Functional significance of intermediate cleavages in the 3′ETS of the pre-rRNA from Schizosaccharomyces pombe

Abstract

Pathways for the maturation of ribosomal RNAs are complex with numerous intermediate cleavage sites that are not always conserved closely in the course of evolution. Both in eukaryotes and bacteria genetic analyses and in vitro studies have strongly implicated RNase III-like enzymes in the processing of rRNA precursors. In Schizosacharomyces pombe, for example, the RNase III-like Pac1 nuclease has been shown to cleave the free 3′ETS at two known intermediate sites but, in the presence of RAC protein, the same RNA also is cleaved at the 3′-end of the 25 S rRNA sequence. In this study normal and mutant 3′ETS sequences were digested with the Pac1 enzyme to further evaluate its role in rRNA processing. Accurate cleavage at the known intermediate processing sites was dependent on the integrity of the helical structure at these sites as well as a more distal upper stem region in the conserved extended hairpin structure of the 3′ETS. The cleavage of mutant 3′ETS sequences also generally correlated with the known effects of these mutations on rRNA production, in vivo. One mutant, however, was efficiently processed in vivo but was not a substrate for the Pac1 nuclease, in vitro. In contrast, in the presence of RAC protein, the same RNA remained susceptible to Pac1 nuclease cleavage at the 3′-end of the 25 rRNA sequence, indicating that the removal of the 3′ETS does not require cleavage at the intermediate sites. These results suggest that basic maturation pathways may be less complex than previously reported raising similar questions about other intermediate processing sites, which have been identified by analyses of termini, and/or processing, in vitro.

INTRODUCTION

The multiple RNA constituents of ribosomes are transcribed as large precursor molecules which are cleaved and modified to form the mature rRNAs. The complex maturation pathways often include numerous intermediate cleavages in the spacer regions. In Escherichia coli, for example, as a primary step in the release of the mature rRNAs, RNase III first makes staggered cuts in both strands of double-helices which form between the termini of the rRNA sequences (1,2). While some intermediate cleavages such as the RNase III cuts in E.coli have been shown to be essential, in many eukaryotes, minor termini also have been included as steps in the maturation pathways without direct experimental evidence.

Over the years, double-strand-specific ribonuclease activities also have been reported in several different eukaryotic cells (3–8). In both Saccharomyces cerevisiae and Schizosaccharomyces pombe, genes encoding such enzymes have been cloned and the proteins purified as Rnt1p and Pac1, respectively. Although mature rRNA termini are not produced, when free RNA is used as a substrate, both enzymes appear to cleave efficiently in their cognate 3′ external transcribed spacer (3′ETS), in vitro. In S.cerevisiae (6) only a single cut is observed 21 nucleotides downstream of the 25S rRNA although, in vivo, two sites have been mapped (+14 and +49) on opposite sides of a predicted stem–loop structure (7). In S.pombe (8,9) two staggered cuts are introduced at known alternate intermediate sites (+41 and +83) in a highly conserved hairpin structure. In S.cerevisiae, Rnt1p also has been reported to cleave the 5′ external transcribed spacer (5′ETS) at a U3 snoRNP-dependent site (10), although this has been questioned experimentally (7). In yeasts the RNase III-like enzymes has been implicated in the maturation of small nuclear or nucleolar RNAs (10–12), including several that have been linked to rRNA maturation. For example, in S.cerevisiae the snR190 and U14 snoRNAs, which are co-transcribed as a dicistronic precursor are separated by Rnt1p cleavage (13).

Recently, a series of studies on interdependences in the processing of eukaryotic ribosomal RNAs (see for example 14–16) have suggested that the formation of the 80–90 S nucleolar precursor particle and the subsequent cleavages of the nascent rRNA transcripts represent, at least in part, a quality control mechanism which helps insure that only functional rRNA is incorporated into ribosomes (17). In search of proteins which may mediate these interdependencies, a large protein complex, ribosome assembly chaperone (RAC) of 20 or more polypeptides has been isolated from S.pombe cells and shown to interact specifically with the spacer regions (see for example 18). Since several studies on the 3′ETS have indicated that the highly conserved hairpin structure interacts with nucleolar protein and this interaction can affect rRNA maturation (15,19), the effect of RAC protein on Pac1 cleavages also was examined (20). This study indicated that, in the presence of the RAC protein/3′ETS complex, cleavage by the RNase III-like homolog is not restricted to the known intermediate sites but also is directed at the 3′-end of the 25 S rRNA sequence.

