HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 3, 1695-1704, January 18, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/3/1695    most recent M109616200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Cherepanov, A. V. Articles by de Vries, S. Search for Related Content PubMed PubMed Citation Articles by Cherepanov, A. V. Articles by de Vries, S. Social Bookmarking                      What's this?

Kinetic Mechanism of the Mg²⁺-dependent Nucleotidyl Transfer Catalyzed by T4 DNA and RNA Ligases^*

Alexei V. Cherepanov and Simon de Vries

From the Kluyver Department of Biotechnology, Delft University of Technology, Julianalaan 67, Delft 2628 BC, The Netherlands

Received for publication, October 4, 2001, and in revised form, October 26, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The Mg²⁺-dependent adenylylation of the T4 DNA and RNA ligases was studied in the absence of a DNA substrate using transient optical absorbance and fluorescence spectroscopy. The concentrations of Mg²⁺, ATP, and pyrophosphate were systematically varied, and the results led to the conclusion that the nucleotidyl transfer proceeds according to a two-metal ion mechanism. According to this mechanism, only the di-magnesium-coordinated form Mg₂ATP⁰ reacts with the enzyme forming the covalent complex E·AMP. The reverse reaction (ATP synthesis) occurs between the mono-magnesium-coordinated pyrophosphate form MgP₂O and the enzyme·MgAMP complex. The nucleotide binding rate decreases in the sequence ATP⁴ > MgATP² > Mg₂ATP⁰, indicating that the formation of the non-covalent enzyme·nucleotide complex is driven by electrostatic interactions. T4 DNA ligase shows notably higher rates of ATP binding and of subsequent adenylylation compared with RNA ligase, in part because it decreases the K_d of Mg²⁺ for the enzyme-bound Mg₂ATP⁰ more than 10-fold. To elucidate the role of Mg²⁺ in the nucleotidyl transfer catalyzed by T4 DNA and RNA ligases, we propose a transition state configuration, in which the catalytic Mg²⁺ ion coordinates to both reacting nucleophiles: the lysyl moiety of the enzyme that forms the phosphoramidate bond and the -bridging oxygen of ATP.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA and RNA ligases from the bacteriophage T4 are ATP-dependent enzymes that catalyze the formation of the phosphodiester bond between the adjacent 3'-OH and 5'-PO₄ ends of two nucleic acid fragments. The first step of catalysis requires a divalent metal cofactor and consists of the binding of ATP with the subsequent formation of the ligase-AMP complex and the release of pyrophosphate (1, 2). This reaction is reversible (3-5). In the bound state the -phosphate of the nucleotide is attached to a conserved lysine residue of the enzyme, forming an (-amino)-linked adenosine monophosphoramidate (6, 7).

Although the T4 DNA ligase (EC 6.5.1.1) and T4 RNA ligase (EC 6.5.1.3) have been first purified more than 30 years ago, relatively few studies have been performed on the interaction between the ligase, ATP, pyrophosphate, and Mg²⁺. In the past, the pyrophosphate exchange reaction catalyzed by these enzymes have been studied, and the apparent K_m values for ATP in the DNA-joining reaction on different DNA substrates have been determined (4, 5, 8, 9). X-ray structures of the related enzymes, T7 DNA ligase and mRNA capping enzyme from Chlorella virus PBCV-1 in complex with the nucleoside triphosphate as well as the enzyme-adenylylate complex of DNA ligase from Chlorella virus PBCV-1 have been solved (10-12). It was shown that the nucleotide is oriented in the binding cleft of the ligase by stacking interactions and by hydrogen bonding with the adenine ring, the ribose moiety, and the -phosphate. However, the position of the metal cofactor in complex with the nucleoside triphosphate and ligase or its stoichiometry remained unknown.

A steady-state kinetic analysis of the nucleotidyl transfer reaction catalyzed by T4 RNA ligase has been performed by previous researchers (13). On the basis of isotope equilibration studies of the Mg²⁺-dependent pyrophosphate exchange reaction, these authors proposed a two-metal ion mechanism in which the di-magnesium-coordinated forms of ATP and pyrophosphate would be the true catalytic substrates.

In this report we present a pre-steady-state kinetic analysis of the interactions of T4 DNA ligase and T4 RNA ligase with ATP, pyrophosphate, and Mg²⁺. By varying the experimental conditions we were able to isolate individual steps of the binding reaction. Using dATP and dCTP instead of ATP we could observe the nucleotide-binding step without the subsequent covalent attachment of the nucleotide to the ligase. In the case of ATP, addition of pyrophosphate to the reaction mixture allowed us to suppress the first-order adenylylation of the enzyme, so that only the binding of ATP was observed. As was previously shown, addition of ATP to the ligase in the absence of Mg²⁺ leads to non-covalent binding of the nucleotide (10). So, we were able to study covalent catalysis separately from nucleotide binding by mixing the pre-formed non-covalent nucleotide-enzyme complex with Mg²⁺.

The dissociation constants of ATP complexes with Mg²⁺, Na1⁺, and Tris⁰ are known (14-17) (see summary in Table III below). Changing the Mg²⁺ concentration at fixed concentrations of enzyme and nucleotide allowed us to determine the binding rates of the different Mg²⁺-coordinated forms of the nucleotide and to obtain information on the stoichiometry of Mg²⁺ in the forward (cleavage of nucleoside triphosphate) and the reverse (ATP synthesis) reaction. Our results indicate that the di-magnesium form, Mg₂ATP⁰, is the true substrate for the adenylylation of T4 DNA ligase and T4 RNA ligase, whereas the MgP₂O form is the true substrate for the ATP synthesis, the latter being opposite to what has been proposed previously (13). In the present report we describe an expanded kinetic scheme of the nucleotidyl transfer catalyzed by T4 DNA ligase and T4 RNA ligase and discuss the mechanism in general terms of the metal-dependent enzymic catalysis of the phosphoryl transfer reactions.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

T4 DNA ligase and T4 RNA ligase were purchased from MBI Fermentas (Vilnius, Lithuania). The protein concentrations were determined using the BCA, protein determination kit (Pierce). Transient state kinetic experiments were performed on the Bio Sequential Stopped-Flow Reaction Analyzer SX-18MV (Applied Photophysics, UK) using an ozone-free 150-watt xenon arc light source.

