RNase YI* was purified from Saccaromyces cerevisiae , strain Py26 ( Mata , ura3-52 , trp1 [Delta], leu2-3 , 112 , prb1-1122 , prc1-407 , pep4-3 , [Delta] nuc1::LEU2 ). Four liter cultures were grown at 30^oC in 1% yeast extract, 2% peptone, 2% glucose to give 20 g wet weight of cells, harvested and extracts prepared in a BioSpec bead beater, as described ( 27 ). Subsequent steps were at 0 to 2^oC. After removal of most nucleic acids by Polymin P precipitation, the supernatant was brought to 45% ammonium sulfate ( 27 ) and the precipitate, containing about half the total protein, was discarded after centrifugation. The supernatant was brought to 90% ammonium sulfate; the resulting precipitate (~1 g) could be stored at -70^oC. The following fractionations used a Pharmacia/LKB HPLC instrument and Pharmacia resins in columns jacketed in ice water with A ₂₂₆ recorded outflow. These steps can be modified, but fortunately any purification can take advantage of the fact that RNase YI* can be bound to both strong cation and anion exchangers with extremely strong binding to the former. Unless noted, RNase YI* was assayed in 10 mM 2-[ N -morpholino]ethanesulfonic acid (MES), pH 6.4 (see below). The ammonium sulfate pellet was resuspended in <5 ml of 40 mM Tris-HCl, pH 8.0. If necessary, residual precipitate was removed by centrifugation (here and in subsequent steps) without loss of activity. The clear supernatant was applied to a HiLoad 16/90 Superdex 75 sizing column (V _t = 122 ml, V ₀ = 44 ml) with elution in 20 mM Tris-HCl, pH 7.4 + 200 mM NaCl at 0.5 ml/min. Fractions of 0.7 ml were collected. RNase YI* eluted after a volume (~55 ml) corresponding to M _r ~70 kDa with a purification of ~15-fold. The active fractions (in ~6 ml) were pooled and diluted 4-fold with 20 mM MES, pH 5.7, then loaded, using a 50 ml Superloop immersed in ice, onto a 22 ml S-Sepharose column. After collecting the flow through (~70% of the A ₂₂₆ ) in 15 mM MES, pH 5.7, 2 ml fractions were collected at 0^oC at 1 ml/min with a 180 ml linear gradient from 0 to 2 M NaCl in 15 mM MES, pH 5.7. RNase YI* eluted in a peak at ~0.7 M NaCl with a 20-fold purification to yield a total purification of ~600-fold with ~60% recovery. The peak fractions were sufficiently pure to give the specificities described in this paper with a yield of ~4000 U (defined below) (~3 U/[mu]g protein). It lost activity with a half-life of ~8 months in 50% glycerol at -20^oC. Further purification was achieved by dialyzing the active fractions (~8 ml) against 20 mM Tris-HCl, pH 7.4 + 20 mM NaCl and then applying it to a 1 ml monoQ column immersed in ice. The enzyme eluted in 1.5 ml at 0.18 M NaCl in a 40 ml linear gradient from 0 to 1 M NaCl in 20 mM Tris-HCl, pH 7.4 to give another 15-fold purification (~50 U/[mu]g). After >7000-fold total purification, this preparation was probably ~30% RNase YI* based on the A ₂₂₆ of the activity peak from the monoQ column relative to the intensity of a Coomassie stained band that migrated at ~75 kDa, after SDS-PAGE, but it has not been proven to be the RNase YI* band. If it is, its size agrees with the size estimated by gel filtration and suggests that RNase YI* is a monomeric protein that is large relative to many other endoribonucleases. Almost all the other protein in the SDS gel was in a single band at ~65 kDa. The latter protein eluted from the monoQ at 0.28 M NaCl and accounted for ~40% of the A ₂₂₆ applied to the column. The enzyme was dialyzed against 30 mM Tris-HCl, pH 7.4, 40 mM NaCl, 50% glycerol (~4-fold reduction in volume) and stored at -20^oC after adding bovine serum albumin to 100 [mu]g/ml.