To further examine the Pac1 RNase cleavage specificity with respect to the intermediate steps and their relationship to rRNA maturation, in this study we have looked at Pac1 nuclease cleavages in both natural and mutant rRNA sequences. The results indicate that these cleavages are not essential either to 25 S rRNA production, in vivo, or the RAC-induced, Pac1 nuclease cleavage at the 3′-end, in vitro.

MATERIALS AND METHODS

Preparation of RNA substrates

Normal and mutant RNA substrates for digestion analyses were prepared by in vitro transcription using T₇ RNA polymerase (21,22). The templates for transcription reactions were prepared by PCR amplification of normal or mutant subclones or the subcloned DNAs were used directly. For most 3′ETS templates a T₇ promoter/3′ETS-specific hybrid sequence (5′-CGGAATTCCGTTAATACGACTCACTATAGGAACCATCATCTTATTTC) was used as the forward primer and a 3′ETS-specific sequence (5′-CCCTTACCTCCTCTCCTC) was used as the reverse primer. In some cases the plasmids were transcribed directly using the plasmid T7 RNA promoter after digestion with EcoRI restriction endonuclease to permit run off transcription. For templates that included the 3′-end of the 25S rRNA, a DNA fragment beginning at G₃₄₄₃ in the 25S rRNA sequence and again ending at G₊₁₄₉ in the 3′ETS was cloned in the pTZR19 plasmid and used for RNA expression. Transcription reactions were performed with 0.4–2 mg of a DNA template in the presence of 2 mM ATP, CTP, GTP and UTP and 100 U of T₇ RNA polymerase using N4 buffer [20 mM MgCl₂, 40 mM Tris–HCl (pH 8.1), 1 mM spermidine, 0.01% Triton X-100] in a final volume of 100 µl. After incubation at 37°C for 4 h the RNA was precipitated with 2.5 vol of cold salted ethanol. The transcripts were then purified on a 6 or 8% denaturing polyacrylamide gel and labeled at the 5′-end using bacteriophage T4 polynucleotide kinase and [γ-³²P]ATP (23), after dephosphorylation with calf intestinal phosphatase (24). The phosphatase was heat inactivated for 10 min at 75°C in presence of 5 mM EDTA (pH 8.0) before the RNAs were labeled and the labeled RNAs were again purified on a 6 or 8% denaturing polyacrylamide gel. Alternatively, in some experiments the transcribed RNAs were labeled at the 3′-end using cytidine [3′, 5′-³²P] bisphosphate and bacteriophage T₄ RNA ligase as described by Peattie (25) and purified on a 6 or 8% denaturing polyacrylamide gel.

Pac1 nuclease digestion assay

Pac1 RNase was prepared and cleavage reactions were performed essentially as described by Rotondo and Frendewey (9). The in vitro synthesized and labeled RNAs were digested with appropriate amounts of the Pac1 RNase for 60 min at 30°C in 10 µl of buffer containing 5 mM MgCl₂, 1 mM DTT and 30 mM Tris–HCl, pH 8.1. Reactions were stopped with the addition of an equal volume of loading buffer (formamide containing 0.05% xylene cyanol and 0.05% bromo-phenol blue), the solution was heated for 2 min at 90°C, and 5 µl aliquots were applied directly to a 6 or 8% (19:1; acrylamide: bisacrylamide) polyacrylamide gel containing 8.3 M urea. After fractionation, the fragments were detected by autoradiography, the images were captured using a Gel Doc 1000 (Bio-Rad Laboratories, Richmond, CA) and the relative amount of the Pac1 digestion product was determined using Molecular Analyst software.

Preparation of RAC protein extract

The RAC protein complex was prepared essentially as described by Lalev et al. (18). Logarithmically growing cells (A_550nm = 0.6, 1 l) were harvested by centrifugation, washed with water, suspended in 10 ml ice-cold breaking buffer (0.4 M KCl, 1.5 mM MgCl₂, 0.5 mM DTT, 0.5 mM PMSF, 10 mM Tris–HCl, pH 7.9) and broken by vortex for 30 min with an equal volume of glass beads (30 s cycles alternating with 30 s on ice). The cell debris and glass beads were removed by centrifugation at 10 000 g in a Beckman (Fullerton, CA) JA20 rotor for 10 min at 4°C, the supernatant was cleared by ultracentrifugation at 100 000 g in a Beckman Ti 70 rotor for 1 h at 4°C, diluted with glycerol (15%, final concentration) and divided into 100 µl aliquots for storage at –85°C. The protein concentration was ∼10 µg/µl when standardized against bovine serum albumin.

