Steady-state kinetics of the glutaminase reaction of CTP synthase from Lactococcus lactis The role of the allosteric activator GTP in coupling between glutamine hydrolysis and CTP synthesis Martin Willemoe¨s 1 and Bent W. Sigurskjold 2 1 Centre for Crystallographic Studies, Department of Chemistry, University of Copenhagen, Copenhagen, Denmark; 2 Department of Biochemistry, August Krogh Institute, University of Copenhagen, Copenhagen, Denmark CTP synthase catalyzes the reaction glutamine + UTP +ATP fi glutamate + CTP + ADP + P i .Therate of the reaction is greatly enhanced by the allosteric activator GTP. We have studied the glutaminase half-reaction of CTP synthase from Lactococcus lactis and its response to the allosteric activator GTP and nucleotides that bind to the active site. In contrast to what has been found for the Escherichia coli enzyme, GTP activation of the L. lactis enzyme did not result in similar k cat values for the glutami- nase activity and glutamine hydrolysis coupled to CTP synthesis. GTP activation of the glutaminase reaction never reached the levels of GTP-activated CTP synthesis, not even when the active site was saturated with UTP and the non- hydrolyzeable ATP-binding analog adenosine 5¢-[c-thio]tri- phosphate. Furthermore, under conditions where the rate of glutamine hydrolysis exceeded that of CTP synthesis, GTP would stimulate CTP synthesis. These results indicate that the L. lactis enzyme differs significantly from the E.coli enzyme. For the E.colienzyme, activation by GTP was found to stimulate glutamine hydrolysis and CTP synthesis to the same extent, suggesting that the major function of GTP binding is to activate the chemical steps of glutamine hydrolysis. An alternative mechanism for the action of GTP on L. lactis CTP synthase is suggested. Here the binding of GTP to the allosteric site promotes coordination of the phosphorylation of UTP and hydrolysis of glutamine for optimal efficiency in CTP synthesis rather than just acting to increase the rate of glutamine hydrolysis itself. Keywords: CTP synthase; isothermal titration calorimetry; glutaminase activity; allosteric regulation; Lactococcus lactis. CTP synthase (EC 6.4.3.2) catalyzes the synthesis of CTP from UTP by amination of the pyrimidine ring at the 4-position. The enzyme has three functionally distinct sites; the glutaminase site where glutamine hydrolysis occurs, the active site where CTP synthesis takes place and the allosteric site that binds GTP. The reaction is thought to proceed via phosphorylation of UTP by ATP to give an activated intermediate 4-phosphoryl UTP and ADP [1,2]. Ammonia then reacts with this intermediate yielding CTP and P i as illustrated in Scheme 1A. Ammonia can either be utilized directly or be generated from the hydrolysis of glutamine in a GTP-activated reaction [3,4]. Similar mechanisms to that shown in Scheme 1A, have been shown now for several amido transferase enzymes [5–7]. Here the binding of an already activated substrate, or activation of the substrate on the enzyme, in this case by phosphorylation, precedes amination. The overall reaction is as follows: * PPPrib O HN N OPO 3 2- PPPrib O O HN N ATP NH 3 PPPrib H 2 N O HN N OPO 3 2- PPPrib NH 2 O N N A ADP P i PPPrib O O HN N PPPrib H 2 N O HN N OPO 3 2- PPPrib NH 2 O N N B * NH 3 ATP ADP P i PPPrib H 2 N O HN N OH Scheme 1. Proposed mechanisms of CTP synthesis. The box indicates an expected transition state like structure. * indicates that the amino donor can either be free ammonia or ammonia generated from hydrolysis of glutamine. Correspondence to M. Willemoe ¨ s, Centre for Crystallographic Studies, Department of Chemistry, University of Copenhagen, Universitetsparken 5, DK-2100 Copenhagen Ø, Denmark. Fax: + 45 35320299, Tel.: + 45 35320239, E-mail: martin@ccs.ki.ku.dk Abbreviations: ADPNP, adenosine 5¢-[b,c-imido]triphosphate; ATP-cS, adenosine 5¢-[c-thio]triphosphate; CPS, carbamoyl phosphate synthase; DON, 6-diazo-5-oxo-norleucine; GDH, glutamate dehydrogenase; ITC, isothermal titration calorimetry. Enzymes: CTP synthase (EC 6.4.3.2). (Received 24 May 2002, revised 29 July 2002, accepted 9 August 2002) Eur. J. Biochem. 