Replacement by site-directed mutagenesis indicates a role for histidine 170 in the glutamine amide transfer function of anthranilate synthase.

Anthranilate synthase is a glutamine amidotransferase that catalyzes the first reaction in tryptophan biosynthesis. Conserved amino acid residues likely to be essential for glutamine-dependent activity were identified by alignment of the glutamine amide transfer domains in four different enzymes: anthranilate synthase component II (AS II), p-aminobenzoate synthase component II, GMP synthetase, and carbamoyl-P synthetase. Conserved amino acids were mainly localized in three clusters. A single conserved histidine, AS II His-170, was replaced by tyrosine using site-directed mutagenesis. Glutamine-dependent enzyme activity was undetectable in the Tyr-170 mutant, whereas the NH3-dependent activity was unchanged. Affinity labeling of AS II active site Cys-84 by 6-diazo-5-oxonorleucine was used to distinguish whether His-170 has a role in formation or in breakdown of the covalent glutaminyl-Cys-84 intermediate. The data favor the interpretation that His-170 functions as a general base to promote glutaminylation of Cys-84. Reversion analysis was consistent with a proposed role of His-170 in catalysis as opposed to a structural function. These experiments demonstrate the application of combining sequence analyses to identify conserved, possibly functional amino acids, site-directed mutagenesis to replace candidate amino acids, and protein chemistry for analysis of mutationally altered proteins, a regimen that can provide new insights into enzyme function.


Replacement by Site-directed Mutagenesis Indicates a Role for Histidine 170 in the Glutamine Amide Transfer Function of Anthranilate Synthase*
(Received for publication, June 24, 1985) Naoki Amuro, Janet L. PaluhS, and Howard Zalkin Anthranilate synthase is a glutamine amidotransferase that catalyzes the first reaction in tryptophan biosynthesis. Conserved amino acid residues likely to be essential for glutamine-dependent activity were identified by alignment of the glutamine amide transfer domains in four different enzymes: anthranilate synthase component I1 (AS 11), p-aminobenzoate synthase component 11, GMP synthetase, and carbamoyl-P synthetase. Conserved amino acids were mainly localized in three clusters. A single conserved histidine, AS I1 His-170, was replaced by tyrosine using site-directed mutagenesis. Glutamine-dependent enzyme activity was undetectable in the Tyr-170 mutant, whereas the NH3-dependent activity was unchanged. Affinity labeling of AS I1 active site Cys-84 by 6-diazo-5-oxonorleucine was used to distinguish whether His-170 has a role in formation or in breakdown of the covalent glutaminyl-Cys-84 intermediate. The data favor the interpretation that His-170 functions as a general base to promote glutaminylation of Cys-84. Reversion analysis was consistent with a proposed role of His-170 in catalysis as opposed to a structural function. These experiments demonstrate the application of combining sequence analyses to identify conserved, possibly functional amino acids, site-directed mutagenesis to replace candidate amino acids, and protein chemistry for analysis of mutationally altered proteins, a regimen that can provide new insights into enzyme function.
Anthranilate synthase is a glutamine amidotransferase that catalyzes the initial step of tryptophan biosynthesis in microorganisms and plants. The microbial enzyme is an oligomer of dissimilar subunits designated AS 1' and AS I1 (1). AS I catalyzes an NH3-dependent synthesis of anthranilate: chorismate + NH3 += anthranilate + pyruvate. AS I1 provides glutamine amide transfer function to the AS I . AS I1 complex: chorismate + glutamine += anthranilate $. pyruvate + glutamate. Dual specificity for glutamine or NH, is representative of all amidotransferases (2) and may reflect acquisition of glutamine amide transfer function by ancestral NH3-dependent enzymes (3)(4)(5)(6)(7)(8). Techniques of enzymology (3, 9-12), protein chemistry (3, [13][14][15][16], and, more recently, site-directed mutagenesis (7) have been utilized to study the structure and function of anthranilate synthase. This laboratory has reported evidence for a mechanism of glutamine amide transfer that involves covalent glutaminyl and glutamyl adducts with an AS I1 active site cysteine (1,3,13). The active site cysteine was identified in the AS I1 subunit from Pseudomonas putida (14) and Serratia marcescens (15). Specific replacement of active site cysteine 84 by glycine in S. marcescens AS I1 confirmed the role of this residue in glutamine amide transfer (7). Using techniques of chemical modification and protein chemistry we could not identify other residues that functioned in the glutamine amide transfer mechanism. ' In this report we present the initial results of an alternative approach to identify residues that are essential for glutamine amide transfer function. Amino acids that are conserved in homologous glutamine amide transfer domains of four different glutamine amidotransferases were identified and are considered likely to have essential structural or catalytic roles. Analysis of a proposed mechanism for glutamine amide transfer suggests a requirement for general base-general acid catalysis. We, therefore, examined the effects on catalysis of replacing the only conserved histidine residue in S. marcescens AS 11. Site-specific replacement of AS I1 His-170 by Tyr inactivates glutamine amide transfer function without affecting NH3-dependent synthesis of anthranilate. Reversion analysis is consistent with a function of His-170 in catalysis as opposed to a structural role.

