INTRODUCTION

Transfer RNA undergoes extensive processing during the production of the mature molecule. The endoribonuclease ribonuclease P (RNase P)¹ cleaves the nascent transcript (pre-tRNA) to generate the mature 5 termini of tRNA in both procaryotes and eucaryotes (1, 2). RNase P is an unusual enzyme in that the holoenzyme is a ribonucleoprotein. The RNA and protein subunits of the bacterial RNase P are required for enzymatic functions in vivo (3, 4, 5) and under physiologic conditions in vitro (6). In buffers containing high concentrations of monovalent and divalent salts, the RNA subunit (P RNA) displays the catalytic activity of the bacterial enzyme (6). The protein subunit binds to the P RNA in the bacterial holoenzyme and serves to promote product release during the catalytic cycle (7). The RNase P holoenzyme of eucaryotic organisms also exists as a ribonucleoprotein; however, the functional roles of RNA and protein subunits of these enzymes are not well defined. The holoenzyme activities have been characterized in many eucaryotes including mammalian systems (8, 9, 10), yeast (11, 12, 13), Xenopus (14), spinach (15), potato (16), and Dictyostelium (17). Many studies have shown that RNA and protein molecules are essential for RNase P function in vivo (18, 19) and in vitro (8, 9, 17), but biochemical and genetic analyses of holoenzyme composition have failed to provide clues as to their functional assignments in RNase P.

Our interest in RNase P of eucaryotes is to develop a greater understanding about the catalytic mechanism for the holoenzyme and to learn the manner in which the RNA and protein subunits interact in the holoenzyme to achieve catalytic activity. Phylogenetic analyses of the components of the bacterial RNase P holoenzyme suggest that both the RNA and protein subunit genes change more rapidly than the corresponding ribosomal RNA gene sequences (1, 20). In this regard, the molecular clocks of genes that encode the subunits of RNase P resemble those of genes that encode typical polypeptide enzymes. We have chosen to focus our studies on the RNase P holoenzymes from ciliate protozoa because these organisms comprise an extremely phylogenetically diverse group of organisms (21). Through the examination of the biochemical and molecular properties of homologous enzymes from phylogenetically diverse origins, considerable insight can be obtained relevant to the catalytic mechanism of RNase P and the role of the molecular components in holoenzyme function. We describe in this report our initial purification and characterization of the RNase P from Tetrahymena thermophila.

EXPERIMENTAL PROCEDURES

T. thermophila nuclei were isolated as described previously (22). Genomic DNA was extracted from sonicated nuclei according to standard procedures (23) to serve as a template for the polymerase chain reaction amplification (24) of the T. thermophila tRNA^Gln and 5-flanking DNA (25). Amplification of tRNA^Gln was accomplished using Vent DNA polymerase (New England Biolabs Inc.) according to manufacturer's instructions and the following synthetic DNA primers (Operon Technologies, Inc.): TGLN3: 5-CGGAATTCGGTCTCGTGGAGGTCCCACTGGGATTCG-3 and TGLN5: 5-CGGGATCCTAATACGACTCACTATAGGAAATAGAAAATATTGTGTGCTGG-3. The resulting DNA amplification product was digested with EcoRI and BamHI restriction endonucleases (New England Biolabs) and cloned into pSP65 (Promega) using T4 DNA ligase (New England Biolabs) to produce the plasmid pTGLN1. The DNA sequence of the cloned insert was verified by double-stranded cycle sequencing procedures (26) using Vent(exo-) DNA polymerase (New England Biolabs).

The plasmid pTGLN1 was linearized with BsaI (New England Biolabs) and transcribed in vitro using purified T7 RNA polymerase (27) to generate a pre-tRNA for use in 5-end-processing studies. The ³²P internally labeled pre-tRNA substrate was generated in a 20-µl transcription reaction that contained 1-2 µg of linearized template DNA, 0.5 mM GTP, 0.5 mM CTP, 0.5 mM UTP, 0.125 mM ATP, 50 mM Tris-HCl (pH 8.0), 10 mM MgCl₂, 10 mM dithiothreitol, 50 µCi of [-³²P]ATP, and T7 RNA polymerase and proceeded for 2 h at 37 °C. The reaction was terminated with the addition of an equal volume of sample loading buffer (SLB: 40 mM EDTA, 8 M urea, 0.2% xylene cyanol, and 0.2% bromphenol blue). The product was resolved on a denaturing 15% polyacrylamide gel, and the full length pre-tRNA was detected by autoradiography. The precursor tRNA was eluted from the gel slice in elution buffer (0.25 M NaCl, 10 mM Tris-HCl (pH 8.0), 1 mM EDTA) overnight at 4 °C. The transcribed pre-tRNA was ethanol-precipitated from the eluate and resuspended in 75 µl of H₂O. Unlabeled pre-tRNA was generated following the procedure of Latham et al. (28).

