Inhibition of acetolactate synthase isozyme II from Escherichia coli by a new azido-photoaffinity sulfonylurea

Article abstract of DOI:10.1006/bioo.1997.1071

The sulfonylurea herbicides are very patent inhibitors of acetolactate synthase (ALS). These compounds have been reported as 'extraneous inhibitors' due to the fact that their inhibition site corresponds to neither the catalytic site nor the regulatory sites of the enzyme. So far, the complexity of the ALS reaction and the reversible binding mode of sulfonylureas have hampered any attempt to locate the inhibitor domain. Toward this goal, a photoactivatable azidosulfonylurea has been synthesized. The azido derivative was analyzed for its photochemical and in vitro biological properties toward the bacterial ALS isozyme II. Similar to other ALS inhibitors, azidosulfonylurea potently inhibited ALS II with estimated initial and final dissociation K(i) constant values of 52 and 300 nM, respectively, and slowly inactivated the enzyme. After inhibition, removal of the free azido inhibitor and precipitation with ammonium sulfate of the azidosulfonylurea/ALS II complex led to complete though slow recovery of the enzyme activity. Following photoreaction of the inhibited complex and removal of the free inhibitor under the same conditions, the bacterial enzyme conversely exhibited stable inactivation. These results suggest that the newly synthesized azidosulfonylurea is capable of undergoing covalent reaction with ALS II, and hence it might be useful, once radiolabeled, to shed light on the inhibitor binding site of ALS.

Full text of DOI:10.1006/bioo.1997.1071

BIOORGANIC CHEMISTRY 25, 261–274 (1997)
ARTICLE NO. BH971071
Inhibition of Acetolactate Synthase Isozyme II from Escherichia
coli by a New Azido-Photoaffinity Sulfonylurea
Florence Ortega1 and Jean Bastide
´
Centre de Phytopharmacie–CNRS URA 461, Universite´ de Perpignan,
52 Avenue de Villeneuve, 66860 Perpignan Cedex, France
Received May 23, 1997
The sulfonylurea herbicides are very potent inhibitors of acetolactate synthase (ALS). These
compounds have been reported as ‘‘extraneous inhibitors’’ due to the fact that their inhibition
site corresponds to neither the catalytic site nor the regulatory sites of the enzyme. So far,
the complexity of the ALS reaction and the reversible binding mode of sulfonylureas have
hampered any attempt to locate the inhibitor domain. Toward this goal, a photoactivatable
azidosulfonylurea has been synthesized. The azido derivative was analyzed for its photo-
chemical and in vitro biological properties toward the bacterial ALS isozyme II. Similar to
other ALS inhibitors, azidosulfonylurea potently inhibited ALS II with estimated initial and
final dissociation Ki constant values of 52 and 300 nM, respectively, and slowly inactivated
the enzyme. After inhibition, removal of the free azido inhibitor and precipitation with
ammonium sulfate of the azidosulfonylurea/ALS II complex led to complete though slow
recovery of the enzyme activity. Following photoreaction of the inhibited complex and removal
of the free inhibitor under the same conditions, the bacterial enzyme conversely exhibited
stable inactivation. These results suggest that the newly synthesized azidosulfonylurea is
capable of undergoing covalent reaction with ALS II, and hence it might be useful, once
radiolabeled, to shed light on the inhibitor binding site of ALS.
1997 Academic Press
INTRODUCTION
The sulfonylurea herbicides are characterized by low use rates (typically in the
range of 2 to 75 g/ha) and low mammalian toxicity. These highly active chemicals
inhibit the first common enzyme in branched-chain amino acid biosynthesis, aceto-
lactate synthase (EC 4.1.3.18) (ALS) (1). ALS catalyzes the thiamine pyrophosphate
(TPP)-dependent condensation of two molecules of pyruvate to form acetolactate
or the condensation of pyruvate and 2-ketobutyrate to form Ͱ-aceto-Ͱ-hydroxybu-
tyrate. In addition the enzyme requires flavine adenine dinucleotide (FAD) and
Mg2 for catalysis although its reaction involves no overall change in redox (2).
ϩ
Sulfonylureas have been shown to exhibit slow, tight-binding properties to both
plant ALS (3, 4) and the isozyme II of ALS from Salmonella typhimurium (5). In
time course experiments, the inhibition of ALS isozyme II from S. typhimurium
by the classical sulfometuron methyl could best be modeled as a biphasic process
involving the rapid formation of an initial relatively weak, enzyme/inhibitor complex
that slowly isomerizes to a more tightly bound form.
1 To whom correspondance should be addressed. Fax: (33) 4 68 66 22 23. E-mail: fortega@univ-perp.fr.
261
0045-2068/97 $25.00
Copyright 1997 by Academic Press
All rights of reproduction in any form reserved.

