The Sulfonylurea Herbicide Sulfometuron Methyl Is an Extremely Potent and Selective Inhibitor of Acetolactate Synthase in Salmonella typhimurium”

The sulfonylurea herbicide sulfometuron methyl inhibits the growth of several bacterial species. In the presence of L-valine, sulfometuron methyl inhibits Salmonella typhimurium, this inhibition can be reversed by L-isoleucine. Reversal of growth retardation by Lisoleucine, accumulation of guanosine 5”diphosphate 3”diphosphate (magic spot), and relA mutant hypersensitivity suggest sulfometuron methyl interference with branched-chain amino acid biosynthesis. Growth inhibition of S. typhimurium is mediated by sulfometuron methyl’s inhibition of acetolactate synthase, the first common enzyme in the branched-chain amino acid biosynthetic pathway. Sulfometuron methyl exhibits slow-binding inhibition of acetolactate synthase isozyme I1 from S. typhimurium with an initial Ki of 660 f 60 nM and a final, steady-state Ki of 65 f 25 nM. Inhibition of acetolactate synthase by sulfometuron methyl is substantially more rapid (10 times) in the presence of pyruvate with a maximal first-order rate  constant  for  conversion  from initial to  final steady-state inhibition of 0.25 f 0.07 min” (minimal half-time of 2.8 min). Mutants of S. typhimurium able to grow in the presence of sulfometuron methyl were obtained. They have acetolactate synthase ctivity that is insensitive to sulfometuron methyl because of mutations in or near ilvG, the structural gene for acetolactate synthase isozyme 11.

Many herbicides of diverse structure interfere with photosynthetic electron transport (3). That these herbicides affect plants and not animals reflects the difference in energy acquisition between plants and animals. These two kingdoms differ in a second profound way. Metaphytes synthesize all * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Proposed common name from the American National Standards Institute. their vitamins and amino acids; metazoans, especially higher animals, usually lack some biosynthetic capabilities. The herbicide glyphosate interferes with aromatic amino acid biosynthesis (4), although its precise target remains uncertain (5- 8).
Since bacteria and plants share many common biochemical pathways, bacteria can provide an expedient means of localizing the site of herbicidal action. Bacteria, unlike plants, have the advantage of well-defined biochemistry and genetics. In this report, we establish that growth retardation of Salmonella typhimurium induced by sulfometuron methyl is a result of inhibition of the branched-chain amino acid biosynthetic enzyme acetolactate synthase (EC 4.1.3.18) isozyme 11. Microbial Techniques-For growth experiments, Vogel and Bcnner medium (11) supplemented with 0.2% glucose was the minimal medium; the rich medium was LB (12).

Materials-~yrimidine-'4C]Sulfometuron
Antibacterial activity was measured by the disc diffusion method on minimal agar plates (13). Washed, overnight cultures were plated in soft agar overlays; zones of inhibition caused by placing a 6-mm paper disc, impregnated with 40 pg of sulfometuron methyl, upon a bacterial lawn were measured after a 16-h incubation at 30 "C.
Large-scale culture (2000 liters) of E. coli HBlOl/pDU9 was carried out with the minimal salts medium of Omerod et al. (14) containing 5.5 g/liter (initially) of dextrose, 1.2 mM L-leucine, 1.2 mM L-proline, 3 mM L-valine (found to be effective in ensuring plasmid maintenance), 20 mg/liter of sodium ampicillin, and 100 mg/liter of thiamine HCl. Additional dextrose (four additions equivalent to the initial) 8753 was added during fermentation as needed. Phosphates used were 2.33 Data Processing-The MLAB data modeling program (23), availg/liter of K,HP04 and 1 g/liter of KHZPO,. Anhydrous ammonia was able from the National Institutes of Health, was used to fit the used to maintain a pH of 7.1 throughout the fermentation and as an following equations to data: additional nitrogen source. Stirring and aeration rates were varied to achieve a dissolved 0, concentration 230% of air saturation. While v = VA/(K + A ) (1) still in exponential growth, 240 kg of ice were added to the 2000-liter culture to reduce the culture temperature rapidly from 35 to 18 "C.
