Cleavage of oligoribonucleotides by the 2',5'-oligoadenylate- dependent ribonuclease L.

RNase L, the 2',5' oligoadenylate-dependent ribonuclease, is one of the enzyme systems important in the cellular response to interferon. When activated in the presence of 2',5'-linked oligoadenylates, RNase L can catalyze the cleavage of synthetic oligoribonucleotides that contain dyad sequences of the forms UU, UA, AU, AA, and UG, but it cannot catalyze the cleavage of an oligoribonucleotide containing only cytosines. The primary site of the cleavage reaction with the substrate C11UUC7 has been defined to be 3' of the UU dyad by labeling either the 5' or the 3' end of the oligoribonucleotide and by examining the reaction products on polyacrylamide sequencing gels. Reaction time courses have been used to determine the kinetic parameters of the cleavage reactions. The effect of the overall length of the oligomeric substrate as well as the sequence of the bases around the position of the cleavage site on the kinetics of the cleavage reaction has been examined. The efficiency with which activated RNase L catalyzes the cleavage of the substrate C11UUC7 is 1.9 x 10(7) m-1 s-1. Because the cleavage of the synthetic oligoribonucleotide can be used to monitor the steady-state kinetics of catalysis by activated RNase L, this method offers an advantage over previous methods of assay for RNase L activity.

One of the enzyme systems whose activity is enhanced on treatment with interferon is RNase L, the 2Ј,5Ј oligoadenylatedependent ribonuclease (Kerr and Brown, 1978). Interferons induce the expression of 2Ј,5Ј oligoadenylate synthetases, which catalyze the formation of oligomers of adenosine linked 2Ј to 5Ј (Hovanessian et al., 1977). These 2Ј,5Ј oligoadenylates bind to and activate the latent RNase L to cleave viral and cellular RNAs at the 3Ј side of UpNp sequences (Floyd-Smith et al., 1981;Wreschner et al., 1981), leading to the inhibition of protein synthesis.
RNase L activity has been shown to be important in the mechanisms of cellular antiviral defense. Overexpression of 2Ј,5Ј oligoadenylate synthetase leads to the inhibition of picornavirus replication (Chebath et al., 1987;Rysiecki et al., 1989). Introduction of an inactive mutant of RNase L caused an increased susceptibility to infection by picornavirus and a loss of the inhibition of cell growth caused by interferon treatment . Introduction of 2Ј,5Ј oligoadenylate into cells (Hovanessian and Wood, 1980) or the expression of 2Ј,5Ј oligoadenylate synthetase (Rysiecki et al., 1989) have been found to cause growth arrest, suggesting a role for the RNase L system in the regulation of cell growth.
The genes for human and murine RNase L have been cloned and sequenced . Overexpression of RNase L employing baculoviral vectors and insect cell lines (Dong et al., 1994) has supplied enough material for protein chemical experiments to be performed. The recent demonstration of the formation of dimers of RNase L in the presence of activators  has opened up new areas of investigation regarding the mechanism of action of the activated complex and has created the need for detailed investigation of the kinetics of catalysis by RNase L.
Current methods of assay for RNase L activity rely on the cleavage of radiolabeled poly(U) (Silverman, 1985) followed by quantification of the remaining substrate with precipitation or on the cleavage of ribosomal RNA (Kariko et al., 1987) followed by analysis of specific cleavage products with agarose gel electrophoresis. These methods are useful for the detection of RNase L activity, but detailed examination of the steady-state kinetics of catalysis by RNase L requires a quantitative method. This work examines the cleavage by activated RNase L of synthetic oligoribonucleotides containing one or two sites of cleavage as a means of quantitatively determining the steady-state activity of the enzyme. The effect of overall length of the oligoribonucleotide and the sequence of bases at the site of cleavage on the catalytic efficiency and position of cleavage are examined.

MATERIALS AND METHODS
Enzyme-The gene encoding RNase L was cloned from a human kidney cDNA library (Clontech) using polymerase chain reaction and the following primers: 5Ј-GAATTCGGATCCAAGCTTCATATG-GAGAGCAGGGATCATAACAACCCC-3Ј and 5Ј-GAATTCGGATC-CAAGCTTTCAGCACCCAGGGCTGGCCAACCCACT-3Ј. The sequences of the oligonucleotides used for polymerase chain reaction were based on the published sequence of RNase L . The polymerase chain reaction-cloned RNase L gene was completely sequenced using dideoxy methodology, which confirmed the published sequence with the exception of one amino acid. The deposited sequence (GenBank accession number L10381) has been corrected.