The RNase P activity was monitored throughout the purification and characterization by assaying the ability of fractions to cleave ³²P internally labeled T. thermophila pre-tRNA^Gln. Reactions were carried out in MBB buffer (50 mM Tris-HCl (pH 7.5), 10 mM MgCl₂) for 1 h at 37 °C. Reactions were terminated with the addition of an equal volume of SLB. These reaction products were separated by electrophoresis on 8 M urea, 6% polyacrylamide gels and analyzed on a Molecular Dynamics PhosphorImager^TM system using Image Quant software. One unit of T. thermophila RNase P activity is defined as the amount of enzyme required to cleave 10% of 28 fmol of T. thermophila pre-tRNA^Gln to mature products in 1 h at 37 °C.

The protein concentrations were determined by the method of Bradford (29) in a microtiter plate format. The absorbance at 600 nm was measured with the use of a Molecular Devices microplate reader. Bovine plasma -globulin (Bio-Rad) was used as a protein standard.

Seventy-two liters of T. thermophila (strain B1968, mating type VII) were grown axenically in modified NEFF media (0.75% proteose peptone (Difco), 0.75% yeast extract (Difco), 1 mM MgSO₄, 2 mM KH₂PO₄ (pH 6.5), 0.05 mM CaCl₂, and 43 mg/liter Sequestrene NaFe13% (Ciba-Geigy; (30)) at 30 °C to a culture density of 3 × 10⁵ cells/ml. During harvesting procedures and all subsequent purification steps the extract material was maintained at 4 °C. Cells were harvested by centrifugation for 10 min at 10,550 × g. The cells were washed once with 20 volumes of ultrapure H₂O and reharvested. Pellets were resuspended in 2 volumes of SB buffer (20 mM Tris-HCl (pH 7.5), 0.1 mM EDTA, 10 mM MgCl₂, 60 mM NH₄Cl, 5% glycerol (v/v), 1 mM dithiothreitol, 1 mM p-hydroxymercuribenzoate (Sigma), 1 mM leupeptin (Sigma)) per volume of wet pellet and stored at 20 °C. Cell suspensions were thawed at 20 °C and was subjected to Dounce homogenization 25 times in a Wheaton glass 15-ml tissue grinder fitted with a Type B pestle. The extract was clarified by centrifugation at 16,000 × g for 20 min. The supernatant from this initial clarification contained the majority of RNase P activity and was stored at 4 °C. A significant amount of RNase P activity remained associated with the cell debris following this lysis procedure. The majority of this trapped activity could be recovered by the incorporation of a wash step. The pelleted cell debris was resuspended in an equal volume of SB buffer and reclarified by centrifugation to enhance the recovery of RNase P activity.

The initial and wash supernatants were combined for a total of 413 ml, which was then subjected to ammonium sulfate precipitation. Solid (NH₄)₂SO₄ (Fisher) was added to the clarified extract sequentially to 20, 40, and 60% saturation. The RNase P activity was distributed equally between the 40 and 60% pellets of the (NH₄)₂SO₄ fractionation, and these pellets were resuspended in PGMK buffer (0.02 M KPO₄ (pH 6.5), 10% glycerol, 0.01 M MgCl₂, 0.3 M KCl) to yield a final volume of 67.5 ml. Sixty-seven milliliters of this activity was applied by batch absorption to 100 ml of BioGel HT hydroxylapatite (Bio-Rad) equilibrated in PGMK buffer. The resin was washed sequentially with 2 volumes of PGMK buffer, 2 volumes of 1 M KCl, and again with 1 volume of PGMK buffer. The resin was then vacuum-packed into a 0.45-µm Nalgene 115-ml filter unit and washed with 2 volumes of 0.06 M PGMK buffer (0.06 M KPO₄ in PGMK, pH 6.5). RNase P was subsequently eluted from the resin with 1 volume of 0.08 M PGMK (0.08 M KPO₄ in PGMK, pH 6.5). The RNase P activity was precipitated from this eluate with the addition of solid (NH₄)₂SO₄ to 60% saturation. Pelleted material from the (NH₄)₂SO₄ precipitation was resuspended in HCB7 buffer (15 mM K-HEPES (pH 7.0), 10% glycerol, 3 mM MgCl₂, 0.1% Tween 20, 25 mM KCl) and dialyzed for 18 h against HCB7 buffer (1.6 liters) to yield a final volume of 10.8 ml.

