Structural Role for a Conserved Region in the CTP Synthetase Glutamine Amide Transfer Domaint

Home , CTP synthetase

JOURNAL OF BACTERIOLOGY, JUIY 1987, p. 3023-3028 Vol. 169, No. 7 0021-9193/87/073023-06$02.00/0 Copyright C 1987, American Society for Microbiology

Structural Role for a Conserved Region in the CTP Synthetase Glutamine Amide Transfer Domaint MANLI WENG AND HOWARD ZALKIN* Department ofBiochemistry, Purdue University, West Lafayette, Indiana 47907 Received 22 December 1986/Accepted 20 April 1987

Site-directed mutations were introduced into a conserved region of the Escherchia coli CTP synthetase glutamine amide transfer domain. The amino acid replacements, valine 349 to serine, glycine 351 to alanine, glycine 352 to proline, and glycine 352 to cysteine, all increased the lability of CTP synthetase. The proline 352 replaceipent abolished the capacity to form the covalent glutaminyl-cysteine 379 catalytic intermediate, thus preventing glutamine amide transfer function; NH3-dependent CTP synthetase activity was retained. In CTP synthetase (serine 349), both glutamine and NH3-dependent activities were increased approximately 30% relative to that of the wild type. CTP synthetase mutants alanine 351 and cysteine 352 were not overproduced because of a1pparent instability and proteolytic degradation. We conclude that the conserved region between residues 346 and 355 in the CTP synthetase glutamine amide transfer domain has an imnportant structural role.

CTP synthetase catalyzes the final reaction in the de novo This paper reports the replacement of conserved residues pathway for pyrimidine nucleotide synthesis: UTP + ATP + in region I of the CTP synthetase GAT domain. One replace- glutamine -* CTP + ADP + Pi + glutamlate. Similar to other ment, Gly 352 to Pro, specifically inactivated GAT function glutamine amidotransferases (4, 28), purified CTP synthetase with little, if any, effect on aminator function. Other replace- can utilize NH3 as a substrate in place of glutamine, in which ments, Val 349 to Ser, Gly 351 to Ala, and Gly 352 to Cys, case the products of the reaction are CTP + ADP + Pi. We caused enzyme lability. These results support the view that recently reported the nucleotide sequence of Escherichia amino acid residues in conserved region I have an essential coli pyrG encoding CTP synthetase (26). By sequence com- structural role in the GAT domain. parisons with other glutamine amidotransferases, functional domains for the NH3-dependent reaction and for glutamine AND amide transfer (GAT) were identified. The CTP synthetase MATERIALS METHODS domain structure is NH2-aminator-GAT-CO2H, in which the Bacteria and plasmids. E. coli JF646 (pyrE pyrG cdd argE first 290 to 300 amino acid residues provide the aminator his4 leu-6 proA thr-J thi-J recA (5), provided by James function (UTP + ATP + NH3 -_ CTP + ADP + Pi) and the distal 245 amino acids contribute the GAT function. The role Friesen, University of Toronto, Toronto, Ontario, Canada, was the host for pyrG plasmids. E. coli JM101, JM105, and of the GAT domain is to bind glutamine, form a covalent GM119 for growth of M13 phage have been previously glutaminyl intermediate, and transfer the amide of glutamine described (13, 17). The growth rates of strain JF646 bearing to the aminator domain as a source of NH3 (15). pyrG+ plasmids were determined in M9 medium (14) supple- Two types ofGAT domain have been identified in different mented with amino acids at 50 of uracil A conserved GAT domain designated required ,ug/ml, 22 ,ug amidotransferases. per ml, 2 ,ug of thiamine per ml, NH4Cl as indicated, and 50 trpG type is found in anthranilate synthase (16, 23), p- ,ug of ampicillin per ml. For enzyme preparation, plasmid- aminobenzoate synthase (6, 7), carbamoyl-phosphate synthetase (19, 27), GMP synthetase (22, 29), and CTP synthe- bearing strain JF646 was grown as previously described (26). Plasmid pMWS (pyrG+), a pUC8 derivative, was previously tase (26). A purF-type GAT domain is in amidophosphoribo- described (26). Mutations in CTP synthetase were con- syltransferase (24), glucosamine 6-phosphate synthase (25), structed by site-directed mutagenesis to yield plasmids des- and asparagine synthetase (I. Andrulis, personal communi- ignated pMW7 (Ser 349), pMW8 (Ala 351), pMW9 (Cys 352), cation). The trpG-type GAT domain has three blocks of conserved and pMW10 (Pro 352) as described below. These plasmids are pUC8 derivatives. The pUC8 recombinant with wild- amino acids (1, 26, 27). Conserved regions II and III con- type pyr7 is pMW6. tribute essential cysteine and histidine residues, respectively, which participate directly in catalysis! (1, 17). In CTP Site-directed mutagenesis. The following deoxyribonucleo- tides, complerientary to the mRNA coding strand, were synthetase, 379 and histidine 515 are presumably in cysteirle synthesized on an Applied Biosystems model 380 DNA close proximrity to bound glutamine and are inferred to in synthesizer (base mismatches underlined): 5'-GCAAT participate formation of the covalent cysteinyl-glutamine CCTCTCACCTGGCGG-3', codon change GTA to TCA, catalytic intermediate. There is presently no information on the Ser 349; 5'-CGTACCTGCCGGTTTCG-3', codon change role of conserved amino acids in region I of the GAT GGC to GCC, Ala 351; 5'-TACCTGGCTGTTTCGGC-3', domain. Mutations in region I have not previously been isolated. codon change GGT to TGT, Cys 352; 5'-GTACCTGGCCC TTTCGGCTA-3', codon change GGT to CCT, Pro 352. Purification of oligodeoxynucleotides was by 20% polyacrylamide-7 M urea gel electrophoresis and gel filtration. * Corresponding author. The gapped heteroduplex method of Bauer et al. (3) was t Journal paper 11003 from the Purdue University Agricultural used for mutagenesis as previously described (1, 12, 17). The Experiment Station. gapped heteroduplex (Fig. 1) was constructed from 3023 3024 WENG AND ZALKIN J. BACTERIOL.

