Design, synthesis and evaluation of transition-state analogue inhibitors of Escherichia coli γ-glutamylcysteine synthetase

BIOORGANIC &

MEDICINAL

CHEMISTRY

Bioorganic & Medicinal Chemistry 6 (1998) 1935±1953

Design, Synthesis and Evaluation of Transition-state Analogue

Inhibitors of Escherichia coli ꢀ-Glutamylcysteine Synthetase

Nobuya Tokutake, Jun Hiratake,* Makoto Katoh, Takayuki Irie,

Hiroaki Kato and Jun'ichi Oda

Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan

Received 8 May 1998; accepted 18 July 1998

AbstractÐPhosphinic acid-, sulfoximine- and sulfone-based transition-state analogues were synthesized and evaluated

as inhibitors of Escherichia coli ꢀ-glutamylcysteine synthetase. These compounds have a carboxyl function at the ꢁ-

carbon to the tetrahedral central hetero atom so as to mimic the carboxyl group of the attacking cysteine in the tran-

sition state. The phosphinic acid- and the sulfoximine-based compounds were found to be potent ATP-dependent

inactivators, both showing a slow-binding kinetics with overall anities and second-order inactivation rates of one to

two orders of magnitude greater than those of l-buthionine (SR)-sulfoximine (l-BSO). The sulfone was a simple

reversible inhibitor without causing ATP-dependent enzyme inactivation, but its anity toward the enzyme was still

®ve times greater than that of l-BSO, indicating that the ꢁ-carboxyl function plays a key role in the recognition of the

inhibitors by the enzyme. The sulfoximine with (S)-ꢁ-carbon to the sulfur was synthesized stereoselectively, and the

two diastereomers with respect to the chiral sulfur atom were separated as a cyclic sulfoximine derivative. The sulfox-

imine with R-con®guration around the sulfur served as an extremely powerful ATP-dependent inactivator with an

overall inhibition constant of 39 nM and an inactivation rate of 6750 M ¹s ¹, which correspond to 1260-fold higher

anity and almost 1400-fold greater inactivation rate as compared with l-BSO. The sulfoximine with (S)-sulfur was a

simple reversible inhibitor with an inhibition potency comparable to that of the sulfone. The synthesis and inhibition

Introduction

glutathionylspermidine synthetase,⁵trypanothione syn-

thetase,^5c,6and a series of amino acid adding enzymes in

bacterial peptidoglycan biosynthesis.⁷Thus the design and

synthesis of speci®c inhibitors of glutathione biosynthetic

enzymes are of crucial importance not only for use as

therapeutic agents, but also for use as mechanistic probes

for the ADP-forming peptide ligases, in which an amide

bond is formed with concomitant hydrolysis of ATP to

ADP and P_ito supply the thermodynamic driving force.

Glutathione (g-l-glutamyl-l-cysteinylglycine) is synthe-

sized in living organisms from its constituent amino

acids by the consecutive actions of two mechanistically

related ATP-dependent peptide ligases, g-glutamyl-

cysteine synthetase (g-GCS, EC 6.3.2.2) and glutathione

synthetase (EC 6.3.2.3).¹These enzymes are of parti-

cular interest because glutathione plays a pivotal role in

the cellular thiol redox system^1band constitutes a front

line of the detoxi®cation of reactive oxygens and other

xenobiotics such as heavy metals and electrophilic alkyl-

ating agents.²In addition, these enzymes are typical

members of the group of ADP-forming peptide ligases

such as glutamine synthetase,³d-Ala: d-Ala ligase,⁴

We recently reported the synthesis and characterization

of a phosphinic acid transition-state analogue 1 as a

potent mechanism-based inactivator of Escherichia coli

glutathione synthetase.⁸The success in the rational

design of the inhibitor of glutathione synthetase

has directed our attention to the development of trans-

ition-state analogue inhibitors of g-GCS based on

the same rationale. This enzyme, which catalyzes the

ATP-dependent coupling of l-Glu and l-Cys to form

Key words: Escherichia coli g-glutamylcysteine synthetase;

transition-state analogues; phosphinic acid; sulfoximine; slow-

binding inhibition.

*Corresponding author.

PII: S0968-0896(98)00142-4

1936

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

g-l-Glu-l-Cys, is the rate-limiting enzyme in glutathione

biosynthesis, and its inhibition causes glutathione

depletion^2a. Hence its inhibitor is highly important for

biochemical and pharmacological studies.⁹Considering

that the reaction catalyzed by g-GCS is thought to pro-

ceed through the initial formation of g-glutamylphos-

phate intermediate followed by the nucleophilic attack

of l-Cys (Scheme 1), we reasoned that tetracoordinated

phosphinic acid⁸and sulfoximine¹⁰moiety would well

mimic the partially sp³-hybridized carbonyl carbon in

the transition state and that the carboxyl group of the

incoming l-Cys as well as the mercaptomethyl side

chain would be a major recognition element by the

enzyme. We therefore designed the phosphinic acid-,

sulfoximine- and sulfone-based transition-state ana-

logues 2, 3, and 4 with a carboxyl function at the b-carbon

to the central hetero atom.¹¹

inhibitor sulfoximine 3 was elaborated further for

stereoselective synthesis, and its inhibition behavior is

discussed in terms of mechanism-based phosphorylation

of the imino nitrogen at the chiral sulfur atom. The

inhibition pro®le and potency of the chemically synthe-

sized N-phosphoryl sulfoximine 21 is also described.

Results and Discussion

Synthesis of phosphinic acid 2, sulfoximine 3 and sulfone 4

The phosphinic acid 2 was synthesized as shown in

Scheme 2. In order to achieve a stepwise construction of

two dierent C-P bonds in the phosphinic acid 2, we

chose ethyl (diethoxymethyl)phosphinate (5)¹²as the

starting material. The diethoxymethyl group attached to

phosphorus serves as `a masked hydrogen',¹³and after

the ®rst C-P bond is formed with the P-H function of 5,

acid hydrolysis unmasks the second P-H for further

construction of the second C-P bond. Thus, the com-

pound 5 was subjected to base-catalyzed conjugate

addition to ethyl 2-ethylacrylate¹⁴to give the phosphi-

nate 6. Complete acid hydrolysis of 6 followed by

methylation with diazomethane gave the hydrogen

phosphinic acid dimethyl ester 7 as an unstable oil. The

construction of the second C-P bond utilized the epi-

merization-free radical addition of a phosphinate to a

vinylglycine derivative reported for the synthesis of

enantiomerically pure phosphinothricin.¹⁵Thus, the

hydrogen phosphinate 7 was allowed to react with N-

(benzyloxycarbonyl)-l-vinylglycine methyl ester (8)¹⁶by

using tert-butyl peroxybenzoate as a radical initiator to

yield the fully protected phosphinate 9. Alkaline hydro-

lysis followed by hydrogenolysis aorded the phosphinic

acid 2 as a 1:1 mixture of diastereomers (³¹P NMR).

In this paper, we describe the synthesis and evaluation

of these transition-state analogues as inhibitors of E.

coli g-GCS and demonstrate that all the compounds

showed very high anity for the enzyme. In particular,

the sulfoximine-based compound 3 was an extremely

powerful slow-binding inactivator with an unprece-

dented anity and inactivation rate. The most potent

The sulfoximine 3 has two chiral carbons and one chiral

sulfur atom. Starting with l-homocystine, we ®rst syn-

thesized a mixture of four diastereomers with respect to

the chiral sulfur and the chiral carbon b to the sulfur

atom. Scheme 3 outlines the synthesis of the sulfoximine

3. According to the reported procedure¹⁷with slight

modi®cation, methyl 3-hydroxypropanoate (10)¹⁸was

alkylated at the a-position and was converted to the

corresponding

iodide

12.

N-(4-Nitrobenzyloxy-

carbonyl)-l-homocystine methyl ester (13) was reduced

by tri-n-butylphosphine in the presence of water. The

reduction was best carried out in dichloromethane

which had been washed to remove a trace of MeOH and

saturated with water to give the homocysteine derivative

14 quantitatively. The thiol 14 and the iodide 12 were

then coupled to aord the diastereomeric sul®de 15 in

77% yield.

Subsequent oxidation with NaIO₄provided the sulf-

oxide 16 quantitatively. In this stage, an additional

Scheme 1. Proposed reaction mechanisms of g-GCS

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

1937

chiral center was generated at the sulfur atom, but the

four diastereomeric mixture of the sulfoxide 16 was

treated as such with O-mesitylsulfonylhydroxylamine

(MSH)¹⁹to introduce the imino nitrogen at the sulfur

atom. A part (ca. 10%) of the product sulfoximine 17 was

found to cyclize spontaneously during the reaction to yield

the cyclic sulfoximine 18. This side reaction, however, was

successfully utilized to eliminate the need of direct alkaline

hydrolysis of the b-methyl ester in 17, which might risk the

C-S bond cleavage by b-elimination.²⁰Thus, the sulfox-

imine 17 was heated in acetic acid to complete the

cyclization to yield 18 in 76% yield. In this stage, two

sets of diastereomers of 18 could be separated in almost

equal amounts by ¯ash column chromatography (vide

Scheme 2. Reagents and conditions: (a) Ethyl 2-ethylacrylate, NaOEt, EtOH; (b) 12 m HCl, re¯ux; (c) CH₂N₂, Et₂O; (d) tert-Butyl

peroxybenzoate (cat.), xylene; (e) NaOH, THF-H₂O; (f) H₂, 10% Pd on carbon, MeOH±AcOH±H₂O then Dowex 50WÂ8 (H⁺form),

eluted with H₂O.

Scheme 3. Reagents and conditions: (a) LDA, iodoethane, DMPU-THF; (b) CH₃SO₂Cl, Net₃, CH₂Cl₂; (c) NaI, acetone; (d) Tri-n-

butylphosphine, CH₂Cl₂, H₂O; (e) 12, K₂CO₃, DMF; (f) NaIO₄, THF±H₂O; (g) O-Mesitylenesulfonylhydroxylamine (MSH), CH₃CN;

(h) AcOH, 50 ^ꢀC; (i) H₂, 10% Pd on carbon; (j) aq. KOH then Dowex 50WÂ8 (NH₄⁺form), eluted with H₂O.

1938

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

infra), but for the present purpose, all the diastereomers

of 18 combined together were subjected to hydro-

genolysis. Unexpectedly, the a-methyl ester was also

cleaved during the hydrogenolysis to aord the cyclic

sulfoximine 19 in one step. The ring opening of 19 by

mild alkaline hydrolysis gave the sulfoximine 3 in

quantitative yield. The sulfoximine 3 was prone to

cyclize under weakly acidic conditions such as in acetic

acid or on a silica gel TLC, and therefore the sulfox-

imine 3 was isolated as an ammonium form to prevent

the spontaneous cyclization. The ammonium salt of 3

was stable and could be stored without any extent of

cyclization.

2 and 3 is most likely to re¯ect a similar inhibition

scheme involving the enzyme-catalyzed phosphorylation

of the phosphinyl oxygen or the sulfoximine nitrogen by

ATP. In fact, both 2 and 3 required ATP for enzyme

inactivation: 5⁰-adenylylimidodiphosphate (AMPPNP),

a non-hydrolyzable ATP analogue, failed to cause the

inactivation of the enzyme, while the combination of 2

or 3 and ATP resulted in enzyme inactivation (Fig. 2).

This is also supported by the fact that the sulfone 4,

which is incapable of phosphorylation at the S O

oxygen, served as a simple reversible inhibitor (Fig. 1).

This result is consistent with the reversible inhibition of

rat kidney g-GCS by buthionine sulfone, in which the

extent of inhibition did not increase with time.²¹

For the synthesis of sulfone 4 (Scheme 4), l-homocys-

tine was reduced by dithiothreitol, followed by base-

catalyzed conjugate addition to ethyl 2-ethylacrylate¹⁴

and subsequent hydrolysis in one ¯ask to aord the

sul®de 20 in 76% overall yield after chromatography on

Dowex 50W (H⁺form). Oxidation with hydrogen per-

oxide and crystallization from AcOH gave the sulfone 4.

The diastereomeric ratio was 1:1 as judged by the ¹H

NMR signal of the a-proton.

It is worthy of note, however, that all the inhibitors 2, 3

and 4 were much more potent inhibitors than l-BSO

(Table 1). The overall anity of the phosphinic acid 2

and the sulfoximine 3 was 10- and 500-times greater

than that of l-BSO, respectively, and the inactivation

rate of enzyme by 2 and 3 was 90- and 290-times,

respectively, as fast as that of l-BSO. The two orders

increase in the g-GCS inhibitory activity of the sulfox-

imine 3 compared to that of l-BSO suggested a sig-

ni®cant interaction of the b-carboxyl function in 3 with

the enzyme active site. Although the sulfone 4 did not

inactivate the enzyme, its inhibition potency was more

than ®ve times higher than that of l-BSO as measured

by the inhibition constant. The dierence is still under-

estimated because the initial inhibition constant of the

sulfone 4 (K_i=9.2 mM) is compared with the overall

inhibition constant of l-BSO (K_i^*=49 mM) in which the

Enzyme inhibition

The phosphinic acid 2 and the sulfoximine 3 served as a

potent slow-binding inhibitor of E. coli g-GCS (Fig. 1a,

b). In view of the mechanism-based phosphorylation of

the phosphinic acid 1 by E. coli glutathione synthetase⁸

and l-buthionine sulfoximine (l-BSO) by rat kidney g-

GCS,²¹the slow-binding inhibition of E. coli g-GCS by

Scheme 4. Reagents and conditions: (a) 1,4-Dithio-dl-threitol, NaOH, EtOH-H₂O; (b) Ethyl 2-ethylacrylate; (c) NaOH then Dowex

50WÂ8 (H⁺form), eluted with 0.7 m HCl; (d) H₂O₂, AcOH.