EXPERIMENTAL PROCEDURES
Strains, Plasmids, and Phage-Escherichia coli strain JMB9, relevant genotype AtrpEGD was used as a recipient for plasmids carrying S. marcescens trp genes (7). The growth rates of strain JMB9 transformants were determined as described (7). Plasmid pJPl7 is a S. marcescens trpE+G+D+ pBR322 recombinant (7) that utilizes a plasmid promoter for trp gene transcription. The AS I and AS I1 subunits are encoded by trpE and trpG, respectively. Plasmid pJPZ0 is a pJPl7 derivative that contains the strong S. murcescens trp operon promoter (7). pJP19 (trpE+GID+) and pJP2l (trpE+GID+) are isogenic with pJPl7 and pJPZ0, respectively, but have an AS I1 Cys-84 to Gly replacement (7). M13 phage for mutagenesis and DNA sequencing have been described (7).
Selection of Reuertunts-Strain JMBS ( AtrpEGD) bearing plasmids pJP19 (trpE+GZD+) or pNA3 (trpE+GZD+) cannot synthesize tryptophan and grow normally in M9 media (22), with 50 pg/ml ampicillin, containing 1 mM NH&1 as a nitrogen source because of defective anthranilate synthase glutamine amide transfer. Strain JMBS bearingplasmid pJP17 (trpE+G+D+) which encodes wild type anthranilate synthase yields colonies or grows to stationary phase in liquid culture within 24 h in low NH3 media containing ampicillin since NH3 is assimilated into glutamine which can be utilized by the wild type anthranilate synthase for tryptophan synthesis. Spontaneous revertants of strain JMBS bearing plasmid pJP19 or pNA3 were selected in liquid media from a series of 10 separate cultures. Revertants were obtained after 3-5 days at 37 "C, and only one revertant was saved from each of the cultures. Spontaneous revertants of JMB9/pJP19 were also isolated directly on solid media after incubation for 3-5 days. Plasmid DNA was isolated (22) from revertants grown in LB (22) plus ampicillin and used to retransform strain JMB9. Transformants were screened on low NH3 plates containing ampicillin, and representative clones were checked by enzyme assay, measurement of growth rate, and DNA sequencing. In some cases a second cycle of plasmid isolation and retransformation was performed to resolve mixtures as determined by DNA sequencing.

RESULTS
Glutamine Amide Transfer Domains-Previous attempts to detect essential AS I1 amino acids, other than Cys-84 (15), were unsuccessful using techniques of protein chemistry.'Side reactions of group-specific reagents with Cys-84 were difficult to exclude. The present objective was, therefore, to identify conserved amino acids that might have essential structural or catalytic roles in AS I1 glutamine amide transfer. A recently reported computer alignment of E. coli AS 11, PABS 11, and amino acids 1-198 of GMP synthetase confirmed the homology of AS I1 and PABS I1 sequences (23) and identified a homologous glutamine amide transfer domain in GMP synthetase (8). A segment of the E. coli carA-encoded small subunit of carbamoyl-P synthetase (24)(25)(26) can be added to the three previously compared sequences. Residues 185-382 of carbamoyl-P synthetase are matched with amino acids 1-198 of GMP synthetase, AS 11, and PABS I1 in Fig. 1. The alignment shown in Fig. 1 has 19 identities and 34 positions with conserved residues in all four sequences. By these criteria 27% of the 193 AS I1 amino acids have been conserved in the trpG-related glutamine amide transfer domains in carA-encoded carbamoyl-P synthetase, GMP synthetase, and PABS 11. Of the 19 identical amino acids, 13 are in three clusters at positions 57-64, 85-102, and 187-192. The single conserved cysteine at position 91, which corresponds to Cys-84 in S. marcescens AS 11, was previously shown by affinity labeling (10,15) and site-directed mutagenesis (7) to function in glutamine amide transfer. Our proposal for the mechanism of glutamine amide transfer (below) predicts a requirement for an amino acid side chain that can function in general acidbase catalysis to donate and abstract protons. The best candidates for such a group are the conserved histidine at position 190 and the conserved glutamate at position 192. Accordingly, we have replaced the histidine at position 190, histidine 170 in S. marcescens AS 11.