Hydroxylapatite eluate (10.5 ml) was applied to a 76-ml Affi-Gel heparin (Bio-Rad) column (2.5 cm diameter) at a flow rate of 0.2 ml/min. The column was washed with HCB7 buffer (approximately 3 column volumes) at a flow rate of 1 ml/min until the absorbance at 280 nm returned to base line. RNase P activity was eluted from the column at a flow rate of 1 ml/min with a 480-ml linear gradient from 25 mM KCl to 250 mM KCl in HCB7 buffer. RNase P eluted in the latter two-thirds of the gradient and was concentrated in an Amicon stirred cell concentrator. The concentrated fractions (16.1 ml) were dialyzed for 8 h against HCB7 buffer (1.6 liters).

RNase P activity (15.8 ml) that was recovered from Affi-Gel heparin fractionation was applied to a 120 ml DEAE-Sepharose (Pharmacia Biotech Inc.) column (2.5 cm diameter) at a flow rate of 1 ml/min. The column was washed with 3 volumes of HCB7 buffer before the RNase P activity was eluted with a 600-ml linear gradient from 25 mM KCl to 500 mM KCl in HCB7 buffer. The RNase P activity began to elute at approximately 250 mM KCl. Fractions containing RNase P activity were pooled and concentrated in an Amicon stirred cell concentrator fitted with an Amicon YM-10 membrane. This activity (11.2 ml) was dialyzed against HCB7 buffer (1.5 liters) for 3 h, and an aliquot of the dialysate (2.25 ml) was applied to a 10-ml spermine-agarose (Sigma) column (1.5 cm diameter) at a flow rate of 0.2 ml/min. The column was washed sequentially with 3 volumes of HCB7 buffer and 140 ml of HCB7 buffer containing 250 mM KCl. RNase P was eluted with a 240-ml linear gradient of KCl from 250 mM to 1 M in HCB7 buffer. The activity began eluting at approximately 300 mM KCl, and the active fractions were pooled (24 ml) and concentrated (5.5 ml). The spermine-agarose-eluted RNase P was dialyzed against HCB7 buffer (1.6 liters) for 3 h.

A tRNA affinity column was constructed by coupling 40 µg of biotinylated tRNA^nscendtRNA (Promega) to 8 mg of streptavidin Dynabeads M-280 (Dynal, Inc.) in phosphate-buffered saline (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na₂HPO₄, 1.4 mM KH₂PO₄, pH 7.3) with shaking at room temperature for 1 h. After binding, the beads were washed twice with HCB7 buffer containing 1 M NaCl and washed twice with HCB7 buffer at room temperature. An aliquot (250 µl) of RNase P activity that had been recovered from the spermine-agarose fractionation was diluted 2-fold in HCB7 buffer and bound to the resin for 1 h at 4 °C with shaking. The unbound RNase P activity was removed after harvesting the tRNA-containing streptavidin beads with a Dynal magnetic particle concentrator. The beads were subsequently washed five times with 1 ml of HCB7 buffer. The RNase P activity was eluted with 200 µl of HCB7 buffer containing 1 M KCl.