Table 1. Inhibition of g-GCS by Phosphinic acid 2, Sulfoximine 3, Sulfone 4 and l-buthionine-(S,R)-sulfoximine (l-BSO)

b

1

Inhibitor

K_i^a[mM]

k_inact[sec

]

k_inact/K_i^c[M ¹sec

]

K_i^*d[mM]

Phosphinic acid 2

Sulfoximine 3

Sulfone 4

28.96.6

ND^e

0.01260.0014

43633

143090

Ð^f

4.950.27

0.09890.0224

Ð^f

ND^e

Ð^f

ND^e

9.231.76

ND^e

l-BSO

4.871.01

49.37.4

^aInitial inhibition constant.

^bInactivation rate constant.

^cSecond-order rate constant for slow-binding inhibition (see Experimental).

^dOverall inhibition constant measured after ATP-dependent binding equilibrium was established.

^eNot determined (see Experimental).

^fNo slow-binding inhibition was observed.

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

1939

ATP-dependent tight binding equilibrium was estab-

lished. This result again substantiates the importance of

the b-carboxyl group in the recognition of the inhibitors

by the enzyme. The K_iof the phosphinic acid 2 was also

smaller than the K_i^*of l-BSO, suggesting that the

phosphinic acid 2 itself is a better transition-state mimic

than the putative phosphorylated l-BSO, provided K_i

and K_i^*re¯ect the anity of the inhibitors before and

after phosphorylation, respectively.

Of particular interest is the dierence in the inhibition

behavior between the phosphinic acid 2 and the sulfoximine

Figure 2. ATP-dependence of enzyme inactivation by: (a) the

phosphinic acid 2; and (b) the sulfoximine 3. The enzyme

(10 mM) was preincubated with the inhibitor in the presence of

ATP (1 mM; curve A) or AMP-PNP (10 mM; curve B) at 25 ^ꢀC

for 30 min in Tris±HCl buer (pH 7.5, 50 mM) containing

MgCl₂(2 mM). The remaining activity was then measured after

1000-fold dilution. Curve C (control): no ATP and AMP-PNP

was added to the preincubation mixture.

Figure 1. Progress curves for the inhibition of g-GCS by: (a)

the phosphinic acid 2; (b) the sulfoximine 3; and (c) the sulfone

4. The reaction was initiated by adding the enzyme (20±56 nM)

to a standard assay mixture (see Experimental) containing each

inhibitor at the ®nal concentrations indicated.

1940

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

3. First of all, the sulfoximine 3 was a much more potent

inactivator than the phosphinic acid 2: the overall an-

ity of 3 was 50 times greater than that of 2, and the

sulfoximine 3 inactivated the enzyme more than three

times faster than 2 as judged by the second-order rate

constant for enzyme inactivation (k_inact/K_i). Another

interesting facet concerns the regain of enzyme activity.

Duration of enzyme inactivation is an important

measure relevant to the inhibition potency of slow-

binding inhibitors. In this regard, the sulfoximine 3 was

also superior to the phosphinic acid 2. No regain of

enzyme activity was observed with 3 under the standard

assay conditions (Fig. 2b, curve A), but a signi®cant

enzyme reactivation (t_1/2<3 min) was noted with the

phosphinic acid 2 by 1000-fold dilution as evidenced by

a concave upward progress curve (Fig. 2a, curve A).

inhibitor may account for the dierence. Thus, the

putative phosphorylated 2, a phosphinic±phosphoric

anhydride, is potentially unstable and is still capable of

transferring the phosphoryl group either to water

(hydrolysis) or to ADP (reverse reaction) to release the

free phosphinic acid 2. On the other hand, the phos-

phoryl group attached to the sulfoximine nitrogen is

expected to be stable²²and remains unchanged, once it

is transferred from ATP. To test this notion, we syn-

thesized the N-phosphoryl sulfoximine 21, and its inhi-

bition behavior was examined.

Synthesis and inhibition properties of N-phosphoryl

sulfoximine 21

The N-phosphoryl sulfoximine 21 was synthesized by a

similar route to that of the sulfoximine 3 (Scheme 5). A

major dierence was that the benzyl-based protecting

groups were used for the protection of the carboxyl as

well as the phosphoryl groups, because the phosphory-

lated sulfoximine nitrogen would be no longer capable

of intramolecular nucleophilic substitution to yield the

corresponding cyclic sulfoximine as observed with the

cyclization of the sulfoximine 17. In addition, the

simultaneous cleavage of the (NO₂)Z and the a-methyl

ester by hydrogenolysis seemed unique to the cyclic

A question may arise here why the enzyme regained

activity so quickly with the phosphinic acid 2, but not

with the sulfoximine 3, although they are structurally

analogous. Since both 2 and 3 failed to inactivate the

enzyme with AMPPNP (Fig. 2) and the sulfone 4 did

not inactivate the enzyme, the phosphorylation of the

phosphinyl oxygen or the sulfoximine nitrogen should

be prerequisite for enzyme inactivation. If this is the

case, then the chemical stability of the phosphorylated

Scheme 5. Reagents and conditions: (a) 23, CsCO₃(cat.), i-PrOH; (b) NaIO₄, THF±H₂O; (c) MSH, CH₃CN; (d) o-Xylylene N,N-

diethylphosphoramidite, 1H-tetrazole, then t-BuOOH; (e) H₂, 10% Pd on carbon, MeOH-H₂O-AcOH, then Dowex 50WÂ8 (H⁺

form), eluted with H₂O.

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

1941

sulfoximine 18 and in fact was not the case with non-

cyclized sulfoximines.²³Thus, N-(4-nitrobenzyloxy-

carbonyl)-l-homocysteine benzyl ester (22) prepared by

the reduction of the corresponding disul®de was allowed

to react with benzyl 2-ethylacrylate (23) prepared from

ethylmalonic acid monobenzyl ester.¹⁴The resulting

sul®de 24 was oxidized stepwise with NaIO₄and MSH

to aord the sulfoximine 26. In this case, no sponta-

neous cyclization of the sulfoximine was observed. The

imino nitrogen of 26 was phosphorylated by o-xylylene

N,N-diethylphosphoramidite²⁴with 1H-tetrazole as a

catalyst followed by oxidation to give the N-phosphoryl

sulfoximine 27. All benzyl-based protecting groups in 27

were successfully cleaved at once by hydrogenolysis to

aord the ®nal product 21. The N-P bonding was

hydrolytically stable even under an acidic condition, and

the compound 21 was puri®ed by passing a Dowex-50W

column (H⁺form) without any extent of decomposi-

tion. The chemical synthesis of 21 thus con®rmed the

stability of the putative N-phosphorylated sulfoximine 3

which is thought to be formed enzymatically within the

enzyme active site.

tight binding of the phosphorylated inhibitor is always

accompanied by ADP as a binding partner. In addition,

the inhibition of a mechanistically related glutamine

synthetase by l-methionine sulfoximine phosphate was

increased signi®cantly in the presence of ADP and

Mg .

^{2+ 22}We reasoned therefore that the tight binding of

21 would also require ADP and that the slow-binding

kinetics observed with 21 would re¯ect the enzymatic

formation of ADP during the assay. To test this notion,

we examined the eect of ADP on the enzyme inactiva-

tion by the N-phosphoryl sulfoximine 21 (Fig. 4). As

expected, the preincubation of the enzyme and 21 in the

presence of ADP caused facile enzyme inactivation,

while a very slow decrease of enzyme activity was

observed without ADP. These results suggested that the

N-phosphoryl sulfoximine 21 alone did not cause rapid

enzyme inactivation, but the presence of ADP resulted

in the formation of a ternary complex consisting of 21,

ADP and Mg²⁺within the enzyme active site to cause

facile enzyme inactivation. If this is the case, then the

ternary complex thus formed in the enzyme active site

should be identical to the complex formed by the sulf-

oximine 3, ATP and Mg²⁺. We therefore measured the

kinetic parameters for the inhibition by 21 and found

that the overall inhibition constant (K_i^*) of 21 was

160 nM and the second-order inactivation rate (k_inact/K_i)

The N-phosphoryl sulfoximine 21 was found to be a

potent inactivator of g-GCS. Interestingly, the extent of

inhibition increased with time (Fig. 3), showing a typical

slow-binding kinetics as observed for the inhibition by

the sulfoximine 3 (Fig. 1b). Since the inhibitor 21 is

already phosphorylated, the inhibition scheme involving

the enzyme-catalyzed phosphorylation by ATP seems

unlikely. However, in view of the X-ray crystal struc-

tures of glutathione synthetase⁸and d-Ala: d-Ala

ligase²⁵complexed with the respective phosphinate, the

Figure 4. Eect of ADP on the inactivation of g-GCS by the

N-phosphoryl sulfoximine 21. The enzyme (1 mM) was incu-

bated with 21 (2 mM) and MgCl₂(2 mM) in Tris±HCl buer

(pH 7.5, 20 mM) in the presence (curve A) or in the absence

(curve B) of ADP (1 mM). Aliquots (50 mL) were withdrawn at

the indicated intervals and were diluted with 2 mL of Tris±HCl

buer (pH 7.3, 20 mM) containing MgCl₂(5 mM) and 10% (v/

v) glycerol. The mixture was subjected to Centricon-10 to

remove unbound 21 and ADP, and the remaining enzyme

activity was measured by the standard assay method (see

Experimental).

Figure 3. Progress curves for the inhibition of g-GCS by the N-

phosphoryl sulfoximine 21. The reaction was started by adding

the enzyme (56 nM) to the standard assay solution (see Experi-

mental) containing 21 at the ®nal concentrations indicated.

1942

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

was 1300 M ¹s ¹. As expected, the K_i^*value of 21 was

comparable to that of 3 (99 nM), indicating that the

enzyme inactivation was most likely to be caused by the

formation of the ternary complex comprised of the N-

phosphoryl sulfoximine 21, ADP and Mg²⁺. Interest-

ingly, the k_inact/K_ivalue of 21 (1300 M ¹s ¹) was also

comparable to that of 3 (1430 M ¹s ¹). However, we

assume this as coincidence, because there is little reason

to expect that the rate at which the ternary complex is

formed in the enzyme active site by enzymatic phos-

phorylation of 3 with ATP should bear any relationship

to the rate at which the association of the enzymatically

formed ADP and 21 in the enzyme active site.

potency was much higher than that of l-BSO. Hence

the stereoselective synthesis of the sulfoximine 3 was

expected not only to produce a g-GCS inhibitor with

much higher potency, but also to yield a transition-state

analogue serving as a better mechanistic probe for the

enzyme catalysis from a stereochemical point of view. In

light of transition-state analogy, the diastereomers with

S-con®guration at the b-carbon would be preferable,

since the (S)-b-carbon corresponds to the l-con®gura-

tion of the attacking Cys in the transition state. We

therefore synthesized two diastereomeric sulfoximines

3a and 3b with (S)-b-carbon and (R) or (S)-chiral sulfur

atom, respectively.

Scheme 6 summarizes the stereoselective synthesis of the

diastereomeric sulfoximine 3a and 3b. The chiral carbon

center was constructed by diastereoselective aldol con-

densation of formaldehyde and the dibutylboron eno-

late prepared from the chiral N-butanoyloxazolidinone

28.²⁶The desired (R)-alcohol 29 was isolated, and the

Stereoselective synthesis and inhibition properties of

sulfoximine 3

Although the sulfoximine 3 synthesized above was a

mixture of four diastereomers with respect to the chiral

b-carbon and the chiral sulfur atom, its inhibition

Scheme 6. Reagents and conditions: (a) Bu₂BOTf, NEt₃, CH₂Cl₂, then HCHO (gas); (b) CH₃SO₂Cl, NEt₃, CH₂Cl₂; (c) NaI, acetone;

(d) 14, K₂CO₃, DMF; (e) NaIO₄, THF±H₂O; (f) MSH, CH₃CN; (g) K₂CO₃, H₂O₂, THF±H₂O, then 2 M HCl; (h) H₂, 10% Pd on

carbon; (i) 1.2 M KOH, H₂O then Dowex 50WÂ8 (NH₄⁺form), eluted with H₂O.

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

1943

hydroxyl group was converted to the iodide. The iodide

31 was coupled with the thiol 14 to yield the sul®de 32.

Stepwise oxidation with NaIO₄and MSH gave the sulf-

oximine 34. At this stage, another chiral center was

generated on the sulfur atom, but the diastereomers of

33 or 34 were not separated by chromatography. We

therefore cyclized 34 by removing the chiral oxazolidi-

none ®rst followed by treatment with dilute acid. Direct

cyclization by heating 34 in AcOH as for the cyclization

of the sulfoximine 17 was unsuccessful, probably due to

the stability of the imide C(O)-N bond compared with

the ester C(O)-O bond.

for comparison. The overall anity and the second-

order inactivation rate of the (R)-sulfoximine 3a were

1260- and 1390-times, respectively, greater than those of

l-BSO. The (R)-sulfoximine 3a was also highly potent in

terms of duration of enzyme inactivation: the enzyme

inactivated by (R)-3a and ATP was still totally inactive

after gel ®ltration, and a very slow regain was observed

with a half life of recovery (t_1/2) of 11 days. On the other

hand, the inhibition pro®le of the less inhibitory (S)-

sulfoximine 3b was quite similar to that of the sulfone 4

in that both compounds inhibited the enzyme in a

reversible manner with comparable inhibition constants

(K_i=9.23 mM for the sulfone 4 and 12.0 mM for the (S)-

sulfoximine 3b). Considering that the enzyme inactivation

The two diastereomers with respect to the chiral sulfur

atom of 35 were separated successfully by ¯ash column

chromatography to aord the single stereoisomer 35a

and 35b. Each stereoisomer was subjected to hydro-

genolysis followed by alkaline hydrolysis to yield the

sulfoximines 3a and 3b. The chirality of the sulfur atom

was stable, but the chiral carbon b to the sulfur atom

was partially racemized during the alkaline hydrolysis to

give ca. 20% of the epimeric isomer as an unseparable

byproduct.

