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

J. Biol. Chem., Vol. 276, Issue 41, 37922-37928, October 12, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/41/37922    most recent M107198200v1 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 Sanker, S. Articles by Kent, C. Search for Related Content PubMed PubMed Citation Articles by Sanker, S. Articles by Kent, C. Social Bookmarking                      What's this?

Negative Cooperativity of Substrate Binding but Not Enzyme Activity in Wild-type and Mutant Forms of CTP:Glycerol-3-Phosphate Cytidylyltransferase^*

Subramaniam Sanker, Heidi A. Campbell, and Claudia Kent

From the Department of Biological Chemistry, University of Michigan Medical Center, Ann Arbor, Michigan 48109-0606

Received for publication, July 29, 2001, and in revised form, August 2, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CTP:glycerol-3-phosphate cytidylyltransferase (GCT) catalyzes the synthesis of CDP-glycerol for teichoic acid biosynthesis in certain Gram-positive bacteria. This enzyme is a model for a cytidylyltransferase family that includes the enzymes that synthesize CDP-choline and CDP-ethanolamine for phosphatidylcholine and phosphatidylethanolamine biosynthesis. We have used quenching of intrinsic tryptophan fluorescence to measure binding affinities of substrates to the GCT from Bacillus subtilis. Binding of either CTP or glycerol-3-phosphate to GCT was biphasic, with two binding constants of about 0.1-0.3 and 20-40 µM for each substrate. The stoichiometry of binding was 2 molecules of substrate/enzyme dimer, so the two binding constants represented distinctly different affinities of the enzyme for the first and second molecule of each substrate. The biphasic nature of binding was observed with the wild-type GCT as well as with several mutants with altered K_m or k_cat values. This negative cooperativity of binding was also seen when a catalytically defective mutant was saturated with two molecules of CTP and then titrated with glycerol-3-phosphate. Despite the pronounced negative cooperativity of substrate binding, negative cooperativity of enzyme activity was not observed. These data support a mechanism in which catalysis occurs only when the enzyme is fully loaded with 2 molecules of each substrate/enzyme dimer.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CTP:glycerol-3-phosphate cytidylyltransferase (GCT)¹ belongs to the family of cytidylyltransferases that includes CTP:phosphocholine cytidylyltransferase and CTP:phosphoethanolamine cytidylyltransferase (1). CTP:phosphocholine cytidylyltransferase is a key regulatory enzyme in the pathway leading to the biosynthesis of phosphatidylcholine (2), and CTP:phosphoethanolamine cytidylyltransferase is involved in the synthesis of phosphatidylethanolamine (3). GCT catalyzes the conversion of CTP and glycerol-3-phosphate to CDP-glycerol, which then serves as the substrate for synthesis of poly- (glycerol phosphate), a cell wall teichoic acid in a number of Gram-positive bacteria. The GCT used in these studies is the product of the tagD gene of Bacillus subtilis (4, 5). Alignment of amino acid sequences comprising the catalytic cores of members of this cytidylyltransferase family reveals several residues that are highly conserved (1). Other properties shared by GCT and CTP:phosphocholine cytidylyltransferase are relatively high K_m values for both substrates and the fact that they are homodimers (6, 7). In light of these similarities and the ease of expression and purification of GCT, we have begun to use GCT as a model to study the catalytic mechanism of this group of cytidylyltransferases (6). Highly conserved residues have been mutated, and the effects of the mutations on catalysis have been determined (1). The recent solution of a crystal structure of GCT with bound CTP has revealed the location of the critical catalytic residues as well as the location of those that appear to be important for structural stabilization (8).

Examination of the crystal structure for residues that interact with the bound CTP shows two highly conserved sequences that appear to be involved in CTP binding. One of these conserved sequence motifs is the HXGH motif, first identified in the class I aminoacyl-tRNA synthetases (9, 10). Site-directed mutagenesis of the two histidines of the HXGH sequence in GCT, His-14 and His-17, resulted in greatly decreased catalytic activity (1). In the crystal structure, these two histidines are in close proximity to the  and  phosphate oxygens of CTP. The second conserved sequence, the RTEGISTT motif, is a signature sequence of this cytidylyltransferase family (1). Mutation of these residues is also detrimental to catalysis, and the structure shows that these residues form a loop that has multiple interactions with the bound CTP (8).

As a step toward understanding the catalytic mechanism of GCT, we have begun to measure affinities of the enzyme for its substrates. For GCT, one might expect that the K_d values for the substrates would be similar to the respective K_m values because for a random order, rapid equilibrium reaction, the K_m for a substrate is equal to the dissociation constant for that substrate from the ternary enzyme complex (11, 12). However, if the binding of one substrate to the enzyme affects the binding of the other substrate, then this relationship of K_d to K_m would not be expected to hold. The possibility that the K_m of GCT for CTP may not be similar to K_d was suggested by finding CTP bound to the active site of crystallized GCT (8). Whereas the procedure used to purify the enzyme used for crystallization involved the use of CTP to elute the enzyme from an affinity matrix, excess free CTP was then removed from the enzyme, and CTP was not included in the crystallization solution. The presence of CTP in the crystal structure therefore suggests that the enzyme has a much higher affinity for CTP than would be indicated by its K_m for CTP of about 1- 3 mM (1, 6).