Cesium sulfate step density gradients were generated using 1.6-ml layers of 25, 31, and 37% Cs₂SO₄ (w/v) in buffer G (50 mM Tris-HCl (pH 8.0), 100 mM KCl, 10 mM MgCl₂, 10% glycerol) as described previously (31). A 500-µl sample of concentrated, partially purified RNase P was applied to the top of the gradient. The gradient was formed by centrifugation at 35,000 rpm for 23 h at 18 °C in a Beckman SW 50.1 rotor. Fractions of 200 µl were drawn from the top of the gradient. A portion of each sample was dialyzed against HCB7 buffer for 2 h at 4 °C and an RNase P activity assay was completed immediately thereafter. The refractive index of each fraction was measured and the density was calculated as described previously (32). The densities of the RNA and protein standards were determined in separate experiments. The buoyant density for the Escherichia coli RNase P standard could not be accurately determined with the step density gradients described above. For the E. coli RNase P, step density gradients were generated using 1.6-ml layers of 31, 37, and 45% Cs₂SO₄ (w/v) in buffer G. After centrifugation, the gradient fractions were collected either by removing 200-µl aliquots from the top of the gradient or by collecting 3-4 drop fractions from the bottom of the gradient through an 18-gauge needle. Similar density values were obtained using either collection method.

RNase P activity enriched through spermine-agarose chromatography was preincubated with 30 units of micrococcal nuclease (Pharmacia) in a 6-µl reaction containing 1 mM CaCl₂ at 37 °C for 30 min. Micrococcal nuclease was inactivated with the addition of EGTA to a final concentration of 3.33 mM. RNase P activity in this mixture was then assayed in a 15-µl reaction by the ability to cleave the pre-tRNA in the presence of MN buffer (0.67 µg/µl polyadenylate (Pharmacia), 50 mM Tris-HCl (pH 7.5), 20 mM MgCl₂) at 37 °C in 30 min (33). A control reaction was performed in which 8.3 mM EGTA was added to the micrococcal nuclease to inactivate the enzyme by the chelation of calcium. RNase P was preincubated with the EGTA-inactivated micrococcal nuclease at 37 °C for 30 min prior to the pre-tRNA processing assay. RNase P activity assays were also performed in MN buffer and under standard conditions using MBB buffer. RNase P was treated with proteinase K (1 µg/µl) for 30 min at 37 °C and RNase P activity was assayed as described previously.

The optimal temperature and pH of the T. thermophila RNase P were determined by analyzing the initial velocity of the cleavage reaction over a range of temperature and pH values. The enzyme activity and the buffering components were preincubated for 5 min at the temperature of assay before the pre-tRNA substrate was added to the mixture to initiate the assay. Aliquots of the reaction were removed and terminated with the addition of SLB, followed by immediate cooling to 0 °C. The optimal temperature was determined in MBB buffer. The optimal pH was determined at 37 °C in a buffer containing 9 mM MgCl₂ and 50 mM MOPS for the pH range of 6.3-8.75.

The effects of monovalent salts (NaCl, CsCl, KCl, and NH₄Cl) and divalent salts (CaCl₂, MgCl₂, MnCl₂) on RNase P function were assessed using activity diluted in HB buffer (15 mM K-HEPES (pH 7.0), 10% glycerol, 0.1% Tween 20) such that upon dilution, the maximum concentrations of MgCl₂ and KCl were 0.015 and 0.125 mM, respectively. The salt(s) and the appropriate amount of water were premixed prior to the addition of enzyme. The enzyme mix was then preincubated at 37 °C for 5 min prior to the addition of pre-tRNA^Gln substrate in MBB buffer, which was also pre-warmed to 37 °C. The initial velocity of the processing reaction was determined as described above.

The apparent K_m was determined by measuring the velocity of the reaction as a function of pre-tRNA^Gln concentration using RNase P activity enriched through spermine-agarose chromatography. The assays were performed at 37 °C in MBB buffer with pre-tRNA^Gln used as the substrate.

Puromycin dihydrochloride (Sigma) was prepared for use as described previously (10). The concentration of puromycin diluted in 0.1 M HCl was determined spectrophotometrically (₂₆₇ = 1.95 × 10⁴ M¹ cm¹) (34). RNase P activity was diluted in HCB7 buffer and preincubated at 37 °C for 5 min with puromycin in MBB buffer that contained a final concentration of 20 mM NaCl. The activity assay was initiated with the addition of pre-tRNA^Gln substrate, and the reaction was incubated at 37 °C for a further 30 min. The assays were analyzed as described above.