In order to determine the absolute stereochemistry of 3a

and 3b, compound 36b was crystallized and its absolute

con®guration was determined by X-ray diraction analy-

sis (Fig. 5). On the basis of the known l-con®guration of

the a-amino group and S-con®guration of the b-carbon,

the con®guration at sulfur was determined to be S.

Hence the stereochemistry of the sulfur atom of the

sulfoximine 3a and 3b is R and S, respectively.

The (R)-sulfoximine 3a was found to be an extremely

powerful slow-binding inhibitor, while the (S)-sulfox-

imine 3b served as a simple reversible inhibitor with

much lower inhibition potency (Fig. 6). Preincubation

of the enzyme with (R)-3a and ATP caused almost

complete inactivation of the enzyme (<2% of residual

activity), but the (S)-sulfoximine 3b caused no enzyme

inactivation (100% of residual activity) under the same

conditions. Table 2 summarizes the kinetic parameters

for the enzyme inhibition by (R)-3a, (S)-3b and l-BSO

Figure 6. Progress curves for the inhibition of g-GCS by: (a)

(R)-sulfoximine 3a; and (b) (S)-sulfoximine 3b. The reaction

was started by adding the enzyme (28 nM) to a standard assay

solution (see Experimental) containing each inhibitor at the

®nal concentrations indicated.

Figure 5. Structure of cyclic sulfoximine 36b as determined by

X-ray diraction analysis.

1944

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

Table 2. Inhibition of g-GCS by the (R)-sulfoximine 3a, (S)-sulfoximine 3b and l-buthionine-(S,R)-sulfoximine (l-BSO)

K_i^*[mM]

k_inact/K_i[M ¹sec

]

Type of inhibition

1

Inhibitor

(R)-Sulfoximine 3a

(S)-Sulfoximine 3b

l-BSO

0.03920.0090

12.01.7

49.37.4

6750200

Ð^a

4.871.01

slow-binding

reversible

slow-binding

^aNo slow-binding inhibition was observed.

by the sulfoximine 3 was due to the mechanism-based

phosphorylation of the sulfoximine nitrogen by ATP,

the active site geometry of E. coli g-GCS is most likely

to allow phosphorylation of (R)-3a solely, while the (S)-

sulfoximine 3b was devoid of enzymatic phosphoryla-

tion probably due to the unfavorable orientation of the

sulfoximine nitrogen in the enzyme active site. Grith et

al. reported the separation of the diastereomers of l-

buthionine (SR)-sulfoximine and examined the inhibi-

tory activity of each diastereomer towards rat kidney g-

GCS.²⁷In this case, l-buthionine-(S)-sulfoximine served

as a tight-binding, mechanism-based inactivator of rat

kidney g-GCS, while l-buthionine-(R)-sulfoximine was

a relatively weak inhibitor without causing enzyme

inactivation. Of particular interest is that the relative

con®guration of the chiral sulfur atom of (R)-3a is the

same as that of l-buthionine-(S)-sulfoximine, although

the sequence rule dictates the opposite sign for these two

chiral sulfur atoms. As far as the mechanism-based

phosphorylation of the sulfoximine nitrogen is con-

cerned, the E. coli g-GCS and rat kidney g-GCS require

the same relative con®guration of the chiral sulfoximine

nitrogen to transfer the g-phosphate of ATP in the

enzyme active site.

4, the sulfoximine 3 served as the most potent, ATP-

dependent inactivator of g-GCS. In particular, the sul-

foximine 3a with (S)-b-carbon and the (R)-sulfur atom

was an extremely powerful mechanism-based inacti-

vator with an overall anity and a second-order inacti-

vation rate of more than 1000 times greater than those

of l-BSO. As far as we know, this is the most potent

transition-state analogue inhibitor of g-GCS. In this

study, the inhibitory activity was examined only with

the E. coli enzyme, but g-GCS from other sources

should also be inactivated by the sulfoximine 3 more

eectively than l-BSO, because the structure of 3 is

structurally much closer to the transition state. Another

important facet is the stereochemistry of the chiral sulfur

atom. The E. coli enzyme was rapidly inactivated by the

(R)-sulfoximine 3a, but not by the (S)-sulfoximine 3b.

Since the stereochemical preference depends on the active

site geometry of individual enzyme, the chiral sulfoximine

3a and 3b may serve as a lead for a species-selective drug

for regulating the physiological level of glutathione.

Experimental

General methods

Recent elucidation of the crystal structures of gluta-

thione synthetase²⁸and d-Ala: d-Ala ligase²⁵has

revealed a marked structural similarity between these

two mechanistically related ligases.²⁹A common fold

may also be found for other members of ADP-forming

peptide ligases such as g-GCS. In this context, the (R)-

sulfoximine 3a should ®nd an important use as a ligand

for structural determination of E. coli g-GCS. This

aspect of study is in progress.

THF was dried by distillation from Na/benzophenone

ketyl. CH₂Cl₂, CH₃CN and NEt₃were distilled from

CaH₂. DMF was distilled from CaH₂under reduced

pressure and stored over 4 A molecular sieves. Ana-

lytical TLC was performed on a silica gel plate (Merck

5715, 0.25 mm). Flash column chromatography was

performed on silica gel 60 (Merck 9385, 230±400 mesh).

Medium-pressure column chromatography was per-

formed using an ULTRA PACK²silica gel column

(Yamazen Co.) with a medium pressure of 2±3 kg/cm².

Melting points were recorded on a Mettler FP62 and

were corrected. Elemental analyses were performed on a

Conclusion

1

The results presented in this paper demonstrate that the

rational design of g-GCS inhibitors on the basis of the

proposed transition state is highly eective in producing

potent g-GCS inhibitors. A key was the introduction of

a carboxyl function at the b-carbon to the central hetero

atom so as to mimic the carboxyl group of the attacking

l-Cys in the transition state. Among three types of

inhibitors, phosphinic acid 2, sulfoximine 3 and sulfone

Yanaco MT-5. H NMR and ³¹P NMR were recorded

on a Varian VXR-200 (200 MHz) or JEOL JNM-AL400

(400 MHz) with tetramethylsilane or 3-(trimethylsilyl)-

propanesulfonic acid sodium salt (for D₂O) as internal

standard. Eighty ®ve per cent phosphoric acid was used

as external standard for ³¹P NMR. Infrared spectra

were recorded on a Hitachi U-215 spectrophotometer.

Optical rotations were measured on a Perkin-Elmer 241

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

1945

polarimeter. Mass spectra were obtained on a JEOL

JMS700 spectrometer.

posed during puri®cation and was used as such for the

next reaction.

Ethyl (2R,S)-3-[ethoxy(diethoxymethyl)phosphinyl]-2-ethyl-

propanoate (6). A solution of NaOEt prepared from Na

(2.04 g, 88.9 mmol) in dry EtOH (200 mL) was added

dropwise via cannula to ethyl (diethoxymethyl)phos-

phinate¹²(15.9 g, 80.9 mmol) in dry EtOH (100 mL) at

0 ^ꢀC. The mixture was stirred at 0 ^ꢀC for 10 min, and

ethyl 2-ethylacrylate¹⁴(11.4 g, 88.9 mmol) was added

dropwise to the mixture at 0 ^ꢀC over 20 min. The mix-

ture was stirred at 0 ^ꢀC for 1 h and at room temperature

for 2 h. The reaction was quenched by adding AcOH

(7.2 g, 121 mmol), and the reaction mixture was evapo-

rated. The residue was dissolved in EtOAc (200 mL),

washed with satd NaHCO₃(200 mL) and satd NaCl

(200 mL), successively, and dried over Na₂SO₄. Solvent

was evaporated, and the product was puri®ed by ¯ash

column chromatography (6:1, EtOAc/hexane) to give

the phosphinate 6 (19.7 g, 75%, diastereomer mixture)

as a colorless oil: IR (NaCl, neat) n_max1725, 1200,

Methyl (2S)-2-(N-benzyloxycarbonylamino)-4-[(R,S)-

(methoxy)[(2R,S)-2-(methoxycarbonyl)butyl]phosphinyl]-

butanoate (9). The phosphinate 7 (1.53 g, 7.88 mmol)

and (N-benzyloxycarbonyl)-l-vinylglycine methyl ester

(8)¹⁶(1.16 g, 4.65 mmol) were dissolved in xylene

(15 mL), and Ar gas was passed through the solution for

10 min to purge oxygen. To this solution was added tert-

butyl peroxybezoate (12.8 mg, 0.066 mmol), and the

mixture was heated at 110 ^ꢀC for 4 h under an Ar

atmosphere. During this period, additional tert-butyl

peroxybenzoate (12.8 mg, 0.066 mmol) was added in

every 1 h (total 38.3 mg, 0.20 mmol), and the mixture

was heated at 110 ^ꢀC for 11 h. Another portion of tert-

butyl peroxybenzoate (25.6 mg, 0.132 mmol) was added,

and the mixture was heated at 120 ^ꢀC for 6 h. The

reaction mixture was evaporated, and the residual oil

was puri®ed by ¯ash column chromatography (4:1 to

2:1, benzene/acetone) to give the compound 9 (454 mg,

1040 cm ¹; H NMR (CDCl₃) d 4.68 (d, J=7.8 Hz) and

22%, diastereomer mixture): [a]_d+8.76^ꢀ(c 1.05,

1

25

4.65 (d, J=7.2 Hz) [1H, PCH], 4.3±4.0 (m, 4H, POCH₂

and COOCH₂), 3.9±3.6 [m, 4H, (CH₃CH₂O)₂C], 2.8 (m,

1H, CHCOOCH₂CH₃), 2.3 and 1.9 (m, 2H, CH₂P), 1.7

(m, 2H, CH₂CH₃), 1.4±1.2 (m, 12H, OCH₂CH₃Â4),

0.92 (t, 3H, J=7.4 Hz, CH₂CH₃); ³¹P NMR (CDCl₃) d

+44.4. MS (EI) 324 (M⁺).

CHCl₃); IR (NaCl, neat) 3250 (br), 1720, 1525, 1203,

1

1035 cm

;

¹H NMR (CDCl₃) d 7.36 (s, 5H, arom),

5.7 (m, 1H, NH), 5.12 (s, 2H, PhCH₂), 4.40 (m, 1H,

a-proton), 3.8±3.6 (m, 9H, POCH₃and COOCH₃Â2),

2.7 (m, 1H, CHCOOCH₃), 2.2 and 2.0 (2Âm, 2H,

CH₂P), 1.9±1.7 (m, 4H, CH₂CH₂P), 1.7 (m, 2H,

CH₂CH₃), 0.90 (t, 3H, J=7.4 Hz, CH₂CH₃); ³¹P NMR

d+55.9 (47%), +56.4 (27%), +56.6 (26%). HRMS

(EI) calcd for C₂₀H₃₀NO₈P (M⁺) 443.1710, found

443.1730.

Methyl (2R,S)-2-ethyl-3-(methoxyphosphinyl)propanoate

(7). Compound 6 (4.40 g, 13.6 mmol) was heated in 12

M HCl (50 mL) under re¯ux for 4 h. The reaction mix-

ture was evaporated to dryness to give a light brown

syrup. H₂O (50 mL) was added to the syrup and evapo-

rated again. The residual syrup was dried in a desiccator

over powdered CaCl₂and NaOH to remove trace HCl

to give 2-carboxybutylphosphorous acid (2.42 g) as a

(2S)-2-Amino-4-[(2R,S)-(2-carboxybutyl)(hydroxy)phos-

phinyl]butanoic acid (2). Compound

9

(381 mg,

0.86 mmol) was dissolved in THF (2 mL), and 3 M NaOH

(1.72 mL, 5.15 mmol) was added to the solution. The

mixture was stirred at 25 ^ꢀC for 7 h. The reaction mix-

ture was made acidic (pHꢁ1) by adding 3 M HCl

(1.8 mL), and the mixture was evaporated to dryness to

give a colorless foam. The residue was extracted with

acetone (15 mL), and insoluble NaCl was removed by

®ltration through Celite. The ®ltrate was evaporated to

give the N-protected free acid as a colorless foam

pale brown syrup: ¹H NMR (D₂O)

d 7.10 (d,

¹J_HP=549 Hz, 1H, PH), 2.7 (m, 1H, CHCOOH), 2.3±

1.8 (m, 2H, CH₂P), 1.7 (m, 2H, CH₂CH₃), 0.92 (t, 3H,

J=7.4 Hz, CH₂CH₃). MS (EI) 166 (M⁺).

The crude phosphorous acid (2.42 g) was treated with an

ethereal solution of CH₂N₂until yellow color persisted.

Excess CH₂N₂was decomposed by adding AcOH, and

the reaction mixture was evaporated below 40 ^ꢀC. The

product was unstable when heated above 50 ^ꢀC and was

quickly puri®ed by ¯ash column chromatography (2:3,

acetone/hexane) to give the phosphinate 7 (1.54 g, 59%,

diastereomer mixture) as a colorless oil: ¹H NMR

1

(346 mg): H NMR (acetone-d₆) d 9.50 (br s, 3H, POH

and COOHÂ2), 7.3 (m, 5H, arom), 6.75 (br s, 1H, NH),

5.10 (s, 2H, PhCH₂), 4.3 (m, 1H, a-proton), 2.8 (m, 1H,

CHCOOH), 2.3±1.8 (m, 6H, CH₂CH₂PCH₂), 1.7 (m,

2H, CH₂CH₃), 0.93 (t, 3H, J=7.4 Hz, CH₂CH₃); ³¹P

NMR (acetone-d₆) d +58.95.

1

(CDCl₃) d 7.15 (d, J_HP=552 Hz, 1H, PH), 3.79 (d,

³J_HP=11.8 Hz, 3H, POCH₃), 3.73 (2Âs, 3H, COOCH₃),

2.8 (m, 1H, CHCOOCH₃), 2.4±1.8 (m, 2H, CH₂P), 1.7

(m, 2H, CH₂CH₃), 0.92 (2Ât, 3H, J=7.4 Hz, CH₂CH₃).

