 4 ml of BSM plus 20% FBS and were treated 48 h later with 50 µg/ml G418 (Gibco BRL). Stable transformants were selected within 10-15 days. Drug-selected transformants were then weaned to growth in BSM plus 5% FBS for DNA or RNA isolation. Two to four stable transfectant cultures were obtained for each experimental plasmid construct.
RESULTS
Two distinct 5[prime]-ETS cleavages occur in T.brucei pre-rRNAs
Early analysis of the T.brucei rRNA promoter and processing pathway unveiled a major 5[prime]-ETS processing site, called A[prime] herein, at +115 relative to the start of transcription (28). This site is normally detected as a doublet in reverse transcriptase assays (19). We examined the entire 5[prime]-ETS region for the presence of additional processing sites by reverse transcription using primers spaced ~200 nt apart (data not shown). Single-stranded DNA probes generated from the same primers were used for side-by-side nuclease S1 protection assays to confirm that primer extension stops indicated the 5[prime]-end of a pre-rRNA. A single new processed site, observed as a doublet at nt A1027 and U1026, was detected using the crSSU5[prime] primer complementary to the first 20 nt of SSU rRNA, or cr9 primer complementary to 20 nt near the 3[prime]-end of the 5[prime]-ETS (Fig. 1A and B); the primer extension signal is stronger using crSSU5[prime] (shown) versus cr9 (not shown). The detected site occurs at the 3[prime]-side of the base of a possible stem-loop structure; the 5[prime]-side of the stem base contains the A1 site that forms the 5[prime]-end of SSU rRNA. Similarly, yeast 5[prime]-ETS sequences may form a stem-loop structure having, on either side of its base, the primary pre-rRNA A0 and the A1 cleavages (35,36). The new cleavage site in the T.brucei 5[prime]-ETS thus appears related to the yeast A0 site and is hereafter referred to as such. Stem-loop structures containing the A1 site can be computer modeled for Trypanosoma cruzi, Leishmania amazona and Crithidia fasciculata sequences (data not shown); it is not yet known if these putative structures contain A0-like cleavages. Additionally, a hexamer sequence, ACUUGA in T.brucei and T.cruzi and AAUUGA in L.amazona and C.fasciculata, occurs 5[prime]-adjacent to the A1 site at the 5[prime]-end of SSU rRNA (Fig. 1D). A different hexamer sequence, conserved in the analogous location in yeast 5[prime]-ETS sequences, has a demonstrated role in directing A1 cleavage in Saccharomyces cerevisiae (37,38). Whereas T.brucei A[prime] cleavage is readily detected in primer extension assays at greater than 10-fold the abundance of the transcription start site, the A0 site produces a weaker experimental signal, of the order of 10-fold less than for the A[prime] site (Fig. 1C). This suggests that intermediates cleaved at A[prime] are more stable than those processed at A0; successive cleavage at A1 seems relatively rapid (see also Fig. 2). These data imply structural similarities between yeast and trypanosomes for the 3[prime]-end of the 5[prime]-ETS containing the A0 and A1 cleavages. Yeast do not have an upstream 5[prime]-ETS cleavage site; indeed, the entire 5[prime]-ETS is released by endonucleolytic cleavage at A0 (39). Trypanosoma brucei is thus far exceptional in having both upstream and downstream 5[prime]-ETS cleavage sites.
Figure 1. Identification of the A0 cleavage site in T.brucei 5[prime]-ETS. (A) Primer extension of (1) 40 and (2) 20 µg of RNA using the crSSU5[prime] oligonucleotide is shown. (B) Nuclease S1 protection of 20 µg of RNA using single-stranded cDNA probes made with the cr9 oligonucleotide and HinfI-digested pG.5[prime]-ETS DNA. Hybridization of probe to RNA was done at the three different temperatures indicated; S1 digestion was performed at 37°C. Probe alone plus S1 digestion and untreated probe controls are shown. Dideoxy DNA sequencing ladders were generated for (A) and (B) using appropriate oligonucleotide primers and pG.5[prime]-ETS DNA template. (C) Comparison of primer extension signals for (lane 1) the transcription start site and the A[prime] cleavage (using cr10) and (lane 2) the A0 cleavage (using cr9). The oligonucleotide primers were labeled to the same specific activity. The samples shown are from the same exposure of the sequencing gel on which they were separated. (D) Computer predicted secondary structure of the 3[prime]-end of the 5[prime]-ETS. A0 and A1 (at the 5[prime]-ETS/SSU rRNA boundary) cleavage sites are positioned on either side of the base of a long stem-loop structure. Similar structures can be modeled for T.cruzi, C.fasciculata and L.amazona 5[prime]-ETS sequences that contain the boxed first 4 or 5 nt of SSU rRNA. The boxed hexamer element preceding A1 is also conserved between trypanosomatid species. (E) Positions of T.brucei 5[prime]-ETS cleavage sites are drawn in relation to sites of crosslink formation with U3 snoRNA (19).