RESULTS

RNase P from T. thermophila was purified nearly 700-fold by the chromatographic steps summarized in Table I. The enzyme activity cleaved the 98-nucleotide pre-tRNA^Gln substrate into a 75-nucleotide mature tRNA and a 23-nucleotide 5-leader sequence (Fig. 1A). The E. coli RNA subunit (M1 RNA) and holoenzyme also cleaved this substrate RNA to generate the same fragments observed with the T. thermophila enzyme (data not shown). The enzyme activity was stable through DEAE-Sepharose chromatography in this purification scheme when stored in HCB7 buffer at 4 °C or at 20 °C. T. thermophila are rich in protease and nuclease activities (35, 36); therefore, co-purifying substances that inactivate RNase P may account for the loss of activity observed in the purification procedures. A contaminating nucleolytic activity co-purified with RNase P activity through the DEAE-Sepharose step and was detected by the appearance of an anomalous cleavage product of pre-tRNA^Gln in activity assays (data not shown). RNase P was resolved from this contaminating nuclease by chromatography on spermine-agarose. RNase P was effectively recovered in three fractions following spermine-agarose chromatography (Fig. 1B) but was unstable if not concentrated immediately.² In a concentrated form (0.48 mg/ml total protein), the RNase P enriched through spermine-agarose fractionation was stable in HCB7 buffer at 4 °C or at 20 °C for months. Specific activity measurements for this enzyme purification fluctuated during steps such as Affi-Gel heparin and spermine-agarose, which apparently concentrate inhibitors of the activity. The inclusion of these fractionations was necessary because they enabled RNase P to be further enriched by affinity chromatography using an immobilized tRNA resin. RNase P enriched through tRNA-agarose chromatography was analyzed by SDS-polyacrylamide gel electrophoresis and Coomassie blue staining to assess the degree of purity. One polypeptide of an estimated molecular mass of 36 kDa was visible by this procedure (data not shown). Several RNA species copurify with RNase P activity through spermine-agarose, as assessed by a 3-end-labeling assay using [-³²P]pCp and T4 RNA ligase. Four RNA species were most abundant and range from 80 to 230 nucleotides in length (data not shown).

Table I. Purification of RNase P from T. thermophila Purification step Volume Total protein Activitya Specific activity Purificationb ml mg units units/mg -fold S10 413.0 2.11  × 104 2.83  × 107 1.34  × 103 1.0 (NH4)2SO4 precipitant 407.0 1.15  × 104 6.84  × 107 5.92  × 103 4.4 Hydroxylapatite 67.0 679.64 2.55  × 107 3.75  × 104 28.0 Affi-Gel heparin 10.5 236.65 6.09  × 105 2.57  × 103 1.92 DEAE-Sepharose 15.8 52.19 2.27  × 107 4.35  × 105 324.6 Spermine-agarose 2.8 2.79 8.58  × 105 3.08  × 105 230.0 tRNA affinity 0.5 7.84  × 103 7.30  × 102 9.31  × 105 694.8 a One unit is defined as the amount of activity required to cleave 2.8 fmol of pre-tRNAGln to mature form in 1 h at 37 °C. b Fold enrichment is based upon specific activity.

Many ribonucleoproteins possess buoyant densities in a Cs₂SO₄ gradient that are greater that observed for protein. In order to investigate this property for RNase P from T. thermophila, density gradient centrifugation was performed. After Cs₂SO₄ gradient centrifugation, fractions were collected, dialyzed briefly to remove the salt, and immediately assayed for activity. The T. thermophila holoenzyme, either from crude extract or in partially purified form, came to an equilibrium between yeast tRNA^Phe and the RNase P holoenzyme of E. coli in the gradient (Fig. 2). The buoyant density of the Tetrahymena RNase P in Cs₂SO₄ was centered around 1.42 g/ml (Fig. 2). This density is greater than that observed for protein in this gradient,³ which suggests that the enzyme is composed of RNA and protein.

The requirement for both RNA and protein in the cleavage of precursor tRNA was determined by selective elimination of the components prior to activity assay. Micrococcal nuclease (MN) effectively eliminated the ability of RNase P from Tetrahymena to generate mature tRNA (Fig. 3, lanes 4 and 5). Pretreatment of RNase P with micrococcal nuclease that was inactivated with EGTA prior to the assay failed to abolish activity (Fig. 3, lane 3). The addition of polyadenylate following micrococcal nuclease digestion was required to assay the ability of RNase P to cleave pre-tRNA but did not restore activity to the MN-digested fraction (33). Concentrations of calcium, EGTA, or polyadenylate greater than those used in these experiments resulted in a decrease in detectable RNase P activity (data not shown). Proteinase K digestion of RNase P prior to the pre-tRNA processing assay also abolished enzyme function (Fig. 3, lane 7). Therefore, the selective removal of either RNA or protein prior to activity assay eliminated the ability of RNase P to generate mature tRNA.