In this study, we report the use of quenching of intrinsic tryptophan fluorescence to determine substrate binding affinities. We found that the K_d values for initial binding of each substrate were >1000-fold lower than the K_m values. Moreover, the enzyme exhibited negative cooperativity with respect to substrate binding, in that the binding of subsequent substrate molecules occurred with markedly decreasing affinities. However, negative cooperativity of enzyme activity is not observed, implying a mechanism in which catalysis does not occur unless the enzyme is "fully loaded," i.e. all four substrate molecules are bound to the enzyme dimer.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- CHAPS, N-acetyl tryptophanamide, CTP, glycerol-3-phosphate, CDP-glycerol, and sodium pyrophosphate were obtained from Sigma. Sequenase v.2, [³H]CTP, and [¹⁴C]glycerol-3-phosphate were obtained from Amersham Pharmacia Biotech. Ni²⁺-nitrilotriacetate-agarose was purchased from Qiagen. Plasmid pET21b and Escherichia coli HMS 174 (DE3) pLysS were from Novagen. Restriction enzymes were from New England Biolabs.

Protein Expression and Purification-- Construction and characterization of N-terminally His-tagged GCT and site-directed mutants thereof were as described previously (1). The wild-type and mutants of GCT were expressed in E. coli HMS 174 (DE3) pLysS. The bacterial pellet from 1 liter of culture was frozen at 80 °C overnight and then resuspended in 40 ml of 100 mM potassium phosphate buffer, pH 8.0, containing 1 mg of bovine deoxyribonuclease. The resuspended pellet was incubated on ice for 10 min before being centrifuged at 38,000 × g for 1.5 h. The supernatant was applied to a 1-ml Ni²⁺-nitrilotriacetate-agarose column, and the protein was purified according to the manufacturer's directions (Qiagen). The eluate was treated with 2 mM EDTA to chelate any residual Ni²⁺ ions and then dialyzed against 10 mM Tris-HCl, pH 8.0. The enzyme preparations were essentially homogeneous as assessed by polyacrylamide gel electrophoresis. About 10-15 mg of purified enzyme were obtained from a 1000-ml culture. The purified enzymes were stored at 4 °C because enzyme activity was lost upon freezing.

The protein concentration was estimated routinely by the method of Lowry et al. (13). For determination of stoichiometry of binding, the protein concentration was determined by amino acid analysis at the University of Michigan protein core on an Applied Biosystems amino acid analyzer.

Generation of the Double Mutant H14A/D94A-- The double mutant was constructed from the pET21b constructs of the single mutants, H14A and D94A (1). The original plasmids were cut at the HindIII restriction sites at 140 base pairs in the coding sequence and in the vector after the 3' stop site. The HindIII fragment containing the D94A mutation was ligated to the plasmid containing the fragment with the H14A mutation. The mutagenic sites were verified in the double mutant by DNA sequencing.

Measurement of Tryptophan Fluorescence-- Tryptophan fluorescence was measured at a GCT concentration (200 nM dimer or 400 nM monomer) to obtain a fluorescence intensity of about 1 million photons/min at a slit width of 4-5 µm. The protein was excited at 280 or 295 nm in 20 mM phosphate, pH 8.0, containing 5 mM CHAPS and 5 mM MgCl₂. The CHAPS was included to prevent the protein from binding to the cuvette. Fluorescence emission spectra were recorded from 320-380 nm on a Photon Technology Inc. spectrofluorimeter or a Fluoromax II spectrofluorimeter. Aliquots of substrate were sequentially added to the cuvette, and the volume was changed by no more than 1-5% total. The contents of the cuvette were constantly stirred using a magnetic pellet for uniform mixing during the experiment. For each concentration, the spectrum was recorded thrice, and the values were averaged. Control experiments showed no change in fluorescence when buffer was added without substrate, and the extent of fluorescence quenching was the same when a bolus of substrate was added as when small aliquots were added sequentially, indicating that the decrease in fluorescence was not due to protein denaturation over time.

Inner filter effects due to the high absorbance of CTP in the excitation range were diminished by use of a short path cuvette (inside width, 4 mm) and the use of 295 nm as the excitation wavelength (14, 15). A control experiment was carried out by titrating N-acetyl tryptophanamide solution with CTP at 295 nm. This control showed that there was no inner filter effect up to a concentration of 180 µM CTP. Hence, in all CTP titrations of the enzyme, the highest concentration of CTP used did not exceed 180 µM.