RNase L was overproduced in insect cell line Tn5B1-4 using a baculoviral vector. Recombinant virus was produced using the Baculogold system (Pharmingen, San Diego, CA) according to the supplier's instructions. Tn5B1-4 cells in spinner flasks in Ex-Cell 401 medium (JRH Bioscience, Lenexa, KS) at 1-2 ϫ 10 6 cells/ml at 27°C were infected at a multiplicity of infection of 5 and were harvested 72 h post-infection. The purification of RNase L was accomplished according to published procedures (Dong et al., 1994) with the following exception. Instead of chromatography on Superose 12, the protein was chromatographed on Biospin-6 columns (Bio-Rad) that had been equilibrated with 25 mM Tris, pH 7.5, 100 mM KCl, and 5.8 mM magnesium acetate. Approximately 4 mg of RNase L were purified from 3.7 ϫ 10 9 cells. The RNase L was ϳ90% pure as judged with SDS-polyacrylamide gels and Coomassie Blue staining.
Protein concentrations were determined using amino acid analysis. Attempts to sequence the RNase L with automated Edman degradation were unsuccessful, suggesting that the N terminus of the enzyme was blocked. Sequencing of peptides generated by trypsin-catalyzed hydrolysis confirmed the presence of the published amino acid sequence.
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
RNA Synthesis-Oligoribonucleotide substrates and 2Ј,5Ј-linked oligoadenylate activators were synthesized by Midland Certified Reagent Company (Midland, TX). Because of the high content of cytosines of the substrates, a 30-h deprotection in tetrabutyl ammonium fluoride was required. RNA substrates were purified by electrophoresis on 20% polyacrylamide, 7 M urea/TBE (90 mM Tris borate and 1 mM EDTA) sequencing gels. Bands were located with UV shadowing and excised from the gel with a scalpel. The gel pieces were crushed with a stirring rod, and the oligoribonucleotide was eluted by soaking in 300 mM sodium acetate, pH 5. The oligoribonucleotide was then precipitated with the addition of 5 volumes of ethanol, centrifuged, and resuspended in 10 mM HEPES, pH 7.0. The C 7 oligoribonucleotide was purified with C-18 Sep-pak cartridges (Waters) because the oligoribonucleotide remained soluble in ethanol. Concentrations of oligoribonucleotides were determined from absorbance spectra using molar extinction coefficients calculated from the sequences. The activator used in the present study, 5Ј-monophosphate 2Ј,5Ј adenosine trimer (p2,5A 3 ), 1 was purified using high pressure liquid chromatography on a C-18 column.
Steady-state Kinetics of Cleavage-Reactions contained 11.5 mM HEPES, 104 mM KCl, 5.8 mM magnesium acetate, 5 mM dithiothreitol, 50 -2000 nM oligoribonucleotide, 0.2% polyethylene glycol 8000, 1.2 mM ATP, 0 -800 nM p2,5A 3 , and 0 -600 pM RNase L in a total volume of 50 l at a pH of 7.6 and at 30°C. To avoid the loss of enzyme activity due to protein adsorption to the surface of the reaction tube, reactions were carried out in polypropylene microfuge tubes that had been incubated for several hours with 1% polyethylene glycol 20,000 in RNase-free H 2 O and then blown dry and incubated at 65°C for 30 min. RNase L was preactivated by incubating enzyme (300 pM to 2.5 nM) and 500 or 800 nM p2,5A 3 on ice for 30 min in the same buffer as in the cleavage reaction except that substrate and ATP were omitted. Reactions were initiated by the addition of activated enzyme and were quenched after the indicated reaction times by the addition of an aliquot (8 l) to an equal volume of formamide gel load buffer containing 10 mM EDTA.
For product analysis, an aliquot of the quenched solution was electrophoresed on 20 or 25% acrylamide/7 M urea/TBE sequencing gels. The presence of HEPES in the reaction buffer caused smearing of the bands on 20% acrylamide gels in the region of the gel where oligomers of lengths of 8 -10 bases migrate. The bands corresponding to substrate and product were quantified with the use of a PhosphorImager (Molecular Dynamics). Conversion of the raw PhosphorImager counts to the concentration of product was accomplished by integrating the product band as a fraction of the total of the counts in the substrate and product bands and multiplying by the initial substrate concentration, thus eliminating errors due to loading different volumes of the samples on the gel.
