Activation of RNase L by 2',5'-Oligoadenylates
KINETIC CHARACTERIZATION*

(Received for publication, February 26, 1997)

Steven S. Carroll , James L. Cole , Tracy Viscount , James Geib , John Gehman and Lawrence C. Kuo

From the Department of Antiviral Research, Merck Research Laboratories, West Point, Pennsylvania 19486

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
REFERENCES


ABSTRACT

Ribonuclease L (RNase L), the 2',5'-oligoadenylate-dependent ribonuclease, is one of the cellular antiviral systems with enhanced activity in the presence of interferon. A reaction scheme has been developed to model the sequence of steps necessary for the activation of RNase L (Cole, J. L., Carroll, S. S., Blue, E. S., Viscount, T., and Kuo, L. C. (1997) J. Biol. Chem. 272, 19187-19192). The model comprises three sequential binding steps: the binding of activator to enzyme monomer, the subsequent dimerization of the activated monomer to form the active enzyme dimer, followed by the binding of substrate prior to catalysis. The model is used to evaluate the activation of RNase L by several synthetic analogs of the native activator. The 5'-phosphate of the activator has been determined to be an important structural determinant for the efficient activation of RNase L, and its loss caused a loss of activator affinity of 2-3 orders of magnitude. The length of activator is not an important determinant of activator potency for the activator analogs examined. The specific activity of the enzyme under conditions of saturation of activator binding and complete dimerization of the activated monomers varies only by about a factor of 3 for the activators examined, indicating that once dimerized in the presence of any of these activators, the enzyme exhibits a similar catalytic activity.


INTRODUCTION

Interferons induce the expression of 2',5'-oligoadenylate synthetases (1), which, in the presence of double-stranded RNA, catalyze the formation from ATP of short oligomers of adenosine with 2' to 5' phosphodiester linkages. 2',5'-oligoadenylates of three nucleotides in length or greater bind to and activate RNase L which can then degrade viral and cellular mRNAs, leading to an inhibition of protein synthesis in virally infected cells (2, 3).

RNase L activity has been shown to be an important antiviral defense mechanism. Overexpression of 2',5'-oligoadenylate synthetase leads to the inhibition of picornavirus replication (4, 5). Introduction of an inactive mutant of RNase L causes an increased susceptibility to infection by picornavirus and a loss of the inhibition of cell growth caused by interferon treatment (6). Transfection of 2',5'-oligoadenylate into cells (2) or the expression of 2',5'-oligoadenylate synthetase (5) has been found to cause growth arrest, suggesting a role for the RNase L system in the regulation of cell growth.

The activation of RNase L on binding of 2',5'-oligoadenylates is correlated with the dimerization of enzyme monomers. Dimerization has been detected with chemical cross-linking and gel filtration techniques (7) and with the use of analytical ultracentrifugation (8). Activator-induced dimerization occurs with a stoichiometry of one activator/RNase L subunit (8). The extent of dimerization correlates with the fraction of activated enzyme, suggesting that most if not all of the enzymatic activity results from the dimeric form of the enzyme (8).

A quantitative description of the relationship among activator concentration, enzyme concentration, and enzymatic activity is required to make valid comparisons of the potency of activation by different analogs of the native activator, ppp-2',5'-A3.1 This work extends the model (9) describing the dimerization of the enzyme on binding of the activator to include the binding of substrate. The activation of RNase L by analogs of the native activator is determined in terms of the activation model. Comparisons are made between the activation parameters as determined from the rates of the RNase L-catalyzed reactions in this work and the equilibrium measurements of the binding of the activator and the dimerization of enzyme monomers as detected with sedimentation equilibrium and fluorescence anisotropy in the accompanying report (9).


EXPERIMENTAL PROCEDURES

Materials

T4 polynucleotide kinase was obtained from U. S. Biochemical Corp. [gamma -32P]ATP (6,000 Ci/mmol) was purchased from NEN Life Science Products.

Enzyme

RNase L was expressed and purified as described previously (10, 11). The concentration of enzyme was determined with amino acid analysis or with absorbance at 280 nm using an extinction coefficient of 84,000 M-1 cm-1 (8).

Substrates

Substrates C11UC8, C11U2C7, and C11U3C7 were prepared and quantified as described previously (11). The substrates were radiolabeled at the 5'-terminus with [gamma -32P]ATP and polynucleotide kinase as described previously (11).

Activators

Activators of RNase L were prepared synthetically via phosphoramidite chemistry either by Midland Certified Reagent Co. (Midland, TX) or by NAPS GmbH (Goettingen, Germany). The activators were purified with reverse-phase high performance liquid chromatography on C18 columns (Vydac) using a gradient of acetonitrile in triethylammonium acetate buffer, pH 6.5. The concentrations of stock solutions of activators were determined by absorbance at 260 nm using extinction coefficients of 45,900 M-1 cm-1 for trimeric activators and 61,200 M-1 cm-1 for tetrameric activators.