HRMS (EI) calcd for C₇H₁₅O₄P (M⁺) 194.0709,

found 194.0708. The product 7 was partially decom-

The product (320 mg) was dissolved in MeOH±AcOH±

H₂O (8:2:1, 11 mL). Hydrogen gas was passed through

the solution in the presence of 10% Pd on carbon

(150 mg) at room temperature. The hydrogenolysis was

monitored with silica gel TLC (n-BuOH±AcOH±H₂O,

1946

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

5:2:2) and was completed in 20 min. The reaction mix-

ture was ®ltered through Celite, and the ®ltrate was

evaporated to give crude 2 as a colorless syrup. The

crude 2 was applied to a Dowex 50WX8 column (H⁺

form, 35 cm³), and the column was eluted with H₂O. A

strongly acidic impurity passed through ®rst, and the

eluate became neutral (75 mL of H₂O). The eluent was

changed to 6% pyridine, and the column was eluted

until the solvent front reached 4 cm above the bottom of

the column (ca. 50 mL of 6% pyridine). The eluent was

changed again to H₂O, and the product 2 was eluted as

a mildly acidic eluate (pH 3±4). Ninhydrin positive

fractions were collected (ca. 70 mL) and lyophilized to

give the phosphinic acid 2 (184 mg, 80%, diastereomer

(c 1.17, CHCl₃); mp 104.5±105.2 ^ꢀC; IR (KBr) n_max

3340, 2960, 1740, 1695, 1520, 1340, 1215 cm

1

;

¹H

NMR (CDCl₃) d 8.21 (d, 4H, J=8.6 Hz, arom), 7.50 (d,

4H, J=8.8 Hz, arom), 5.60 (d, 2H, J=8.6 Hz, NHÂ2), 5.21

(s, 4H, ArCH₂Â2), 4.5 (m, 2H, a-protonÂ2), 3.78 (s, 6H,

COOCH₃Â2), 2.73 (t, 4H, J=7.4 Hz, CH₂CH₂SÂ2), 2.2

(m, 4H, CH₂CH₂SÂ2). Anal. calcd for C₂₆H₃₀N₄O₁₂S₂:

C, 47.70; H, 4.62; N, 8.56. Found: C, 47.80; H, 4.62; N,

8.57.

(N-4-Nitrobenzyloxycarbonyl)-L-homocysteine methyl ester

(14). The homocystine derivative 13 (2.30 g, 3.51 mmol)

was dissolved in CH₂Cl₂(30 mL) which had been

washed and saturated with H₂O. To this solution was

added tri-n-butylphosphine (0.71 g, 3.51 mmol) at once

at room temperature. The mixture was stirred for 20 min

at room temperature and was concentrated to dryness.

The residue was puri®ed by ¯ash column chromatog-

raphy (3:2, hexane/EtOAc) to give the thiol 14 (2.36 g,

100 %) as a yellow oil: ¹H NMR (CDCl₃) d 8.23 (d, 2H,

J=8.8 Hz, arom), 7.52 (d, 2H, J=8.8 Hz, arom), 5.45

(d, 1H, J=7.6 Hz, NH), 5.22 (s, 2H, ArCH₂), 4.5 (m,

1H, a-proton), 3.78 (s, 3H, COOCH₃), 2.6 (m, 2H,

CH₂SH), 2.0 (m, 2H, CH₂CH₂S), 1.56 (t, 1H, J=8.3 Hz,

SH).

25

mixture) as a colorless non-hygroscopic solid: [a]_d

+11.8^ꢀ(c 0.53, H₂O); IR (KBr) n_max3400 (br), 2950 (br),

1

1710, 1638, 1510, 1220, 1100, 1008 cm

;

¹H NMR

(400 MHz, D₂O) d 3.99 (2Ât, 1H, J=6.05 Hz, a-proton),

2.7 (m, 1H, CHCOOH), 2.1 (m, 2H, CH₂CH₂P), 2.0 and

1.7 (m, 2H, PCH₂CH), 1.8 (m, 2H, CH₂CH₂P), 1.6 (m,

2H, CH₂CH₃), 0.91 (t, 3H, J=7.33 Hz, CH₂CH₃);

³¹P NMR (D₂O) d +45.6. HRMS (FAB, glycerol)

calcd for C₉H₁₉NO₆P (MH⁺) 268.0951, found

268.0972.

Methyl 2-ethyl-3-hydroxypropanoate (11). This com-

pound was prepared by the alkylation¹⁷of methyl 3-

hydroxypropanoate (10) in the presence of two equiv of

N,N⁰-dimethylpropyleneurea (DMPU) and was puri®ed

by ¯ash column chromatography (1:2, EtOAc/hexane)

to give a colorless liquid (56%): IR (NaCl, neat) n_max

Methyl (2S)-4-[(2R,S)-2-(methoxycarbonyl)butylthio]-2-

(N-4-nitrobenzyloxycarbonylamino)butanoate (15). The

thiol 14 (8.17 g, 24.9 mmol) and the iodide 12 (7.23 g,

29.9 mmol) were dissolved in dry DMF (70 mL). To the

solution was added powdered K₂CO₃(3.78 g,

27.4 mmol), and the mixture was stirred at 60 ^ꢀC for 1 h.

The reaction mixture was poured into H₂O (600 mL)

and extracted with EtOAc (100 mLÂ3). A dilute solu-

tion of I₂in EtOAc was added to the combined EtOAc

extracts until brown color persisted to oxidize the resi-

dual thiol. The mixture was then washed with 5%

Na₂S₂O₃, satd NaCl, successively, and dried over

Na₂SO₄. Evaporation gave a syrup. The product was

puri®ed by ¯ash column chromatography (1:2, EtOAc/

hexane) to give the sul®de 15 (8.50 g, 77%, diastereomer

mixture) as a pale yellow syrup: IR (NaCl, neat) n_max

1

3350 (br), 2940, 1710, 1430, 1370, 1180, 1040 cm ¹; H

NMR (CDCl₃) d 3.8 (m, 2H, OCH₂), 3.73 (s, 3H,

COOCH₃), 2.5 (m, 1H, CH), 2.08 (br s, 1H, OH), 1.6

(m, 2H, CH₂CH₃), 0.95 (t, 3H, J=7.5 Hz, CH₂CH₃).

HRMS (FAB, glycerol) calcd for C₆H₁₃O₃(MH⁺)

133.0865, found 133.0874.

Methyl 2-(iodomethyl)butanoate (12). The alcohol 11

was converted to the iodide 12 via the mesylate by the

general method (NaI, re¯ux in acetone). Puri®cation by

¯ash column chromatography (1:9, Et₂O/hexane) gave

12 as a colorless liquid (70%): IR (NaCl, neat) n_max

2950, 1735, 1200 cm ¹; ¹H NMR (CDCl₃) d 3.74 (s, 3H,

COOCH₃), 3.36 (dd, 1H, J=8.0 and 9.8 Hz, CH₂I), 3.26

(dd, 1H, J=5.6 and 9.8 Hz, CH₂I), 2.7 (m, 1H, CH), 1.7

(m, 2H, CH₂CH₃), 0.93 (t, 3H, J=7.5 Hz, CH₂CH₃).

HRMS (EI) calcd for C₆H₁₁IO₂(M⁺) 241.9806, found

241.9807.

3300, 2930, 1720, 1600, 1510, 1430, 1340, 1210, 1045,

1

740 cm

;

¹H NMR (400 MHz, CDCl₃) d 8.23 (d, 2H,

J=8.8 Hz, arom), 7.52 (d, 2H, J=8.8 Hz, arom), 5.53

(d, 1H, J=7.8 Hz, NH), 5.21 (2Âd, 2H, J=13.9 Hz,

ArCH₂), 4.48 (dt, 1H, J=5.1 and 7.8 Hz, a-proton),

3.77 (s, 3H, NHCHCO₂CH₃), 3.71 (s, 3H, CO₂CH₃),

2.77 (dd, J=2.7 and 13.2 Hz) and 2.75 (dd, J=2.9 and

13.2 Hz) [1H, SCH₂CH], 2.64 (dd, J=3.9 and 13.2 Hz)

and 2.63 (dd, J=3.7 and 13.2 Hz) [1H, SCH₂CH], 2.6±

2.5 (m, 3H, CH₂CH₂S and SCH₂CH), 2.2 and 2.0 (2Âm,

2H, CH₂CH₂S), 1.7 (m, 2H, CH₂CH₃), 0.91 (t, 3H,

J=7.6 Hz, CH₂CH₃). Anal. calcd for C₁₉H₂₆N₂O₈S: C,

51.57; H, 5.92; N, 6.33. Found: C, 51.43; H, 5.90; N,

6.34.

(N-4-Nitrobenzyloxycarbonyl)-L-homocystine methyl ester

(13). (N-4-Nitrobenzyloxycarbonyl)-l-homocystine was

treated with methanol containing 1% (v/v) H₂SO₄

(room temperature, 12 h). Puri®cation by ¯ash column

chromatography (1:1, EtOAc/hexane) aorded the

methyl ester 13 (65%) as white powder: [a]_d+12.82^ꢀ

27

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

1947

Methyl (2S)-4-[(2R,S)-2-(methoxycarbonyl)butyl-(R,S)-

sul®nyl]-2-(N-4-nitrobenzyloxycarbonylamino)butanoate

(16). A solution of NaIO₄(4.87 g, 22.7 mmol) in H₂O

(50 mL) was added to a solution of the sul®de 15 (8.40 g,

18.9 mmol) in THF±H₂O (7:2, 90 mL) at 0 ^ꢀC. The mix-

ture was stirred at 25 ^ꢀC for 24 h and evaporated to

remove THF, and the resulting mixture was extracted

with EtOAc (100 mLÂ2). The combined extracts were

washed twice with H₂O and satd NaCl, successively,

and dried over Na₂SO₄and concentrated. This residual

syrup was puri®ed by ¯ash column chromatography

(2:3, acetone/hexane) to give the sulfoxide 16 (7.93 g,

91%, diastereomer mixture) as a pale yellow syrup: IR

Diastereomers of methyl (2S)-4-[(1RS,4RS)-4-ethyl-4,5-

dihydro-1,3-dioxo-1,2-thiazol-1-yl]-2-(N-4-nitrobenzyloxy-

carbonylamino)butanoate (18a and 18b). A solution of

the sulfoximine 17 (1.10 g, 2.32 mmol) in AcOH (30 mL)

was stirred at 50 ^ꢀC for 5 h and was concentrated. The

residue was puri®ed by medium-pressure column chro-

matography (6% i-PrOH in CHCl₃) to give the diastereo-

meric 18a (382 mg, 37%) and 18b (393 mg, 38%) as a

colorless foam and solid, respectively. Each compound

was a mixture of two diastereomers. Compound 18a: IR

(KBr) n_max3280, 2950, 1700 (br), 1600, 1510, 1440,

1

1340, 1220 (br), 1055, 1000, 740 cm

;

¹H NMR

(400 MHz, CDCl₃) d 8.23 (d, 2H, J=8.4 Hz, arom), 7.52

(d, 2H, J=8.4 Hz, arom), 5.96 and 5.90 (2Âbr s, 1H,

NH), 5.2 (m, 2H, ArCH₂), 4.5 (m, 1H, a-proton), 3.81

and 3.80 (2Âs, 3H, CO₂CH₃), 3.67 and 3.43 (2Âm, 2H,

SCH₂CH), 3.6 (m, 2H, CH₂CH₂S), 3.2 (m, 1H,

SCH₂CH), 2.6 and 2.3 (2Âm, 2H, CH₂CH₂S), 2.1 and

1.6 (2Âm, 2H, CH₂CH₃), 1.00 and 0.98 (2Ât, 3H,

J=7.2 Hz, CH₂CH₃). Anal. calcd for C₁₈H₂₃N₃O₈S: C,

48.97; H, 5.25; N, 9.52. Found: C, 48.79; H, 5.25; N,

9.31. Compound 18b: IR (KBr) n_max3300, 2930, 1725,

(NaCl, neat) n_max3220, 2950, 1720, 1600, 1520, 1430,

1

1340, 1210, 1010, 740 cm

;

¹H NMR (400 MHz,

CDCl₃) d 8.23 (d, 2H, J=8.6 Hz, arom), 7.52 (d, 2H,

J=8.6 Hz, arom), 6.0 (m) and 5.9 (d, J=7.1 Hz) [1H,

NH], 5.21 (2Âd, 2H, J=13.6 Hz, ArCH₂), 4.5 (m, 1H,

a-proton), 3.79 (s, 3H, NHCHCO₂CH₃), 3.74±3.72

(4Âs, 3H, CO₂CH₃), 3.1 and 2.7 (2Âm, 2H, SCH₂CH),

3.0 and 2.9 (2Âm, 1H, SCH₂CH), 2.82 (t, 2H,

J=7.6 Hz, CH₂CH₂S), 2.4 and 2.2 (2Âm, 2H,

CH₂CH₂S), 1.8 (m, 2H, CH₂CH₃), 0.95 (4Ât, 3H,

J=7.4 Hz, CH₂CH₃). Anal. calcd for C₁₉H₂₆N₂O₉S: C,

49.78; H, 5.72; N, 6.11. Found: C, 50.12; H, 5.70; N,

6.06. HRMS (FAB, glycerol) calcd for C₁₉H₂₇N₂O₉S

(MH⁺) 459.1437, found 459.1428.