Cleavage of substrate C 11 UUC 7 by RNase Phy M (Sigma) was carried out according to the supplier's instructions using 3 units of enzyme in a 30-l reaction. Base-catalyzed hydrolysis of oligoribonucleotides was accomplished by incubating the 5Ј-labeled oligoribonucleotide in 100 mM Na 2 CO 4 , pH 9, at 85°C for 1 h.
Data Analysis-All nonlinear regression calculations were performed as described previously (Carroll et al., 1993). Rate saturation data were fit directly to the Michaelis-Menten equation to determine k cat and K m . The values for k cat reported in Table I were calculated based on the concentrations of monomer in the reactions, assuming that dimer formation was completed during the preincubation with activator.

Cleavage of Oligoribonucleotides by RNase L-
The sequence of the substrate C 11 UUC 7 was based on studies of the cleavage of viral mRNAs by RNase L that demonstrated a preference for cleavage after UU and UA sequences (Wreschner et al., 1981;Floyd-Smith et al., 1981). For the present study the sequence flanking the cleavage dyad included only cytosines to ensure the lack of self-complementarity. The addition of activator p2,5A 3 to a reaction containing 5Ј-32 P-C 11 UUC 7 and RNase L led to the formation of one major radiolabeled product band that migrated on a 20% acrylamide/7 M urea/TBE gel at approximately the position of a 12-mer (Fig. 1, lane 7). In reaction time courses, the intensity of the PhosphorImage of the radiolabeled product band increased with longer reaction time. In the absence of either activator or RNase L, no product band was detected (Fig. 1, lanes 5 and 6, respectively). The addition of a higher concentration of RNase L or an increase in the reaction time resulted in the appearance of another product band migrating one nucleotide shorter on the gel, after almost all of the substrate had first been converted to the initial product (Fig. 1, lane 8).
Site of Cleavage-Comparison of the position on the gel of the product band corresponding to the first cleavage reaction with substrate C 11 UUC 7 against synthetic oligoribonucleotide markers and against the products resulting from the cleavage of 5Ј-32 P-C 11 UUC 7 with RNase Phy M is also shown in Fig. 1. RNase Phy M catalyzes the cleavage of RNA preferentially to the 3Ј side of U and A generating RNA products with a terminal 3Ј phosphate (Donis-Keller, 1980). Activated RNase L generates on cleavage of C 11 UUC 7 a major product that migrates to the same position as the larger of the two major products generated by RNase Phy M (Fig. 1, lanes 8 and 9). Because this product migrates between 5Ј-32 P-C 11 U (Fig. 1, lane 10) and 5Ј-32 P-C 11 UU (Fig. 1, lane 11), it most likely corresponds to 5Ј-32 P-C 11 UU with a 3Ј phosphate.
Further evidence to pinpoint the site of cleavage of C 11 UUC 7 comes from comparison of base-catalyzed hydrolysis of 5Ј-32 Plabeled oligoribonucleotides. Base-catalyzed hydrolysis of 5Ј-32 P-C 11 UUC 7 and of 5Ј-32 P-C 11 (2Ј-fluoroU)UC 7 creates a ladder of 3Јphosphate-terminated oligos (Fig. 1, lanes 1 and 2). The hydrolysis product that is present in the base-catalyzed hydrolysis of C 11 UUC 7 but is absent in the hydrolysis of C 11 (2Ј-fluoroU)UC 7 corresponds to C 11 U(3Ј-p). Because the major cleavage product of C 11 UUC 7 by RNase L is one nucleotide longer than this position, the product corresponds to C 11 UU(3Ј-p).
That a single cleavage event leads to the production of 5Ј end-labeled C 11 UUp is demonstrated by examination of the reaction products from RNase L-catalyzed cleavage of 3Ј endlabeled C 11 UUC 7 cytidine 3Ј,5Ј-bisphosphate (Fig. 1, lane 14). A single radiolabeled product is formed that migrates to the same position as the smaller of the two products generated by cleavage of C 11 UUC 7 cytidine 3Ј,5Ј-bisphosphate by RNase Phy M (Fig. 1, lane 15). This radiolabeled product migrates with the same mobility as authentic C 7 cytidine 3Ј,5Ј-bisphosphate produced by RNA ligase-catalyzed ligation of C 7 and cytidine 3Ј,5Јbisphosphate (Fig. 1, lane 13).