´
ORTEGA AND BASTIDE
262
One intriguing feature of sulfonylureas, although not unprecedented (cf. 6), is
that these herbicides do not bear any striking resemblance to any of the cofactors
or substrates bound by the enzyme. The key to identifying the nature of the binding
site for these inhibitors came with the determination of the sequence of poxB gene
which encodes for pyruvate oxidase from Escherichia coli (7). This sequence shares
substantial homology with ilvB, ilvG, and ivll, the genes that code for the large
subunits of the three major isozymes of bacterial ALS. Unlike ALS, pyruvate
oxidase uses its flavin for normal redox chemistry and binds one additional redox
cofactor, ubiquinone-40. Based on considerations of the expected chemistry of the
pyruvate oxidase reaction, it has been shown that water soluble quinones also bind
to ALS, as do sulfonylureas, and inhibit the enzyme, though not in a time-dependent
manner (8). Thus, it would appear that the unusual herbicide binding site of ALS
is derived from an evolutionary vestige of a quinone binding site of pyruvate oxidase.
Therefore sulfonylureas were termed ‘‘extraneous site’’ inhibitors (2) in that they
bind in large part to a site outside the enzyme’s active site; such extraneous site
does not serve a known regulatory function. The major drawback of extraneous
site inhibitors is that their mode of action upon the enzyme is hardly understood,
hence in the absence of a high-resolution X-ray crystal structure of the enzyme/
inhibitor complex, the rational design of new, more potent inhibitors is hampered.
Therefore, the most reasonable approach to obtaining topographical information
is covalent modification by specific reagents.
In an effort to locate the ALS binding domain for sulfonylureas, we have synthe-
sized an azido-photoaffinity sulfonylurea. Photoaffinity azido derivatives are known
to undergo covalent reactions with proteins through reactive nitrene intermediates
that are easily generated by low-intensity ultraviolet light (9). The affinity probe
was analyzed for its inhibition potency, selectivity, and reversibility toward the
activity of ALS isozyme II from E. coli.
MATERIALS AND METHODS
Commercial Reagents and Materials
Reagent grade chemicals were purchased from Sigma Chemical Co. -Tryptophan
L
was purchased from Merck. Cyanuric chloride, o-tolylsulfonylisocyanate, and so-
dium azide were obtained from Aldrich-Chemie. PD-10 Sephadex G25 minicolumns
were purchased from Pharmacia LKB. IR spectra were recorded on a Perkin–Elmer
1
1600 FTIR series model instrument and H and 13C NMR spectra on a Jeol EX
400 spectrophotometer.
2,4-Dichloro-6-methoxy-1,3,5-s-triazine 2
Sodium hydrogenocarbonate (33.6 g, 0.4 mol) was dissolved in a water/methanol
(200/25, v/v) solvent mixture (225 ml) at 0ЊC. Cyanuric chloride (36.8 g, 0.2 mol) was
then carefully added. The reaction mixture was stirred for 1 h at room temperature
(Յ20ЊC). A large volume of water was then added to precipitate the formed
methoxy-s-triazine. After washing twice with water, the compound was removed