269, 4772–4779 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03175.x ATP þ UTP þ glutamine ! ADP þ P i þ CTP þ glutamate CTP synthase from Escherichia coli has been shown to perform the following two partial reactions: glutamine ! glutamate þ NH 3 ½8 and ATP ! ADP þ P i where the latter reaction takes place only in the presence of UTP [1]. Levitzki and Koshland originally suggested a mechanism as shown in Scheme 1B [9]. This mechan- ism involves another intermediate, a carbinol amine, formed prior to the phosphorylation step. It was found that the E.coli enzyme would hydrolyze glutamine at a steady-state rate similar to the rate of GTP-activated CTP synthesis, if the enzyme was incubated in the presence of GTP, UTP and the nonhydrolyzable ATP analog ADPNP [8]. Furthermore, it was shown that the fold activation of the glutaminase activity by GTP was similar to that of the overall CTP synthesis reaction. It was concluded that the effect of GTP was mainly to enhance the rate of chemical steps of the glutaminase reaction. The finding that the k cat value was similar for glutaminase activity in the presence of GTP, UTP and ADPNP and CTP synthesis with ATP replacing ADP- NP seems in agreement with the mechanism in Scheme 1B, but maybe less so with the mechanism in Scheme 1A. We have previously characterized the CTP synthase from Lactococcus lactis [4]. This enzyme appears to be a more stable tetramer than the E.coli [10], yeast [11] and mammalian [12,13] enzymes that all require the presence of UTP and/or ATP to form tetramers. Therefore, the L. lactis CTP synthase is an attractive candidate for mechanistic and structure–function studies since equilibria between different oligomeric forms of the enzyme will not interfere with the interpretation of the data. In this research, we have analyzed the steady state kinetics of the glutaminase reaction of CTP synthase from L. lactis in order to distinguish between the effects of GTP on the glutaminase reaction and the CTP synthesis reaction. The results from this work suggests that there are major differences between E.coli and L. lactis CTP synthase with respect to the regulation of glutamine hydrolysis. EXPERIMENTAL PROCEDURES Materials Bovine GDH, nucleotides, and all other chemicals were obtained from Sigma, except ATP-cS which was obtained from Roche. CTP synthase from L. lactis was purified as described previously [4]. 6-Diazo-5-oxo-norleucine (DON)- labeled CTP synthase was obtained by incubating overnight at 25 °C, with 100 lLof8mgÆmL )1 of enzyme in 50 m M Hepes, pH 8.0, 2 m M dithiothreitol and 5 m M DON. The protein concentration was determined by the bicinchoninic acid procedure with reagents provided by Pierce Chemical Company and with bovine serum albumin as a standard. The enzyme concentration was calculated using an M r of 60 000 per subunit [4]. Spectrophotometric assay of CTP synthesis and glutaminase activity Assays were performed at 30 °Cin50m M Hepes, pH 8.0, 2m M dithiothreitol. For CTP synthesis, the conversion of UTP to CTP with De 291 ¼ 1338 cm )1 Æ M )1 was measured as described previously [4]. For glutaminase activity, a con- tinuous coupled assay was used where the glutamate produced by CTP synthase was oxidized by GDH and monitored by the reduction of NAD + to NADH with De 340 ¼ 6300 cm )1 Æ M )1 as described previously [14]. Unless otherwise stated, the MgCl 2 concentration was 20 m M ,the glutamine concentration was 15 m M , and the concentration of each nucleotide when present in the assay was 1 m M . Isothermal titration calorimetry (ITC) assay of glutaminase activity and CTP synthesis CTP synthase at concentrations between 0.029 and 1.16 l M was loaded in the reaction cell of an MCS active temperature compensation isothermal titration calorimeter from Micro- Cal, LLC (Northampton, MA, USA). The steady-state heat evolvement from successive injections of glutamine or GTP was recorded as a displacement of the baseline as illustrated in Fig. 1A. The power value of the baseline and its 02468101214 30 32 34 Time (min) Power (µcal s -1 ) A B 60 70 80 90 100 110 120 130 140 150 30 31 32 33 Time (min) Power (µcal s -1 ) Fig. 1. Isothermal titration calorimetry of the glutaminase activity of L. lactis CTP synthase. (A) Enthalpogram showing the recording of steady-state rates for the hydrolysis of glutamine at increasing substrate concentrations measured as the displacement of the baseline. The peaks observed at each injection time are derived from the heat of dilution of glutamine into the reaction cell. (B) The heat generated by hydrolysis of 0.15 l M of glutamine used to determine DH of the reaction. Ó FEBS 2002 Glutaminase activity of CTP synthase (Eur. J. Biochem. 