Rapid reduction of the culture temperature ensured maximal preservation of acetolactate synthase activity during harvest of the cells by continuous-flow centrifugation. The final cell paste (16.5 kg; 8.2 X where is velocity (the rate of enzymic reaction or the rate of io5 units of acetolactate synthase isozyme 11) was frozen (as a finely transition between initial and final inhibition for slow-binding inhidivided gravel) and stored in liquid nitrogen. bition), A is substrate or inhibitor concentration, V is maximal Assay Of Acetolactate Synthase-Assays were conducted at 37 O C velocity, K is the concentration of substrate or inhibitor which gives by a modification of the method of Bauerle et al. (15). Each assay (1 half-maximal velocity, p is product, po is product present initially ml) contained 0.1 mmol ofTricine2/NaOH, PH 7.8950 Pmol of sodium (zero-time absorbancy), V , is the final, steady-state rate, v,, is the Pyruvate, 10 pmol ofMd& 0.1 Pmol of thiamine PYroPhosPha% and initial rate, k is the first-order rate constant for transition between 0.1 rmol of FAD, Enzymic reactions were Wenched with 0.25 ml of initial and final rates, vi is the inhibited velocity (initial or final for 12 N and S'JbSequentlY incubated at 80 "c for 5 min. After slow binding), I is inhibitor concentration, and Ki is the concentration cooling samples to 37 "c, 0.16 ml Of 50% (W/W) NaOH, 0.3 ml O f 0.5% of inhibitor which gives 50% inhibition.
(w/v) creatine, and 0.3 ml of 5% a-naphthol in 2.5 N NaOH were added in rapid succession with immediate mixing. Samples were incubated at 37 "C for 1 h with several intervening mixinns to ensure RESULTS efficient aeration (essential to color development). After low-speed centrifugation to remove turbidity, absorbance at 530 nm was determined. Formation of 0.1 mM acetolactate in the assay resulted in 0.65 A unit at 530 nm. One unit of activity is the amount required to form 1 pmol of acetolactate (consume 2 pmol of pyruvate) per min under these assay conditions. Protein was determined by the biuret (16) procedure where practical or by the Coomassie Brilliant Blue binding assay (Bio-Rad) of Bradford (17).
Genetics-Independent mutants were selected on minimal medium supplemented with sulfometuron methyl (33 pg/ml) and L-valine (83 pglml). Crosses were mediated via P22 transduction (18). Sulfometuron methyl resistance in the presence of L-valine (83 pg/ml) was an unselected marker in all crosses; the selected marker was either tetracycline resistance associated with TnlO or prototrophy conferred by ilvE+. The response of strains to sulfometuron methyl can be determined only in the absence of L-isoleucine. Thus, Tnl0:iluE transductants were converted to IIvE+ derivatives. Excision of TnIO from Tnl0:ilveE-bearing strains was accomplished by plating -10' cells from a washed, overnight culture on unsupplemented minimal medium (19). The growth response of these IlvE+ strains to medium containing 33 pg/ml of sulfometuron methyl and 83 pg/ml of L-valine was determined. Standard genetic nomenclature was followed (20).
Preparation of Acetoluctate Synthase-Extracts of E. coli HB101/ pDU9 (35 mg of protein/ml) could be dialyzed at 4 "C (twice, 12 h each) against 0.1 M Taps/KOH, pH 8.8, at 4 "C, containing 1 mM EDTA, 0.1 mM dithiothreitol, and 20% (v/v) glycerol, with little or no loss of enzymic activity. These dialyzed extracts did not require M T , thiamine pyrophosphate, or FAD in short-term (1 min) assays. However, upon extended assay at high dilution (2000-fold), the enzyme lost activity with a half-time of 12 min in the absence of M$+, thismine pyrophosphate, and FAD. After preincubation at high dilution (100-fold) for several hours at 37 "C (assay buffer without thiamine pyrophosphate, FAD, or Me), the enzyme had an absolute dependence on MgZ+ or MnZ+, thiamine pyrophosphate, and FAD, with complete restoration of initial levels of enzymic activity by 10 mM MgCl,, 0.1 mM thiamine pyrophosphate, and 0.1 mM FAD (added immediately prior, <1 min, to assay). Dialyzed extracts could be stored at -20 "C for several months without noticeable loss of enzymic activity. Preliminary efforts to purify acetolactate synthase from E. coli HBlOl/pDU9 have established that the enzyme comprises 2.7% of the total extracted protein? A specific activity of 25 units/m$ and molecular weight of 59,300 (21) were used to estimate the concentration of acetolactate synthase in extracts.