Mutant Isolation-A CAC (His) to TAC (Tyr) change at codon 170 in S. marcescens trpG was constructed by incorporation of a synthetic oligonucleotide into a recombinant M13mpll-trpG heteroduplex. The nucleotide and amino acid sequence of the pertinent region of trpG together with the sequence of the synthetic 17-mer are shown in Fig. 2. The mutagenic 17-mer was annealed to the complementary region of the trpG non coding strand (Fig. 2), generating a C:A mismatch at nucleotide 508. Following primer extension, ligation, and transformation, trpG phage were screened by "A lane" sequencing (Fig. 3). One mutant was identified among 48 phage screened. The Tyr-170 mutation was designated trpG2. The trpG2 mutation was transferred in vitro to plasmid pJP17 (trpE+G+D+) to yield pNA3 (trpE+G2D+) in which trp genes are transcribed from a pBR322 plasmid promoter.
The nucleotide sequence of trpG2 was checked between nucleotides 370-520. The only change from wild type was C to T at position 508. A portion of the sequencing gel that verifies the mutation is shown in Fig. 4.
Enzyme Actiuity-In order to examine the consequences of the trpG2 mutation, anthranilate synthase activity was determined in extracts of plasmid-bearing cells. Specific activities of 9.8 units/mg and 13.1 units/mg were obtained for the glutamine-and NH3-dependent activities, respectively, for wild type anthranilate synthase from cells bearing plasmid pJPl7 (trpE+G+D+). Glutamine-dependent anthranilate synthase was undetectable from cells bearing plasmid pNA4 (trpE+G2D+), whereas the NH3-dependent activity was 18.7 units/mg. To confirm that the Tyr-170 replacement inactivated the glutamine amide transfer function of anthranilate synthase, the trpG2 mutant and wild type enzymes were purified to homogeneity. The NH3-dependent activities were 2150 and 2280 units/mg for the wild type and trpG2 mutant, respectively. Glutamine-dependent anthranilate synthase was 3200 units/mg for the wild type and less than 0.01 unit/mg for the Tyr-170 mutant enzyme (trpE+G2+) encoded by pNA4. These data are formally similar to those obtained for the trpE+Gl mutant anthranilate synthase having an AS I1 Cys-84 to Gly replacement and, therefore, suggest that His-170 is required in addition to Cys-84 for glutamine amide transfer function.
I n Vivo Function-Paluh et al. (7) reported that the NH3dependent activity of trpE+Gl+ (Gly-84)-encoded anthranilate synthase supported tryptophan synthesis and growth in media containing 50 mM NH4C1 but not in media containing 1 mM NH4C1. Quantitatively similar results were obtained for the trpE+G2 Tyr-170 mutant. The doubling time of strain JMBS (AtrpEGD) bearing plasmid pNA3 (trpEfG2D+) in minimal media containing 50 mM NH4C1 was 1 h, in contrast to a 15-h doubling time in media containing 1 mM NH4C1. These results further support the conclusion (7) that the glutamine-dependent activity of anthranilate synthase provides selective advantage in media having limiting levels of Role of His-170"Alkylation of AS I1 Cys-84 by the glutamine affinity analog DON mimics formation of the catalytic glutaminyl-cysteine 84 covalent intermediate (3,10). Affinity labeling of the trpE+G2 mutant enzyme was conducted to NH3.

PARS (1) R I L L I D N Y D S F T Y N L Y U Y -F C E L 6 A D V L V K R N D
. . . . .  . * + + + * .

A P L F D H F I E L I E O Y R K T A K R M L -E K F V R U I C O C E A O L L A b -R L L -O T L A V A -U H K L F L H K 210
* . determine whether His-170 functions in a step prior to or after formation of the covalent glutaminyl enzyme intermediate.