For the kinetic experiments T4 DNA ligase or T4 RNA ligase were diluted yielding final concentrations after mixing of 2.5 ± 0.1 µM in 50 mM Tris-HCl buffer, pH = 7.8, 1 mM dithioerythritol in the presence of 0.025 mg/ml bovine serum albumin (buffer A).

Two types of stopped-flow experiments have been performed as follows.

Binding of the Nucleoside Triphosphate to the Ligase in the Presence of Mg²⁺ or Mg²⁺ Plus Pyrophosphate-- For these experiments a 5 µM solution of T4 DNA ligase or T4 RNA ligase was prepared in buffer A supplemented with 5 mM MgCl₂. The nucleotide solution was prepared in the same buffer with MgCl₂ at a concentration of ~200 µM. Concentrations of nucleotide were determined optically (with ₂₅₉ = 15.4 mM¹ cm¹ for ATP, ₂₅₉ = 15.2 mM¹ cm¹ for dATP, and ₂₇₁ = 9.3 mM¹ cm¹ for dCTP). Stocks with lower concentration were prepared by 2-fold serial dilutions of a 200 µM stock solution into buffer A, containing 5 mM MgCl₂. To calculate the dilution coefficients with high accuracy, components were weighed on an analytical balance (±0.2-mg precision). Pyrophosphate was added to both the nucleotide stock and the dilution buffer to a final concentration of ~620 µM in case we wanted to suppress the adenylylation of the ligase. All enzyme and nucleotide stocks were stored on ice and used within 1 h after preparation. For the stopped-flow experiments, enzyme and nucleotide stocks were incubated in the syringes at 20 ± 0.05 °C for 5 min and mixed at 1:1 ratio.

Binding of Mg²⁺ or Mg²⁺ Plus Pyrophosphate to the Ligase·ATP Complex-- Ligase solution (5 µM) was prepared in the presence of ~150 µM ATP in buffer A without Mg²⁺. Mg²⁺ or Mg²⁺ and pyrophosphate stocks were prepared separately by serial dilutions as mentioned above. The reaction was started by mixing the ligase solution with the Mg²⁺ solution.

The SX-18MV software package for single-wavelength operation mode was used for the optical measurements. Tryptophan emission was excited at 280 nm and measured as the light passed through a <320-nm cut-off filter. Kinetic traces were obtained by averaging 3-10 shots (protein fluorescence emission) or 10-15 shots in case of optical absorbance at 260 nm.

Numerical solving of kinetic equations, fitting, and minimization was performed in the Mathematica v. 4.0 software package (Wolfram Research), Scientist v. 2.0 (MicroMath), and Igor Pro v. 3.15 (Wavemetrics). Error estimates for the data values in graphs and tables represent 95% confidence intervals calculated using the Student distribution function.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Characterization of the Nucleotide Binding-- In the first set of experiments (see "Experimental Procedures"), the enzyme was mixed with different amounts of nucleotide in the presence of Mg²⁺, and changes of the tryptophan fluorescence emission and the optical absorbance at 260 nm were followed (Figs. 1 and 2). As expected from the structures of the related nucleotidyltransferases (10-12, 18), we observed a decrease of absorbance at 260 nm (Fig. 2), which corresponds to the -stacking between the adenine ring and the aromatic residue in the active site of the ligase. In addition, the fluorescence emission of both enzymes was partially quenched, implying that the optically active tryptophan residue is located in the ligase active site. Kinetic traces indicate that the binding of the nucleotide is a biphasic process. In the case of T4 DNA ligase, a decrease of emission is followed by its increase, and the final emission is lower than that of the free enzyme (Fig. 1A, Table I). In the case of T4 RNA ligase, we observed a decrease of the fluorescence emission for both phases (Fig. 1C, Table I). For both enzymes, the rate of the faster phase increases proportionally to the concentration of ATP, corresponding to second order kinetics of the nucleotide binding (Fig. 3, A and B). The rate of the following slower phase, which at low concentrations of ATP is determined by the rate of the initial binding phase, reaches its maximum at ~30 µM ATP (Fig. 3C). To verify that the first phase represents binding and the second represents the formation of the covalent enzyme-AMP complex, we used dCTP and dATP instead of ATP, the former two nucleotides known to be poor substrates in the nucleotidyl transfer reaction (4, 5). From Fig. 4 it follows that the second slower process is absent in the case of dCTP and negligible in the case of dATP, which is expected if this process would correspond to the formation of the covalent enzyme·nucleotide complex. Further evidence for this conclusion was obtained by changing the conditions of the experiment: The ligase was mixed with both ATP and pyrophosphate pre-equilibrated with Mg²⁺. The concentration of pyrophosphate was chosen high enough to drive the nucleotidyl transfer reaction backwards, so that the enzyme would remain in the non-adenylylated form. As a result, we observed only (non-covalent) binding of ATP, whereas the second slower process disappeared (Fig. 1, B and D). The concentration of ATP was varied in a range to maintain the pseudo-first order conditions (keeping the concentration of Mg²⁺ and pyrophosphate fixed). The values of the observed binding rate constants in the presence and in the absence of pyrophosphate were so similar (Fig. 3, A and B), that one may conclude that pyrophosphate and nucleotide do not compete for binding in the active site of the ligase. Taking this into account, the nucleotidyl transfer reaction catalyzed by T4 ligases as observed in our experiments can be sufficiently described by Scheme 1. 