The final preparation was free of detectable DNase, phosphatase or RNA exonuclease activities. Single-strand DNase activity was assayed by incubating 10 [mu]g oligo(dC) ₃₃ with 1 U enzyme in a 20 [mu]l reaction containing 20 mM Tris-HCl, pH 7.4 and 10 mM MgCl ₂ at 37^oC for 2 h before loading on a mono(Q) column and eluting in a gradient from 0 to 2 M NaCl in 20 mM Tris-HCl, pH 7.4 ( 21 , 23 ). All A ₂₅₄ was centered in a sharp peak that eluted at ~0.6 M NaCl at which intact oligo(dC) ₃₃ elutes. Double-stranded DNase activity assay and reaction conditions were the same using lambda Hin dIII DNA fragments as substrates. Phosphatase activity was assayed by incubating 1 U enzyme with [[gamma]- ³² P]ATP in 20 mM Tris-HCl, pH 7.4 at 37^oC for 2 h and separating any released ³² P _i from [[gamma]- ³² P]ATP by ascending chromatography at 22^oC on a cel 300 PEI sheet (Machery-Nagel) with a solvent of 2 M sodium formate, pH 3.5 ( 25 ). There was no evidence of a processive RNA exonuclease activity based on the size distributions of RNA homopolymer products, separated on a mono(Q) column ( 29 ).

RNase I was purified from E.coli , strain DK552, containing a plasmid with the rna gene( 22 , 23 ), and was assayed in 20 mM Tris-HCl, pH 7.4. One unit of RNase I or RNase YI* degrades 1 [mu]g single-stranded homopolymer RNA (~300 nt) to <6mers per min at 37^oC in a 20 [mu]l reaction. (This definition would be equivalent to ~2 U based on a definition of conversion to products soluble in 5% trichloroacetic acid). Reactions with S1 nuclease or mung bean nuclease (Boehringer Mannheim) were in 30 mM sodium acetate buffer, pH 5.0, with 1 mM ZnSO ₄ .

Assay for dinucleotide degradation

Two kinds of assays were used. The first used 5'- ³² P-XpY-OH-3' as substrate. RNase YI* generates 5'- ³² P-Xp plus 5'-OH-Y-OH reaction products which were separated on cel 300 PEI sheets (Machery-Nagel) by ascending chromatography at 22^oC with a solvent of 2 M sodium formate, pH 3.5. The sheets had been pre-washed in 2 M formic acid brought to pH 2.2 with pyridine followed by a final rinsing with water ( 30 ). The ³² P-X-OH migrated slower than did the original 5'- ³² P-dimer. The PEI sheets were dried and the positions of labeled products identified by autoradiography, then cut out and counted.

The second assay used unlabeled substrate (5'-OH-XpY-OH-3'). Products of the reaction are 5'-OH-Xp and 5'-OH-Y-OH and were fractionated in a salt gradient on a monoQ column. The 5'-OH-Xp mononucleotide eluted at ~0.14 M NaCl and the dialcohol dimer with only an internal phosphate at 40 mM NaCl. The 5'-OH-Y-3'-OH nucleoside product was not bound.

Assay for release of 5 ' -nucleotide from a random mix of RNA oligonucleotides

Total yeast RNA was partially digested with 0.5 N NaOH for times varying from 0.5 to 15 min at 37^oC to produce a uniform distribution of sizes with 5'-OH ends and then 5'- ³² P labeled with [[gamma]- ³² P]ATP in the polynucleotide kinase reaction. The ³² P-oligonucleotides were separated by 20% PAGE and an oligonucleotide band of ~25 nt length was eluted to use as substrate. To obtain data for kinetic analysis each reaction contained the same amount of ³² P plus a known amount of unlabeled yeast RNA carrier which defined the relative concentration of RNA. Reactions were in 20 [mu]l with 14 U RNase YI* for 6 min at 23^oC and stopped by adding 1 M citric acid, pH 3.5, to 25 mM and ZnCl ₂ to 1 mM. Rates were linear with all concentrations during this time. The mononucleotides were separated by PAGE in a pH 3.5, 10% citrate gel, cut out and counted. AMP and CMP did not separate sufficiently and were counted together.

Structural analysis of E.coli 5S rRNA

Escherichia coli 5S rRNA (Boehringer Mannheim) was treated with heat labile alkaline phosphatase before phenol extraction and ethanol precipitation and then 5' end-labeled with [[gamma]- ³² P]ATP. The RNA was brought to 100^oC for 2 min to dissociate any molecules with `nicks' before loading on a 20% polyacrylamide gel. After electrophoresis, the gel band containing the full-length molecule, was excised, crushed and the RNA eluted by overnight incubation in a small volume of water. The acrylamide particles were spun out and the supernatant precipitated with ethanol and the RNA resuspended in water.

RNase YI* reactions were in 20 [mu]l of 10 mM MES, pH 6.4, at 23^oC for 30 min with 1 to 2 [mu]g ³² P-RNA and stopped by addition of 1 M citric acid to give 30 mM, pH 3.5. After adding 10 [mu]g poly(A), they were brought to 100^oC for 1 min and then chilled to 0^oC before adding 8 [mu]l of 10 M urea plus 0.2 [mu]l of 0.1 M ZnSO ₄ . The tracking dyes, bromophenol blue and xylene cyanol and glycerol (to 10%) were added and the samples loaded and run by 20% PAGE at 23^oC.