S1 nuclease mapping of RNA termini

Termini of mature and precursor ribosomal RNA transcripts were detected by S1 nuclease mapping as previously described (14,19). For the 3′-end labeled DNA probes, the mutant rDNA was digested with NotI and HindIII restriction endonuclease to prepare a fragment extending from the 28S rRNA into the 3′ETS region. This was labeled by extending the protruding ends of the NotI restriction sites with [α-³²P]dCTP and Klenow enzyme. The labeled probe was incubated with whole cell extracted RNA (20 µg) at 30°C for 12 h in 50 µl of 3 M NaTCA, 5 mM, Na₂EDTA, 50 mM Pipes (pH 7.0), rapidly chilled on ice and digested for 30 min at 37°C with S1 nuclease (100 U) in 15 mM ZnCl₂, 250 mM NaCl, 40 mM NaOAc (pH 5.5: 200 µl, total volume). Digestions were terminated with SDS/EDTA, extracted with phenol/chloroform and analyzed on 6% polyacrylamide sequencing gels. Standard chain termination sequence reaction products also were applied to the analytical gels as fragment length markers.

RESULTS AND DISCUSSION

In past studies on the role of the S.pombe 3′ETS in rRNA processing, mutations were systematically introduced into a conserved, extended hairpin structure, immediately distal to the 3′-end of the 25S rRNA (14,19). The results demonstrated a range of effects on the maturation of both the 3′-end of the 25S rRNA as well as the 5.8S rRNA sequence, located more than 3000 nucleotides upstream of the 3′ETS. Most intriguing was the fact that the effects were equally dramatic with both RNAs, suggesting a common dependence on the 3′ETS structure. In the present study, the importance of the 3′ETS structure with respect to Pac1 cleavage and any relationship to this interdependency was examined further when RNAs containing the same mutations were cleaved in vitro. The normal and mutant RNA substrates, consisting of a complete extended hairpin with short stretches of adjacent sequence at each end (A₊₁₈-G₊₁₄₆), were prepared by transcription in vitro, using a PCR amplified template and T₇ RNA polymerase (19,21). As previously reported (9) and illustrated in Figure 1, over a wide range of concentrations (lanes a–c), the enzyme effectively cleaves the 5′-end labeled 3′ETS sequence producing a specific fragment that terminates at U₊₄₁. This cleavage site consistently has been observed in vivo (14). Additional, less specific cuts were observed, but only at high enzyme concentrations (lane c). In the assay of mutant substrates, a relatively large amount (100 ng) of unlabeled normal 3′ETS sequence was added to the labeled mutant RNA in each digest in order to maintain a constant RNA/enzyme ratio. Under these conditions ∼25% of the normal RNA is cleaved (lane d).

Figure 1.

Pac1 nuclease cleavage of a 5′-labeled 3′ETS substrate from the S.pombe pre-rRNA. A normal 3′ETS substrate was prepared in vitro by transcription with T₇ RNA polymerase, labeled at the 5′-end (Ctl) and aliquots (10 000 c.p.m.) were digested with 0.4 (a), 1.2 (b) or 3.6 ng (c and d) of Pac1 RNase in the absence (a–c) and presence (d) of 100 ng of unlabeled 3′ETS substrate. Digests were fractionated on a 10% denaturing polyacrylamide gel and the fragments were detected by autoradiography. Labeled substrate, which was partially digested with T1, ribonuclease (T₁) or mild base (B) was included as nucleotide markers. The positions of known intermediate processing sites (13) are indicated on the right.

As illustrated with the examples shown in Figure 2, the degree of cleavage varied dramatically with mutant RNA substrates depending on the change that had been introduced. For example, relative to normal RNA (Wt), a deletion (3′ETSΔ12) in the terminal loop of the extended hairpin structure (mutant c) had a relatively modest effect (∼84% of normal), but a larger deletion (3′ETSΔ22) which included a portion of the upper stem (mutant b), almost eliminated the Pac1-mediated cleavage (15% of normal). As might be anticipated, mutant RNAs in which the cleavage site itself was disrupted (e.g. 3′ETSG₃₉G₄₁G₄₂ and 3′ETSC₇₀A₇₂U₇₃C₈₁ C₈₂C₈₇) revealed little (24% of normal) or no cleavage (mutants a and g, respectively).

Figure 2.