View larger version (35K): [in this window] [in a new window]   Fig. 1.   Changes of protein fluorescence emission upon mixing of T4 DNA ligase (A, B) or T4 RNA ligase (C, D) with ATP (A, C) or ATP and pyrophosphate (B, D) in the presence of Mg2+. Concentrations after mixing 2.5 ± 0.1 µM enzyme (A-D); 0, 2.55, 7.66, 17.81, 26.3, 39.43, 74.68, and 96.67 µM ATP (A); 5.27, 10.63, 21.19, 42.57, 85, 140.7, 344.8 µM ATP and 351.24 µM pyrophosphate (B); 0, 4.84, 9.39, 20.07, 39.63, 74.46, and 100.03 µM ATP (C); 0, 3.13, 6.32, 12.29, 24.95, 50.12, 74.87, 101.01 µM ATP and 362.74 µM pyrophosphate (D). The reaction was carried out in the presence of 5 ± 0.02 mM Mg2+ at 20 ± 0.1 °C and pH = 7.8.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Changes of absorbance at 260 nm upon mixing of T4 DNA ligase (A) or T4 RNA ligase (B) with ATP in the presence of Mg2+. Concentrations after mixing 2.5 ± 0.1 µM enzyme (A, B); 0, 2.55, 7.66, 17.81, 26.3, and 39.43 µM ATP (A); 4.84, 9.39, and 20.07 µM ATP (B). Reaction conditions were as in the legend to Fig. 1.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Dependence of the observed rate constant for ATP binding (k) or adenylylation (k) on the concentration of ATP. A and B, binding of ATP (k). Closed circles, fluorescence emission data; open circles, 260 nm absorbance data; open crossed circles, binding of ATP in the presence of pyrophosphate, fluorescence emission data. C, adenylylation of the enzyme (k). The values were determined by fitting double exponentials to the stopped-flow traces. Fits were obtained from the Scheme 1: T4 DNA ligase: k = 8.7 ± 0.11 × 105 M1 s1, k = 10, 1, or 0.1 s1 (Fits 1-3). T4 RNA ligase: k = 3.09 ± 0.13 × 105 M1 s1, k = 3, 0.3, or 0.03 s1 (Fits 1-3).

View larger version (35K): [in this window] [in a new window]   Fig. 4.   Changes of protein fluorescence emission upon mixing of T4 DNA ligase with dATP (A) or dCTP (B) in the presence of Mg2+. Dependence of the k (C), and the amplitude of fluorescence changes (D) on the concentration of the nucleotide. Concentrations after mixing 2.5 ± 0.1 µM enzyme (A, B); 5.12, 10.16, 20.36, 40.3, 80.37, 162.16, and 325.95 µM dATP (A); 5.31, 10.69, 21.16, 42.04, 83.49, and 166.07 µM dCTP (B). Traces have been offset slightly for clarity. Reaction conditions were as in the legend to Fig. 1. Fits (dotted lines) shown in C were obtained from Scheme 1, taking both k, k = 0. Fitted values are shown in Table II.

SCHEME 1. Mechanism of ATP binding and self-adenylation catalyzed by T4 DNA ligase or T4 DNA ligase. wz are the mean stoichiometries at a given [Mg²⁺], k₁^app, k₁^app, k₂^app, k₂^app are the apparent rate constants, and k₁^obs, k₂^obs are the observed rate constants obtained by fitting a sum of two exponentials to the kinetic traces.

From Figs. 3 and 4 and Table II it can be seen that T4 DNA ligase binds ATP, dATP, and dCTP with similar k rate constants, and that its affinity to the nucleotide is dependent solely on the stability of the enzyme·nucleotide complex. The dissociation rate constant of ATP, k is less than 1 s¹ (taking into account equilibrium competition binding studies (19), this value is even lower, <0.1 s¹), whereas for dATP and dCTP the k values are higher by one or two orders of magnitude, respectively. For comparison, T4 DNA ligase binds ATP approximately three times faster than T4 RNA ligase at the same Mg²⁺ and ATP concentrations, whereas the equilibrium dissociation constants for ATP for these two enzymes are similar (Table II).

Nucleotidyl Transfer Reaction-- Formation of the covalent enzyme-AMP product and the release of the pyrophosphate occur as a slow process, following binding of the ATP coenzyme. The observed rate constant of this process, k, increases with the ATP concentration reaching its maximal value at ~30 µM ATP (Fig. 3C). To separate the nucleotide binding phase from the adenylylation of the ligase, we modified the experimental conditions: Mg²⁺-free ligase was pre-equilibrated with excess ATP so that the enzyme·nucleotide complex would be formed without further adenylylation. Subsequently, this complex was mixed with Mg²⁺, triggering a first order nucleotidyl transfer (Fig. 5). In this set of experiments, similar to Fig. 1, the intensity of the fluorescence emission of T4 DNA ligase increases during the reaction, whereas that of T4 RNA ligase decreases. Because at the [Mg²⁺] used here the binding of Mg²⁺ to the nucleotide occurs significantly faster (Table III), the observed emission changes could result only from the adenylylation of the enzyme. The same can be concluded from the pronounced dependence of k on the pyrophosphate concentration (see below), and on pH (not shown). The k drastically increases at alkaline pH, which is expected, because only the deprotonated lysine residue participates in the nucleophilic substitution (3). In addition, the adenylylation of the enzyme is as fast as that observed in the experiments with the Mg²⁺-pre-equilibrated nucleotides (for example, compare the first and second traces from the top in Fig. 5). The adenylylation occurs without any delay, confirming that the second order binding of Mg²⁺ to the ligase·ATP is, indeed, faster and, therefore, can not be seen as a separate kinetic phase. It also implies that binding of Mg²⁺ to ATP does not require the dissociation of the enzyme·ATP complex. Otherwise, the adenylylation would be limited by k for ATP from the enzyme·ATP complex, which is quite low (<1 s¹, Table II).