M13mpll (wild type) replicative form, cut at the BamHI and 350 355 PstI polylinker sites, and M13mpll viral DNA containing a 1,981-base-pair (bp) BamHI-PstI fragment of pyrG cloned into the corresponding polylinker sites. Mutant phage were screened by single-lane dideoxy sequencing with a synthetic 17-mer oligodeoxynucleotide sequencing primer that an- A GMPS I I L S G G P E S T nealed approximately 50 bases upstream of the region sub- ject to mutagenesis. All mutations were confirmed by dide- PABS II I V I S P G P C T P oxy sequencing (21) for about 40 bp upstream and 150 bp downstream of the site of replacement. Intact pyrG was reconstructed in pUC8 in two steps (Fig. 1). (i) The 461-bp ASII L M L S P G P G V P NruI-BamHI 5' segment of pyrG plus flanking DNA was isolated from pMW5 and ligated into the SmaI and BamHI CPS I F L S N G P G D P sites of pUC8 to yield a plasmid designated pUC8-NB461. (ii) The 1,981-bp BamHI-PstI distal region ofpyrG (wild type G V L I P G G F S Y or mutant) was isolated from the M13mpll recombinant and FGAMS ligated into the BamHI and PstI sites of pUC8-NB461 to regenerate intact pyrG in pUC8. In this construction, tran- CTPS A I L V P G G F G Y scription is likely from the pyrG promoter (26) and translation is from the pyrG translation initiation site. B Ser 349 S Enzyme purification and assay. CTP synthetase was purified to electrophoretic homogeneity from 1-liter cultures of strain JF646 bearing plasmid pMW6, pMW7, or pMW10 Ala 351 A (26). CTP synthetase activity was determined at 23°C by the spectrophotometric method of Long and Pardee (11) by Cys 352 C using the assay mixture specified by Anderson (2). A unit of activity corresponds to production of 1 ,umol of CTP/min at 230C. Pro 352 P FIG. 2. Alignment of amino acids in GAT domain conserved Reaction with DON. Enzyme was incubated at 0°C with a region I and amino acid replacements. (A) Alignment of amino acids in E. coli GMP synthetase (GMPS) (29), E. coli p-aminobenzoate H P R synthase component II (PABS II) (7), E. coli anthranilate synthase l component II (ASII) (16), E. coli carbamoyl phosphate synthetase Gapped (CPS) (19), B. subtilis formylglycinamidine ribonucleotide synthe- Heteroduplex + tase (FGAMS) (D. J. Ebbole, unpublished data), and E. coli CTP Th) synthetase (CTPS) (26). The numbering refers to positions in CTPS. (B) Amino acid replacements in CTP synthetase.

