Báo cáo khoa học: Lid L11 of the glutamine amidotransferase domain of CTP synthase mediates allosteric GTP activation of glutaminase activity docx

Lid L11 of the glutamine amidotransferase domain of CTP synthase mediates allosteric GTP activation of glutaminase activity Martin Willemoe ¨ s 1 , Anne Mølgaard 1,2 , Eva Johansson 1,3 and Jan Martinussen 4 1 Centre for Crystallographic Studies, Department of Chemistry, University of Copenhagen, Denmark 2 Center for Biological Sequence Analysis, BioCentrum-DTU, The Technical University of Denmark, Lyngby, Denmark 3 European Synchrotron Radiation Facility, Grenoble Cedex, France 4 Microbial Physiology and Genetics, BioCentrum-DTU, The Technical University of Denmark, Lyngby, Denmark CTP synthase (EC 6.3.4.2) catalyses the synthesis of CTP by amination of the 4-position of the pyrimidine moiety of UTP [1]. The enzyme is a homotetramer, each subunit consisting of two domains; the N-terminal synthase domain where CTP formation takes place and the C-terminal GATase domain Keywords allosteric regulation; flexible loop; Lactococcus lactis; nucleotide metabolism; oxy-anion hole Correspondence M. Willemoe ¨ s, Centre for Crystallographic Studies, Department of Chemistry, University of Copenhagen, Universitetsparken 5, Copenhagen DK-2100 Ø, Denmark Fax: +45 3532 0299 Tel: +45 3532 0239 E-mail: martin@ccs.ki.ku.dk (Received 25 November 2004, revised 8 December 2004, accepted 10 December 2004) doi:10.1111/j.1742-4658.2004.04525.x GTP is an allosteric activator of CTP synthase and acts to increase the k cat for the glutamine-dependent CTP synthesis reaction. GTP is suggested, in part, to optimally orient the oxy-anion hole for hydrolysis of glutamine that takes place in the glutamine amidotransferase class I (GATase) domain of CTP synthase. In the GATase domain of the recently published structures of the Escherichia coli and Thermus thermophilus CTP synthases a loop region immediately proceeding amino acid residues forming the oxy-anion hole and named lid L11 is shown for the latter enzyme to be flexible and change position depending on the presence or absence of glutamine in the glutamine binding site. Displacement or rearrangement of this loop may provide a means for the suggested role of allosteric activation by GTP to optimize the oxy-anion hole for glutamine hydrolysis. Arg359, Gly360 and Glu362 of the Lactococcus lactis enzyme are highly conserved residues in lid L11 and we have analyzed their possible role in GTP activation. Characteri- zation of the mutant enzymes R359M, R359P, G360A and G360P indicated that both Arg359 and Gly360 are involved in the allosteric response to GTP binding whereas the E362Q enzyme behaved like wild-type enzyme. Apart from the G360A enzyme, the results from kinetic analysis of the enzymes altered at position 359 and 360 showed a 10- to 50-fold decrease in GTP activation of glutamine dependent CTP synthesis and concomitant four- to 10-fold increases in K A for GTP. The R359M, R359P and G360P also showed no GTP activation of the uncoupled glutaminase reaction whereas the G360A enzyme was about twofold more active than wild-type enzyme. The elevated K A for GTP and reduced GTP activation of CTP synthesis of the mutant enzymes are in agreement with a predicted interaction of bound GTP with lid L11 and indicate that the GTP activation of glutamine dependent CTP synthesis may be explained by structural rearrangements around the oxy-anion hole of the GATase domain. Abbreviations ATPcS, adenosine 5¢-[c-thio]triphosphate; GATase, class I glutamine amidotransferase; PDB ID, Protein Data Bank entry. 856 FEBS Journal 272 (2005) 856–864 ª 2005 FEBS responsible for the hydrolysis of glutamine [2]. In addi- tion the enzyme has a site for GTP that acts as an allosteric activator. The reaction proceeds via the intermediate 4-phosphoryl UTP generated by ATP- dependent phosphorylation [3,4]. The amino group transferred to this activated intermediate is either obtained from glutamine hydrolysis or ammonia pre- sent in the solution [1]. The rate of the glutamine dependent CTP synthesis reaction is greatly stimulated by the allosteric binding of GTP and this activation has been the focus of several reports [5–9]. CTP syn- thase is the only member of the family of GATase domain harboring enzymes where an allosteric effector regulates the glutaminase activity. However, only recently has it been possible functionally to associate individual amino acid residues, Thr-431 and Arg433 in Lactococcus lactis CTP synthase [10] and Arg429 in the Escherichia coli enzyme [11], with this property of the glutamine dependent CTP