View larger version (36K): [in this window] [in a new window]   Fig. 5.   Changes of the fluorescence emission of ligase·ATP complexes upon mixing with Mg2+ and pyrophosphate. For comparison, the top traces in each graph are from Fig. 1, A and C. Pre-formed complexes were prepared by mixing 5 ± 0.2 µM ligase with 153.5 ± 0.08 µM ATP (T4 DNA ligase) or 158.52 ± 0.14 µM ATP (T4 RNA ligase). The reaction was started by 1:1 mixing of the pre-formed non-covalent complex ligase·ATP with a solution containing 10 mM Mg2+ and pyrophosphate. Concentration of pyrophosphate after mixing (from the second top trace down): 0, 1.56, 3.12, 6.25, 12.5, 25, 50, and 100 µM.

To separate the kinetic phases of the forward (cleavage of ATP) and the reverse (ATP synthesis) reaction, different amounts of pyrophosphate pre-equilibrated with Mg²⁺ were mixed with the enzyme·ATP complex. Upon increasing the pyrophosphate concentration, the observed reaction rate increases while the amplitude of the fluorescence decreases (Fig. 5). In fact, the fluorescence jump can be suppressed nearly completely by addition of a 200-fold excess of pyrophosphate over the enzyme. This shows that the equilibrium of the reaction under these conditions is shifted to the non-covalent enzyme·MgATP complex and that most of the enzyme remains in the non-adenylylated form even in the presence of free Mg²⁺. The rate of ATP synthesis catalyzed by both T4 DNA and RNA ligases is proportional to the concentration of pyrophosphate (Figs. 6 and 7).

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Dependence of the k constant on the pyrophosphate concentration, determined at different [Mg2+]. The pre-formed non-covalent complex ligase·ATP was mixed with pyrophosphate and Mg2+. Concentrations after mixing 2.5 ± 0.1 µM enzyme; 0, 0.78, 1.56, 3.12, 6.25, 12.5, 25, 50, 68, 100 µM pyrophosphate at 0.4, 5, or 50 mM Mg2+ (T4 DNA ligase); 0, 1.56, 3.12, 6.25, 12.5, 25, 32.5, 50, 75, 100 µM pyrophosphate at 5 mM Mg2+ and 0, 3.12, 6.25, 12.5, 25, 50, 108, and 200 µM pyrophosphate at 50 mM Mg2+ (T4 RNA ligase). k values were obtained by fitting single exponentials to the kinetic traces. Ligase·ATP complexes were prepared as described in the legend to Fig. 5. Fits were obtained from Scheme 2 using Equation 6. The values of the apparent rate constants k, k are reproduced in Table II.

View larger version (49K): [in this window] [in a new window]   Fig. 7.   Observed rate constant k as a function of both [Mg2+] and [P2O7]. Data points were obtained from the traces shown in Figs. 5 and 6 and similar additional measurements. The fit surfaces were obtained from Scheme 3 and from the values in Tables III and IV.

Taking this into account, the reaction of adenylyl transfer under these experimental conditions can be sufficiently described as follows (see Scheme 2).
SCHEME 2. Mechanism of self-adenylation catalyzed by T4 DNA ligase or RNA ligase.

The corresponding kinetic equation for the formation of the enzyme-adenylylate has the form, dB/dt = k (A₀  B)  k B(B + C₀), where B is the enzyme-adenylylate, A₀ is the initial concentration of the non-covalent complex ligase·Mg_xATP, and C₀ is the initial concentration of pyrophosphate.

After integration we obtain,

(Eq. 1)

where

(Eq. 2)

and

(Eq. 3)

In our experiments A₀ K_d (Table II), therefore, we can apply a pseudo-first order approximation and reduce the expression for B(t) to the following,

(Eq. 4)

where

(Eq. 5)

and

(Eq. 6)

The k value here is the observed rate constant obtained by fitting single exponentials to the kinetic traces.

For comparison, both ligases have very similar apparent pyrophosphate dissociation constants, around 30 µM at 5 mM Mg²⁺, or 600-800 µM at 50 mM Mg²⁺. The individual rate constants, however, differ more than an order of magnitude, with T4 RNA ligase being the slower enzyme (Table II).

Stoichiometry of Mg²⁺ in the Reaction of the Nucleotidyl Transfer-- Covalent binding of AMP by T4 ligases is a reversible enzymic reaction involving ATP, pyrophosphate, and Mg²⁺. In aqueous solution in the presence of Mg²⁺ both ATP and pyrophosphate are known to form different protonated and/or metal-coordinated complexes (Table III). To estimate the binding rate constants for each nucleotide form, we varied [Mg²⁺] at fixed concentrations of enzyme and ATP. As can be seen in Fig. 8 (A and B), an increase of [Mg²⁺] leads to a decrease of the binding rate of the nucleotide. At 50 mM Mg²⁺ both T4 DNA ligase and T4 RNA ligase bind ATP roughly 10- to 20-fold slower than at 0.4 mM Mg²⁺ (Fig. 9A, Table II). This fact implies that the binding rate is determined by electrostatic attraction between the nucleotide and the ligase, in agreement with the previous results on T4 RNA ligase (20). This conclusion is supported by the structures of the related enzymes, which show that the hydrophobic nucleotide-binding pocket is surrounded by a large positively charged area where the DNA/RNA substrate binds (10-12, 18).