Mismatches in duplex nucleic acids

Plasmid pSP73 (Promega) was linearized by Bam HI digestion and used as a template to synthesize a runoff 58 nt RNA from the T7 promoter in vitro in the T7 RNA polymerase (Boehringer Mannheim) reaction in 50 [mu]l containing 40 mM Tris-HCl, pH 8.0, 6 mM MgCl ₂ , 10 mM dithiothreitol, 2 mM spermidine, 2 mM of each of the four nucleoside triphosphates, 1.25 [mu]g linearized plasmid and 50 U enzyme for 60 min at 37^oC. A separate reaction in 20 [mu]l contained the same concentrations of reactants plus 20 [mu]Ci [[alpha]- ³² P]GTP to label the RNA as a marker. The labeled and unlabeled samples were fractionated by 20% PAGE. The 58 nt RNA was identified in the lanes with ³² P and the unlabeled bands at the same position in adjacent lanes were eluted, as described in the preceding section. The T7 RNA was then treated at pH 7.5 with heat sensitive calf alkaline phosphatase (Boehringer Mannheim) for 5 min at 37^oC. The alkaline phosphatase was inactivated by heating at 100^oC for two min. The RNA was 5' labeled with 50 [mu]Ci [[gamma]- ³² P]ATP in the polynucleotide kinase reaction. In order to eliminate any possible contaminating ³² P fragments, the ³² P-RNA was purified by another 20% PAGE and prepared by the preceding steps, except that 10 [mu]g carrier poly(A) was added before ethanol precipitating, and the final sample (~1 [mu]g with ~10 ⁶ c.p.m.) was used to form an RNA-DNA hybrid with a specific 39mer DNA oligonucleotide that contained one or more base mismatches to the first 39 nt at the 5' end of the RNA. The DNA oligonucleotides had been synthesized in the Medical School's Protein Chemistry Laboratory. The RNA-DNA hybrids were purified by electrophoresis through a 20% native polyacrylamide gel. Reactions were at 23^oC for 60 min (37^oC for 12 min gave the same results). RNase I reactions were stopped with SDS to 0.1% and RNase YI* reactions with final concentrations of 3 M urea, 1 mM ZnCl ₂ and 30 mM citric acid, pH 3.5. In both cases, the stop mixes included glycerol and dyes plus a 10-fold excess of DNA oligonucleotide that had the same sequence as the 39 nt of ³² P-RNA. Subsequently, each mix was brought to 100^oC for 90 s to dissociate the RNA-DNA hybrid and then to 37^oC for 5 min to convert all the released DNA to a DNA-DNA duplex. The ³² P-RNA ran as single strands in the subsequent PAGE containing 7 M urea plus 2.5 mM EDTA.

RESULTS

Effectors of RNase YI* activity

pH optimum. The reaction rate of RNase YI* was dependent on pH over a fairly narrow range from 6.0 to 8.0 with a peak of activity at pH 6.2 (Fig. 1 ). However, since the activity decreased precipitously at pH <6.0, we have used MES, pH 6.4 buffer in most assays. The marked decline in reaction rate at pH 6.0 did not result from inactivation but rather from a slower reaction. The reaction rate was the same in 10 and 50 mM MES and with phosphate or Tris buffers in the pH ranges that overlapped.

Figure 1 . The pH dependence of RNase YI* in 50 mM MES, 20 mM Tris-HCl or phosphate buffers and heat inactivation of RNase YI*. Activity was estimated from the extent of degradation after fractionation on a monoQ column. ( A ) The maximum activity is at pH ~6.2 and falls off rapidly at lower pH. ([squ]), MES; (u), phosphate; ([squ]), (Tris-HCl). The reactions contained 20 [mu]g poly(C) and 10 U enzyme in 100 [mu]l at 23^oC. ( B ) Ten units of RNaseYI* in 100 [mu]l of 10 mM MES, pH 6.4, was incubated for the time and at the temperatures shown. Samples were brought to 0^oC before adding 20 [mu]g poly(C) and incubating for an additional 10 min at 37^oC.

Inhibitors of RNase YI*. The activity was inhibited by various divalent cations, such as 10 mM Mg ²⁺ , 5 mM Mn ²⁺ (>50%) or 0.1 mM Zn ⁺² (>95%). Conversely, it was fully active with 1 mM EDTA, although 10 mM EDTA inhibited 30%. It retained almost full activity in the presence of 200 mM NaCl or KCl salts. Heat inactivation. RNase YI* is very heat labile. The half life at 37^oC in the absence of substrate was ~8 min and at 60^oC ~24 s (Fig. 1 ) but might be longer in the presence of substrate, as is the case with RNase II ( 29 ), and it was much more stable to heat in 50% glycerol. This sensitivity to heat could obviate the need for phenol extractions to eliminate enzyme activity.