Reaction Kinetics

Reactions were initiated, unless otherwise indicated, by the addition of an aliquot of 10-fold concentrated substrate (5 µl) to a solution (45 µl) of enzyme which had been preincubated with activator at various concentrations for 20 min at 22 °C in reaction buffer (11.5 mM HEPES, pH 7.5, 104 mM KCl, 5.8 mM magnesium acetate, 1.2 mM ATP, 0.2% polyethylene glycol, 8000 MW, and 5 mM dithiothreitol). Under these conditions the enzyme was maximally activated, and enzyme activity was stable for at least 60 min. Reaction time courses were followed by quenching an aliquot (5 µl) of the reaction with an equal volume of electrophoresis gel load buffer containing 90% formamide and 10 mM EDTA. Substrates and products were separated with the use of denaturing 20% acrylamide, 7 M urea gels and were quantified with the use of a PhosphorImager (Molecular Dynamics).

Determination of Ka, the Dissociation Constant for Activator

Enzyme (0.3 nM in terms of enzyme monomers) was preincubated with different fixed concentrations of activator (1 nM-15 µM), and reaction rates were determined from reaction time courses as described. The rate of the reaction as a function of the activator concentration was fit to Equation 7 (below) with Kd fixed at 3 nM, and Ks fixed at 200 nM to determine an initial value for Ka. After values for Kd and Ks were determined experimentally using the initial value for Ka, the data were fit again to determine the final value for Ka. The experimentally determined value for Kd was then checked using the final Ka value.

Determination of Specific Activity as a Function of Enzyme Concentration

Enzyme (50 pM-8 nM in monomers) was preincubated with a near saturating concentration of activator using the initial value for Ka determined as described above. Cleavage reactions were then initiated with the addition of substrate C11U2C7 to a final concentration of either 1.5 or 2.5 µM, and the rate of cleavage of the substrate was determined. The plot of specific activity versus [RNase L] was fit to Equation 7 (divided by the concentration of enzyme monomers to convert the units to those of the specific activity determination) using the initial value of Ka for the given activator and the value of Ks determined as described below to determine the value of the dimerization equilibrium constant, Kd. The enzyme specific activity as a function of enzyme concentration was also determined in reactions containing enzyme monomers at concentrations from 0.5 to 20 nM in reactions including 2.5 µM C11UC8 as substrate.

Determination of Km for Substrate C11U2C7

Reactions contained RNase L at a final concentration of 0.3 nM, either 4 or 500 nM p-2',5'-A3 and from 50 to 1,500 nM C11U2C7 in enzyme reaction buffer. Reactions were initiated by the addition of substrate to preactivated enzyme as described. Reaction rate data were fit directly to the Michaelis-Menten equation to determine kcat and Km.

Determination of Ks for Substrates C11U2C7 and C11UC8

The values for Ki for substrate C11U2C7 and for substrate C11UC8 as inhibitors of the cleavage of substrate C11U3C7 were determined in reactions containing 0.3 nM RNase L and a near saturating concentration of activator in reaction buffer with 5'-32P-C11U3C7 as the radiolabeled substrate at final concentrations of 125-600 nM. Reactions also contained either C11U2C7 (0-600 nM) or C11UC8 (0-1.8 µM). Reaction rate data were fit to a competitive mechanism. Replots of the slopes of the double-reciprocal plot were used to determine Ki. Under these conditions, the competitive inhibition constant, Ki, is equal to the binding constant for substrate, Ks.

Reequilibration of Enzyme Activity upon Dilution of the Enzyme-Activator Complex

The effect of the dilution of enzyme on the rate of subsequent RNA cleavage reactions was determined with preincubation of RNase L (1 nM) and p-2',5'-A3 (500 nM) in reaction buffer for 20 min. Reactions were then initiated by the addition of an aliquot (5 µl) of the preincubation solution in an RNA cleavage reaction (total volume 200 µl) such that the concentration of p-2',5'-A3 was maintained at 500 nM, but the enzyme was diluted to 25 pM. Control reactions included preincubation of 27.5 pM RNase L and 500 nM p-2',5'-A3 and initiation of the reaction by addition of substrate such that the enzyme concentration decreased only by 10%. Reaction time courses were monitored as described above. The effect of dilution of the concentration of activator on the rate of subsequent cleavage reaction was determined by preincubation of RNase L (20 nM) and p-2',5'-A3 (20 nM) in reaction buffer for 20 min followed by a 20-fold dilution into an RNA cleavage reaction such that the concentration of activator was either decreased 20-fold or was maintained at the same concentration as in the preincubation. Reaction time courses were then monitored. Control reactions in which the enzyme concentration was either diluted 20-fold or 1.1-fold to a final concentration of 1 nM did not display any difference in the rate of cleavage, indicating that no detectable reequilibration due to dilution of the enzyme occurred under these conditions.