View larger version (41K): [in this window] [in a new window]   Fig. 8.   Mg2+-dependence of the nucleotidyl transfer reaction catalyzed by T4 DNA ligase and T4 RNA ligase. A and B, fluorescence emission changes after mixing ligase with ATP and Mg2+. Concentrations after mixing 2.5 ± 0.1 µM enzyme; 77 ± 0.04 µM ATP; 0, 0.6, 1.2, 5, 20, 50 mM Mg2+ (T4 DNA ligase); 0, 0.3, 0.6, 1.2, 3.1, 5, 20, 50, 100 mM Mg2+ (T4 RNA ligase). C and D, kinetic traces obtained after mixing the pre-formed ligase·ATP complexes with Mg2+. Concentration of Mg2+ after mixing, 0, 0.3, 0.6, 1.2, 2.4, 5, 20, 50, 100 mM (T4 DNA ligase); 0, 0.3, 0.6, 3.1, 5, 20, 50, 100 mM (T4 RNA ligase). Ligase·ATP complexes were prepared as described in the legend to Fig. 5.

View larger version (17K): [in this window] [in a new window]   Fig. 9.   Dependence of the rate constants of the nucleotidyl transfer reaction on [Mg2+]. Solid circles, T4 DNA ligase; opened circles, T4 RNA ligase. A, binding of ATP to the T4 DNA and RNA ligase. Fits 1 and 2 are obtained from Scheme 3 and the values in Table IV. Parameter variation evaluation: Kd of the reaction 2  1 + Mg2+ is: fit 2, 2 × 103 M; fit 3, 102 M; fit 4, 5 × 102 M. B, k determined in the absence of pyrophosphate. Fits 2 and 4 are obtained from Scheme 3 and values in Table IV. Fits 3 and 5 are obtained allowing additionally the routes 11  16 + 18 and 11  13 + 19, with the koff = 1 s1 for T4 DNA ligase and 0.4 s1 for T4 RNA ligase, which is 5% of the value of the koff for the 12  16 + 19. Parameter variation evaluation: Kd of the reaction 12  11+ Mg2+ is: fit 1, 1.67 × 102 M; fit 2, 1.4 × 103 M. C, k for the formation of the enzyme·Mg2ATP from the enzyme-MgAMP and MgP2O. Small symbols with error bars represent the fitted values obtained from data in Fig. 7. Large symbols without error bars represent the values estimated from two concentrations of pyrophosphate (Fig. 7). For fits 1-6 see comments in the text. D, equilibrium concentrations of different Mg2+ forms of the enzyme and nucleotide. Pyrophosphate traces are nearly identical to that of ATP and omitted from the graph for clarity. [ATP]tot = 100%, and [E]tot = 100%. This graph gives an idea about the rates of binding of MgATP2 and Mg2ATP0 to the enzyme, and rates of adenylylation. For example, at 0.3 mM Mg2+ most of ATP is present in the mono-magnesium form, whereas the enzyme is Mg2+-free for 90%. Therefore, the binding rate observed at this [Mg2+] reflects the reaction 1 + 6  11. For further comments see in the text.

We also studied the dependence of the rate of adenylylation on [Mg²⁺]. In this set of experiments, the enzyme was first pre-equilibrated with ATP and then mixed with different amounts of Mg²⁺, triggering adenylylation. In contrast to the ATP binding, formation of the enzyme·adenylylate is strongly stimulated by Mg²⁺, reaching its maximum rate above 100 mM Mg²⁺ (Figs. 8C, 8D, and 9B). Because binding of Mg²⁺ to the nucleotide is significantly faster than the adenylylation, the reaction rate should change due to the different reactivity of the mono- and di-magnesium-coordinated ATP forms. Above 100 mM Mg²⁺ more than 80% of the nucleotide is present in the Mg₂ATP⁰ form, and the reaction rate is nearly 10-fold higher than at 0.4 mM Mg²⁺, when roughly 90% of ATP is present as MgATP². This fact implies that Mg₂ATP⁰ and not MgATP² is the substrate in the nucleotidyl transfer reaction, in agreement with the previously reported results on T4 RNA ligase (13).

The relevant Mg²⁺ coordination state of pyrophosphate, which participates in the reaction of ATP synthesis, was also determined. For example, at 0.4 mM Mg²⁺, nearly all pyrophosphate is present in the mono-magnesium-coordinated form MgP₂O, and the rate constant for this particular substrate was determined by varying the pyrophosphate concentration. At 50 mM Mg²⁺, the pyrophosphate is present mainly in the di-magnesium form, and, likewise, the rate constants for this substrate were obtained. So, both [Mg²⁺] and [P₂O] were systematically varied (Fig. 7). It was found that pyrophosphate participates in the reverse reaction only at relatively low [Mg²⁺], showing a pronounced Mg²⁺-optimum around 3 mM (Fig. 9C). An increase of [Mg²⁺] above this value leads to a decrease of the reverse reaction rate, which becomes negligible at 100 mM Mg²⁺, when pyrophosphate is present mainly as Mg₂P₂O. This finding disagrees with the earlier proposal (13) and implies that Mg₂P₂O is not involved in the ATP synthesis catalyzed by T4 DNA and T4 RNA ligase, and that the MgP₂O is the true catalytic substrate.