RSBPH SoN B four- to fivefold molar excess (relative to the concentration pUC8 loc?9l pMW5 NOR= of enzyme subunits of 14C-labeled 6-diazo-5-oxonorleucine (DON; specific activity, 1,260 cpm/nmol) in a 20-,ul reaction SinaI Nru I mixture containing 50 mM KPO4 buffer [pH 7.4], 10 mM Bm HI Bom HI MgCl2, 0.2 mM GTP, 0.89 nmol of enzyme subunit, and DON. The reaction was stopped with 50 ,ul of 1.6 mM B PH unlabeled DON. Centrifugal gel ifitration (18) was used to pUC8-NB461 separate the radioactive enzyme from free DON. Immunoprecipitation. Rabbit antiserum was prepared with Ban HI purified CTP synthetase as previously described (20). For Pst I enzyme purity, see Fig. 5, lane 5. For radiolabeling, 1.0 ml of add Ban HI-PstI ftamet from M13 RF log-phase cells was grown with 100 ,Ci of [35S]methionine (800 Ci/mmol) for 45 min as previously described (12) and sonically disrupted, and CTP synthetase was immunoprecip- B P H S/N I I itated. Proteins were resolved by sodium dodecyl sulfate- pMW6-pMW10O ,apoR polyacrylamide gel electrophoresis (8). FIG. 1. Schematic diagram for site-directed mutagenesis ofpyrG in an M13mpll heteroduplex and construction of pyrG-pUC8 re- RESULTS combinants. The top part shows the gapped heteroduplex that was used for site-directed mutagenesis. An oligonucleotide (solid line) Effects of amino acid replacements on enzyme activity. The with a one- or two-base mismatch (black dot) was annealed to a data in Fig. 2 and Table 1 summarize the four amino acid complementary segment ofpyrG. The two strands ofthe M13 vector replacements that were made in CTP synthetase region I. are shown, along with cross-hatched polylinker. The gap was filled Val 349 to Ser makes CTP synthetase identical at position in with DNA polymerase Klenow fragment. For the p1.smid deriv- 349 to four of the other amidotransferases having a trpG-type atives, the dark box is pyrG coding DNA and the open box is E. coli Ser flanking DNA. Only selected restriction enzyme sites are shown. GAT domain. The specific activity of purified 349 Abbreviations: H, HindIII; P, PstI; B, BamHI; R, EcoRI; S, SmaI; enzyme was increased approximately 30% over that of the Sa, SaII; N, NruI; RF, replicative form. The virgule represents a wild-type CTP synthetase (Table 1). Position 349 in the CTP junction in which the two sites are no longer present. Transcription synthetase GAT domain can thus accommodate hydrophilic of pyrG is from left to right in plasmids pMW6 to pMW10. as well as nonpolar amino acid side chains. The change of VOL. 169, 1987 CTP SYNTHETASE GAT DOMAIN 3025

TABLE 1. Activities of purified wild-type E. coli CTP synthetase a and mutant enzymes 100 Amino acid Enzyme activity (U/mg) replacement in CTP Gln CTP NH3 CTP 50 synthetase synthetase synthetase

None (wild type) 1.4 0.48 > 20 Ser 349 1.8 0.64 Ala 351 Np'a NP > 10 Cys 352 NP NP > Pro 352 0 0.38 w a NP, the Ala 351 and Cys 352 enzymes were not purified. 5[