Substrate specificities of RNase YI*

Homopolymers. RNA homopolymers of A, C or U were degraded by RNase YI* at the same rates ( 21 ). Dimers. The reaction velocities of many broad-specificity endo- or exo-ribonucleases are slower with smaller substrates and may not even react with substrates below some minimal size, e.g. the processive exoRNase II reaction slows markedly when the RNA substrate is less than ~12 nt and the limit digest yields a dimer ( 29 ). The 16 possible RNA dimers were tested as substrates for RNase YI* by two different assays which also differed by the presence of 5'-P in one and not the other (Fig. 2 ). In both cases, RNase YI* was inactive with all dimers except those that had a 5' G residue (5'-G-X) with no obvious preferences among the four 3'-nucleotides as shown by Lineweaver-Burk plots.

Figure 2 . The cleavage of dinucleotides by RNase YI*. ( A ) is dimer A-A and ( B ) is A-A after RNase YI* reaction, ( C ) is dimer G-A and ( D ) is G-A after RNase YI* reaction. Four [mu]g dimer was reacted with 15 U RNase YI* in 20 [mu]l of 10 mM MES, pH 6.4, for 20 min at 20^oC. Separation was on a monoQ column (Pharmacia) in 20 mM Tris-HCl, pH 7.3, with a NaCl gradient from 0 to 2 M. ( E ) Double reciprocal plots of substrate ( S ) versus reaction velocity ( V ) for degradation of dimers of 5'- ³² P-G-X. The assay used chromatography of ³² P products on PEI sheets (Materials and Methods). 5'- ³² P-G-G gave poorly resolved products, but G-G was degraded at approximately the same rate as the other G dimers in the monoQ fractionation assay.

5 ' -End bond preferences in total cell RNA. The observed bond preferences seen with dimers could carry over to give similar preferences for the 5'-end bonds of longer oligonucleotides. A random mix of oligonucleotides of specific size (~25 nt) were 5'- ³² P labeled. Equivalent amounts of ³² P were incorporated into each of the four 5'-nucleotides as shown by a complete digest with RNase I, or alkali. 5'- ³² P-A and 5'- ³² P-C were not resolved, but the sum of their counts was about twice that of 5'- ³² P-G or 5'- ³² P-U. However, at low substrate concentrations RNase YI* generated 5'- ³² P-U four to five times slower than each of the other 5'-nucleotides. The reaction rates for their appearance became closer as the substrate concentrations increased to give the same V _max values and a higher K _m value for cleavage of 5'-U (Fig. 3 ). These results show that 5'-U residues are cleaved more slowly by RNase YI* than are 5'-A, 5'-C or 5'-G. However, the striking preference for dimers starting with 5'-G was not seen when the 3' nt was followed by a length of nucleotides.

Figure 3 . Double reciprocal plots of substrate ( S ) versus reaction velocity ( V ) for release of the 5' nucleotide from a random mixture of RNA oligonucleotides from total yeast RNA in a RNase YI* reaction. The 5'- ³² P-oligonucleotides were added to carrier yeast RNA that was used to define the amount of RNA in 20 [mu]l reactions from 0.1 to 10 [mu]g with the same cts/min in each reaction. The kinetic parameters are relative since molarity of the yeast RNA is not defined.

Nucleic acid inhibitors of RNase YI*

Single- or double-stranded DNA were potent inhibitors. As little as 0.4 [mu]g oligo dC ₂₅ (s.s.) inhibited >90% the degradation of 20 [mu]g poly(C) and 1 [mu]g [lambda] Hin d III DNA fragments (d.s.) also inhibited RNase YI* to the same degree. Highly structured tRNA was degraded very slowly by RNase YI* (or RNase I*) ( 21 ). This specificity could reflect a low affinity to structured RNA, or conversely, the enzyme might bind well but be unable to hydrolyze bonds. In the latter case, structured RNA could inhibit degradation of homopolymer substrates. In experiments with limiting RNase YI*, tRNA was a potent inhibitor; 0.8 [mu]g inhibited by >90% in the same poly(C) assay. However, very high concentrations of poly(C) (300 [mu]g in a 100 [mu]l reaction) were degraded at the same rate with or without 0.8 [mu]g tRNA. The results suggest that tRNA is a competitive inhibitor that binds to the enzyme without being converted to product. It is likely that rapid release of the enzyme requires a hydrolytic reaction. Since duplex RNA or single- or double-stranded DNA are not substrates for cleavage, enzyme remains associated much longer to make them competitive inhibitors at concentrations much lower than that of the substrate.