The lengths of the 5Ј cleavage products of the other sub- C 11 UUC 7 3.4 180 1.9 ϫ 10 7 C 5 UUC 4 7.9 1400 5.6 ϫ 10 6 C 17 UUC 13 3.7 110 3.4 ϫ 10 7 C 11 UUUC 7 7.6 150 5.0 ϫ 10 7 C 11 UC 8 0.54 670 8.0 ϫ 10 5 C 11 UAC 7 1.0 160 6.1 ϫ 10 6 C 11 UGC 7 1.9 140 1.4 ϫ 10 7 C 11 AAC 7 0.7 500 1.4 ϫ 10 6 C 11 AUC 7 1.1 175 6.2 ϫ 10 6 strates used in this study are determined by comparison with the product from the cleavage of C 11 UUC 7 and are shown in schematic form in Fig. 2. Substrates containing the dyad sequence UU with overall lengths of 11, 20, and 32 nucleotides are cleaved in the presence of activated RNase L at the same position relative to the UU dyad. The addition of a third U, as in substrate C 11 UUUC 7 , does not change the primary cleavage site. Activated RNase L catalyzes the cleavage of C 11 AAC 7 and C 11 AUC 7 between the nucleotides of the dyads. Activated RNase L catalyzes the cleavage of C 11 UC 8 and C 11 UGC 7 in two positions. Initial Velocity-A lag in product formation is observed in reaction time courses that are initiated by the addition of RNase L to a solution containing activator, p2,5A 3 , and substrate, as shown in Fig. 3. Reaction time courses that show a linear increase in product with increasing reaction time are obtained in reactions that are initiated by the addition of RNase L that is preactivated by incubation in the presence of p2,5A 3 . Control experiments demonstrate that preincubation of 2.5 nM RNase L with p2,5A 3 (800 nM) on ice for 30 min in reaction buffer (11.5 mM HEPES, pH 7.5, 104 mM KCl, 5.8 mM MgOAc 2 , 5 mM dithiothreitol) containing 0.2% polyethylene glycol 8000 result in the maximal rate of reaction. Further control experiments show that enzyme activity in the activated stock solution is stable for at least 1 h when stored on ice (data not shown).
Steady-state Kinetics-The rate of cleavage of C 11 UUC 7 as a function of the concentration of enzyme is shown in Fig. 4. For this experiment the enzyme was preactivated by preincubation at 2.5 nM in the presence of 800 nM p2,5A 3 . The plot of the rate of cleavage of substrate, C 11 UUC 7 , as a function of the concentration of preactivated RNase L in the reaction, is linear over the enzyme concentration range tested (100 -600 pM). The rates of the individual reactions are linear, indicating that the enzymatic activity in the reaction is constant over the reaction time observed (2.5 min).
As shown in Fig. 5, the rate of cleavage of C 11 UUC 7 increases with increasing concentrations of p2,5A 3 and approaches a maximal rate at about 10 nM with half-maximal activation at a concentration of p2,5A 3 of 1 nM. For this experiment reactions were initiated by the addition of RNase L that had been incubated with the same concentration of p2,5A 3 that was included in the subsequent cleavage reaction. Reaction time courses were monitored for product formation for 2.5 min. Higher concentrations of p2,5A 3 , up to 1 M, gave the same maximal reaction rate.
Effect of the RNA Sequence on the Rate of Cleavage-For substrates that are cleaved at more than one position, all of the radiolabeled reaction products have been added together to determine the total rate of cleavage of the substrate. RNase L catalyzes the cleavage of oligoribonucleotides containing either 2 or 3 sequential Us with efficiencies of catalysis that are approximately 20-and 50-fold higher, respectively, than the efficiency with which the enzyme cleaves a substrate containing a single U. The increased efficiencies of catalysis are due to increases in k cat and decreases in K m . RNase L catalyzes the cleavage of C 11 UGC 7 with the same efficiency that it cleaves C 11 UUC 7 . RNase L is also capable of cleaving oligoribonucleotides with a UA or an AU dyad with efficiencies that are approximately a factor of 3 less than the efficiency of cleavage of C 11 UUC 7 . RNase L catalyzes the cleavage of oligoribonucleotides containing an AA dyad with ϳ10-fold lower efficiency than the enzyme cleaves C 11 UUC 7 . DISCUSSION In the presence of an activator, RNase L is capable of efficiently catalyzing the cleavage of oligoribonucleotide substrates containing dyads of bases other than cytosine. The ability of RNase L to cleave synthetic RNA substrates has been utilized to develop a quantitative assay for enzyme activity.