INHIBITION OF ALS BY A NEW AZIDOSULFONYLUREA
263
from the reaction by filtration, purified by crystallization in 300 ml n-heptane, and
finally isolated by filtration (21.3 g, 0.12 mol, 60% yield), mp 89ЊC.
2-Chloro-4-amino-6-methoxy-1,3,5-s-triazine 3
The dichloro-s-triazine 2 (21.3 g, 0.12 mol) was dissolved in acetone (130 ml) at
0ЊC. Aqueous ammonia (28% of 34.6 ml, 0.27 mol) was added dropwise into the
stirred solution and allowed to react with the dichloro-s-triazine for 30 min in a
ice bath to keep the temperature below 45ЊC. The resulting 2-chloro-4-amino-6-
methoxy-1,3,5-s-triazine was filtered off and washed twice with water (16.15 g, 0.1
mol, 83% yield), mp 240ЊC. Anal. calcd: C, 29.9; H, 3.11; N, 34.89. Found: C, 30.0;
H, 3.08; N, 34.92.
2-Amino-4-azido-6-methoxy-1,3,5-s-triazine 4
The monochloro-s-triazine 3 (3.21 g, 0.02 mol) was dissolved in dimethylform-
amide (70 ml) before the addition of sodium azide (1.3 g, 0.02 mol). The reaction
mixture was heated to 110ЊC with stirring for 3 h. After cooling, the sodium chloride
side product was removed by filtration. The addition of a large volume of water
gave a precipitate which was filtered. The filtered compound was redissolved in
warmed chloroform to separate the insoluble monochloro-s-triazine precursor (as
shown with the negative test of Belstein). After filtration, the chloroformic fraction
was dried over sodium sulfate and evaporated under vacuum (2.4 g, 0.014 mol, 72%
1
yield), mp 225ЊC, IR (CHCl3 , cmϪ ); ␯ NH2 (3520, 3410), ␯ CH3 (2930), ␯ N3 (2120),
1
␯ CuC (1610); H RMN (CDCl3) ͳ ppm 3.95 (s, 3H, OCH3), 5.4 (m, 2H, NH2);
13C RMN (CDCl3) ͳ ppm 55.3 (1C, OCH3), 169.3, 171.2, 172.6 (3C, CuN).
[(4-Azido-6-methoxy-1,3,5-s-triazine-2-yl)aminocarbonyl]-o-tolylsulfonamide 6
Azido-s-triazine 4 (50 mg, 0.3 mmol) was dissolved in toluene (8 ml) at 35–40ЊC
before addition of o-tolylsulfonylisocyanate (91 Ȑl, 0.118 g, 0.6 mmol). The reaction
mixture was stirred and allowed to react for 15 min. The excess triazine was then
removed by filtration and the filtrate solvent was evaporated in the vacuum. The
azidosulfonylurea was purified by crystallization in toluene and isolated by filtration
1
(40 mmg, 0.1 mmol, 36–62% yield). IR (CHCl3 , cmϪ ) : ␯ NH (3390), ␯ N3 (2147),
1
␯ CuO (1736), ␯ NuN (1583), ␯ CuN (1567); H RMN (CDCl3) ͳ ppm 1.65
(s, 3H, CH3), 4.15 (s, 3H, OCH3), 7.2–8.35 (m, 4H, aromatic protons).
Isolation of AHAS II from E. coli
The E. coli strain JA 221 (rkϪ mkϩ trpE5 galϪ leuϪ recAϪ), containing the pRL
106 plasmid (genotype ilvG468 [IIvGϩ]) (10), was kindly provided by R. P. Lawther
(University of South Carolina, Columbia, SC). This strain was grown at 30ЊC for 20
h in a 1-L fermentor containing minimal medium, i.e., 7 g/L K2HPO4 , 3 g/L KH2PO4 ,
0.1 g/L NaCl, 1 g/L (NH4)2SO4 , and 0.1 g/L MgSO4 , adjusted to pH 7.0 and supple-
mented with glucose at a final concentration of 0.7%. -Tryptophan and ampicillin
L
were added at concentrations of 15 and 50 mg/L, respectively. Cells were collected
by centrifugation(15,000g for 20min) and resuspended in 0.1 M potassium phosphate