Figure 2. Northern analysis of pre-SSU rRNA intermediates. (A) Northern membranes correspond to 25 µg of total T.brucei RNA separated on the same formaldehyde-1% agarose gel. Membranes were hybridized with specific transcribed spacer probes: the 5[prime] 115 nt of the 5[prime]-ETS or intact 5[prime]-ETS, ITS1 and ITS2 sequences, as indicated. Sizes of hybridizing RNAs are indicated in kb, determined by ethidium bromide staining of RNA molecular weight markers (Bio-Rad) and mature rRNA species before RNA transfer to nylon membranes. (B) The T.brucei rRNA processing pathway suggested by analysis of stable processing intermediates indicated by this and previous work (28,34).
5[prime]-ETS cleavage is not the requisite first step of processing in T.brucei
Previous analysis of the T.brucei rRNA processing pathway identified several stable processing intermediates by northern analysis, as well as 5[prime]-ETS A[prime] cleavage by reverse transcription and nuclease protection analyses (19,28). The relative abundance of the primary transcript to 5[prime]-ETS-containing pre-SSU intermediates was equivalent to the ratio of uncleaved to cleaved A[prime], thus it seemed likely that A[prime] represented the primary cleavage of pre-rRNAs as described for higher eukaryotes. We have re-examined early T.brucei rRNA processing events using specific transcribed spacer probes, generated by PCR, to detect new intermediates by northern analysis of RNAs separated on agarose (Fig. 2) or polyacrylamide gels (data not shown).
Our analysis demonstrated that the T.brucei A[prime] cleavage is not a requisite first pre-rRNA processing event. A probe specific for the first 115 nt of the primary transcript detected not only the 9.6 kb primary transcript, but also the new 3.7 kb pre-SSU intermediate that is predicted to extend to the end of the ITS1. The 3.6 kb pre-SSU rRNA apparently contains sequences extending from the A[prime] cleavage to the end of ITS1. It is not yet known if alternative processing of A[prime] in the primary transcript also occurs, but it is clear that cleavage at the ITS1/5.8S boundary can precede this event.
Another newly identified pre-SSU RNA is a 2.6 kb RNA that was detected by the full-length 5[prime]-ETS probe (Fig. 2) and also faintly detected by small 206 nt probes generated by primer extension from the cr9 oligonucleotide (antisense to 5[prime]-ETS bases 902-1117) (data not shown). This intermediate likely contains the A0-A1 5[prime]-ETS sequences, in addition to the SSU and ITS1 sequences that comprise the smaller, and more stable, 2.5 kb pre-SSU rRNA species. An ~100 nt sized RNA that weakly hybridizes to 5[prime]-ETS sequences in small RNA northern analysis may represent an A0-A1 endonucleolytic cleavage product (data not shown).
Together with 5[prime]-end analysis of 5[prime]-ETS intermediates, these data indicate that T.brucei pre-rRNA processing can initiate by cleavage at/near the ITS1/5.8S boundary, followed by two successive cleavages, A[prime] then A0, within the 5[prime]-ETS. Subsequent processing at A1 and then A2 at/near the SSU/ITS1 boundary results in SSU rRNA maturation. It has not yet been determined if an alternative pathway exists, wherein A[prime] cleavage precedes processing at ITS1/5.8S, followed by cleavage at A0 and ensuing 5[prime]-ETS removal. It is clear, however, that the ITS1/5.8S event precedes cleavages at A0, A1 and A2.