Derivation of the Rate Equation Dictated by the Activation Model

The equilibrium dissociation constants describing the three equilibria included in the model for activation of RNase L (Scheme I under "Results") may be written as follows,
K<SUB>a</SUB>=[E][<UP>A</UP>] / [E<UP>A</UP>] <UP>or</UP> [E<UP>A</UP>]=[E][<UP>A</UP>] / K<SUB>a</SUB> (Eq. 1)
K<SUB>d</SUB>=[E<UP>A</UP>]<SUP>2</SUP> / [E<SUB>2</SUB><UP>A</UP><SUB>2</SUB>] <UP>or</UP> [E<SUB>2</SUB><UP>A</UP><SUB>2</SUB>]=[E<UP>A</UP>]<SUP>2</SUP>/K<SUB>d</SUB> (Eq. 2)
   K<SUB>s</SUB>=[E<SUB>2</SUB><UP>A</UP><SUB>2</SUB>][<UP>S</UP>] / [E<SUB>2</SUB><UP>A</UP><SUB>2</SUB> · <UP>S</UP>] <UP>or</UP> [E<SUB>2</SUB><UP>A</UP><SUB>2</SUB> · <UP>S</UP>]=[E<SUB>2</SUB><UP>A</UP><SUB>2</SUB>][<UP>S</UP>] / K<SUB>s</SUB> (Eq. 3)
where E is enzyme monomer, A is activator, EA is enzyme with activator bound, E2A2 is the active dimer with two activator molecules bound, S is the substrate, and E2A2·S is the active dimer with substrate bound. The equation describing the total concentration of enzyme species is given by,
[E]<SUB><UP>o</UP></SUB>=[E]+[E<UP>A</UP>]+2*[E<SUB>2</SUB><UP>A</UP><SUB>2</SUB>]+2*[E<SUB>2</SUB><UP>A</UP><SUB>2</SUB> · <UP>S</UP>] (Eq. 4)
where [E]o is the original concentration of enzyme monomers, and [E] is the concentration of monomers at equilibrium. Rearranging Equations 1-3 to solve for [EA], [E2A2], and [E2A2·S] and substituting into Equation 4 followed by rearrangement gives a quadratic equation in terms of [E]. A solution is reached when:
[E]=[<UP>−</UP>b±(b<SUP>2</SUP>−4ac)<SUP>0.5</SUP>] / 2a (Eq. 5)
where: a is {2[A] / KaKd}{1 + [S] / Ks}, b is {1 + Ka / [A]}, and c is {-[E]oKa / [A]}.

In practice only the positive operator in Equation 5 gives a physically meaningful result. The rate equation is determined by substituting Equation 5 into Equation 1 to yield a description of [EA] which may then be substituted into Equation 2 to give an equation describing [E2A2]. The reaction rate will be proportional to the total concentration of dimeric species,
v=2v<SUB><UP>o</UP></SUB>{[E<SUB>2</SUB><UP>A</UP><SUB>2</SUB>]+[E<SUB>2</SUB><UP>A</UP><SUB>2</SUB> · <UP>S</UP>]}=2v<SUB><UP>o</UP></SUB>{1+[<UP>S</UP>] / K<SUB>s</SUB>}{[E<SUB>2</SUB><UP>A</UP><SUB>2</SUB>]} (Eq. 6)
where vo is the specific activity of the activated dimeric enzyme under the defined reaction conditions written in terms of [E]o. Therefore the rate equation is:
v=2v<SUB><UP>o</UP></SUB>{1+[<UP>S</UP>] / K<SUB>s</SUB>}{{[<UP>−</UP>b±(b<SUP>2</SUP>−4ac)<SUP>0.5</SUP>] / 2a}{[<UP>A</UP>] / K<SUB>a</SUB>]}}<SUP>2</SUP> / K<SUB>d</SUB> (Eq. 7)
with a, b, and c as described for Equation 5.

Data Analysis

Reaction rate data were fit to Equation 7 in which the concentration of active dimeric RNase L at equilibrium is expressed in terms of the initial concentration of enzyme monomers, the concentration of substrate, the concentration of activator in the reaction, and the equilibrium dissociation constants associated with the model of activation which is presented under "Results." Fitting of the data to Equation 7 was carried out using the commercial fitting program, Kaleidagraph, which is based on the algorithm of Marquardt (12).