´
ORTEGA AND BASTIDE
264
buffer, pH 8.0, containing 20 ȐM FAD, 10 mM EDTA, 1 mM DTT, and 20% (v/v)
glycerol. Isolation and partial purification of ALS II was carried out at 4ЊC or, as
indicated, at ice temperature. Ice-cooled cells were sonicated (Heat Systems Ultra-
sonics, Inc., W-375 sonicator; 70% duty cycle) three times for 30 s and then for 1 min.
The homogenate was centrifuged at 10,000g for 30 min, then the pellet was discarded
and the supernatant brought to 60% saturation with ammonium sulfate. The solution
was centrifuged at 18,000g for 10 min and the ammonium sulfate-precipitated pellet
was dissolved in 0.1 M potassium phosphate buffer, pH 8.0, containing 20% (v/v)
glycerol and stored at Ϫ20ЊC. Samples were used within 2 weeks of storage. Specific
activities of preparations varied between 0.5 and 1.5 U/mg.
Protein concentrations were determined by the Coomassie brilliant blue binding
assay (Bio-Rad) of Bradford (11) using bovine serum albumin (BSA) as a standard.
Irradiation Procedure
Photolysis was carried out with a ultraviolet lamp (␭ 254 nm, 2 ϫ 15 W) 10 cm
from the reaction watch glasses, which were held on ice. Azidosulfonylurea (5 ȐM)
in 10 ml 0.1 M potassium phosphate buffer, pH 8.0 [20% (v/v) glycerol], containing
0.5 mM TPP, 20 ȐM FAD, 0.1 mM MgCl2 , and 50 mM pyruvate was irradiated in
the presence or absence of BSA (50 mg). Aliquots (1.5 ml) were removed at 15-s
intervals for 1 min then kept in the dark. The residual azidosulfonylurea incubated
without BSA was analyzed spectrophotometrically at ␭ ranging from 230 to
300 nm using a Hewlett Packard 8452A spectrophotometer. The spectroscopic
characteristics of the light-modified azidosulfonylurea were determined after photol-
ysis was complete. Kinetics were monitored by injecting the irradiated solution
(20 Ȑl) onto an Ultrabase C18 HPLC column (5 Ȑm, 4.6 mm ϫ25 cm) eluted with
a linear gradient of acetonitrile/water/acetic acid (70/30/0.5, v/v/v) at a flow rate
of 1 ml/min (retention time 4.8 min). Solutions containing BSA were first acidified
to pH 3.0 with 6 N H2SO4 (35 Ȑl) to precipitate proteins. Sulfonylurea was then
extracted from the mixture using dichloromethane (2 ϫ 2.5 ml) and filtered prior
to evaporation of the solvent under vacuum. The extract was redissolved in acetoni-
trile (300 Ȑl) and analyzed by HPLC as described above.
Deactivation of ALS II
ALS (0.02 U/ml) in 0.1 M potassium phosphate buffer, pH 8.0 [20% (v/v) glycerol,
5 g/L BSA], was irradiated as described before. Aliquots (50 Ȑl) were withdrawn
at 12-s intervals and tested for enzyme activity (see below) following addition of
0.5 mM TPP, 20 ȐM FAD, 0.1 mM MgCl2 , and 50 mM pyruvate after reaction for
30 min at 37ЊC. Photolysis was otherwise routinely performed by illuminating for
1 min the enzyme solution held on ice in the aforementioned potassium phosphate
buffer either containing the azidosulfonylurea at the appropriate concentration or
not for the control.
Enzyme Assays
ALS assays were carried out at 37ЊC in 200 Ȑl of 0.1 M potassium phosphate
buffer, pH 8.0 [20% (v/v) glycerol, 5 g/L BSA], containing 0.5 mM TPP, 20ȐM
FAD, 0.1 mM MgCl2 , and 50 mM pyruvate. The assays were stopped by the addition

INHIBITION OF ALS BY A NEW AZIDOSULFONYLUREA
265
of 20 Ȑl of 6 N H2SO4 and incubated at 60ЊC for 10 min. The reaction product
(acetoin) was then determined according to the method of Westerfeld (12) modified
as follows: 350 Ȑl of 0.5% (w/v) creatine in distilled water, 350 Ȑl of 5% (w/v)
Ͱ-naphthol in 2.5 N NaOH, and 300 Ȑl distilled water were added to the solution
prior to mixing and heating at 60ЊC for 15 min. The absorbance of the colored
complex was measured at 530 nm. Assays were carried out in triplicate.
ALS inhibition assays were performed in 0.1 M potassium phosphate buffer, pH
8.0 [20% (v/v) glycerol, 5 g/L BSA], containing all the cofactors and substrate as
described above and azidosulfonylurea at the appropriate concentration. Aliquots
were withdrawn at definite times and tested for enzyme activity. Experimental
details are given in the legends of the figures.
Experiments for Recovery of ALS Activity
E. coli AHAS II (0.054 U) in 0.1 M potassium phosphate buffer, pH 8.0 (2 ml)
[20% (v/v) glycerol]), containing 0.5 mM TPP, 20 ȐM FAD, 0.1 mM MgCl2 , and
50 mM pyruvate, was incubated in the dark with either 200 nM or 5 ȐM azidosulfo-
nylurea. Controls were the same but without inhibitor. After incubation for 2 h at
20ЊC an aliquot of each solution (10 to 50 Ȑl) was taken for enzyme assay. Half of
the incubation volume was then removed from the solutions and exposed for 1 min
to ultraviolet light, while the other part was kept protected from light. Each fraction
(0.8 ml) was then exchanged by gel filtration down a PD-10 column of Sephadex.
PD-10 Sephadex G25-column buffer was 0.1 M potassium phosphate buffer, pH
8.0, containing 20 ȐM FAD and 20% (v/v) glycerol. The protein eluates (2 ml)
were then collected and supplemented with 20 mg BSA. The solutions were brought
to 50% saturation with ammonium sulfate and stood for 1 h on ice. Precipitated
protein was then collected by centrifugation (15,000g for 10 min) and redissolved
in 0.1 M potassium phosphate buffer (2 ml), pH 8.0 [20% (v/v) glycerol]. Protein
and recovered enzyme activity were then measured in aliquots (50 to 200 Ȑl)
following reaction vs time at 37ЊC with 50 mM pyruvate, 0.5 mM TPP, 20 ȐM FAD,
and 0.1 mM MgCl2 .
Data Processing
Data were fitted by a non-linear regression analysis using the Deltagraph Pro3
program (Midvision, Inc.).
RESULTS AND DISCUSSION
Synthesis of Azidosulfonylureas
The photolabile azido group can be bound on either the aromatic cycle or the
heterocycle of sulfonylurea. Based on the structure–activity relationships of the
sulfonylurea chemical class and criteria of easiness for synthesis, we have chosen
to introduce the marker function on the amino-s-triazine cycle. The synthesis of
a series of different amino-s-triazines from cyanuric chloride was well described