[I4C]DON was rapidly incorporated into the wild type enzyme (Fig. 5) and concomitantly inactivated the glutaminedependent activity (not shown). The stoichiometry for incor-

FIG. 4. Nucleotide sequence of a portion of wild type (left)
and trpC2 (right). The sequence of the coding strand from nurleotides 500 to 520 is shown. The GTG and A'I'C. triplets which are complementary to the His (CAC) and Tyr ITAC) codons. respectively, are boxed. Arrows set off the sequence of the synthetic 17-mer. Arrowheads point to the hase change.
poration was approximately 0.7 eq of DON/AS 1. A S I1 protomer. Mutant enzyme with the A S I1 Gly-84 replacement was not alkylated verifying the specificity of DON for AS I1 Cys-84. Using the same conditions as for the wild t+ype enzyme, the Tyr-170 mutant also was not alkylated. However, a 10-fold increase in DON concentration permitted incorporation of DON a t a rate approximately 2% that of the wild type (Fig. 5). The simplest interpretation of this result is that His-170 has an essential role in formation of the covalent Anthranilate synthase exhibits a glutaminase activity that reflects uncoupling of glutamine hydrolysis from anthranilate synthesis (3,10). The glutaminase activity presumably employs the same steps for formation and breakdown of the covalent glutaminyl and glutamyl intermediates that are normally used in amide transfer. Data in Fig. 6 show a specific activity for glutamine hydrolysis of 5160 units/mg at 37 "C catalyzed by the wild type enzyme. Glutamine hydrolysis was not detected for anthranilate synthase having the Gly-84 or Tyr-170 replacements, thus supporting the conclusion that the two mutant enzymes are defective in glutamine utilization.
Reversion Analysis-Evidence is required to distinguish whether an essential amino acid residue functions in catalysis or exerts a structural role. Analysis of revertants can potentially distinguish between these two cases. For the case of an amino acid that functions in catalysis, reversion of a missense mutation must restore that amino acid. For an amino acid that exerts a structural role, a different amino acid could be functional at the initial site or at a second site (27,28). E. coli strain JMBS (AtrpEGD) bearing plasmids trpE+GlD+ (AS I1 Gly-84) or trpE+G2D+ (AS I1 Tyr-170) does not grow on minimal agar media containing 1 mM NH&l because mutant anthranilate synthase with defective glutamine amide transfer function cannot utilize glutamine for tryptophan synthesis. A low concentration of NH3 can be used by glutamine synthetase but not by anthranilate synthase. We selected spontaneous revertants of strains JMBS (AtrpEGD)/pJP19 (trpE+GID+) Gly-84 anthranilate synthase and JMBS (AtrpEGDlpNA3 (trpE+G2D+) Tyr-170 anthranilate synthase that utilized 1 mM NH&1 as a nitrogen source for tryptophan synthesis as described under "Experimental Procedures." For strain JMBS (AtrpEGD)/pJP19 (trpE+GID+) 54 revertant colonies of varied sizes were obtained on solid media after 5 days. The reversion frequency was 1 in 4 x lo7. Twelve colonies of different sizes were picked for further analysis. Following plasmid isolation and retransformation all colonies appeared to grow at the same rate consistent with a single class of revertants. Furthermore, the growth of revertants initially isolated in liquid and on solid media was indistinguishable on agar plates. The growth rate in low NH3 liquid culture of four revertant strains was measured and was identical to that of the parental JMBS (AtrpEGDlpJPl7 (trpE+G+D+). Likewise, glutamine-dependent anthranilate synthase activity was restored to the wild type level in extracts of the four strains examined. Finally, DNA sequence analysis indicated that 12/ 12 revertants had a GGC (Gly) t o TGC (Cys) reversion at codon 84. In an identical manner, Tyr-170 revertants were obtained. Representative strains (10/10) grew in low NH3 minimal media at the wild type rate, regained the wild type glutamine-dependent enzyme activity (10/10), and restored CAC (His) at codon 170 (lO/lO). These results support a role of Cys-84 and His-170 in catalysis.

DISCUSSION
All glutamine amidotransferases, including anthranilate synthase from nine prokaryotic and eukaryotic organisms (14,(29)(30)(31)(32), utilize an active site cysteine in glutamine amide transfer function (2, 6, 33). S. marcescens AS I1 cysteine 84 was identified as the essential residue by affinity labeling (10,15) and was confirmed by replacement with glycine using sitedirected mutagenesis (7). Attempts to identify other residues that function in glutamine amide transfer using chemical modification by histidine and arginine reagents were unsuccessful because of the difficulty in eliminating the possibility that inactivation was due to modification of the highly reactive AS I1 Cys-84. 2 We have now utilized newly available techniques to identify residues likely to be important for glutamine amide transfer structure and function and have provided evidence for a role of AS I1 His-170 in catalysis.