RESULTS

Enzyme Activation

RNase L activity has been monitored by following the cleavage of radiolabeled synthetic oligoribonucleotides after separation of substrate and product on denaturing polyacrylamide gels, as described (11). The activation of RNase L by 2,5A analogs is allowed to reach equilibrium during a preincubation step that included activator and enzyme in reaction buffer at 22 °C for at least 20 min. Control reactions indicate that 20 min is a sufficient time to allow the association of activator and enzyme and the formation of enzyme dimers to reach equilibrium. The activity of the preactivated enzyme stock is stable for at least 60 min. Substrate is added to the enzyme-activator solution in a volume equal to 0.10 of the final reaction volume to initiate the reaction. Addition of the substrate in a small volume is designed to minimize the effect of volume increase on the equilibria of activator binding and enzyme dimerization.

At low enzyme concentrations (<0.3 nM) after a 20-min preincubation with activator, a lag in the reaction time course is observed lasting ~1 min followed by a linear phase. The lag in the time course is not evident at high enzyme concentrations (>1 nM). Longer preincubation times (up to 1 h) do not eliminate the lag phase at low enzyme concentrations. The lag is consistent with the binding of substrate to dimers which would serve to increase the concentration of enzyme dimers under low enzyme concentrations where a higher fraction of the enzyme exists as activated monomers. Care has been taken to measure reaction rates during the final linear phase of the reaction time course in reactions containing low concentrations of the enzyme.

Reversibility of Activation

The reversibility of the activation process has been investigated with the use of dilution experiments. In one set of experiments, enzyme is preincubated with 20 nM p-2',5'-A3. The concentrations of the enzyme and activator are then decreased 20-fold by dilution, and the reaction time course is monitored. The results shown in Fig. 1 indicate that the activation of the enzyme is reversible as evidenced by a slow decrease in reaction rate after dilution of activator compared with a reaction in which the enzyme was diluted from 20 nM to 1 nM but the activator concentration remained at 20 nM. The observed rate of activity loss is 0.004 min-1, indicating a slow reequilibration under the conditions of the reaction.


Fig. 1. Reequilibration of enzyme activity on dilution of the enzyme-activator complex. Enzyme (1.1 nM, diamond ; or 20 nM, square  or open circle ) was preincubated with activator p-2',5'-A3 (20 nM, square  or open circle ; or 1.1 nM, diamond ) in reaction buffer for 20 min. Cleavage reactions were initiated such that activator was diluted 20-fold (open circle ) or by 10% (diamond ) to a final concentration of 1 nM or was not diluted (square ). Reaction time courses were monitored for product formation. Time courses were fit to a linear increase in product (square  and diamond ) or to the following equation (open circle ), P(t) = [(mi - mf)*e-kt + mf]*t, where P(t) is the product at time t, mi is the initial rate, mf is the final rate, and k is the observed rate of loss of activity. Best fits were obtained with an mi of 1.7 nM/min and k of 0.004 min-1. mf was set to 0.16 nm/min, the rate of the reaction that was preincubated at 1.1 nM p-2',5'-A3.
[View Larger Version of this Image (15K GIF file)]

In another set of experiments enzyme is preactivated at a concentration that is 40-fold (1 nM) higher or at a concentration that was 1.1-fold (27.5 pM) higher than the final reaction concentration. A cleavage reaction is initiated by the addition of substrate to the preincubation containing enzyme at 27.5 pM or by the addition of an aliquot of the 40-fold concentrated enzyme solution to a reaction mixture. In each experiment the concentration of activator, p-2',5'-A3, remains constant (500 nM). As shown in Fig. 2 the reaction in which enzyme is diluted 40-fold initially shows a higher rate than that of a reaction preincubated at 1.1-fold higher than the final concentration, but the rate then decreases to approximately the same rate as the reaction in which enzyme is diluted 1.1-fold, indicating that the specific activity of the enzyme decreases upon 40-fold dilution of the enzyme to a final concentration of 25 pM.