At [Mg²⁺] below 1 mM, most of the pyrophosphate is already present in the mono-magnesium form (Fig. 9D). If the rate of the reverse reaction would be determined only by the different reactivity of the Mg₁- and Mg₂-pyrophosphate forms, one would expect the observed reaction rate to decrease from its maximal value at 1 mM Mg²⁺ to zero above 100 mM Mg²⁺, where pyrophosphate is present in the non-reactive form Mg₂P₂O. The existence of an optimum around 3 mM Mg²⁺ indicates a more complicated relationship, suggesting that the adenylylated enzyme can also bind Mg²⁺, and that the Mg²⁺-bound enzyme binds pyrophosphate with a different affinity relative to the Mg²⁺-free enzyme (see Scheme 3).

Modeling and Simulation of the Experimental Data-- To express the results in terms of a reaction mechanism that would account for the effect of Mg²⁺, Scheme 1 was rewritten in the expanded form (Scheme 3) as follows.

View larger version (13K): [in this window] [in a new window]   Scheme 3.   Expanded mechanism of ATP binding and self-adenylation catalyzed by T4 DNA ligase or RNA ligase.

The kinetic roots 8  13 and 9  14 were a priori excluded from the reaction scheme, because in the absence of Mg²⁺ the nucleotidyl transfer does not proceed.

The observed rate constants k, k were obtained by fitting a sum of two exponentials to the experimental kinetic traces. To correlate obtained values to the rate constants in Scheme 3, it was solved numerically and a sum of two exponentials were fitted to the resulting analytical traces to obtain the calculated rate constant values, k, k. The parameters of Scheme 3 were varied until the values of k, k and k, k converged. In addition, the number of non-zero kinetic parameters was varied to obtain the minimal set of reaction roots that would fit the data. Three characteristic examples of the fitting procedure are shown below.

One of the assumptions was that the ligase does not bind Mg²⁺ in the absence of the nucleoside triphosphate, excluding Mg²⁺·enzyme complexes 2 and 19. In addition, we assumed that the enzymes do not alter the stability of the bound complexes Mg²⁺_xATP (i.e. the equilibrium constant for the reaction 11 + Mg²⁺  12 is equal to that of 6 + Mg²⁺  7; for 8 + Mg²⁺  11 to that of 3 + Mg²⁺  6, etc.). This assumption proved untenable, because it was unable to reproduce an optimum for the rate of the reverse reaction at 3 mM Mg²⁺ (Fig. 9C, fits 1 and 4). Neither was it consistent with the steep decrease of the ATP binding rate with increasing of [Mg²⁺] (Fig. 9A, compare fits 2, 3, and 4). A better fit of the experimental data was obtained when we took into account binding of Mg²⁺ to the free enzyme and to the enzyme-adenylylate forming 2 and 19 with K_d = 2 ± 0.6 mM, (Fig. 9A, fits 1 and 2; Fig. 9C, fits 2 and 5). To our knowledge, no data is available on the Mg²⁺ binding to the DNA or RNA ligases in the absence of ATP; on the other hand, it is known that the related nucleotidyltransferases do, indeed, form weak complexes with Mg²⁺ after the covalent binding of the nucleotide. The crystal structure of the guanylylated mRNA capping enzyme from Chlorella virus PBCV-1 shows no bound Mg²⁺ when the crystals are soaked in 5 mM Mg²⁺ solution, but only after increasing the concentration to 100 mM (11). According to the authors in Ref. 12, the crystal structure of the adenylylated Chlorella virus DNA ligase shows a lutetium atom at a site, where Mg²⁺ is expected to coordinate to the non-bridging -phosphate oxygen of the adenylylate moiety. Finally, free AMP in solution complexes Mg²⁺ with K_d between 2 and 13 mM, depending on the medium composition (21).

The minimization showed that all roots connecting II and III could be excluded, except for one, 12  16 + 19 (for the rate constants see Table IV). In this case, the increase of the rate of the reverse reaction at [Mg²⁺] between 0.4 and 3 mM would occur due to an increase of the concentration of the ligase-MgAMP complex (19), whereas the decrease of the rate at [Mg²⁺] above 3 mM would occur as a consequence of the decrease of [MgP₂O] (16) (Fig. 9D, see traces ligase-MgAMP and MgATP).

Additional improvement of the fitting was achieved by allowing the enzyme to alter the stability of the enzyme-bound Mg₂ATP⁰ (Fig. 9B, fits 2 and 4; Fig. 9C, fits 3 and 6). For example, the Mg²⁺ dependence of the adenylyl transfer rate for T4 DNA ligase could be fitted only under the condition that the enzyme would stabilize the bound Mg₂ATP⁰, decreasing K_d for the reaction 12  11 + Mg²⁺ by a factor of ten relative to the reaction in solution 7  6 + Mg²⁺ (Fig. 9B, compare fits 1 and 2, Table IV).

In conclusion, all experimental data on the nucleotidyl transfer catalyzed by T4 ligases obtained in this work can be sufficiently described by Scheme 4 (the rate constants are summarized in Table IV):

SCHEME 4. Effective mechanism of ATP binding and self-adenylation catalyzed by T4 DNA ligase or RNA ligase.