formation. For the L. lactis enzyme, allosteric binding of GTP acts in syn- ergy with the 4-phoshorylated UTP intermediate to activate glutamine dependent CTP synthesis [12]. In addition, GTP appears to promote channeling of NH 3 derived from glutamine hydrolysis to the synthase site [9]. The nature of the formation of this channel and the role GTP may play in forming this, has recently been suggested from analysis of the crystal structure of the E. coli enzyme. Based on structural homology to GTP binding enzymes, GTP was modeled into the apo-structure of CTP synthase and was suggested to bind in a cleft between the GATase domain and the synthase domain [13] (Fig. 1A). Recently, an E. coli mutant enzyme altered at position 109 (L109A) that is impaired in the coupling between ammonia derived from glutamine hydrolysis and CTP formation due to a constricted or leaky ammonia tunnel [14], has provi- ded evidence for this putative binding site for GTP and the role GTP plays in coupling of glutamine hydrolysis and CTP synthesis [13]. In a work on inhibition of the E. coli enzyme by the analogue glutamate c-semialdehyde that mimics intermediates of glutamine hydrolysis, Bearne and coworkers [6] suggested that the oxy-anion hole is a target for GTP regulation of the glutaminase activity of CTP synthase. A comparison of part of the CTP synthase GATase domain (Fig. 1B) with other known structures of enzymes incorporating this catalytic domain (Fig. 1C) showed that whereas the central beta-sheet and the alpha helices superimpose well, the loop regions differ, including the region that proceeds from the oxy-anion hole into a common alpha helix. This region with a variable loop structure has been named lid L11 in the E. coli CTP synthase structure [13] and we will use this nomenclature when referring to this structural region in the following. The apo- structure of the E. coli enzyme [13] and the structure of the Thermus thermophilus CTP synthase [15] in a complex with glutamine are virtually superimposable. However, the apo-structure and the structure in com- plex with sulfate ions of the T. thermophilus enzyme deviates significantly in the Ca-chain of this region (Fig. 1B). In fact some residues of the lid L11 were not detected in the electron density maps [15] indica- ting that this region is highly flexible and can adapt different structures. Even though there is currently no detailed structural information that explains the activation of glutamine hydrolysis in GATase domains upon binding of amino acceptor substrates [16], it is likely to involve small movements or re-organizations of the environment around the oxy-anion hole [6,17–19]. The importance of the loop corresponding to lid L11 has previously been demonstrated for the E. coli carbamoyl phos- phate synthase that has a cysteine residue (Cys248) within this loop. When this residue is labeled with N-ethylmaleimide [20] or changed to more bulky resi- dues (Arg, Asp, Phe and Trp) [17], it results in uncoupling of glutaminase activity from carbamoyl phosphate synthesis. In the case of the C248D enzyme, glutamine-dependent carbamoyl phosphate synthesis is completely abolished and large increases in glutami- nase activity and K M for glutamine are observed [17]. The dramatic effect on glutamine hydrolysis and coup- ling of the reactions can be explained by local changes in the structure of the region corresponding to lid L11 [21]. Residues 354–357 of the L. lactis enzyme forms part of the oxy-anion hole and homologues residues are found with little variation in all known GATase domains [22]. The corresponding residues from the GATase domain of E. coli CTP synthase [23] were previously subjected to mutational analysis. One mutant enzyme where Gly352 was changed to a pro- lyl (Gly355 in the L. lactis enzyme) had lost detect- able glutamine-dependent CTP synthase activity and could not be labeled by [ 14 C]6-diazo-5-oxonorleucine, but was still active with NH 4 Cl as a substrate. The result from this study may be interpreted in terms of the recent CTP synthase structure, as this residue is positioned at a pivotal point at which the preceding rigid part of the oxy-anion hole connects with the flexi- ble lid L11 (Fig. 1B). Flexibility or displacement of lid L11 may be a prerequisite for the activation of the glutaminase activity hampered by a glycine to a proline substitution at position 352 (355 in the L. lactis enzyme). M. Willemoe ¨ s et al. GTP activation of CTP synthase FEBS Journal 272 (2005) 856–864 ª 2005 FEBS 857 Taken together, the above observations all point to the oxy-anion hole and its immediate surroundings as a target