Placement of sequence tags within SSU and LSU1 rRNAs
To study the functional necessity of T.brucei 5[prime]-ETS sites implicated in contacting U3 snoRNA by RNA crosslinking (19), it was first necessary to develop an in vivo assay system for analysis of cis-acting elements in rRNA processing. Using the yeast tagged rRNA system as a model (40,41), benign sequence tags were placed within the E10 region of SSU (tag1) and the D1 region of LSU1 (tag2) rRNA coding sequences (see Materials and Methods). Shown in Figure 3, tags were positioned within rRNA variable regions expected to be unnecessary for conserved rRNA functions, as alteration of analogous sequences in yeast rRNAs did not noticably affect ribosome function. These new sequences altered the ends of stem-loop structures predicted to occur in T.brucei and other trypanosomatids, as well as in yeasts (42). Tagged, intact rRNA gene sequences were placed in plasmids containing drug-selectable markers for production in both bacteria and T.brucei. The resultant plasmid, pX.rRNAt1t2, contained a unique MluI restriction site in the 3[prime]-end of LSU rRNA sequences. This site was digested to promote targeted integration of pX.rRNAt1t2, and its later constructed derivatives, into rRNA loci of transfected T.brucei procyclic forms. In drug-selected transformants, integrated DNAs were examined by southern analysis; it was estimated that tagged rRNA genes represented less than 1% of the rRNA genes in the cell (data not shown). Due to this low level of expression, deleterious mutations within tagged rRNA expression units would not be expected to interfere with cell health. RNAs isolated from stable transformant populations contained mature, tagged SSU and LSU1 rRNAs. These had normal 5[prime]-ends, detectable by primer extension from the sequence tags (Fig. 4). Tagged, mature SSU and LSU1 rRNAs were readily detected by northern analysis using antisense tag1 and tag2 sequence probes (Fig. 5, lane 2), but the levels of tagged pre-rRNA intermediates produced in trypanosomes were too low to identify. This system at least provided a means of examining the effects of pre-rRNA mutations on maturation of SSU and LSU rRNAs, as production of tagged rRNAs from mutated pre-rRNAs could readily be monitored above the background of the abundant wild-type rRNA genes (about 300 diploid genes/cell).
Figure 3. Tagged rRNA gene constructs for targeted integration and expression in T.brucei procyclics. (A) The integrating plasmid pX.rRNAt1t2 is drawn. It contains the neomycin drug-selectable cassette flanked by the T.brucei PARP promoter and tubulin intergenic regions and the 12.1 kb rRNA gene-containing fragment with positions of tag1 and tag2 and the unique MluI restriction site indicated. pBlueScript sequences are not diagrammed. (B) The locations and predicted structure of sequence tag1 in SSU rRNA and tag2 in LSU1 rRNA regions are shown, within thick lined boxes, in comparison to predicted structures for wild-type rRNAs (42). Introduced restriction site sequences are indicated by fine lined boxes.
Figure 4. Tagged SSU and LSU1 rRNAs have appropriate 5[prime] ends. Shown are primer extension reactions using 5[prime]-radiolabeled ctag1 (A) and ctag2 (B) primers and template RNAs from control, non-transfected cells (lanes C) and two stable transfectant populations containing pX.rRNAt1t2 (lanes 1 and 2). Marker dideoxy sequencing reactions were done using the same primers and pX.rRNAt1t2 DNA as template. Sequence corresponding to the 5[prime] 10 bases of each transcribed rRNA is indicated.
Figure 5. Deletion analysis of 5[prime]-ETS sequences implicated in interactions with U3 snoRNA. (A) Possible base paired interactions between U3 snoRNA and each 5[prime]-ETS region which contained a site of crosslinking with U3 are shown. Deletions of each 5[prime]-ETS sequence tested are indicated by name and by lines above the pairing schemes. The filled circles and ovals refer to positions of psoralen adducts noted in each RNA moiety in crosslinking studies; U3-site1b, U3-site2a and U3-site2b contained experimentally localized U3-5[prime]-ETS crosslinks, as indicated by dotted lines. The U3-site1a region contains the A[prime] cleavage site, indicated by arrows between nt 114/115 and 115/116. (Refer to the text for a discussion of U3 sequences implicated in each pairing scheme shown.) (B) Northern analysis of RNAs isolated from stable transformants containing integrated pX.rRNAt1t2 or its derivatives. Two similar membranes having 25 µg RNA/lane were probed with antisense ctag1 or ctag2 probes, as indicated. Lane 1 contains wild-type strain 427 RNA, lane 2 contains RNA from transfectants containing pX.rRNAt1t2 and lanes 3-9 contain RNA from transfectants containing the indicated 5[prime]-ETS deletion derivatives of pX.rRNAt1t2. The positions of SSU, LSU1 and LSU2 rRNAs are shown.