´
266
ORTEGA AND BASTIDE
SCHEME 1. Synthesis of 2-azido-4-amino-6-methoxy-1,3,5-s-triazine 4.
previously (13). Actually, cyanuric chloride is better classed as an acyl halide than
an alkyl halide, since it readily reacts with amines, alcohols, and sodium azide.
However, special care must be taken with both the order of stepwise replacement
and the nature of the nucleophilic group that substitutes the chloride atoms of the
s-triazine. Indeed, the introduction of new substituents progressively decreases the
reactivity of the chlorides on the cyanuric nucleus so that the nucleophilicity of the
function to be introduced and the temperature both need to be correspondingly
increased. Based on these considerations, we have synthesized the 2-azido-4-amino-
6-methoxy-1,3,5-s-triazine as follows: (a) cyanuric chloride 1 in the presence of
sodium hydrogenocarbonate reacted with methanol at a controlled temperature
that did not exceed 20ЊC to produce the 2,4-dichloro-6-methoxy-1,3,5-s-triazine 2,
(b) the dichloro-s-triazine 2 was then reacted with ammonia at a higher temperature
although not above 45ЊC to yield the 2-chloro-4-amino-6-methoxy-1,3,5-s-triazine
3, and (c) the s-triazine 3 finally was heated at 110ЊC in DMF to allow the displace-
ment of the last chloride by sodium azide and give the expected 2-azido-4-amino-
6-methoxy-1,3,5-s-triazine 4 (Scheme 1).
Among the numerous variations on the method of obtaining sulfonylureas, one
successful synthetic method is the coupling of an arylsulfonamide to an aminohetero-
cycle as follows (14): (1) reaction of an amino-s-triazine with the arylsulfonylisocya-
nate of a sulfonamide, (2) coupling of the methyl- or the phenylcarbamate of an
aminoheterocycle to a sulfonamide in the presence of a tertiary amine, or (3)
thermal reaction of the phenylcarbamate of a sulfonamide to an amino-s-triazine.
Given the location of the azido substituent on the s-triazine cycle, we considered

INHIBITION OF ALS BY A NEW AZIDOSULFONYLUREA
267
SCHEME 2. Synthesis of azidosulfonylurea 6.
the method with sulfonylisocyanate best. We used commercial o-tolylsulfonylisocya-
nate. The expected sulfonylurea 6 was readily obtained by reaction in toluene of
the arylsulfonylisocyanate 5 with the azido-s-triazine 4 (Scheme 2). Preliminary
enzymatic tests indicated that the azidosulfonylurea bearing a methyl group on its
phenyl cycle potently inhibited the isozyme II of ALS from E. coli (data not shown).
Photoactivation Properties of the Azidosulfonylurea
The azido groups can be easily activated by low-intensity ultraviolet light (␭ 254
nm) to yield highly reactive nitrene radicals and azote molecules (9). Once in
solution, the free radicals are powerful nucleophiles that undergo covalent reaction
with the amino acid side chains of proteins. The photoincorporation of azido com-
pounds depends on both the duration of illumination and the effect of proteins
that absorb light radiations. Furthermore, illumination must not be damaging to
the proteins, so that suitable conditions have to be established first. Toward this
goal, we investigated the photolysis vs time of the azidosulfonylurea (5 ȐM) exposed
to ultraviolet light both in the absence and in the presence of BSA (5 mg/ml) in
standard enzyme reaction medium at 0ЊC. When proteins were omitted from the
solution, the photoreaction was monitored either spectrophotometrically or by
HPLC. The spectroscopic characteristics of the initial and the photoreacted azidosul-
fonylurea are reported in Table 1. The solution containing BSA was analyzed by
HPLC following acidic precipitation of proteins, extraction with dichloromethane
and concentration in acetonitrile of the sulfonyulurea. The photoreaction kinetics
(ϮBSA) monitored by HPLC indicate that in either instance the photoactivation
adequately follows a first-order process which is slowed down in the presence of
proteins (Fig. 1A). Examination of the half-life of the photoreactant gave estima-
tions of t1/2(N3) Ȃ 5 s and t1/2(N3)ϩBSA Ȃ 25 s, indicating that the rate of activation
is five times lower in the solution containing the proteins. Yet, the azido group was
totally reacted after 1-min exposure to ultraviolet light for the latter.