for regulation of GATases. This prompted us to investigate whether highly conserved amino acid residues surrounding the oxy-anion hole plays a role in GTP activation of glutamine hydrolysis by the L. lactis CTP synthase. We chose to analyze by site-directed mutagenesis the role of the highly conserved residues Arg359, Gly360 and Glu362 (Fig. 1D), that are part in lid L11. Arg359 was changed to methionyl to maintain the hydrophobic part of the arginyl side chain while deleting the guanidinium group. The Gly360 to alanyl and Glu362 to glutaminyl substitutions are both allowed for at these positions (Fig. 1D), but do reduce the backbone flexibility or delete the charge of the side chain, respectively. In addition, we replaced Arg359 and Gly360 with a prolyl to test the importance of the flexibility of lid L11 based on the assumption that a prolyl would restrict the flexibility of the Ca-chain at these positions. AB CD Fig. 1. The structure and role of lid L11 in CTP synthase. (A) The proposed binding of GTP modelled into the apo-structure of the E. coli CTP synthase by Endrizzi et al. [13]. The residues are numbered according to the E. coli enzyme while numbers in subscript are according to the L. lactis sequence. Shown are: residues 346–380 including the catalytic Cys379 of the GATase domain, residues 51–55 and residues 104–110 of the synthase domain. (B) Comparison of the structures for T. thermophilus CTP synthase (residues 358–392) in complex with glutamine (PDB ID; 1VCO, light grey) or sulfate (PDB ID; 1VCN, dark grey). In the latter structure residues 365 and 366 (corresponding to 356 and 357 in the L. lactis enzyme) are missing from the structure as indicated by the dotted line. The residues of lid L11 are numbered according to the T. thermophilus sequence while numbers in subscript are according to the L. lactis sequence. The side chain of the catalytic Cys391 is shown for orientation. (C) Comparison of the oxy-anion hole and surroundings of GATase domains from CTP synthase (residues 346–380, PDP ID; 1S1M), carbamoyl phosphate synthase small domain (residues 235–270, PDB ID; 1BXR), GMP synthase (residues 53–87, PDB ID; 1GPM) and anthranilate synthase (residues 48–85, PDB ID; 1QDL) and imidazole glycerol phosphate synthase (residues 47–87, PDB ID; 1JVN).The oxy-anion hole and the loops corresponding to L11 in CTP synthase are encircled. (A), (B), and (C) were prepared with MOLSCRIPT [30] and RASTER3D [31]. (D) A sequence logo [32] was generated of the sequence region of interest (L. lactis CTP synthase num- bering), based on a BLAST [33] alignment of 43 full-length sequences annotated as CTP synthase. The height of each residue is proportional to its frequency, and the most common residue is on top at each position. The total height of each position is adjusted to signify the extent of conservation at that position [34]. GTP activation of CTP synthase M. Willemoe ¨ s et al. 858 FEBS Journal 272 (2005) 856–864 ª 2005 FEBS Results Kinetic constants of substrate nucleotide- and NH 4 Cl saturation All mutant CTP synthases displayed kinetics of satura- tion with ATP, UTP and NH 4 Cl similar to wild-type enzyme (Table 1). However, the G360P enzyme had about a fourfold lower k cat for the NH 4 Cl-dependent CTP synthesis reaction. The fact that the G360P enzyme was partially insoluble in crude cell extracts as mentioned in Experimental procedures could indicate that the reduced NH 4 Cl-dependent activity was due to inactive protein in the enzyme preparation reducing the specific activity. However, the soluble fraction of the G360P enzyme that purified like wild-type enzyme was shown from independent preparations to have a specific activity of 822 ± 56 nmolÆmin )1 Æmg )1 , the same reproducibility of specific activity between enzyme preparations as found for wild-type enzyme. Based on this observation we conclude that the reduced k cat for the NH 4 Cl-dependent activity of G360P (Table 1) reflects the amino acid replacement and not an artefact due to the presence of denatured protein. Kinetic constants of the glutaminase half-reaction Except the R359P and G360P enzymes, all mutant CTP synthases displayed kinetics of glutamine hydro- lysis in the absence of GTP similar to wild-type enzyme (Table 2). The R359P enzyme showed a two- fold lower k cat whereas the G360P enzyme displayed negative cooperativity of glutamine binding (Fig. 2B). By fitting the data to Eqn 2 for the G360P enzyme we calculated a k cat of 0.21 ± 0.04, an S 0.5 of 75±35mm and an n-value of 0.68 ± 0.04. Kinetics of the glutaminase reaction in the presence of GTP (Table 2) revealed that the enzymes