1665, 1600, 1510, 1430, 1340, 1220, 1060, 1000,

1

740 cm

;

¹H NMR (400 MHz, CDCl₃) d 8.23 (d, 2H,

J=8.7 Hz, arom), 7.52 (d, 2H, J=8.7 Hz, arom), 5.63

and 5.98 (2Âbr s, 1H, NH), 5.2 (m, 2H, ArCH₂), 4.5 (m,

1H, a-proton), 3.9 (m, 1H, SCH₂CH), 3.81 and 3.80

(2Âs, 3H, CO₂CH₃), 3.6 (m, 2H, CH₂CH₂S), 3.2 (m,

1H, SCH₂CH), 3.0 (m, 1H, SCH₂CH), 2.5 and 2.3

(2Âm, 2H, CH₂CH₂S), 2.0 and 1.8 (2Âm, 2H,

CH₂CH₃), 1.00 (t, 3H, J=7.5 Hz, CH₂CH₃). Anal. calcd

for C₁₈H₂₃N₃O₈S: C, 48.97; H, 5.25; N, 9.52. Found: C,

48.79; H, 5.19; N, 9.49.

Methyl (2S)-4-[(2R,S)-2-(methoxycarbonyl)butyl-(R,S)-

sulfonimidoyl]-2-(N-4-nitrobenzyloxycarbonylamino)butano-

ate (17). To a solution of the sulfoxide 16 (7.35 g,

16.0 mmol) in dry CH₃CN was added O-mesitylene-

sulfonylhydroxylamine (MSH; CAUTION!: explosive

especially when stored in a glass vial) (3.45 g, 16.0 mmol)

at once under an Ar atmosphere, and the mixture was

stirred at room temperature for 24 h. An additional

portion of MSH (1.03 g, 4.78 mmol) was added, and the

solution was stirred for 72 h to complete the reaction.

The reaction mixture was evaporated, and the residual

colorless foam was dissolved in EtOAc. The solution

was washed with satd NaHCO₃, H₂O, and satd NaCl,

successively, and was dried over Na₂SO₄and con-

centrated. The residual syrup was puri®ed by medium-

pressure column chromatography (3±5% MeOH in

CHCl₃) to give the sulfoximine 17 (4.95 g, 65%, dia-

stereomer mixture) as a colorless foam: IR (NaCl, neat)

n_max3270, 2950, 1720, 1600, 1510, 1430, 1340, 1200,

1050, 990, 730 cm ¹; ¹H NMR d 8.22 (d, 2H, J=8.6 Hz,

arom), 7.52 (d, 2H, J=8.6 Hz, arom), 6.3±6.0 (m, 1H,

CHNH), 5.21 (s, 2H, ArCH₂), 4.46 (m, 1H, a-proton),

3.78 (s, 3H, NHCHCO₂CH₃), 3.74 (s, 3H, CO₂CH₃), 3.7

(m, 1H, SCH₂CH), 3.2±2.8 (m, 5H, CH₂SCHHCH and

S=NH), 2.5 and 2.3 (2Âm, 2H, CH₂CH₂S), 1.7 (m, 2H,

CH₂CH₃), 0.95 (t, 3H, J=7.5 Hz, CH₂CH₃). HRMS

(FAB, glycerol) calcd for C₁₉H₂₈N₃O₉S (MH⁺)

474.1546, found 474.1551.

(2S)-2-Amino-4-[(1RS),(4RS)-4-ethyl-4,5-dihydro-1,3-di-

oxo-1,2-thiazol-1-yl]butanoic acid (19). The combined

sulfoximine 18a and 18b (1:1, 709 mg, 1.61 mmol) were

dissolved in AcOH (15 mL). To the solution was added

10% Pd on carbon (Aldrich, wet, Degussa type E101

NE/W), and H₂gas was passed through the mixture at

30 ^ꢀC for 2.5 h. The reaction mixture was ®ltered

through Celite, and the ®ltrate was concentrated to

dryness. The residue was applied to a Dowex 50WÂ8

column (H⁺form, 25 cm³). The column was washed

with H₂O (250 mL) and 5% pyridine (200 mL), succes-

sively, and was eluted with 3% NH₄OH. The ninhydrin

positive fractions were collected and lyophilized to give

the cyclic sulfoximine 19 (349 mg, 88%, diastereomer

mixture) as a colorless powder: IR (KBr) n_max3350 (br),

1

2930 (br), 1665, 1650±1600, 1400, 1210, 1000 cm ¹; H

NMR (400 MHz, D₂O) d 4.4, 4.0 and 3.6 (3Âm, 2H,

SCH₂CH), 3.9±3.8 (m, 3H, a-proton and CH₂CH₂S),

3.4 and 3.2 (2Âm, 1H, SCH₂CH), 2.4 (m, 2H,

CH₂CH₂S), 1.8 (m, 2H, CH₂CH₃), 0.93 (t, 3H,

J=7.5 Hz, CH₂CH₃). HRMS (FAB, glycerol) calcd for

C₉H₁₇N₂O₄S (MH⁺) 249.0909, found 249.0915.

1948

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

(2S)-2-Amino-4-[(2R,S)-2-carboxybutyl-(R,S)-sulfonimido-

yl]butanoic acid (3). To solution of the cyclic

(2S)-2-Amino-4-[(2R,S)-(2-carboxybutyl)sulfonyl]butanoic

acid (4). The sul®de 20 (830 mg, 3.05 mmol) was oxi-

dized with 30% H₂O₂(1.07 mL, 13.8 mmol) in AcOH

(6.5 mL) at 50 ^ꢀC for 7 h. After the oxidation was com-

pleted (TLC, n-BuOH:AcOH:H₂O=5:2:2, ninhydrin),

the reaction mixture was evaporated to give a colorless

syrup. AcOH (5 mL) was added to the syrup, and the

solution was allowed to stand at room temperature for

30 min and then at 4 ^ꢀC overnight to give the sulfone 4

(144 mg, 23%) as colorless crystals complexed with ca. 1

equiv of AcOH: mp 210.1±211.4 ^ꢀC (decomp.); IR (KBr)

n_max3600±3200 (br), 3100±2700 (br), 1710, 1620, 1460,

a

sulfoximine 19 (326 mg, 1.31 mmol) in H₂O (20 mL) was

added dropwise 1 m KOH (3.28 mL, 3.28 mmol)

at 0 ^ꢀC. The mixture was stirred for 5 days at 4 ^ꢀC and

+

was applied to a Dowex 50WÂ8 column (NH₄form,

25 cm³). The column was eluted with H₂O, and the

ninhydrin positive fractions were collected and lyophi-

lized to give the ammonium salt of the sulfoximine

3 (370 mg, 100%) as a hygroscopic solid: IR (KBr) n_max

1

3600±2400 (br), 1580 (br), 1390, 1190, 1000 cm

¹H NMR (400 MHz, D₂O)

;

d

3.9±3.8 (4Ât, 1H,

1

J=6.4 Hz, a-proton), 3.6 (m, 1H, SCH₂CH), 3.4±3.2

(m, 3H, CH₂CH₂SCHH), 2.7 (m, 1H, SCH₂CH), 2.4±

2.2 (m, 2H, CH₂CH₂S), 1.6 (m, 2H, CH₂CH₃), 0.89 (t,

3H, J=7.6 Hz, CH₂CH₃). HRMS (FAB, glycerol)

Calcd for C₉H₁₉N₂O₅S (MH⁺) 267.1016, found

267.1019.

1285, 1223, 1105 cm ¹; H NMR (D₂O+CF₃COOD) d

4.28 (t, 1H, J=6.2 Hz, a-proton), 3.73 (dd, J=10.3 and

14.7 Hz) and 3.45 (dd, J=3.3 and 14.7 Hz) [2H,

SO₂CH₂CH], 3.5 (m, 2H, CH₂CH₂SO₂), 3.0 (m, 1H,

CHCOOH), 2.5 (m, 2H, CH₂CH₂SO₂), 2.08 (s, 3H,

CH₃COOH), 1.74 (quintet, 2H, J=7.3 Hz, CHCH₂CH₃),

0.96 (t, 3H, J=7.5 Hz, CH₂CH₃). Anal. calcd for

.

(2S)-2-Amino-4-[(2R,S)-2-carboxybutylthio]butanoic acid

hydrochloride (20). l-Homocystine (3.50 g, 13.0 mmol)

was dissolved in degassed EtOH±H₂O (1:1, 50 mL) con-

taining NaOH (11.4 g, 27.4 mmol). 1,4-Dithio-dl-thre-

itol (DTT; 2.41 g, 15.7 mmol) was added to the solution

at once, and the mixture was stirred at room tempera-

ture for 1 h under an Ar atmosphere. Ethyl 2-ethylacry-

late¹⁴(4.01 g, 31.3 mmol) was added to the solution with

vigorous stirring, and a small amount of NaOH (ca.

50 mg) was added as a catalyst. The mixture was stirred

under an Ar protection for 2 h at room temperature.

The reaction mixture was evaporated, and the residue

was dissolved in 2 m NaOH (60 mL). The solution was

washed with ether (50 mLÂ3), and the aqueous solution

was allowed to stand at room temperature for 18 h.

After the hydrolysis was completed (TLC; n-BuOH:

AcOH:H₂O=5:2:2, ninhydrin), the mixture was made

acidic (pH=2) by adding 6 m HCl. The mixture was

evaporated to dryness, and the residual semisolid was

triturated with EtOH (200 mL). Insoluble salt (NaCl)

was ®ltered o, and the ®ltrate was evaporated to give

crude 20 as a pale yellow semisolid. The crude product

was applied to a Dowex 50WÂ8 column (H⁺form,

100 cm³), and the column was washed with H₂O. The

column was then eluted with 0.7 M HCl, and ninhydrin

positive fractions were collected and evaporated to dry-

ness to give the sul®de 20 (5.39 g, 76%) as a colorless

C₉H₁₇NO₆S CH₃COOH 0.3H₂O: C, 39.71; H, 6.54; N,

4.21. Found: C, 39.73; H, 6.43; N, 4.44. HRMS (FAB,

glycerol) calcd for C₉H₁₈NO₆S (MH⁺) 268.0855, found

268.0864.

(N-4-Nitrobenzyloxycarbonyl)-L-homocysteine benzyl ester

(22). (N-4-Nitrobenzyloxycarbonyl)-l-homocystine benzyl

ester was reduced by tri-n-butylphosphine as for the

synthesis of the thiol 14. Puri®cation with ¯ash column

chromatography (1:2, EtOAc/hexane) gave the thiol 22

1

(2.10 g, 91%) as a pale yellow oil: H NMR (CDCl₃) d

8.21 (d, 2H, J=8.6 Hz, arom), 7.50 (d, 2H, J=8.8 Hz,

arom), 7.36 (s, 5H, Ph), 5.4 (m, 1H, NH), 5.2 (m, 4H,

ArCH₂Â2), 4.5 (m, 1H, a-proton), 2.6 (m, 2H, CH₂CH₂S),

2.1 (m, 2H, CH₂CH₂S), 1.52 (t, 3H, J=8.2 Hz, SH).

Benzyl 2-ethylacrylate (23). Compound 23 was prepared

from benzyl hydrogen ethylmalonate according to the

reported procedure.¹⁴The crude product was puri®ed

by ¯ash column chromatography (5:95, Et₂O/hexane) to

aord the acrylate 23 (81%) as a colorless oil: IR (NaCl,

neat) n_max2970, 1715, 1150 cm ¹; H NMR (400 MHz,

1

CDCl₃) d 7.3 (m, 5H, Ph), 6.2 (m, 1H, C CH₂), 5.6 (m,

1H, C CH₂), 5.20 (s, 2H, PhCH₂), 2.4 (m, 2H,

CH₂CH₃), 1.09 (t, 3H, J=7.4 Hz, CH₂CH₃). Anal. calcd

for C₁₂H₁₄O₂: C, 75.76; H, 7.42. Found: C, 75.45; H,

7.57. HRMS (EI) calcd for C₁₂H₁₄O₂(M⁺) 190.0994,

found 190.1010.

solid: mp 172.7±173.9 ^ꢀC; [a]_d+11.3^ꢀ(c 1.01, H₂O);

24

IR (KBr) n_max3200±2300 (br), 1720, 1695, 1604, 1490,

1

1215 cm ¹; H NMR (D₂O) d 4.17 and 4.16 (2Ât, 1H,

Benzyl (2S)-4-[(2R,S)-2-(benzyloxycarbonyl)butylthio]-2-

(N-4-nitrobenzyloxycarbonylamino)butanoate (24). The

thiol 22 (4.0 g, 9.89 mmol), benzyl 2-ethylacrylate (23)

(2.63 g, 13.5 mmol) and CsCO₃(48.3 mg, 0.148 mmol)

were mixed in degassed i-PrOH (30 mL) under an Ar

atmosphere. The resulting mixture was stirred for 1.5 h

at room temperature under an Ar atmosphere. The

reaction was quenched by adding AcOH (0.32 g,

J=6.4 Hz, a-proton), 2.8 (m, 4H, CH₂SCH₂), 2.7 (m,

1H, CHCOOH), 2.3 (m, 2H, CH₂CH₂S), 1.65 (quintet,

2H, J=7.4 Hz, CH₂CH₃), 0.92 (t, 3H, J=7.4 Hz,

CH₂CH₃). Anal. calcd for C₉H₁₉Cl₂NO₄S: C, 35.07; H,

6.21; N, 4.54. Found: C, 35.24; H, 5.90; N, 4.64. HRMS

(EI) calcd for C₉H₁₇NO₄S (M⁺) 235.0879, found

235.0909.

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

1949

5.3 mmol), and the solvent was removed by evaporation.

The residue was dissolved in EtOAc (150 mL), washed

with H₂O (150 mLÂ2) and sat. NaCl (100 mL), succes-

sively, dried over Na₂SO₄, and the solvent was evapo-

rated to dryness. The crude product was puri®ed by

¯ash column chromatography (7:3 to 2:1, EtOAc/hex-

Benzyl (2S)-2-(N-4-nitrobenzyloxycarbonylamino)-4-

[(2R,S)-2-(benzyloxycarbonyl)butyl-(R,S)-N-(o-xylylene-

phosphoryl)-sulfonimidoyl]butanoate (27). The sulfox-

imine 26 was dried by co-evaporation with dry CH₃CN.