Data Analysis-- The fluorescence quanta obtained at 340 nm were converted to F/F_max. To estimate F_max, a linear plot of 1/F versus 1/S for the highest seven data points was used, with the Y intercept taken as 1/F_max. The F/F_max values were then plotted against S_total using the software Kaleidagraph (Synergy) and fitted to an equation for a two-site model as the sum of two rectangular hyberbolae, F/F_max = {m1 × S^h1/(K_d1 + S^h1)} + {m2 × S^h2/(K_d2 + S^h2)}, where m1 and m2 are the maximal F/F_max values for site 1 and site 2, respectively. For CTP binding data, the average value for m1 was 0.51 ± 0.1, and the average value for m2 was 0.47 ± 0.08. For glycerol-3-phosphate binding, the average value for m1 was 0.58 ± 0.09, and the average value for m2 was 0.44 ± 0.08. The values h1 and h2 were Hill-type exponents required to fit the data. The need for these exponents was not clear but was possibly due to energy transfer among tryptophans within the protein. For CTP binding data, the average value for h1 was 1.6 ± 0.8, and the average value for h2 was 1.7 ± 0.8. For glycerol-3-phosphate binding, the average value for h1 was 1.2 ± 0.2, and the average value for h2 was 1.3 ± 0.7. The K_d values obtained by fitting against S_total were then used to estimate S_free by a quadratic equation. The plot of F/F_max versus S_free was then used to obtain new K_d values. This process was repeated until the new and old K_d values differed by <10%. This iterative process resulted in K_d1 values that differed from those obtained by plotting against S_total by an average factor of 1.4 ± 0.9 for CTP and 1.1 ± 0.6 for glycerol-3-phosphate. The K_d2 values differed by 1.2 ± 0.4 for CTP and 1.1 ± 0.4 for glycerol-3-phosphate.

Stoichiometry of Binding of the Substrates to the Enzyme-- A binding assay with radioactive CTP or glycerol-3-phosphate was used to measure stoichiometry of binding to the enzyme. GCT protein (4.75 nmol; 15.8 µM) was mixed with 0.4-30 nmol (1-100 µM) of the substrate (CTP mixed with [³H]CTP or glycerol-3-phosphate mixed with [¹⁴C]glycerol-3-phosphate as radioactive tracer) in 300 µl of 20 mM Tris-HCl, pH 8.0, and 6 mM MgCl₂. The unbound substrate was separated from the bound substrate by centrifugation in a Microcon-10 microconcentrator (Millipore), and the radioactivity in the filtrate was measured. Care was taken to remove no more than 10% of the total volume by centrifugation. The amount of bound substrate was plotted against free substrate, and the data were fitted to the equation for a rectangular hyperbola. Stoichiometry of binding was also determined by quenching of tryptophan fluorescence. This experiment was carried out with 10 µM GCT protein in otherwise the same conditions as described above for quenching of tryptophan fluorescence. The excitation wavelength was set at 295 nm, and the slit width was 2 µm. In this experiment, the concentration of substrate bound was taken to be the concentration at which fluorescence stopped decreasing.

Kinetic Analysis-- Steady-state kinetic analysis was performed as described previously (1), except that the substrate concentrations were varied from 0.3 µM to 10 mM. The V_max apparent values for each fixed concentration were determined by fitting the data to the Michaelis-Menten equation with Kaleidagraph software (Synergy), and then the V_max apparent values were plotted as a function of fixed substrate, and these data were fitted to the Michaelis-Menten equation to obtain true V_max and K_m values.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Purification of CTP-free Enzymes-- To measure the binding of substrates to GCT, the enzyme must be free of the substrate. The purification procedure we use for native GCT, however, involves elution of the enzyme from an affinity matrix with CTP. As mentioned above, enzyme purified by this procedure contains a molecule of CTP at each active site (8). To purify the substrate-free enzyme, we used His-tagged constructs of wild-type and mutant enzymes. Purification of these enzymes by Ni²⁺-chelate chromatography does not utilize either substrate. These His-tagged constructs have been characterized previously (1). The His-tagged wild-type enzyme exhibits kinetic constants similar to those of non-His-tagged GCT, and the His-tagged wild-type and mutant GCT proteins were shown to retain their tertiary structure by two-dimensional NMR (1). No phosphorus peaks were associated with the His-tagged GCT as indicated by ³¹P NMR spectroscopy, indicating that the enzyme did not contain appreciable quantities of substrates or products.²