´
268
ORTEGA AND BASTIDE
TABLE 1
Spectroscopic Characteristics of the Azidosulfonylurea
Sulfonylurea
Nonactivated
Photoactivated
␭max (nm)
238
42,200
21,600
234
39,900
11,600
1
1
␧
(
(MϪ cmϪ
)
␭
max)
1
1
␧
(
(MϪ cmϪ
)
␭
254 nm)
Exposure of ALS to ultraviolet light for the same period of time can still alter
its activity. Therefore, we examined the enzymatic deactivation under the same
conditions. The residual activity vs time of ALS was measured after incubation of
proteins at 0ЊC and irradiation (Fig. 1B). Clearly, ultraviolet light induces a strong
decay in the enzyme activity. The assay inactivation curve was described by a first-
order transition between the native and the UV-inactivated enzyme. The analysis
of the process gave an estimate t1/2(ALS II) Ȃ 115 s, showing that the rate of
ALS inactivation is four to five times slower than that of the photoactivation of
azidosulfonylurea in the presence of a similar concentration of proteins. Exposure
of the enzyme to ultraviolet light for 1 min actually resulted in less than 35% loss
of activity.
Inhibition of ALS II by the Azidosulfonylurea
Azido derivatives can be analyzed for either their biological activity or their
chemical activity depending upon photoactivation. In order to investigate this aspect,
we carried out experiments in which the azidosulfonylurea/ALS II complex was
FIG. 1. Time courses of azidosulfonylurea photolysis and ALS II photoinactivation induced by 254-
nm ultraviolet light. (A) Azidosulfonylurea (5 ȐM) was irradiated in the absence (᭡) or presence (᭝)
of BSA (5 mg/ml) at 0ЊC and tested for photoreaction vs time by HPLC as described under Materials
and Methods. (B) Photodeactivation vs time of ALS II in solution with BSA (5 mg/ml) (᭹) was carried
out at 0ЊC and analyzed for activity as indicated under Materials and Methods.

INHIBITION OF ALS BY A NEW AZIDOSULFONYLUREA
269
FIG. 2. Time dependency of the inhibition of ALS II by azidosulfonylurea. ALS II (0.01 U/ml) in
solution with substrate and cofactors at standard concentrations (see Materials and Methods) was
incubated in the absence (control) or presence of azidosulfonylurea (200 nM) at 37ЊC. Acetoin produced
was determined from aliquots (180 Ȑl) every 10 min of reaction as given under Materials and Methods.
After 15 min of inhibition, half of the reaction mixture was withdrawn, cooled down on ice, and exposed
to 254-nm ultraviolet light for 1 min. The same treatment was applied to the other part of the incubation
mixture except for irradiation. Samples were then processed for further incubation at 37ЊC, after which
aliquots (120 to 25 Ȑl) were taken at definite times (20 min to 2 h, respectively) for determination of
enzyme activity. Comparison between the activity of control and that of solution containing azidosulfonyl-
urea gave the estimated inhibition vs time for the nonirradiated (᭹) and the irradiated (᭡) sample.
illuminated in the course of inhibition. Figure 2 depicts the progress time curves
of ALS inhibition induced by azidosulfonylurea with the assay solution either kept
away from light or submitted to illumination at an early stage of inhibition (15
min). Assay time courses were markedly biphasic over a period of 2 h, as is typically
observed with sulfonylureas. These results also show that (1) the free activated
derivative was no longer biologically active, as indicated by the fact that the inhibi-
tion process was readily interrupted following illumination, thereby leading to a
lower steady-state level than that of the inhibition that proceeded in the dark, and
(2) the level of enzymatic inhibition obtained prior to photolysis was kept constant
for an extended period of time after photolysis of the sulfonylurea, i.e., ALS did
not readily recover its activity. It thus seems possible to discriminate between
the effect of the initial azidosulfonylurea and its corresponding photoactivated
derivative, as defined by their different abilities to inhibit the enzyme activity. When
nonspecific binding of nitrenes to proteins occurred, it did not affect the basic
catalytic properties of ALS.
In an additional experiment, we conducted the characterization of the inhibition
of E. coli ALS II by azidosulfonylurea protected from light. A detailed analysis