R359M and R359P were not activated by GTP under these conditions where the wild-type and E362Q enzymes showed an activation of about 2.5-fold. Compared to wild-type enzyme, the G360A mutant CTP synthase showed an increased GTP-dependent activation of uncoupled glutamine hydrolysis of about fivefold. This increase in GTP activation was further analyzed by varying GTP in the presence of glutamine. How- ever, to prevent inhibition of the glutaminase reaction when increasing the GTP concentration above 1 mm, we included the nucleotides ATPcS and UTP that has been shown to relieve the GTP inhibition while Table 1. Steady state kinetic constants for the NH 4 Cl dependent CTP synthesis reaction of the wild-type and mutant CTP synthases. Assay conditions were as described in Experimental procedures. ATP or UTP were fixed at a concentration of 2 mM. The NH 4 Cl concentration was 100 mM. The concentration of ATP and UTP was 1 mM each. Enzyme Varied substrate ATP UTP NH 4 Cl S 0.5 (lM) nS 0.5 (lM) nK M (mM) I 0.5 (mM) k cat (s )1 ) Wild-type 199 ± 11 2.1 ± 0.2 275 ± 14 1.62 ± 0.05 54 ± 4 225 ± 4 6.3 ± 0.2 R359M 206 ± 6 2.09 ± 0.07 241 ± 4 1.78 ± 0.02 57 ± 6 223 ± 6 6.9 ± 0.3 R359P 175 ± 5 2.5 ± 0.1 178 ± 3 2.6 ± 0.1 92 ± 22 196 ± 12 5.1 ± 0.7 G360A 271 ± 5 1.72 ± 0.03 239 ± 3 1.76 ± 0.02 56 ± 7 212 ± 6 7.5 ± 0.4 G360P 201 ± 11 1.75 ± 0.09 221 ± 4 2.67 ± 0.09 92 ± 22 162 ± 9 1.8 ± 0.2 E362Q 165 ± 3 2.24 ± 0.07 189 ± 4 1.98 ± 0.05 78 ± 8 232 ± 6 10.1 ± 0.5 Table 2. Steady state kinetic constants for the uncoupled hydrolysis of glutamine by wild-type and mutant CTP synthases in the absence or presence of GTP. Assay conditions were as described in Experimental procedures. The GTP concentration was 1 mm. Enzyme Glutaminase – GTP Glutaminase + GTP K M (mM) k cat (s )1 ) K M (mM) k cat (s )1 ) Wild-type 0.94 ± 0.04 0.134 ± 0.002 0.85 ± 0.02 0.334 ± 0.003 R359M 1.29 ± 0.06 0.150 ± 0.003 1.01 ± 0.05 0.148 ± 0.002 R359P 0.74 ± 0.05 0.072 ± 0.001 0.95 ± 0.02 0.0785 ± 0.0006 G360A 0.96 ± 0.02 0.125 ± 0.001 1.05 ± 0.03 0.608 ± 0.006 E362Q 0.87 ± 0.03 0.135 ± 0.002 0.87 ± 0.05 0.389 ± 0.007 M. Willemoe ¨ s et al. GTP activation of CTP synthase FEBS Journal 272 (2005) 856–864 ª 2005 FEBS 859 stimulating the glutaminase reaction about twofold [12]. The results from this experiment are shown in Fig. 3 where GTP activation of uncoupled glutamine hydrolysis of the G360A enzyme is compared to that of wild-type enzyme. Whereas the K A is similar for both mutant and wild-type CTP synthases, the G360A enzyme has an additional 1.6-fold increase in k cat under these conditions compared to wild-type CTP synthase. The presence of GTP did not influence the kinetics of glutamine hydrolysis by the G360P enzyme (Fig. 2B) and we calculated a k cat of 0.22 ± 0.03, a S 0.5 of 78 ± 26 mm and a n of 0.70 ± 0.03 from Eqn (2) (see Experimental procedures). GTP activation of glutamine-dependent CTP synthesis The kinetic constants for GTP activation of the gluta- mine dependent CTP synthesis reaction of wild-type and the various mutant CTP synthases are shown in Table 3. GTP activation of all mutant enzymes was hyperbolic and apart from the G360A and E362Q enzymes that showed normal GTP activation kinetics, the R359M, R359P and G360P enzymes were altered in both GTP binding and activation. The R359M enzyme displayed about fourfold and 10-fold decreases in k cat1 and k cat2 , respectively, with a concomitant increase in K A for GTP of about fivefold. The mutant CTP synth- ases R359P and G360P were similar as k cat1 and k cat2 were reduced about 10-fold and more than 25-fold, respectively, and K A for GTP was increased by about 10-fold. The determined k cat1 and k cat2 for the G360P enzyme should be regarded as ‘apparent’, as this enzyme is probably not saturated with glutamine under the experimental conditions judged from the apparent negative cooperativity of glutamine binding mentioned above. The R359P and G360P enzymes could only be fully analyzed with respect to glutamine-dependent CTP synthesis using the calorimetric assay due to the requirement for high GTP concentrations that are incompatible with the spectrophotometric assay. How- ever, CTP synthesis measured with the spectrophoto- metric assay was found to occur at similar rates as determined in the calorimeter at up to 0.4 mm GTP. Discussion The role of GTP in enhancing the rate of glutamine hydrolysis, both in the absence and presence of sub- strate nucleotides and analogues hereof, has been known for decades [5]. The allosteric regulation of the glutamine dependent reaction is one of the most inter- esting properties of CTP synthase since, as mentioned