Intrinsic Tryptophan Fluorescence-- Tryptophan residues in proteins exhibit intrinsic fluorescence with fluorescence maxima between 340 and 350 nm when the protein solution is excited at about 280-300 nm. Change in fluorescence intensities can be monitored as a measure of ligand binding to the protein (16). GCT has three tryptophan residues in its coding sequence, Trp-15, Trp-74, and Trp-95. Trp-15 is located within the nucleotide-binding motif, the ¹⁴HWGH sequence. Trp-74 and Trp-95 appear to be near the active site of GCT (8), and the W74A mutation exhibits a 10-25-fold decrease in k_cat/K_m values (1). The proximity of Trp-15, Trp-74, and Trp-95 to the active site suggested that these tryptophan residues might be sensitive to substrate binding, allowing the use of fluorescence quenching or enhancement to measure the affinities of the enzyme for the substrates. Emission spectra for GCT fluorescence are shown in Fig. 1. The spectra were taken for the unliganded enzyme (top curve) and for the enzyme with several levels of glycerol-3-phosphate. It is clear that the addition of this substrate caused an appreciable quenching of fluorescence without a noticeable change in emission maximum.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Quenching of intrinsic tryptophan fluorescence upon binding of glycerol-3-phosphate to wild-type GCT. The spectra were recorded on a Fluoromax II spectrofluorimeter at an excitation wavelength of 280 nm. The fluorescence emission was recorded from 325-380 nm. The spectra (from top to bottom) are: GCT alone and then GCT in the presence of 0.05, 0.15, 3.0, 20, 90, 110, 200, and 400 µM glycerol-3-phosphate.

Substrate Binding to Wild-type GCT-- When either CTP or glycerol-3-phosphate was added to wild-type GCT, quenching of tryptophan fluorescence was observed. Plots of fluorescence intensity at 340 nm versus the substrate concentration showed biphasic curves, indicating the existence of two binding sites with distinctly different affinities (Fig. 2). Two K_d values were obtained for the binding of either CTP (0.24 and 38 µM) or glycerol-3-phosphate (0.14 and 22 µM) to GCT (Table I). The lower K_d value will be referred to as K_d1, and the higher K_d value will be referred to as K_d2. The K_d1 values were more than 1000-fold lower than the K_m for each of the substrates, and the K_d2 values were about 40-50-fold lower than the K_m. Therefore, the substrates do bind to the enzyme with much higher affinities than predicted by the K_m values. Moreover, the fact that there are two distinct K_d values indicates that substrate binding is negatively cooperative, i.e. binding to one site decreases the affinity of the other active site for the substrate (17).

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Determination of affinity of wild-type GCT for substrates. Substrates were added sequentially; the fluorescence emission data were recorded and analyzed as indicated under "Experimental Procedures." The lines drawn are the results of fitting the data to the sum of two rectangular hyperbolae as described under "Experimental Procedures." This experiment was performed three times with similar results. The binding constants determined in the three experiments were averaged and are listed in Table I.

To determine whether the two different binding affinities for each substrate actually corresponded to the two active sites per dimer, it was necessary to determine the stoichiometry of substrate binding. The binding of radioactive substrates individually to GCT was therefore measured by filtration using microconcentrators. Plots of bound substrate versus free substrate revealed that binding leveled off at 1.2 molecules of CTP and 0.84 molecules of glycerol-3-phosphate per active site (Fig. 3, A and B). (This experiment does not reveal the very low binding affinity of K_d1 because it utilized a very high enzyme concentration, 16 µM, and was not sufficiently sensitive in the range of K_d1.) The stoichiometry of binding was also determined by tryptophan fluorescence quenching, titrating a high level of enzyme with each substrate, and taking the concentration at which fluorescence stopped decreasing as the maximum bound (Fig. 3, C and D). This experiment confirmed that one substrate bound per monomer. Thus, a total of two molecules of each substrate were binding per dimer. In summary, the binding of either substrate to the enzyme dimer is represented by two distinct affinities, where K_d1 is about 0.1-0.3 µM, and K_d2 is about 20-40 µM.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Stoichiometry of substrate binding to wild-type GCT. A and B, free substrate was separated from bound substrate by filtration as described under "Experimental Procedures." A control experiment was carried out for each of the data points without protein to estimate recovery. The protein concentration was determined by amino acid analysis. This experiment was performed twice with similar results. C and D, fluorescence quenching was used to measure stoichiometry as described under "Experimental Procedures."

Substrate Binding to Mutant D94A-- The results presented above indicated that the K_m value for either substrate in the GCT reaction is not the same as the K_d of the enzyme for the first substrate to bind. This would predict that those mutant enzymes with defective K_m or k_cat values might have the same K_d1 values as seen for the wild-type enzyme. It was therefore of interest to determine these binding constants for several mutant enzymes. The D94A mutant was previously shown to have very high K_m values for both substrates; the K_m values for CTP and glycerol-3-phosphate are about 70- and 130-fold higher, respectively, than those obtained for the wild-type enzyme (1) (Table I). The k_cat value for this mutant is about 5-fold lower than that of the wild-type (1). Analysis of tryptophan fluorescence quenching for D94A in response to substrate binding (Fig. 4, A and B) revealed complex substrate binding curves similar to those seen for the wild-type. The K_d values were 0.1 and 140 µM for CTP and 0.2 and 18 µM for glycerol-3-phosphate (Table I). Whereas the K_d2 for CTP for D94A was 3-fold higher than that of the wild-type, it was still about 600-fold lower than the K_m of D94A for CTP. The other K_d values for D94A were similar to those seen for wild-type GCT and much lower than the K_m values for D94A.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Substrate binding to GCT mutants. Fluorescence emission data were obtained and analyzed as described under "Experimental Procedures." A and B, mutant D94A; C and D, mutant H14A. Each experiment was performed three times with similar results.