´
ORTEGA AND BASTIDE
270
FIG. 3. Inhibition of ALS II by azidosulfonylurea with respect to pyruvate. (A) ALS II (0.11 U/ml)
in solution containing all cofactors at standard concentrations (cf. Materials and Methods) and pyruvate
(15 to 50 mM) was incubated for 5 min at 37ЊC with various concentrations of azidosulfonylurea: 0 nM
(᭿), 200 nM (᭹), 500 nM (᭡), 1 ȐM (᭜), and 2 ȐM (᭺). Acetoin formed was determined from aliquots
(200 Ȑl) as described under Materials and Methods. (B) ALS II (0.006 U/ml) was incubated for 2 h as
outlined above with azidosulfonylurea at the following concentrations: 0 nM (᭿), 40 nM (᭹), 70 nM
(᭡), 100 nM (᭜), and 200 nM (᭺). Acetoin formed was determined from aliquots (100 Ȑl) as given
under Materials and Methods.
of inhibition of ALS II as a function of pyruvate is shown in Figs. 3A and 3B.
Lineweaver–Burk double-reciprocal plots indicate a noncompetitive (or ‘‘mixed-
type’’) inhibition with respect to the substrate for both the initial and the final
inhibition by azidosulfonylurea. Analysis of these data gave estimated Ki values of

INHIBITION OF ALS BY A NEW AZIDOSULFONYLUREA
271
300 and 52 nM, respectively. For comparison, inhibition of the enzyme by the classic
thifensulfuron methyl led to the same type of patterns, with initial and final Ki
values of 101 and 8.5 nM, respectively.
Reversibility of the Inhibition of ALS II Induced by the Azidosulfonylurea
Nonspecific binding of photoactivated derivatives to proteins depends on both
the interactions between the enzyme and its ligand and the location of the photoacti-
vatable function on the marker molecule. In the case of ALS II, inhibition appeared
to be specific to the nonactivated azidosulfonylurea (cf. Fig. 2). However, despite
the fact that the enzyme activity was not readily recovered after photolysis of the
inhibitor bound to protein, evidence was not provided for (1) covalent reaction
between the ALS II and the azidosulfonylurea or (2) specific photoincorporation
assuming covalent attachment had occurred. As reported previously (4), sulfonyl-
ureas induce a quite stable inactivation of the enzyme without involving covalent
interaction between these molecules and ALS. Actually, since azidosulfonylurea
inhibits ALS II in a time-dependent manner, the reversed enzyme activity recovery
reaction is predicted to occur through a slow reactivation process following removal
of the active inhibitor from the reaction. Stable enzyme inactivation might then
simply reflect the alteration of the enzyme conformation due to the original inhibi-
tion rather than to covalent interaction between ALS II and the photoactivated in-
hibitor.
In a previous paper (15), we showed that the activity of ALS II initially inhibited
by the thifensulfuron methyl sulfonylurea could not be recovered after removal of
this compound from reaction, unless proteins were precipitated with ammonium
sulfate. Therefore we used this method to investigate the stability of the inhibited
ALS II/azidosulfonylurea complex. The enzyme was inhibited until a steady-state
residual activity was reached, then the solution was split into two parts: one was
directly exchanged by gel filtration for fresh buffer down a rapid minicolumn (G25
PD-10), while the other was exposed to ultraviolet light prior to gel filtration.
Proteins were precipitated with ammonium sulfate and tested for enzyme reactiva-
tion. The kinetics vs time of ALS activity are shown in Figs. 4A and 4B for the
nonactivated and the photoactivated sulfonylurea, respectively. Figure 4A clearly
indicates that ALS II slowly recovers its activity when the solution containing the
inhibited complex was not submitted to ultraviolet light. The progress curve data
of Fig. 4A were adequately fitted to Eq. [1] in accordance with a slow activation
model (cf. 16), where P is product formed, Vs is the final steady-state velocity, and
kr is the apparent first-order rate constant of reactivation. Reactivation occurred
faster when ALS II was inhibited by a low compared to a high concentration of
1
sulfonylurea (kr Ȃ 0.05 and 0.03 minϪ with initial concentrations of 200 nM and
5 ȐM, respectively), in agreement with the fact that the observed kr value is inversely
related to the concentration of the inhibited complex and hence, to the initial
sulfonylurea concentration.
k t
r
P ϭ Vst Ϫ (Vs/kr)(1 Ϫ e )
[1]
On the contrary, ALS II shows a stable residual activity after the azidosulfonyl-

´
ORTEGA AND BASTIDE
272
FIG. 4. Progress curves for the recovery of the activity of ALS II after inhibition by azidosulfonylurea.
(A) ALS II was incubated in the absence (᭿) for control activity and the presence of 200 nM (᭹) and
5 ȐM (᭡) azidosulfonylurea for 2 h at 20ЊC. After removal by rapid gel filtration and precipitation with
ammonium sulfate of unbound azidosulfonylurea from samples, enzyme activity was initiated by the
addition of substrate and cofactors at standard concentrations and was monitored for 2 h. (B) The
experimental procedure was as described for (4) except for irradiation by 254-nm ultraviolet light of
the reaction mixture for 1 min after inhibition and prior to removal of azidosulfonylurea from samples.
Experimental details are given under Materials and Methods. Data refer to 200-Ȑl reaction aliquots.