The biochemical properties of the T. thermophila RNase P were determined in order to compare this enzyme's characteristics to RNase P from other organisms. The optimal temperature for the Tetrahymena RNase P was 40 °C under standard assay conditions (Fig. 4A). The pH dependence of the reaction varied with the buffer used in the assay. The optimal pH for the activity in MOPS buffer is about 6.8 (Fig. 4B). The enzyme had a higher pH optimum (pH 7.5) in reaction buffers that contain 50 mM Tris-HCl rather than 50 mM MOPS.⁴

To determine the behavior of the enzyme in different ionic conditions, the pre-tRNA processing activity was assessed in the presence of various monovalent and divalent salts. The T. thermophila enzyme was inhibited significantly, even at low concentrations, by all monovalent salts tested (Fig. 4C). Pre-tRNA^Gln processing was not detected for T. thermophila RNase P preparations equilibrated in HCB7 buffer lacking divalent cations (data not shown). Fig. 4D illustrates that the pre-tRNA processing activity displayed by RNase P preparations equilibrated in buffer containing an optimal concentration of MgCl₂ (5 mM) was approximately 30% greater than that observed by RNase P preparations equilibrated in HCB7 buffer containing an optimal concentration of MnCl₂ (2 mM). When the divalent salt concentration is increased 5-fold relative to the optimal concentration of each divalent cation, RNase P activity is reduced by approximately 30-50% (Fig. 4D). Pre-tRNA^Gln processing activity was not detected for RNase P preparations equilibrated in buffer containing CaCl₂ as the sole divalent salt at any concentration tested (Fig. 4D).

The inability of Ca(II) to support RNase P function prompted us to investigate whether processing buffers containing Ca(II) in the presence of Mg(II) or Mn(II) could support holoenzyme activity. The addition of Ca(II) to processing buffer containing suboptimal concentrations of Mn(II) stimulated holoenzyme pre-tRNA^Gln processing activity to levels greater than that observed for RNase P equilibrated in buffers containing the optimal concentration of Mn(II) only (Fig. 5). This stimulatory effect was not observed for RNase P equilibrated in buffer containing Ca(II) and Mg(II). Equimolar amounts of Ca(II) neither stimulated nor inhibited holoenzyme activity in buffer mixtures that contain suboptimal concentrations of Mg(II) (data not shown). Calcium(II) inhibited RNase P function when present at a molar excess greater than 4-fold relative to Mg(II) (Fig. 5; data not shown).

Polyamine supplementation has been demonstrated to enhance the ability of RNase P to generate mature tRNA in some systems in vitro (37, 38). The inclusion of spermidine in processing assay buffers did not support pre-tRNA cleavage by the T. thermophila RNase P in the absence of divalent ions. Furthermore, spermidine did not enhance activity in processing assays when buffers contained divalent ions at concentrations insufficient for optimal holoenzyme activity (data not shown).

The apparent K_m for the processing of pre-tRNA^Gln by partially purified RNase P was determined by an Eadie-Hofstee plot to be 160 nM (Fig. 6).

7

Puromycin is a potent inhibitor of ribosome function by virtue of its ability to interfere with tRNA binding at the A site (39). To assess whether puromycin inhibits the T. thermophila RNase P, the pre-tRNA^Gln processing assay was done in the presence of 1-8 mM puromycin. A concentration of 2 mM puromycin in the activity assay resulted in approximately 50% inhibition of cleavage of pre-tRNA substrate (Fig. 7).