Gly 352 to Pro also makes CTP synthetase identical to four of the other amidotransferases at this position. However, in contrast to the Ser 349 enzyme, GAT function was abolished in Pro 352 CTP synthetase. The enzymatic defect is ascribed to a structural alteration since neither glycine nor proline has TIME (minues) TIME (minutes) FIG. 4. Probe of enzyme conformation by heat inactivation. (a) functional groups that could participate in catalysis. The NH3-dependent CTP synthetase. Wild-type and mutant enzymes structural defect in Pro 352 CTP synthetase is in the GAT were incubated at 50°C and assayed at 22'C as for Fig. 3 but with domain since the NH3-dependent activity of Pro 352 CTP omission of trypsin. Symbols: 0, wild type; 0, Ser 349 mutant; 2, synthetase was similar to that of the wild-type enzyme Pro 352 mutant. (b) CTP synthetase was inactivated at 50°C and (Table 1). Neither glutamine- nor NH3-dependent CTP syn- assayed for glutamine-dependent activity at 22°C. Symbols: 0, wild thetase activity was detected in extracts of strain JF646 type; 0, Ser 349 mutant. bearing plasmid pMW9 or plasmid pMW8 encoding the Cys 352 and Ala 351 mutant enzymes, respectively. Probes of enzyme structure. Two empirical probes of that expose lysine and arginine residues on the surface of the enzyme structure, proteolysis by trypsin and heat inactiva- enzyme. Heat inactivation of NH3-dependent CTP synthe- tion, were used to characterize the Ser 349 and Pro 352 tase gave similar results, although biphasic inactivation mutant enzymes. The data (Fig. 3) show that proteolytic kinetics were obtained for the mutant enzymes. Both CTP inactivation of NH3-dependent CTP synthetase for the wild- synthetase with Pro 352 and that with Ser 349 were 99o type and Pro 352 mutant enzymes was apparently a first- inactivated at 50°C in 20 min or less, whereas wild-type CTP order process. The Pro 352 substitution in the CTP synthe- synthetase retained more than 70% of its initial NH3- tase GAT domain increased the rate of inactivation fourfold, dependent activity under these conditions (Fig. 4a). Further from 0.18 to 0.77 min-'. The rate of proteolytic inactivation evidence for structural alteration in CTP synthetase (Ser of CTP synthetase (Ser 349) was marginally increased. 349) came from heat inactivation of its glutamine-dependent Increased rates of proteolysis result from structural changes activity. Figure 4b shows biphasic inactivation of glutamine- dependent CTP synthetase (Ser 349). In 30 min, the Ser 349 enzyme was 99% inactivated compared with about 36% 10 5 inactivation of the wild type. We conclude that the Ser 349 and Pro 352 mutations alter the structure of the GAT 50 domain, and these alterations can be transmitted to the aminator domain. 5.0 Detection of CTP synthetase protein. CTP synthetase was

2 readily resolved and visualized by sodium dodecyl sulfate- > polyacrylamide gel electrophoresis of crude extracts from plasmid-bearing cells. The photograph of a Coomassie blue- I0 stained gel (Fig. 5) shows overproduction of CTP synthetase Pro 352 (lane 1) and Ser 349 (lane 4) in crude extracts. Clearly, CTP synthetase is the major protein in the extracts of these two plasmid-bearing strains. An extract of strain JF646 carrying plasmid pMW8 appears to contain lesser 2- amounts of the Ala 351 enzyme (lane 3). The Cys 352 enzyme is not visible in lane 2. We conclude that these replacements may result in structural alterations that render 0 5 10 the Ala 351 and Cys 352 mutant enzymes hypersensitive to TIME (minutes) proteolytic degradation. The latter two mutant enzymes FIG. 3. Probe of enzyme conformation by trypsin digestion. were not purified. Mutant and wild-type CTP synthetases were treated with trypsin in To the a 60-Il reaction mixture containing 0.06 M HEPES (N-2- verify synthesis of unstable mutant enzymes, hydroxyethylpiperazine-N'-2-ethanesulfonic acid; pH 8.0), 0.01 M plasmid-bearing cells were labeled with [35S]methionine. MgCl2, 0.005 M EDTA, 74 ,ug of CTP synthetase, and 1.0 ,ug of CTP synthetase was immunoprecipitated and resolved by trypsin. Samples (12 FLl) were removed into a 1.0-ml CTP synthetase sodium dodecyl sulfate-polyacrylamide gel electrophoresis. assay mixture for determination of the initial rate of NH3-dependent The fluorograph (Fig. 6) shows that the level of Ala 351 CTP synthase. The incubation and assay temperature was 22°C. enzyme (lane 4) was comparable to those of the Ser 349 and Symbols: 0, wild type; 0, Ser 349 mutant; O, Pro 352 mutant. Pro 352 enzymes. However, only a trace level of the Cys 352 3026 WENG AND ZALKIN J. BACTERIOL.