INHIBITION OF ALS BY A NEW AZIDOSULFONYLUREA
273
urea in complex with the enzyme had been photoactivated (Fig. 4B). These results
indicate that irreversible inactivation of the enzyme, and hence covalent reaction,
has most likely occurred. Analysis by linear regression of data in Fig. 4B gave
estimated residual inhibitions of 31 and 54% for corresponding initial inhibition
values, i.e., before gel filtration values of 57 and 97% with sulfonylurea concentra-
tions of 200 nM and 5 ȐM, respectively. It thus appears that nearly half of the
ALS inactivation is maintained due to the photoactivation of the azido derivative.
However, if this residual inhibition potency were to reflect the specific yield of the
photoaffinity reaction between the azidosulfonylurea and ALS II, a slow activity-
recovery rate profile would still be observed. These results rather suggest that part
of the azido derivative is lost due to covalent binding outside but close to the
herbicide site, thereby retarding enzymatic reactivation, and/or ALS II is stabilized
by the azidosulfonylurea against the light-dependent inactivation, i.e., the residual
activity of the inhibited enzyme after light exposure is proportionally higher than
that of the control.
CONCLUSION
The present work opens a wide scope for future research on photoaffinity labeling
of ALS. ALS II from E. coli proved to be sufficiently stable toward UV light to
allow us to characterize its biochemical interaction with the azido derivative. Our
results support the idea that the newly synthesized azidosulfonylurea can be used
as a specific and potent tool for the identification of the sulfonylurea binding domain
of the enzyme. Experiments with the ALS from barley (Hordeum vulgare) also led
to same type of results (data not shown). The next step will be to carry on with
the synthesis of the radiolabeled azidosulfonylurea in order to directly investigate
the binding of the inhibitor to the enzyme.
REFERENCES
1. Ray, T. B. (1984) Plant Physiol. 75, 827–831.
2. Schloss, J. V. (1990) Pestic. Sci. 29, 283–292.
3. Hawkes, T. R., Howard, J. L., and Pontin, S. E. (1989) in Herbicides and Plant Metabolism (SEB
Seminar Series) ( Dodge, A. D., Ed.), pp. 113–117, Cambridge Univ. Press, Cambridge, UK.
4. Durner, J., Gailus, V., and Goger, P. (1991) Plant Physiol. 95, 1144–1149.
¨
5. La Rossa, R. A., and Schloss, J. V. (1984) J. Biol. Chem. 259, 8753–8557.
6. Schloss, J. V. (1989) in Target Sites of Herbicide Action (Boger and Sandman, Eds.), pp. 165–245,
¨
CRC Press, Boca Raton, FL.
7. Grabau, C., and Cronan, J. E., Jr. (1986) Nucleic Acids Res. 14, 5449–5460.
8. Schloss, J. V., Ciskanik, L. M., anf Van Dyk, D. E. (1988) Nature 331, 360–362.
9. Bell, J. E., and Bell, E. T. (1988) in Proteins and Enzymes, Chap. 8, pp. 184–204, Prentice Hall,
Englewood Cliffs, NJ.
10. Lawther, R. P., Calhoun, D. H., Gray, J., Adams, C. W., Hauser, C. A., and Hatfield, G. W. (1982)
J. Bacteriol. 149, 294–298.
11. Bradford, M. (1976) Anal. Biochem. 72, 248–254.

´
ORTEGA AND BASTIDE
274
12. Westerfeld, W. W. (1945) J. Biol. Chem. 161, 495–502.
13. Dudley, J. R., Thurston, J. T., Schaffer, F. C., Holm-Hansen, D., Hull, C. J., and Adams, P. (1951)
J. Am. Chem. Soc. 73, 2986–2990.
14. Hay, J. V. (1990) Pestic. Sci. 29, 247–261.
15. Ortega, F., Bastide, J., and Hawkes, T. R. (1996) Pestic. Biochem. Physiol. 56, 231–242.
´
16. Morrison, J. F., and Cleland, W. W. (1983) Biochemistry 22, 5507–5513.