Two 5[prime]-ETS regions that crosslink with U3 are required for SSU rRNA accumulation
Three distinct 5[prime]-ETS regions were previously crosslinked to U3 snoRNA, two of which were studied in detail to identify both 5[prime]-ETS and U3 snoRNA bases involved (19; see Figs 1 and 5). To determine whether these crosslinks might indicate functionally relevant U3-5[prime]-ETS interactions, each 5[prime]-ETS component was tested for roles in rRNA processing using the tagged rDNA system (Figs 5 and 6). 5[prime]-ETS sites of interaction with U3 were individually deleted from pX.rDNAt1t2 sequences and these mutant derivatives were transfected into trypanosomes. Constructs were examined by DNA sequence analysis to ensure the presence of appropriate mutations before transfection into trypanosomes. RNAs isolated from drug-selected transformants were assayed by northern analysis for production of tagged rRNAs from mutated pre-rRNAs (Fig. 5).
Figure 6. Substitution analysis of 5[prime]-ETS site1a sequences. (A) The base pairing schemes for wild-type and substituted site1 sequences with U3 snoRNA are shown. (B) Northern analysis of RNAs from transfectants containing sub1a1 and sub1a2 derivatives of pX.rRNAt1t2. The site1a substitution samples are present in lanes 6 and 7, as indicated. They are shown in comparison to wild-type RNA (lane 1) and RNAs from transfectants containing pX.rRNAt1t2 (lane 2) and [Delta]s1a, [Delta]s2a and [Delta]s3 derivatives. See also Figure 5B legend.
The 5[prime]-ETS site1 region contains several interesting features. The A[prime] cleavage site at 115/116 and abutting sequences share base complementarity with U3 5[prime] hinge sequences. The 3[prime]-adjacent sequences contain two blocks of contiguous base pairing potential with the entire U3 hinge region; one of these contains U140 and U142 that crosslinked to U3 3[prime] hinge sequences. These three blocks of sequence were each deleted from tagged rRNA gene constructs and the effects on rRNA processing of transcribed pre-rRNAs examined (Fig. 5, lanes 3-5). [Delta]s1a removed 21 nt including the A[prime] site and bases complementary to U3 5[prime] hinge bases; [Delta]s1b removed 12 nt of sequence containing U140 and U142 that crosslinked to, and could pair with, U3 3[prime] hinge bases; [Delta]s1c removed 15 bases 3[prime] to site1b sequences sharing complementarity to U3 5[prime] hinge bases. Examination of tagged rRNAs by northern analysis showed that deletion of site1a and site1b elements abolished SSU rRNA accumulation, whereas LSU1 rRNA levels appeared unaffected; site1c sequences were not required for rRNA maturation. These results were intriguing as [Delta]s1a provides the first demonstration that specific deletion of a major 5[prime]-ETS cleavage site can abolish SSU rRNA maturation. [Delta]s1b showed that sequences containing bases that crosslink with U3 are required for SSU rRNA maturation. [Delta]s1c showed that although U3 hinge bases have the potential to pair across site1b and site1c sequences, if this interaction occurs, it is non-essential for rRNA processing. This deletion analysis of 5[prime]-ETS site1 region showed that two different site1 elements are required for SSU rRNA accumulation. At least for site1b sequences, which crosslink to U3 bases, this is consistent with deletion of a functionally important U3 binding site required for correct rRNA processing.
The 5[prime]-ETS site2 region contains base U945 that crosslinks with U3 box A bases. An extended base pairing scheme between U3 5[prime] sequences and the site2 region can be drawn. Three deletions within this region were tested for roles in rRNA processing (Fig. 5, lanes 6-8). [Delta]s2a removed 9 nt including the crosslinked base and [Delta]s2c removed 21 upstream nt having complementarity to U3 5[prime] sequences; [Delta]s2b removed both site2a and site2c sequences. None of these deletions affected SSU or LSU1 rRNA maturation. Thus, if U3 truly pairs with 5[prime]-ETS site2 sequences, it is a non-requisite interaction for processing by this analysis. Perhaps the earlier detected crosslink between U3 box A and site2 bases formed fortuitously as a consequence of another, spatially nearby 5[prime]-ETS interaction with U3, such as the 5[prime]-proximal site3 (see below).
The 5[prime]-ETS site3 region contains U862, previously detected as a base that crosslinked to U3 snoRNA (19). A specific crosslinked complex was not isolated and dissected to determine the exact U3 bases involved, as was done for U3-site1 and U3-site2 complexes. Site3 sequences can form 10 perfect base pairs with the 5[prime] hinge region of U3 snoRNA. [Delta]s3 removed 21 nt containing this pairing motif (smaller deletions created new possible U3 binding sites), resulting in loss of SSU, but not LSU, rRNA maturation. This result was intriguing, as yet another possible, necessary U3-5[prime]-ETS contact was identified. Furthermore, the U3-site3 pairing motif involved bears strong similarity to that seen for the functional yeast U3-5[prime]-ETS interaction (24), which also involves 10 bp of complementarity between U3 hinge and 5[prime]-ETS sequences. The putative trypanosome U3-site3 paired interaction is located 164 nt upstream of the A0 cleavage, whereas the yeast interaction is similarly positioned 139 nt upstream. Thus, the trypanosome U3-5[prime]-ETS site3 interaction is possibly the functional homolog of the well-studied yeast U3-5[prime]-ETS interaction required for A0, A1 and A2 processing and SSU rRNA maturation.