269) 4773 displacement with each injection was directly obtained from the data files as the value recorded just prior to the subsequent injections. Dividing the power with the molar reaction enthalpy gives the steady-state rate. The assay conditions were as described above for the spectrophoto- metric assays. The molar reaction enthalpy, DH,ofgluta- mine hydrolysis or CTP synthesis under the experimental conditions as outlined above, was determined by recording the complete hydrolysis of between 0.05 and 0.15 lmol of glutamine injected into the reaction cell containing between 1.7 and 17 l M CTP synthase and integrating the entire heat evolvement over time. For the measurement of DH for CTP synthesis the enzyme was incubated with ATP, UTP, GTP and MgCl 2 as described above. Although hydrolysis of ATP will take place under these conditions before and after injection of glutamine and prior to kinetic experi- ments, it will not interfere with the measurements as this hydrolysis is slow and steady-state will prevail for at least 8 h, greatly exceeding the time required for the experiments (about 20–40 min). The values of DH were found to be )29.7 ± 0.8 kJÆmol )1 and )47.3 ± 0.3 kJÆmol )1 for the glutaminase reaction and the CTP synthesis reaction, respectively. The first value agrees well with the value of Kishore et al.[15]. The steady-state rate, or initial velocity, v j for each injection, j, was determined from v j ¼ DP j DH½E ð1Þ where DP is the change in compensation power of the calorimeter necessary to maintain the temperature in the reaction cell upon injection of substrate. This is represented by a shift in the baseline position. DH is the molar enthalpy of the reaction under the chosen experimental conditions, and [E] is the enzyme concentration. The correction for the enzyme dilution and liquid displacement upon substrate injection was calculated from ½E j ¼½E jÀ1 exp À V inj V cell  ð2Þ where V inj and V cell are the volumes of the injectant solution and the reaction cell, respectively. The corrections for the dilution of substrate already present in the reaction cell upon further substrate injection and for the decrease in substrate concentration with time were calculated from ½S j ¼½S jÀ1 exp À V inj V cell  À v jÀ1 t jÀ1  þ½S syr 1 À exp À V inj V cell  ð3Þ where [S] j is the accumulated substrate concentration at the time of injection j,[S] j)1 is the substrate concentration at the time of injection j)1, v j)1 is the steady-state rate of enzyme activity prior to injection j, t j)1 is the time between injection j)1andj,[S] syr is the concentration of the injectant in the syringe. Analysis of initial velocity data Analysis of saturation curves was performed by nonlinear regression using v ¼ k cat ½E½S K M þ½S ð4Þ where k cat is the turnover number for the enzyme, [S] is the substrate concentration and K m is the Henri–Michaelis– Menten constant. Partial inhibition of the glutaminase activity induced by GTP was analysed using a modification of the equation by LiCata and Allewell [16] v ¼ k cat ½Eþk cat;inh ½Eð½A=I 0:5 Þ n 1 þ K A =½Aþð½A=I 0:5 Þ n ð5Þ where k cat,inh is the turnover number for the enzyme fully complexed with the activator A, with an activation constant, K a , that in turn shows cooperative partial inhibition with an inhibition constant for half-maximal inhibition, I 0.5 ,anda Hill-coefficient, n. Initial velocity data from the activation of CTP synthesis or glutaminase activity by GTP as measured spectrophotometrically was analysed using v ¼ k cat;1 ½Eþ k cat;2 ½E½A K A þ½A ð6Þ where k cat,1 and k cat,2 are the turnover numbers for the enzyme in the absence of or fully saturated with the activator, respectively. GTP inhibition of the ammonia dependent CTP synthesis reaction for DON labeled CTP synthase was performed using % Inhibition ¼ 100% ½I n K n I þ½I n ð7Þ where K I is the concentration of inhibitor that gives rise to 50% inhibition by the inhibitor I and n is the Hill- coefficient. The standard errors presented are those given by the computer program ( ULTRAFIT for the Macintosh vs. 3.0, BioSoft). RESULTS Steady state kinetics using ITC The use of a calorimetric assay for the study of the L. lactis CTP synthase proved very useful, since a complete satura- tion curve for glutamine or GTP could be recorded quickly and reproducibly without complicating interference from the presence of added ligands (e.g. the GTP absorbance at 291 nm that prevents the use of higher concentrations than about 0.3 m M GTP in the assay). A calorimetric assay has been shown recently to be generally applicable to most enzyme systems [17]. However, the technique requires that DH for the studied reaction is significantly different from zero. Another condition that needs to be fulfilled in order to determine k cat foranenzymeisthatDH must be measured under experimental conditions in which complete turnover of the substrate takes place. Alternatively, one has to determine the equilibrium constant of the reaction to calculate DH. Both the glutaminase reaction and the CTP synthesis reaction of CTP synthase are virtually irreversible and present no problem with respect to determining DH. Figure 1A shows the raw calorimetric data (enthalpo- gram) obtained for glutamine hydrolysis by CTP synthase. The complete hydrolysis of glutamine necessary for calcu- lating the molar enthalpy DH, and subsequently the rate of 4774 M. Willemoe ¨ s and B. W. Sigurskjold (Eur. J. Biochem. 