previously, no other GATase requires an allosteric effector for the hydrolysis of glutamine. However, the mechanisms for activation of glutamine hydrolysis by a second substrate or a reaction intermediate found with other GATases appear allosteric in nature and are crucial for the coupling between glutamine hydro- lysis and synthesis of the aminated product [18,19,24]. Fig. 3. Increase in initial rates for GTP activation of uncoupled gluta- mine hydrolysis determined for the wild-type and G360A enzymes. Experiments were performed as described in Experimental proce- dures. GTP varied as indicated in the presence of 1 m M UTP, 1 mM ATPcS, 20 mM MgCl 2 and 10 mM glutamine. The calorimetric assay only measures the increase in rate from addition of GTP as the basal level of glutamine hydrolysis is part of the baseline. Data were fitted to Eqn (2) by replacing K M with K A . The calculated kin- etic constants for wild-type CTP synthase (d) were; K A ¼ 1.53 ± 0.05 m M and k cat ¼ 1.65 ± 0.03 s )1 and for the G360A enzyme (s) were; K A ¼ 1.38 ± 0.06 mM and k cat ¼ 2.68 ± 0.06 s )1 . Fig. 2. Initial rates of uncoupled glutamine hydrolysis determined for the wild-type and G360P enzymes. Experiments were per- formed as described in Experimental procedures. Glutamine varied as indicated for wild-type (A) and G360P (B) enzymes in the absence (s) or presence (d)of1m M GTP and 20 mM MgCl 2 .The calculated kinetic constants for wild-type enzyme are given in Table 2. For the G360P enzyme see Results. GTP activation of CTP synthase M. Willemoe ¨ s et al. 860 FEBS Journal 272 (2005) 856–864 ª 2005 FEBS The recent publication of the E. coli [13] and T. thermophilus [15] CTP synthase structures in combi- nation with previously published mutational analysis of CTP synthases [10,14,25] has allowed for a much more detailed understanding of the contribution of various amino acid residues to the catalysis and regula- tion of this enzyme. From analysis of the structure of the T. thermophilus enzyme [15], it is suggested that the binding of ATP and UTP to the enzyme brings together the GATase and the synthase domains and enables the formation of the putative ammonia tunnel. An observation that supports this analysis is that conserved sequence motifs of the primary structure involved in GTP activation [10] and coupling of gluta- mine hydrolysis and CTP formation [25] are brought closer together in the structure. Other reports also sup- port such a structural transition and have demonstra- ted the importance of GTP in both ammonia tunnel formation and coupling between glutamine hydrolysis and CTP formation [6,9,12]. Apart from Bearne and co-workers [6] that suggested a link between GTP activation and the oxy-anion hole of CTP synthase and despite the progress in the under- standing of the mechanism of CTP synthase provided by the recent crystal structures [13,15], no detailed explanation for the way GTP binding stimulates the glutamine hydrolysis reaction has been offered. In this work we have performed a mutational analysis of lid L11 to test whether the presence of highly conserved residues in this region, close to the oxy-anion hole, could be involved in the regulation of CTP synthase. The E362Q enzyme did not show major changes in catalytic properties compared to wild-type enzyme. Substitution of the residues Arg359 and Gly360 affected the GTP activation of both the uncoupled glutaminase activity (Table 2 and Fig. 2B) and gluta- mine dependent CTP formation (Table 3). Replacing Gly352 of the E. coli enzyme with a prolyl (Gly355 in the L. lactis enzyme) abolish glutamine-dependent CTP synthesis [23]. This finding agrees well with the results from kinetic characterization of the L. lactis R359P and G360P enzymes and may indicate that mobility or flexibility within this region is crucial to both glutamine hydrolysis and GTP activation of this reaction. The uncoupled glutaminase activity of both these mutant enzymes no longer responds to the pres- ence of GTP (Table 2) and the activation mediated by GTP on glutamine-dependent CTP synthesis (Table 3) is almost absent in the R359P and G360P enzymes. In addition, both enzymes show a concomitant increase in K A for GTP of about 10-fold for the glutamine dependent CTP formation compared to wild-type enzyme, indicating that the affinity of these enzymes for the activator has been reduced due to the changes in lid L11. The apparent negative cooperativity of glu- tamine binding and greatly lowered affinity for this substrate (Fig. 2B) together with the reduced activity of the NH 4 Cl-dependent CTP synthesis reaction (Table 1) found for the G360P