T4 ligase (1) binds non-covalently the MgATP² (6), but not Mg₂ATP⁰ (7). Subsequently, the enzyme·MgATP² complex (11) binds the second Mg²⁺ ion, forming the catalytic enzyme·Mg₂ATP⁰ intermediate (12). In the adenylylation reaction the Mg₂ATP⁰ is the predominant substrate (route 12  16 + 19); the possible contribution to the overall rate by the MgATP² (routes starting from 10 and 11) is below 5% (Fig. 9B, compare fits 2 with 3 and fits 4 with 5). In the reaction of ATP synthesis, the MgP₂O (16) is the main catalytic substrate, reacting with the Mg²⁺-bound enzyme adenylylate (19). The participation of the Mg₂P₂O (17) reacting with the Mg²⁺-free enzyme adenylylate (18) is estimated to contribute less than 5% to the overall rate of the reaction, i.e. within experimental error.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The transient-state kinetic analysis performed in this work aimed to elucidate the mechanism of the nucleotidyl transfer reaction catalyzed by T4 DNA ligase and T4 RNA ligase. In our experiments, three enzyme species with different levels of tryptophan emission were observed, including the nucleotide-free enzyme (1, 2), and the non-covalent enzyme·ATP complexes (8-12). The third species appears only in the presence of Mg²⁺ (see Fig. 1, A and C) and is ascribed to the covalent complexes enzyme-AMP (18) and enzyme-MgAMP (19). Because these three enzyme species are optically active, the overall reaction can be monitored as a superposition of two processes: non-covalent binding of ATP and the formation of the covalent enzyme·adenylylate. By varying the experimental conditions, either of them can be observed separately as well. However, even though we were able to obtain a good kinetic description of the reaction, one should bear in mind that Schemes 1 to 4 are minimal schemes sufficient to describe the data obtained in the current set of experiments, and more complex mechanisms of catalysis cannot be excluded.

Besides the kinetic characterization of T4 DNA and RNA ligases, the objective of this work was to determine the Mg²⁺ stoichiometry of the nucleotidyl transfer reaction. This issue had not been covered in recent extensive structural analyses of DNA ligases and related nucleotidyltransferases (10-12, 22-24). Apparently, no further studies on Mg²⁺ stoichiometry have been performed since Zagrebel'nyi et al. (13) introduced the two-metal-ion concept in the ligase catalysis. An indirect indication that T4 DNA and RNA ligases might employ Mg²⁺ with a stoichiometry greater than one is suggested by the finding that these enzymes are able to synthesize dinucleoside polyphosphates, such as Ap₄A (25-28). These compounds are the substrates for a superfamily of dinucleoside polyphosphates hydrolases (Nudix hydrolases), enzymes, which employ two to three Mg²⁺ ions to cleave Ap₄A to ATP and AMP (29, 30). Furthermore, the two-metal-ion mechanism of nick-closure was proposed in a recent paper on the NAD⁺-dependent DNA ligase from Thermus filiformis, but only by analogy with DNA polymerases (18). So, what could be the role of the second Mg²⁺ ion in the nucleotidyl transfer reaction catalyzed by T4 ligases?

Mechanism of the Nucleotidyl Transfer Catalyzed by T4 DNA and RNA Ligases-- Despite the high intracellular concentration of magnesium (~30 mM) the concentration of free Mg²⁺ is below 1 mM (31). The typical intracellular concentration of ATP is around 1-3 mM (32), indicating that the main fraction of nucleotide is present in the mono-magnesium-coordinated form. From the values shown in Table III, the intracellular concentration of the di-magnesium-coordinated ATP can be estimated as 60-170 µM, which is either higher or in the order of the K_m of many ATP-dependent enzymes, including T4 DNA and RNA ligases (20, 33-39). It has been shown that the mono-magnesium-coordinated form of ATP exists in aqueous solution as a mixture of rapidly exchanging mono-, bi-, and tridentate structures, with - and -phosphoryls being the main ligands (40, 41). The K_d for Mg²⁺ in these complexes is around 10 µM (Table III). The solution structure of the di-magnesium-coordinated nucleotide is unknown; however, it is known that the second Mg²⁺ binds to the nucleotide three orders of magnitude less tightly (Table III). All enzymes performing the nucleotidyl transfer reactions essentially require the metal cofactor for catalysis. The role of the metal includes charge neutralization, electron withdrawal, adjustment of the conformation of the polyphosphate chain, and orientation of the reactive groups in a favorable position for catalysis (42-44). To achieve that, the metal must coordinate to the reacting phosphoryl group, which in fact has been observed in many crystal structures of nucleotide-dependent enzymes catalyzing the cleavage of the - (cf. Refs. 33, 45, 46) and of the - phosphoanhydride bond of the nucleotide (cf. Refs. 47-50). The - phosphoanhydride bond-cleaving enzymes make use of two Mg²⁺ ions: the first ion binds to the high affinity site on the nucleotide, namely between the - and -phosphates, whereas the second contacts the -phosphate (33, 45, 46). Because T4 DNA and RNA ligases are also - phosphoanhydride bond-cleaving enzymes, the requirement for two Mg²⁺ ions in catalysis may serve the same purpose. In this case, the first Mg²⁺ ion could bind in the high affinity non-catalytic site between the - and - phosphates of ATP allowing the second Mg²⁺ ion to bind in the low affinity catalytic site at the -phosphoryl group and promote the bond cleavage and adenylyl transfer.

Even in the absence of an enzyme, Mg²⁺ is known to accelerate the hydrolysis of nucleoside triphosphates to NDP and phosphate. According to the proposed mechanism, the metal ion coordinates to the reacting -phosphoryl of the nucleotide and to the attacking nucleophile, the water molecule (51). A similar metal coordination is proposed in the case of the enzymic hydrolysis of pyrophosphate (52). The estimated pH- and Mg²⁺-independent rate constant of the adenylylation by T4 DNA ligase is between 10³ and 10⁴ s¹,¹ which is significantly slower than the rate of the Mg²⁺ inner-outer complex exchange (2 × 10⁵ s¹, a rate which is practically independent of the type of the exchanging ligand (44)). We thus suggest that T4 DNA and RNA ligases make use of the same mode of catalysis. In the ligase·Mg₂ATP complex, the first Mg²⁺ ion chelates the - and -phosphates and has no obvious catalytic role, except for occupying the high affinity binding site. The second Mg²⁺ ion coordinates to the low affinity binding site at the -phosphoryl moiety of ATP and, at the same time, forms an inner-sphere complex with the -amino group of the catalytic lysine, thus promoting the nucleophilic attack by an appropriate positioning of the reacting groups.