269) Ó FEBS 2002 the reaction, is shown in Fig. 1B. Finally, the calorimetric data converted to rate data can be fitted as conventional enzyme kinetic data. From Fig. 2 it can be seen that there is an excellent agreement between measured initial velocities independently of the assay method used. Steady state kinetics of uncoupled- and CTP synthesis-coupled glutamine hydrolysis The ATP analogs, ATP-cS and ADPNP, did not serve as substrates (data not shown) but inhibited the CTP synthesis reaction (Fig. 3). ADPNP, reported to inhibit the E.coli enzyme with a K i similar to the dissociation constant for ATP [9], was a poor inhibitor of L. lactis CTP synthase compared to ATP-cS. On the basis of these results, ATP-cS was chosen as a binding analog of ATP. CTP synthesis requires the presence of both the nucleotide substrates ATP and UTP. When phosphorylation of UTP is hindered by the absence of ATP, only glutamine hydrolysis takes place. GTP alone or in combination with UTP and ATP-cS gave a Glutamine, mM Glutamine, mM A B 0 1 2 3 4 0 0.05 0.1 0.15 0 1 2 3 4 0 1 2 3 ν , s -1 ν , s -1 Fig. 2. Comparison of the ITC assay with spectrophotometric assays. Initial velocities obtained from varying glutamine. (A) Glutaminase (circles) and CTP synthesis (squares). Open symbols represent data obtained from the assay of glutamate production coupled to NAD + reduction by GDH. Closed symbols represent data obtained from ITC. (B) CTP synthesis in the absence (circles) and presence (squares) of 0.1 m M GTP. Open symbols represent data obtained from spec- trophotometric measurement of UTP to CTP conversion. Closed symbols represent data obtained from ITC. In panel (A) and (B) closed (squares) and closed (circles), respectively, represent the same data- points. 0 2 4 6 8 0 0.25 0.5 0.75 1 1.25 Relative Activity Inhibitor, mM Fig. 3. Inhibition of L. lactis CTP synthase by ADPNP and ATP-cS. CTP synthesis was measured spectrophotometrically as described in Experimental Procedures. Inhibition was by ADPNP (circles) and by ATP-cS (triangles). Table 1. Kinetic constants for L. lactis CTP synthase from varying glutamine in the presence of various nucleotides a . Reaction b Nucleotides present K m (m M ) k cat (s )1 ) Glutaminase none 1.10 ± 0.06 0.084 ± 0.002 Glutaminase GTP c 1.15 ± 0.05 0.259 ± 0.004 Glutaminase UTP, ATP-cS 1.26 ± 0.09 0.082 ± 0.002 Glutaminase GTP c , UTP, ATP-cS 1.0 ± 0.1 0.56 ± 0.02 CTP synthesis UTP, ATP 0.95 ± 0.03 0.139 ± 0.001 CTP synthesis GTP d , UTP, ATP 0.75 ± 0.05 2.85 ± 0.06 CTP synthesis GTP c , UTP, ATP 0.34 ± 0.03 5.7 ± 0.1 a Assays were performed by ITC as described in Experimental procedures. Glutamine varied from 0–4 m M . b Glutaminase refers to hydrolysis of glutamine without CTP synthesis taking place. c The GTP concentration was 1 m M . d The GTP concentration was 0.1 m M . Ó FEBS 2002 Glutaminase activity of CTP synthase (Eur. J. Biochem. 269) 4775 threefold or sixfold increase in k cat , respectively, compared to the absence of nucleotides, whereas the K m for glutamine was similar for all conditions (Table 1). Adding ATP and UTP, so that CTP synthesis could take place, gave a 1.7-fold increase in k cat when compared to glutamine hydrolysis in the absence of nucleotides (Table 1). The rate of CTP synthesis was dramatically influenced by the presence of GTP, and 20 and 41-fold increases in k cat were obtained with GTP concentrations of 0.1 m M and 1m M , respectively. However, only a modest decrease in K m for glutamine was observed when compared to the absence of GTP (Table 1). Allosteric GTP activation of uncoupled- and CTP synthesis-coupled glutamine hydrolysis As was already indicated by the results in Table 1 and discussed above, the kinetics of GTP activation of the glutaminase half-reaction differed markedly on whether the reaction was coupled to CTP synthesis or not (Fig. 4A). The glutaminase activity in the absence of GTP, represented by k cat,1 (Eqn 6) is not obtainable with the calorimetric assay where GTP is varied, since the heat evolved representing this activity is included in the baseline of the experiment. Therefore, the assay only measures the rate increase due to the addition of GTP with a resulting k cat that represents k cat,2 (Eqn 6). When the ITC assay was used, the basal glutaminase activity was therefore calculated for the experimental conditions used for GTP activation on the basis of the kinetic constants in Table 1, and those obtained for the glutaminase reaction in the presence of 0.1 m M each of UTP and ATP-cS (data not shown). Apparently, the value of K a , k cat,1 and k cat,2 were similar regardless of the active site being saturated with UTP and ATP-cS(1m M each) or not (0.1 m M each) (Table 2). However, GTP concentrations above 1 m M appeared to partially inhibit glutamine hydrolysis when UTP and ATP-cS were present at only 0.1 m M each. This inhibition by GTP seemed to be relieved by increasing the concentration of UTP and ATP- cSto1m M each (Fig. 4B). In either case, as judged from the values of k cat,1 and k cat,2 , the maximal GTP activation of uncoupled glutamine hydrolysis was about 14-fold (Table 2). At UTP and ATP concentrations of 1 m M each, a 49-fold increase in k cat was observed with a concomitant decrease in K a for GTP of about sevenfold compared to uncoupled glutamine hydrolysis (Fig. 4A and Table 2). GTP-activated CTP synthesis in the presence of low concentrations (0.1 m M each) of ATP and UTP showed a sevenfold activation and a K a value three orders of magnitude lower than for uncoupled glutamine hydrolysis where ATP-cS replaced ATP (Fig. 4C and Table 2). In another experiment similar to that in Fig. 4C, the GTP activation of the glutaminase reaction and CTP synthesis was compared (Fig. 5). In the absence of GTP, the rate of glutamine hydrolysis was significantly higher than the rate of CTP synthesis so that the reactions appeared uncoupled in terms of stochiometry. However, GTP stimulated CTP synthesis and apparently acted to coordinate or couple the two reactions. For comparison, the glutaminase activity calculated from the kinetic constants of GTP activation in the presence of 0.1 m M each of ATP-cS and UTP (Table 2) is indicated by the straight line in Fig. 5. L. lactis CTP synthase was incubated with the glutamine affinity analog, DON, that covalently labels an active site cysteine residue and thereby prevents the use of glutamine as amino donor for CTP synthesis [18]. However, even though the enzyme had no detectable activity with glutamine, the Fig. 4. GTP activation of L. lactis CTP synthase. Data (A,B) were obtained by ITC, or by (C) the spectrophotometric CTP synthesis assay. (A) GTP activation (squares) of CTP synthesis. For comparison GTP activation of glutaminase activity in the presence of 0.1 m M (open circles) or 1 m M (closed circles) each of UTP and ATP-cSis shown within the same concentration range. (B) GTP activation of the glutaminase activity in the presence of 0.1 m M (open circles) or 1 m M (closedcircles)eachofUTPandATP-cS. (C) GTP activation of CTP synthesis at 0.1 m M each of UTP and ATP. 4776 M. Willemoe ¨ s and B. W. Sigurskjold (Eur. J. Biochem. 269) Ó FEBS 2002 ammonia-dependent activity was fully retained, as was also found for the E.colienzyme [9]. GTP has previously been reported to inhibit the NH 4 Cl-dependent CTP synthesis reaction of the DON-labeled E.colienzyme, but not the unmodified enzyme [8]. This was also the case for the L. lactis enzyme (Fig. 6). When this inhibition was analysed as a function of the GTP concentration using Eqn 7 we obtained a K I ¼ 0.40 ± 0.05 and n ¼ 0.39 ± 0.02, results that are very similar to those found for the E.colienzyme [8]. Interestingly, the inhibitory response to GTP binding in this experiment shows negative cooperativity in contrast to the activation experiments presented above, where cooper- ativity is not observed. DISCUSSION The original model for the mechanism of GTP activation of the E.coliCTP synthase was rather complex, involving both negative and positive cooperativity of GTP binding [8]. However, we have not found cooperativity associated with GTP activation in these or previous studies [4] of the L. lactis enzyme. Also, for the E.colienzyme, it appears that the cooperative phenomena that have been observed previously are only associated with equilibrium binding, but seem irrelevant in terms of kinetic activation by GTP [19]. Since there appears to be no effect of GTP on the CTP synthesis reaction where NH 4 Cl is the amino donor, the activation by GTP seems exclusively associated with catalytic properties unique to the utilization of glutamine. This observation also implies that there is no effect of GTP on the rate of phosphorylation of UTP by CTP synthase, as k cat for CTP synthesis with NH 4 Cl as amino donor is similar or higher than for the glutamine-dependent reaction satur- ated with GTP. These observations, valid for both E.coli[8] and L. lactis CTP synthase [4], are important when dissecting the effect of GTP on glutamine-dependent CTP synthesis. As mentioned in the Introduction, the GTP activation of the E.coli enzyme is proposed to be due mainly to an increase in the rate of glutamine turnover [8]. In agreement with this proposal the fold activation by GTP for this enzyme is largely