The 5[prime]-ETS site1a element potentially pairs to U3 snoRNA
The deletion of site1b and site3a sequences resulted in lack of SSU rRNA production from deleted pre-rRNAs, probably as a consequence of removal of U3 binding sites indicated in crosslinking experiments. Deletion of site1a sequences also led to disruption of processing events leading to SSU rRNA accumulation, yet [Delta]s1a removed both the A[prime] site and sequences sharing complementarity with U3 hinge bases. Site1a sequences were thus substituted to discriminate between the possibilities that loss of the A[prime] cleavage or loss of a U3 binding site led to the observed defects in rRNA processing (Fig. 6). In the sub1a1 construct, nine of 11 contiguous bases were substituted to remove the native A[prime] sequence yet maintain base pairing potential with U3, whereas in the sub1a2 construct, 11 of 11 bases were substituted to maintain the native A[prime] cleavage sequence but to disrupt base pairing potential to U3. Analysis of rRNAs produced in trypanosomes from these constructs showed that a small amount of tagged SSU rRNA was present in sub1a1 strains, whereas none was present in sub1a2 strains; LSU1 rRNA maturation appeared unaffected. Trypanosome populations containing sub1a1 constructs produced ~5% the amount of SSU rRNAs detected from the parent pX.rRNAt1t2 construct. These results implied that the cleavage site itself is important for efficient processing, but is not absolutely necessary, whereas sequences with the potential to pair with U3 are required. Alternatively, deletion and substitution mutagenesis of site1a sequences may have led to disruption of a 5[prime]-ETS secondary structure required for correct pre-rRNA processing. Further molecular genetic and 5[prime]-ETS RNA structural analyses must be done to distinguish between possibilities.
The experiments in Figures 5 and 6 identified three 5[prime]-ETS sequence elements required primarily for SSU rRNA production. To determine whether our 5[prime]-ETS mutations might subtly affect LSU1 rRNA maturation, as certain 5[prime]-ETS mutations appear to in Schizosaccharomyces pombe (36), we attempted to normalize amounts of tagged rRNAs to NEO mRNA produced from integrated pX.rRNAt1t2 derivatives. This proved impractical as, for unascertained reasons, NEO mRNA accumulation varied greatly between transfectant populations (data not shown). The genomic context of a particular integrant probably influenced the level of NEO gene transcription. Therefore, a clear appraisal of 5[prime]-ETS sequence requirements for maximal LSU1 production was not feasible.
DISCUSSION
In this work, we developed a T.brucei tagged rDNA system necessary for analysis of cis-acting sequences in rRNA processing. Using this, we tested the roles of three distinct 5[prime]-ETS regions implicated in pairing with U3 snoRNA 5[prime] sequences by in vivo psoralen RNA crosslinking. Site1b and site3 sequences, which crosslink to U3 hinge bases, are essential for SSU, but not LSU, rRNA accumulation, consistent with roles for these as functionally important U3 binding elements. Site2, which may contact U3 box A bases, is not required for rRNA maturation. Two distinct site1 elements, found 3[prime]-adjacent to the major A[prime] cleavage site, are requisite for SSU rRNA accumulation. Each of these can be modeled to pair with different U3 hinge bases suggesting that multiple interactions between U3 and site1 5[prime]-ETS sequences may affect A[prime] cleavage and/or SSU rRNA production. The site3 sequence could form 10 contiguous base pairs with U3 hinge sequences in a manner that strongly resembles the functional U3-5[prime]-ETS interaction described in yeast. As well, a new T.brucei 5[prime]-ETS cleavage site was identified that may be the homolog of the yeast A0 cleavage. Early pre-rRNA processing in T.brucei is unusual, both in that two internal 5[prime]-ETS events occur and that these can be preceded by processing near the ITS1/5.8S boundary.