Substrate Binding to Mutants H14A and H17A-- His-14 and His-17 in the HWGH sequence motif are implicated in transition state stabilization during catalysis (1). Mutation of these histidines does not appreciably alter K_m but does cause a decrease in k_cat for H14A and H17A by factors of 1.8 × 10⁴ and 2.6 × 10³, respectively. Because these residues are so close to CTP in the active site, it was of interest to determine whether the K_d values were altered by mutagenesis. The quenching of tryptophan fluorescence in response to substrate binding was again complex for both H14A (Fig. 4, C and D) and H17A (data not shown). Two binding constants for each substrate were obtained for each mutant (Table I). Although K_d1 and K_d2 were altered somewhat for the two mutants, they were reasonably close to the K_d values for the wild-type and considerably lower than the K_m values.

Binding of Glycerol-3-phosphate to the Enzyme-CTP Complex-- The results presented above indicated that GCT exhibits negative cooperativity with respect to binding of a single substrate. It was possible that further negative cooperativity existed with respect to binding the other substrate species to the enzyme-substrate complex. It was therefore of interest to follow quenching of tryptophan fluorescence of the enzyme-CTP complex in response to the addition of glycerol-3-phosphate. This experiment would be impossible to do with wild-type GCT because the catalytic activity observed upon addition of the second substrate species would complicate interpretation of the data. However, the H14A mutant catalyzes the reaction so slowly that turnover would be negligible within the time required for the experiment (1). Moreover, the K_d values for binding of a single substrate to H14A were similar to those for wild-type GCT (Table I), thus it was reasonable to expect that the K_d values for binding of the second substrate species would be similar for H14A and wild-type.

This experiment was performed with GCT to which 100 µM CTP had been added to saturate the two CTP binding sites. Quenching of fluorescence was then followed as glycerol-3-phosphate was added sequentially. The curve obtained for glycerol-3-phosphate binding as the second substrate species was biphasic, showing two binding affinities (Fig. 5). The calculated K_d values were 2.6 and 1380 µM (Table I); the lower of these values will be referred to henceforth as K_d3, and the higher of these values will be referred to henceforth as K_d4.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Determination of affinity for the second substrate in the presence of the first substrate. CTP was first added to mutant H14A to a concentration of 100 µM; glycerol-3-phosphate was added sequentially, and the fluorescence emission was recorded.

It is notable that the K_d4 value of 1.3 mM is very similar to the K_m values of 1.0-1.4 mM for glycerol-3-phosphate in the GCT reaction (Table I). This suggests that K_m for this reaction is K_d4, the dissociation constant for the enzyme complexed with four substrate molecules rather than two substrate molecules.

Binding of Glycerol-3-phosphate to a High K_m Enzyme-CTP Complex-- The hypothesis that K_m reflects K_d4 can be tested by determining whether K_d4 is altered by a mutation that alters K_m. As indicated above, mutant D94A has high K_m values for both substrates. However, the k_cat of D94A is only 5-fold lower than that of the wild-type (1), therefore this mutant enzyme is too active to be used with natural substrates for determining the K_d values for glycerol-3-phosphate binding to the enzyme-CTP complex. We therefore constructed a double mutant combining the H14A and D94A mutations. It was assumed that this mutant exhibits the phenotypes of both the single mutants, i.e. the substrate binding properties of D94A and the catalytic properties of H14A. Before performing the dual substrate experiment, quenching of tryptophan fluorescence was followed as a function of binding a single substrate to the H14A/D94A double mutant. Negative cooperativity of binding was observed, with two K_d values (Fig. 6, A and B). The binding constants of 0.05 and 22 µM for CTP and 0.17 and 25 µM for glycerol-3-phosphate were as low as those obtained for wild-type GCT and either of the single mutants (Table I).

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Determination of affinities of high Km mutant. A and B, binding curves for each substrate in the absence of the other to the H14A/D94A double mutant. C and D, CTP was first added to mutant H14A/D94A to a concentration of 180 µM; glycerol-3-phosphate was added sequentially, and the fluorescence emission was recorded. The data in C and D were each fitted to the equation for a rectangular hyperbola. This experiment was performed twice (A, C, and D) or once (B).