independent of the occupancy of the active site, i.e. whether the ATP binding analog, ADPNP, and Table 2. Kinetic constants for L. lactis CTP synthase from varying GTP in the presence of glutamine and the indicated nucleotides. Reaction a Nucleotides present (m M ) K a (mM) k cat,1 b (s )1 ) k cat,2 c (s )1 ) Glutaminase UTP 0.1 2.43 ± 0.08 0.078 e 1.20 ± 0.03 (Fig. 4B) d ATP-cS 0.1 Glutaminase UTP 1 1.62 ± 0.03 0.076 f 1.054 ± 0.007 (Fig. 4B) ATP-cS1 CTP synthesis UTP 0.1 0.0027 ± 0.0004 0.028 ± 0.005 0.195 ± 0.007 (Fig. 4C) ATP 0.1 CTP synthesis UTP 1 0.22 ± 0.01 0.130 f 6.4 ± 0.1 (Fig. 4A) ATP 1 a Assays were performed as described in the legend to the indicated Fig. Glutaminase refers to hydrolysis of glutamine without CTP synthesis taking place. b In the absence of GTP. c Saturated with GTP. d The other kinetic parameters of Eqn 7 were k cat,inh ¼ 0.353 ± 0.006 s )1 , I 0.5 ¼ 1.65 ± 0.03 m M , and n ¼ 2.83 ± 0.1. e Value calculated from experiment (data not shown) where glutamine was varied in the presence of 0.1 m M ATP-cS and 0.1 m M UTP. f Value calculated from the data in Table 1. GTP, mM % inhibition 0 0.1 0.2 0.3 0.4 0 10 20 30 40 50 Fig. 6. GTP inhibition of DON-labeled L. lactis CTP synthase. The data were fitted to Eqn 7 and kinetic constants are given in the text. GTP, mM 0 0.025 0.05 0.075 0.1 0 0.05 0.1 0.15 0.2 0.25 v, s -1 Fig. 5. Coupling of glutaminase activity and CTP synthesis at 0.1 m M each of UTP and ATP. Spectrophotometrical measurement of CTP synthesis (squares) and glutaminase activity (circles) was performed as described in Experimental procedures. The solid line is calculated on the basis of data (Table 2) from GTP activation of the glutaminase activity in the presence of 0.1 m M each of UTP and ATP-cSandis shown for comparison. Ó FEBS 2002 Glutaminase activity of CTP synthase (Eur. J. Biochem. 269) 4777 UTP are present or not. Also, it has been shown that the E.colienzyme in the presence of GTP, UTP and ADPNP will hydrolyze glutamine with a k cat similar to CTP synthesis, where ATP replaces ADPNP. This latter obser- vation suggests that the rate of glutaminase activity of the E.colienzyme is independent of the UTP-phosphorylation reaction. The results presented here for the L. lactis enzyme seem to indicate significant mechanistic differences between this enzyme and the E.colicounterpart. From Tables 1 and 2 it is seen that GTP would activate the uncoupled glutaminase reaction, but not to the extent observed for CTP synthesis. Even though GTP activation of the uncoupled glutaminase activity was clearly sensitive to the occupation of the active site by the nucleotides UTP and ATP-cS (Fig. 4B), the kinetic constants deviated significantly from those obtained for CTP synthesis under similar conditions (Tables 1 and 2). Together, these results seem to indicate that allosteric binding by GTP alone or in combination with UTP and ATP-cS, is not capable of activating glutamine hydrolysis in terms of k cat to the level of CTP synthesis, where ATP replaced ATP-cS. As ATP- cS could not replace ATP in this coactivation with GTP, we find it reasonable to suggest that the true coactivator is the 4-phosphorylated UTP intermediate in a mechanism as that of Scheme 1A. From Fig. 4 and Table 2 it can be seen that the effect of saturating the active site with ATP-cS and UTP was a relief of partial inhibition by GTP at higher concentrations than 1m M (Fig. 4B). The exact mechanism behind this inhibition cannot be resolved from our data, but apart from this inhibition the kinetics were similar when ATP-cSandUTP were present at 1 m M or 0.1 m M (Table 2). The results presented in Fig. 4B seem to exclude that the lower fold of GTP activation of the uncoupled glutaminase reaction was due to subsaturation with nucleotides binding to the active site. That CTP synthesis in the absence of GTP occurs with a k cat that is higher than for the glutaminase activity under similar conditions, except that ATP-cS replaced ATP or in the complete absence of nucleotides (Table 1), seems to indicate an activation of the glutaminase reaction by the substrate nucleotides alone. A similar observation was made with the E.colienzyme except that UTP and ADPNP also activated the glutaminase reaction, though not to the same extent as ATP and UTP [9]. For the L. lactis enzyme, a plausible explanation may be that 4-phosphorylated UTP by itself acts as a weak activator of glutamine hydrolysis. This activation is greatly enhanced by GTP binding to the enzyme. From Table 2 it can be seen that the degree of saturation with ATP and UTP, unlike ATP-cS and UTP, has a large influence on K a for GTP. This correlation of a decrease in K a for GTP with the lowering of the concentration