Fig. 2. Reequilibration of enzyme activity on dilution of enzyme. Enzyme (1 nM, bullet , or 27.5 pM, black-square) was preincubated with 500 nM p-2',5'-A3 in reaction buffer. An aliquot of either preincubation was diluted to a final enzyme concentration of 25 pM in reactions containing 1.5 µM C11UC8 as substrate. Reaction time courses were monitored for the formation of product. The reaction in which the enzyme was preincubated at the higher concentration shows an initial rapid formation of product followed by a slower linear phase, indicating that on dilution to a concentration much less than Kd, the enzyme dimers dissociate. The reaction progress curves were fit to a linear increase in product (black-square) or to the same equation as described in the legend to Fig. 1 (bullet ). The curve fit shown is obtained with mi at 0.75 nM/min, k at 0.029 min-1, and mf set to 0.42 nM/min, the rate of the reaction with enzyme preincubated at 27.5 pM.
[View Larger Version of this Image (13K GIF file)]

Model for Activation of RNase L

The following model (Scheme I) has been developed as a basis for the quantitative description of the kinetics of catalysis by RNase L. 
E+<UP>A</UP>   &cjs0420; E<UP>A</UP>           K<SUB>a</SUB>
E<UP>A</UP>+E<UP>A</UP>      &cjs0420; E<SUB>2</SUB><UP>A</UP><SUB>2</SUB>        K<SUB>d</SUB>
E<SUB>2</SUB><UP>A<SUB>2</SUB> + S</UP>     &cjs0420; E<SUB>2</SUB><UP>A</UP><SUB>2</SUB>·<UP>S</UP>   K<SUB>s</SUB>
<UP>S<SC>cheme</SC> I</UP>
E represents enzyme monomer, A is the activator, EA is the activated monomer, E2A2 is the enzyme dimer, and S and E2A2·S are substrate and the enzyme dimer-substrate complex, respectively (9). Ka represents the dissociation constant for activator binding to enzyme monomers, Kd is the dissociation constant for enzyme dimers, and Ks is the dissociation constant for substrate binding to enzyme dimers.

The first two steps of this binding scheme represent a subset of the complete coupled ligand binding-enzyme dimerization scheme (13, 14) and are minimally required to explain the observed dependence of the reaction rate on both activator concentration and enzyme concentration. It is necessary to extend the model developed in the preceding paper (9) by including the binding of substrate because substrate binding influences the equilibrium concentrations of activated monomer (EA) and dimer (E2A2). The model assumes that only activated dimer binds substrate. Under "Experimental Procedures," a derivation is given of the rate equation (Equation 7) dictated by the model in terms of enzyme monomer concentration, activator concentration, substrate concentration, and the dissociation constants Ka, Kd, and Ks.

Saturation of Reaction Rate with Activator

As shown in Fig. 3, the rate of cleavage of the substrate C11U2C7 is saturable with increasing concentrations of p-2',5'-A3. The concentration of p-2',5'-A3 at which half-maximal activation is obtained (AC50) under the reaction conditions is 10 nM. A previous determination of AC50 for p-2',5'-A3 with preincubation of enzyme and activator at 0 °C was 1 nM, indicating a temperature dependence of AC50 (11). The increase in the rate of the reaction with increasing concentrations of activator could be due to a change in kcat or Km or both. The value of Km for substrate C11U2C7 has been determined at a near saturating concentration of activator p-2',5'-A3 (500 nM) and at a subsaturating concentration of p-2',5'-A3 (4 nM). The Km values are 230 nM and 200 nM, respectively, indicating that the concentration of activator does not have a significant effect on the Km for substrate.


Fig. 3. Saturation of the rate of cleavage with increasing concentrations of activator. Enzyme (300 pM in monomers) was preincubated with activator, p-2',5'-A3 (1-100 nM), for 20 min at 22 °C in reaction buffer in a total of 45 µl. Reactions were then initiated by the addition of 5 µl of 5'-32P-C11U2C7 to a final concentration of 1.5 µM. Reaction time courses were monitored. The rate of the reaction as a function of activator concentration was fit to Equation 7 with Kd of 3 nM and Ks of 170 nM to determine a Ka of 11 nM.
[View Larger Version of this Image (13K GIF file)]

Rate saturation data have been fit to Equation 7 to determine Ka, the activator dissociation constant, by first fixing values for Kd and Ks to give an approximate value for Ka. After determining Kd and Ks using near saturating concentrations of activator based on the initial Ka value, a true fit for the rate data with varying concentrations of activator is carried out to determine the value of Ka.

Specific Activity as a Function of Enzyme Concentration

A near saturating concentration of activator is included in the preincubation with enzyme at concentrations that vary over a wide range. Reaction time courses are subsequently measured. A plot of the enzyme specific activity as a function of the concentration of the enzyme in the reaction is shown in Fig. 4 for activation in the presence of 500 nM p-2',5'-A3. The specific activity increases with increasing concentrations of enzyme and reaches a maximum, consistent with the dissociation of enzyme dimers at low enzyme concentrations. The plot of specific activity as a function of enzyme concentration has been fit to a modified form of Equation 7, dividing by the enzyme monomer concentration (to convert rate data to specific activity data), and using the value of Ka determined as described previously, to determine Kd.