Analysis of stable pre-rRNA intermediates has shown that 5[prime]-ETS cleavage is not the requisite first step of processing in T.brucei, as it generally appears to be in later evolved eukaryotes (1-3). U3 has long been postulated to establish the initial pre-rRNA processing complex within the 5[prime]-ETS, but this is clearly not compulsory for LSU rRNA processing. U3 depletion from S.cerevisiae or Xenopus oocytes specifically impedes cleavage events required for SSU rRNA maturation (7,10). Yet proficient SSU and LSU rRNA production appears linked, at least in S.pombe, as illustrated by mutation analyses of 5[prime]-ETS and ITS1 sequences (36,43). Models based on numerous experimental observations in S.cerevisiae propose that U3 may influence the overall efficiency of pre-rRNA processing by establishment of multi-snoRNP complexes that communicate between 5[prime]-ETS and ITS1 cleavage events (4,44). Relatedly, depletion of U3 from Xenopus oocytes leads to slow processing at the ITS1/5.8S boundary (10,11). In some oocytes, this cleavage preceded removal of 5[prime]-ETS sequences, although internal 5[prime]-ETS cleavages were not noted (45). The variation seen in T.brucei, whereby processing at the ITS1/5.8S boundary is succeeded by the two internal 5[prime]-ETS A[prime] and A0, and the A1, cleavages, illustrates further diversity in eukaryotic rRNA processing pathways. The mechanism whereby trypanosome ITS1 sequences are recognized and digested early in pre-rRNA processing might be distinctive, as a single 5.8S RNA species is produced (unpublished data). In higher eukaryotes, two 5.8S species occur that have 5[prime]-ends of different lengths; these are products of alternative ITS1 processing pathways in yeast and presumably in other eukaryotes (46,47).
The T.brucei A[prime] cleavage site bears resemblance to vertebrate primary pre-rRNA cleavage sites in its 5[prime] location in the 5[prime]-ETS and it could serve as an early assembly point for a pre-rRNA processing complex. Early recognition of pre-rRNA transcripts may be especially important in T.brucei, where RNA polymerase I apparently transcribes both rRNA and mRNA genes (48). The experimentally indicated interactions between U3 snoRNAs and the A[prime]-containing site1 region are so far novel to T.brucei. Our previous RNA crosslinking analysis (19) combined with our present work suggests that U3 may functionally bind to at least one and possibly two sequence elements 3[prime]-adjacent to A[prime]. Deletion of the small site1a and site1b elements from pre-rRNAs expressed in T.brucei showed that these are required for processing events culminating in SSU rRNA accumulation, consistent with the expected phenotype for cellular U3 depletion. Preliminary data from our laboratory indicate that deletion of site1a sequences results in no A[prime] cleavage, suggesting that this 5[prime]-ETS cleavage is required for SSU rRNA processing. Substitution analysis of site1a sequences implied that U3 pairing here was more important for SSU rRNA production than the A[prime] site itself. The low level of SSU rRNA produced from transcripts containing substituted A[prime] sequences may have arisen due to essential, low efficiency cleavage at a cryptic site, yet clear analysis of this possibility awaits development of more sensitive assays. Our current model proposes that one or two U3 molecules interact, via distinct hinge region bases (see Fig. 5), at the two A[prime] proximal elements to affect A[prime] cleavage and/or downstream processing events.
If processing of the T.brucei A[prime] site and vertebrate primary cleavage sites are related, then U3 interactions should be detectable adjacent to vertebrate primary cleavage site sequences. In favor of this, base pairing motifs between U3 hinge and 5[prime]-ETS sequences have been noted for mouse, human and Xenopus (23). Although the locations of these pairing motifs with respect to related 5[prime]-ETS sequences are not conserved between systems, the human and mouse models do indicate potential interactions at 35-41 and 16-22 nt downstream from the cleavage site; the Xenopus laevis possible interaction is at 207-214 nt downstream. The possibility exists that interactions between U3 and at least mammalian 5[prime]-ETS sequences occur that are related to those indicated for the T.brucei A[prime] region. Evidence against this idea comes from analysis of processing complexes that assemble on synthetic 5[prime]-ETS substrates in vertebrate cell extracts (8,9). U3 was present in these complexes when large, but not minimally sized, 5[prime]-ETS substrates were used, attesting that U3 was not stably bound near the cleavage site. The mechanism by which U3 affects vertebrate 5[prime]-ETS cleavage and whether processing of 5[prime]-positioned 5[prime]-ETS cleavage sites is related between organisms remains to be established.