RNase YI* as a probe of RNA structure

The 120 nt 5S rRNA of E.coli has been a model for numerous structure studies since its sequence was first determined ( 30 , 31 ). The generally accepted structure, based on estimated thermodynamic stability and evolutionary conservation of specific segments ( 32 ), is shown in Figure 4 . The following observations suggest that RNase YI* reactions predict this structure.

Figure 4 . Structure of 5S rRNA based on energy considerations and the conservation of secondary structures in a broad range of species. The nucleotides are numbered starting at the 5' end. G to U bonds are designated by a ([squf] ) . The cleavage sites are determined from Figure 5. Those from the 5'- ³² P end are designated by pointed arrows; closed arrows show ends that persisted at the highest reaction conditions, and open arrows for bands that declined with increasing enzyme activities. Cleavage sites from 3'- ³² P ends are shown by the ball arrows with dashed tails showing bands that declined with increasing activities. The arc in the large loop region designates ³² P bands that have not been resolved but must include cuts at all bonds. The darker arrows after G-9, G-10 and U-111 designate the most prominent bands which were also the most persistent with increasing activities. The calculated total net free energy is -35.5 kcal/mol (kindly computed by Michael Zuker).

Cleavages by RNase YI* from the 5 ' end. (The numbers refer to nucleotides from the 5' end.) A very intense pair of bands at G-9 and G-10 appeared at the lowest enzyme activities and not only persisted but increased at higher activities (Fig. 5 A). These bands must result from cleavages at the end of `stem I'. NMR analysis of isolated stem I has indicated that G-9, paired with U-111, is not only stacked with G-10 but also with G-112 on the opposite strand rather than to its adjacent C-8 ( 33 ). This could distort the phosphodiester bonds of G-9 and U-111 as well as G-10 and C-110 to account for lability at G-9.

Figure 5 . ³² P-labeled molecules resulting from reactions of 5S-rRNA over a range of RNase YI* activities after fractionation by 20% PAGE. 5'- ³² P-label in (A) and 3'- ³² P in (B). From the uniform ladder of sizes from alkali digestions [lanes 1-3 in (A) and 1 and 2 in (B)], the log( M _d ) was plotted versus R _f (distance of bands from well) and used to estimate nucleotide positions of bands. Except for lanes 13-15, the films were highly exposed to detect minor bands. The numbers in the right columns give the nt position from the 5' end. Note that the 3' end was labeled in the RNA ligase reaction which added a C residue (121) to that end. ( A ) Lane 4, no enzyme; lanes 5-12, RNase YI*, 0.07, 0.14, 0.36, 0.72, 1.4, 3.6, 7.2 and 14 U respectively. Lanes 13-15 are the same gel lanes as 10-12 whose film was exposed a much shorter time to show the only visible bands at the 9-10 junction and the large loop at 35-47. ( B ) Lane 3, no enzyme; lanes 4-10 have the same units of RNase YI* as lanes 6-12 in panel (A).

There were no bands seen between C-3 and G-9 even at the highest RNase YI* activity indicating that stem I is extremely stable except for `end-nibbling' to give some dimers and trimers. The dimer 5'-UG-3' accumulated at the highest enzyme activity which is consistent with the preceding observation that 5'-UG-3' is not a substrate. In contrast, there were a series of strong bands from G-9 to A-15 that increased with activity and then declined at the highest activities. These bands resulted from molecules that were cleaved before any cleavages more proximal to the 5' end and were then lost as subsequent cuts trimmed the molecule to the `stop' at G-9 and G-10. The next pairs of persistent bands were at G-23 and G-24 and at C-26 and C-27.

The most prominent larger molecules appeared as a smear of bands that corresponded to cleavages in the large single-stranded region from A-34 to U-48. They appeared at low enzyme activity and only declined at the highest activity when cleavages occurred in the same molecules more proximal to the 5'- ³² P end. It was not determined if all the bonds in this region were hydrolyzed with equal rates. However, bands resulting from cleavages near the ends of the loop, i.e. near A-34 and U-48, were stronger which could reflect a rapid loss of single-strand `tails' from internal loop cuts. Cleavages from the 3 ' end. When the 5S RNA was labeled at the 3' end (Fig. 5 B), the most prominent bands were in the single-stranded region from G-112 to G-107. In this case the smallest product seen at the highest activity was the trimer, 5'-A-U- ³² P-C-OH. Since no evidence of dimer or monomer was visible, this oligo might not be degradable by RNase YI*. A single declining band appeared at G-102 which is consistent with the single strand that terminates at the stem starting at that base. Finally, there was a very strong band just below the full-length size. It was too high in the gel to calculate an exact nucleotide site, but it results from cleavages in the large loop at A-34 to U-48.