Fig. 4. Enzyme specific activity as a function of enzyme concentration. Enzyme (0.05-4.4 nM in monomers) was preincubated with 500 nM p-2',5'-A3 for 20 min at room temperature. Substrate 5'-[32P]C11U2C7 was added to a final concentration of 1.5 µM, and reaction time courses were monitored. Specific activity data as a function of enzyme concentration were fit to Equation 7 divided by enzyme monomer concentration with Ka set to 11 nM and Ks set to 170 nM. Kd was determined to be 2 nM, and the maximal specific activity was determined to be 140 min-1.
[View Larger Version of this Image (14K GIF file)]

Determination of Ki for Substrate C11U2C7

The value of Ks for substrate C11U2C7 has been determined by measuring a Ki for C11U2C7 as an alternate substrate-inhibitor of the cleavage of substrate C11U3C7 in reactions containing from 0 to 1,800 nM C11U2C7, from 125 to 600 nM C11U3C7 and a near saturating concentration of the activator. Reaction rate data are fit to a competitive mechanism, and the values for Ks shown in Table I are determined from a replot of the slopes of the double-reciprocal plots. The values for Ks shown in Table II for substrate C11UC8 as an alternate substrate inhibitor of the cleavage of substrate C11U3C7 have been determined in a similar manner.

Table I. Activation parameters for analogs of 2,5A


Activator Kaa Kda Ksa Maximal specific activityb

nM nM nM min-1
ppp-2',5'-A3 1.0 2 250 101
p-2',5'-A3 11 2 170 140
HO-2',5'-A3 7,700 5 300 121
HO-2',5'-A4 210 6 160 107
p-2',5'-A4 6.6 7 320 131
p-2',5'-A5 15 12c 1,000 150c
HO-2',5'-A3-7HC 820 1.1 500 46

a Binding constants as defined in Scheme I are determined as described under "Experimental Procedures."
b Maximal specific activity is defined as the specific activity under conditions of saturating concentrations of activator and complete dimerization at a given substrate concentration.
c Average of two trials.

Table II. Activation parameters for analogs of 2,5A with substrate C11UC8


Activator Kaa Kda Ksa Maximal specific activity

nM nM nM min-1
p-2',5'-A3 3.7 9.6 665 7.2
HO-2',5'-A3 2,800 24 870 9.6

a Binding constants as defined in Scheme I are determined as described under "Experimental Procedures."
b Maximal specific activity is defined as the specific activity under conditions of saturating concentrations of activator and complete dimerization at a given substrate concentration.

Effect of Substrate Sequence on Ka and Kd

The activation parameters Ka and Kd have been determined for the activators p-2',5'-A3 and HO-2',5'-A3 using C11UC8 as substrate in a manner similar to that used to determine the same parameters with substrate C11U2C7. As shown in Table II, the values of Ka for the activators p-2',5'-A3 and HO-2',5'-A3 determined using C11UC8 as substrate are both ~3-fold lower than the values of Ka determined using C11U2C7 (Table I). The values for Kd determined in reactions included C11UC8 (Table II) in the case of activation by p-2',5'-A3 or by HO-2',5'-A3 are 5-fold higher than the Kd determined in reaction containing substrate C11U2C7 (Table I).

Activation by Analogs of 2',5'A

The activation of RNase L by structural analogs of 2,5A has been analyzed according to Scheme I. The results are shown in Table I. The most potent activator of RNase L examined in this study is a synthetic version of the native activator, ppp-2',5'-A3, which activates the enzyme with a Ka of 1.1 nM. The least potent activator characterized is the HO-2',5'-A3, with a Ka of 7.7 µM. The values determined for the dimer dissociation constant, Kd, also show variation depending on the structure of the activator, although the Kd values show less variation than do the Ka values. The values for Ki which are equal to the equilibrium binding constant for the substrate, Ks, are independent of the nature of the activator with the exception of activation by p-2',5'-A5 which has a 4-fold increased Ks. For 5'-monophosphate activators the optimal length in terms of the lowest value for Ka is four adenosines with slight increases in Ka for activation by the oligonucleotide with five adenosines. The maximal specific activity, the activity of the enzyme under conditions of saturating activator and complete dimerization, does not show a significant dependence on the structure of the activators shown in Table I, varying by 3-fold maximally.


DISCUSSION

Reaction rate kinetics have been used to measure the kinetic parameters governing the activation of RNase L. A model for the activation of the enzyme was developed (9) as the basis for the quantitative evaluation of activation by structural analogs of the native activator. The rates of cleavage reactions catalyzed by RNase L show saturation kinetics as a function of the concentration of the activator in the reaction (Fig. 3). The specific activity of the enzyme decreases at low concentrations of enzyme consistent with the dissociation of active enzyme dimers at low enzyme concentrations (Fig. 4). Both the binding of activator and the dimerization of the enzyme are reversible processes as evidenced by a slow loss of activity on dilution of either the activator below Ka or dilution of the enzyme to a concentration well below Kd. These observations are consistent with the proposed model for activation wherein enzyme monomers first bind activator to yield activated monomers, followed by dimerization of the monomers to give the active dimeric enzyme species (9).