For the dual substrate experiment, CTP was first added to the H14A/D94A mutant enzyme to a concentration of 180 µM. Addition of glycerol-3-phosphate to the double mutant-CTP complex did not cause any change in fluorescence up to 50 µM. Quenching was then observed in the range of 50 µM to 30 mM glycerol-3-phosphate (Fig. 6C). Addition of a higher concentration of glycerol-3-phosphate (30-240 mM) caused a further change in fluorescence (Fig. 6D), although it should be noted that the latter change was an enhancement of fluorescence rather than quenching. In control experiments, only a marginal decrease in fluorescence was observed upon substitution of Na₂SO₄ for glycerol-3-phosphate, indicating that the fluorescence changes seen in Fig. 6, C and D, were not due to ionic strength effects. In addition, titration of the H14A single mutant with glycerol-3-phosphate at a concentration up to 240 mM showed appreciable fluorescence quenching only up to 100 µM. The fluorescence quenching of the double mutant between 50 µM and 30 mM could be interpreted as binding of a single molecule of glycerol-3-phosphate to the enzyme-CTP complex, and the fluorescence enhancement from 30-240 mM could be interpreted as binding of the second molecule. These titration data were plotted separately and fitted by regression analysis. The calculated K_d3 value was 1.8 mM for the quenched fluorescence, and a K_d4 value of 109 mM was calculated for the enhanced fluorescence (Table I). The K_d4 value of 109 mM approximates the K_m of the reaction for D94A of 141 mM. This result supports the concept that K_m is the same as K_d4.

Reanalysis of Kinetics-- The previous kinetic analyses in which the K_m values for native GCT and His-tagged GCT were determined did not show evidence of negative cooperativity (1, 6). However, the lowest level of substrate used in those studies was 0.125 mM, which is considerably above the K_d1 and K_d2 values determined in the present study. To determine whether there was appreciable enzyme activity with lower substrate concentrations, which would appear as negative cooperativity in the analysis, we repeated the kinetic analysis for His-tagged wild-type GCT and used 0.3 µM as the lowest level of both substrates. As before (1), the new analysis yielded K_m values for each substrate in the low millimolar range (1.3 mM for both glycerol-3-phosphate and CTP). The secondary data fit well to a Michaelis-Menten equation with no Hill exponent (r = 0.9987 for glycerol-3-phosphate and r = 0.9954 for CTP). If fitted to an equation with a Hill exponent (v = V_max × Sⁿ/[K + Sⁿ]), the exponent n was actually slightly positive (1.2 for glycerol-3-phosphate and 1.3 for CTP), further supporting the lack of negative cooperativity with respect to enzyme activity.

The similarity of K_m and K_d4 suggests a mechanism in which catalysis does not occur until both molecules of each substrate are bound to the enzyme dimer. To determine whether our kinetic data were consistent with this mechanism, we wanted to model the appearance of data that would be observed if there were appreciable activity when only one monomer was occupied with substrates. The reaction was considered to proceed as indicated in Fig. 7, with random binding of substrates A and B to the enzyme dimer. The binding constants referred to above as K_d1 and K_d2 are K_A1, K_B1, K_A2, and K_B2 in the scheme. The binding constants referred to above as K_d3 and K_d4 would be gK_A1 and iK_A2 when A is glycerol-3-phosphate and B is CTP. This kinetic scheme would lead to four enzyme-substrate complexes in which one of the monomers has both substrates bound. Using rapid equilibrium assumptions (18), we derived a velocity equation for the reaction as indicated in the legend to Fig. 7. We then compared the actual enzymatic activity observed with the activity expected if only the EA₂B₂ species is active, the activity expected if all monomers with both A and B bound are equally active, and the activity expected if EAB, EA₂B, and EAB₂ are 10% as active as the EA₂B₂ complex (Fig. 8). The curve obtained with the expectation that EA₂B₂ is the only active complex (squares in Fig. 8) closely resembles the actual data (circles in Fig. 8) at all substrate concentrations. The expectation that the monomers in the EAB, EA₂B, and EAB₂ complexes are as active as those in the EA₂B₂ complex leads to curves that do not fit the data (Xs in Fig. 8). Assuming that the EAB, EA₂B, and EAB₂ complexes are 10% as active as EA₂B₂ leads to a close fit at high fixed substrate levels, but the fit at low fixed substrate levels is poor (triangles in Fig. 8). Thus it can be estimated that very little, if any, activity is obtained until all four substrate sites in the enzyme dimer are occupied.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Kinetic scheme for the GCT reaction. The velocity of the reaction, v, is equal to [EAB]k + [EA2B]k + [EAB2]k + [EA2B2]k, where k- are rate constants for the complexes. The use of random equilibrium assumptions allows the formulation of the following equation for this reaction pathway: v/[Et] = (([A][B]k/cKA1KB1) + ([A]2[B]k/dKA1KA2KB1) + ([A][B]2k/gKA1KB1KB2) + ([A]2[B]2k/dhKA1KA2KB1KB2))/(1 + ([A]/KA1) + ([B]/KB1) + ([A]2/KA1KA2) + ([B]2)/KB1KB2) + ([A][B]/cKA1KB1) + ([A]2[B]/dKA1KA2KB1) + ([A][B]2/gKA1KB1KB2) + ([A]2[B]2/dhKA1KA2KB1KB2)).