The alignment of glutamine amide transfer domains in E. coli carbamoyl-P synthetase, GMP synthetase, AS 11, and PABS component 11, shown in Fig. 1, identifies three regions having highly conserved primary structure. It is likely that conserved residues in the homologous glutamine amide trans-Role of His-1 70 in Anthranilate Synthase fer domains are critical for structure or catalysis. Indeed, S. marcescem AS I1 Cys-84 is conserved at position 91. In addition to the essential cysteine, the mechanism for glutamine amide transfer likely requires a group to abstract and donate protons (see below). AS I1 His-170, position 190 in Fig. 1, is a good candidate for such a residue, and the replacement of His-170 by tyrosine was shown to inactivate glutamine amide transfer without having any effect on NH3-dependent activity. Affinity labeling of Cys-84 by DON was used to provide evidence for the role of His-170 in glutamine amide transfer. The reactions shown in Fig. 7 summarize our view of glutamine amide transfer. The overall reaction (3) is divided into three steps, formation of the glutaminyl-cysteine enzyme adduct, amide transfer, and thioester hydrolysis. We visualize that in each step a general base (:B) or its conjugate general acid (HB) is required for proton abstraction or donation, respectively. Previous evidence supports the conclusion that affinity labeling by glutamine analogs mimics step 1 and results in alkylation of the AS I1 active site cysteine (11-SH) ( 3 )~ Under reaction conditions that uncouple glutamine hydrolysis from product formation (3)  nonprotonated histidyl residue. According to these interpretations of the mutagenesis experiment, protein folding must position Cys-84 and His-170 to allow their interaction at the catalytic site.
There are several possibilities to explain why affinity labeling of the Tyr-170 enzyme occurred at a slow rate, but glutaminase activity was not detected. DON may be more reactive to nucleophilic attack by Cys-84 than the carboxamide group of glutamine. Alternatively, step 1 may not be rate determining for glutamine hydrolysis.
The proposed role of His-170 in the glutaminylation of AS I1 Cys-84 (Fig. 7) is analogous to the function of His-159 in the acylation of Cys-25 by peptide substrates in papain (34). Furthermore, if the conserved AS I1 Glu-172, position 192 in Fig. 1, were to also participate in proton transfer, the triad Cys-84, His-170, and Glu-172 would be similar in function to the chymotrypsin charge-relay system composed of Ser-195, His-57, and Asp-102 (35). In this regard, it is of interest that chemical conversion of Ser-195 to alanine in chymotrypsin decreased the reactivity of His-57 to affinity labeling (36). In AS I1 replacement of His-170 decreased the reactivity of Cys-84 to affinity labeling (Fig. 5).
We recognize two main qualifications that bear upon the proposed role of AS I1 His-170 and the mechanism suggested in Fig. 7. First, the scheme in Fig. 7 describes the simplest possible case. Other amino acids could function together with AS I1 His-170 in abstracting and donating protons. Further mutagenesis experiments are required to clarify this possibility. Second, we need to evaluate the evidence that His-170 has a direct role in catalysis and not a structural role. As discussed below, reversion analysis provided preliminary evidence for a role of His-170 in catalysis.
In previous replacements of the active site cysteine in two glutamine amidotransferases, enzymatic, chemical, and physical properties of the mutant and wild type enzymes were compared and found to be indistinguishable (6,7). This evidence supported the conclusions derived from affinity labeling that a cysteine residue is required for glutamine amide transfer. However, it is uncertain whether measurements of NH3dependent activity, allosteric inhibition, proteolytic inactivation, and circular dichroism are adequately sensitive to detect putative small local changes in conformation that could obstruct Cys-84 and disrupt glutamine amide transfer. Reversion analysis should provide an alternative sensitive distinction between residues that are required for catalysis or for structure. Amino acids that participate in catalysis should be irreplaceable whereas other amino acids should be able to correct a missense mutation by incorporation at the primary or secondary positions. Previous studies of E. coli tryptophan synthase provide examples for restored function of second site reversionq in missense mutants (27,28). In our experil ments all trpE+GlD+ and trpE+G2D+ revertants were of a single class in which the original trpG mutation was corrected to restore the wild type codon. These results are consistent with a role of AS I1 His-170 in catalysis. Further evaluation of this approach is required to determine the relative frequency that missense mutations causing structural alterations can be corrected by the original amino acid compared to a structurally acceptable amino acid at the primary or secondary sites.