An apparent link between pre-rRNA processing in the evolutionarily distant trypanosomes and yeasts was provided by analysis of T.brucei 5[prime]-ETS site3 sequences, which crosslink to U3, and of pre-SSU rRNA intermediates. The functional interaction between yeast U3 hinge bases and 5[prime]-ETS sequences (24), required for cleavages leading to SSU rRNA production, and the conceivably comparable interaction between trypanosome RNAs are related in that 10 perfect base pairs of complementarity are shared between U3 hinge and 5[prime]-ETS sequences. Predictably, the T.brucei 5[prime]-ETS bases involved are required for SSU rRNA accumulation. In each organism, the 5[prime]-ETS U3 binding motifs are located at similar distances upstream from A0 sequences. The 3[prime]-ends of their 5[prime]-ETSs exhibit structural congruence: A0 and A1 cleavage sites could occur on either side of the base of a long stem-loop structure (35,36). Interestingly, yeast- and trypanosome-specific hexamer sequences are found 5[prime]-adjacent to the A1 cleavage; these elements conceivably have related roles in each group of organisms. In yeast, accurate processing of the A1 site requires the hexamer sequence and a 3 nt distance to the 5[prime] stem-loop pseudoknot structure conserved in all SSU rRNAs (37,38). Cleavage of both A0 and A1 in yeast is endonucleolytic (36) and preliminary data for T.brucei has provided a candidate for an endonucleolytically released A0-A1 fragment. A0 equivalents have not yet been identified for other organisms, but its broad conservation is suggested, as stem-loop structures 5[prime]-adjacent to the A1 site can be computer modeled for X.laevis and human 5[prime]-ETS sequences (data not shown). However, whereas processing of the 5[prime]-end of SSU rRNA in yeast occurs by simple endonuclease digestion, the 5[prime]-end of human 18S rRNA is formed, at least in vitro, by a different mechanism: endonucleolytic cleavage occurs 3 nt upstream of the mature end of 18S rRNA, followed by exonucleolytic digestion (49,50).
One difference between the putative trypanosome U3-site3 interaction and the yeast U3-5[prime]-ETS interaction is that distinct U3 hinge regions are utilized in each. Phylogenetic co-variation has indicated that the 5[prime]-most U3 hinge bases (yeast U3 nt 39-48) pair to 5[prime]-ETS sequences in distinct yeast species (24,51), whereas the 3[prime]-most (trypanosome U3 nt 54-63) are implicated for the apparently related interaction in T.brucei. This U3 sequence varies between trypanosome species (31), and U3 hinge sequences generally are not conserved between phylogenetically distinct organisms. This variability in U3 hinge bases may reflect the sequence diversity noted for 5[prime]-ETS sequences; U3 hinge bases may co-vary to pair with divergent 5[prime]-ETS sequences. Species-specific U3-5[prime]-ETS interactions could then occur that ultimately lead to the same end product: removal of 5[prime]-ETS sequences and production of SSU rRNAs.
The highly conserved U3 box A bases are expected to have important roles in rRNA processing. Crosslinking studies in a number of systems have suggested that U3 box A, or nearby bases, contact the 5[prime]-ETS (18-22). The T.brucei U3 box A-site2 5[prime]-ETS crosslink was readily detected, implying a stable contact between RNAs, yet deletion of site2 did not affect pre-rRNA processing. This suggests that either the noted contact was artifactual or that the U3 box A-5[prime]-ETS interaction was important, but a cryptic, or functionally redundant, interaction compensated for deletion of the normal binding site. It is also conceivable that U3 snoRNA scans the 5[prime]-ETS via its 5[prime]-most sequences; this scenerio potentially explains non-essential U3-5[prime]-ETS crosslinks, noted in T.brucei and other systems, that occur downstream from supposed U3 snoRNA loading sites (22,23). A tangible role for U3 box A bases in interactions with 18S rRNA 5[prime] sequences has recently been demonstrated in yeast; the interaction is required for A1 and A2 cleavages and formation of the SSU rRNA (27). The high conservation of the target SSU rRNA pseudoknot sequences correlates well with the high conservation of the U3 box A sequences (25).
In T.brucei, processing of 5[prime]-ETS sequences is a three-step process, versus the two-step process noted for yeast (3,52). Examination of stable pre-SSU rRNA processing intermediates indicates that A[prime] cleavage precedes that at A0, followed by A1 and then A2 cleavage. Similar U3 3[prime] hinge region bases may interact with the critical site1b and site3 5[prime]-ETS elements and 5[prime] hinge bases might contact site1a. The proximity of proposed U3 binding sites to the cleavage sites they may affect varies greatly. Site1 elements are 3[prime]-adjacent to A[prime] and site3 lies 162 nt upstream of A0. By analogy to the yeast system, the predicted trypanosome U3-site3 interaction may be required for A0 and subsequent A1 and A2 cleavages. This interaction is not necessary for A[prime] cleavage as the first 260 nt of the T.brucei pre-rRNA are properly cleaved in vivo (A.Chapman and N.Agabian, personal communication). The mode of action of U3 snoRNA at each site may be distinctive, though U3 may be affiliated with a processing complex at each location. A U3-associated complex assembled at the A[prime] site may translocate to the A0 region or individual complexes assembled at each site may communicate with one another to coordinate the ordered pre-SSU rRNA cleavages. We are currently developing a T.brucei U3 snoRNA expression system so that each putative U3-5[prime]-ETS interaction can be tested by co-variation and the roles of U3-5[prime]-ETS interactions in particular 5[prime]-ETS cleavages can be further investigated.