To a solution of the sulfoximine 26 (2.73 g, 4.36 mmol)

and o-xylylene N,N-diethylphosphoramidite (1.04 g,

4.35 mmol) in dry CH₂Cl₂(50 mL) was added 1H-tetra-

zole (351 mg, 5.01 mmol). The mixture was stirred for

3 h at room temperature under an Ar atmosphere. tert-

Butyl hydroperoxide (70% aqueous solution, 600 mL,

4.38 mmol) was then added to the solution. The result-

ing solution was stirred for 20 min at room temperature.

The reaction mixture was washed with satd NaHCO₃

(100 mL), H₂O (100 mLÂ2) and satd NaCl (100 mL),

successively, dried over Na₂SO₄, and the solvent was

evaporated to dryness. The crude product was puri®ed

by ¯ash column chromatography (1:1, acetone/hexane)

to give the N-phosphoryl sulfoximine 27 (2.88 g, 81.7%,

ane) to give the sul®de 24 (5.60 g, 95%, diastereomer

1.24^ꢀ(c 1.13, CHCl₃);

27

mixture) as a yellow oil: [a]_d

IR (NaCl, neat) n_max3320, 3030, 2950, 1720, 1510, 1340,

1

1215 cm ¹; H NMR (400 MHz, CDCl₃) d 8.18 (d, 2H,

J=8.3 Hz, arom), 7.47 (d, 2H, J=8.3 Hz, arom), 7.3 (m,

10H, PhÂ2), 5.59 (d, 1H, J=8.3 Hz, NH), 5.23±5.10 (m,

6H, ArCH₂Â3), 4.5 (m, 1H, a-proton), 2.74 (dd, J=8.7

and 12.8 Hz) and 2.73 (dd, J=8.9 and 13.1 Hz) [1H,

SCH₂CH], 2.6±2.5 (m, 4H, CH₂SCHHCH), 2.1 and 2.0

(2Âm, 2H, CH₂CH₂S), 1.7 (m, 2H, CH₂CH₃), 0.87 (t, 3H,

J=7.6 Hz, CH₂CH₃). Anal. Calcd for C₃₁H₃₄N₂O₈S: C,

62.61; H, 5.76; N, 4.71. Found: C, 62.61; H, 5.79; N,

4.74.

26

diastereomer mixture) as a pale yellow syrup: [a]_d

7.87^ꢀ(c 0.94, CHCl₃); IR (NaCl, neat) n_max3250,

3040, 2950, 1720, 1520, 1340, 1240, 1050, 1020 cm ¹; ¹H

NMR (400 MHz, CDCl₃) d 8.1 (m, 2H, arom), 7.5 (m,

2H, arom), 7.3 (m, 12H, Ph), 7.2 (m, 2H, Ph), 6.5 (m,

1H, NH), 5.2±5.0 [m, 10H, ArCH₂Â3 and Ph(CH₂)₂],

4.4 (m, 1H, a-proton), 3.8 (m, 1H, SCH₂CH), 3.4 (m,

2H, CH₂CH₂S), 3.2±3.1 (m, 2H, SCHHCH), 2.5 and 2.4

(2Âm, 2H, CH₂CH₂S), 1.7 (m, 2H, CH₂CH₃), 0.9 (m,

3H, CH₂CH₃); ³¹P NMR (CDCl₃) d 0.045 (53%),

0.428 (47%). Anal. calcd for C₃₉H₄₂N₃O₁₂PS: C,

57.99; H, 5.24; N, 5.20. Found: C, 57.82; H, 5.31; N,

5.22. HRMS (EI) calcd for C₃₉H₄₂N₃O₁₂PS (M⁺)

807.2227, found 807.2234.

Benzyl (2S)-4-[(2R,S)-2-(benzyloxycarbonyl)butyl-(R,S)-

sul®nyl]-2-(N-4-nitrobenzyloxycarbonylamino)butanoate

(25). Compound 25 was prepared by the same proce-

dure as for the sulfoxide 16. Puri®cation by ¯ash col-

umn chromatography (3:2 to 3:1, EtOAc/hexane) gave

27

25 (85%, diastereomer mixture) as a yellow oil: [a]_d

4.24^ꢀ(c 0.99, CHCl₃); IR n_max(NaCl, neat) 3250,

1

3040, 2940, 1720, 1520, 1340, 1215, 1020 cm

;

¹H

NMR (400 MHz, CDCl₃) d 8.2 (m, 2H, arom), 7.5 (m,

2H, arom), 7.3 (m, 10H, PhÂ2), 6.1±5.9 (m, 1H, NH),

5.2±5.1 (m, 6H, ArCH₂Â3), 4.5 (m, 1H, a-proton), 3.1±2.8

(m, 2H, SCHHCH), 2.7±2.6 (m, 3H, CH₂SCHHCH),

2.4 and 2.2 (2Âm, 2H, CH₂CH₂S), 1.8 (m, 2H,

CH₂CH₃), 0.9 (m, 3H, CH₂CH₃). Anal. Calcd for

C₃₁H₃₄N₂O₉S: C, 60.97; H, 5.61; N, 4.59. Found: C,

60.25; H, 5.53; N, 4.69. HRMS (FAB, NBA) Calcd for

C₃₁H₃₅N₂O₉S (MH⁺) 611.2063, found 611.2067.

(2S)-2-Amino-4-[(2R,S)-2-carboxybutyl-(R,S)-(N-phosphor-

yl)-sulfonimidoyl]butanoic acid (21). The N-phosphoryl

sulfoximine 27 (1.24 g, 2.80 mmol) and 10% Pd on car-

bon (Aldrich, wet, Degussa type E101 NE/W, 1.36 g)

were suspended in a mixture of MeOH±H₂O±AcOH

(5:3:2, 40 mL), and H₂gas was passed through the sus-

pension for 24 h at 35 ^ꢀC. The resulting mixture was ®l-

tered through Celite, and the ®ltrate was evaporated to

dryness. The residual colorless syrup was diluted with

H₂O (40 mL), washed with Et₂O (20 mLÂ2), and con-

centrated to dryness. The residue was applied to a

Dowex 50WÂ8 column (H⁺form, 60 cm³), and the

column was washed with H₂O (200 mL). The eluent was

changed to 5% pyridine, and the column was eluted

until the solvent front reached 6 cm above the bottom of

the column (ca. 100 mL of 5% pyridine). The eluent was

changed again to H₂O, and the acidic fractions (ca. pH

<3.0) were combined, concentrated and lyophilized to

aord the N-phosphoryl sulfoximine 21 (369 mg, 67%,

Benzyl (2S)-4-[(2R,S)-2-(benzyloxycarbonyl)butyl-(R,S)-

sulfonimidoyl]-2-(N-4-nitrobenzyloxycarbonylamino)butano-

ate (26). Compound 26 was synthesized by the same

procedure as for the sulfoximine 17. The product was

puri®ed by ¯ash column chromatography (2:3, acetone/

hexane) to give 26 (3.22 g, 68%, diastereomer mixture)

2.45^ꢀ(c 0.94, CHCl₃); IR

26

as a pale yellow syrup: [a]_d

(NaCl, neat) n_max3300, 3040, 2940, 1720, 1520, 1340,

1

1200, 1050, 1010 cm ¹; H NMR (400 MHz, CDCl₃) d

8.2 (m, 2H, arom), 7.5 (m, 2H, arom), 7.3 (m, 10H,

PhÂ2), 6.2±6.0 (m, 1H, NH), 5.2±5.1 (m, 6H,

ArCH₂Â3), 4.5 (m, 1H, a-proton), 3.7 (m, 1H,

SCH₂CH), 3.2±3.0 (m, 4H, CH₂SCHHCH), 2.54 (br s,

1H, S NH), 2.4 and 2.2 (2Âm, 2H, CH₂CH₂S), 1.71

(m, 2H, CH₂CH₃), 0.90 (m, 3H, CH₂CH₃). Anal. Calcd

for C₃₁H₃₅N₃O₉S: C, 59.51; H, 5.64; N, 6.72. Found: C,

59.06; H, 5.64; N, 6.71. HRMS (EI) Calcd for

C₃₁H₃₅N₃O₉S (M+) 625.2094, found 625.2080.

26

diastereomer mixture) as a colorless powder: [a]_d

+3.48^ꢀ(c 1.04, H₂O); IR (KBr) n_max3600±2300 (br),

1

1720, 1520, 1340, 1240, 1050, 1020 cm

;

¹H NMR

(400 MHz, D₂O) d 4.07 (m, 1H, a-proton), 3.9 (m, 1H,

1950

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

SCH₂CH), 3.7±3.4 (m, 3H, CH₂SCHH), 3.1 (m, 1H,

SCH₂CH), 2.5 (m, 2H, CH₂CH₂S), 1.74 (q, 2H,

J=7.3 Hz, CH₂CH₃), 0.95 (t, 3H, J=7.4 Hz, CH₂CH₃);

³¹P NMR (200 MHz, D₂O+K₂CO₃, pH 8.9) d 1.31

(CDCl₃) d 7.4 (m, 5H, Ph), 5.74 (d, 1H, J=7.2 Hz,

OCHPh), 4.82 (dq, 1H, J=6.6 and 7.2 Hz, CH₃CHN),

4.54 (dd, J=9.4 and 9.4 Hz) and 4.34 (dd, J=4.2 and

9.6 Hz) [2H, CH₂SO₂], 4.2 (m, 1H, CHCH₂), 3.05 (s,

3H, SO₂CH₃), 1.8 (m, 2H, CH₂CH₃), 1.01 (t, 3H,

J=7.5 Hz, CH₂CH₃), 0.90 (d, 3H, J=6.6 Hz, CHCH₃).

Anal. calcd for C₁₆H₂₁NO₆S: C, 54.07; H, 5.96; N, 3.94.

Found: C, 54.24; H, 6.11; N, 3.79.

(36%),

1.46 (19%),

1.73 (21%),

1.97 (24%).

HRMS (FAB, glycerol) calcd for C₉H₂₀N₂O₈PS (MH⁺)

347.0678, found 347.0687.

(4R,5S)-3-[(2R)-2-Ethyl-3-hydroxypropanoyl]-4-methyl-5-

phenyl-2-oxazolidinone (29). The oxazolidinone 28

(4.0 g, 16.2 mmol) prepared from (1S,2R)-norephe-

drine^26bwas dissolved in dry CH₂Cl₂(40 mL), and the

solution was cooled in an ice salt bath at 15 ^ꢀC.

Dibutylboron tri¯ate (1.0 m solution in CH₂Cl₂,

19.4 mL, 19.4 mmol) was added via syringe to the solu-

tion under an Ar protection, followed by the addition of

dry NEt₃(2.12 g, 21.0 mmol). The resulting solution was

stirred for 20 min at 15 ^ꢀC and then cooled to 78 ^ꢀC.

Formaldehyde gas, which was generated by heating

paraformaldehyde (2.43 g, 80.9 mmol), was passed

through the solution with a stream of Ar gas for 10 min.

After the resulting mixture was stirred for 20 min at

78 ^ꢀC, the ¯ask was warmed to 0 ^ꢀC with continuous

stirring. After 1 h, the reaction was quenched by adding

a phosphate buer (pH 7.0, 1.0 m, 18.0 mL) and MeOH

(54 mL). A mixture of MeOH±30% H₂O₂(2:1, 60 mL)

was added to the mixture at such a rate as to keep the

internal temperature below 10 ^ꢀC. After the reaction

mixture was stirred for additional 1 h, the solvent was

removed by evaporation. The residue was extracted with

Et₂O (50 mLÂ3), and the combined Et₂O layers were

washed with satd NaHCO₃(100 mL) and satd NaCl

(100 mL), successively, dried over Na₂SO₄, and was

concentrated to dryness. The crude adduct had a dia-

stereomeric purity of 75.3% as determined by HPLC

(ChemcoPak¹, silica gel column 4.6 mmÂ250 mm, 3:4,

EtOAc/hexane). The desired diastereomer was puri®ed

by ¯ash column chromatography (1:2, EtOAc/hexane)

to aord the alcohol 29 (2.25 g, 50%) as a colorless oil:

(4R,5S)-3-[(2S)-2-(Iodomethyl)butanoyl]-4-methyl-5-phenyl-

2-oxazolidinone (31). The mesylate 30 (1.52 g, 4.3 mmol)

was converted to the iodide by the general method (NaI,

re¯ux in acetone) and puri®ed by ¯ash column chroma-

tography (1:9, EtOAc/hexane) to give the iodide 31

57.1^ꢀ(c 1.00, MeOH);

21

(84%) as a colorless oil: [a]_d

1

IR (NaCl, neat) n_max2960, 1780, 1690, 1240, 1120 cm

;

¹H NMR (CDCl₃) d 7.3 (m, 5H, Ph), 5.72 (d, 1H,

J=7.4 Hz, OCHPh), 4.86 (dq, 1H, J=6.8 and 6.8 Hz,

CH₃CHN), 4.2 (m, 1H, CHCH₂), 3.43 (dd, J=9.6 and

9.6 Hz) and 3.28 (dd, J=4.6 and 9.6 Hz) [2H, CH₂I], 1.7

(m, 2H, CH₂CH₃), 0.98 (t, 3H, J=7.5 Hz, CH₂CH₃),

0.91 (d, 3H, J=6.6 Hz, CHCH₃). Anal. calcd for

C₁₅H₁₈INO₃: C, 46.53; H, 4.69; N, 3.62. Found: C,

46.45; H, 4.68; N, 3.50.