An obvious difference between the 3' end versus 5' end labeling was that many fewer cleavages were observed on the 3' half of the molecule. This difference could result from tertiary interactions between that side of the molecule (U-48 to C-97) that make it less accessible to RNase YI* activity. Probing 5S rRNA structure with S1 and mung bean nucleases. The results with S1 and mung bean nucleases were in agreement with each other but were markedly different from the results with RNase YI* (Fig. 6 ). Briefly, they showed only one major cleavage region at all activities used: the large loop at C-34 to U-48. There was a secondary 3'- ³² P-band mapping at G-96 which would result from cleavage between two consecutive G-U bonds. Both primary and secondary bands declined with increasing enzyme activities, but there were no major stops even at G-9 or G-112. Apparently, these enzymes are specific for extended single-stranded regions but are not sensitive to short disruptions of a duplex with the exception of the cleavage at G-96.

The detection of base pair mismatches by RNases

The apparently successful use of RNase YI* as a probe of RNA secondary structure suggested that it has a strong preference for bonds of unpaired nucleotides. This conclusion was tested with RNA-DNA duplexes with known base pair mismatches. A defined sequence of RNA was synthesized in vitro from the T7 promoter of plasmid pSP73 and hybrid RNA-DNA complexes formed with DNA oligonucleotides containing specific base mismatches. The DNA-RNA hybrids with ³² P at the 5' end of the RNA were reacted with a range of enzyme activities under fairly relaxed hybrid conditions (37^oC and no added salt). RNase I has been suggested for detecting base pair mismatches (e.g. by Promega Corp., Ambion, Inc.). It was tested in parallel with RNase YI*, for identifying either four contiguous mismatched bases or a single mismatched base pair, by the appearance of a strong band in a polyacrylamide gel. Even with a 4 nt `bubble', RNase I degraded it barely faster than it broke phosphodiester bonds of nucleotides in perfect base pairs. The specific band was only observed in a very narrow range of enzyme activities (Fig. 7 A). In contrast, RNase YI* easily identified the four base pair mismatch over a wide range of activities. The single base pair mismatch was also identified by RNase YI* but was completely missed by RNase I with the activities tested (Fig. 7 B).

DISCUSSION

Endonucleases have marked variability for single versus double strands

Figure 6 . The ³² P-labeled molecules resulting from reactions of 5S-rRNA over a range of S1 or mung bean nuclease activities after fractionation by PAGE. Reactions were in 20 [mu]l at 23^oC for 60 min in 30 mM sodium acetate, pH 5.0, plus 1 mM ZnSO ₄ and contained 1 to 2 [mu]g 5'- ³² P-RNA in ( A ) and 3'- ³² P-RNA in ( B ). Lane 7, no enzyme; lanes 1-6, S1 nuclease at 20, 8, 4, 2, 0.8 and 0.4 U respectively. Lanes 8-12, mung bean nuclease at 25, 10, 5, 2.5 and 1.0 U respectively. Lanes 13-16, S1 nuclease with 8, 20, 40 and 80 U respectively and lanes 17-19, mung bean nuclease with 5, 10 and 25 U respectively. Reactions were stopped by addition of EDTA to 10 mM before adding SDS to 0.1% and 10 M urea to 3 M. The samples were heated at 100^oC for 1 min, cooled rapidly to 0^oC before loading onto the gel.

Figure 7 . Recognition of base pair mismatches in a RNA-DNA duplex. The plasmid pSP73 with a T7 promoter was linearized at the single Bam HI site and then transcribed in vitro in reactions with T7 RNA polymerase to give the following 58 nt RNA product: 5'- GGGAG ACCGG CCTCG AGCAG CTGAA GCTTG CATGC CTGCA GGTCG ACTCT AGAGG ATC-3'. The RNA was purified by PAGE for duplex formation with different DNA oligonucleotides. Hybrids were formed with the two DNA oligonucleotides below that were complementary to the first 39 nt of the RNA except for four consecutive mismatches at positions 19-22 of the RNA ( A ) 5'-GCAGG CATGC AAGCT TC TCG A GCTC GAGGC CGGTC TCCC-3' or one base mismatch at position 23 ( B ) 5'-GCAGG CATGC AAGCT T C AGC TGCTC GAGGC CGGTC TCCC-3'. Each nucleotide in bold underlined was a mismatch of identical bases and is depicted above its panel as a nucleotide in italics and subscript in the RNA in order to coincide with the gel patterns of the RNA. The columns with RNase YI* have no enzyme in the far right side, and in the other lanes from right to left: RNase YI*, 14, 7, 3.5, 1.4 and 0.7 U, respectively. RNase I in (A) from left to right: 25, 50, 100, 200, 0 and 12.5 U respectively; RNase I in (B): no enzyme in the far left lane, and 200, 100, 50, 25 and 12.5 U in the next lanes from left to right, respectively.