Pac1 nuclease cleavage of mutant 5′-labeled 3′ETS substrates from S.pombe pre-rRNA. Normal or mutant 3′ETS substrates were prepared in vitro by transcription with T₇ RNA polymerase, labeled at the 5′-end and aliquots (10 000 c.p.m.) were digested with 3.6 ng of Pac1 RNase (8.0 nM) in the presence 100 ng of unlabeled normal 3′ETS substrate. Undigested normal 3′ETS substrate (Ctl) and digests of normal 3′ETS (Wt), 3′ETSG₃₉G₄₁G₄₂ (a), 3′ETSΔ22 (b), 3′ETSΔ12 (c), 3′ETSG₃₉G₄₁G₄₂ G₆₈C₇₂C₈₁C₈₂C₈₄ (e), 3′ETSG₃₉G₄₁G₄₂C₈₁C₈₂C₈₄ (f), 3′ETSC₇₀A₇₂U₇₃C₈₁ C₈₂C₈₄ (g), 3′ETSC₇₀A₇₂U₇₃ (h), 3′ETSA₂₇A₂₈C₃₉ (i) and 3′ETSG₈₆ G₈₇G₈₈G₈₉G₉₀ (j) were fractionated on a 10% denaturing polyacrylamide gel and the fragments were detected by autoradiography. Labeled substrate which was partially digested with T1 ribonuclease (T1) or mild base (B) was included as nucleotide markers. The position of the known intermediate processing sites at U₊₄₁ (13) is indicated on the right. The mutant lanes are labeled to correspond with the structural estimates presented in Figures 3 and 4.

To assess any relationships between mutant RNA cleavage, in vitro, and the effect of the same mutations, in vivo, the degree of cleavage with each mutant RNA was quantified and compared with the effect of the same mutation on rRNA production as previously observed (19) or determined in this study. In our summary of results for changes in the upper stem region (Fig. 3), rRNA production and the Pac1 cleavage efficiency usually were affected in a similar fashion, being influenced by two helical regions, the helix in which cleavage occurs as well as the apical helix, above it. In contrast, most changes to the lower third of the stem (Fig. 4) had only modest effects on both the efficiency of rRNA production, in vivo, and Pac1 cleavage, in vitro. Despite this general correlation, the severity of the effects did not always match. For example, three of the mutants (Fig. 3a, b and d), which essentially abolished rRNA synthesis in vivo, allowed some correct, although inefficient, Pac1 cleavage in vitro. Equally, with a deletion in the lower stem (Fig. 4l), Pac1 cleavage was only modestly affected (83% of normal), but the yield of plasmid-derived 5.8S rRNA was reduced by 70%. Furthermore, the effects with one of the mutants clearly were divergent. As shown in Figure 4 (mutant j), a change in the cleavage site region, which still maintained the actual cleaved residues in a helical structure, had no effect on rRNA production, but this mutant RNA was essentially uncleaved by the Pac1 enzyme (2% of normal). Of equal surprise was the fact that as long as the helical structure was maintained, the cleaved sequence could be dramatically altered (Fig. 3f) with little or no effect on RNA cleavage or rRNA production.

Figure 3.

Effect of the upper hairpin region on Pac1 nuclease cleavage of the 3′ETS in S.pombe pre-rRNA. Normal or mutant 3′ETS substrates were prepared in vitro, labeled at the 5′-end and digested with Pac1 RNase in the presence of 100 ng of unlabeled normal RNA; the digests were fractionated as described for Figures 1 and 2. The degree of digestion, indicated as an average for three experiments, was determined as a percentage of the digestion with the normal 3′ETS substrate. The changed residues are indicated by shading and deleted regions are indicated by the number of residues. The normal structure is shown on the left as previously reported (14). The amounts of plasmid-derived 25S rRNA, as determined by S1 mapping previously (19) or in the course of this study, are indicated as strong band (++) or none detectable (n.d.), respectively; the amount of plasmid-derived 5.8S rRNA as a percentage of that observed with a normal 3′ETS structure was taken from the same study or determined by electrophoretic fractionation.

Figure 4.

Effect of the lower hairpin region on Pac1 nuclease cleavage of the 3′ETS in S.pombe pre-rRNA. Normal or mutant 3′ETS substrates were prepared in vitro, labeled at the 5′-end and digested with Pac1 RNase in the presence of 100 ng of unlabeled normal RNA; the digests were fractionated as described for Figures 1 and 2. The degree of digestion, indicated as an average for three experiments, was determined as a percentage of the digestion with the normal 3′ETS substrate. The changed residues are indicated by shading and deleted regions are indicated by the number of residues. The normal structure is shown on the left as previously reported (14). The amounts of plasmid-derived 25S rRNA, as determined by S1 mapping previously (19) or in the course of this study, are indicated as strong band (++) or weak band (+), respectively; the amount of plasmid-derived 5.8S rRNA as a percentage of that observed with a normal 3′ETS structure was taken from the same study or determined by electrophoretic fractionation.