Nucleotide Leuek-A qualitative determination of nucleotide levels was made by the method of Bochner and Ames (22). Two-dimensional thin-layer chromatography on polyethylene imine plates (Brinkmann Instruments) was conducted utilizing solvent Tb in the first dimension and Sb in the second dimension (22).
Initial Observations-Sulfometuron methyl inhibits the growth of C. freundii and Acinetobucter sp. on minimal medium. Only in the presence of valine, however, is growth inhibition of wild-type S. typhimurium by sulfometuron methyl observed. Inhibition in the presence of valine is reversed by isoleucine (but not by the other 18 common amino acids), suggesting sulfometuron methyl inhibits a step in the biosynthesis of branched-chain amino acids (Table I). Thiamine partially and thiamine together with a-ketobutyrate protect against sulfometuron methyl inhibition (data not shown). ppGpp and pppGpp accumulated upon treatment of S. typhimurium concomitantly with sulfometuron methyl (100 pg/ml) and L-valine (83 pg/ml). This accumulation is expected if amino acid limitation is a primary event in growth inhibition by sulfometuron methyl.
relA mutants are defective in generalized stimulation of amino acid biosynthetic operons (13); thus, these enzymes (including those involved in the biosynthesis of branchedchain amino acids (24)) exhibit low activities in relA mutants on both minimal and valine-supplemented media. In contrast to wild-type S. typhimurium, growth of the relA mutant TA2439 in the presence of sulfometuron methyl is rescued by inclusion of isoleucine, methionine, or pantothenate. This growth requirement mimics the phenotype of iluG mutants in which auxotrophy is satisfied by isoleucine, methionine, or pantothenate (25). These observations suggest that sulfometuron methyl inhibits the iluG gene product, acetolactate synthase isozyme 11.
Partial reversal of sulfometuron methyl inhibition of the relA mutant TA2439 was mediated by L-leucine (data not shown). In the presence of sulfometuron methyl and valine, however, growth of this relA mutant is only permitted by inclusion of isoleucine in the medium. The pattern of valinepromoted growth inhibition by sulfometuron methyl and reversal by isoleucine is consistent with selective inhibition of acetolactate synthase isozyme I1 by sulfometuron methyl, with valine-sensitive (feedback inhibited) isozyme I being resistant.
Analogous results have been obtained for the r e a l mutant TR3381 and the isogenic re&+ strain TR3379. The relA::TnlO strain TT7542 displays a phenotype indistinguishable from the relAl and relA2 strains in terms of growth inhibition by sulfometuron methyl and reversal of inhibition by pantothenate, methionine, or isoleucine.
Growth inhibition of wild-type C. freundii is reversed by methionine, pantothenate, or isoleucine and potentiated by valine (data not shown). Thus, sulfometuron methyl produces an iluG phenocopy in wild-type C. freundii. In the presence of valine and sulfometuron methyl, only isoleucine prevents growth inhibition. Thus, the responses to sulfometuron methyl of the S. typhimurium relA derivative are quite similar to those of C. freundii.
Genetics-Spontaneous mutants of S. typhimurium capable of growth in the presence of L-valine and sulfometuron methyl have been selected. These strains were used as donors in transductional crosses with the iluE recipient CBS501. The five sulfometuron methyl-resistant mutations were all COtransduced with iluE+ at frequencies ranging from 0.5 to 1.00 (Table 11). This cotransduction suggests that these sulfometuron methyl-resistant mutations lie within the ilu operon.