An ideal enzyme, or combination of enzymes, for elucidating RNA structure should (i) show high preference for any unpaired nucleotides, (ii) recognize the same phosphodiester bonds at all enzyme activities, and (iii) be active in reaction conditions that do not disrupt the normal RNA structure. As noted, most of the endoRNases used for probing structure have bond preferences that can be a function of the enzyme activity. Of the enzymes that have no obvious bond preferences, the periplasmic RNase I of E.coli has a preference for single strands, but that preference is not sufficiently high, so that the levels of double strand nicks are unacceptable. A single base pair mismatch was not detected and even a mismatch of four contiguous bases was degraded only slightly faster than the perfect RNA-DNA duplexes (Fig. 7 ). As a result, relatively low activities of RNase I are able to degrade total cell RNA, but this weak preference for single strands makes the enzyme unsuitable for structure analysis or base pair mismatch detection. At the other extreme, S1 and mung bean nucleases have a very stringent preference for single strands; only bonds in the presumptive long single-stranded loop of the 5S rRNA were hydrolyzed. Short disruptions of double strands were not detected (Fig. 6 ). Enzymes with such stringent specificity are ideal for eliminating long single-stranded regions without disturbing imperfect duplexes ( 34 ) but they are not good candidates for detecting small loops, `bubbles' or single mismatches, and as such, are not good probes for these unpaired nucleotides.

RNase YI* specificity appears intermediate between the extremes of RNase I and these nucleases for single versus double stranded RNA preference. Escherichia coli RNase I* is also in this category, but its value as a probe of RNA structure has not been studied sufficiently. RNase YI* showed little activity against stable duplexes but was able to recognize and cleave short single-stranded regions of the 5S rRNA and recognized the base pair mismatches tested quite well, i.e. it degraded any phosphodiester bonds of nucleotides that were not perfectly bonded in a duplex. This sensitive discrimination for imperfect versus perfect duplexes was also shown by the stability of the latter structures even after multiple cleavages of the 5S rRNA. When a substrate is in great excess, all cleavages should be initial ones. At sufficiently high enzyme activities, some molecules are cleaved more than once, and larger ³² P-molecules decline as a result of second cleavages closer to the ³² P-end. When RNase YI* was used to degrade 5S rRNA, the progression of band appearance and disappearance appeared to be quite ordered, i.e. only specific bands appeared over the entire range of activities. If one or more hits had disrupted the structure sufficiently, alternative patterns of sizes from the newly generated structures should have been observed with increasing activities. However, new bands did not appear at the highest activities, e.g. no 5'- ³² P bands were ever seen between 3 and 9 nt or 15 and 23 nt. This suggests that the same sites that were vulnerable in the full-length substrate were the vulnerable ones in molecules that may have had two or more cleavages with the separated duplex structures retaining sufficient stability. A case is stem I which remained intact at the highest reaction activities, i.e. some cleavages were occurring at G-9 or G-10 in molecules that may have already had several more distal cleavages. This conclusion is apparent because RNase YI* is sufficiently selective; it does not hydrolyze bonds in perfect duplexes even at activities much greater than are needed to cleave bonds in imperfect base pairs.

This specificity makes it an ideal enzyme for probing RNA structure as well as a candidate for use in detecting base pair mismatches. It has the added advantages of optimal activity near neutral pH, no requirement for cations that might affect RNA structures, activity in the presence of salt, and the generation upon cleavage of 5'-OH ends for easy labeling.

ACKNOWLEDGEMENTS

We thank Peter M. J. Burgers of the Department of Biochemistry and Molecular Biophysics for the original ammonium sulfate supernatant from which RNase YI* was purified and the Boehringer Mannheim Corporation for some of the supplies used in this research.

REFERENCES

1 ten Dam,E., Pleij,K. and Draper,D. (1992) Biochemistry 31, 11665-11676.