Experimental support for the model comes from sedimentation equilibrium experiments, which fail to detect any dimerization of enzyme monomers in the absence of activator at concentrations up to 18 µM (8). Further support for the proposal that the primary active species is the enzyme dimer comes from the observation that the extent of dimerization of the enzyme at high enzyme concentrations (greater than Kd) is a linear function of the stoichiometry of activator to enzyme monomers. Because the stoichiometry of dimerization closely parallels the stoichiometry of activation, most if not all of the enzyme activity results from the dimeric form of the enzyme (8). The binding of substrate is included in the model for activation since substrate binding can influence the extent of dimerization by binding solely or preferentially to the dimer. The Ki of the substrates used in reactions for determination of Ka and Kd as alternate substrates for the cleavage of the substrate C11U3C7 has been determined experimentally because the Ki of the competitive inhibitor will be equivalent to the binding constant of the substrate Ks.

Prior activation conditions involved preincubation of enzyme and activator at 0 °C (11). Because both Ka and Kd show a temperature dependence, preincubation at 0 °C gives a greater degree of dimerization and higher activity than does preincubation at the same concentrations of activator and enzyme at 22 °C, the conditions used in this work. Further support for a decrease in Ka and Kd with lower temperatures comes from sedimentation equilibrium measurements of enzyme dimerization (9).

A change in the structure of the activator can cause changes in the activation parameters. Three main conclusions may be drawn from the data presented in Table I. First, comparison of activation by ppp-2',5'-A3 to activation by p-2',5'-A3 indicates that loss of the 5'-diphosphate causes a 10-fold increase in Ka and only minor changes in Kd and Ks. Further loss of the 5'-phosphate in the case of HO-2',5'-A3 causes another 700-fold increase in Ka and minor increases in Kd and Ks, indicating that the 5'-phosphate moiety is an important determinant of the potency of activator binding. Second, the values of Ka for the series of activators with a 5'-monophosphate and lengths 3, 4, or 5 adenosines show only minor variations, indicating that the length of these activators is not a strong determinant for potency of activation. However, addition of a fourth base in the 5'-OH analogs is associated with a significant decrease in Ka, from 7.7 µM for the trimer to 210 nM for the tetrameric activator. Third, despite the greater than 3 orders of magnitude variation in Ka, the maximal specific activity of the enzyme, under conditions of saturating concentrations of activator and complete dimerization, does not show a large dependence on the structure of the activator, indicating that once the enzyme has dimerized it retains a similar efficiency of catalysis.

These values of Ka are consistent with activator binding as determined with competitive binding affinity measurements. The binding constant of ppp-2',5'-A3 with the murine enzyme has been reported variously as 40 pM (15) or 1 nM (16), with the differences probably being due to differences in experimental conditions and the inherent difficulties of measuring so tight a binding interaction. Competitive binding measurements with HO-2',5'-A3 indicate a binding constant in the low micromolar range (17), a value consistent with the Ka determined in this work using substrate C11UC8 (2.8 µM, Table II).

The activation parameters Ka, Kd, and Ks have been determined for activation by p-2',5'-A3 and HO-2',5'-A3 in reactions containing two different substrates, C11U2C7 (kcat/Km = 1.9 × 107 M-1 s-1) and C11UC8 (kcat/Km = 8 × 105 M-1 s-1). The value of Ka for each activator decreases by about a factor of 3 in reactions containing the substrate C11UC8 relative to reactions with substrate C11U2C7 (Tables I and II). The values for Kd determined with each activator also show a 5-fold increase in reactions containing substrate C11UC8 relative to the Kd determined in reactions with the substrate C11U2C7. These changes in Ka and Kd with the sequence of the substrate may indicate that the binding of substrate produces an effect on the conformation of the enzyme such that the activation parameters are intrinsically changed slightly from those of the enzyme in the absence of substrate. Comparison of the Kd determined by measurement of reaction kinetics with activation by HO-2',5'-A3, for example (5 and 24 nM for reactions containing C11U2C7 and C11UC8, respectively), with the Kd determined by analytical ultracentrifugation experiments (17 nM; Ref. 9) supports the possibility of a decrease in Kd in the presence of the better substrate, C11U2C7. The small change in the energy of binding (0.6-1 kcal/mol) for the activator with different substrates is not unreasonable.