To substantiate this conclusion, a reciprocal cross was performed. The donor was strain CBS501; the recipients were the sulfometuron methyl-resistant mutants. The iluE mutation in CBS501 is due to an insertion of TnlO; thus, transduction of the ilu region is mediated by selection of tetracycline resistance. The growth response to sulfometuron methyl of the resulting i1uE::TnlO transductants cannot be directly analyzed because of the isoleucine requirement of iluE mutants. Ilv+ derivatives (possibly the product of precise excision events) of at least 19 such tetracycline-resistant, Ilv-transductants arising from each cross were obtained. That the responses to sulfometuron methyl of all derivatives from a single transductant were identical allowed deduction of its sulfometuron methyl phenotype. The sulfometuron methylsensitive phenotype arose at a frequency of 0.72 to 1.00 (Table  11). The cotransduction indicates close linkage of iluE and mutations resulting in sulfometuron methyl resistance. Assay of acetolactate synthase isozyme I1 (in the presence of 1 mM L-valine) in extracts from these mutants showed the enzyme to be far less sensitive to sulfometuron methyl (negligible inhibition by 100 PM sulfometuron methyl) than the wildtype enzyme.
Interaction of Sulfometuron Methyl with Acetolactate Synthase-Acetolactate synthase activity in extracts from wildtype s. typhimurium LT2 was completely inhibited by 1 mM sulfometuron methyl. A detailed characterization of the mode of inhibition was conducted with extracts from E. coli HB101/ pDU9. Assay time courses were markedly biphasic in the presence of sulfometuron methyl (Fig. 2). Although long-term enzymic instability under assay conditions (half-time = 3 h) somewhat distorts the results shown in Fig. 2, it is clear that a final steady-state rate is achieved. Thus, despite the timedependent nature of the inhibition of acetolactate synthase by sulfometuron methyl, it is not irreversible. To determine the kinetic constants which define the interaction of sulfometuron methyl with acetolactate synthase, a higher concentration of enzyme (10 nM) and a shorter-time interval (10 min) were used. Under these conditions, instability of acetolactate synthase was negligible. Assay progress curves could be adequately defined by first-order transients in which there were both an initial (weak) level and a final, steady-state (more potent) level of inhibition (Equation 2). Initial inhibition, final inhibition, and the first-order rate constant for slow binding were all dependent on the concentration of sulfometuron methyl. Analyses of initial and final inhibitions (Equa- the absence of pyruvate. Examination of the degree of initial inhibition, final inhibition, and the first-order transient rate constant a t various pyruvate concentrations (1 p~ sulfometuron methyl) has given similar results over the concentration range of 1-50 mM. In this context, the enzyme displays hyperbolic saturation by pyruvate (0.25-50 mM) with a Michaelis constant of 2.7 f 0.4 mM. Similarly, the concentration of thiamine pyrophosphate (2-200 p~) or FAD (20-100 p~) used in the assay did not affect the degree of inhibition of acetolactate synthase by sulfometuron methyl.
Incubation of 26 p~ acetolactate synthase with 20 p~ sulfometuron methyl in 0.1 M Taps/KOH, pH 8.8 (at 4 "C), 1 mM EDTA, 0.1 mM dithiothreitol, 20% glycerol on ice resulted in a decline of the enzymic activity to 31% of its initial value (within 21 h). Gel filtration (Sephacryl S-200; room temperature; 0.1 M Tricine/KOH, pH 8, as eluant) of incubation mixtures that contained [phenyl-14C]-or [pyrimidine-'*C]sulfometuron methyl resulted in complete reactivation of the enzyme and resolution of the enzyme from radiolabel. However, elution profiles of 14C-labeled sulfometuron methyl from enzymic incubation mixtures were skewed to an earlier elution position than was observed for sulfometuron methyl passed through the gel filtration column alone. It would appear that sulfometuron methyl was initially bound by the enzyme, but was slowly released during gel filtration. When the rapid, centrifugal gel filtration technique of Penefsky (26) was employed to resolve enzyme and unbound ['4C]sulfometuron methyl, 0.8 mol of radiolabel/mol of enzyme eluted coincident with the enzyme (with or without 50 mM pyruvate present in the incubation mixture).