2 Shen,L.X. and Tinoco,I.,Jr. (1995) J. Mol. Biol. 247, 963-978. MEDLINE Abstract

3 Chamorro,M., Parkin,N. and Varmus,H.E. (1992) Proc. Natl. Acad. Sci. USA 89, 713-717. MEDLINE Abstract

4 Chen,S., Chamorro,M., Lee,S.I., Shen,L.X., Hines,J.V., Tinoco,I.,Jr. and Varmus,H.E. (1995) EMBO J. 14, 842-852.

5 Simons,R.W. and Kleckner,N. (1988) Annu. Rev. Genet. 22, 567-600. MEDLINE Abstract

6 Feng,S. and Holland,E.C. (1988) Nature 334, 165-167. MEDLINE Abstract

7 Varani,G., Cheong,C. and Tinoco,I.,Jr. (1991) Biochemistry 30, 3280-3289. MEDLINE Abstract

8 Heus,H.A. and Pardi,A. (1991) Science 253, 191-194. MEDLINE Abstract

9 Shen,L.X., Cai,Z. and Tinoco,I.,Jr. (1995) FASEB J. 9, 1023-1033. MEDLINE Abstract

10 Ehresmann,C., Baudin,F., Mougel,M., Romby,P., Ebel, J-P. and Ehresmann,B. (1987) Nucleic Acids Res. 15, 9109-9128. MEDLINE Abstract

11 Knapp,G. (1989) Methods Enzymol. 180, 192-212. MEDLINE Abstract

12 McLennan,B.D. and Lane,B.G. (1967) Can. J. Biochem. 46, 93-107.

13 Witzel,H. and Barnard,E.A. (1962) Biochem. Biophys. Res. Commun. 7, 295-299.

14 Uchida,T., Arima,T. and Egami,F. (1970) J. Biochem. 67, 91-102. MEDLINE Abstract

15 Uchida,T. and Egami,F. (1967) Methods Enzymol. 12, 239-247.

16 Levy,C.C. and Karpetsky,T.P. (1980) J. Biol. Chem. 255, 2153-2159. MEDLINE Abstract

17 Boguski,M.S., Hieter,P. and Levy,C.C. (1980) J. Biol. Chem. 255, 2160-2163. MEDLINE Abstract

18 Vassilenko,S.K. and Rythe,V.C. (1975) Biokhimiya 40, 578-582.

19 Boldyreva,L.G., Boutorine,A.S. and Vassilenko,S.K. (1978) Eur. J. Biochem. 121, 587-595.

20 Lowman,H.B. and Draper,D.E. (1986) J. Biol. Chem. 261, 5396-5403. MEDLINE Abstract

21 Cannistraro,V.J. and Kennell,D. (1991) J. Bacteriol. 173, 4653-4659. MEDLINE Abstract

22 Meador,J. III and Kennell,D. (1990) Gene 95, 1-7.

23 Srivastava,S.K., Cannistraro,V. J. and Kennell,D. (1992) J. Bacteriol. 174, 56-62. MEDLINE Abstract

24 Cannistraro,V.J. and Kennell,D. (1993) Eur. J. Biochem. 213, 285-293. MEDLINE Abstract

25 Cannistraro,V.J. and Kennell,D. (1986) J. Mol. Biol. 192, 257-274. MEDLINE Abstract

26 Ando,T. (1966) Biochem. Biophys. Acta 114, 158-168.

27 Burgers,P.M.J. (1995) Methods Enzymol., 262, 49-62.

28 Meador,J., III, Cannon,B., Cannistraro,V.J. and Kennell,D. (1990) Eur. J. Biochem. 187, 549-553. MEDLINE Abstract

29 Cannistraro,V.J. and Kennell,D. (1994). J. Mol. Biol. 243, 930-943.

30 Brownlee,G.G. (1972) Determination of Sequences in RNA. North-Holland, Amsterdam and Elsevier, New York.

31 Noller,H.F. (1984) Annu. Rev. Biochem., 53, 119-162. MEDLINE Abstract

32 Fox,G.E. and Woese,C.R. (1975) Nature 256, 505-507.

33 White,S.A., Nilges,M., Huang,A., Brunger,A.T. and Moore,P.B. (1992) Biochemistry 31, 1610-1621. MEDLINE Abstract

34 Berk,A.J. and Sharp,P.A. (1977) Cell 12, 721-732. MEDLINE Abstract

---

Return

^* To whom correspondence should be addressed. Tel: +1 314 362 2751; Fax: +1 314 362 1232; Email: kennell@borcim.wustl.edu
CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				D. Kennell Processing Endoribonucleases and mRNA Degradation in Bacteria J. Bacteriol., September 1, 2002; 184(17): 4645 - 4657. [Full Text] [PDF]

				G. C. MacIntosh, P. A. Bariola, E. Newbigin, and P. J. Green Characterization of Rny1, the Saccharomyces cerevisiae member of the T2 RNase family of RNases: Unexpected functions for ancient enzymes? PNAS, January 30, 2001; 98(3): 1018 - 1023. [Abstract] [Full Text] [PDF]

This Article Abstract Print PDF (143K) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (8) Request Permissions Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Cannistraro, V. J. Articles by Kennell, D. Search for Related Content PubMed PubMed Citation Articles by Cannistraro, V. J. Articles by Kennell, D. Social Bookmarking          What's this?