Methyl (2S)-4-[(2S)-2-[[(4R,5S)-4-methyl-5-phenyl-2-

oxazolidinone-3-yl]carbonyl]butylthio]-2-(N-4-nitrobenzyl-

oxycarbonylamino)butanoate (32). The sul®de 32 was

prepared from thiol 14 and the iodide 31 by the same

procedure as for the sul®de 15. The crude product was

puri®ed by ¯ash column chromatography (1:4, EtOAc/

52.8^ꢀ(c 1.04, MeOH); IR

26

hexane) (100%): [a]_d

(NaCl, neat) n_max3340, 2940, 1780, 1720, 1695, 1605,

1520, 1340, 1220 cm ¹; ¹H NMR (CDCl₃) d 8.18 (d, 2H,

J=8.8 Hz, arom), 7.50 (d, 2H, J=8.8 Hz, arom), 7.4 (m,

3H, Ph), 7.3 (m, 2H, Ph), 5.72 (d, 1H, J=7.2 Hz,

OCHPh), 5.64 (d, 1H, J=8.0 Hz, NH), 5.21 (s, 2H,

ArCH₂), 4.86 (dq, 1H, J=6.8 and 7.0 Hz, CH₃CHN),

4.5 (m, 1H, a-proton), 4.1 (m, 1H, SCH₂CH), 3.78 (s,

3H, COOCH₃), 2.86 (dd, 1H, J=9.6 and 13.4 Hz,

SCH₂CH), 2.7±2.5 (m, 3H, CH₂SCHH), 2.1 (m, 2H,

CH₂CH₂S),1.7 (m, 2H, CH₂CH₃), 1.0 (m, 6H, CHCH₃

and CH₂CH₃). Anal. calcd for C₂₈H₃₃N₃O₉S: C, 57.23;

H, 5.66; N, 7.15. Found: C, 56.98; H, 5.61; N, 7.13.

[a]_d+27.7^ꢀ(c 1.00, CHCl₃); IR (NaCl, neat) n_max

22

3430 (br), 2960, 1780, 1690, 1450, 1340, 1230,

1120 cm ¹; ¹H NMR (CDCl₃) d 7.4 (m, 5H, Ph), 5.70 (d,

1H, J=7.4 Hz, OCHPh), 4.85 (dq, 1H, J=6.6 and

7.4 Hz, CH₃CHN), 4.0±3.8 (m, 3H, COCH and

CH₂OH), 2.66 (br s, 1H, OH), 1.7 (m, 2H, CH₂CH₃),

0.98 (t, 3H, J=7.4 Hz, CH₂CH₃), 0.90 (d, 3H,

J=6.6 Hz, CHCH₃). Anal. calcd for C₁₅H₁₉NO₄: C,

64.97; H, 6.91; N, 5.05. Found: C, 65.00; H, 6.96; N,

4.93.

Methyl (2S)-4-[(2S)-2-[[(4R,5S)-4-methyl-5-phenyl-2-

oxazolidinone-3-yl]carbonyl]butyl-(R,S)-sul®nyl]-2-(N-4-

nitrobenzyloxycarbonylamino)butanoate (33). The sul®de

32 was oxidized with NaIO₄by the same procedure as

for the synthesis of 16. The product was puri®ed by

¯ash column chromatography (8:1, EtOAc/hexane) to

give the sulfoxide 33 (84%, diastereomer mixture) as a

(4R,5S)-3-[(2R)-2-Ethyl-3-(methylsulfonyloxy)propanoyl]-

4-methyl-5-phenyl-2-oxazolidinone (30). The alcohol 29

(1.19 g, 4.29 mmol) was mesylated by the standard pro-

cedure and puri®ed by ¯ash column chromatography

(1:2, EtOAc/hexane) (100%): IR (NaCl, neat) n_max

pale yellow oil: IR (NaCl, neat) n_max3300, 2960, 1780,

1

1720, 1695, 1520, 1340, 1200, 1040 cm

;

¹H NMR

(CDCl₃) d 8.20 (d, 2H, J=8.6 Hz, arom), 7.51 (d, 2H,

J=8.6 Hz, arom), 7.4 (m, 3H, Ph), 7.3 (m, 2H, Ph), 6.07

1

2970, 1780, 1695, 1345, 1210, 1170, 960 cm ¹; H NMR

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

1951

(d, J=7.8 Hz) and 5.80 (d, J=7.6 Hz) [1H, NH], 5.72 (d,

J=7.4 Hz) and 5.70 (d, J=7.4 Hz) [1H, OCHPh], 5.2

(m, 2H, ArCH₂), 4.8 (m, 1H, CH₃CHN), 4.5 (m, 1H, a-

proton), 4.3 (m, 1H, SCH₂CH), 3.80 (s) and 3.79 (s)

[3H, COOCH₃], 3.41 (dd, J=10.6 and 13.0 Hz) and 3.35

(dd, J=10.2 and 12.8 Hz) [1H, SCH₂CH], 3.0±2.7 (m,

3H, CH₂SCHH), 2.3 (m, 2H, CH₂CH₂S), 1.8 (m, 2H,

CH₂CH₃), 1.00 (t, 3H, J=7.4 Hz, CH₂CH₃), 0.89 (d, 3H,

J=6.6 Hz, CHCH₃). Anal. calcd for C₂₈H₃₃N₃O₁₀S: C,

55.71; H, 5.51; N, 6.96. Found: C, 55.53; H, 5.62; N,

6.76.

6.12 (d, 1H, J=7.7 Hz, NH), 5.21 (2Âd, 2H, J=13.6 Hz,

ArCH₂), 4.5 (m, 1H, a-proton), 3.80 (s, 3H, COOCH₃),

3.68 (dd, J=8.8 and 13.9 Hz) and 3.48 (dd, J=8.4 and

13.9 Hz) [2H, SCH₂CH], 3.6 (m, 2H, CH₂CH₂S), 3.1 (m,

1H, SCH₂CH), 2.6 and 2.3 (2Âm, 2H, CH₂CH₂S), 2.1

and 1.6 (2Âm, 2H, CH₂CH₃), 0.97 (t, 3H, J=7.3 Hz,

CH₂CH₃). Anal. calcd for C₁₈H₂₃N₃O₈S: C, 48.97; H,

5.25; N, 9.52. Found: C, 49.01; H, 5.24; N, 9.54. 35b;

[a]_d+50.3^ꢀ(c 0.99, CHCl₃); mp 140.3±141.4 ^ꢀC; IR

25

(KBr) n_max3340, 2940, 1740, 1680, 1520, 1350, 1205,

1

1010 cm ¹; H NMR (400 MHz, CDCl₃) d 8.20 (d, 2H,

J=8.8 Hz, arom), 7.51 (d, 2H, J=8.8 Hz, arom), 6.40

(d, 1H, J=8.0 Hz, NH), 5.20 (s, 2H, ArCH₂), 4.5 (m,

1H, a-proton), 4.01 (dd, J=9.9 and 14.3 Hz) and 3.19

(dd, J=4.4 and 14.3 Hz) [2H, SCH₂CH], 3.78 (s, 3H,

COOCH₃), 3.7 (m, 2H, CH₂CH₂S), 3.0 (m, 1H,

SCH₂CH), 2.5 and 2.3 (2Âm, 2H, CH₂CH₂S), 1.9 and

1.7 (2Âm, 2H, CH₂CH₃), 0.98 (t, 3H, J=7.3 Hz,

CH₂CH₃). Anal. calcd for C₁₈H₂₃N₃O₈S: C, 48.97; H,

5.25; N, 9.52. Found: C, 48.84; H, 5.23; N, 9.49.

Methyl (2S)-4-[(2S)-2-[[(4R,5S)-4-methyl-5-phenyl-2-

oxazolidinone-3-yl]carbonyl]butyl-(R,S)-sulfonimidoyl]-2-

(N-4-nitrobenzyloxycarbonylamino)butanoate (34). The

sulfoxide 33 was treated with 1.8 equiv of MSH as for

the synthesis of 17 to give the sulfoximine 34 (100%,

diastereomer mixture) as a colorless foam: ¹H NMR

(CDCl₃) d 8.18 (d, 2H, J=8.8 Hz, arom), 7.51 (d, 2H,

J=8.6 Hz, arom), 7.4 (m, 3H, Ph), 7.3 (m, 2H, Ph), 6.07

(d, J=8.0 Hz) and 5.85 (d, J=8.0 Hz) [1H, NH], 5.73 (d,

1H, J=7.2 Hz, OCHPh), 5.2 (m, 2H, ArCH₂), 4.8 (m,

1H, CH₃CHN), 4.5 (m, 1H, a-proton), 4.3 (m, 1H,

SCH₂CH), 3.80 (s) and 3.79 (s) [3H, COOCH₃], 3.39

(dd, 1H, J=10.7 and 13.1 Hz, SCH₂CH), 3.1±2.6 (m,

3H, CH₂SCHH), 2.3 (m, 2H, CH₂CH₂S), 1.8 (m, 2H,

CH₂CH₃), 1.00 (t, 3H, J=7.4 Hz, CH₂CH₃), 0.89 (d,

3H, J=6.6 Hz, CHCH₃). The sulfoximine 34 was used

for the next reaction without puri®cation.

(2S)-2-Amino-4-[(1R,4S)-4-ethyl-4,5-dihydro-1,3-dioxo-1,2-

thiazol-1-yl]butanoic acid (36a). The cyclic sulfoximine

35a (1.24 g, 2.80 mmol) and 10% Pd on carbon (123 mg)

were suspended in a mixture of MeOH±EtOAc (1:1,

100 mL). A few drops of AcOH was added to the solu-

tion, and H₂gas was passed through the mixture at

room temperature. After 4 h, the solution was ®ltered

through Celite, and the ®ltrate was evaporated to dry-

ness. The pale yellow residue was applied to a Dowex

50WÂ8 column (H⁺form, 30 cm³), and the column was

eluted successively with H₂O, 5% pyridine and 3%

NH₄OH. The ninhydrin positive fractions eluted with

3% NH₄OH were combined and lyophilized to aord

the crude product 36a (634 mg) as a colorless powder.

This material was recrystallized from H₂O (8 mL)±

Methyl (2S)-4-[(1R,4S)-4-ethyl-4,5-dihydro-1,3-dioxo-1,2-

thiazol-1-yl]-2-(N-4-nitrobenzyloxycarbonylamino)butano-

ate (35a)

Methyl (2S)-4-[(1S,4S)-4-ethyl-4,5-dihydro-1,3-dioxo-

1,2-thiazol-1-yl]-2-(N-4-nitrobenzyloxycarbonylamino)

butanoate (35b).

A solution of K₂CO₃(616 mg,

4.46 mmol) in H₂O (4 mL) was added to a solution of

the sulfoximine 34 (6.90 g, 11.1 mmol) in THF±H₂O

(10:3, 130 mL) at 0 ^ꢀC. After 20 min, 30% H₂O₂aqueous

solution (2 mL) was added to the mixture at 0 ^ꢀC. The

resulting mixture was stirred for 3 h at 0 ^ꢀC. To the

solution was added 2 m HCl (10 mL) dropwise, and the

mixture was evaporated. The residue was dissolved in

EtOAc (200 mL), and the solution was washed with

0.3 m HCl (100 mLÂ3), H₂O (100 mL) and satd NaCl

(100 mL), successively, dried over Na₂SO₄, and was

evaporated to dryness to give the cyclic sulfoximine 35a

and 35b, as a colorless foam. The two diastereomers

were separated by ¯ash column chromatography (7:1,

EtOAc/hexane, then 2:1, EtOAc/acetone) to aord the

cyclic sulfoximine 35a (1.15 g, 23.4%) and the cyclic

sulfoximine 35b (1.35 g, 27.5%) as colorless crystals:

EtOH (3 mL) to give the pure cyclic sulfoximine 36a

25

(479 mg, 69%): [a]_d

77.71^ꢀ(c 0.96, H₂O); mp 203.1±

203.5 ^ꢀC (dec.); IR (KBr) n_max3100±2400 (br), 2960,

1670, 1580, 1390, 1220, 1010 cm ¹; ¹H NMR (400 MHz,

D₂O) d 4.1±3.8 (m, 5H, a-proton and CH₂SCH₂), 3.4

(m, 1H, SCH₂CH), 2.5 (m, 2H, CH₂CH₂S), 1.9 and 1.6

(2Âm, 2H, CH₂CH₃), 0.94 (t, 3H, J=7.4 Hz, CH₂CH₃).

Anal. calcd for C₉H₁₆N₂O₄S: C, 43.54; H, 6.50; N,

11.28. Found: C, 43.28; H, 6.51; N, 11.30. HRMS

(FAB, glycerol) calcd for C₉H₁₇N₂O₄S (MH⁺)

249.0909, found 249.0933.