Inhibition could be completely reversed (94%) by incubation of the diluted (100-fold) enzyme at 37 "C (0.1 M Tricine/ KOH, pH 7.8, 1 mM EDTA) for several hours prior to addition of 10 mM MgC12, 0.1 mM thiamine pyrophosphate, 0.1 mM FAD, and 50 mM sodium pyruvate to initiate the assay. Similarly, dilution (4000-fold) of the inhibited enzyme (26 p~ acetolactate synthase, 20 p~ sulfometuron methyl as above) directly into an assay mixture results in substantial reversal (45%) of inhibition over the course of a 1-h assay.

DISCUSSION
Clearly, sulfometuron methyl inhibits growth of S. typhimurium via acetolactate synthase isozyme 11. Sulfometuron methyl is an exceptionally potent inhibitor of acetolactate synthase isozyme 11. Sulfometuron methyl does not inhibit the growth of E. coli, which lacks a functional acetolactate synthase isozyme 11, and only potently inhibits S. typhimurium in the presence of L-valine, which selectively blocks other acetolactate synthase isozymes. Mutations that confer resistance to sulfometuron methyl are in the iluG region and result in an acetolactate synthase isozyme I1 resistant to sulfometuron methyl. An S. typhimurium iluG mutant (lacking acetolactate synthase isozyme 11) is an isoleucine auxotroph since the remaining acetolactate synthase isozyme inefficiently catalyzes aaceto-a-hydroxybutyrate formation (25). This mutant accumulates high levels of a-ketobutyrate, which is toxic due to its interference with pantothenate formation. This toxicity is overcome by supplementation of the growth medium with either pantothenate or methionine. Thus, the iluG auxotrophy is satisfied by isoleucine, methionine, or pantothenate (25). In both rek4 mutants of S. typhimurium and wild-type C.
freundii, sulfometuron methyl inhibition is prevented by inclusion of pantothenate, methionine, or isoleucine in the growth medium. This suggests that in both organisms, aaceto-a-hydroxybutyrate synthesis (isoleucine pathway) is impeded, while acetolactate formation (valine-leucine pathway) proceeds. That valine potentiates growth inhibition by sulfometuron methyl in both organisms implies a comparable division of function between acetolactate synthase isozymes in 5' . typhimurium and C. freundii.
Lack of inhibition of acetolactate synthase isozyme I by sulfometuron methyl4 would seem to suggest the lack of a mechanistic (active site directed) basis for the inhibition. The absence of marked protection of acetolactate synthase by thiamine pyrophosphate (2-200 p~) , FAD (20-100 p~) , or pyruvate (1-50 mM) is inconsistent with the interaction of sulfometuron methyl at the active site. The complementary inhibitors, L-valine and sulfometuron methyl, of isozymes I and 11, respectively, may both inhibit acetolactate synthase via allosteric sites. Unlike isozyme I, however, no allosteric regulation of isozyme I1 has been reported.
Slow-binding inhibition (27), which is seen with sulfometuron methyl, is becoming recognized as a typical feature of potent, reversible inhibitors (28). The unusually slow onset of potent inhibition of acetolactate synthase by sulfometuron methyl is more characteristic of allosteric inhibitors (29) than the slow-binding inhibition associated with reaction-intermediate analogs (28). The slow phase of inhibition could be due to a slow change in the oligomeric state of acetolactate synthase, as is induced in isozyme I by FAD (30), or it could simply be due to a "tightening" of the interaction between the enzyme and sulfometuron methyl.
Since there is no net oxidation or reduction in the acetolactate synthase reaction, the absolute dependence on FAD is unusual. Two similar examples of carbon-carbon lyases that require FAD but catalyze reactions that involve no net oxidation have been reported, tartronate-semialdehyde synthase (EC 4.1.1.47) (31) and mandelonitrile lyase (EC 4.1.2.10) (32). Presumably, the FAD serves as a transient carbanion trap (electron sink) during the acetolactate synthase reaction, as has been proposed for a number of flavoproteins (33), but without concomitant oxidation of the carbanion intermediate.