View larger version (34K): [in this window] [in a new window]   Fig. 8.   Comparison of experimental and calculated activity of wild-type GCT as a function of substrate concentration. The experimental data for three fixed concentrations of each substrate are represented by open circles and a solid line. The other curves were calculated from the equation in the Fig. 7 legend, with KA1 = 0.14 µM, KA2 = 22 µM, KB1 = 0.24 µM, and KB2 = 38 µM. (These binding constants are from Table I for wild-type enzyme, where A is glycerol-3-phosphate, and B is CTP.) It is also assumed that c = d = g = 158, e = f = 1, h = i = 45, only half of the EAB species contain both substrates in the same monomer, both monomers in the EA2B2 complex are active, and the total catalytic rate = 14s1. The values for c-i were based on the relative values in Table I. The curves with open squares and dashed lines were calculated assuming that k = k = k = 0 and k = 14s1. The curves with Xs and solid lines were calculated assuming that all monomers that contain both substrates are equally active, so that k = 1.56, k = k = 3.11, and k = 6.22. The curves with open triangles and dashed lines were calculated assuming that the monomers in the EA2B2 complex were 10 times as active as the other monomers containing both substrates, or k = 0.31, k = k = 0.62, and k = 12.4.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

These studies have shown that the intrinsic tryptophan fluorescence of GCT is sensitive to binding of substrates to the enzyme and that measurement of the fluorescence quenching is a sensitive technique for estimating the affinities of substrates to this enzyme. It is clear that each substrate binds to the enzyme in the absence of the other substrate, supporting a random order mechanism, as observed previously by kinetic analysis (6).

Several conclusions can be drawn from the results presented here. First, the enzyme binds a single substrate with a much higher affinity than represented by the K_m values in the millimolar range. K_d1 values, in fact, were >1000-fold lower than K_m values. The observation that K_d1 is so much lower than K_m can be seen even with the D94A mutant, in which the initial binding constant is very similar to that of wild-type, but the K_m is much higher than that of wild-type. The K_d1 values for the mutants H14A and H17A were also very similar to those of wild-type GCT, indicating that initial substrate binding is not involved in the dramatic changes in k_cat values for these mutants.

A second conclusion is that there are two distinct binding affinities for each substrate. It is likely that these two binding sites are actually the two active sites of the enzyme dimer because the overall stoichiometry of binding is approximately 2 molecules/enzyme dimer. Furthermore, the crystal structure of GCT reveals CTP bound in each active site of the dimer (8). Therefore, the enzyme exhibits negative cooperativity, in that binding of a substrate molecule to one active site decreases the affinity of the enzyme for the second substrate molecule.

A third notable observation is that binding of the second substrate species to the enzyme-substrate complex also exhibits negative cooperativity. In a situation in which both active sites of mutant H14A were saturated with CTP, the affinity of the enzyme for the first molecule of the second substrate, glycerol-3-phosphate (K_d3), was quite similar to the affinity of the enzyme for the second molecule of the first substrate (K_d2) (Table I). The second molecule of glycerol-3-phosphate then bound with a much lower affinity (K_d4). Most significantly, K_d4 approximated the K_m value for the substrates. That K_d4 approximated K_m was seen with both the H14A enzyme, in which K_m approximated that of the wild-type, and the H14A/D94A double mutant, in which K_m was much higher than that of the wild-type and similar to that of the D94A mutant (1).

Despite the strong negative cooperativity observed upon substrate binding, negative cooperativity is not observed with measurements of enzyme activity. Fitting substrate concentration curves for GCT to a Hill equation does not yield a Hill coefficient of <1. The lack of negative cooperativity with enzyme activity coupled with the similarity of K_m and K_d4 suggests that catalysis does not occur until the enzyme is fully loaded with all four substrate molecules bound to the dimer. This would mean that there is not appreciable activity if only one monomer of the enzyme dimer contains both substrates. In fact, modeling the activity expected if there were appreciable activity of EAB, EA₂B, and EAB₂ complexes results in velocity versus substrate profiles that definitely do not fit the experimental data.

Many enzymes exhibit negative cooperativity of enzymatic activity. A mechanism, however, in which negative cooperativity of substrate binding is not accompanied by negative cooperativity of activity is highly unusual, but not unique. A similar mechanism has been noted with the F₁-ATPase of E. coli. Weber et al. (19) demonstrated that the affinity of the ATPase trimer for ATP decreased after the first molecule was bound but that appreciable ATP hydrolysis was observed only after all three molecules bound. The K_m for ATP hydrolysis was similar to K_d3, the constant for ATP binding to the third catalytic site.

This mechanism may explain the fact that each of the three mutations in GCT that cause an increased K_m for one substrate (D38A, W74A, and D94A) also causes an increased K_m for the other substrate (1). Of these three mutations, only D94A is within 10 Å of CTP in the crystal structure, so it certainly does not seem that the other two residues, Trp-74 and Asp-38, are involved in CTP binding per se. However, the observation that K_m is a complex value determined by binding of both substrate species suggests that some mutations that alter the binding site for one substrate would also affect the K_m value measured kinetically for the other substrate.

The negative cooperativity of substrate binding suggests that the enzyme follows a sequential model in which binding of substrate to one monomer induces a change in the other monomer (20). Alternatively, the data could be explained if the unliganded enzyme dimer were asymmetric. Although the structure of the enzyme with no ligand has not been determined, the structure of the CTP-bound enzyme does not indicate any obvious asymmetry that could account for such distinct binding affinities.