In general, a good agreement exists between the values for Ka and Kd as determined with enzyme kinetics for HO-2',5'-A3 using substrate C11UC8 with Ka and Kd as determined with the use of sedimentation equilibrium methods (9). The activation by a fluorescent analog of HO-2',5'-A3, HO-2',5'-A3-7HC, described in the previous paper, was also examined. The values for Ka and Kd as determined by examination of enzyme kinetics (0.8 µM, 1.1 nM, respectively, Table I) are in reasonable agreement with Ka and Kd as determined with fluorescence anisotropy (1.8 µM, 1.1 nM, respectively; Ref. 9).

The proposed model for the activation of RNase L takes into account the three equilibria that lead to the formation of the activated complex. The model should prove useful for the quantitative evaluation of the activation of RNase L by different activators.


FOOTNOTES

*   The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1   The abbreviations used are: ppp-2',5'-A3, a trimer of adenosines linked 2' to 5' with a 5'-triphosphate; p-2',5'-A3, a trimer of adenosines linked 2' to 5' with a 5'-monophosphate; 2,5A, an oligomer of adenosine linked via 2',5'-phosphodiester groups; AC50, concentration at which half-maximal activation is obtained under specified reaction conditions; HO-2',5'-A3, a trimer of adenosines linked 2' to 5' with a 5'-hydroxyl; HO-2',5'-A3-7HC, a conjugate of HO-2',5'-A3-2'-NH2 and 7-hydroxycoumarin-3-carboxylic acid, succinimidyl ester.

REFERENCES

  1. Hovanessian, A. G., Brown, R. E., and Kerr, I. M. (1977) Nature 268, 537-539 [Medline] [Order article via Infotrieve]
  2. Hovanessian, A. G., and Wood, J. N. (1980) Virology 101, 81-90 [Medline] [Order article via Infotrieve]
  3. Kerr, I. M., and Brown, R. E. (1978) Proc. Natl. Acad. Sci. U. S. A. 75, 256-260 [Abstract]
  4. Chebath, J., Benech, P., Revel, M., and Vigneron, M. (1987) Nature 330, 587-588 [CrossRef][Medline] [Order article via Infotrieve]
  5. Rysiecki, G., Gewert, D. R., and Williams, B. R. G. (1989) J. Interferon Res. 9, 649-657 [Medline] [Order article via Infotrieve]
  6. Hassel, B. A., Zhou, A., Sotomayor, C., Maran, A., and Silverman, R. H. (1993) EMBO J. 12, 3297-3304 [Abstract]
  7. Dong, B., and Silverman, R. H. (1995) J. Biol. Chem. 270, 4133-4137 [Abstract/Free Full Text]
  8. Cole, J. L., Carroll, S. S., and Kuo, L. C. (1996) J. Biol. Chem. 271, 3979-3981 [Abstract/Free Full Text]
  9. Cole, J. L., Carroll, S. S., Blue, E. S., Viscount, T., and Kuo, L. C. (1997) J. Biol. Chem. 272, 19187-19192 [Abstract/Free Full Text]
  10. Dong, B., Xu, L., Zhou, A., Hassel, B. A., Lee, X., Torrence, P. F., and Silverman, R. H. (1994) J. Biol. Chem. 269, 14153-14158 [Abstract/Free Full Text]
  11. Carroll, S. S., Chen, E., Viscount, T., Geib, J., Sardana, M. K., Gehman, J., and Kuo, L. C. (1996) J. Biol. Chem. 271, 4988-4992 [Abstract/Free Full Text]
  12. Marquardt, D. M. (1963) J. Soc. Indust. Appl. Math. 11, 431-441
  13. Wyman, J., and Gill, S. J. (1990) Binding and Linkage, pp. 203-212, University Science Books, Mill Valley, CA
  14. Wong, I., and Lohman, T. (1995) Methods Enzymol. 259, 95-127 [Medline] [Order article via Infotrieve]
  15. Silverman, R. H., Jung, D. D., Nolan-Sorden, N. L., Dieffenbach, C. W., Kedar, V. P., and SenGupta, D. N. (1988) J. Biol. Chem. 263, 7336-7341 [Abstract/Free Full Text]
  16. Knight, M., Cayley, P. J., Silverman, R. H., Wreschner, D. H., Gilbert, C. S., Brown, R. E., and Kerr, I. M. (1980) Nature 288, 189-192 [Medline] [Order article via Infotrieve]
  17. Kariko, K., Li, S., Sobol, R. W., Suhadolnick, R. J., Charubala, R., and Pfleiderer, W. (1987) Biochemistry 26, 7136-7142 [Medline] [Order article via Infotrieve]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.