It is known that neutralization of the charge repulsion between the reacting phosphates is one of the functions of Mg²⁺ in catalysis of the phosphoryl transfer. Because it leads to the decrease of the activation energy, it concerns both directions of the reaction (42, 44). To the authors it seems unlikely that the first Mg²⁺ ion, which binds between the - and -phosphates would perform this task, as was suggested for DNA polymerases in (46). Instead, we propose that the second Mg²⁺ ion compensates the negative charge on the bridging oxygen when the phosphoanhydride bond is being broken or, in the reverse reaction, is about to be made. In this way one Mg²⁺ ion would coordinate to both reacting nucleophiles, the catalytic lysine residue and the bridging oxygen of ATP, as shown in Fig. 10. In the transition-state structure at least two out of five -phosphorane substituents would form inner-sphere complexes with the catalytic Mg²⁺. The phosphorane would not have its usual trigonal-bipyramidal (TBP)² symmetry; instead, the phosphorane is proposed to adopt a square-pyramidal (SP) symmetry, in which the TBP apical bonds (to the -amino group of the lysine residue and to the -bridging oxygen) are bent toward each other (see deviation from TBP axis, Fig. 10). In solution in the absence of Mg²⁺ phosphorane SP symmetry is energetically less favorable than the TBP, but only slightly (53). The proposed close disposition of the -phosphorane and Mg²⁺ also suggests that one of the inner-sphere-coordinated water molecules could be replaced by the outer-sphere-coordinated non-bridging -phosphorane oxygen, so that Mg²⁺ would ligand three out of five -phosphorane substituents. In this case, the phosphorane geometry would become similar to a 30° barrier conformation of the turnstile rotation mechanism (54), which energetically is very similar to the SP conformation (53). The role of the enzyme in assembling the transition state configuration shown in Fig. 10 would be to provide the nucleophile (-amino group of the lysine residue), to precisely position the nucleotide and, possibly, to move the pyrophosphate leaving group to the apical position with respect to the catalytic lysine residue by changing to a closed conformation (cf. Refs. 11, 12, 23, 55). In addition, the enzyme can stabilize the transition state by replacing one or more of the Mg²⁺-coordinated water molecules with protein ligands, leading to an increase of the affinity of the enzyme·Mg₂ATP complex for Mg²⁺ compared with that of Mg₂ATP in solution. This we actually observed in the case of T4 DNA ligase (cf. Fig. 9B, compare fits 1 and 2).

View larger version (18K): [in this window] [in a new window]   Fig. 10.   Proposed structure for the transition state of metal-assisted nucleotidyl transfer catalyzed by T4 DNA ligase and T4 RNA ligase. See text for explanation. Details of the structure of Mg2+ coordinated to the - and -phosphates in the high affinity non-catalytic binding site are not shown. The octahedral geometry of the Mg2+ coordination complex is indicated with gray lines. The reacting -phosphorane is shown in boldface. The phosphorane TBP symmetry axis directed along the apical bonds is shown with a dashed line.

To generalize our proposal, the transition state configuration shown in Fig. 10, in which the catalytic Mg²⁺ coordinates to two substituents of the reacting phosphorane, namely the attacking nucleophile and the nucleophile of the leaving group, could serve as the catalytic intermediate in all metal-promoted phosphoryl transfer reactions.

	ACKNOWLEDGEMENT

We thank Prof. W. R. Hagen for critically reading the manuscript.

	FOOTNOTES

^* This work was supported by the Association of Biotechnology Centers in The Netherlands (project I.2.8), by the Netherlands Research Council for Chemical Sciences, and by the Netherlands Technology Foundation (grant 349-3565).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.

To whom correspondence should be addressed: Tel.: 31-15-278-5139; Fax: 31-15-278-2355; E-mail: S.deVries@tnw.tudelft.nl.

Published, JBC Papers in Press, October 30, 2001, DOI 10.1074/jbc.M109616200

¹ A. V. Cherepanov and S. de Vries, unpublished results.

	ABBREVIATIONS

The abbreviations used are: TBP, trigonal-bipyramidal; SP, square-pyramidal.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				A. V. Cherepanov, E. V. Doroshenko, J. Matysik, S. de Vries, and H. J. M. de Groot The associative nature of adenylyl transfer catalyzed by T4 DNA ligase PNAS, June 24, 2008; 105(25): 8563 - 8568. [Abstract] [Full Text] [PDF]

				K. E. Omari, J. Ren, L. E. Bird, M. K. Bona, G. Klarmann, S. F. J. LeGrice, and D. K. Stammers Molecular Architecture and Ligand Recognition Determinants for T4 RNA Ligase J. Biol. Chem., January 20, 2006; 281(3): 1573 - 1579. [Abstract] [Full Text] [PDF]

				W. Sun, A. Pertzev, and A. W. Nicholson Catalytic mechanism of Escherichia coli ribonuclease III: kinetic and inhibitor evidence for the involvement of two magnesium ions in RNA phosphodiester hydrolysis Nucleic Acids Res., February 7, 2005; 33(3): 807 - 815. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/3/1695    most recent M109616200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Cherepanov, A. V. Articles by de Vries, S. Search for Related Content PubMed PubMed Citation Articles by Cherepanov, A. V. Articles by de Vries, S. Social Bookmarking                      What's this?