DISCUSSION

This study describes the isolation and characterization of RNase P from a ciliate protozoan. T. thermophila was chosen as a candidate organism for these studies because the organism represents one of the few ciliate protozoa that can be grown under axenic conditions. Many of the unique properties of the RNase P from T. thermophila were observed during our attempts to purify the activity from crude whole organism lysates. RNase P from other eucaryotic organisms has been purified by chromatography on successive anion and cation exchange resins (11, 13, 14, 16, 17). We were unable to adsorb the RNase P from T. thermophila to either phosphocellulose or SP-Sepharose cation exchange resins under any condition tested, which suggests that the enzyme has very strong anionic character. The application of hydrophobic resins, glycerol gradient sedimentation, and Cs₂SO₄ density gradient centrifugation, each of which had been successfully used during the purification of other eucaryotic RNase P enzymes (9, 31), failed to afford us either substantial or stable enrichment of the T. thermophila enzyme. Other techniques that also proved unsuccessful in the purification of the RNase P from T. thermophila include the use of dye-coupled resins and tRNA-agarose resin (14). The enzyme activity could be adsorbed to the dye-coupled resin Cibacron blue under mildly acidic conditions (pH <6.0); however, the enzymatic activity was unstable if the activity was maintained under these conditions for prolonged periods of time (72 h). The chromatographic behavior of the T. thermophila enzyme during purification suggests that the RNase P enzyme from ciliate protozoa may differ substantially from RNase P holoenzymes of other eucaryotic organisms.

Despite its unique properties that were revealed during purification, the RNase P of T. thermophila possesses many enzymatic properties that are shared with RNase P from other organisms. The Tetrahymena RNase P displays a pH optimum and temperature optimum characteristic of RNase P enzymes isolated from other organisms. The K_m of the T. thermophila enzyme for pre-tRNA^Gln is 1.6 × 10⁷ M, which is comparable to the values reported for other examples of RNase P, 2.5 × 10⁷ M for Sulfolobus acidocaldarius RNase P (40), 2.3 × 10⁷ M for Saccharomyces cerevisiae mitochondrial RNase P (13), 4.2 × 10⁸ M for E. coli RNase P (41, 42), 2.0 × 10⁷ M for Bacillus subtilis RNase P (7), and 2.4 × 10⁷ M for Dictyostelium discoideum RNase P (17).

RNase P enzymatic activities from other sources are mildly stimulated by low concentrations (<100 mM) of monovalent salts; however, RNase P activity from these organisms are inhibited when the enzyme activity assays are performed in the presence of elevated concentrations (>100 mM) of monovalent salts (13, 17, 40). The Tetrahymena enzyme activity is reduced by the presence of monovalent salts in the reaction mixture. The molecular significance of this remarkable ionic strength effect on the Tetrahymena RNase P activity remains unclear. It is possible that monovalent salts alter the holoenzyme conformation, displace a required divalent cation, or alter the manner in which the enzyme interacts with substrates and products during the catalytic cycle.

Almost all RNase P enzymes require divalent cations for their activity (1, 2, 13, 37, 38). One reason for this metal ion requirement stems from the hydrolytic mechanism of the pre-tRNA processing reaction. Divalent metal ions are presumed to have two roles during the hydrolytic cleavage of the scissile bond in a pre-tRNA substrate; one metal ion acts as a Lewis acid to generate a suitable nucleophile from water, and the other metal ion stabilizes the negative charge that develops at the phosphodiester bond during nucleophilic attack (43, 44). Divalent cations also can serve structural roles in catalytic function, possibly participating in folding enzyme components into a conformation required for activity (45) or allowing for the substrate to fit into the active site (46). The Tetrahymena RNase P, like RNase P from other organisms, possesses an absolute divalent metal ion requirement for activity. The divalent cations Mg(II), Mn(II), and Ca(II) were found to stimulate other RNase P activities to varying degrees (13, 37, 38, 43). The RNase P activity from T. thermophila was supported by either Mg(II) and Mn(II) but not by Ca(II). This remarkable divalent ion preference displayed by the Tetrahymena RNase P is more closely related to the corresponding divalent ion preferences observed with the large ribosomal RNA self-splicing group I intron RNA from T. thermophila (47) than to the corresponding divalent ion preferences observed with RNase P from other organisms.