a .: *'X't;A :,t

Or4 4. *9.AM

.,z. '. ... .-i

FIG. 5. Content of CTP synthetase in crude cell extracts of plasmid-bearing strains. Electrophoresis was on sodium dodecyl sulfate-10% polyacrylamide gels. Lanes: 1, 245 jig of protein, Pro FIG. 6. Immunoprecipitation of mutant and wild-type CTP syn- 352 enzyme; 2, 240 jig of protein, Cys 352 enzyme; 3, 310 pg of thetase. To each lane was added approximately 2.1 x 105 cpm of protein, Ala 351 enzyme; 4, 255 ju.g of protein, Ser 349 enzyme; 5, 80 [35S]methionine-labeled cell extract. Lanes: 1, wild type; 2, Pro 352 ,ug of purified CTP synthetase. The gel was stained with Coomassie mutant; 3. Cys 352 mutant; 4, Ala 351 mutant; 5, Ser 349 mutant. blue. The molecular masses of the protein standards are shown in kilodaltons. enzyme was detected (lane 3). A common pattern of faintly visible protein bands, presumed to be contaminants, was transfer function. Poor growth of cells having CTP synthe- present in all samples. In addition, several unique protein tase (Cys 352) in medium with 50 mM NH4Cl is consistent bands were detected in lanes 3 and 4, containing the Cys 352 with the very low enzyme content noted by immunoprecip- and Ala 351 enzymes, respectively. These proteins, marked itation (Fig. 6). by arrowheads in lanes 2 and 3, are candidates for proteo- Affinity labeling with DON. The glutamine affinity analog lytic degradation products. DON inactivates glutamine-dependent CTP synthetase by Growth of wild-type and mutant strains. In vivo function of alkylation (9, 10) of the inferred essential cysteinyl residue at mutant CTP synthetase was probed by measurements of the position 379 (26) in region II of the glutamine amide transfer growth rate of pyrG strain JF646 bearing plasmids encoding domain. This alkylation reaction mimics formation of the the wild-type or mutant enzyme. The growth rate was normal glutaminyl enzyme catalytic intermediate. In a con- determined in minimal media containing either 50 mM trol reaction with the wild-type enzyme, glutamine-depen- NH4Cl or 0.2 mM NH4Cl as a nitrogen source. With the dent CTP synthetase was completely inactivated in 20 min at higher level of NH4Cl, NH3 as well as glutamine is available 23°C by a 4.5-fold molar excess of [14C]DON. Measurement as a substrate for CTP synthesis. With 0.2 mM NH4Cl as a of radioactivity indicated incorporation of 0.15 equivalent of nitrogen source, the level of NH3 is expected to be limiting ['4C]DON per enzyme subunit. The low stoichiometry for for NH3-dependent CTP synthetase because of the high Km inactivation is not understood. A stoichiometry of approxi- for NH3 (9). These growth conditions are similar to those mately 0.4 to 0.5 equivalent of DON per enzyme subunit was used previously to evaluate the in vivo function of other mutant glutamine amidotransferases (1, 17). Growth rates are tabulated in Table 2. Plasmids encoding the wild-type, TABLE 2. Growth rates of plasmid-bearing pyrG straina Ser 349, and Ala 351 enzymes supported similar growth rates CTP Doubling time (h) with NH4Cl at: in the two media, suggestive of normal function of the CTP Plasmid synthetase synthetase GAT domain. In vivo GAT function for the Ala 50 mM 0.2 mM 351 enzyme can be reconciled with undetectable activity in pMW6 Wild type 2.0 3.2 extracts by assuming that this labile mutant enzyme was pMW7 Ser 349 2.8 3.0 inactivated during preparation of the extract. High levels of pMW8 Ala 351 2.5 2.5 Ala 351 enzyme were not accumulated (Fig. 5, lane 3), pMW9 Cys 352 13 20 presumably because of proteolysis. Slow growth of plasmid- pMW10 Pro 352 5.3 20 bearing cells in medium with low NH4Cl confirms that the a Strain JF646 (pyrG), harboring a plasmid