enzyme, demonstrates a close link between the glutamine- and NH 4 Cl-depend- ent reactions. This has previously been demonstrated by the exclusion of external ammonia from the active site by the combination of glutamine, or glutamate c-semialdehyde, and GTP [6,9]. In the G360P enzyme lid L11 may be ‘frozen’ in a position that hinders both the access to the glutamine binding site and the entry for external ammonia. This is not unlikely as the resi- dues in lid L11 (Fig. 1A) are found to close over the glutamine binding site in the GATase domain (Phe356 in L. lactis CTP synthase) and to be flanking the entry hole for externally provided ammonia that overlaps with the putative GTP binding site (Phe356, Gly357 and Arg359 in L. lactis CTP synthase) [13]. The lack of GTP activation of the glutaminase activ- ity (Table 2), the elevated K A for GTP binding and the reduced k cat2 (Table 3) for glutamine-dependent CTP Table 3. Steady state kinetic constants for GTP activation of the glutamine dependent CTP synthesis reaction of wild-type and mutant CTP synthases. Assay conditions were as described in Experimental procedures. The concentration of glutamine was 10 mM. The concentration of ATP and UTP were each 1 mM. Values in parentheses are obtained from the calorimetric assay. The calorimetric assay only measures the response of the reaction to the addition of GTP. For this reason k cat1 can not be determined as it is zeroed as part of the baseline. ND, not determined. Could not be determined for technical reasons (see text). Enzyme Kinetic constant K A (lM) k cat1 (s )1 ) k cat2 (s )1 ) Wild-type 108 ± 5 (132 ± 3) 0.11 ± 0.03 5.18 ± 0.07 (5.20 ± 0.05) R359M 495 ± 62 (462 ± 30) 0.024 ± 0.001 0.44 ± 0.03 (0.53 ± 0.02) R359P ND (1239 ± 64) 0.012 ± 0.002 ND (0.118 ± 0.004) G360A 89 ± 10 (90 ± 4) 0.2 ± 0.1 6.6 ± 0.2 (5.53 ± 0.08) G360P ND (1468 ± 35) 0.010 ± 0.001 ND (0.215 ± 0.003) E362Q 117 ± 18 (161 ± 5) 0.1 ± 0.1 6.0 ± 0.3 (6.13 ± 0.08) M. Willemoe ¨ s et al. GTP activation of CTP synthase FEBS Journal 272 (2005) 856–864 ª 2005 FEBS 861 formation that is observed upon deletion of the charge of the side chain in the R359M enzyme, is in accord- ance with the current model of GTP binding to CTP synthase (Fig. 1A). The results support that Arg359 directly interacts with the activator [13] and may act as a lever inducing a conformational change in lid L11 to an activating position upon GTP binding. The sensitiv- ity of GTP regulation of uncoupled glutaminase activ- ity to replacement of amino acid residues in lid L11 is also demonstrated by the observation that increased activation by GTP for this reaction was found for the enzyme G360A (Table 2 and Fig. 3). We interpret this result in terms of a small favorable conformational change in lid L11 of the G360A enzyme, which enhan- ces the effect of binding the activator. In conclusion, our results support the model of GTP binding at the predicted site (Fig. 1A) at the interface between the GATase domain and the synthase domain [13]. Amino acid residues in the synthase domain and residues of lid L11 forms the platform at the domain interface for binding of GTP. The interaction of amino acid residues of lid L11 with bound GTP mediates the activation of glutaminase activity. At the same time, GTP also serves to close the ammonia tunnel and opti- mize this for channeling of ammonia from the GATase domain to the synthase domain [14]. Experimental procedures Site-directed mutagenesis and DNA sequencing Site-directed mutagenesis of the L. lactis pyrG gene was performed using the QuickChange method (Stratagene, AH Diagnostics, Aarhus, Denmark) and the comple- mentary deoxy-oligonucleotides LL5-R359M; GGCTTTG GTCAAATGGGAACAGAAGGT, LL3-R359M; ACCTTC TGTTCCCATTTGACCAAAGCC, LL5-R359P; GGCTTT GGTCAACCGGGAACAGAAGGT, LL3-R359P; ACCTT CTGTTCCCGGTTGACCAAAGCC, LL5-G360A; TTTGG TCAACGTGCAACAGAAGGTAAG, LL3-G360A; CTTA CCTTCTGTTGCACGTTGACCAAA, LL5-G360P; TTTGG TCAACGTCCAACAGAAGGTAAG, LL3-G360P; CTTAC CTTCTGTTGGACGTTGACCAAA, LL5-E362Q; CAACG TGGAACACAAGGTAAGATTGCA and LL3-E362Q; TGCAATCTTACCTTGTGTTCCACGTTG for construc- tion of the alleles encoding the R359M, R359P, G360A, G360P and E362Q enzymes, respectively. Lettering in italics indicates the base changes introduced by the oligonucleo- tides. Plasmid pMW602 [26] served as a template for muta- genesis. The mutations were verified by sequencing of the entire coding region of L. lactis pyrG using an ABI PRISM 310 DNA Sequencer as recommended by the supplier (Perk- inElmer, Danmark A ⁄ S, Hvidovre, Denmark). Protein purification and enzyme assays All chemicals were purchased from Sigma-Aldrich Denmark A ⁄ S (Brondby, Denmark). The L. lactis wild- type and mutant pyrG alleles were expressed and the enco- ded CTP synthases were purified to homogeneity [26] as judged by SDS ⁄ PAGE [27]. About 50% of the total syn- thesized G360P protein, was insoluble and was sedimented with the cell debris after sonication of the over expressing culture. The soluble fraction of G360P could be purified by the same procedure as for wild-type enzyme [26]. All enzymes were stable during storage and could be handled like wild-type enzyme while performing the various assays. Assays were performed at 30 ° Cin50mm Hepes, pH 8.0, 2mm dithiothreitol. The used concentrations of the assay components ATP, ATPcS, UTP, GTP, MgCl 2 , glutamine or NH 4 Cl were as described under Results. For the spec- trophotometric measurement of CTP synthesis, conversion of UTP to CTP with De 291 ¼ 1338 cm )1 Æm )1 was recorded as previously described [26,28]. The isothermal titration calorimetric assay for CTP synthesis or glutamine hydro- lysis was performed as described in detail elsewhere [12]. Analysis of enzyme kinetic data Calculation of kinetic constants was performed by fitting the initial velocities to one of the four equations below using the computer program ultrafit (BioSoft, Cambridge, UK, ver- sion 3.01). The reported standard errors are those calculated by the computer program. Equations (1) and (2) apply to hyperbolic and sigmoid substrate saturation kinetics, respect- ively. Equation (3) applies to hyperbolic activation kinetics. Equation (4) applies to cooperative substrate inhibition: v ¼ k cat ½S=ðK M þ½SÞ ð1Þ v ¼ k cat ½S n =ðS n 0:5 þ½S n Þð2Þ v ¼ k cat1 þ k cat2 ½A=ðK A þ½AÞ ð3Þ v ¼ k cat ½S=ðK M þ½Sþ½Sf½S=I 0:5  n Þð4Þ where v is the initial rate, k cat is the apparent catalytic con- stant, K M and K A are the apparent Michaelis–Menten con- stants for substrate S or activator A, respectively, S 0.5 is the concentration of substrate S at apparent half maximal velo- city, n is the Hill coefficient, k cat1 and (k cat1 +k cat2 ) are the apparent catalytic constants in the absence and presence of saturating concentrations of activator, respectively, I 0.5 is the substrate concentration for half maximal substrate inhi- bition. For Eqn (4), n was fixed at a value of 4 as deter- mined from analysis of initial velocity data obtained by varying NH 4 Cl at several equimolar concentrations of ATP and UTP [29]. GTP activation of CTP synthase M. Willemoe ¨ s et al. 862 FEBS Journal 272 (2005) 856–864 ª 2005 FEBS Acknowledgements The authors wish to thank Enoch P. Baldwin (Univer- sity of California, Davies, CA) for providing the structural coordinates (1S1M) for the E. coli CTP syn- thase in advance of publication as well as the coordi- nates for the modelled GTP molecule. We appreciate the expert technical assistance by Dorthe Boelskifte. Sine Larsen is thanked for support to MW through a grant from the Danish National Research Foundation. References 1 Zalkin H (1985) CTP synthetase. Methods Enzymol 113, 282–287. 2 Weng M, Makaroff CA & Zalkin H (1986) Nucleotide sequence of Escherichia coli pyrG encoding CTP synthe- tase. J Biol Chem 261, 5568–5574. 3 von der Saal W, Anderson PM & Villafranca JJ (1985) Mechanistic investigations of Escherichia coli cytidine- 5¢-triphosphate synthetase: detection of an intermediate by positional isotope exchange experiments. J Biol Chem 260, 14993–14997. 4 Lewis DA & Villafranca JJ (1989) Investigation of the mechanism of CTP synthetase using rapid quench and isotope partitioning methods. Biochemistry 28, 8454– 8459. 5 Levitzki A & Koshland DE Jr (1972) Role of an allo- steric effector: guanosine triphosphate activation in cytosine triphosphate synthetase. Biochemistry 11, 241– 246. 6 Bearne SL, Hekmat O & Macdonnell JE (2001) Inhibi- tion 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. 7 Levitzki A, Stallcup WB & Koshland DE Jr (1971) Half-of-the-sites reactivity and the conformational states of cytidine triphosphate synthetase. Biochemistry 10, 3371–3378. 8 Kizaki H, Ohsaka F & Sakurada T (1982) Role of GTP in CTP synthetase from Ehrlich ascites tumor cells. Biochem Biophys Res Commun 108, 286–291. 9 Willemoes M (2004) Competition between ammonia derived from internal glutamine hydrolysis and hydroxyl- amine present in the solution for incorporation into UTP as catalysed by Lactococcus lactis CTP synthase. Arch Biochem Biophys 424, 105–111. 10 Willemoes M (2003) Thr-431 and Arg-433 are part of a conserved sequence motif of the glutamine amidotrans- ferase domain of CTP synthases and are involved in GTP activation of the Lactococcus lactis enzyme. J Biol Chem 278, 9407–9411. 