The behavior of the Tetrahymena RNase P in the presence of different divalent cations suggest that two types of divalent cation binding sites exist in the holoenzyme, sites whose occupancy is necessary for catalysis (catalytic sites) and sites whose occupancy is required for an auxiliary function (structural sites). When RNase P is equilibrated in buffers containing only one type of divalent ion, Mg(II) and to a lessor extent Mn(II) can fulfill the roles of both catalytic and structural sites. Calcium(II) is deficient in one of these roles, since holoenzyme activity was not evident in buffers containing this metal ion only. The results from the enzymatic activity assays for the buffers that contain mixtures of Ca(II) with either Mg(II) or Mn(II) suggest that Ca(II) can compete for, and in some cases function in, a site that is normally occupied by Mg(II) or Mn(II) in an active holoenzyme. Divalent cations have been rigorously shown to possess similar roles in catalysis by a shortened form of the Tetrahymena group I intron RNA (47). Unlike the metal ion studies completed with the group I intron RNA, however, we know little about the exact role of these metal ions in relation to Tetrahymena RNase P function. Metal ions may participate directly in catalysis in the Tetrahymena enzyme or may promote binding of the pre-tRNA substrate to the enzyme, as observed for the bacterial enzyme (37, 38, 43). The purified components of the holoenzyme must be obtained to rigorously test these roles and other functions of divalent metal ions for the Tetrahymena RNase P.

Structural aspects of transfer RNA are recognized by at least three types of macromolecules in the cell: RNase P, aminoacyl tRNA synthetase, and the ribosome. The sensitivity that both ribosomes and the Tetrahymena RNase P display toward puromycin suggests that the antibiotic may interfere with tRNA binding events in both macromolecules (39). Puromycin also has been shown to inhibit RNase P holoenzyme activity from mouse and bacteria (10, 48) and the catalytic activity of the E. coli M1 RNA (48). Since an RNA-catalyzed reaction can be inhibited by puromycin, Vioque suggested that similar substrate recognition mechanisms exist for ribosomal RNA and for the E. coli M1 RNA (48). The inhibitory effects of puromycin on Tetrahymena RNase P function could not be alleviated by arginine,⁴ an amino acid which can bind to RNA (49) and inhibit puromycin binding to ribosomes (50). This suggests that while there may be similarities in the recognition of tRNA and puromycin by ribosomes and RNase P, there are still fundamental differences in their recognition or affinity.

The Tetrahymena RNase P is composed of both RNA and protein. The buoyant density of the Tetrahymena RNase P is greater than that reported for HeLa (human) nuclear RNase P ( = 1.28 g/ml (31)), the Xenopus laevis RNase P ( = 1.34 g/ml (14)), the S. cerevisiae mitochondrial RNase P ( = 1.28 g/ml (13)), the spinach chloroplast RNase P ( = 1.28 g/ml in CsCl (15)), and the D. discoideum RNase P ( = 1.23 g/ml (17)) as eucaryotic representatives and is greater than that reported for the S. acidocaldarius RNase P ( = 1.27 g/ml (40)) as an Archaea representative. All of these enzymes, including the Tetrahymena RNase P, display buoyant densities on Cs₂SO₄ gradients that are considerably less than that observed with the RNase P from E. coli ( = 1.55 g/ml (8, 51)). The buoyant density of the E. coli enzyme corresponds to an enzyme composed of 10% protein and 90% RNA, in accordance with its known 119-amino-acid protein (52) and a 377-nucleotide RNA (53, 54). The buoyant density of the Tetrahymena RNase P presumably corresponds to a holoenzyme composed of about 50% protein and 50% RNA (55).

The RNase P holoenzyme from T. thermophila possesses RNA and protein components that are essential for catalytic activity in vitro. RNA and protein of RNase P from other organisms have been shown to play essential, although not always well understood, roles in holoenzyme function. In the bacterial holoenzymes, the RNA subunit harbors the catalytic activity (6), while the protein subunit serves to promote product release during the catalytic cycle (7). The division of labor is not so clearly established for examples of RNase P from organisms of either Eucarya or Archaea. The genetic isolation of mutants in RNase P function has provided important clues that point to the in vivo relevance of both protein and RNA subunits in pre-tRNA processing in S. cerevisiae (18, 19). Eucaryotic P RNA has not been demonstrated to maintain the ability to cleave substrate pre-tRNA in the absence of its protein counterpart(s). Whether similar observations will hold true for the P RNA subunit of Tetrahymena may be addressed once the candidate RNA subunit(s) for the RNase P holoenzyme of this organism is isolated.