ACKNOWLEDGEMENTS
We thank D. Tollervey and anonymous reviewers for their helpful comments on this manuscript. This work was funded by NIH grant AI34093 to T.H.
REFERENCES
*To whom correspondence should be addressed. Tel: +1 518 262 6654; Fax: +1 518 262 5689; Email: toinette_hartshorne{at}ccgateway.amc.edu
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. Kent, Y. R. Lapik, and D. G. Pestov The 5' external transcribed spacer in mouse ribosomal RNA contains two cleavage sites RNA, January 1, 2009; 15(1): 14 - 20. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Barth, B. Shalem, A. Hury, I. D. Tkacz, X.-h. Liang, S. Uliel, I. Myslyuk, T. Doniger, M. Salmon-Divon, R. Unger, et al. Elucidating the Role of C/D snoRNA in rRNA Processing and Modification in Trypanosoma brucei Eukaryot. Cell, January 1, 2008; 7(1): 86 - 101. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Barth, A. Hury, X.-h. Liang, and S. Michaeli Elucidating the Role of H/ACA-like RNAs in trans-Splicing and rRNA Processing via RNA Interference Silencing of the Trypanosoma brucei CBF5 Pseudouridine Synthase J. Biol. Chem., October 14, 2005; 280(41): 34558 - 34568. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. C. Jensen, D. L. Brekken, A. C. Randall, C. T. Kifer, and M. Parsons Species Specificity in Ribosome Biogenesis: a Nonconserved Phosphoprotein Is Required for Formation of the Large Ribosomal Subunit in Trypanosoma brucei Eukaryot. Cell, January 1, 2005; 4(1): 30 - 35. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. V. BOROVJAGIN and S. A. GERBI Xenopus U3 snoRNA docks on pre-rRNA through a novel base-pairing interaction RNA, June 1, 2004; 10(6): 942 - 953. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. MARMIER-GOURRIER, A. CLERY, V. SENTY-SEGAULT, B. CHARPENTIER, F. SCHLOTTER, F. LECLERC, R. FOURNIER, and C. BRANLANT A structural, phylogenetic, and functional study of 15.5-kD/Snu13 protein binding on U3 small nucleolar RNA RNA, July 1, 2003; 9(7): 821 - 838. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. V. Borovjagin and S. A. Gerbi Xenopus U3 snoRNA GAC-Box A' and Box A Sequences Play Distinct Functional Roles in rRNA Processing Mol. Cell. Biol., September 15, 2001; 21(18): 6210 - 6221. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. J. Greenwood, M. N. Schnare, J. R. Cook, and M. W. Gray Analysis of intergenic spacer transcripts suggests 'read-around' transcription of the extrachromosomal circular rDNA in Euglena gracilis Nucleic Acids Res., May 15, 2001; 29(10): 2191 - 2198. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.A. GERBI, A.V. BOROVJAGIN, M. EZROKHI, and T.S. LANGE Ribosome Biogenesis: Role of Small Nucleolar RNA in Maturation of Eukaryotic rRNA Cold Spring Harb Symp Quant Biol, January 1, 2001; 66(0): 575 - 590. [Abstract] [PDF]
|
|
|
|
|
|
M. N. Schnare, J. C. Collings, D. F. Spencer, and M. W. Gray The 28S-18S rDNA intergenic spacer from Crithidia fasciculata: repeated sequences, length heterogeneity, putative processing sites and potential interactions between U3 small nucleolar RNA and the ribosomal RNA precursor Nucleic Acids Res., September 15, 2000; 28(18): 3452 - 3461. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Antal, A. Mougin, M. Kis, E. Boros, G. Steger, G. Jakab, F. Solymosy, and C. Branlant Molecular characterization at the RNA and gene levels of U3 snoRNA from a unicellular green alga, Chlamydomonas reinhardtii Nucleic Acids Res., August 1, 2000; 28(15): 2959 - 2968. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||