The precise position of the site(s) of cleavage of the RNA substrates is determined at least in part by the sequence of the bases surrounding the cleavage site. For the substrates containing dyads of UU or UA, the primary cleavage site is 3Ј of the second nucleotide of the dyad. Substrates containing either an AA or AU dyad are cleaved between the nucleotides of the dyad. Substrates with a UG or a single U are cleaved 3Ј of the U and 3Ј of the next base to the 3Ј side of the U.
For most of the experiments, RNase L and the activator p2,5A 3 were preincubated on ice prior to initiating subsequent cleavage reactions. The preactivation of the enzyme was carried out to eliminate the possibility of any rate associated with the activation of RNase L from contributing to the observed rate of cleavage of substrates. The subsequent reaction time courses displayed a linear increase in the amount of product and a replot of the rate of the reactions, as a function of enzyme concentration, was also linear over the range of enzyme concentrations tested. These conditions are satisfied to allow the use of steady-state equations to determine the kinetic parameters, k cat and K m . Reactions that were initiated by the addition of RNase L that had not been preincubated with activator resulted in time courses displaying a lag in the first few minutes of the reaction (Fig. 3). One possible reason for the lag is a slow formation of active dimerized enzyme on the time scale of the reactions. A more detailed investigation of the kinetics of dimerization of RNase L and the effect of dimerization on enzyme activity is currently under way.
The efficiency with which activated RNase L catalyzes the cleavage of a substrate depends on the length of the oligoribonucleotide and the number of non-C bases around the cleavage site. The efficiency of cleavage of the oligoribonucleotide substrates containing one UU dyad increases by a factor of 3 with an increase in the length of the substrate from 11 to 20 nucleotides. When the length of the substrate is increased from 20 to 30 nucleotides, a further 2-fold increase in efficiency of cleavage is found. The increases in efficiency of catalysis for cleavage of the longer oligoribonucleotides are primarily due to decreases in the values of K m . Activated RNase L cleaves substrates containing a UU dyad approximately 20-fold more efficiently than a substrate with a single U. A substrate with three consecutive Us is cleaved 2.3 times more efficiently than the substrate with the UU dyad. The increases in the efficiency of cleavage are primarily due to increases in the rate of catalysis. Thus, for a given length, the information determining the rate of cleavage of an oligoribonucleotide containing two or more non-C bases primarily resides in two bases of the substrate. Substrates containing the dyad sequences UU, UA, or UG are cleaved by RNase L with high efficiency relative to the substrate containing a single U, in agreement with published reports of the cleavage of viral RNAs (Wreschner et al., 1981;Floyd-Smith et al., 1981).
In vivo, secondary structural characteristics associated with individual RNA sequences may mask potential cleavage sites or possibly enhance the efficiency of cleavage of specific sites. However, the value of k cat (7.6 s Ϫ1 ; see Table I) for the cleavage of C 11 UUUC 7 is comparable with the rate of cleavage of a FIG. 4. Reaction rate as a function of the concentration of RNase L. Enzyme (2.5 nM) was incubated with p2,5A 3 (800 nM) in reaction buffer in the absence of substrate for 30 min on ice. The preactivated enzyme was used to initiate reactions at the final enzyme concentrations as indicated. The amount of product formed was monitored as described under "Materials and Methods." The line drawn represents a linear fit to the data through the origin.
FIG. 5. Activation of RNase L as a function of concentration of p2,5A 3 . RNase L (500 pM) was incubated in the presence of p2,5A 3 (0.5-20 nM) in reaction buffer on ice for 30 min. The preactivated enzyme was then used to initiate cleavage reactions containing the same concentration of p2,5A 3 that was used in the preincubation and that included 1500 nM C 11 UUC 7 as substrate. Reaction time courses were monitored for 2 min, and the rate of each reaction was determined relative to the rate of the reaction that included 15 nM p2,5A 3 . The line represents a saturation curve as dictated by the data points. specific mRNA catalyzed by RNase L that has been activated by a 2Ј,5Ј adenosine oligomeric DNA chimera (7 s Ϫ1 ; Maitra et al., 1995).
Recent investigations of the quaternary structure of the activated RNase L have indicated that the enzyme is dimerized in the presence of activator Cole et al., 1996). The existence of a quantitative assay for RNase L activity will make possible the determination of the dynamic relationship between the oligomeric/liganded states of the activated complex and its catalytic activity.