Although the mutants were not actually designed to evaluate the cleavage specificity of the Pac1 nuclease, the results of the Pac1 digestion studies were consistent with past studies on the specificity of the E.coli RNase III. In general, the experiments indicated that the region around the cleavage sites must be maintained in a double helical structure for accurate Pac1 cleavage. Disruption of the helix at the processing sites greatly reduced or abolished Pac1 cutting (Fig. 3a and g), while restoration of the helix by compensating mutations (Fig. 3f) restored accurate and efficient cleavage. Pac1, like RNase III, can make a single strand cut in a helix containing non-Watson–Crick base pairs (9). The difference in severity of the two helix-breaking mutations in the 3′ETS on Pac1 cleavage (compare Fig. 3a with g) probably reflect the unequal potential to accommodate G-A versus U-C pairs in the helix. Disruption of the helices above the +41/+83 processing sites also severely reduced Pac1 cleavage in vitro (Fig. 3b, d and e). In contrast, alterations that were more distal to the cleavage sites, such as deletion of most of the apical loop (Fig. 3c) or mutation in the lower half of the hairpin (Fig. 4i and k–m), had very little effect on cleavage efficiency. Although subtle primary sequence antideterminants or ‘disfavored’ sequences have been shown to have a negative effect on E.coli RNase III recognition (26), secondary structure is considered the most important determinant of substrate recognition and cleavage specificity (27). The effects of 3′ETS mutations on Pac1 cleavage support this conclusion. In addition, the results are consistent with a model in which a homodimeric enzyme requires approximately one turn of helix above and below the cleavage site for productive substrate binding (28).

While a correlation between the effects, in vivo and in vitro, would not necessarily be anticipated, a closer examination of factors that affect rRNA processing, in vivo, appears to provide a reasonable explanation for the correlations that actually were observed. At least three factors, the structure surrounding the cleavage sites, additional sequence or structure recognized by the Pac1 nuclease and the binding and influence of RAC protein, independently or in combination, should be considered when the effects are compared. As noted above, the in vitro analyses show that, in addition to the helix at the processing site, disruption above the +41/+83 sites could severely reduce cleavage, in vitro, consistent with further structure recognition by the Pac1 enzyme. This area borders and probably overlaps the RAC protein-binding site, which has been localized in the upper region of the stem (19). As a result, mutations in the upper region of the stem could have effects in vitro and in vivo for entirely different reasons or multiple factors may have cumulative effects, in vivo. The differences in severity, which were observed in this study, are likely reflecting this. For example, when mutant e is compared with mutant f (see Fig. 3), the additional changes in the upper stem have about a 2-fold effect on Pac1 nuclease digestion in vitro (36 versus 87%), as a result of a reduced recognition by the Pac1 protein, but essentially eliminated rRNA production likely due to the reduced RAC protein affinity (19). Whatever the explanation for the differences in severity, the observation that mutant j had no effect on pre-rRNA production, in vivo, but prevented Pac1 RNase cleavage in vitro, suggested clearly that the intermediate cleavages were dispensable. As shown in Figure 4, in this mutant five adenosines (A_86–90) in one strand of the helix just beneath the +G ₈₃ Pac1 cleavage site were replaced with five guanosines. This change could be thought of as replacing five A-U base pairs with five less stable G-U pairs; however, the most stable computer generated estimate for this mutant RNA that is shown in Figure 4, predicts a bulged disruption just below the cleavage site, which would explain the poor Pac1 cleavage efficiency with this substrate. On the other hand, without a change in the RAC protein-binding site, the RAC protein-induced maturation (20) probably would not be eliminated and would remain normal if the intermediate cleavages were not required. The possibility that in the cell, auxiliary factors such as rRNA binding proteins or snoRNPs assist Pac1 RNase cleavage of the mutant 3′ETS by promoting and stabilizing a structure that resembles the wild type RNA cannot be eliminated entirely, but it appears more likely that the intermediate cleavage is not critical to rRNA maturation.