11 Simard D, Hewitt KA, Lunn F, Iyengar A & Bearne SL (2003) Limited proteolysis of Escherichia coli cytidine 5¢-triphosphate synthase: identification of residues required for CTP formation and GTP-dependent activa- tion of glutamine hydrolysis. Eur J Biochem 270, 2195– 2206. 12 Willemoes M & Sigurskjold BW (2002) Steady-state kinetics of the glutaminase reaction of CTP synthase from Lactococcus lactis. Eur J Biochem 269, 4772– 4779. 13 Endrizzi JA, Kim H, Anderson PM & Baldwin EP (2004) Crystal structure of Escherichia coli cytidine tri- phosphate synthetase, a nucleotide-regulated glutamine amidotransferase ⁄ ATP-dependent amidoligase fusion protein and homologue of anticancer and antiparasitic drug targets. Biochemistry 43, 6447–6463. 14 Lunn FA & Bearne SL (2004) Alternative substrates for wild-type and L109A E. coli CTP synthases. Eur J Bio- chem 271, 4204–4212. 15 Goto M, Omi R, Nakagawa N, Miyahara I & Hirotsu K (2004) Crystal structures of CTP synthetase reveal ATP, UTP, and glutamine binding sites. Structure (Camb) 12, 1413–1423. 16 Thoden JB, Huang X, Raushel FM & Holden HM (1999) The small subunit of carbamoyl phosphate synthetase: snapshots along the reaction pathway. Biochemistry 38, 16158–16166. 17 Mareya SM & Raushel FM (1994) A molecular wedge for triggering the amidotransferase activity of carbamoyl phosphate synthetase. Biochemistry 33, 2945–2950. 18 Miles BW, Banzon JA & Raushel FM (1998) Regula- tory control of the amidotransferase domain of carbamoyl phosphate synthetase. Biochemistry 37, 16773–16779. 19 Myers RS, Jensen JR, Deras IL, Smith JL & Davisson VJ (2003) Substrate-induced changes in the ammonia channel for imidazole glycerol phosphate synthase. Biochemistry 42, 7013–7022. 20 Wellner VP & Meister A (1975) Enhancement of the glutaminase activity of carbamyl phosphate synthetase by alterations in the interaction between the heavy and light subunits. J Biol Chem 250, 3261–3266. 21 Thoden JB, Huang X, Kim J, Raushel FM & Holden HM (2004) Long-range allosteric transitions in carbamoyl phosphate synthetase. Protein Sci 13, 2398–2405. 22 Chittur SV, Klem TJ, Shafer CM & Davisson VJ (2001) Mechanism for acivicin inactivation of triad glutamine amidotransferases. Biochemistry 40, 876–887. 23 Weng ML & Zalkin H (1987) Structural role for a con- served region in the CTP synthetase glutamine amide transfer domain. J Bacteriol 169, 3023–3028. 24 Kim JH, Krahn JM, Tomchick DR, Smith JL & Zalkin H (1996) Structure and function of the glutamine phos- phoribosylpyrophosphate amidotransferase glutamine site and communication with the phosphoribosylpyro- phosphate site. J Biol Chem 271, 15549–15557. M. Willemoe ¨ s et al. GTP activation of CTP synthase FEBS Journal 272 (2005) 856–864 ª 2005 FEBS 863 25 Iyengar A & Bearne SL (2003) Aspartate-107 and leucine-109 facilitate efficient coupling of glutamine hydrolysis to CTP synthesis by Escherichia coli CTP synthase. Biochem J 369, 497–507. 26 Wadskov-Hansen SL, Willemoes M, Martinussen J, Hammer K, Neuhard J & Larsen S (2001) Cloning and verification of the Lactococcus lactis pyrG gene and characterization of the gene product, CTP synthase. J Biol Chem 276, 38002–38009. 27 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 28 Long CW & Pardee AB (1967) Cytidine triphosphate synthetase of Escherichia coli B. I. Purification and kinetics. J Biol Chem 242, 4715–4721. 29 Willemoes M & Larsen S (2003) Substrate inhibition of Lactococcus lactis cytidine 5(¢)-triphosphate synthase by ammonium chloride is enhanced by salt-dependent tetramer dissociation. Arch Biochem Biophys 413, 17–22. 30 Kraulis PJ (1991) Molscript – a program to produce both detailed and schematic plots of protein structures. J Appl Crystallogr 24, 946–950. 31 Merrit EA (1994) Raster3D version 2.0. A program for photorealistic molecular graphics. Acta Crystallogr D Biol Crystallogr 50, 869–873. 32 Schneider TD & Stephens RM (1990) Sequence logos: a new way to display consensus sequences. Nucleic Acids Res 18, 6097–6100. 33 Altschul SF, Gish W, Miller W, Myers EW & Lipman DJ (1990) Basic local alignment search tool. J Mol Biol 215, 403–410. 34 Shannon CE (1948) A mathematical theory of commu- nication. Bell System Tech J 27, 379–423. GTP activation of CTP synthase M. Willemoe ¨ s et al. 864 FEBS Journal 272 (2005) 856–864 ª 2005 FEBS . Lid L11 of the glutamine amidotransferase domain of CTP synthase mediates allosteric GTP activation of glutaminase activity Martin Willemoe ¨ s 1 ,. at the domain interface for binding of GTP. The interaction of amino acid residues of lid L11 with bound GTP mediates the activation of glutaminase activity.