(2S)-2-Amino-4-[(1S,4S)-4-ethyl-4,5-dihydro-1,3-dioxo-1,2-

thiazol-1-yl]butanoic acid (36b). Compound 36b was

synthesized according to the same procedure as for the

compound 36a. After passing through a Dowex 50WÂ8

column (H⁺form, 30 cm³) in the same way, the ninhy-

drin positive fractions eluted with 3% NH₄OH were

combined and lyophilized to aord the crude product

36b (490 mg) as a colorless powder. This material was

35a; [a]_d+51.9^ꢀ(c 1.03, CHCl₃); mp 142.3±143.3 ^ꢀC;

25

IR (KBr) n_max3360, 2960, 1735, 1705, 1680, 1540, 1350,

1205, 1010 cm ¹; ¹H NMR (400 MHz, CDCl₃) d 8.22 (d,

2H, J=8.8 Hz, arom), 7.52 (d, 2H, J=8.8 Hz, arom),

1952

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

recrystallized from 0.5 m HCl (5 mL)±EtOH (3 mL) to

give the pure cyclic sulfoximine 36b (265 mg, 35%):

[a]_d²⁵+78.02^ꢀ(c 1.01, 0.5 N HCl); IR (KBr) n_max3100±

(PK)-lactate dehydrogenase (LDH) coupled enzyme

assay.³⁰A standard assay mixture consisted of 0.75 mM

l-Glu, 150 mM l-2-aminobutyric acid, 1 mM ATP,

10 mM MgSO₄, 1 mM phosphoenolpyruvate, 0.24 mM

NADH, 10 unit PK, 25 unit LDH, 0.1 m KCl in 0.1 m

Tris±HCl buer (pH 7.5). The second-order rate con-

stants for slow-binding inhibition (k_inact/K_i) were calcu-

lated from a plot of the pseudo ®rst-order rate constants

(k_obs) of enzyme inactivation versus inhibitor con-

centration ([I]) by regression to k_obs=k_inact[I]/(K_i+[I])

or k_obs=(k_inact/K_i)[I] when saturation or no saturation

was observed, respectively. In the case of the phosphinic

acid 2, saturation was observed and the values of K_iand

k_inactwere determined separately. On the other hand, no

saturation was observed with the sulfoximine 3 and l-

BSO up to 10 mM and 1 mM, respectively, beyond which

rapid enzyme inactivation did not enable the initial

reaction rate to be measured by the conventional

method. The values of k_obswere calculated from the

progress curves³¹in which the reaction was initiated by

adding the enzyme to the standard assay solution con-

taining varying concentrations of the inhibitor. The

overall inhibition constants (K_i^*) for slow binding were

obtained from a plot of the steady state inhibited velo-

cities (v_s) versus [I] by nonlinear least-squares analysis.

The values of v_swere obtained by incubating the enzyme

and various concentrations of the inhibitor in the stan-

dard assay mixture without l-Glu at 37 ^ꢀC for 30 min to

establish the binding equilibrium, followed by adding l-

Glu (®nal concn 0.75 mM) to start the assay for enzyme

activity.

2300 (br), 2960, 1680, 1580, 1520, 1400, 1220,

1

1010 cm

;

¹H NMR (400 MHz, D₂O) d 4.42 (dd,

J=10.1 and 15.2 Hz) and 3.64 (dd, J=4.6 and 15.4 Hz)

[2H, SCH₂CH], 4.0±3.8 (m, 3H, a-proton and

CH₂CH₂S), 3.3 (m, 1H, SCH₂CH), 2.4 (m, 2H,

CH₂CH₂S), 1.8 (m, 2H, CH₂CH₃), 0.94 (t, 3H,

J=7.4 Hz, CH₂CH₃). Anal. calcd for C₉H₁₆N₂O₄S: C,

43.54; H, 6.50; N, 11.28. Found: C, 43.30; H, 6.50; N,

11.29.

(2S)-2-Amino-4-[(2S)-2-carboxybutyl-(R)-sulfonimidoyl]

butanoic acid (3a). To a cooled solution of the cyclic

sulfoximine 36a (152.1 mg, 0.613 mmol) in H₂O (2 mL)

was added 1.16 m KOH aqueous solution (1.06 mL,

1.23 mmol) dropwise at 0 ^ꢀC. The mixture was stirred

for 16 h at 4 ^ꢀC. The reaction mixture was directly

+

applied to a Dowex 50WÂ8 column (NH₄form,

20 cm³). The column was eluted with H₂O, and the nin-

hydrin positive fractions were collected and lyophilized

to aord the sulfoximine 3a (172 mg, 99%) as a colorless

powder: IR (KBr) n_max3500±2400 (br), 1595 (br),1400,

1

1200, 1000 cm

;

¹H NMR (400 MHz, D₂O) d 3.89 (t,

0.8H, J=6.2 Hz, a-proton in major isomer), 3.85 (t,

0.2H, J=6.6 Hz, a-proton in minor isomer), 3.68 (dd,

0.2H, J=10.1 and 14.7 Hz, SCH₂CH in minor isomer),

3.63 (dd, 0.8H, J=9.5 and 14.7 Hz, SCH₂CH in major

isomer), 3.5±3.2 (m, 3H, CH₂SCHH), 2.7 (m, 1H,

SCH₂CH), 2.4 (m, 2H, CH₂CH₂S), 1.6 (m, 2H, CH₂CH₃),

0.92 (t, 3H, J=7.6 Hz, CH₂CH₃). HRMS (FAB, glycerol)

calcd for C₉H₁₉N₂O₅S (MH⁺) 267.1016, found

267.1013.

Acknowledgements

We thank Yumiko Ohkame (Kinki University) for her

assistance in the synthesis of l-homocystine. This study

was supported in part by a Grant-in-Aid for Scienti®c

Research on Priority Areas from the Ministry of Edu-

cation, Science, Sports and Culture of Japan to which

we are grateful.

(2S)-2-Amino-4-[(2S)-2-carboxybutyl-(S)-sulfonimidoyl]

butanoic acid (3b). Compound 3b was synthesized

according to the same procedure as for the compound

3a. The crude product was puri®ed by a Dowex 50WÂ8

+

column (NH₄form, 20 cm³) eluted with H₂O, and the

ninhydrin positive fractions were combined and lyophi-

lized to aord the sulfoximine 3b (43 mg, 97%) as a

colorless powder: IR (KBr) n_max3500±2400 (br), 1590

References and Notes

1

(br), 1400, 1200, 1000 cm ¹; H NMR (400 MHz, D₂O)

1. (a) Meister, A. In The Enzymes; Boyer, P. D., Ed.; Aca-

demic: New York, 1974; Vol. 10, pp 671±697. (b) Meister, A. In

Glutathione; Chemical, Biochemical, and Medical Aspect; Dol-

phin, D.; Poulson, R.; Avramovic, O., Eds; John Wiley: New

York, 1989; Vol. III, Part A, pp 367±474.

d 3.89 (t, 0.8H, J=6.2 Hz, a-proton in major isomer),

3.85 (t, 0.2H, J=6.4 Hz, a-proton in minor isomer), 3.68

(dd, 0.8H, J=10.2 and 14.6 Hz, SCH₂CH in major iso-

mer), 3.62 (dd, 0.2H, J=9.8 and 14.6 Hz, SCH₂CH in

minor isomer), 3.4±3.2 (m, 3H, CH₂SCHH), 2.7 (m, 1H,

SCH₂CH), 2.4 (m, 2H, CH₂CH₂S), 1.6 (m, 2H,

CH₂CH₃), 0.92 (t, 3H, J=7.5 Hz, CH₂CH₃). HRMS

(FAB, glycerol) calcd for C₉H₁₉N₂O₅S (MH⁺)

267.1016, found 267.1023.

2. (a) Meister, A. Pharmac. Ther. 1991, 51, 155±194. (b) Ishi-

kawa, T., Sies, H. In Glutathione; Chemical, Biochemical, and

Medical Aspect; Dolphin, D.; Poulson, R.; Avramovic, O., Eds;

John Wiley: New York, 1989; Vol. III, Part B, pp 85±109. (c)

Smith, C. V., Mitchell, J. R. In Glutathione; Chemical, Bio-

chemical, and Medical Aspect; Dolphin, D.; Poulson, R.; Avra-

movic, O., Eds; John Wiley: New York, 1989; Vol. III, Part B,

pp 1±43.

Enzyme assay. Enzyme activity was determined by

measuring ADP formation with a pyruvate kinase

N. Tokutake et al./Bioorg. Med. Chem. 6 (1998) 1935±1953

1953

3. (a) Meister, A. In The Enzymes; Boyer P. D., Ed.; Academic

Press: New York; 1974; Vol. 10, pp 699±754. (b) Meister, A.

Methods Enzymol. 1985, 113, 185±199.

14. Stetter, H.; Kuhlmann, H. Synthesis 1979, 29±30.

15. Zeiss, H. J. German patent application, DE 3817956,

Hoechst. Date of publication: Dec. 7, 1989. CA 1990, 113, 6806r.

4. Mullins, L. S.; Zawadzke, L. E.; Walsh, C. T.; Raushel, F.

M. J. Biol. Chem. 1990, 265, 8993±8998.

16. Hanessian, S.; Sahoo, S. P. Tetrahedron Lett. 1984, 25,

1425±1428.

5. (a) Bollinger, Jr. J. M.; Kwon, D. S.; Huisman, G. W.;

Kolter, R.; Walsh, C. T. J. Biol. Chem. 1995, 270, 14031±14041.

(b) Kwon, D. S.; Lin, C.-H.; Chen, S.; Coward, J. K.; Walsh,

C. T.; Bollinger, Jr. J. M. J. Biol. Chem. 1997, 272, 2429±2445.

(c) Smith, K.; Nadeau, K.; Bradley, M.; Walsh, C. T.; Fair-

lamb, A. H. Prot. Sci. 1992, 1, 874±883.

17. Herrmann, J. L.; Schlessinger, R. H. Tetrahedron Lett.

1973, 2429±2432.

18. Ogawa, T.; Nakazato, A.; Sato, M.; Hatayama, K. Syn-

thesis 1990, 459±460.

19. Johnson C. R.; Kirchho R. A.; Corkins H. G. J. Org.

Chem. 1974, 39, 2458±2459.

6. Henderson, G. B.; Yamaguchi, M.; Novoa, L.; Fairlamb,

A. H.; Cerami, A. Biochemistry 1990, 29, 3924±3929.

20. Direct alkaline hydrolysis of a related sulfoximine deriva-

tive without ethyl side chain resulted indeed in the C±S bond

cleavage by b-elimination to yield methyl acrylate and a num-

ber of byproducts.

7. (a) Mengin-Lecreulx, D.; Flouret, B.; van Heijenoort, J.

J. Bacteriol. 1982, 151, 1109±1117. (b) Falk, P. J.; Ervin, K. M.;

Volk, K. S.; Ho, H.-T. Biochemistry 1996, 35, 1417±1422.

21. Grith, O. W.; J. Biol. Chem. 1982, 257, 13704±13712.

8. Hiratake, J.; Kato, H.; Oda, J. J. Am. Chem. Soc. 1994, 116,

12059±12060.

22. Rowe, W. B.; Ronzio, R. A.; Meister, A. Biochemistry

1969, 8, 2674±2680.

9. (a) Yokomizo, A.; Kohno, K.; Wada, M.; Ono, M.; Mor-

row, C. S.; Cowan, K. H.; Kuwano, M. J. Biol. Chem. 1995,

270, 19451±19457. (b) Bergelson, S.; Pinkus, R.; Daniel, V.

Cancer Research 1994, 54, 36±40. (c) Koizumi, T.; Negishi, M.;

Ichikawa, A. Biochem. Pharmacol. 1993, 45, 2457±2464. (d)

Lautier, D.; Luscher, P.; Tyrrell, R. M. Carcinogenesis 1992,

13, 227±232. (e) Andersen, J. K.; Mo, J. Q.; Hom, D. G.; Lee,

F. Y.; Harnish, P.; Hamill, R. W.; McNeill, T. H. J. Neuro-

chem. 1996, 67, 2164±2171. (f) Hussein, A. S.; Walter, R. D.

Parasitol. Res. 1996, 82, 372±374. (g) Moncada, C.; Repetto,

Y.; Aldunate, J.; Letelier, M. E.; Morello, A. Comp. Biochem.

Physiol. 1989, 94, 87±91. (h) Arrick, B. A.; Grith, O. W.;

Cerami, A. J. Exp. Med. 1981, 153, 720±725.

23. The a-methyl ester of a related N-phosphoryl sulfoximine

was not cleaved by hydrogenolysis.

24. Watanabe, Y.; Komoda, Y.; Ebisuya, K.; Ozaki, S. Tetra-

hedron Lett. 1990, 31, 255±256.

25. Fan, C.; Moews, P. C.; Walsh, C. T.; Knox, J. R. Science

1994, 266, 439±443.

26. (a) Evans, D. A.; Bartroli, J.; Shih, T. L. J. Am. Chem. Soc.

1981, 103, 2127±2129. (b) Gage, J. R.; Evans D. A. Org. Synth.

1989, 68, 83±91.

27. Campbell, E. B.; Hayward, M. L.; Grith, O. W. Anal.

Biochem. 1991, 194, 268±277.

28. Yamaguchi, H.; Kato, H.; Hata, Y.; Nishioka, T.; Kimura,

A.; Oda, J.; Katsube, Y. J. Mol. Biol. 1993, 229, 1083±1100.

10. (a) Grith, O. W.; Meister, A. J. Biol. Chem. 1979, 254,

7558±7560. (b) Grith, O. W. Meth. Enzymol. 1981, 77, 59±

64.

29. Fan, C.; Moews, P. C.; Shi, Y.; Walsh, C. T.; Knox, J. R;

Proc. Natl. Acad. Sci. USA 1995, 92, 1172±1176.

11. For a preliminary report, see: Katoh, M.; Hiratake, J.;

Kato, H.; Oda, J. Bioorg. Med. Chem. Lett. 1996, 6, 1437±

1442.

30. Brolin, S. E. In Methods of Enzymatic Analysis; Bergmeyer,

H. U.; Bergmeyer, J.; Graûl, M., Eds.; Verlag Chemie: Wein-

heim, 1983; Vol. 3, 3rd edn, pp 540±559.

12. Gallagher, M. J.; Honegger, H. Aust. J. Chem. 1980, 33,

287±294.

31. Morrison, J. F.; Walsh, C. T. In Advances in Enzymology;

Meister, A., Ed.; John Wiley: New York, 1988; Vol. 61, pp

201±301.

13. Dingwall, J. G.; Ehrenfreund, J.; Hall, R. G. Tetrahedron

1989, 45: 3787±3808.

Article Doi

Design, synthesis and evaluation of transition-state analogue inhibitors of Escherichia coli γ-glutamylcysteine synthetase

DOI: 10.1016/S0968-0896(98)00142-4

Source and publish data:

Authors:

Article abstract of DOI:10.1016/S0968-0896(98)00142-4

Full text of DOI:10.1016/S0968-0896(98)00142-4

Products guided by the article

R&D Labs maybe for 219506-89-5

Relevant to this article

Hot Product