of nucleotide substrates, has been reported previously [4]. Even though a full description of the mechanism of GTP activation is not yet available, the difference in K a for GTP observed when ATP replaced ATP-cS appears to involve structural changes exerted by formation of 4-phosphoryl UTP, that in turn increases the affinity of the enzyme for GTP. An interesting observation was that when ATP and UTP were present at concentrations that give rise to CTP synthesis at a rate lower than the rate of glutamine hydrolysis, GTP still acted as an activator (Figs 4C and 5). One might have expected that if the role of GTP was solely to increase the rate of glutamine hydrolysis, there would have been no effect of adding GTP when the rate of CTP synthesis was limited by the concentration of nucleotide substrates, and not hydrolysis of glutamine. In the narrow concentration interval from 0 to about 2 l M GTP, the rate of CTP synthesis is stimulated 3–4-fold from a level below to the level of uncoupled glutamine hydrolysis (Fig. 5). This directly illustrates that GTP plays G N G N GN GN A C B GTP Fig. 7. A working model for the structural movements in the L. lactis CTP synthase monomer. (A) uncoupled glutaminase activity, (B) CTP synthesis in the absence of GTP and (C) CTP synthesis in the presence of GTP. (A) Glutamine hydrolysis in the absence of nucleotides takes place on the enzyme without any large structural changes required. (B) CTP synthesis in the absence of GTP only occurs at a slow rate due to an equilibrium between the inactive and active form of the monomer with respect to CTP synthesis that involves rearrangements of the monomer that brings together the glutaminase site and the active site. (C) GTP locks the enzyme in the active form for CTP synthesis and thereby stimulates the k cat of the reaction. Since GTP actually will activate the uncoupled glutaminase reaction the structure of the monomer represented by (C) must also include a minor but highly important rearrangement of sidechains in the active site in response to the formation of 4-phosphoryl UTP. The ammonia dependent CTP synthesis could in this model proceed via an enzyme form similar to (A). G, glutaminase site; N, CTP synthesis site. 4778 M. Willemoe ¨ s and B. W. Sigurskjold (Eur. J. Biochem. 269) Ó FEBS 2002 an additional role in activation of L. lactis CTP synthase apart from stimulating the chemical steps of glutamine hydrolysis. The results presented in Fig. 6 further support a dual role for GTP activation of CTP synthesis. A mechanism in which GTP only acts to increase the rate of hydrolysis of glutamine has some flaws in explaining the inhibition of DON-labeledenzymewhenassayedwithNH 4 Cl as amino donor. Therefore, Levitzki and Koshland suggested that upon GTP binding, small structural changes took place in theDON-labelledenzymethatalsoaffectedtheactivesitein an inhibitory manner [8]. It may be that these structural changes in the active site postulated by Levitzki and Koshland are related to the GTP activation of CTP synthesis at excess glutamine hydrolysis in the 0 to about 2 l M concentration range of the activator in Fig. 5. It seems plausible that the active site gets shielded from the environ- ment upon GTP binding to the DON-labelled enzyme in a way so that ammonia can no longer enter the active site. The DON-labeled enzyme with GTP bound would then mimic the active form of the enzyme with hydrolyzed glutamine in the glutaminase site and nucleotide substrates in the active site. This is similar to the GTP-sensitive competitive inhibition of NH 4 Cl utilization, exerted by glutamate c-semialdehyde as found with the E.coli enzyme [20]. Glutamate c-semialdehyde is an analog of glutamine that mimics a tetrahedral reaction intermediate [20]. In our current model (Fig. 7), GTP may act to close a lid over the active site, a lid that in turn holds or rearranges catalytically important residues, and residues that enable the enzyme to perform a concerted glutamine hydrolysis with the formation of 4-phosphoryl UTP. Maybe GTP could play a role in the formation of a tunnel for passing ammonia from the glutaminase site to the active site. Such tunnels have been demonstrated in several glutamine amidotransferases [5]. Another enzyme, that also catalyzes amino transfer from glutamine, is carbamoyl phosphate synthase (CPS) which is the first enzyme of the de novo pyrimidine biosynthesis. For CPS, glutamine hydrolysis has been shown to be greatly stimulated by bicarbonate- dependent ATP hydrolysis, indicating that for this enzyme the phosphorylated amino acceptor intermediate, carbonyl phosphate, triggers an allosteric signal to the glutaminase site [5]. We imagine the same type of allosteric