A mechanism in which chemistry does not occur until all four substrate binding sites are occupied necessitates that, prior to the binding of the fourth substrate, one active site is fully occupied with both substrates, but no reaction takes place. In fact, it is likely that this is the usual state of the enzyme in the cell. The physiological concentration of GCT substrates in bacteria appears to be rather high: glycerol-3-phosphate in E. coli is reported to be in the range of 0.2-2 mM (21), and the range of CTP levels in E. coli and Salmonella typhimurium is 0.3-1.4 mM (22-24). These concentrations are well above K_d3, hence even at the lower end of the substrate pool sizes, most of the GCT in the cell would exist in the EA₂B and EAB₂ states. The activity of the enzyme would therefore be very sensitive to changes in cellular substrate levels. If GCT were a regulatory enzyme for teichoic acid biosynthesis, it is possible that regulatory mechanisms might alter the K_m of GCT for one or both of its substrates. The role of GCT in regulation of teichoic acid biosynthesis has not been studied extensively, but the enzyme activity varies greatly in response to external phosphate availability (25). The changes in activity are governed by repression and derepression (26-28) and possibly enzyme inhibition or inactivation (26, 28). Thus, the activity of GCT in vivo is controlled by a variety of mechanisms that include altered substrate levels, altered enzyme levels, and possibly inhibition or activation of the enzyme.

A mechanism similar to that described here may exist for eukaryotic CTP:phosphocholine cytidylyltransferases, given the high degree of similarity of other properties of these enzymes, although substrate-binding measurements have not yet been performed. It is well known that mammalian CTP:phosphocholine cytidylyltransferase is a regulatory enzyme, and this mechanism could render the enzyme quite sensitive to substrate levels. In fact, stimulation of phosphatidylcholine synthesis in response to poliovirus infection is accomplished by alteration of cellular CTP levels (29). It has been reported that activation of CTP:phosphocholine cytidylyltransferase by lipids occurs by a mechanism in which the K_m of the enzyme for CTP is reduced (30), although that mechanism of activation by lipids has been questioned (31). It will be of considerable interest, therefore, to investigate the possibility of negative cooperativity of substrate binding in the eukaryotic CTP:phosphocholine cytidylyltransferases.

	ACKNOWLEDGEMENTS

We thank Drs. David Ballou, Tom K. Kerppola, Michael A. Marletta, and Ari Gafni for use of their spectrofluorimeters, Drs. Eric Schurter, Shawn Stevens, and Erik R. P. Zuiderweg for sharing unpublished results, and Drs. Zuiderweg, Stevens, and Martha Ludwig for frequent discussion.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants RO1 CA64159 and GM60510. The work was supported in part by core services funded by NIH Grant P60DK-20572.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: Dept. of Biological Chemistry, 4417 Medical Science I, University of Michigan Medical School, Ann Arbor, MI 48109-0606. Tel.: 734-764-6118; Fax: 734-763-4581; E-mail: ckent@umich.edu.

Published, JBC Papers in Press, August 3, 2001, DOI 10.1074/jbc.M107198200

² E. Schurter and E. R. P. Zuiderweg, unpublished observations.

	ABBREVIATIONS

The abbreviations used are: GCT, CTP:glycerol-3-phosphate cytidylyltransferase; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid.

	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:

				J. W. Schertzer, A. P. Bhavsar, and E. D. Brown Two Conserved Histidine Residues Are Critical to the Function of the TagF-like Family of Enzymes J. Biol. Chem., November 4, 2005; 280(44): 36683 - 36690. [Abstract] [Full Text] [PDF]

				S. Jackowski and P. Fagone CTP:Phosphocholine Cytidylyltransferase: Paving the Way from Gene to Membrane J. Biol. Chem., January 14, 2005; 280(2): 853 - 856. [Full Text] [PDF]

				K. A. Pattridge, C. H. Weber, J. A. Friesen, S. Sanker, C. Kent, and M. L. Ludwig Glycerol-3-phosphate Cytidylyltransferase: STRUCTURAL CHANGES INDUCED BY BINDING OF CDP-GLYCEROL AND THE ROLE OF LYSINE RESIDUES IN CATALYSIS J. Biol. Chem., December 19, 2003; 278(51): 51863 - 51871. [Abstract] [Full Text] [PDF]

				J. W. Schertzer and E. D. Brown Purified, Recombinant TagF Protein from Bacillus subtilis 168 Catalyzes the Polymerization of Glycerol Phosphate onto a Membrane Acceptor in Vitro J. Biol. Chem., May 9, 2003; 278(20): 18002 - 18007. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 276/41/37922    most recent M107198200v1 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 Sanker, S. Articles by Kent, C. Search for Related Content PubMed PubMed Citation Articles by Sanker, S. Articles by Kent, C. Social Bookmarking                      What's this?