In S.pombe, the PAC gene is essential for cell viability (5) and past studies in both S.cerevisiae and S.pombe have indicated that depletion of the RNase III-like enzyme, results in the accumulation of unprocessed rRNA precursors (6,11). Generally these observations have been taken as evidence that the intermediate cuts in the 3′ETS are important steps in the maturation of the ribosomal RNA precursors. As noted above, however, the present observations with respect to mutant j are not consistent. Since recent studies in the presence of RAC protein have shown that the Pac1 nuclease can completely remove the 3′ETS (20), it seemed reasonable to speculate that the intermediate cuts are not necessary for rRNA maturation, in vivo, because the RAC protein-induced cleavage actually represents the basic processing step. Indeed, S1 nuclease mapping of termini in the 3′ETS region of RNA from S.pombe cells expressing mutant j does indicate a normal 3′end for the 25S rRNA and the +21 intermediate (14), but no evidence of either the +41 or +83 Pac1 nuclease cleavages (Fig. 5, right panel). To examine this possibility further, the same mutant RNA was treated with Pac1 nuclease in the presence of RAC protein. As also shown in Figure 5, the results were entirely consistent with this suggestion. With normal RNA and RAC protein both the intermediate cut at U + 41 and the second cut at the 3′-end are clearly present. In contrast, with the mutant RNA the cleavage at U + 41 again is clearly absent but the cleavage at the 3′-end remains unaffected. Taken together, the results both in vivo and in vitro strongly suggest that 3′ETS maturation does occur without cleavage at the intermediate steps.

Figure 5.

RAC protein-induced Pac1 nuclease cleavage at the 3′-end of the 25S rRNA sequence in S.pombe pre-rRNA. Normal (left panel) or mutant (3′ETSG₈₆G₈₇G₈₈G₈₉G₉₀) RNA substrate (middle panel) consisting of the last 55 nucleotides in the mature 25S rRNA and the conserved extended hairpin structure in the S.pombe 3′ETS was prepared in vitro by transcription with T₇ RNA polymerase, labeled at the 5′-end (3′ETS) and aliquots (10 000 c.p.m.) were digested with Pac1 RNase in the absence (Pac) and presence (Pac + RAC) of ∼250 ng of RAC protein extract. Digests were fractionated on a 10% denaturing polyacrylamide gel and the fragments were detected by autoradiography. Labeled substrate that was partially digested with T1 ribonuclease (T1) or mild base (OH^–) was included as nucleotide markers. The position of the known PAC1 intermediate cleavage site as described in Figure 1 and the mature 3′-end of the 25S rRNA are indicated on the right. Termini in the 3′ETS region of RNA from S.pombe cells expressing the same mutant RNA also were mapped with S1 nuclease (right panel) as described previously described (14,19) and fractionated on a 6% denaturing polyacrylamide gel. The positions of the known termini, based on standard dideoxy sequencing reactions as residue markers, are indicated on the right.

In another recent study by Raue (29) on the pathway for ITS1 processing in S.cerevisiae, cleavages at two normal intermediate steps, A2 and A3, were virtually undetectable with deletions in the S1 domain of Rrp5p, a protein factor that has been linked with intermediate cuts in the ITS1 region. Instead still another site, A4, was suggested as a substitute site (for A3 cleavage) in an alternate pathway. As observed in the present study with the 3′ETS mutants, the growth rate with the RRP5 deletion mutations also was essentially unaffected.

All of these observations also raise an interesting question about the numerous termini in various organisms that have been linked into complex eukaryotic maturation pathways. Many of the termini such as the U + 41 and U + 83 sites in the S.pombe 3′ETS are present in only small amounts. In addition, there appears to be little conservation of the actual positions for some of these intermediate cleavages and the number of sites may differ significantly from organism to organism. For example, when the pathway in S.pombe is compared with that of S.cerevisiae the positions of the intermediate cleavage sites in the 3′ETS are different (6–8,14). This is equally true for the other nuclease activities and spacer regions including the 5′ETS (16,30,31) and ITS1 regions (29,30). Although generally the assumption has been that the relatively low amounts of many intermediate precursors reflect the rapid processing of the nascent rRNA precursor, the present data place emphasis on alternate possibilities. Since the Pac1 nuclease cleavages, in vitro, appear to be relatively efficient with naked RNA, the minor termini in some of the proposed pathways could actually represent fortuitous cuts rather than real steps in an obligatory sequence. They may also represent alternate pathways that provide flexibility or fail-safe mechanisms. Indeed, at least with respect to the ITS1 region, alternate pathways have been recognized (29,31). Whatever the case, the present results underline the possibility that the basic maturation pathway for the eukaryotic rRNAs may be less complex than generally believed and should be reexamined in this light.