as indicated, was grown at 37°C Pro 352 enzyme was defective in vivo in glutamine amide in M9 medium containing either 50 or 0.2 mM NH4CI as a nitrogen source. VOL. 169, 1987 CTP SYNTHETASE GAT DOMAIN 3027 previously ascribed to negative cooperativity (10). The co- ing experiments indicate that the Ala 351 enzyme was valent attachment of DON to one subunit of an oligomeric synthesized at about the same rate as the Ser 349 and Pro 352 enzyme is thought to prevent interaction of DON or gluta- enzymes and functioned in vivo at least as well as the latter mine with unoccupied subunits. Reaction of DON with the yet was not accumulated. Lack of in vivo accumulation and glutamine amide transfer domain was specific, as 88% of the in vitro activity is therefore ascribed to lability and suscep- NH3-dependent CTP synthetase was retained. Under iden- tibility to proteolysis. By contrast, labeling and immuno- tical reaction conditions, there was no incorporation of precipitation detected only a very low level of Cys 352 [14C]DON into the Pro 352 mutant enzyme, indicating that enzyme. We consider that the structural change is so drastic the capacity to form the glutaminyl covalent intermediate in the Cys 352 enzyme that it is either largely degraded was abolished by the mutation. before folding or not efficiently immunoprecipitated. An alternative possibility is that overproduction of Cys 352 CTP DISCUSSION synthetase is harmful to the cell and a mutation that reduces expression was required to permit maintenance of plasmid Previous identification of conserved amino acid residues pMW9 (Cys 352). Although this possibility was not ex- in several glutamine amidotransferases has provided evi- cluded, there were no indications of plasmid instability dence for a homologous GAT domain in a subgroup of these during construction of pMW9 (Cys 352). enzymes (1, 6, 26, 27, 29). We have designated this subgroup In attempts to identify nucleotide binding sites, we con- trpG type after the prototype trpG-encoded GAT domain in structed six other single and double missense mutations at anthranilate synthase (16, 29). Conserved amino acids are other positions in the CTP synthetase GAT domain. Amino mainly clustered in three regions of the approximately 200 acid replacements at positions 483, 487, 489, 490, and 494 amino acids of the domain. In the case of CTP synthetase, had no effect on either glutamine- or NH3-dependent enzyme we have proposed that a GAT gene, derived from a trpG activity. Thus, other positions in the enzyme can be altered ancestor, was fused onto the 3' end of a gene encoding the without detectable structural alterations. Mutations in GAT CTP synthetase aminator function (26). Mutational analysis domain region I have a significant effect on structure and indicates that each amino acid replacement in CTP synthe- function. tase conserved region I adversely affects enzyme structure and supports the view that this region has a functional role. ACKNOWLEDGMENTS The Val 349 to Ser replacement was least drastic. It provides CTP synthetase with the same amino acid that is present at This work was supported by Public Health Service grant this position in the GAT domain of four other GM24658 from the National Institutes of Health. amidotransferases. The only functional consequence of this We thank Elinor Darbishire and Jack Dixon for synthesis of mutation was increased lability of the highly active Ser 349 oligonucleotides. enzyme relative to the wild type. The lability was initially LITERATURE CITED detected after at Even the storage -20°C. though enzyme 1. Amuro, N., J. L. Paluh, and H. Zalkin. 