activation of glutamine hydrolysis takes place by the phosphorylation of UTP on CTP synthase, only that for CTP synthase this allosteric effect exerted by the amino acceptor is strongly controlled by GTP. ACKNOWLEDGEMENTS This work was supported by the Danish National Research Foundation. We gratefully acknowledge the expert technical assist- ance by Dorthe Boelskifte. We wish to express our gratitude to Sine Larsen for support to M. W. and for comments to the manuscript. REFERENCES 1. von der Saal, W., Anderson, P.M. & Villafranca, J.J. (1985) Mechanistic investigations of Escherichia coli cytidine-5¢-triphos- phate synthetase. Detection of an intermediate by positional iso- tope exchange experiments. J. Biol. Chem. 260, 14993–14997. 2. Lewis, D.A. & Villafranca, J.J. (1989) Investigation of the mechanism of CTP synthetase using rapid quench and isotope partitioning methods. Biochemistry 28, 8454–8459. 3. Zalkin, H. (1985) CTP synthetase. Methods Enzymol. 113, 282– 287. 4. Wadskov-Hansen, S.L., Willemoes, M., Martinussen, J., Hammer, K., Neuhard, J. & Larsen, S. (2001) Cloning and verification of the Lactococcus lactis pyrG gene and characteri- zation of the gene product, CTP synthase. J. Biol. Chem. 276, 38002–38009. 5. Raushel, F.M., Thoden, J.B. & Holden, H.M. (1999) The ami- dotransferase family of enzymes: molecular machines for the production and delivery of ammonia. Biochemistry 38, 7891–7899. 6. Deras, M.L., Chittur, S.V. & Davisson, V.J. (1999) N2-hydro- xyguanosine 5¢-monophosphate is a time-dependent inhibitor of Escherichia coli guanosine monophosphate synthetase. Biochem- istry 38, 303–310. 7. Kim, J.H., Krahn, J.M., Tomchick, D.R., Smith, J.L. & Zalkin, H. (1996) Structure and function of the glutamine phospho- ribosylpyrophosphate amidotransferase glutamine site and com- munication with the phosphoribosylpyrophosphate site. J. Biol. Chem. 271, 15549–15557. 8. Levitzki, A. & Koshland, D.E. Jr (1972) Role of an allosteric effector. Guanosine triphosphate activation in cytosine triphos- phate synthetase. Biochemistry 11, 241–246. 9. Levitzki, A. & Koshland, D.E. Jr (1971) Cytidine triphosphate synthetase. Covalent intermediates and mechanisms of action. Biochemistry 10, 3365–3371. 10. Robertson, J.G. (1995) Determination of subunit dissociation constants in native and inactivated CTP synthetase by sedimen- tation equilibrium. Biochemistry 34, 7533–7541. 11. Pappas, A., Yang, W.L., Park, T.S. & Carman, G.M. (1998) Nucleotide-dependent tetramerization of CTP synthetase from Saccharomyces cerevisiae. J. Biol. Chem. 273, 15954–15960. 12. Thomas, P.E., Lamb, B.J. & Chu, E.H. (1988) Purification of cytidine-triphosphate synthetase from rat liver, and demonstration of monomer, dimer and tetramer. Biochim. Biophys. Acta 953, 334–344. 13. van Kuilenburg, A.B., Elzinga, L., van den Berg, A.A., Slingerland, R.J. & Van Gennip, A.H. (1994) A fast and novel assay of CTP synthetase. Evidence for hysteretic properties of the mammalian enzyme. Anticancer Res. 14, 411–415. 14. Beismann-Driemeyer, S. & Sterner, R. (2001) Imidazole glycerol phosphate synthase from Thermotoga maritima.Quaternary structure, steady-state kinetics, and reaction mechanism of the bienzyme complex. J. Biol. Chem. 276, 20387–20396. 15. Kishore, N., Tewari, Y.B. & Goldberg, R.N. (2000) A thermo- dynamic study of the hydrolysis of 1-glutamine to (1-glutamate + ammonia) and of 1-asparagine to (1-aspartate + ammonia). J. Chem. Thermodynamics 32, 1077–1090. 16. LiCata, V.J. & Allewell, N.M. (1997) Is substrate inhibition a consequence of allostery in aspartate transcarbamylase? Biophys. Chem. 64, 225–234. 17. Todd, M.J. & Gomez, J. (2001) Enzyme kinetics determined using calorimetry: a general assay for enzyme activity? Anal. Biochem. 296, 179–187. 18. Levitzki, A., Stallcup, W.B. & Koshland, D.E. Jr (1971) Half-of- the-sites reactivity and the conformational states of cytidine tri- phosphate synthetase. Biochemistry 10, 3371–3378. 19. Robertson, J.G. & Villafranca, J.J. (1993) Characterization of metal ion activation and inhibition of CTP synthetase. Biochem- istry 32, 3769–3777. 20. Bearne, S.L., Hekmat, O. & Macdonnell, J.E. (2001) Inhibition of Escherichia coli CTP synthase by glutamate gamma- semialdehyde and the role of the allosteric effector GTP in glutamine hydrolysis. Biochem. J. 356, 223–232. Ó FEBS 2002 Glutaminase activity of CTP synthase (Eur. J. Biochem. 269) 4779 . Steady-state kinetics of the glutaminase reaction of CTP synthase from Lactococcus lactis The role of the allosteric activator GTP in coupling between. glutaminase activity, (B) CTP synthesis in the absence of GTP and (C) CTP synthesis in the presence of GTP. (A) Glutamine hydrolysis in the absence of nucleotides