1985. Replacement by was catalytically competent, heat inactivation demonstrates site-directed mutagenesis indicates a role for histidine 170 in the that the structural alteration in the GAT domain was trans- glutamine amide transfer function of anthranilate synthase. J. mitted to the aminator domain. It is interesting to speculate Biol. Chem. 260:14844 14849. that a serine to valine mutation at position 349 may have 2. Anderson, P. M. 1983. CTP synthetase from Escherichia coli: an evolved to enhance the stability of an ancestral CTP synthe- improved purification procedure and characterization of hys- tase. teric and enzyme concentration effects on kinetic properties. For all GAT domain mutations at positions 349, 351, and Biochemistry 22:3285-3292. 352, structural perturbations were transmitted to the amina- 3. Bauer, C. E., S. D. Hesse, D. A. Waechter-Brulia, S. P. Lynn, tor residues 1 to 300. This result the R. I. Gumport, and J. F. Gardner. 1985. A genetic enrichment domain, demonstrates for mutations constructed by oligodeoxynucleotide-directed interactions between GAT and aminator domains which mutagenesis. Gene 37:73-81. must be important for coupling glutamine hydrolysis with 4. Buchanan, J. M. 1973. The amidotransferases. Adv. Enzymol. utilization of the resultant nascent NH3. Levitzki and Kosh- Relat. Areas Mol. Biol. 39:91-183. land (9) previously concluded that externally added NH3 and 5. Friesen, J. D., G. An, and N. P. Fiil. 1978. Nonsense and the nascent NH3 released from glutamine occupy the same insertion mutants in the relA gene of E. coli: cloning relA. Cell enzyme site and that nascent NH3 cannot dissociate from the 15:1187-1197. enzyme and equilibrate with a pool of external NH3. The 1:1 6. Kaplan, J. B., W. K. Merkel, and B. P. Nichols. 1985. Evolution stoichiometry of the products CTP and glutamate indicates of glutamine amidotransferase genes. Nucleotide sequences of that 100% of the nascent the pabA genes from Salmonella typhimurium, Klebsiella aero- NH3 originating from the amide of genes and Serratia marcescens. J. Mol. Biol. 183:327-340. glutamine must be transferred to the aminator domain. There 7. Kaplan, J. B., and B. P. Nichols. 1983. Nucleotide sequence of are at least two general possibilities to explain loss of GAT Escherichia coli pabA and its evolutionary relationship to function in CTP synthetase (Pro 352). The mutation either (i) trp(G)D. J. Mol. Biol. 168:451-468. prevents formation of the covalent glutaminyl intermediate 8. Laemmli, U. K. 1970. Cleavage of structural proteins during the or (ii) disrupts transfer of nascent NH3 from the glutaminyl assembly of the head of bacteriophage T4. Nature (London) intermediate. CTP synthetase (Pro 352) was not affinity 227:680-685. labeled by DON, indicating that the mutant enzyme had lost 9. Levitzki, A., and D. E. Koshland, Jr. 1971. Cytidine triphos- the capacity to form the covalent glutaminyl intermediate phate synthetase. Covalent intermediates and mechanism of with 379. At we cannot action. Biochemistry 10:3365-3371. cysteine present, distinguish whether 10. Levitzki, A., W. B. Stalicup, and D. E. Koshland. 1971. Half-of- the structural alteration has abolished the capacity to (i) bind the-sites reactivity and the conformational states of cytidine glutamine or (ii) attach bound glutamine to cysteine 379. triphosphate synthetase. Biochemistry 10:3371-3378. Although neither CTP synthetase (Ala 351) nor Cys 352 11. Long, C. W., and A. B. Pardee. 1967. Cytidine triphosphate enzyme was active in cell extracts and neither could be synthetase of Escherichia coli B. I. Purification and kinetics. J. purified, the enzymes exhibit interesting properties. Label- Biol. Chem. 242:4715-4721. 3028 WENG AND ZALKIN J. BACTERIOL.

12. Makaroff, C. A., J. L. Paluh, and H. Zalkin. 1986. Mutagenesis anthranilate synthetase. J. Bacteriol. 123:620-630. of ligands to the [4Fe-4S] center of Bacillus subtilis glutamine 21. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc- phosphoribosylpyrophosphate amidotransferase. J. Biol. Chem. ing with chain terminating inhibitors. Proc. Natl. Acad. Sci. 261:11416-11423. USA 74:5463-5467. 13. Messing, J. 1983. New M13 vectors for cloning. Methods 22. Tiedeman, A. A., J. M. Smith, and H. Zalkin. 1985. Nucleotide Enzymol. 101:20-78. sequence of the guaA gene encoding GMP synthetase of Esch- 14. Miller, J. H. 1972. Experiments in molecular genetics, p. 431. erichia coli K12. J. Biol. Chem. 260:8676-8679. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 23. Tso, J. Y., M. A. Hermodson, and H. Zalkin. 1980. Primary 15. Nagano, H., H. Zalkin, and E. J. Henderson. 1970. The structure of Serratia marcescens anthranilate synthase compo- anthranilate synthetase-anthranilate-5-phosphoribosylpyro- nent II. J. Biol. Chem. 255:1451-1457. phosphate phosphoribosyltransferase aggregate. On the reac- 24. Tso, J. Y., H. Zalkin, M. van Cleemput, C. Yanofsky, and J. M. tion mechanism of anthranilate synthetase from Salmonella Smith. 1982. Nucleotide sequence of Escherichia coli purF and typhimurium. J. Biol. Chem. 245:3810-3820. deduced amino acid sequence of glutamine phosphoribosylpyro- 16. Nichols, B. P., G. F. Miozzari, M. van Cleemput, G. N. Bennett, phosphate amidotransferase. J. Biol. Chem. 257:3525-3531. and C. Yanofsky. 1980. Nucleotide sequences of the trpG 25. Walker, J. E., N. J. Gay, M. Saraste, and A. N. Eberle. 1984. regions of Escherichia coli, Shigella dysenteriae, Salmonella DNA sequence around the Escherichia coli unc operon. Com- typhimurium and Serratia marcescens. J. Mol. Biol. 142:503- pletion of the sequence of a 17 kilobase segment containing 517. asnA, oriC, unc, glmS and phoS. Biochem. J. 224:799-815. 17. Paluh, J. L., H. Zalkin, D. Betsch, and H. L. Weith. 1985. Study 26. Weng, M., C. A. Makaroff, and H. Zalkin. 1986. Nucleotide of anthranilate synthase function by replacement of cysteine 84 sequence of Escherichia coli pyrG encoding CTP synthetase. J. using site-directed mutagenesis. J. Biol. Chem. 260:1889-1894. Biol. Chem. 261:5568-5574. 18. Penefsky, H. 1977. Reversible binding of Pi by beef heart 27. Werner, M., A. Feller, and A. Pierard. 1985. Nucleotide se- mitochondrial adenosine triphosphate. J. Biol. Chem. 252:2891- quence of yeast gene CPA1 encoding the small subunit of 2899. arginine-pathway carbamoylphosphate synthetase. Homology 19. Piette, J., H. Nyunoya, C. J. Lusty, R. Cunin, G. Weyens, M. of the deduced amino acid sequence to other glutamine Crabeel, D. Charlier, N. Glansdorff, and A. Pierard. 1984. DNA amidotransferases. Eur. J. Biochem. 146:371-381. sequence of the carA gene and the control region of carAB: 28. Zalkin, H. 1973. Anthranilate synthetase. Adv. Enzymol. Relat. tandem promoters, respectively controlled by arginine and the Areas Mol. Biol. 38:1-39. pyrimidines, regulate the synthesis of carbamoyl-phosphate 29. Zalkin, H., P. Argos, S. V. L. Narayana, A. A. Tiedeman, and synthetase in Escherichia coli K-12. Proc. Natl. Acad. Sci. USA J. M. Smith. 1985. Identification of a trpG-related glutamine 81:4134-4138. amide transfer domain in Escherichia coli GMP synthetase. J. 20. Reiners, J. J. Jr., and H. Zalkin. 1975. Immunological study of Biol. Chem. 260:3350-3354.